sign in sign up
<<
Home , Collis gastroplasty, Hill repair

The Society for of the Alimentary Tract

52ND ANNUAL MEETING

Program Book Abstract Supplement

May 6 – 10, 2011 McCormick Place Chicago, Illinois THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT

Table of Contents

Schedule-at-a-Glance ...... 2

Sunday Plenary Video, and Quick Shot Session Abstracts ...... 4 PROGRAM BOOK ABSTRACT SUPPLEMENT Monday Plenary, Video, and Quick Shot Session Abstracts ...... 17

Tuesday Plenary, Video, and Quick Shot Session Abstracts ...... 38 FIFTY-SECOND ANNUAL MEETING

Sunday Poster Session Abstracts ...... 55 McCormick Place Chicago, Illinois Monday Poster Session Abstracts ...... 95 May 6–10, 2011

Tuesday Poster Session Abstracts ...... 138

PLEASE BRING THIS PROGRAM BOOK ABSTRACT SUPPLEMENT WITH YOU TO THE ANNUAL MEETING. THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Schedule-at-a-Glance Schedule-at-a-Glance

SATURDAY, MAY 7, 2011 SUNDAY, MAY 8, 2011 MONDAY, MAY 9, 2011 TUESDAY, MAY 10, 2011 S504 Other S501a S501bcd S503 S504 Hall A Other S501a S501bcd S504 S505 Hall A Other S501a S501bcd S504 Hall A Other

7:30 AM 7:30 AM 7:45 AM 7:45 AM OPENING SESSION 8:00 AM 8:00 AM 8:15 AM 8:15 AM PRESIDENTIAL 8:30 AM 8:30 AM PLENARY A 8:45 AM 8:45 AM THE MOVIES BREAKFAST BREAKFAST AT (PLENARY SESSION I) SYMPOSIUM (PLENARY (PLENARY Liver VIDEO SESSION II: WARD WARD 9:00 AM 9:00 AM SESSION VI) SSAT/ISDS JOINT SCI PLENARY CLINICAL Resection TRANSLATIONAL

9:15 AM PRESIDENTIAL 9:15 AM Lap ROUNDS: PANEL NAFLD Approach to DDW CCS: DDW CCS: ADDRESS: Happy

9:30 AM and Obesity, 9:30 AM Multidisciplinary PUBLIC POLICY , Surgery, Bariatric 9:45 AM Mother's Day 9:45 AM Barrett's 10:00 AM 10:00 AM 10:15 AM 10:15 AM III: JOINT JOINT 10:30 AM PRESIDENTIAL 10:30 AM SYMPOSIUM SSAT/AHPBA VIDEO QUICK

10:45 AM PLENARY B 10:45 AM SHOTS Emerging Emerging PLENARY SESSION I SESSION SESSION IV SESSION 11:00 AM (PLENARY SESSION II) 11:00 AM Technologies - QUICK 11:15 AM 11:15 AM SHOTS SESSION III SESSION DDW CCS: 11:30 AM GUEST ORATION 11:30 AM DDW CCS: Resection of GE GE of Resection PLENARY SESSION VII SESSION PLENARY Junction Dysplasia Junction & Surgical Mgmt Endoscopic Mucosal Mucosal Endoscopic Familial Colorectal Cancer: Screening THE-ART LECTURE

11:45 AM 11:45 AM STATE-OF 12:00 PM 12:00 PM 12:15 PM 12:15 PM 12:30 PM 12:30 PM 12:45 PM 12:45 PM 1:00 PM 1:00 PM 1:15 PM 1:15 PM MEET-THE- MEET-THE- MEET-THE- PROFESSOR LUNCHEONS PROFESSOR LUNCHEONS 1:30 PM PROFESSOR LUNCHEONS 1:30 PM 1:45 PM 1:45 PM 2:00 PM 2:00 PM

2:15 PM 2:15 PM JOINT SSAT/SAGES MAINTENANCE COURSE: CERTIFICATION OF 2:30 PM 2:30 PM SYMPOSIUM LUNCHEON 2:45 PM 2:45 PM How to Surgery Academic 3:00 PM 3:00 PM Succeed in Robotics GI SURGERY GI SURGERY SURGERY GI DDW CCS: DDW CCS: Approach to GI GI to Approach Regionalization Multidisciplinary 3:15 PM 3:15 PM (NETs) Tumors Mgmt of PPI Neuroendocrine Neuroendocrine The Surgeon in of Gastric the Management and Esophageal Diseases Fundoplication & & Fundoplication Definition and Definition GERD; GI Surgery Surgery GI GERD; PLENARY PLENARY POSTER SESSION I (authors available @ posters 12:00 PM - 2:00 PM) 2:00 - PM 12:00 @ posters available (authors I SESSION POSTER Complex ; POSTER SESSION II (authors available @ posters 12:00 PM - 2:00 PM) 2:00 - PM 12:00 @ posters available (authors II SESSION POSTER DEBATES A: Nissen POSTER SESSION III (authors available @ posters 12:00 PM - 2:00 PM) 2:00 - PM 12:00 @ posters available SESSION III (authors POSTER Failure in GERD Failure SESSION V SESSION CONTROVERSIES IN CONTROVERSIES IN

3:30 PM 3:30 PM Rare and B: DEBATES 3:45 PM 3:45 PM Specialty Videos Clinical Practice Clinical CONFERENCE: CONFERENCE: 4:00 PM I: SESSION VIDEO 4:00 PM JOINT JOINT STATE-OF-THE-ART STATE-OF-THE-ART 4:15 PM 4:15 PM Show Potential Applications in SYMPOSIUM Gastrointestinal Cancer: II SSAT/ASCRS y 4:30 PM Personalized Medicine in 4:30 PM WARD WARD PLENARY SESSION III SESSION PLENARY QUICK SHOTS CLINICAL Carcinoma CLINICAL CLINICAL ROUNDS: SESSION Appendicitis Complicated Complicated WARD RNDS: RNDS: WARD 4:45 PM 4:45 PM Hepatocellular Anatom

5:00 PM 5:00 PM the Duct in the Surgically Altered Me the Way: Finding Finding Way: the Me 5:15 PM 5:15 PM DDW CCS: 5:30 PM 5:30 PM ANNUAL MEETING 5:45 PM 5:45 PM BUSINESS

2 3 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

SSAT PLENARY, VIDEO, AND 306 RESULTS: Rates of pancreatic fi stula, delayed gastric emp- tying and mortality were not different between patients

Laparoscopic for Cancer: Abstracts who underwent Lap or Open PD. Estimated blood loss was QUICK SHOT ABSTRACTS Margin Status, Adequacy of Resection and 90 Day Sunday less in the Lap PD group, and operative time was longer Outcomes Printed as submitted by the authors. in the Lap PD group. Percentage of patients undergoing indicates a paper that is also being presented at the Residents & Fellows Research Conference. Craig P. Fischer1,3, Bridget N. Fahy1,2, Brian J. Dunkin1,2, a margin negative resection was equivalent amongst the / Patrick R. Reardon 1,4 1,2 groups, and mean tumor size and percentage of regional Participation in and attendance at this conference is by invitation only. , Barbara L. Bass 1Surgery, The Methodist Hospital, Houston, TX;2Surgery, Weill lymph node metastasis were also not different. Signifi cant šindicates a video presentation scheduled during a Plenary Session. Medical College of Cornell University, New York, NY; 3Division differences were noted regarding the likelihood of begin- of Surgical Oncology, The Methodist Hospital, Houston, TX; ning adjuvant therapy at 6 weeks. 87% of patients in the Lap PD group achieved this benchmark, and only 53% in 4Division of Minimally Invasive and Forgut Surgery, The Methodist Sunday, May 8, 2011 the open PD group. Lastly, patients undergoing open PD Hospital, Houston, TX were more likely to be debilitated after surgery than those INTRODUCTION: Laparoscopic pancreaticoduodenectomy 8:15 AM – 9:15 AM who underwent the minimally invasive approach. (Lap PD) has been shown to have similar outcomes as the con- CONCLUSIONS: S504 This study is the fi rst to compare a con- ventional open approach (Open PD), in selected patients. temporary cohort of patients undergoing Lap PD and Open PRESIDENTIAL PLENARY A The cancer-specifi c outcomes in patients with periampul- PD for cancer. Lap PD in this study was associated with less lary malignancy have not been reported in suffi cient num- (PLENARY SESSION I) blood loss, longer operative time, equal rates of complica- bers to allow conclusions. We report an unselected series tions, and improved likelihood receiving adjuvant ther- of 53 patients who underwent Lap PD for periampullary apy in a timely manner following surgery. The benefi ts of malignancy and compare these to matched patients who minimal access surgery regarding faster recovery may allow underwent Open PD at our institution. patients with periampullary malignancy to receive timely /305 and those receiving elective resections were twice as likely METHODS: (Table1). When evaluating racial and socioeconomic fac- From February 2009 to October 2010, 109 pan- adjuvant chemotherapy. Disparities in the Use of Minimally-Invasive Surgery tors, patients within the highest income quartile were more creaticoduodenectomies were performed for periampullary for Colorectal Disease likely to undergo minimally invasive surgery than those in malignancy. With approval from the institutional IRB, 53 307 1,2 3 2,3 patients underwent a total laparoscopic approach, without Celia N. Robinson, Shubhada Sansgiry, Courtney J. Balentine, the lowest income groups. In addition, Medicaid and Medi- Gastro-Intestinal Metabolic Surgery for the Treatment David H . Berger 2,3 care patients were signifi cantly less likely to undergo MIS. the use of hand-port or incision other than for specimen extraction. Exclusion criteria Lap PD included tumors >4 of Diabetic Patients: A Multi-Instituional International 1Michael E. DeBakey Department of Surgery, Baylor College of Lastly, race was not a signifi cant predictive factor for under- cm or vascular invasion determined by preoperative three Study Medicine, Houston, TX;2 Operative Care Line, Michael E. DeBakey going MIS for colorectal disease at a high volume center. phase CT scan. During the study period, 6 patients did not Wei J. Lee1, Kyung Yul Hur2, Muffazal Lakdawala3, Kazunori Kasama4, Veterans Affairs Medical Center, Houston, TX;3 Houston VA meet criteria established for Lap PD and are excluded fromSimon K. Wong 5 Health Services Research and Development Center of Excellence, Table 1: Multivariate Logistic Regression Analyzing Predictive the analysis. Propensity score matching was used to exam- 1 2 Michael E. DeBakey Veterans Affairs Medical Center, Houston, TX Factors for Undergoing MIS vs. Open Resection Surgery, Min Sheng General Hospital, Taoyuan, Taiwan;Surgery, (Ref. = Open Surgery) ine the institutional database, and select patients for the Soonchunhyang University Hospital, seoul, Korea, Democratic BACKGROUND: control group based on Age, Sex, Ca 19-9 and key preop- Morbidity and mortality rates for major People’s Republic of; 3Surgery, Saifee Hospital, Mumbai, India; surgical procedures are improved in high volume hospitals. Variable OR Confi dence Interval P-VALUE erative comorbidities. The institution practices a uniform 4 5 approach to adjuvant chemotherapy for patients resected Surgery, Yotsuya Medical Cube, Tokyo, Japan;Surgery, Prince of Additionally, high volume centers are often leaders in the Colorectal Cancer 3.21 2.51–4.10 <0.001 Wales Hospital, Hong Kong, Hong Kong utilization of novel surgical technology such as minimally Elective Resection 2.26 1.82–2.80 <0.001 with intent to cure, with a goal benchmark of beginning of BACKGROUND: invasive surgery (MIS) for colorectal disease. Although high RACE therapy 6 weeks following surgery. Perioperative outcomes Gastro-intestinal metabolic surgery has been volume centers often serve diverse patient populations, it is Caucasian Ref. were tracked for 90 days following surgery. proposed for the treatment of not well controlled type 2 unknown if there are disparities in the application of new African American 0.97 0.76–1.24 0.79 Diabetes Mellitus (T2DM) patients with a BMI < 35 Kg/m2. surgical technologies within these hospitals. We sought Hispanic 1.24 0.88–1.75 0.22 Perioperative and Cancer Specifi c Outcomes for LaparoscoicThis and study aims to describe recent experience with surgical to determine if ethnic and socioeconomic disparities in PRIMARY PAYER Open Pancreaticoduodenectomy treatment of T2DM in Asian centers. the use of MIS for colorectal disease exist at high volume Private Ref. METHODS: Medicare 0.86 0.74–1.0 0.05 Laparoscopic Patients aged 20 to 70 years, with not well centers. 2 Medicaid 0.48 0.33–0.68 <0.001 n = 53 Open n = 53 p Valuecontrolled T2DM (HbA1C > 7.0%) and BMI < 35 Kg/m METHODS: Estimated Blood Loss, ml. 369( 50–1400) 964 (250–1600) <0.05 Using the 2008 National Inpatient Sample MEDIAN HOUSEHOLD INCOME were included at 5 institutes between 2007 and 2010. The Operative Time, minutes 387 (245–689) 248 (157–512) <0.05 database, a retrospective review of laparoscopic colecto- I: \$1–\$35,999 0.72 0.58–0.91 0.005 end point is T2DM remission, defi ned by fasting plasma Grade B/C Pancreatic Fistula* 11/53 (20) 13/53 (25) NS mies performed at high volume centers (case volume >200 II:\$36,000–\$44,999 0.73 0.60–0.90 0.003 glucose <110 mg/dl and HbA1C <6.0%. Delayed Gastric Emptying 2/53 (4) 1/53 (2) NS year) was conducted. ICD-9 codes were used to identify III: \$45,000–\$58,999 0.87 0.74–1.02 0.086 RESULTS: Mortality, No. (%) 2/53 (4) 1/53 (2) NS 163 patients with a mean BMI 30.4 (22–34), age minimally invasive (MIS) colorectal resections. Multivari- IV: • \$59,000 Ref. Tumor Size, mean mm. 3.2 (0.5–4.0) 3.5 (1.0–4.0) NS46.6 (29–67), duration of T2DM 8.8 years (0.5–20), C-pep- ate logistic regression including ethnic and socioeconomic Margin Negative, R o resection 45/53 (85) 41/53 (77) NStide 2.8 ng/ml (0.9–14) and HbA1C of 9.1% (7.0–15) were variables was used to identify independent predictive fac- No. (%) recruited up to now. Among them, 140 (85.6%) received tors for undergoing MIS. Regional Lymph Node 39/53 (75) 38/53 (72) NSlaparoscopic gastric bypass (LGB), 19 (11.7%) received lap- RESULTS: CONCLUSION: There are signifi cant socioeconomic dis- Metastasis, No. (%) A total of 98,047 colorectal resections were % of patients beginning 46/53 (87) 28/53 (53) <0.05aroscopic sleeve gastrectomy (LSG) and 4 (2.4%) received performed at high volume centers in 2008. Overall, only parities in the use of minimally invasive surgery for colorec- adjuvant therapy at 6 weeks laparoscopic adjustable gastric banding (LAGB). Up to now, 7950 (8.1%) colorectal resections were performed using tal disease at high volume centers. Future studies should be Karnofsky Performance Status 1/53 (2) 18/53 (34) <0.0574 patients had been follow-up for at least 12 months. The a minimally invasive approach. Patients with malignant aimed at identifying access barriers to MIS in the treatment <70 at 6 weeks. BMI decreased postoperatively by 19.4% to 24.5. The fast- of colorectal disease. *Bassi et al, 2005. Karnovsky performance status <70 - patient unable to work ingor blood glucose level decreased from 212 mg/dL to 117.5 neoplasms were three times more likely to undergo MIS cary on normal activites of daily living without assistance. mg/dL, and the HbA1C decreased to 6.3%. The overall T2DM remission rate at 12 months after surgery was 77% (55/74), and was 86.5% for LGB, 61.6% for LSG, 50% for

4 5 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

LAGB. Type of surgery, duration of T2DM and C-peptide METHODS: Between August 2008 and August 2010, 42 recorded using the same catheter confi guration. We defi ned recurrence rate as determined by upper GI barium x-rays predict the remission of diabetes after surgery. patients with acute appendicitis were offered a pure TVA. this patient set as “true LPR.” (UGI) was >50%. This study analyzes the impact of her-

Patients who did not wish to undergo a TVA underwent nia recurrence after LPEHR on clinical symptoms at 5-year Abstracts CONCLUSION: Gastro-intestinal metabolic surgery safely RESULTS: Forty (18 male, 22 female, mean age 33 years) Sunday and effectively eliminated T2DM in patients with a BMI < a LA and served as controls. Both groups were provided a subjects were enrolled. Six subjects were excluded because follow-up. 35 Kg/m2. Larger and longer study is needed for this trial FSFI before surgery and at regular intervals for up to 1 year. of a positive DeMeester score or sensor noise. Thirty-four METHODS: From 2002 to 2005 we enrolled 108 patients to confi rm the benefi t. Pre and post-operative FSFI results were compared between subjects completed one, and 26 completed both HMII test- with PEH who agreed to participate in the study and cohorts using unpaired t-tests, and between individuals ing periods off and on PPI. There was no difference in the undergo LPEHR and with or without within each cohort pre and post operatively using paired /308 total number of total refl ux events between off and on PPI the use of biologic mesh. UGI and symptom questionnaire t-tests. (Median, 22 and 24, respectively, p = 0.89). One subject (frequency and severity of 9 commonly associated foregut Female Sexual Function After Pure Transvaginal RESULTS: 18 patients underwent a pure TVA, 22 under- had 1 LPR event off PPI. On PPI, two subjects had 1 and symptoms) were performed at baseline, 6 months post- : A Cohort Study went LA, and 2 refused participation in this study. Pre-oper- 3 LPR events, respectively. The 95th percentile occurrence operatively, and approximately 5 years after the operation, Daniel Solomon1,2, Rachel Lentz2, Andrew Duffy1,2, Robert Bell1,2, ative and >60 day post-operative FSFI data was available for of LPR off and on PPI were 0 and 1, respectively; 95th per- during 2009. 1,2 Kurt E. Roberts 21 patients (11 TVA, 10 LA). Baseline FSFI scores were not centile for proximal refl ux was 4 for both off and on PPI. RESULTS: Of 108 patients, 72 could be contacted and sur- 1Department of Surgery, Yale-New Haven Hospital, New Haven, signifi cantly different between groups (TVA: 18.0 ± 1.46, LA: Patients with treatment responsive LPR had pre-treatment th veyed and 60 of them had an UGI. The median follow-up CT; 2Surgery, Yale University School of Medicine, New Haven, CT 15.0 ± 2.1, p = 0.92). FSFI scores at greater than 60 days post- HMII values greater than the 95 percentile of normal was 58 months (42–78) and the average age at follow-up was operatively did not differ signifi cantly from pre-operatively patients in all criteria except total number of refl ux events BACKGROUND: The impact of transvaginal appendec- 68 ± 10 years. Overall radiographic recurrence (≥20 mm) in either group. (TVA: 18.0 ± 1.46 to 16.7 ± 2.4; p = 0.49, (Table). was 14% at 6 months and 57% at 5 years, and the aver- tomy on female sexual function is unknown. We therefore LA:15.0 ± 2.1 to 16.4 ± 2.1; p = 0.49). No FSFI domain in performed a prospective cohort study of women undergo- age size of the recurrence was 40 ± 10 mm. All symptoms either cohort was signifi cantly changed post-operatively. Comparison between normal subjects and Treatment Responsivecontinued to be signifi cantly better than before repair at 5 ing pure transvaginal (TVA) versus tradi- CONCLUSIONS: LPR patients tional laparoscopic appendectomies (LA). Using a validated, Neither TVA nor LA affected female sex- years, and with the exception of heartburn appeared to be 19 point, female sexual function index questionnaire (FSFI) ual function scores. This suggests that TVA does not have unaffected by the presence or size of the recurrence (Table I). Normal Treatment Responsive assessing 6 domains of sexual function: desire, arousal, negative effects on female sexual function. The results Two patients had a re-operation during this follow-up for of this study may prove benefi cial in consultations with Subjects LPR Patients symptoms related to their recurrent . Ten patients lubrication, orgasm, satisfaction and pain with intercourse, HMII Measurements (off PPI): N = 34 (off PPI): N = 15 p-value pre and post-operative sexual function were compared. patients concerning the sexual sequelae of transvaginal DMS 3.1 (1.2–5.4) 9.3 (5.8–19.5) had 0.002 died, but none as a complication of their LPEHR. surgery. # Total refl ux events 22 (8.3–32) 24.5 (14.5–31) 0.26 # Proximal refl ux events 1 (0–2.75) 10.5 (6.3–16.8) <0.001 (2 cm distal to UES) #LPR events 0 (0–0) 1 (1–2) <0.001 Predominate position Upright Upright 10:30 AM – 11:15 AM of events S504 Variables were compared using Pearson’s chi-squared test and a p-value <0.05 was considered signifi cant. The values are expressed as the median and PRESIDENTIAL PLENARY B interquartile range (IQR). (PLENARY SESSION II) CONCLUSION: This study establishes the normative data CONCLUSION: Despite an overall high rate of radiologic for LPR. LPR events are extremely rare in patients with- recurrence following LPEHR, patient satisfaction remains out GERD or LPR symptoms as measured with HMII. One high, radiologic recurrences are well tolerated, and the 389 METHODS: This is an IRB approved prospective trial. or more LPR events should be considered abnormal in need for reoperation is very low. Participants were contacted through community advertis- How Much Pharyngeal Acid Is “Normal?”: Normative patients with LPR symptoms regardless of whether there is ing. Patients with past or current history of GERD, LPR or Data for Laryngopharyngeal Refl ux Events Using a positive DeMeester score. antisecretory medication use were excluded. Absence of š391 Hypopharyngeal Multichannel Intraluminal symptoms was confi rmed using the refl ux symptom index Impedance (HMII) 390 Hiatal Hernia Recurrence Following Mesh and and GERD-HRQoL questionnaires. Subjects underwent Non-Mesh Repair Toshitaka Hoppo1, Alejandro F. Sanz1, Katie S. Nason1, unsedated transnasal ; patients with esophageal Laparoscopic Paraesophageal : Defi ning Long-Term Clinical and Anatomic Outcomes Hubert Kim, Kevin M. Reavis, Ninh T. Nguyen Thomas Carroll 2, Clark Rosen3, Daniel P. Normolle4, fi ndings of refl ux were excluded. 24-hour HMII was then 1 1 2 UCI Medical Center, Orange, CA Nicholas J. Shaheen 5, James D. Luketich1, Blair A. Jobe1 performed before and after a 2-week course of proton pump Brant K. Oelschlager, Rebecca P. Petersen, L.Michael Brunt, 1Cardiothoracic Surgery, University of Pittsburgh Medical Center, inhibitors (PPI). All subjects with a positive DeMeester Score Nathaniel J. Soper 3, Brett C. Sheppard4, Lee Mitsumori5, Paraesophageal are common in Western countries were excluded from the fi nal analysis. Catheter confi gura- 5 6 1 Pittsburgh, PA;2 Otolaryngology, Tufts Medical Center, Boston, Charles Rohrmann , Lee L. Swanstrom, Carlos A. Pellegrini and are also more frequently seen in the elderly and the tion consisted of pH probes located in the hypopharynx 1 2 obese. Despite mesh reinforcement of hiatal hernias, recur- MA; 3Otolaryngology, University of Pittsburgh Medical Center, Surgery, University of Washington, Seattle, WA;Surgery, and 5 cm proximal to the esophagogastric junction; imped- rence is high (up to 40% in large hernias). This is a video 4 Washington University, St. Louis, MO; 3Surgery, Northwestern Pittsburgh, PA; Biostatistics, University of Pittsburgh, Graduate ance electrode pairs were located in the hypopharynx and presentation of the pattern of hiatal hernia recurrence fol- 5 University, Chicago, IL; 4 School of Public Health, Pittsburgh, PA; Medicine, University of proximal and distal esophagus. LPR events were considered Surgery, Oregon Health & Science 5 lowing mesh and non-mesh repair and how to approach North Carolina, Chapel Hill, NC present with retrograde bolus transit across all ring sets and University, Portland, OR; Radiology, University of Washington, this challenging problem. Seattle, WA; 6Surgery, The Oregon Clinic, Portland, OR BACKGROUND: Laryngopharyngeal refl ux (LPR) with or ultimately reaching the hypopharynx. Assuming a Pois- th without aspiration can cause atypical symptoms, asthma son distribution, 95 percentiles for proximal refl ux and OBJECTIVE: We recently determined in a multi-institu- and pulmonary fi brosis. No method has been accurate or LPR events were calculated and used as boundaries defi n- tional, prospective, randomized study of laparoscopic para- sensitive for the diagnosis of LPR. The aim of this study was ing for normal vs. pathologic LPR. Subsequently, a cohort esophageal hernia repair (LPEHR) that the 5-year anatomic to establish the normative data for LPR and proximal refl ux of 15 LPR patients who had a complete response to medi- events in an asymptomatic cohort using hypopharyngeal cal or surgical therapy was used for comparison with the multichannel intraluminal impedance-pH (HMII). normative data; all patients had pre-treatment LPR events

6 7 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

2:15 PM – 4:30 PM 543 545

S503 Laparoscopic-Assisted Frey Procedure: A New OptionLaparo-Endoscopic Single Site (LESS) Morgagni HerniaAbstracts for Treatment of Pain in Chronic Pancreatitis Repair and Toupet Fundoplication Sunday VIDEO SESSION I: SPECIALTY VIDEOS Sujit Kulkarni, Dilip Parekh, Rick Selby, Kaylene Barrera Sharona B. Ross1,2, Harold Paul1,2, Kenneth Luberice1,2, Surgery, University of Southern California, Los Angeles, CA Farhaad C. Golkar 1,2, Alexander S. Rosemurgy1,2 1 2 The Frey procedure is an established procedure for treat- Surgery, University of South Florida, Tampa, FL;Tampa General Hospital, Tampa, FL 538 are shown. It was decided to approach this patient surgi- ment of pain in chronic pancreatitis. A laparoscopic tech- cally as snare polypectomy was felt unsafe. : nique for this procedure has not been described previously. In this video we have expanded the applications of LESS Minimally Invasive for Carcinoma of the polyp is located under intraoperative endoscopic guid- In this video we describe a laparoscopic-assisted redo Frey surgery to treat an uncommon condition. The operation Esophagus After Neoadjuvant Chemoradiation ance. An anterior longitudinal gastrotomy is made. The procedure with a peustow-type pancreaticojejunostomy was undertaken through a 12mm incision at the umbilicus Daniel K. Tong, Simon Law, Fion S. Chan, Kam H. Wong polyp is exteriorized on its stalk into the peritoneal cavity performed with the Da Vinci robot in a 46 y old female using a multi-trocar port and a 5mm defl ectable tip laparo- Surgery, The University of Hong Kong, Hong Kong, Hong Kong and resected with a linear stapler. The closure is with familial pancreatitis who previously underwent a peu- scope. Loops of intrathoracic small bowel and colon were stow procedure to the body and tail of the . The reduced. The diaphragmatic defect was repaired using non- Minimally invasive esophagectomy in patients with performed. Postop evolution is described. patient remains asymptomatic 2 y after surgery. Laparo- absorbable mesh and a laparoscopic tacking device. The esophageal cancer and prior chemoradiation is challeng- scopic Frey procedure may provide a new treatment option distal esophagus was circumferentially dissected and the ing because of post-radiation fi brosis. We present such a 541 for intractable pain from chronic pancreatitis. gastric fundus was mobilized by dividing the short gastric patient who was operated via combined thoracoscopy and Laparoscopic Re-Do Myotomy for Recurrent Achalasia vessels. After reconstructing the esophageal hiatus, a 270° laparoscopy. After Heller-Dor Procedure 544 fundoplication was constructed over a bougie. The umbili- In the thoracic phase, the esophagus was mobilized together Mario Costantini, Renato Salvador, Lisa Zanatta, Cristina Longo, Laparoscopic Pancreatic Head Resection with Hybridcal defect was closed and the skin approximated with excel- with mediastinal lymphadenectomy. In the abdominal Elena Finotti, Tiziana Morbin, Giovanni Zaninotto, Ermanno Reconstruction via Pancreatogastrostomy lent cosmesis. phase, the stomach was mobilized with preservation of the Ancona vascular arcades. The cervical esophagus was divided and Tobias Keck, Simon Küsters, Ulrich F. Wellner, Frank Makowiec, Clinica Chirurgica 1, University of Padua, Padova, Italy 546 pulled down into the peritoneal cavity. Gastric transection Ulrich T. Hopt, Wojciech K. Karcz was performed extracorporeally via a 5-cm mini-laparot- Failures of laparoscopic myotomy are usually managed Department for General and Visceral Surgery, University of Robot-Assisted Laparoscopic Biliopancreatic Diversion omy. A narrowed gastric tube was created and railroaded with complimentary dilations. Sometimes, however, a new Freiburg, Freiburg, Germany with via the orthotopic route to the neck for esophagogastric operation is necessary. as in the case we present. At re-lapa- Sapan S. Desai, Ranjan Sudan Since 2010, we performed 11 laparoscopic pylorus preserv- anastomosis. roscopy, the myotomy was clearly short on the gastric side. General Surgery, Duke University Medical Center, Durham, NC A new myotomy on the right side of the gullet, prolonged ing pancreatic head resections. The operation is performed as hybrid operation with complete laparoscopic resection Our patient is a 33-year-old female with a BMI of 42 and an 539 for 2 cm on the lesser curvature, was then performed. A partial antirefl ux wrap completed the operation. With this and reconstruction via a small retrieval incision of 5–7 cm. excess body weight of 110 pounds who underwent a robot- Minimally Invasive Esophagectomy After Nissen technique we operated 11 patients (7 were referrals): all The video demonstrates the surgical technique and pos- assisted laparoscopic biliopancreatic diversion with duode- Fundoplication were cured from their residual dysphagia, even if GERD sible pitfalls. The complication rate was not increased in nal switch (RA LBPD/DS). This procedure begins with an appendectomy, followed by creation of the ileo- Hubert Kim, Ninh T. Nguyen, Kevin M. Reavis symptoms developed in 2. Laparoscopic revision of unsuc- comparison to open surgery. The average operation time anastomosis. A is completed followed by Surgery, University of California, Irvine, Orange, CA cessful myotomy is feasible, and should be considered an was not signifi cantly increased in comparison to open (528 option in recurrent patients who do not respond to pneu- vs 540 min). the sleeve . The duodeno-ileal anastomosis is Nissen fundoplication is a common operation reserved for matic dilations. made with a retrocolic Roux limb via a robot-assisted hand- gastroesophageal refl ux refractory to medical management. We demonstrate that pancreatic head resections can be sewn anastomosis. Mesenteric defects are closed to prevent Despite surgical management, patients are still at risk for safely performed via laparoscopy. The combination with an internal hernias. A methylene blue leak test ensures the developing esophageal cancer. This is a video presentation 542 open reconstruction allows a standardized reconstruction anastomosis is patent and secure. Our patient has lost 30% which demonstrates that minimally invasive Ivor-Lewis Laparoscopic Truncal and Antrectomy with no concessions to safety and operation time. of her excess weight in one month. esophagectomy can be safely peformed after prior Nissen Mariah Alexander, Aaron M. Hurd, Gregory J. Mancini fundoplication. University of Tennessee Medical Center Knoxville, Knoxville, TN 540 The advent of PPI’s and the improved treatment of H. Pylori has greatly reduced the frequency that peptic ulcer Laparoscopic Transgastric Resection of a surgery is performed. PPI’s have helped to sub-select the Large Intussuscepted Gastric Polyp most severe cases of ulcer disease presenting for surgical Etienne Auger-Dufour1, Liane S. Feldman1, Gerald M. Fried1, evaluation. Chronic peptic ulcer disease refractory to medi- Lorenzo E. Ferri 1 cal treatment necessitates surgical intervention. The pri- Surgery, McGill University Health Centre, Montreal, QC, Canada mary indication for gastric resection in the setting of PUD is chronic obstruction caused by scarring. Surgical inter- An 89-year-old female patient presented with melena, vention must achieve acid suppression, resection of con- intermittent vomiting and anemia. A large pedunculated cerning gastric pathology and reconstruction. We present hyperplastic polyp is found in the proximal part of the laparoscopic truncal vagotomy with antrectomy as a safe, stomach body. Its long stalk containing gastric wall extends effective option that leads to excellent functional results. across the pylorus into the duodenum. Preoperative studies

8 9 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

2:15 PM – 5:15 PM tive visit if their symptoms are better and for persistent or Review of surgical pathology revealed that only 54 of 141 recurrent symptoms they are frequently seen by Primary (38.3%) specimens were normal while 75 of 141 showed

S501a Abstracts Care Providers and subsequently referred to Gastroenter- mild chronic infl ammation. The only factor that was sig- Sunday PLENARY SESSION III ologists (GI). We did an extensive Pubmed search using the nifi cantly different between the two groups of patients words “biliary dyskinesia” and could not fi nd any study with and without resolution of symptoms after cholecys- which took this into consideration. We attempted to study tectomy was whether the pathology was normal versus if this pattern and assess which factors if any predict patients there was infl ammation, chi square p value = 0.02. that will benefi t from cholecystectomy. We benefi t from CONCLUSION: Although our study has the limitation an extremely homogenous patient population our hospital /547 of being a retrospective analytical study it raises it some being the only tertiary care hospital serving a rural popula- important issues. Cholecystectomy does not appear to be Prospective Study for Selective Management of Patients tion for a considerable distance. as effective for biliary dyskinesia as most studies suggest with Acute Biliary Pancreatitis: Interim Results METHOD: Retrospective analysis of medical records of when long term follow up including that with Gastroen- Dana A. Telem, Modesto Colon, Nalin Ranasinghe, patients who underwent cholecystectomy for biliary dyski- terology is accounted for. If operative pathology shows Michelle K. Kim nesia from February 2001 to January 2010 with minimum chronic infl ammation the patients are signifi cantly more Surgery, The Mount Sinai Hospital, New York, NY postoperative follow-up of 6 months. likely to have resolution of symptoms. BACKGROUND: Detection of persistent common bile RESULTS AND DISCUSSION: After an initial analysis of duct (CBD) stones in patients presenting with acute biliary 274 patients only 141 patients were included in the study /549 pancreatitis (ABP) remains challenging, resulting in high based upon our inclusion and exclusion criteria. Of the 141 Validation of the Updatedth 7Edition AJCC TNM rates of negative endoscopic retrograde cholangiopancrea- patients 117 (83%) were female and 24 (17%) male. Symp-Staging Criteria for Gastric Adenocarcinoma tography (ERCP) and magnetic resonance cholangiopan- toms for which patient underwent cholecystectomy were Lee J. McGhan1, Barbara A. Pockaj1, Richard J. Gray1, Sanjay P. Bagaria2, creatography (MRCP). In a study previously conducted by persistent or recurred in 61 of 141 (43.3%) patients postop- Nabil Wasif 1 our institution, an algorithm for selective management eratively. Although most patients on initial surgical follow- 1 2 of patients with ABP was developed. In this study, multi- up said symptoms had improved however 58/141 patients General Surgery, Mayo Clinic Arizona, Scottsdale, AZ;General variate analysis identifi ed 5 clinical variables which sig- were seen by the GI service with persistent symptoms over Surgery, Mayo Clinic Florida, Jacksonville, FL nifi cantly correlated with persistent stone. Optimal values a time range of 1–63 months (median at 5 months, mean INTRODUCTION: The recently published 7th edition of were: CBD size on ultrasound ≥9 mm; alkaline phoshatase 9 months). An upper GI endoscopy was performed in 40 of the American Joint Committee on Cancer (AJCC) TNM ≥250 U/L; gamma-glutamyl transferase ≥ 350U/L; total bili- 141 (28.4%) patients for persistent symptoms. staging criteria for gastric adenocarcinoma contains some rubin ≥3 mg/dl; and direct bilirubin ≥2 mg/dl. We hypoth- important changes including a re-classifi cation of tumor esized that grouping variables would increase reliability RESULTS: Of the 34 patients, 17 were in-protocol and Operative pathology depth (T) and lymph node (N) status; it also marks the of detection. Receiver operator curve (ROC) curve analy- introduction of new stages IIA, IIB and IIIC as well as reclas- 17 were controls. Twenty patients were low probability Normal 54 38.3% sis and linear regression were applied to obtain optimal (0–2), 10 patients intermediate (3–4) and 4 patients high sifi cation of stage IV as patients with distant metastases and equitable predictive values, and variables combined. Mild chronic infl ammation 70 49.6% probability (5). CBD stones were present in 15% of low Cholesterolosis 12 8.5%only. The goal of this study was to validate these new stag- Based on results, a 5-point scoring system was devised and probability, 30% intermediate and 100% of high probabil- Cholesterolosis and infl ammation 5 3.5%ing criteria in a national population registry and compare recommendations for biliary evaluation suggested. Table 1 ity patients (p < 0.01). A total of 13 MRCP and 6 ERCP were Total 141 100%it to the prior staging in terms of survival discrimination. demonstrates recommendations according to probability performed. Overall, 69% of MRCP performed were nega- METHODS: of CBD stone as assessed by score. A retrospective review of all gastric cancer tive for CBD stone and 33% of ERCP were negative. Com- patients from Surveillance, Epidemiology and End Results parison of patients demonstrated decreased performance (SEER) registry data from 2004 to 2007 was conducted. The Table 1: Positive Predictive Value (PPV), Associated Odds Ratio of MRCP (23% vs. 53%) and rate of negative MRCP (50% same dataset was used to stage patients according to both (OR) with 95% Confi dence Interval [CI] and Recommendationsvs. for 78%) as well as decreased performance of ERCP (12% Biliary Eevaluation Based on Score: 6th and 7th edition criteria. Three-year disease-specifi c sur- vs. 23%) and decreased rate of negative study (0% vs. 50%) vival was compared using Cox proportional regression. for protocol versus control patients. For the low probability RESULTS: Recommended protocolBiliary patients with CBD stone, all stones were detected A total of 13,547 patients with gastric adeno- Score PPV OR +95% [CI] P-value Evaluationat time of cholangiogram and extracted via ERCP. No post- carcinoma were identifi ed; 8193 patients (60%) underwent 0 0% 0 [0–0.16] <0.001 IOC* operative complications ensued. surgery, and 3486 (26%) received radiation therapy. The 1 11% 0.33 [0.09–1.1] 0.076 IOC mean patient age was 66 years. There was a marked increase CONCLUSION: 2 25% 1.05 [0.3–3.4] 1.00 IOC Preliminary results demonstrate that the in the number of patients classifi ed as stage III using the 3 39% 2.30 [0.8–6.4] 0.125 MRCP algorithm correlates with probability of CBD stone. This updated criteria (23% vs. 13%), coupled with a decrease in 4 50% 3.83 [1.3–11.1] ”0.01 MRCP algorithm may be utilized to stratify patients to appropriate the number of patients classifi ed as stage IV (47% vs. 53%). 5 92% 53.1 [6.8–581.2] ”0.01 ERCP initial testing; decreasing the amount of unnecessary stud- *IOC = intraoperative cholangiogram ies and interventions. Re-staging the same patient population according to the 7th edition criteria improved survival discrimination from the 6th edition staging [Figures 1 and 2]. On Cox regres- PURPOSE: 548 To prospectively study an algorithm developed / sion multivariate analysis, signifi cant variables predicting for evaluation for CBD stones in patients presenting with ABP. Biliary Dyskinesia: Are We Treating It Right? poor survival included high tumor grade (HR 1.17, CI: METHOD: Following institutional board approval, all Vikas Singhal, Patrick Szeto, Heather Norman, Nan Walsh, 1.10–1.26; p < 0.001) and distal location of primary tumor patients presenting with a diagnosis of ABP were offered Thomas J. VanderMeer The mean ejection fraction (EF) on radionuclide imaging vs. proximal location (HR 1.78, CI: 1.64–1.93; p < 0.001). enrollment into a prospective study. Informed consent was Surgery, Guthrie-Robert Packer Hospital, Sayre, PA in patients with persistent symptoms was 18.0% (95% Hazard ratios by stage (vs. stage IA) showed linear progres- obtained for all enrolled patients. Patients not enrolled sion and a statistically signifi cant survival difference from INTRODUCTION: CI range 13.8–22.2) compared to almost similar mean EF were considered controls. Enrollment was at the discretion The results in almost all studies on bili- 18.04% (95% CI range 15.0–21.0) in patients who had res- the prior stage: Stage IB, 2.16 (CI: 1.56-2.98); IIA, 3.25 (CI: of the patient and attending physician. To date, 45 patients ary dyskinesia are based on short term surgical follow up. olution of symptoms. In the group of patients who had 2.44–4.32); IIB, 6.14 (CI: 4.70-8.02); IIIA, 8.76 (CI: 6.75– have been captured of which data collection is complete They do not take into consideration that most patients are reproduction of symptoms with cholecystokinin, 49 of 82 11.4); IIIB, 11.8 (CI: 9.20–15.2); IIIC, 16.7 (CI: 13.0–21.5); for 34. Figure 1 demonstrates study design. discharged from surgical follow up after the fi rst postopera- (59.8%) had resolution of symptoms. and stage IV, 20.2 (CI: 15.9–25.8) (all p < 0.001).

10 11 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

CONCLUSIONS: The AJCC 7th edition staging criteria dem- trolled at 1%. Biological functions of selected microRNAs 551 552 onstrates better survival discrimination than previous crite- of interest were evaluated by overexpressing transfections

National Trends in Survival for Pancreatic Adenocarci-Pain Control and Quality of Life After Abstracts ria. A smaller proportion of patients were classifi ed as stage and subsequent in vitro invasion assays. noma Based on Location of the Primary Lesion with Islet Cell Autotransplantation for Chronic Sunday IV using the updated criteria, with 13% of stage IV patients RESULTS: Primary fresh frozen tissues from 65 patients Gautam K. Malhotra1, Lynette M. Smith2, Quan P. Ly3, Pancreatitis down-staged to IIIC. The revisions refl ect better progno- (40 male and 25 female) with a mean age 65 ± 13 years Aaron R. Sasson 3 3 1 1 2 sis and allow for more appropriate selection of therapeu- and with AJCC Stage I (n = 7), Stage II (n = 22), Stage III (n , Chandrakanth Are Katherine A. Morgan, Stefanie M. Owczarski, Jeffrey J. Borckardt, 1 2 1 tic options, of particular importance in the modern era of = 18) and Stage IV (n = 18) colon cancers, underwent RNA College of Medicine, University of Nebraska Medical Center,Alok Madan , David B. Adams multi-modality therapy. Our study therefore validates the Omaha, NE; 2Department of Biostatistics, College of Public 1Surgery, MUSC, Charleston, SC;2Psychiatry and Behavioral th extraction and microRNA array analysis. We identifi ed a recent revisions made to the T and N stages in the 7 edi- seven-microRNA expression signature to differentiate early Health, University of Nebraska Medical Center, Omaha, NE; Services, MUSC, Charleston, SC tion AJCC staging criteria for gastric adenocarcinoma. 3 cancers (Stage I) from those associated with metastatic dis- Division of Surgical Oncology, Department of Surgery, University INTRODUCTION: In selected patients with chronic pan- ease (Stage IV) (Table). We then demonstrated that this of Nebraska Medical Center, Omaha, NE creatitis, total pancreatectomy with islet autotransplanta- /550 signature was able to allow discrimination between Stage BACKGROUND: The management and prognosis of pan- tion can be effective for the treatment of intractable pain II and III primary colon cancers (Table). Interestingly 6 of Acquisition of Metastatic Potential in Colonic creatic adenocarcinoma varies depending on the location while ameliorating post-operative diabetes. Improved quality the 7 differentially expressed microRNAs were downregu- Adenocarcinomas Is Associated with Downregulation of the primary lesion within the pancreas. There is a pau- of life scores and decreased daily narcotic use, as indica- lated with tumor progression, suggesting tumor suppres- of Complementary Strand MicroRNAs city of data on survival trends in pancreatic adenocarci- tors of successful pain relief, are expected after total pan- sive functions. Moreover, all 7 microRNAs were noted to be 1,2 3 2 noma based on the location. The aim of this study was to createctomy. These outcomes and their relationship have Jonathan M. Hernandez, Dung-Tsa Chen, Susan McCarthy, complementary strand microRNAs (designated by *) which, 2 2 2 analyze the national trends in survival for patients with not been previously well examined in this patient group. Leigh Ann Humphries , Domenico Coppola, Timothy J. Yeatman, until only recently, were thought to be non-functional pancreatic adenocarcinoma based on the location of the METHODS: David Shibata 2 byproducts of microRNA metabolism. In order to evalu- A prospectively collected database of patients primary lesion. 1Surgery, University of South Florida, Tampa, FL;2Gastrointestinal ate the biologic function of the downregulated microR- undergoing extensive pancreatectomy with islet auto METHODS: Oncology, Moffi tt Cancer Center, Tampa, FL; 3Bioinformatics, NAs of our signature, we separately transfected HCT-116, The Surveillance, Epidemiology, and End transplantation for pancreatitis was reviewed. Data pertain- ing to daily oral Morphine Equivalents (MEs) and Quality Moffi tt Cancer Center, Tampa, FL a highly invasive colon cancer cell line, with correspond- Results (SEER) database (SEER 17, 1973–2007) was used to identify patients with pancreatic cancer located in the head of Life (QOL), as measured by the SF-12 or SF-36v2 ques- BACKGROUND: ing precursor microRNAs and confi rmed overexpression by Altered expression of certain microRNAs quantitative RT-PCR. We demonstrated that overexpressing (C25.0), body (C25.1), or tail (C25.2). SEER*Stat 6.6.2 was tionnaire, in the preoperative and postoperative period is known to occur during colorectal carcinogenesis, and transfections of 3 of the microRNAs (miR200c*, miR143*, used to calculate 5-year survival data. To stabilize rates, we were reviewed. Approval from the IRB for the evaluation of they have been demonstrated to result in the acquisition and miR424*) signifi cantly abrogated invasive potential in used 5-year cohorts when calculating trends. human subjects was obtained. of more aggressive phenotypes. However, little is known Matrigel assays. RESULTS: RESULTS: Over a 20 month period, 33 patients (25 women, about the genome-wide alterations in microRNAs during A total of 89,733 patients were included in the CONCLUSION: median age 42) underwent extensive pancreatectomy with the neoplastic progression of colorectal cancers. With broad high-throughput evaluation study. The overall 5-year survival, inclusive of all locations and treatment approaches was is 5.4% (95% CI 5.2–5.6%). islet autotransplantation for pancreatitis. Mean follow-up METHODS: of microRNA expression across the spectrum of colon can- Using an ABI microRNA array platform, we cer stages, we have identifi ed a seven-microRNA signature For lesions in the head, body and tail the overall 5-year was 9 months with a range of 6–12 months. Postoperative simultaneously evaluated the expression of 668 microRNAs that is associated with more aggressive disease. In addition, survival, inclusive of all treatment approaches, was 5.2% complications occurred in 20 patients (60%). Preopera- in fresh frozen primary colonic adenocarcinomas across forced overexpression of 3 of the microRNAs resulted in an (95% CI 5–5.5%), 3.9% (95% CI 3.4–4.4%) and 7.8% (95% tive QOL scores were a mean 26 for Physical component the spectrum of AJCC stages (I–IV). Prediction analysis for attenuation of in vitro invasion, suggesting direct tumor CI 7.2–8.5%) respectively. For patients that underwent sur- and 35 for Mental Health component. Post-operatively, microarrays (PAM) was used to compare microRNA expres- suppressive function and further supporting the biologic gery, the 5-year survival for patients with lesions in all loca- Physical component scores averaged 32 at 6 months and sion across AJCC stages and build a microRNA stage classi- importance of complementary strand microRNAs. tions, head, body and tail was 18.8%, 15.9%, 23.8% and 35 at 12 months (mean increase of 9); the Mental Health fi er. Cross-validation was used to evaluate performance of 33.8% respectively. When analyzing trends we noted a sig- component scores averaged 42 at 6 months and 44 at 12 prediction error rate. False discovery rate (FDR) was con- nifi cant improvement in survival for patients that under- months (mean increase of 7). Pre-operative MEs averaged went operative intervention. Comparing the time periods 357 mg daily. At discharge from the hospital, this number Table from 1977–1981 versus 2002–2006, the 5-year survival for increased to 536 mg average MEs daily, a 50% increase, as surgical patients improved for all locations (9.2% to 22.6%, expected after major surgery in the chronic pain patient. AJCC AJCC AJCC Correlation with Correlation with Stage At 6 months, 15 out of 31 patients (48%) required less AJCC Stage AJCC Stage Stage AJCC Stage Stage Stage Stage Progression Progression p < 0.05) and also for lesions in the head (8.3% to 19.3%, IV vs. I IV vs. I IV vs. I III vs. II III vs. II III vs. II (I to IV) (I to IV) p < 0.05), body (6.7% to 26%, p < 0.05) and tail (14.7% to daily MEs than pre-operatively and averaged 161 mg daily (–55%). By 12 months, 11 out of 17 patients (65%) required Regulation Fold Change p-Value* ROC Fold Change p-Value* ROC r † p-Value 40.7%, p < 0.05). CONCLUSION: less daily MEs than pre-operatively and averaged 128 mg 7-microRNA signature 0.05 0.0002 0.98 0.31 0.0081 0.80 –0.63 <0.0001 The results of our study demonstrate an daily (–64%); 4 were narcotic free (23%). Of the 6 patients miR-200c* Down 0.07 0.0004 0.99 0.77 0.5396 0.53 –0.38 0.0016 improvement in survival for patients with pancreatic can- who did not decrease their analgesic requirements at one miR-15b* Down 0.02 0.0002 0.96 0.12 0.0021 0.75 –0.59 <0.0001 cer treated by operative intervention. This improvement in year, 5 (83%) still had an improved Physical QOL score miR-424* Down 0.08 0.0036 0.91 0.45 0.0445 0.67 –0.53 <0.0001 survival was more pronounced for patients with primary (one patient was unchanged) and all 6 had an improved miR-143* Down 0.05 0.0220 0.84 0.33 0.1155 0.63 –0.48 0.0001 lesions located in the body and tail. Mental Health QOL. miR-135a* Up 11.05 0.0026 0.94 2.06 0.0902 0.66 0.49 <0.0001 miR-378* Down 0.10 <0.0001 0.98 0.16 0.0001 0.84 –0.60 <0.0001 CONCLUSION: Total pancreatectomy with islet autotrans- miR-106b* Down 0.05 <0.0001 0.96 0.11 0.0007 0.80 –0.55 <0.0001 plant is an effective surgery for end stage chronic pancre- * Two-sample test atitis. Quality of Life improves early post operatively while † Pearson correlation test r = Correlation coeffi cient ROC = Receiver operator characteristic curve improved pain control and decreased prescription narcotic analgesia requirements occur later. Both improved Qual- ity of Life and decreased narcotic analgesia requirements continue to occur at least up to 1 year postoperatively. Further investigation is needed to assess the durability of total pancreatectomy with islet auto-transplantation for severe chronic pancreatitis with respect to pain relief and improved quality of life.

12 13 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

555 Percutaneous local ablative therapies (LAT) such as radio- /553 /554 frequency ablation (RFA) and ethanol injection (PEI) for

Budd-Chiari Syndrome Revisited: Thirty-Eight Years’ Abstracts HCC gained consent for its effi cacy and safety. Patients Superior Prognostic Importance of Perineural The Fate of Barrett’s Epithelium Under Acid Suppression Experience with Surgical Portal Decompression Sunday Invasion vs. Lymph Node Involvement After Obtained by Medical or Surgical Therapy submitted to resection (LR) or LAT have large differences Marshall J. Orloff1, Jon I. Isenberg1, Henry O. Wheeler1, on their characteristics. Propensity case matched studies Curative Resection of Duodenal Adenocarcinoma Renato Salvador1, Paola Parente3, Nicola Passuello1, Mario Costantini1, Pat O. Daily 1, Barbara Girard1 proved to reduce selection bias of retrospective analyses 1,3 1 Lisa Zanatta 1 1 1 Stefano Cecchini, Camilo Correa-Gallego, Vikram , Tieppo Chiara, Loredana Nicoletti, Francesco 1Surgery, University of California, San Diego, San Diego, CA and allow comparison between different therapies. Deshpande 2,4, Abdulmetin Dursun1, Jennifer A. Wargo1,4, Cavallin 4, Fabio Farinati1, Ermanno Ancona1, Massimo Rugge2, OBJECTIVE: AIMS: The aim of this study was to evaluate survival com- Carlos Fernandez del-Castillo 1,4, Andrew L. Warshaw1,4, Giovanni Zaninotto 2 To assess the validity of our observations regarding Budd-Chiari syndrome (BCS) reported 10 years paring LR and LAT in two groups of cirrhotic patients with Cristina R. Ferrone 1,4 1Department of Surgical and Gastroenterological Sciences, Clinica ago by expansion of our series and close follow-up for an HCC matched with propensity score analysis. 1General Surgery, Massachusetts General Hospital, Boston, MA; Chirurgica I, University of Padova, Padua, Italy;2Department of additional decade. METHODS: 2Pathology, Massachusetts General Hospital, Boston, MA;3General General Surgery, SS Giovanni e Paolo Hospital, ULSS 12, Venice, Four-hundred and seventy-eight cirrhotic BACKGROUND: patients with HCC treated with LR or LAT with cura- Surgery, Parma University, Parma, Italy, Italy; 4Surgery, Harvard Italy; 3Department of Medical Diagnostic Sciences & Special 10 years ago we reported our results with tive intent between January 1995 and December 2009 Medical School, Boston, MA Therapies, Pathology Unit, University of Padova, Padova, Italy; what remains the largest clinical experience in Western were included in the study. One hundred and eighty-one 4 countries with surgical portal decompression for BCS. Since BACKGROUND: Lymph node involvement has not been Surgical Oncology, Istituto Oncologico Veneto, IOV-IRCCS, then our series has expanded to 77 patients and there has patients underwent LR and 297 patients were treated with consistently shown to be a negative prognostic index for Padova, Italy been a worldwide explosion of interest in and publications LAT. Tumor stage and severity liver impairment was evalu- survival of patients with duodenal adenocarcinoma. Our BACKGROUND: Barrett’s Epithelium (BE) is the most about BCS. ated in all patients. To balance the covariates in the two aim is to identify stronger prognostic factors in survival of groups, a one-to-one propensity case-matched analysis was serious complication of GERD and it is associated with METHODS: 77 BCS patients were allocated to 3 groups: I – patients after resection of duodenal adenocarcinoma. used. A multivariate model based on age, sex, etiology of an increased risk to develop adenocarcinoma. Anecdotal 39 had hepatic vein occlusion alone, treated by side-to-side cirrhotic, Child-Pugh class, number of nodules, maximum METHODS: Of 102 patients with duodenal adenocarci- BE regression after profound acid suppression have been portacaval shunt (SSPCS); II – 26 had inferior vena cava diameter of nodules and serum alpha-fetoprotein level, noma evaluated between 1/1993–1/2010, 72 underwent reported. The aim of the study was to evaluate the modifi - occlusion treated by mesoatrial shunt in 8 and combined was used to obtain one-to-one match, using the nearest curative resection and 30 were explored and bypassed. Clin- cations induced by surgical and medical therapy on clinical SSPCS/cavoatrial shunt (CAS) in 18; III – 12 had decompen- neighbor matching model for LR and LAT. After propensity ico-pathologic factors were evaluated in resected patients. fi ndings, phenotype of BE and expression of the Cdx2 pro- sated cirrhosis too late for portal decompression, listed for matched analysis 176 patients were compared, 88 patients RESULTS: tein, as a gene involved in the metaplastic differentiation. (LT). An extensive diagnostic workup The 72 patients (33 females and 39 males) had in each group. PATIENTS AND METHODS: 89 Barrett’s patients with a included angiography with pressures and . Fol- a median age of 67 years. An R0 resection was performed in RESULTS: 62 patients. The majority of patients (51.6%) had Stage III follow-up endoscopy performed at least 12 months after low-up was 100%, lasting 5–38 years. Median survival was 65,1 months (95% CI 48.5–81.7) after LR and 37.3 months (95% CI 29.3–45.3) disease (anyT, N1). Median tumor size was 38 mm (range surgery or medical therapy were enrolled. Group A (n = 45) RESULTS: Group I, long-term survival is 95% with 36 after LAT (p = 0,008). For patients in Child-Pugh class A 2–130 mm) and a median of 13 (1–38) lymph nodes was received antirefl ux surgery Group B (n = 44) medical ther- free of ascites, leading lives of good quality 5 to 38 years with single HCC with diameter <5 cm, median survival examined. Perineural invasion was identifi ed in 18 patients apy. Patient symptoms were evaluated using a detailed post-SSPCS. Group II, mesoatrial shunt was discontinued was 65 months (95% CI 58.4–71.6) for LR group and 63.7 (29%). Median follow up was 25 months. The 5-year dis- questionnaire; endoscopy with biopsy was performed because of a high failure rate, but combined SSPCS-CAS has months (95% CI 31.8–95.7) for LAT group (p = 0.73). For ease specifi c survival (DSS) was 78% vs 54% for pN0 and according to Seattle; histologic assessment. of BE pheno- 100% survival for 5–25 years. Group III, 6 patients (50%) patients in Child-Pugh class A with single HCC with diameter pN+ patients (p < .001) and 78% vs 13% for absence and type modifi cations induced by therapy (presence of I.M., are alive and well for over 5 years post-LT. Serial liver biop- ≥5 cm, median survival was 79.9 months (95% CI 40.1– presence of perineural invasion (p < .001) respectively. type of I.M., expression of Cdx2) were quantifi ed by means sies following portal decompression show long-term rever- 119,8) for LR group and 21,5 months (95% CI 10.8–32.1) Lymph node ratio (LNR), type of resection (whipple vs. seg- of histology (H&E), histochemistry (HID) and himmuno- sal of BCS lesions. mental resection), and size of tumor failed to stratify prog- histochemistry. Barium swallow, manometry and 24-h pH for LAT group (p = 0.023). For patients in Child-Pugh class CONCLUSIONS: Long-term survival following surgical nosis. Multivariate analysis indentifi ed perineural invasion were performed pre and 6 months after surgery. Surgical A with 2–3 nodules and <3 cm of maximum diameter, portal decompression of BCS has not been equaled by other as the most powerful signifi cant independent predictor of failures were defi ned as: (1) abnormal 24-h pH monitoring; mean survival was 69.3 months (95% CI 48.7–89.9) for LR forms of therapy, medical or surgical. It is imperative to survival (HR 5.8, p = .003 CI: 1.792–18.749). (2) recurrence of endoscopic esophagitis; (3) recurrence of group and 45,7 months (95% CI 22.8–68.7) for LAT group perform surgical portal decompression early in the course (p = 0.168). For patients in Child-Pugh class A with 2–3 CONCLUSIONS: hiatal hernia/slipped fundoplication at endoscopy/barium Perineural invasion is a stronger pre- swallow. of BCS to avoid irreversible liver damage. nodules and maximum diameter >3 cm, median survival dictor for survival than tumor size, T stage, lymph node was 82.9 months (95% CI 52–113.7) for LR group and 18,9 RESULTS: The median of follow up was 34 months (IQR involvement, LNR and type of resection. 556 months (95% CI 6.3–31.4) for LAT group (p = 0.001). 23–48). The symptom score decreased after both therapies: 14.5 (10–21) pre vs 0 (0–6) post (p = 0.0001). The decrease Surgical Resection Versus Local Ablation for HCC on CONCLUSION: Our propensity case-matched study con- of endoscopic length of BE was more signifi cant in group Cirrhosis: Results from a Propensity Case-Matched fi rmed that survival is similar after LR and LAT for single A than in group B (p = 0.006), and generally BE length Study HCC smaller than 5 cm and for oligofocal HCC (up to 3 nodules) smaller than 3 cm; instead, for HCC larger than decreased more in patients with short segment BE (SSBE, Andrea Ruzzenente, Tommaso Campagnaro, Alessandro <3 cm BE), than in those with long segment BE (LSBE, 5 cm or oligofocal HCC (up to 3 nodules) larger than 3 cm Valdegamberi, Simone Conci, Fabio Bagante, Gianni Turcato, surgical resection improves signifi cantly long term survival ≥3 cm BE), (p = 0.02). Regression of I.M. was observed in Calogero Iacono, Alfredo Guglielmi 12 of 89 patients after therapy: all cases occurred in SSBE patients. In patients in which I.M. was no longer evident Department of Surgery, University of Verona Medical School, after therapy, Cdx2 expression was also absent. Two SSBE Verona, Italy patients had recurrence of refl ux after surgery. In the 13 BACKGROUND: Surgery for hepatocellular carcinoma SSBE patients in whom surgery was effective BE regression (HCC) had great improvements in last decades with low was observed in 7, signifi cantly different from the medical morbidity and mortality and with good long-term results. group. (p = 0.02). CONCLUSIONS: Medical and Surgical therapies are both effective in controlling GERD symptoms. Patients with SSBE (but not LSBE) may regress with adequate acid sup- pression. Surgery—when effective—leads to regression more frequently than medical treatment.

14 15 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

PATIENTS AND METHODS: A questionnaire-based study MONDAY, MAY 9, 2011 š557 was proposed to 143 morbidly obese patients attending bar- Pure Laparoscopic for HCC Patients withiatric surgery. Data included demographics, past medical, Severe Liver Cirrhosis surgical and obstetrics histories, as well as obesity related 7:30 AM – 9:15 AM co-morbidities. Wexner Constipation Score (WCS) and the Zenichi Morise Faecal Incontinence Severity Index (FISI) questionnaires S504 Department of Surgery, Fujita Health University Banbuntane were used to evaluate and constipation incontinence. For VIDEO SESSION II: Houtokukai Hospital, Nagoya, Aichi, Japan the purpose of this study we considered abnormal a WS ≥5 BREAKFAST AT THE MOVIES The patients with severe liver cirrhosis who undergo hepa- and a FISI score ≥10. Same questionnaires were completed tectomy often develop post-operative liver failure, even if after bariatric surgery at 3 and 6 months follow-up. the hepatectomy is very limited. RESULTS: A total of 114 patients accepted the study since For the patients, pure laparoscopic hepatectomy (PLH) March 2010, 72 F, 42 M, mean age of 43 years (range 17–73).

611 613 Abstracts minimize destruction of the collateral blood and lymphatic Mean BMI before surgery was 46 ± 8 kg/h2 (range 35–67 Monday fl ow caused by and mobilization of the liver kg/h2). Mean WS was 4.2 ± 4 (range 0–17) while mean FISI Complete Traumatic Rupture of the Pancreas— Laparoscopic Frey Procedure for Chronic Calcifi c and, also, mesenchymal injury caused by compression of score was 9.0 ± 8 (range 0–38). Overall 61% of the patients Pancreas Preserving Reconstruction: A Non-Resectional Pancreatitis reported DDs according to our scores. Twenty-six patients the liver. It restrains the complications, which lead to the Procedure Using Pancreatogastrostomy and a Modifi Jessicaed M. Gutierrez, Michael L. Kendrick postoperative serious liver failure, such as massive ascites. (23%) had WS ≥ 5, 27 patients (24%) a FISI ≥ 10, while 17 “Blumgart” Anastomosis Mayo Clinic, Rochester, MN Severe cirrhotic patients with tumors on the surface of the patients (15%) reported combined abnormal scores. These liver, in case of diffi cult adaptation of RFA and/or local percentages correlates with BMI ranging from 47% with Hannes P. Neeff, Frank Makowiec, Tobias Keck, Ulrich T. HoptBACKGROUND: The laparoscopic approach for the Frey recurrence after repeat treatments, are the good candidates BMI ≤40 kg/h2 to 62% with BMI ≥50 kg/h2. Department of Surgery, University of Freiburg, Freiburg, Germany procedure has not been previously reported. for PLH. Forty-four patients underwent surgery with a minimum We present a case of a 23 yo female with an isolated cen- AIM: This video demonstrates the feasibility of the lapa- follow-up of 3 months. Mean BMI decreased signifi cantly tral pancreatic rupture. Since diagnosis was made early, a roscopic approach for the Frey procedure. Reproduction 558 from 46 kg/h2 to 38 ± 8.5 kg/h2 and to 35 ± 9 kg/h2 respec- non-resectional procedure could be performed. The rem- of the same key steps utilized in the open approach is nant of the pancreatic tail was managed using pancreato- Prevalence of Defaecatory Disorders in Morbid Obesetively at 3 and 6 months after surgery. Accordingly to the demonstrated. BMI decrease, WS improved, but signifi cantly, from 4.2 to for exocrine drainage. The central surface of Patients Before and After Bariatric Surgery RESULTS: The procedure was completed in 308 minutes 3.9 ± 5.2 (NS) at 3 months and to 3.1 ± 4.1 (NS) at 6 months the pancreatic head was treated using a serosal patch of and with an estimated blood loss of 50 ml. The patient Pierpaolo Sileri, Paolo Gentileschi, Ida Camperchioli, after surgery. Similarly, the FISI score improved from 9.0 jejunum in order to prevent leakage. The anastomosis was postoperative course was uncomplicated and she was dis- Elisabetta De Luca, Domenico Benavoli, Luana Franceschilli, to 2.7 ± 5.5 after 3 months (NS) and to 0.3 ± 1.1 after 6 carried out in a modifi ed “Blumgart” fashion. Recovery was missed on postoperative day 4. Giulio P. Angelucci, Sara Lazzaro, Achille Gaspari months (p = .006). uneventful and exocrine and endocrine function could be completely preserved in this young female adult. CONCLUSIONS: The laparoscopic Frey procedure is fea- Surgery, University of Rome Tor Vergata, Rome, Italy CONCLUSIONS: defaecatory disorders are more frequent sible and can be performed with the same surgical steps as BACKGROUND: in morbidly obese patients compared to the general popu- The video was recorded from the surgeon’s point of view Morbid obesity is associated with defae- for the open approach. catory disorders (DDs) such as faecal incontinence and lation. The risk of DDs increases with BMI. Bariatric surgery with an overhead camera. constipation. However data on their prevalence as well reduces DDs, mainly faecal incontinence, and these fi nd- as effectiveness of bariatric surgery on their correction is ings correlated with BMI reduction. 612 614 scant. The primary objective of this study was to estimate Minimally Invasive Esophageal Stripping for Early Adrenocortical Carcinoma with Intracaval Extension the effect morbid obesity on DDs in a cohort of patients Esophageal Cancer to the Right Atrium: Resection on Cardiopulmonary waiting for bariatric surgery. We also evaluated preliminar Bypass results of the effects of bariatric surgery on these disorders Daniel K. Tong, Simon Law, Fion S. Chan, Kam H. Wong Ryan Z. Swan, Kwan N. Lau, David Sindram, David A. Iannitti, in a subgroup of patients who already underwent surgery. Surgery, The University of Hong Kong, Hong Kong, Hong Kong John B. Martinie A 48-year-old lady had an early, mid-third diffuse esopha- Surgery, Carolinas Medical Center, Charlotte, NC, geal squamous cell carcinoma. The stomach was mobilized United States laparoscopically. The esophago-gastric junction was tran- sected using a linear stapler. The cervical esophagus was Resection of advanced adrenocortical carcinoma is the only divided. A varicose vein stripper was inserted via the proxi- potentially curative treatment option. A 28 year-old female mal esophageal stump and the olive tied around it. The tip presents with dyspnea on exertion and is found to have of the stripper was brought through the transected lower a large (19 × 18 cm) right adrenal mass invading the liver esophagus. The esophagus was stripped into the peritoneal and inferior vena cava (IVC) with tumor thrombus extend- cavity, and brought outside the abdomen via a mini-right ing into the right atrium. While on cardiopulmonary subcostal wound. A hand-sewn cervical anastomosis was bypass, the patient underwent tumor resection, extraction fashioned after bringing the stomach to the neck ortho- of tumor thrombus from the IVC and right atrium, and topically. Final pathology showed a pT1a cancer. partial hepatectomy. The abdomen was closed with a tem- porary closure device then closed primarily the following day. The patient was extubated on post-operative day 3, discharged home on day 12, and is undergoing adjuvant chemoradiation.

16 17 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

615 ICG-15 test was 8%. The patient underwent to right trisec- of patients with advanced stage disease, poor pathologic examined the impact of surgery alone with adequate nodal tionectomy with “en-bloc” resection and reconstruction of features such as vascular or perineural invasion, or who evaluation (≥15 lymph nodes) on survival in non-meta- Laparoscopic Two-Staged Hepatic Resection portal vein bifurcation extended to common . recurred after resection were PCR+ as compared to CYT+. static gastric cancer. Farah A. Husain, Kalyana Nandipati, Edward Lin, METHODS: Juan M. Sarmiento The surgical specimen confi rmed the radicality of surgery An R0 resection was performed in 85 (87%) patients: only one Using the 2001–08 California Cancer Regis- and the diagnosis of cholangiocarcinoma with involve- (1%) was CYT+ while 13 (15%) were PCR+. Of this group, try, we identifi ed 2,229 patients who underwent gastrec- General Surgery, Emory University Hospital, Atlanta, GA ment of portal vein bifurcation and of the hilar bile duct, PCR+ demonstrated a worse survival than PCR- patients tomy with adequate nodal evaluation (≥15 lymph nodes) A 27 yo female presented with acute onset of RUQ pain. AJCC TNM 7th ed.: pT3 N1 (1/18) M0. (p = .02). (Figure 1A). Further analysis showed that of the for American Joint Committee on Cancer stage (6th edi- She was diagnosed with multiple hepatic adenomas in the stage III/IV, CYT– subgroup, PCR+ had a trend towards tion) I-IVM0 gastric adenocarcinoma. Patient, tumor and left lateral segments and the right lobe, with hemorrhage. 617 worse survival (p = .09) than PCR- patients (Figure 1B). treatment-variables were compared by type of adjuvant The patient underwent a completely laparoscopic left lat- therapy. Cox proportional hazard analysis models were eral segmentectomy. Following discharge and stable recov- Ligation of Intersphincteric Fistula Tract: A Simple and used to evaluate the impact of surgery alone on overall sur- ery, the right portal vein was embolized. A hand-assisted Safe Technique to Treat Anal Fistulas vival (OS) and cancer specifi c mortality (CSM) adjusting for Abstracts

laparoscopic right liver lobectomy was done 6 weeks later Pierpaolo Sileri, Luana Franceschilli, Giulio P. Angelucci, covariates. Monday to remove residual adenomas. The patient was discharge Sara Lazzaro, Emanuele Picone, Achille Gaspari RESULTS: Of our total cohort, 70% had T1/T2 disease and in three days. This case reveals a new approach to patients Surgery, University of Rome Tor Vergata, Rome, Italy 30% had N0 disease. Up to 49% of our cohort underwent with multiple hepatic lesions who may otherwise require surgery alone. Those patients were more likely to be older, transplantation or multiple open resections for cure. A A novel approach through the intersphincteric plane for the treatment of anal fi stulas, known as LIFT (Ligation of black, carriers of Medicare coverage, with T1, and N0 dis- staged laparoscopic approach can minimize morbidity and ease (p ≤ 0.05). Overall, persons who underwent surgery hasten recovery. Intersphincteric Fistula Tract)as been recently proposed. It shows low recurrence rates and neglectable effects on anal alone experienced worse OS and CSM than those who continence. This technique consists in the identifi cation of received adjuvant chemoradiotherapy. However, when 616 the fi stula tract, its ligation and section as close as possible stratifi ed by AJCC-stage and nodal involvement, persons Right Trisectionectomy with “en-bloc” Resection of to the internal sphincter associated to the excision of the CONCLUSION: RT-PCR for CEA increases detection of sub- with AJCC stage I or N0 disease who had adequate nodal Portal Vein for Peri-Hilar Cholangiocarcinoma infected cryptoglandular tissue. In this video we present microscopic peritoneal disease and is more sensitive than evaluation experienced more favorable survival outcomes with surgery alone than those who received adjuvant Alfredo Guglielmi, Andrea Ruzzenente, Alessandro Valdegamberi,the LIFT technique resuming its surgical steps and, briefl y, cytology. PCR positivity was associated with decreased sur- our initial experience. vival. Further followup is required to determine if PCR can chemoradiotherapy. These results persisted on our Cox Tommaso Campagnaro, Calogero Iacono regression analysis to show that in AJCC stage I or N0 dis- Department of Surgery, University of Verona Medical School, be used as an independent predictor of poor survival in gastric cancer. ease, surgery alone predicted more favorable OS and CSM Verona, Italy than when combined with adjuvant therapies (Table). Survival distribution curves of R0 resected patients (A) and This video shows the clinical case of a patient with peri- CONCLUSION: R0, stage III/IV, cytology negative patients (B) by PCR sta- Surgery alone with adequate nodal evalu- hilar cholangiocarcinoma (10 cm) of the right hepatic lobe tus. In patients who underwent R0 resection, PCR+ patients ation may have a role in the treatment of localized gastric with portal vein involvement. had a signifi cantly worse survival than PCR– patients cancer. To corroborate these fi ndings, surgery alone with The patient was a 79 years-old female with no previous his- (p = .02). In the R0 resected, stage III/IV, cytology nega- adequate nodal evaluation (as a treatment arm) deserves tory of liver disease. Preoperative liver blood tests showed tive patients, PCR+ had a trend towards worse survival than consideration in the design of future gastric cancer trials only increased GGT and CA 19.9 level. FRLV was 43% and PCR- patients (p = .09). to provide patients effective yet resource-conserving, rather than maximally tolerated, treatments. /681 10:00 AM – 11:15 AM Is There a Role for Surgery Alone with Adequate Nodal S501a Evaluation in Gastric Adenocarcinoma? PLENARY SESSION IV Vikas Dudeja1, Elizabeth Habermann1, Anasooya Abraham1, Wei Zhong 1, Helen M. Parsons1, Jennifer F. Tseng2, Waddah B. Al-Refaie 1 1University of Minnesota and Minneapolis VAMC, Minneapolis, with biopsy proven gastric cancer undergoing staging lap- MN; 2 680 University of Massachusetts Medical School, Surgical / aroscopy. These washings were analyzed by both Papani- Outcomes Analysis & Research (SOAR), Worcester, MA colaou staining and real-time polymerase chain reaction RT-PCR Increases Detection of Submicroscopic INTRODUCTION: Peritoneal Metastases in Gastric Cancer and Has (RT-PCR) for the tumor marker carcinoembryonic antigen The extent of lymphadenectomy and Prognostic Signifi cance (CEA). protocol design in gastric cancer trials limit the interpre- RESULTS: tation of survival benefi t of adjuvant therapy for those Joyce Wong, Kaitlyn J. Kelly, Arjun Mittra, Mithat Gonen, Peritoneal disease was discovered at laparos- who undergo surgery with adequate nodal evaluation. We Peter J. Allen, Yuman Fong, Daniel G. Coit copy in 38 of the 156 patients (24%), (LAP+). Cytology was positive (CYT+) in 23 of these patients while RT-PCR was Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY positive (PCR+) in 30. The sensitivity of CYT for detection BACKGROUND: Positive peritoneal cytology confers of disease was 61% vs. 79% for PCR (p = .02). the same prognosis as stage IV disease in gastric cancer. Non-stage IV disease was identifi ed at laparoscopy in 118 Conventional cytology, however, has low sensitivity. We (76%) patients. Eight (7%) were CYT+ while 28 (24%) were propose that RT-PCR may have increased sensitivity and PCR+. Long term follow-up demonstrated worse survival provide more accurate staging information. of PCR+CYT– (p = .0003) and PCR+CYT+ (p = .0004) com- METHODS: From 2/2007 to 4/2009, peritoneal lavage pared to CYT-PCR- patients. Resection was performed in samples were collected prospectively from 156 patients 98 of these patients. A signifi cantly greater proportion

18 19 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Table: Multivariate Analysis of Surgery Alone with Adequate Nodal Evaluation on OS and CSM* CONCLUSION: Our results confi rm a pattern of expres- GLUT2 to glucose absorption was 27% at 10 mM but more sion of YKL-40 in colon cancer that may be a potential pre- than double, 56%, at 100 mM (p < 0.01). SUMMARY: Aug- Surgery alone and Mortality (Stratifi ed by AJCC Stage) dictive biomarker to identify high risk patients with Stage mented, carrier-mediated glucose absorption at infusions Stage I HR (95% CI) Stage II HR (95% CI) Stage III HR (95% CI) Stage IV HR (95% CI) II disease who warrant adjuvant therapy. Furthermore, of 100 mM glucose appears mediated by GLUT2 (SGLT1 Cancer Specifi c Mortality Surgery alone 0.28 (0.15–0.53) 1.69 (1.14–2.53) 1.56 (1.17–2.07) 1.78 (1.31–2.40 the uniquely prognostic importance of YKL-40 expression Km ≈ 3–6 mM and would be saturated at ≥20 mM glucose Surgery + Chemotherapy 0.70 (0.25–1.92) 0.90 (0.51–1.58) 1.12 (0.81–1.55) 1.05 (0.74–1.49) within the stroma may be hypothesis-driven by empha- solutions). The marked inhibition of glucose absorption at Surgery + Chemoradiotherapy Ref. Ref. Ref. Ref. sizing the role of the tumor stroma in cancer progression. 100 mM glucose by SGLT1 inhibition (phlorizin) implicates Overall Mortality The pro-angiogenic mechanism of YKL-40 could provide a SGLT1 activity as being necessary in this GLUT2-mediated Surgery alone 0.45 (0.26–0.77) 1.68 (1.21–2.35) 1.51 (1.17–1.95) 1.95 (1.49–2.54) more targeted approach towards the use of anti-angiogenic process. There is a GLUT2-mediated component of glucose Surgery + Chemotherapy 0.74 (0.31–1.73) 0.97 (0.63–1.50) 1.07 (0.79–1.44) 1.12 (0.83–1.52) therapy for colon cancer. Our study warrants further vali- absorption (~27%) even at 10 mM glucose. Surgery + Chemoradiotherapy Ref. Ref. Ref. Ref. dation and supports closer investigation into the tissue- Surgery alone and Mortality(stratifi ed by AJCC-N status)** specifi c mechanism of YKL-40.

Node negative Node Positive Abstracts

High YKL-40 expression predicts worse disease-free sur- Monday Cancer specifi c Mortality Overall Mortality Cancer specifi c Mortality Overall Mortality vival. Level of expression (Low, Intermediate and High) Surgery alone 0.49 (0.26–0.92) 0.66 (0.39–1.12) 1.71 (1.42–2.04) 1.71 (1.46–2.00) Surgery + Chemotherapy 0.65 (0.27–1.60) 0.56 (0.25–1.26) 1.20 (0.97–1.49) 1.22(1.01–1.47) defi ned by a scoring method incorporating intensity and Surgery + Chemoradiotherapy Ref. Ref. Ref. Ref. percent of cells stained by IHC. *After adjusting for age, race, year of diagnosis, insurance, tumor grade, and extent of gastrectomy. ** We also adjusted for AJCC T-stage. Abbreviations: HR, hazard ratio; CI, confi dence intervals; ref, referent. /684 Acute Enterocyte Adaptation to Luminal Glucose: A Posttranslational Mechanism for Rapid Apical Recruit- ment of the Transporter GLUT2 Rizwan M. Chaudhry, Jeffrey S. Scow, Srivats Madhavan, a unique pattern of expression using immunohistochem- Judith A. Duenes, Michael G. Sarr š682 istry (IHC) on formalin-fi xed paraffi n-embedded material Gastrointestinal and General Surgery, Mayo Clinic, Rochester, MN Laparoscopic and Robotic Distal Gastrectomy with collected from 282 patients diagnosed with Stage II to IV colon cancer. Furthermore, our analysis identifi es YKL-40 Glucose absorption early after a meal increases markedly to Lymphadenectomy for Gastric Adenocarcinoma in levels far greater than possible by the classic glucose trans- an Obese Patient tissue expression to be highly predictive of prognosis in Stage II node-negative cancer. porter sodium-glucose cotransporter 1 (SGLT1). Figure 1A: Inhibition of in vivo glucose absorption by phloretin (PT) Joshua Ellenhorn and phlorizin (PZ) at 10 mM glucose (black arrow at 60 min represents METHODS AND MATERIALS: HYPOTHESIS: High luminal concentrations of glucose City of Hope National Medical Center, Duarte, CA From our cohort, 216 administration of inhibitor). patients were identifi ed with Stage II disease. Employing lead to rapid (within minutes) phenotypic, non-genomic In Asia, laparoscopic assisted gastric resection is com- a YKL-40 antibody, IHC was performed on archival tissue related adaptations by the enterocyte to recruit another monly performed with open reconstruction. In the United from primary tumors. Expression was measured by a scor- transporter, glucose transporter 2 (GLUT2), to increase glu- States, minimally invasive techniques have been adapted ing method based on intensity of staining and percentage cose absorption. to enable gastric resection with intracorporeal reconstruc- of stained cells. Cox regression analysis correlated YKL-40 AIM: To determine if glucose absorption increases early tion. The aim of this video presentation is to demonstrate tissue expression with disease-free survival, controlling for postprandially by a GLUT2-mediated mechanism. an approach to minimally invasive gastric cancer surgery. A stage and lymphatic invasion. METHODS: hybrid laparoscopic and robotic approach is used to accom- With IACUC approval we perfused in vivo 30 plish a radical distal gastrectomy with lymphadenectomy cm jejunal segments in Lewis rats (n = 36 rats) for 120 min in a typical obese Western habitus patient with a mid antral with isoosmolar solutions of low (10 mM) and high (100 gastric malignancy. Standard laparoscopy is used for omen- mM) glucose concentrations (n = 6 rats in each) containing tectomy, gastric transaction and reconstruction. Lymphad- 14C-D-glucose and 3H-L-glucose to quantitate separately enectomy is accomplished with the use of a surgical robot. carrier-mediated (stereospecifi c) and passive (non-stereo- specifi c) absorption, respectively. We also evaluated effects of 1 mM phlorizin (SGLT1 inhibitor) and 1 mM phloretin 683 (GLUT2 inhibitor) at low and high glucose concentrations. YKL-40: A Predictive Biomarker for Stage II Colon RESULTS: Total glucose absorption (carrier-mediated and Cancer passive absorption) was much less in 10 mM than the 100 Richard B. Arenas1,3, Hannah Swayze-Quinn1, Christopher mM glucose solutions (mean ± SEM, 2.8 ± 2.0 vs 14.8 ± N. Chapman 2, Frida Rosenblum2, Jane Garb1, Rong Shao3, 3.7 µmol/30 cm/min, respectively; n = 6 rats in each; p < Figure 1B: Inhibition of in vivo glucose absorption by phloretin (PT) Q. Jackie Cao 2 0.001). Importantly, stereospecifi c carrier-mediated uptake and phlorizin (PZ) at 100 mM glucose (black arrow at 60 min represents 1Surgery, Baystate Medical Center, Springfi eld, MA;2Pathology, increased from 1.9 ± 1.3 to 13.9 ± 2.8 µmol/min (p < 0.001), administration of inhibitor). RESULTS: YKL-40 is uniquely expressed in both epithelial Baystate Medical Center, Springfi eld, MA; 3Pioneer Valley Life while passive (non-stereospecifi c) uptake increased, as and stromal cells from colon cancers. With a median fol- CONCLUSION: A small amount of GLUT2 is present api- Sciences Institute, Springfi eld, MA expected by much less, from 0.7 ± 2.0 to 1.3 ± 1.9 µmol/30 low-up of 48 months, high expression of YKL-40 within the cm/min (p < 0.001). In separate experiments (n = 6 rats cally in the enterocyte (constitutively expressed), but when INTRODUCTION: YKL-40, a 40-kDa secreted glycoprotein stroma of the tumor correlated with disease-free survival each) carrier-mediated glucose absorption in 10 and 100 exposed to high luminal concentrations of glucose, the also known as human cartilage-glycoprotein 39 or Chitin- (p = 0.036) while controlling for other known prognostic mM glucose solutions was decreased by phlorizin (SGLT1 enterocyte changes its phenotype by recruiting GLUT2 ase 3-like 1, has recently been shown to promote cancer variables. High expression within the tumor epithelium inhibitor) from 2.5 ± 0.2 to 0.6 ± 0.1 µmol (p < 0.02) and apically (within minutes- thus mediated through a non- growth and progression through angiogenesis. Despite did not correlate with prognosis (p = 0.241). Our results 9.4 ± 2.3 to 2.0 ± 0.9 µmol (p < 0.01), respectively, and genomic, posttranslational mechanism) to markedly aug- the correlation of high serum levels of YKL-40 with poor demonstrated that patients with high stromal expression by phloretin (GLUT2 inhibitor) from 2.4 ± 0.3 to 1.7 ± ment glucose absorption. This early postprandial process prognosis in colon cancer, the signifi cance of tissue expres- of YKL-40 in their tumors were more than twice likely to 0.2 µmol (p <0.02) and 10.0 ± 1.2 to 4.4 ± 1.0 µmol/min, represents a form of “acute phenotypic adaptation”. sion of YKL =40 is yet to be described. We have identifi ed relapse (hazard ratio 2.4 ± 1.0). respectively (p < 0.03) (Fig.1). The relative contribution of

20 21 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

10:00 AM – 11:15 AM tis and one patient with stent migration. The presence of /688 S505 SEMS did not interfere with pancreaticoduodenectomy in any patients who underwent surgery. Investigation for Biomarkers of Barrett’s Esophagus QUICK SHOTS SESSION I Jawad Ahmad1,2, Ken Arthur3, Perry Maxwell2, Helen G. Mulholland1, J.A. Kennedy 2, Liam Murray2, Brian T. Johnston2, Damian T. McManus 2 1School of Medicine, Dentistry and Biomedical Sciences, Queens 685 CONCLUSION: Preoperative plasma PGE2 levels were University Belfast, Belfast, United Kingdom;2Belfast Health and evaluated in CRC pts and compared to those in benign 3 Plasma Levels of Prostaglandin E2 (PGE2), a Protein Social Care Trust, Belfast, United Kingdom; Tissue CTU, Belfast, colorectal disease pts. The mean plasma PGE2 level in CRC with Proangiogenic Effects, Are Elevated in Colorectal United Kingdom pts was 43% higher than the mean level in benign pts. Cancer Patients INTRODUCTION: The higher levels in CRC pts could be attributed to tumor- The incidence of oesophageal adeno- Abstracts M.C. Shantha Kumara H, Joon H. Jang, Sajith A. Herath, related infl ammation, progression, and angiogenesis. Fur- carcinoma (OAC) has increased dramatically over recent Monday Daniel D. Kirchoff, Xiaohong Yan, Vesna Cekic, ther studies with a larger population of cancer free control years and Barrett’s esophagus (BE) is the most established Richard L. Whelan and CRC pts, including more stage-4 patients, are neces- risk factor for its development. Endoscopic surveillance of BE has been widely advocated but hinges on assessment Surgery, St. Luke Roosevelt Hospital, New York, NY sary to better determine whether PGE2 levels correlate with Cancer stage and prognosis. of repeated endoscopic biopsies, which is problematic and INTRODUCTION: Prostaglandin E2 (PGE2), a major not reliably accurate. The use of biomarkers presents an metabolite produced by cylooxygenase 2(COX-2), plays a 687 opportunity to reduce sampling bias and improve our abil- role in tumor development and progression. PGE2 is the ity to risk-stratify these patients. ligand for 4 prostaglandin type E (EP) receptor subtypes, Effi cacy of Self Expanding Metal Biliary Stents in AIMS AND OBJECTIVES: We examined three novel bio- EP1 to EP4. The expression patterns of these receptors on Patients Receiving Preoperative Chemoradiation for markers namely P504S, CD133 and Twist, in the setting of a variety of cell types accounts for the diverse functions Resectable Pancreatic Cancer BE, low grade dysplasia (LGD) and OAC to evaluate their of PGE2. Over-expression of COX2 and PGE2 has been 1 1 1 2 A. Aziz Aadam, Abdul H. Khan, Young Oh, Doug B. Evans, potential to differentiate between benign, dysplastic and observed in colorectal cancer (CRC) and PGE2 binding Kulwinder S. Dua 1 malignant Barrett’s tissue in the EDRN exploratory Phase I to EP2 receptors on endothelial cells in blood vessels of 1 & II study. tumors directly impact tumor angiogenesis by enhanc- Division of Gastroenterology and Hepatology, Medical College 2 ing cell survival. Also, COX-2/PGE2 contributes to VEGF of Wisconsin, Milwaukee, WI;Department of Surgery, Medical MATERIALS AND METHODS: After ethical approval, and CXCL1 mediated angiogenesis which is necessary for College of Wisconsin, Milwaukee, WI 25 cases each of BE, LGD and OAC were included along- tumor growth. There is also evidence that COX2 derived INTRODUCTION: Recent studies have shown a survival with 25 cases of oesophagectomy resections for Barrett’s PGE2 promotes colorectal cancer growth via activation of OAC. The biomarkers were immunostained on automated benefi t with preoperative chemoradiation for resectable ® epidermal growth factor receptor (EGFR) signaling. Ele- and borderline resectable pancreatic adenocarcinoma. Bili- Ventana immunostainer. The archived biopsy materials were assessed for biomarkers expression by two indepen- vated levels of EGF, VEGF, and CXCL1 have been noted in ary stent occlusion as high as 60–93% in prior series* has CONCLUSION: SEMS should be considered to relieve bili- dent observers using a QScore method. Any inter-observer CRC patients; however, plasma PGE2 levels in CRC patients been a major issue with plastic stents during the 2 to 4 ary obstruction in patients with resectable pancreatic can- score discrepancy of ≥2 was discussed for a consensus. have not been well studied. This study’s purpose was to months of neoadjuvant therapy. cer receiving neoadjuvant therapy in view of their superior compare preoperative plasma PGE2 levels in populations AIM: patency rates and fewer re-interventions for recurrent bili- RESULTS: P504S did not express in any case of BE. How- of CRC and benign colonic disease patients. The aim of our study was to evaluate the patency of self-expanding metal stents (SEMS) during the pre-opera- ary obstruction in contrast to the historical experience with ever, its expression was signifi cant in LGD, OAC and resec- METHODS: Patients (pts) undergoing colorectal resec- tive chemoradiation period. plastic stents. The presence of SEMS does not interfere with tions. CD133 also did not express in any of the BE or LGD. Its comparative expression was up-regulated in cases of tion for benign colonic disease or adenocarcinoma were METHODS: subsequent pancreaticoduodenectomy and removal prior prospectively enrolled in an IRB approved tissue and data We reviewed the outcomes of consecutive to surgery is not required. OAC and resections. Twist expression was weak in BE and patients from 2008 to 2010 with resectable, biopsy-proven LGD but over-expressed in cases of OAC and resections. banks. Preoperative blood samples, basic demographic *Mullen et al. J Gastrointest Surg. 2005;9:1094–1105. data, type of resection performed, other clinical informa- pancreatic adenocarcinoma and resulting biliary obstruc- (See Table below). tion. Patients were categorized as resectable or borderline *Wasan et al. Am J Gastroenterol. 2005;100:2056–61. tion, and pathology results were prospectively collected. DISCUSSION: This cross sectional study has shown Blood samples were processed in a timely manner and resectable based on CT imaging criteria. SEMS were placed in all patients and the shortest length stent required to increased expression of P504S, CD133 and Twist in the plasma was stored at –800 C until further use. Plasma PGE2 metaplasia-dysplasia-adenocarcinoma sequence and has was determined in duplicate via ELISA, and results are bridge the stricture was selected. All patients received neo- adjuvant therapy. Data on demographics, stent patency, suggested their possible role as potential biomarkers of Bar- reported as mean ± SD. A t-test was used to compare PGE2 rett’s progression. levels between groups. Signifi cance was defi ned as p < 0.05. complications and re-intervention rates were prospectively collected. Further longitudinal and prospective (EDRN III & IV) stud- RESULTS: A total of 148 pts were enrolled; 78 CRC (83% RESULTS: Fifty-two (21 resectable, 31 borderline resect- ies are required to validate these results. colon, 17% rectal) and 70 benign condition pts (51% diver- RESULTS: ticulitis, 44% adenoma). The mean male to female ratios in able) patients underwent SEMS (Wallfl ex, Boston Scientifi c, the groups were comparable, but the CRC pts were older 8 covered, 44 uncovered) placement. Median time from (p = 0.002). The fi nal cancer staging in the CRC group: stent placement to surgery was 99 days for all patients BE BE LGD LGD OAC OAC Resections Resections stage 1, 24 (31%); stage 2, 28 (36%); stage 3, 18 (23%); stage (resectable 83 days, borderline 133 days, p = 0.10). SEMS Biomarker Positive Negative Positive Negative Positive Negative Positive Negative P504S 0 25 14 (56%) p 11 10 (40%) p 15 15 (60%) p <0.001 10 4, 8 (10%). The mean PGE2 level was signifi cantly higher in occlusion occurred in 6 out of 25 patients (24%) who ultimately underwent surgery and in 3 out of 27 patients <0.001 <0.001 the CRC group than the benign group (676.8 ± 582.5 pg/ml CD133 0 25 0 25 6 (24%) p 0.022 19 17 (68%) p <0.001 8 (11%) who did not undergo surgery (Figure 1). No statisti- vs. 472.5 ± 424.8 pg/ml, p = 0.002). No signifi cant differ- Twist 1 24 3 (12%) p 22 14 (56%) p 11 9 (36%) p 0.005 16 ences in PGE2 levels were noted between the different stage cally signifi cant differences in stent patency were observed 0.609 <0.001 groups, however, plasma levels were higher in stage-4 pts. between the borderline and resectable groups (Figure 2). Complications included two patients with mild pancreati-

22 23 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

689 691 Table 1: Median Survival in Months Based on Treatment /690 How Do Adults and Adolescents Compare in Improvement Impact of Therapeutic Intervention on Survival in Treatment Type Median Overall Survival of Biochemical Cardiac Risk Factors S/P Roux-en-Y MiR-675-Induced Mesenchymal to Epithelial Transition Patients with Cholangiocarcinoma No treatment 1.5 in HCC is Associated with a Reduction in Twist1 Chemotherapy only 7.5 Gastric Bypass? Rebecca Wiatrek, Rebecca Nelson, Brian Mailey, Wendy Lee, Julio Whalen Clark1, Abul Elahi1, Jonathan M. Hernandez1, Radiation only 5.5 Dylan Gwaltney, Shushmita Ahmed, John M. Morton Garcia-Aguilar, Gagandeep Singh, Joseph Kim Chemoradiation 7.5 Jian Wang 1, Leigh Ann Humphries1, Bryan C. Fuchs2, Surgery, Stanford University, Stanford, CA City of Hope National Medical Center, Duarte, CA Surgery Only 23.5 2 1 Kenneth Tanabe , David Shibata Surgery and radiation 23.5 BACKGROUND: Roux-en-Y gastric bypass (RYGB) low- 1 BACKGROUND: Cholangiocarcinoma is a rare cancer Gastrointestinal Oncology, Moffi tt Cancer Center, Tampa, FL; Surgery and chemotherapy 17.5 ers cardiac risk through weight loss and improvement of with typically poor overall survival. Previous reports of 2Surgical Oncology, Massachusetts General Hospital, Boston, MA Surgery, radiation and chemotherapy 21 biochemical cardiac risk factors (BCRF). We hypothesize clinical outcomes are limited to single-institution or small Entire Cohort 4.5 improvement of BCRF for adolescents and adults alike. INTRODUCTION: Epithelial-to-mesenchymal transition cohort studies. Our objective was to examine surgical and (EMT) is a process by which cellular subpopulations in pri- METHODS: At a single academic institution, 2004 to 2010, medical outcomes for patients with cholangiocarcinoma Abstracts

mary epithelial tumors acquire a mesenchymal phenotype Monday we measured BCRF in both adults and adolescents undergo- using a population-based cancer registry. and the ability to metastasize. Subsequently, the metasta- ing RYGB. Markers included: hemoglobin A1C (hA1C) total METHODS: Patients with cholangiocarcinoma treated Table 2: Univariate and Multivariate Analysis of Factors Predictive ses may revert back to the original epithelial phenotype, cholesterol, triglycerides, high-density lipoprotein (HDL), in Los Angeles County from 1988 to 2006 were identifi ed of Overall Survival by mesenchymal to epithelial transition (MET). Certain low-density lipoprotein (LDL), triglyceride/HDL ratio, lipo- using the Los Angeles County Cancer Surveillance Program. microRNAs have been shown to induce EMT, although Univariate HR Univariate Multivariate HR Multivariate protein A (LpoA), high-sensitivity C-reactive protein (CRP), Patients were evaluated by standard clinical and patho- their role in MET remains poorly defi ned. AFP-secreting Factors (95% CI) p-value (95% CI) p-value and homocysteine and were measured preoperatively and logic factors, including race/ethnicity, extent of disease, hepatocellular carcinomas (HCC’s) are thought to exhibit Age 12 months postoperatively. The two groups were matched and treatment received (surgery, radiation therapy, chemo- a more aggressive behavior and also express high levels of <60 — — — — by body mass index (BMI), gender and number of compli- therapy, and none). Overall survival differences between the pro-proliferative miR-675. We sought to evaluate the 60–79 1.23 (1.05–1.43) 0.0086 1.18 (1.00–1.37) 0.0468 cations. Data were analyzed using Student t-tests, Pearson’s treatment arms were assessed by Kaplan-Meier method and effects of miR-675 manipulation on known EMT markers correlations and ANOVAs where appropriate. signifi cance determined using log-rank test. Cox propor- ≥80 1.82 (1.51–2.18) <0.0001 1.42 (1.16–1.74) 0.0007 and inducers in HCC cell lines. Race/ethnicity tional hazard modeling was used in multivariate analysis. RESULTS: For 99 adults: mean age at surgery was 44.4 Caucasian — — — — METHODS: MiR-675 was transfected into the low miR- years; mean BMI was 52.3% kg/m2; 73.7% were female; RESULTS: Of the 1040 patients with cholangiocarci- African American 1.38 (1.10–1.73) 0.0061 1.31 (1.04–1.65) 0.228 675-expressing HCC cell line HepG2 and inhibited by 68.7% were hypertensive; 47.5% were diabetic; 58.8% noma, 48% were male and 52% were female. The median Asian 0.90 (0.75–1.08) 0.2439 0.10 (0.83–1.20) 0.9714 anti-miR in the high-expressing Huh-1 cell line. Successful had hyperlipidemia; and 60% had obstructive sleep apnea age of patients was 69. The majority of the patients were Hispanic 0.98 (0.84–1.14) 0.7789 0.95 (0.81–1.12) 0.5201 transfections were confi rmed by measurement of miR-675 (OSA). The average number of comorbidities was 5.5. For white (47%), followed by Hispanics (26%), Asians (18%), Unknown 0.80 (0.30–2.13 0.6471 1.53 (0.56–4.15) 0.4057 levels using real-time RT-PCR. Western Blot and/or RT-PCR Extent of 33 adolescents: mean age at surgery was 17 years; mean and blacks (9%). Most patients had distant disease (27%), analyses were used to evaluate changes in the expression of Disease BMI was 52.7% kg/m2; 75.8% were female; 81.8% were whereas lower rates were observed for local (17%) and known EMT markers (E-cadherin, Vimentin) and inducers Local — — — — hypertensive; 12.1% were diabetic; 78.7% were non dia- regional (17%) disease. There was no record of disease stage (Twist1, Zeb1, Zeb2). Regional 1.19 (0.95–1.48) 0.1302 1.23 (0.98–1.54) 0.0701 betic but hyperinsulinemic; 51.5% had hyperlipidemia; in 400 patients, but only 23 of these patients underwent Distant 1.89 (1.55–2.31) <0.0001 1.74 (1.41–2.14) <0.0001 and 73% had OSA. Preoperatively, between adults and ado- RESULTS: MiR-675 transfection into HepG2 cells resulted surgical resection and 294 patients had no treatment. Cura- Unknown 0.96 (0.85–1.09) 0.0005 1.02 (0.84–1.24) 0.8397 lescents respectively, there was a signifi cant difference in in an upregulation of E-cadherin and a concomitant reduc- tive-intent surgical resection was performed in 126 patients Sex hA1C (6.36 vs 5.59, p < 0.001), triglyceride (152.9 vs 112.8, tion in Vimentin, fi ndings consistent with MET. Con- of the overall cohort. Of these, 80 patients had surgery Male — — — — p = 0.005), triglyceride/HDL ratio (4.05 vs 3.00, p = 0.018), versely, silencing of miR-675 in the Huh-1 overexpressing alone, whereas 46 patients had combination therapy (sur- Female 0.96 (0.85–1.09) 0.5686 0.86 (0.76–0.98) 0.0242 Surgery CRP (13.4 vs 8.9, p = 0.036) and homocysteine (10.5 vs 8.3, cells resulted in a mesenchymal (EMT) profi le with an gery + chemotherapy, n = 21; surgery + radiation, n = 4; and upregulation of Vimentin. Among the EMT inducers, Twist No — — — — p < 0.001). There was a trending difference in preopera- surgery + chemotherapy + radiation, n = 21). When com- Yes 0.33 (0.26–0.41) <0.0001 0.34 (0.27–0.43) <0.0001 tive total cholesterol levels between adults and adolescents 1 was downregulated in HepG2 miR-675 transfectants but paring outcomes for treatment, overall survival was highest Chemotherapy (182.9 vs 164.8, p = 0.007). At 12 months postop, there upregulated in the anti-miR treated Huh-1 cells. There was for the surgical arms (Table 1). On multivariate analysis, No — — — — was a signifi cant difference only in LpoA levels between no impact of miR-675 manipulation on Zeb1 and Zeb2 in younger age, female gender, limited extent of disease, sur- Yes 0.65 (0.57–0.75) <0.0001 0.64 (0.55–0.75) <0.0001 adults and adolescents (38.6 vs 11.7, p < 0.001). There was either cell line. gery, and chemotherapy were all associated with improved Radiation No — — — — a trending difference in hA1C (5.54 in adults vs 5.35 in CONCLUSION: Despite being associated with increased survival (Table 2). adolescents, p = 0.079) and triglyceride/HDL ratio (2.21 in CONCLUSIONS: Yes 0.68 (0.56–0.82) <0.0001 0.83 (0.67–1.01) 0.0623 proliferation, we have shown that miR-675 induces MET To our knowledge this is the largest HR = hazard ratio; CI = confi dence interval; (—) denotes reference group adults and 1.74 in adolescents, p = 0.074). Between preop with an attenuation of Twist1, suggesting that the pro- report of clinical outcomes in patients with cholangiocar- and 12 months postop, there were signifi cant differences cesses of tumor growth and metastasis may be regulated by cinoma. Our results suggest that surgical resection remains in the percent change of the following markers (compared different pathways. MiR-675 may regulate cellular epithe- the most important variable associated with improved sur- adults vs adolescents): HDL (19.3% in adults vs –4.8%, lial reprogramming that allows further growth following vival. The role of chemotherapy and radiation with surgery p = 0.009); homocysteine (–26.7% vs –3.6%, p=0.050); and the completion of the metastatic process. is not clear but warrants further examination in the multi- fasting insulin (8.48% vs –408%, p < 0.001). Between adults disciplinary management of this deadly cancer. and adolescents, there was a trending difference in the per- cent change in LDL (–17.8 vs –37.7, p = 0.075) and CRP (–1880% vs –882%, p = 0.084) from preop to 12 months post op. CONCLUSION: Our study demonstrates signifi cant differ- ences in both the preop levels and percent change (from preoperative to 12 month time points) of BCRF between adults and adolescents. Between preop and 12 months post op, adolescents were more likely to experience a greater improvement in HDL, homocysteine, and fasting insulin than adults of similar BMI and comorbidities.

24 25 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

692 histological origin of AC may determine survival. Specifi - 694 higher scores postoperatively at 3 weeks (72 vs 58*) and 2 cally, the pancreatobiliary subtype of AC displays worse months (81 vs 68*). Patients who developed complications p53 Gene Mutation Predicts Lymph Node Disease A Novel Measure of Recovery After Abdominal Surgery prognosis compared to the intestinal subtype. However, had lower scores preoperatively (64 vs 70†), at 3 weeks (60 Following Neoadjuvant Chemoradiation Therapy in 1,3 1 1,3 2 knowledge of inherent molecular characteristics of differ- Tung T. Tran , Pepa Kaneva, Gerald M. Fried, Nancy E. Mayo, vs 71*) and 2 months (66 vs 78*). The measure had moder- Rectal Cancer Patients 3,1 ent periampullary tumors and their effects on prognosis Liane S. Feldman ate correlations with all instruments (r = 0.3–0.8*). 1 Marjun Philip N. Duldulao, Zhenbin Chen, Wenyan Li, has been limited. Steinberg Bernstein Centre for Minimally Invasive Surgery, CONCLUSIONS: A novel measure of recovery after abdom- 2 Wendy Lee, Joseph Kim, Julio Garcia-Aguilar METHODS: Gene expression profi ling was used in order McGill University, Montreal, QC, Canada;Division of Clinical inal surgery was developed encompassing a broad range of General Oncologic Surgery, City of Hope, Duarte, CA to screen for differential gene expression between six PDAC Epidemiology, McGill University, Montreal, QC, Canada;3 Surgery, domains. Results are compatible with clinically observed BACKGROUND: The presence of lymph node (LN) metas- cases and 12 AC cases. Among others, hepatocyte nuclear McGill University, Montreal, QC, Canada postoperative recovery trajectories. This method illustrates tasis after neoadjuvant chemoradiation therapy (CRT), factor 4D (HNF4D) mRNA over-expression was observed in OBJECTIVE: Innovations in surgery are advocated on that recovery can be quantifi ed with mathematical units. which does not always correlate with primary tumor AC cases. Nuclear HNF4D protein expression was assessed the basis of “enhanced recovery”. Recovery is a complex response,is the main obstacle to the implementation of using tissue microarrays consisting of 99 individual AC process encompassing multiple domains including symp- 695 Abstracts CRT followed by local excision or even a watch and wait samples. Correlation of HNF4D expression with clinico- toms and emotional, social, economic and physical func- Monday pathological data and survival was calculated. Oncolytic Viruses Effectively Target and Kill Pancreatic approach for the treatment of rectal cancer. We investi- tion. Individual aspects of recovery are currently measured Cancer Stem Cells gated whether cancer gene alterations are associated with RESULTS: HNF4D mRNA is 7.61 fold up-regulated in AC using narrowly focused administrative indicators (length of LN disease status after CRT in rectal cancer patients. compared to that in PDAC. Bioinformatics analyses indi- stay) or patient-reported multiple-item questionnaires (e.g., Joyce Wong, Allison Schulman, Arjun Mittra, Yuman Fong Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY METHODS: Pre-treatment biopsies were obtained from cated its key role in dysregulated signaling pathways. quality of life). Questionnaire length and the diffi culty of 127 patients with Stage II/III rectal cancer enrolled in a Nuclear HNF4D expression correlates with histological sub- integrating these different view points limit accurate mea- INTRODUCTION: Cancer stem cells are highly resistant multi-center prospective Phase II clinical trial. All patients type, grading, CDX2 positivity, MUC1 negativity and pres- surement. The objective of our study was to integrate mul- to chemotherapy and radiotherapy, and these cells are received neoadjuvant CRT followed by total mesorectal ence of adenomatous components in the carcinoma. The tiple instruments used to estimate recovery into a single thought to be the basis of failure of palliative or adjuvant excision. Pathologic staging was determined according presence of HNF4D is a univariate predictor of survival in quantitative measure that would be sensitive to expected therapies for pancreatic cancers. Many characteristics of to AJCC criteria and patients were grouped according to AC mean survival 50 vs. 119 months, P = .002). Multivari- differences in recovery after abdominal surgery. cancer stem cells, including their antiapoptotic nature and pathologic LN status. Tumor DNA was extracted from pre- ate analysis revealed that HNF4D negativity (HR = 17.95, METHODS: The measure was developed and validated presence of multiple bypass pathways for nucleotide syn- treatment biopsies and analyzed by PCR and sequencing 95% CI: 2.35–136.93, P = .005) and lymph node positivity using data from patients undergoing scheduled abdominal thesis, make them attractive targets for replication-competent, for K-ras and p53 gene mutations. Fisher’s Exact Test was (HR = 3.33, 95% CI 1,36–8.18, P = .009) are independent surgery. Patients were interviewed preoperatively, 3 weeks oncolytic viruses. negative predictors of survival. performed to assess the association of gene mutation status and 2 months postop using 93 items from 5 indices cur- METHODS: To determine if oncolytic viruses can specifi - with LN metastasis after CRT. rently used to assess domains of recovery: health-related cally target and kill pancreatic cancer stem cells, a stem RESULTS: Overall, 31 (24%) patients had LN metastasis quality-of-life (SF-36), Quality of Recovery score (QoR), cell population from the MiaPaca-2 pancreatic adenocarci- after CRT. p53 and K-ras mutations were detected in 63 symptoms (visual analog scale), physical activity(CHAMPS noma cell line was isolated by FACS according to cell sur- (50%) and 47 (37%) patients, respectively; and 23 patients questionnaire) and general health perception (EQ-5D). face markers CD133 and CXCR4 and cultured in serum-free had both mutations. Patients with p53 mutations had an Rasch analysis combined items across domains of the vari- media. The stem cell spheroids were infected with NV1066, increased risk of LN metastasis compared to patients who ous instruments to develop the new measure on a logit a third-generation herpesvirus, or NDV-F3aa-GFP, a New- had wild-type p53 (33% vs. 13%; p = 0.015). However, the scale which was transformed to a score from 0–100 (SD). castle Disease virus mutant. Both viruses carried the marker detection of K-ras mutation alone or combined with p53 Longitudinal, known-groups convergent and convergent gene green fl uorescent protein (GFP), which allowed moni- mutation was not associated with LN metastasis after CRT. construct validity were assessed to demonstrate the psy- toring by fl uorescent microscopy. Cell cycle analysis and CONCLUSIONS: chometric properties of the new measure. *p < 0.01, †p < cell migration assay were also performed. Our results suggest that detection of 0.05. RESULTS: Viral infection of cancer stem cells was rapid mutant p53 may predict LN metastasis after CRT. There- RESULTS: fore, routine p53 genotyping may help guide appropriate Data from 177 patients, 67% male with mean (GFP expression was seen by 24 hours). The viruses from therapy in rectal cancer patients undergoing CRT. age 55 (15) were collected. 61% underwent ambulatory sur- both families each produced effi cient infection and killing gery while 39% were admitted. 20% developed at least one of cancer. At doses of multiplicity of infection (MOI, num- complication. A 24-item measure met all model require- 693 ber of viruses per tumor cell) of 0.5 or 1, >95% of cells were ments and included items from various domains. Reliabil- dead by day 6. Infection with virus also produced decreased Hepatocyte Nuclear Factor (HNF) 4D Expression ity was excellent (0.88). Mean score returned to baseline migratory capacity of the cancer stem cells and shifted the Distinguishes Ampullary Cancer Subtypes and from 69 (13) preop to 67 (14) at 3 weeks and increased population to a higher fraction in S phase. Prognosis After Resection above baseline to 76 (14) at 2 months*. Compared to CONCLUSION: Multiple types of oncolytic viruses effec- Robert Grützmann1, Florian Ehehalt1, Petra Ruemmele2, Stephan patients undergoing higher intensity procedures requiring admission, ambulatory patients had faster recovery, return- tively target the stem cell subpopulation of pancreatic Kersting 1, Christian Pilarsky1 ing to preop scores at 3 weeks (72 vs 70†) and increasing cancer cells. Infection decreases metastatic potential and 1 CONCLUSIONS: Surgery, University Hospital Dresden, Dresden, Germany; Immunohistochemical determination above baseline at 2 months (81 vs 70*). Scores for inpa- effects killing of such stem cells. These data support clinical 2 Department of Pathology, University Hospital Regensburg, of HNF4D expression is an effective tool for distinguishing tients were decreased at 3 weeks (58 vs 67†) but returned to studies of oncolytic viruses in the treatment of chemo- and Regensburg, Germany different AC subtypes. Likewise, HNF4D protein expression baseline at 2 months (69 vs 67). Ambulatory patients had radioresistant tumors. OBJECTIVE: is an independent predictor of favorable prognosis in carci- To investigate biological differences and noma of the papilla of Vater. prognostic indicators of different ampullary cancer (AC) subtypes. SUMMARY BACKGROUND DATA: AC is associated with a favorable prognosis compared to other periampul- lary carcinomas. Aside from other prognostic factors, the

26 27 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

10:00 AM – 11:15 AM 700 endoscopic surgery (NOTES) to develop techniques which will eliminate the potential complications of transabdomi- S501bcd NOTES Transoral Remnant Extraction (TORE) for nal organ extraction. Here we demonstrate the world’s fi rst VIDEO SESSION III: case of the unisex approach to transoral remnant extrac- EMERGING TECHNOLOGIES Sheetal Nijhawan, Saniea F. Majid, Toshio Katagiri, tion or TORE. Takayuki Dotai, Michael Sedrak, Bryan J. Sandler, Garth R. Jacobsen, Mark A. Talamini, Alan Wittgrove, Santiago Horgan University of California at San Diego, San Diego, CA 696 698 Laparoscopic sleeve gastrectomy has rapidly become the Minimizing MIS Using Magnetically Anchored and Use of a Magnetically Coupled Camera and Novel Mini procedure of choice for morbid obesity. We have utilized our considerable experience in natural orifi ce translumenal

Percutaneous Needlescopic Instruments for Basic Laparoscopic Instruments to Perform Minimally Inva- Abstracts and Complex Procedures sive Sigmoid Colon Resection Monday

1 2 2 1 Nabeel Arain, Sara Best, Jeffrey A. Cadeddu, Deborah C. Hogg, Calvin D. Lyons, Rohan A. Joseph, Nilson Salas, 2:15 PM – 4:00 PM Richard Bergs 3, Raul Fernandez3, Lauren B. Mashaud1, Patrick R. Reardon, Barbara L. Bass, Brian J. Dunkin Daniel J. Scott 1 Department of Surgery, The Methodist Hospital, Houston, TX S501a 1 Department of Surgery, University of Texas Southwestern Medical Natural orifi ce translumenal endoscopic surgery (NOTES) PLENARY SESSION V Center, Dallas, TX;2 Department of Urology, University of Texas and single site laparoscopy (SSL) have led surgeons to seek Southwestern Medical Center, Dallas, TX;3 Texas Manufacturing less invasive methods of performing even standard opera- Assistance Center/Automation and Robotics Research Institute, tions by reducing port number and size. This video dem- University of Texas Arlington, Arlington, TX onstrates how novel technology can be used to perform 800 RESULTS: A total of 386 benign adenoma pts underwent reduced port surgery, even during complex operations, This video shows techniques for minimizing MIS using OCR during the period assessed. The distribution of large without compromising visualization, triangulation, or A Standard “Oncologic” Segmental Colorectal Resection magnetically anchored and percutaneous needlescopic bowel polyps was: right, 263 (68.1%); transverse, 33 (8.6%); instrument function. Use of MAGS in combination with Is Indicated for Dysplastic Adenomas That Come instruments for both basic and complex procedures which sigmoid, 38 (9.8%); , 23 (6.0%), and multiple sites, percutaneous internally assembled instruments enables to Surgery: ESD and EMR Are Best Avoided in These included a laparoscopic cholecystectomy (porcine model) 13 (3.4%). The preop pathologic diagnosis was adenoma reduced port surgery while maintaining the visualization Patients and a laparoscopic Roux-en-Y Gastric Bypass (human alone for 288 (74.6%) and adenoma with dysplasia for and function of 4 port laparoscopic surgery. The ergonomic 1 3 1 1 cadaver). We combined instrumentation including a mag- Joon H. Jang, Emre Balik, Michael J. Grieco, Tromp Wouter, 98 pts (25.4). Final post resection pathology revealed 62 benefi t of this technology will aid in the performance both netically anchored camera, 3mm percutaneous instru- Daniel D. Kirchoff 1, Anjali S. Kumar2, Daniel L. Feingold2, invasive cancers (16.1% of total). Thirty fi ve percent of pts reduced port surgery and SSL. ments, and a single 12 mm working port. This combination Richard L. Whelan 1 (34/98) with a preop diagnosis of dysplasia had an invasive afforded excellent triangulation and minimized the num- 1Surgery, St. Luke Roosevelt Hospital, New York, NY;2Surgery, cancer whereas 9.7% (28/288) of the adenoma alone pts 699 proved to have a malignancy (p < 0.0001). The mean lymph ber and size of incisions. Furthermore, this strategy may Washington Hospital Center, Washington DC, DC; 3Surgery, node harvest for the entire group was 16.0 ± 10.2; there was offer comparable or even better cosmesis compared to sin- Laparoscopic Rectosigmoid Resection with Transvaginal Columbia University, New York, NY gle incision laparoscopic procedures without the associated Rectopexy and Extraction for Rectal Prolapse no difference in lymph node recovery for the cancer and INTRODUCTION: Endoscopic submucsal dissection benign adenoma groups. The cancer stage breakdown for technical diffi culties. 1 2 2 Patricia Sylla, Samantha J. Pulliam, May Wakamatsu (ESD) and endoscopic mucosal resection (EMR) methods the 62 pts with a malignant polyp was: Stage 1, 73%; Stage 1Surgery, Massachusetts General Hospital, Cambridge, MA; 697 are now being used for benign colorectal polyps judged 2, 8.1%; Stage 3, 16%; Stage 4, 3.2%. The mean polyp size 2Obstetrics and Gynecology, Massachusetts General Hospital, not removable using standard colonoscopic methods. Also, for benign lesions was 3.0 ± 1.9cm whereas for cancers it Endoscopic Treatment of Weight Regain in the Boston, MA laparoscopic partial circumference “wedge” resections of was 3.9 ± 2.4 cm (p = 0.0008). There was a higher incidence Post-Bypass Patient The patient is a 39-year-old G4P4 female with full-thick- the colon wall (+ polyp) are being done in an effort to of Stage 3 cancers in pts with a preop diagnosis of dysplasia Rabindra R. Watson1, David B. Lautz2, Christopher C. Thompson1 ness rectal prolapse, constipation, symptomatic cystocele, avoid a standard “Oncologic” Colorectal Resection (OCR) (p = 0.008). 1Division of Gastroenterology, Brigham & Women’s Hospital, rectocele, and stress urinary incontinence. Laparoscopic and its attendant morbidity. Unfortunately, a subset of CONCLUSION: Ten percent of pts with the preop diagno- these “benign” polyps contain invasive adenocarcinomas; Boston, MA; 2Surgery, Brigham and Women’s Hospital, rectosigmoid mobilization was performed using 4 abdomi- sis of adenoma alone and over 1/3 of pts with the preop in these patients (pts) an OCR is indicated. This retropsec- Boston, MA nal trocars and a 12 mm port was inserted through a poste- diagnosis of dysplastic polyp had invasive cancers. Larger rior colpotomy. The rectum was transected transvaginally tive review of benign polyp pts that underwent OCR’s was polyps were more likely to contain a cancer. A standard Weight regain affects a signifi cant proportion of post- with a stapler and suture rectopexy was performed through undertaken with the hope of identifying polyp characteris- segmental is advisable for pts with dysplastic bypass patients. Pouch and stoma dilation have been the vaginal trocar. The rectosigmoid was exteriorized trans- tics that would allow stratifi cation of these lesions into low polyps because of the high likelihood of cancer. ESD, EMR, implicated as important factors in this process. The signifi - vaginally, the anvil was placed, and intracorporeal stapled and high risk categories which might then guide treatment and wedge resection should be reserved for selected ade- cant risk of morbidity with surgical revision makes endo- anastomosis was completed. Anterior and posterior colpor- choices. noma pts without dysplasia after thorough evaluation. scopic therapy an attractive treatment modality. This video rhaphy with tension-free vaginal sling were performed per- METHODS: All patients with the preoperative (preop) diag- reviews current endoscopic techniques including sclero- ineally. The patient was discharged on postoperative day 4. nosis of adenoma (± dysplasia) who underwent a standard therapy, tissue anchor placement, and endoscopic sutur- colorectal resection over an 18 year period were included ing. Current evidence regarding these techniques is sparse, in the study. Hospital and offi ce records, operative reports, though compelling data are forthcoming. Given the grow- and pathology reports were reviewed. The ing bariatric population, endoscopic treatment of surgical polyp size and location for each patient was noted; likewise failure will play an important role in the management of the preop and fi nal pathologic diagnoses were recorded these patients. and compared. The unpaired t-test and Fisher’s Exact test were used to analyze the results and a p-value less than 0.05 considered signifi cant.

28 29 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

patient in supine position on positive pressure ventilation. Over the previous 5 years, the percentage of older patients š801 With a GIF-180 (Olympus, Tokyo, Japan) positioned at 10 /804 undergoing bariatric surgery has increased from 1.92% in Single Port Transanal Surgery of a Giant Rectal cm above the GEJ, a mucosotomy was performed at the Bariatric Surgery Outcomes in the Elderly Population:2005 to 4.77% in 2009 (p < 0.001). There were only 72 Adenoma 2 O’clock position, and a submucosal space was developed An ACS NSQIP Study deaths throughout the entire study period, 8 of which were

1 2 caudally creating a controlled submucosal tunnel extend- in the ≥65 years cohort. The incidence of 30-day mortality Renée M. Barendse, Pascal G. Doornebosch, ing 2 cm distal to the GEJ. Upon completion of this tun- Robert B. Dorman, Anasooya Abraham, Waddah B. Al-Refaie, in those 35–49, 50–64 and ≥ 65 years was 0.12%, 0.21% 3 1 1 Willem A. Bemelman , Evelien Dekker, Paul Fockens, nel the gastro-esophageal lumen was inspected for mucosal Helen M. Parsons, Sayeed Ikramuddin, Elizabeth Habermann and 0.40%, respectively. Adjusting for confounders, multi- Thomas M. Van Gulik 3, Eelco J. de Graaf2 integrity. The scope was then reinserted into the submu- Surgery, University of Minnesota, Minneapolis, MN variate regression analysis demonstrated advancing age to 1 Gastroenterology and Hepatology, Academic Medical Center, cosal tunnel and using a triangle-tip knife, myotomy was INTRODUCTION: Bariatric surgery has been shown to be trend towards becoming a predictor of mortality but it was Amsterdam, Netherlands;2 Surgery, IJsselland Hospital, Capelle performed starting at 5 cm above the GEJ and ending at 2 benefi cial in achieving weight loss and in decreasing long- not signifi cant (Table 1). Age ≥65 was a signifi cant predic- aan den IJssel, Netherlands; 3Surgery, Academic Medical Center, cm bellow the GEJ. During this process the circular muscle term mortality. To date, however, evidence has suggested an tor of PLOS for both open and laparoscopic procedures. For Amsterdam, Netherlands layer of the esophagus was carefully divided with preserva- increased mortality following bariatric surgery in patients those who underwent laparoscopic procedures, odds ratios

tion of the longitudinal layer. At the end of the procedure ≥65 years. With adjustment of confounding variables, we were similar for PLOS for the 50–64 years and ≥65 years’ Abstracts Single port surgery of rectal tumours may be associated with the mucosal incision was closed longitudinally with endo- cohorts. While major adverse events were not predicted by Monday shorter learning curve and fewer costs than transanal endo- hypothesized that the short-term operative outcome pro- scopic clips and surgical glue. fi le in those ≥65 years undergoing bariatric surgery would age ≥65 for either open or laparoscopic procedures, BMI scopic microsurgery (TEM). A semicircumferent adenoma 2 RESULTS: Four patients underwent TEEM, with no peri- be comparable to that of younger persons. ≥55 kg/m , severe ASA score, cardiac co-morbidities, albu- located 6–12 cm from the anus was successfully resected min <3 and creatinine >1.5 were all predictors of major operative complication. All patients reported signifi cant METHODS: 2 using a single-site access system transanally. Endoluminal Patients with BMI ≥35 kg/m who underwent adverse events (not shown). features were comparable to TEM. Full thickness, en bloc improvement of their dysphagia immediately after the open and laparoscopic Roux-en-Y gastric bypass, open duo- resection was possible. The defect was sutured transversally procedure. On the fi rst post-op day, all barium swallows denal switch, laparoscopic adjustable gastric banding and CONCLUSIONS: This large, multi-hospital study demon- with continuous self-anchoring sutures. Operating time showed disappearance of the classical bird beak taper, rapid vertical banded gastroplasty in the 2005–2009 American strates older age predicts short-term PLOS but not major was 55 min. Blood loss was nil. The patient was hospital- emptying of contrast into the stomach and absence of College of Surgeons National Surgical Quality Improve- events. The overall low death rates likely explain the ized overnight. Histopathology showed tubulovillous ade- leaks. All patients were discharged on the second post-op ment Program were identifi ed. Controlling for confound- observed trend toward signifi cance in operative mortality noma with focal high grade dysplasia, radically resected. day on liquid diet. Two patients reported transient heart- ers and stratifying by open versus laparoscopic surgery, in older patients. Thus surgeons should continue to pro- Single port transanal surgery may become an alternative burn, which was well controlled with medications. The multivariate regression was used to predict the impact of mote caution when considering whether to operate on to TEM. average pre-op GERD-HRQL was 20, which improved to age (≥65 years) on mortality, major events and prolonged this patient population. Once corroborated, these results 11.3 at 7 days post-op and 2 at 30 days post-op. To date, length of stay (PLOS, >90th percentile) at 30 days. provide important information to patients, surgeons, hos- 802 two patients have already returned for their 6-month fol- RESULTS: pitals, and payers prior to performing bariatric surgery in low-up, reporting adequate swallowing and low LES pres- We identifi ed 48,378 patients who underwent older persons with obesity. Transesophageal Endoscopic Myotomy (TEEM) for sures on esophageal manometry (their mean pre-op LES the above bariatric procedures between 2005 and 2009. the Treatment of Achalasia: The United States Humanresting pressure was 33 mmHg and residual pressure was Experience 41.35 mmHg, whereas the 6-months follow-up mean LES Ozanan R. Meireles1,2, Garth R. Jacobsen1, Toshio Katagiri1, resting pressure was 10.85 mmHg and residual pressure was Kari Thompson 1, Abraham Mathew4,1, Noam Belkind1, –0.8 mmHg). Table 1: Predictors of 30-day Mortality and Prolonged Length of Stay 1 1 1 CONCLUSION: Michael Sedrak , Bryan J. Sandler, Takayuki Dotai, TEEM seems to be safe, feasible and effec- 30 Day Mortality OR (95% CI) Prolonged Length of Stay OR (95% CI) Thomas J. Savides 3, Saniea F. Majid1, Sheetal Nijhawan1, tive for the treatment of patients with achalasia. Long-term Factor Open N = 5017 Laparoscopic N = 43361 Open N = 4992 Laparoscopic N = 43314 Mark A. Talamini 1, Santiago Horgan1 data is still necessary for wide-spread utilization of this Age 35–49 yrs 50–64 yrs • Ref. 1.2 (0.4, 3.1) 3.3 (0.8, Ref. 1.2 (0.6, 2.4) 1.5 (0.5, 4.6) Ref. 1.2 (1.0, 1.5) 2.1 (1.5, Ref. 1.2 (1.2, 1.3) 1.2 (1.1, 1.4) 1 novel technique. 65 yrs 14.3) 2.9) Department of Surgery, University of California San Diego, BMI (kg/m San Diego, CA;2 Department of Surgery, Massachusetts General 2) 45–49 50–54 Ref. 1.8 (0.4, 8.1) 2.6 (0.6, Ref. 2.4 (0.9, 6.0) 2.3 (0.8, 6.7) Ref. 1.2 (0.9–1.5) 1.4 (1.1, Ref. 1.1 (1.0–1.2) 1.35 (1.2–1.5) 1.7 (1.6, Hospital—Harvard Medical School, Boston, MA;3Department š803 55-59 • 60 11.7) 5.3 (1.4, 20.1) 4.0 (1.5, 11) 1.8) 1.5 (1.2, 1.9) 1.9) C-index 0.78 0.73 0.65 0.63 of Gastroenterology, University of California San Diego, San Single Incision Laparoscopic and Dor Percentage of Events, % (N) 0.50 (25) 0.11 (47) 17.8 (888) 20.1 (8694) Diego, CA; 4Department of Gastroenterology, Penn State Milton S. Fundoplication Multivariate regression analysis, controlling for confounders, was used to formulate the above odds ratios. Confounders adjusted for include, but are not limited to, age, Hershey Medical Center, Hershey, PA John Afthinos, Koji Park, James McGinty, Ninan BMI, sex, race, ASA score, diabetes, pulmonary and cardiac comorbidities, albumin, creatinine, and admission year. From our early experience with NOTES, our group has Koshy, Julio Teixeira acquired familiarity with transesophageal submucosal dis- General Surgery, St. Luke’s Roosevelt Hospital Center, New York, NY section and myotomy in swine model, which allowed us to INTRODUCTION: We describe our technique for single perfect a model to perform purely endoscopic trans-esoph- incision laparoscopic (SILS) Heller myotomy. The mini- ageal myotomy (TEEM) for the treatment of achalasia and mally invasive nature of a SILS approach may appeal to apply it into clinical practice. This study was designed to patients considering conservative, non-surgical therapy for assess the safety, feasibility and effi cacy of TEEM in a series achalasia, like endoscopic dilatation or botulinum toxin of patients with achalasia. injection. METHODS: Under IRB approval, patients were enrolled TECHNIQUE: A Heller myotomy and partial anterior fun- on our study, where TEEM was offered as an alternative doplication are performed through an umbilical SILS port. to laparoscopic or robotic Heller myotomy. The inclusion A penrose drain facilitates retraction. Intraoperative endos- criteria were patients with achalasia confi rmed by esoph- copy confi rms adequacy of the myotomy. ageal manometry, age between 18 and 50 years old, and CONCLUSION: ASA class 2 or lower. The exclusion criteria were pregnancy, Heller myotomy and Dor fundoplication prior esophageal surgery, immunosuppression, coagu- is technically feasible through a SILS approach, and offers a lopathies, and severe medical co-morbidities. The proce- competitive, minimally invasive alternative to non-opera- dures were performed under general anesthesia, with the tive treatment options.

30 31 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

805 806 4:00 PM – 5:00 PM Counter-Clockwise Rotation of Roux-en-Y Limb Factors Predictive of Recurrence and Mortality After S501a Signifi cantly Reduces Internal Herniation in Defi nitive Surgical Repair of Enterocutaneous Fistula QUICK SHOTS SESSION II Laparoscopic Roux-en-Y Gastric Bypass (LRYGB) Jose L. Martinez, Enrique Luque-de-León Kalyana Nandipati, Edward Lin, Farah A. Husain, Gastrocirugía, UMAE Hospital Especialidades Centro Médico Jahnavi Srinivasan, John F. Sweeney, S.S. Davis Nacional SXXI, Mexico DF, Mexico General Surgery, Emory University, Decatur, GA BACKGROUND: Most enterocutaneous fi stulae (ECF) unrestricted ability to replicate. By expressing IFND locally INTRODUCTION: Internal hernias continue to be a require operative treatment. Although recent advances /807 in the tumor environment, toxicity from systemic delivery signifi cant source of morbidity after LRYGB. Literature have widened therapeutic options, recurrence after surgi- Assessment of In Vivo Functionality of a Novel seen in current therapy should be avoided. addressing the technique of Roux limb construction as a cal repair has not changed substantially. Assessment of out- Cancer-Targeting Adenovirus Expressing Interferon predisposing factor is sparse. The objective of this study comes specifi cally regarding recurrence and mortality after Alpha in an Immunocompetent Model Abstracts is to evaluate the impact of Roux limb construction tech- surgical repair has not been studied extensively. Monday Leonard Armstrong, Julia Davydova, Eric J. Brown, nique on the development of internal hernias. AIM: To determine factors associated with recurrence and Selwyn M. Vickers, Masato Yamamoto METHODS: In this study we included 444 consecutive mortality in patients submitted to surgical repair of ECF. University of Minnesota, Minneapolis, MN patients (367 [82.7%] females and 77 [17.3%] males) from MATERIAL AND METHODS: We analyzed prospectively bariatric database who underwent LRYGB at our Institu- collected databases on all consecutive patients submitted More effective systemic therapy is clearly needed for pan- tion. Variables collected include demographics, body mass to surgical repair of ECF during a 5 year period. Several creatic adenocarcinoma. Interferon alpha (IFND) is promis- index (BMI) before and after the procedure, and postopera- patient, disease and operative variables were assessed as fac- ing in multimodality therapy, but has a short half-life and tive small bowel obstruction secondary to internal hernia- tors related to recurrence and mortality. Univariate statisti- strong side effects. We hypothesize that by expressing IFND tion. Technical details collected include; type of roux-en-y cal (UA) comparisons were made using Students T Test for locally in the tumor environment using a replicating tumor- limb construction, Peterson’s defect closure at initial opera- continuous variables and Fischers exact test for categorical specifi c adenovirus, systemic side effects can be greatly tion and reoperative fi ndings. Roux limbs were constructed variables. Multivariate analyses (MA) were also performed. minimized while allowing for focused tumor delivery and a powerful treatment effect. We generated a novel adenovirus in 291 patients by a clockwise rotation of the bowel and RESULTS: jejuno- performed on the right side of the axis A total of 71 patients were included. Median age using several strategies to enhance effi cacy while reducing of the mesentery (Group 1). In 153 patients the Roux limb was 52 y. (range, 17–81). Operative indications included toxicity. The COX2 promoter is used to drive viral replica- was constructed by a counterclockwise rotation of the Roux ECF persistence (38), sepsis (17), eversion of mucosa (12), tion due to its known overexpression in pancreatic tumor limb resulting in the jejuno-jejunostomy on the left side and others (4). Surgical treatment included resection and cells, while simultaneously avoiding replication toxicity to of the axis of the mesentery (Group 2). We also analyzed anastomosis (37), resection and ostomy (21), oversew (4), normal organs such as liver. Our design uses capsid fi ber 808 the impact of Peterson’s space closure on internal hernias. bypass (3), and catheter placement (2). ECF recurred in modifi cation for optimal infectivity, and adenoviral death Difference in Clinical Characteristics and Survival Fisher’s exact test and Chi-square test used for the analysis. 22 patients (31%). Medical treatment was established in 9 protein for enhanced apoptosis and viral spread. Most Outcomes in Asian and American Patients with (with ECF closure in 7); surgical repair was re-attempted RESULTS: Of a total 444 [mean age – 43.7 + 10.3 years, importantly, localized IFND expression will allow focused Hepatocellular Carcinoma in 13 others (attained in 11). Thus, management of 22 tumor treatment and limited systemic toxicity. Previously, mean BMI pre op was 46.4 + 5.1, BMI after median fol- 1,2 1 2 2 patients with recurrent ECF was successful in 18 (82%). UA we have described a virus targeting human pancreatic can- Taejin Song , Yuman Fong, Sung-Ock Suh, Sang-Yong Choi, low up of 12 months was 34.5 + 6.98] patients included 1 1 1 disclosed non-colonic ECF origin (p = 0.04) and high out- cer with such modifi cations, but available in vivo modelsWilliam R. Jarnagin , Mithat Gonen, David S. Klimstra in the study, 21 (4.7%) internal hernias were identifi ed. put (p = 0.001) as risk factors for recurrence. This latter was 1Surgery, MSKCC, New York, NY;2Surgery, Korea University Of 21 internal hernias, 17 (81%) were through Peterson’s do not allow an assessment of immune response. For this the only one that prevailed after MA (p = 0.01). Although reason we have designed a structurally analogous viral vec- College of Medicine, Seoul, Republic of Korea space and 4 (19%) were through mesenteric defect. Group not statistically signifi cant, management with an open 1 patients had signifi cantly higher overall internal hernias tor for use in immunocompetent hamsters, a system which INTRODUCTION: Hepatocellular carcinoma (HCC) is one abdomen (p = 0.06) and enteroatmosferic fi stulae (p = 0.07) allows human adenoviral replication and an assessment of (20/291, 6.9% vs 1/153, 0.7%; P = 0.0018) and Peterson’s had a tendency to favor recurrence. A total of 14 patients of the most common malignancies seen in Far East Asia hernias (16/275, 5.8% vs 1/152, 0.6%; P = 0.0089) com- the contribution of IFND on antitumor effect. Qualitative but with a relatively lower incidence in the United States. died (20%). UA revealed several risk factors for mortality assessment of effect was done with crystal violet staining at pared to Group 2. measured at diagnosis or referral including malnutrition Because of the variety of tumor characteristics, factors asso- serial time points. Quantitative cell viability was assessed ciated with prognosis and survival have been diffi cult to In addition, no signifi cant difference was noted in the inci- (p = 0.03), sepsis (p = 0.01), hydroelectrolytic imbalance colorimetrically using a commercially-available kit. COX2 determine. dence of Peterson’s hernia whether the defect was closed (p = 0.001), and serum albumin <3 g/dl (p = 0.02). Other promoter function was assessed in cell lines by a luciferase METHODS: or not closed (closed group – 4/117, 3.4% vs. not closed signifi cant variables were interval from diagnosis to opera- gene reporter assay with known COX2-positive and nega- 71 patients with HCC resected at an America – 13/327, 4%; P = 1.00). Within Peterson’s defect closed tion ≤20 weeks (p = 0.03), preop. serum albumin <3 g/dl tive controls. IFND expression levels were quantitiated by institution were compared with 51 patients resected in Asia group, clockwise rotation and anastomosis to the right (p = 0.001), and age ≥ 55 years (p = 0.03). Only the latter ELISA from cell culture supernatant. In vivo, hamsters were using clinical data analyses. These results and demographic side of axis of mesentery was associated with signifi cantly two remained signifi cant after MA. A slight tendency was injected on each fl ank with 4 × 106 HP1 hamster pancreatic data were then correlated with results from immunohisto- higher incidence of Peterson’s hernias compared to counter observed for female gender (p = 0.07) and non-colonic ECF tumor cells, then intratumorally with our novel virus or chemical staining of microarray sections for signifi cance of clockwise rotation (4/54 vs 0/63, P = 0.043). In not closed origin (p=0.09). Interestingly recurrence after surgical treat- controls lacking various of its design features at 3 × 1010 prognostic and survival parameters. group, clockwise rotation was associated with higher inci- ment was not associated with mortality (p = 0.75). viral particles/50μl or with PBS when tumors were 10–15 RESULTS: When comparing the Asian and American, dence of internal hernias that did not reach statistical sig- CONCLUSIONS: Among several studied variables, recur- mm in size. In vitro a robust effect on hamster pancreatic signifi cant differences were found in expression of p53 nifi cance (12/237, 5.1% vs 1/90, 1.1%; P = 0.12). rence was only independently associated with high output. cancer cell lines is seen. Analysis of COX2 expression in and MDM2. Higher p53-positive staining was seen in the SUMMARY: This study demonstrates that the technique Interestingly, once ECF recurred its management was as suc- pancreatic cancer cell lines shows that hamster cell lines Asian group (P = 0.04), while MDM2-positive staining was for construction of the Roux limb probably is a major fac- cessful as non-recurrent fi stulas in our series (closure rate under investigation are COX2-positive. In vivo experi- higher in the American group (P = 0.0003) (in press). Sur- tor in the development of internal hernias. Construction of 82%). Mortality was associated to previously reported ments show a strong suppression of tumor growth from the vival rates of patients according to p53-positive or -nega- of the Roux limb with a counter-clockwise rotation of the bad prognostic factors at diagnosis or referral. Timing of new virus (Figure 1). In summary, a new cancer-targeting tive status indicated that 42 of 85 (49%) p53-negative bowel, such that both jejuno-jejunostomy anastomosis and operation (>20 weeks) seems relevant in order to optimize oncolytic adenovirus was successfully designed to selec- patients were alive, with median survival of 49.9 months, Ligament of Treitz are to the left of the axis of the mesen- patients functional and nutritional status. tively replicate in cancer cells with a powerful antitumor while 8 of 17 (47%) p53-positive patients were alive, with tery signifi cantly reduces the incidence of internal hernias. effect. There is no loss of potency compared to viruses with median survival of 38.5 months (p = 0.28). Survival rates of

32 33 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

patients with positive or negative MDM2 staining showed type (WT) mice were sacrifi ced seven days after tamoxifen square. Individuals with depression were identifi ed from METHODS: Following IRB approval, the 2005–2008 48 of 89 (54%) of MDM2-negative patients were alive, administration. The was harvested and the the Severity File and compared to non-depressed patients National Surgical Quality Improvement Program’s Par- with median survival of 53.0 months, while 3 of 14 (21%) intestinal crypt cells were isolated and fractionated. Cyto- using multiple linear regression accounting for cluster- ticipant User File was used to identify patients undergo- MDM2-positive patients were alive, with median survival chrome c was quantifi ed by Western blotting in cytoplasmic ing by hospital and adjusting for clinical and sociodemo- ing LAPR and OAPR for the treatment of rectal cancer. of 17.7 months (p = 0.23). For MDM2, two long-term sur- fractions. Activated Bax was immunoprecipitated using an graphic confounders. Logistic regression was used to determine which factors vivors unduly infl uenced the results. This is unlikely to be activated Bax isoform-specifi c antibody (6A7) followed by RESULTS: The 2008 NIS cohort included 292,191 patients infl uenced selection of an open vs. laparoscopic surgery. resolved by longer statistical follow-up. HepPar1 diagnostic Western blot analysis. To investigate Bax activation during who underwent colon or rectal surgery. A preoperative diag- Chi-square and multivariable analysis was used to com- stain showed that, for negatively stained patients, 11 of 29 adaptation, mice underwent either proximal SBR or sham nosis of depression was present in 20,039 (6.9%) patients. pare the incidence of 30-day postoperative complications (38%) were alive, with a median survival of 38.5 months, operation. Mice were sacrifi ced 3 days after the operation Individuals with depression were more likely to be younger between the two surgeries and to identify factors predictive while for positively stained patients, 39 of 73 (53%) were and activated Bax was quantifi ed as described above. (mean age 63.2 vs 62.1, p < 0.001), female (70% vs 52%, of complications. alive, with median survival of 54.3 months (P = 0.46). Sur- RESULTS: Rates of apoptosis were signifi cantly lower in p < 0.001), white (87% vs 80%, p < 0.001), and to have had RESULTS: A total of 1197 OAPR’s and 143 LAPR’s were vival as indicated by prognostic value of HepPar1 stain the p38D-defi cient mice compared to their WT littermates preoperative weight loss (11% vs 8%, p < 0.001) compared identifi ed. Compared to OAPR, patients undergoing LAPR

presented with signifi cant differences. In HepPar1 prog- Abstracts

both at baseline and after SBR (baseline: 4.9 ± 0.6 vs 8.9 ± to patients without a diagnosis of depression. Mean length were more likely to be female (p = 0.02), non-Caucasian Monday nostic stain-negative patients, 12 of 43 (28%) were alive, 1.3 apoptotic bodies/100 crypts, p < .05; after SBR 6.9 ± 0.6 of stay for those with depression (10.4 days, 95% CI 10.04- (p = 0.02) and were less likely to have a body mass index with median survival of 29.5 months, but for patients with vs. 10.9 ± 0.8 apoptotic bodies/100 crypts, p < .05). Con- 10.76) was signifi cantly longer than for patients without (BMI) of ≥ 30 (p = 0.04). LAPR’s were associated with longer positive stain, 38 of 59 (64%) were alive, with median sur- sistent with lower apoptotic rates, there were signifi cantly depression (9.64 days, 95% CI 9.48–9.81). After adjusting operative times (p<0.001), with 57% requiring 4–8 hours. vival of 65.4 months (P = 0.01). Multivariate analysis of lower levels of cytoplasmic cytochrome c in the p38D- for comorbidities using the Deyo Index, sociodemographic No difference in 30-day postoperative complication rates variables, which showed signifi cant differences, served as defi cient mice compared to WT littermates (57% reduction, factors, age, insurance status, race, and surgical and hospi- between LAPR’s and OAPR’s was found with the exception predictors of mortality; out of these variables, only vascular p < .05). In addition, the amount of activated Bax present tal variables, depression still predicted an increase in length of a higher odds of sepsis with LAPR (OR 3.12; p = 0.04). invasion was a signifi cant predictor of mortality (p = 0.05, in intestinal crypts was reduced by 67% in mice after p38D of stay. Additionally, depressed patients were less likely to LAPR and OAPR were found to have similar rates of sur- hazard ratio = 2.2). MAPK deletion (p < .05). Finally, activated Bax levels were resume normal function at the end of their prolonged stay, gical site infections (p = 0.13), transfusion requirements CONCLUSION: These data suggest differences in survival two-fold higher in mice after SBR (n = 3) relative to sham as 40% required either home health or time in a skilled (p = 0.17), myocardial infarction (p = 0.48), and need for according to the molecular pathogenesis of HCC based on levels (p = .08) facility following discharge from the acute care hospital. re-operation within 30 days (p = 0.20). There was no differ- racial and regional differences. The results to p53, MDM2, CONCLUSIONS: Deleting p38D MAPK in an inducible, CONCLUSIONS: Among patients undergoing colorectal ence in the number of postoperative complications between and HepPar1 showed a direct correlation between positiv- intestine-specifi c manner leads to reduced levels of entero- surgery, depression is associated with a signifi cantly pro- either group (Table 1). Based on multivariable analysis, the ity of these proteins and a poorer prognosis for both racial cyte apoptosis, attenuated cytoplasmic cytochrome c, and longed hospital stay. At the same time, patients with depres- only factor predictive of postoperative complications with groups. This could explain disparities in clinical outcomes diminished levels of the activated isoform of Bax. These sion are more likely to require assistance after discharge in LAPR was neoadjuvant radiotherapy (p = 0.04). Smoking seen for different ethnic cohorts. It also indicates that adju- results support our hypothesis that p38D MAPK regulates the form of home health services or time in skilled rehabili- (p < 0.001) and >10% weight loss prior to surgery (p = 0.01) vant and palliative systemic therapies for general use can- Bax activation in small intestine enterocytes during resec- tation facilities. The need for assistance after discharge adds predicted complications with OAPR. The odds of undergo- not be directly extrapolated from data derived from one tion-induced enterocyte apoptosis. to an already protracted recovery and can signifi cantly ing an OAPR was higher with Caucasian race (OR = 1.54; ethnic group. increase cost of care. Further research into the mechanism p = 0.02) and BMI ≥25 (OR 1.48; p = 0.03). The single factor associated with higher odds of an LAPR was the absence of 810 underlying these differences and potential treatment strat- /809 egies among depressed patients is warranted. neoadjuvant radiotherapy (OR = 0.37; p = 0.04). Depression Is Associated with Prolonged and Intestinal Resection-Induced Enterocyte Apoptosis Is Complicated Recovery Following Colorectal Surgery Number of Postoperative Complications Between Laparoscopic Mediated by p38D MAPK-Directed Activation of Bax 811 and Open Abdominoperineal Resections Courtney J. Balentine1,2, Aanand D. Naik1,2, Celia N. Robinson2,3, Derek Wakeman1, Jun Guo1, Jethrina A. Santos2, John Schneider3, Laparoscopic Versus Open Abdominoperineal Resection Jesus H. Hermosillo-Rodriguez 4, David H. Berger1,2 Jennifer A. Leinicke 1, Kathryn J. Rowland1, Christopher R. for Rectal Cancer: Is There a Short-Term Advantage Number of Postoperative Open APR Laparoscopic APR 1Houston VA HSR&D Center of Excellence, Houston, TX;2Michael Complications (n = 1197) p = 0.18 (n = 143) p = 0.18 Erwin 1, Brad W. Warner1 in Complication Rates with a Minimally Invasive E. DeBakey VA Hospital, Houston, TX; 3Michael E. DeBakey 0 789 (65.9%) 103 (72.0%) 1Surgery, Washington University School of Medicine, Saint Louis, Approach? Department of Surgery, Baylor College of Medicine, Houston, TX; 1 258 (21.5%) 27 (18.8%) MO; 2Texas A&M University, College Station, TX;3Washington 2 89 (7.4%) 11 (7.6%) 4Internal Medicine, Baylor College of Medicine, Houston, TX David B. Stewart, Christopher S. Hollenbeak, Melissa M. Boltz University, Saint Louis, MO Surgery/Division of Colon and Rectal Surgery, Penn State Hershey 3+ 61 (5.1%) 2 (1.4%) BACKGROUND: Although depression is increasingly INTRODUCTION: Medical Center, Hershey, PA Increased crypt apoptosis is an impor- common among hospitalized patients, little data exists CONCLUSIONS: Short-term complication rates, includ- tant aspect of intestinal adaptation after massive small regarding the impact on surgical outcomes. Since recov- BACKGROUND: Though complication rates following ing surgical site infections, were similar between LAPR and (SBR). We have previously shown expres- ery from major abdominal surgery is both physically and laparoscopic colon cancer resections have been described, OAPR. Few patients are offered LAPR for cancer, which sion of the pro-apoptotic factor Bax is required for resec- emotionally challenging, preexisting depression may com- these rates are less certain for laparoscopic rectal cancer appears due to surgeon preference rather than consistently tion-induced apoptosis. In addition, in vitro experiments plicate this process by prolonging hospital stays and delay- surgery. The objective of this study was to identify fac- identifi ed patient factors. have demonstrated p38D MAPK is required for activation of ing functional recovery. We hypothesized that co-morbid tors predicting short-term complication rates for patients Bax and apoptosis. The present study was designed to test depression would be associated with prolonged hospital undergoing laparoscopic (LAPR) and open abdominoperi- the hypothesis that p38D MAPK is a key regulator of Bax stay and changes in discharge disposition for patients neal resections (OAPR) for rectal cancer, and to identify activation during adaptation after SBR in vivo. undergoing colon and rectal surgery. factors associated with choosing either surgical approach. METHODS: p38D MAPK was deleted in mice specifi cally METHODS: Patients age 18 or older undergoing colon and in the intestinal mucosa by tamoxifen injection via a rectal surgery were identifi ed from the 2008 Nationwide Cre/LoxP recombination system. p38D-defi cient and wild Inpatient Sample. Proportions were compared using chi-

34 35 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

812 813 CONCLUSIONS: Prolonged duration of pre-operative medi- At 7.4 weeks post op, QOLRAD scores improved from 4.03 cal treatment correlates with poor post-operative outcomes to 6.01 (p < 0.0001) and GERD-HQRL improved from 20.5 Ex-Vivo Intra-Arterial Methylene Blue Dye Injection Prolonged Pre-Operative Hospitalization Correlates following colectomy for acute, fulminant UC. Additionally, to 6.8 (p < 0.0001). At 9.3 months, QOLRAD improved Augments Pathology Lymph-Nodes Harvest Followingwith Worse Outcomes After Colectomy for Acute sustaining a major post-operative complication does not from 4.07 to 6.63 (p < 0.0001) and GERD-HRQL improved Colorectal Cancer Surgery Fulminant Ulcerative Colitis prevent patients from eventually undergoing ileal pouch from 20.0 to 4.6 (p < 0.0001). The Dysphagia Severity Score Pierpaolo Sileri1, Luana Franceschilli1, Marco D’Eletto1, Brian A. Coakley, Dana A. Telem, Scott Q. Nguyen, anal anastomosis. worsened at 7.4 weeks from 34.5 to 28.4 (p < 0.01) but then Giulio P. Angelucci 1, Sara Lazzaro1, Mara Capperucci1, Giampiero Celia M. Divino improved at long-term to 42.5 (p < 0.002). Palmieri 2, Nicola Di Lorenzo1, Vincenzo Formica3, Achille Department of Surgery, The Mount Sinai Medical Center, New 814 CONCLUSIONS: Laparoscopic Nissen-Hill hybrid is an 1 Gaspari York, NY The Laparoscopic Nissen—Hill Hybrid Anti-Refl ux effective repair for patients with PEH and/or BE. In this 1 2 Surgery, University of Rome Tor Vergata, Rome, Italy;Pathology, BACKGROUND: Although colectomy has long been con- Repair Is Effective for Managing Paraesophageal diffi cult group the clinical recurrence rate was 2% and no University of Rome Tor Vergata, rome, Italy; 3Oncology, University sidered defi nitive treatment for fulminant ulcerative colitis Hernias and Barrett’s Esophagus patient had resumed antisecretory medication at a mean of Rome Tor Vergata, rome, Italy (UC) unresponsive to medical management, debate persists follow up of 12.8 months. Long-term outcomes and com-

Ralph W. Aye, Alia Qureshi, Alexander S. Farivar, Abstracts

parison with other repairs are needed. Monday INTRODUCTION: Lymph node (LN) assessment has a as to the proper timing of surgery. Early surgical interven- Oliver J. Wagner, Ariel W. Knight, Brian E. Louie critical role in staging colorectal cancer (CRC) thus infl u- tion may be benefi cial to patients with acute UC. Our goal was to compare outcomes of those undergoing colectomy Esophageal and Thoracic Surgery, Swedish Cancer Institute 815 encing 5-year survival rates. However, several studies have and Medical Center, Seattle, WA shown that nodal harvest is highly variable and often inad- for acute UC and to identify pre-operative factors that Routine Contrast Radiology After Oesophagectomy and BACKGROUND: equate. In this prospective study we evaluated if ex-vivo might predict poor outcome. We previously reported, in a pilot trial Total Gastrectomy METHODS: of 20 patients, the feasibility and safety of a new antire- intra-arterial methylene blue dye injection results in a bet- The charts of 107 patients treated by subtotal Paul Finn, S. Dresner, H. Wescott, Peter A. Davis ter and more accurate lymph-nodes harvest and staging. fl ux operation that combines components of the Nissen colectomy with ileostomy for severe fulminant ulcerative Upper Gastrointestinal Surgery, James Cook University Hospital, We also compared these results with data prospectively col- colitis from January 2004 to June 2009 were retrospec- fundoplication with the . This larger study evalu- Middlesbrough, United Kingdom lected from a cohort of patient who underwent CRC sur- tively reviewed. Twenty-nine patients sustained a major ates the clinical and objective effectiveness of the repair in gery prior the beginning of this study. post-operative complication while 78 patients recovered patients with paraesophageal hernia (PEH) and/or Barrett’s AIM: Routine contrast radiography following subtotal PATIENTS AND METHODS: uneventfully. Post-operative length of stay, OR time, esti- metaplasia (BE), where higher rates of recurrence have been oesophagectomy (STO) and total gastrectomy (TG) is still Between August 2009 and reported with traditional repairs. September 2010, a total of 105 consecutive patients with mated blood loss, 30-day readmission/reoperation rates as widely practiced in the UK. This study aimed to evaluate a primary resectable CRC were enrolled for this study. Patients well as peri-operative morbidity and mortality were com- METHODS: An IRB approved, prospective study was per- protocol of the selective use of contrast radiography after with previous colorectal surgery, evidence of advanced or pared using student t-test, Fisher exact and chi-square anal- formed on patients with symptomatic PEH or GERD with surgery. metastatic disease at surgery, were excluded. Same surgeons ysis when appropriate. non-dysplastic BE. Manometry, EGD and 48-hour pH test- METHODS: 179 patients (median age 65 years, M:F 3.1:1) and same pathologists were involved in the study. After ing were obtained pre-operatively and 6 months post- underwent STO (n = 120) and TG (n = 59) for malignancy surgery, retrieved specimens were ex-vivo injected (56) or Post-Operative Complications (Within 30 Days) operatively. Quality of life was measured with QOLRAD, from 01-April-2006 to 30-September-2010. 127 patients not (49) with methylene blue dye and sent for standard GERD-HRQL and Dysphagia Severity Scores pre-opera- (routine group) underwent contrast swallows on days 3-7 COMPLICATION N (%) tively, at 6 weeks, 6 months and one year. pathology. The two groups were well-matched for age, Wound infection 6 (6) post-operatively, whereas 52 patients (selective group) only gender, BMI, neoadjuvant therapy and type of surgery. Intra-Abdominal Abscess Percutaneous drainage 9 (8) 7 (7) 1 (1) RESULTS: Fifty-fi ve patients were enrolled in the study had contrast radiography if clinically indicated. Lymph-nodes were grouped into four categories according Operative drainage with 50 available for analysis. Thirty patients with PEH RESULTS: 125 (98.4%) patients in the routine group Small bowel obstruction Transient Reoperation 10 (9) 7 (7) 3 (3) to the size: <1 mm, 1–2 mm, 2–4 mm and greater than (mean hernia size = 6 cm), 15 with BE and 5 with both received contrast swallows whereas only six (11.5%) patients 4mm. These results were also compared to data obtained Duodenal ulcer 2 (2) underwent Hybrid repair and were followed for a median Clostridim diffi cile infection 3 (3) in the selective group underwent contrast radiography. Ten from a cohort of 146 patients who underwent CRS surgery of 12.8 months (6.4–39.1 months). patients (5.6%) developed anastomotic leaks, four of which between September 2005 and July 2009. Mann-Whitney Dehiscence 4 (4) Pneumonia 2 (2) were not identifi ed at initial routine radiography (sensitiv- and Student t-test were used for statistical analysis. There were no peri-operative mortalities. There were four Sepsis 2 (2) major complications, including 1 intra-operative bougie ity 55.6%, specifi city 99.2%). Evidence of a leak occurred in RESULTS: Methylene blue injection failed in 3 cases (5.7%) perforation, 1 myocardial infarction, 1 intra-operative a further four patients after an apparently normal contrast mainly due to extravasation of colorant, mis-identifi cation bleed and 1 early reoperation for esophageal obstruction. examination. Anastomotic leak rates did not differ signifi - of the artery and incomplete mapping. The average lymph- RESULTS: Admission WBC was signifi cantly higher among cantly between STO (6.7%) and TG (3.4%), or between the There was one clinical recurrence (2%), requiring reop- node harvest was 17 ± 6 (range 10–33) in the stained and patients who developed a post-operative complication, routine (7.1%) and selective (1.9%) group. Length of stay eration at 3. 5 years, and two (4%) asymptomatic fundic 15 ± 7 (range 4–34) in the unstained group. Methylene blue but there were no other differences in patient character- (median 14 days) and in-hospital mortality also did not dif- herniations with an intact GEJ, all in the PEH group. No injection reduced the risk of inadequate LNs harvest (LNs istics (age, gender, ASA class, chronic steroid use), other fer signifi cantly between the routine (3.1%) and selective patient has resumed antisecretory medication. Control of < 8: 14% to 0, p = 0.05; LNs < 12: 32% to 21%, p = 018). measures of acute illness (hemoglobin, albumin) or extent group (0%). Post-operative radiological costs were signifi - preoperative symptoms was achieved in 96%. Two patients The largest difference was found in size groups between of disease involvement (left-sided versus extensive). Uni- cantly higher in the routine group. required post-operative dilatation for dysphagia, with reso- <1 and 4 mm causing a shift in size distribution toward variate analysis showed that patients who developed post- CONCLUSION: lution. Mean DeMeester scores improved from 51.4 to 8.5 Routine contrast radiography cannot be smaller lymph-nodes retrieval. This difference was signifi - operative complications were taken to surgery signifi cantly (p < 0.001) and mean percent time pH < 4 decreased from recommended after STO or TG. The leak rate is low and cantly more evident for LN <1 mm (9% vs. 5%, p = 0.007). later (3.6 vs. 6.6 days, p = 0.01) than those who recovered 11.9% to 2.1% (p < 0.01). Abnormal DeMeester scores were a highly selective approach to the use of contrast radio- When we compare these results to data obtained before the uneventfully. While laparoscopic colectomy took signifi - noted in 4 asymptomatic patients (3 PEH and 1 BE). Lower graphy is both safe and cheaper. beginning of this study, we confi rmed a signifi cant reduc- cantly longer (168 vs. 223 minutes, p < 0.01) than open esophageal sphincter pressures (LESP) increased from 16.2 tion of inadequate harvesting when methylene blue dye is surgery, it did not convey an elevated risk of morbidity. to 22.8 mmHg (p < 0.03) with residual LESP increasing used (LN<8; from 19% vs. 0%; p=0.002) (LN<12; from 40% No other parameter studied, including administration of from 4.1 to 10.4 mmHg (p < 0.002). vs. 21%,p=0.004). The failed injection did not signifi cantly several specifi c medical therapies (antibiotics, intravenous infl uence the mean number of retrieved LNs. steroids, immunosuppressive agents or anti-infl ammatory CONCLUSIONS: Ex-vivo intra-arterial methylene blue dye drugs), was signifi cant for predicting postoperative com- injection augments lymph-nodes retrieval thus allowing a plications. Similar percentages of patients in both groups more accurate colorectal cancer staging and possibly the eventually underwent ileal pouch anal anastomosis (68% oncologic outcome. vs. 77%, p = 0.5).

36 37 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

We divided the 316 operations into two halves, the fi rst and mucosal levels of IL-1E, IL-6 and TNF-D were measured tamine/H1R, rho = 0.841, p < 0.0001 and Histamine/H2R, /816 158 patients (Group 1) were operated on from 2/1/1992 with immunometric assays. Faecal lactoferrin was analyzed by rho = 0.806, p < 0.0001). The results of increased H1R and Changes in Re-Operative Intervention for Failed to 6/5/2006, the second half (Group 2) was operated on quantitative ELISA. After a median follow up of 23 (IQR 20-24) H2R expression in complicated sigmoid diverticulitis were Anti-Refl ux Surgeries Over the Past 19 Years in Our from 6/15/2006 to 11/2/2010. The most common indica- months the patients were contacted for a reassessment of cur- confi rmed on mRNA-level (RT-PCR, p = 0.009). Practice tions for a re-operative procedure were heartburn (27% and rent and past disease activity. CONCLUSIONS: Our fi ndings suggest that allergic predis- 26%) and dysphagia (33% and 28% in each group) and did RESULTS: During the follow up chronic/relapsing pouchi- position is another important risk factor for complicated Arpad Juhasz, Masato Hoshino, Abhishek Sundaram, not change over the years. Pre-operative upper endoscopy Tommy H. Lee, Charles J. Filipi, Sumeet K. Mittal tis was diagnosed in 6 patients. Mucosal level of D-5 and courses of acute sigmoid diverticulitis and linked with revealed a recurrent sliding hiatus hernia (>2 cm) in 61 and D-6 defensins correlated with chronic/relapsing pouchi- histamine receptor expression. Supportive therapies with Department of Surgery, Creighton University Medical Center, 71 patients, disrupted fundoplication in 74 and 89 patients, tis onset (tau = 0.30, p = 0.034 and tau = 0.28, p = 0.053, antihistaminic drugs might also become an option. Aller- Omaha, NE twisted fundoplication or two-compartment stomach in 20 respectively). High levels of D-5 defensin resulted predic- gic predisposition might be worth considering (similar to BACKGROUND: Parallel with the increasing number and 26 cases, slipped fundoplication in 56 and 54 patients tive of chronic/relapsing pouchitis (AUC = 74% [95% CI = immunosuppression and steroid intake) when indicating of laparoscopic anti-refl ux operations there is increasing and a paraesophageal hernia in 32 and 37 patients in each 53–89%], p = 0.052). Patients with high level of D-5 and D-6 surgery for sigmoid diverticulitis. number of re-operative interventions for failed procedures. group respectively. There was no signifi cant difference in defensins had earlier pouchitis relapses (p = 0.009 and p = the pattern of failure over the years. Re-operation consisted We have performed 316 re-operative procedures from 0.034, respectively). High levels of D-5 defensin were asso- 887 Feb/1992 to Nov/2010. The aim of this study was to com- of redo-fundoplication in 141 and 81 patients; Roux-en-Y ciated to a signifi cant risk of chronic/relapsing pouchitis pare presenting symptoms, endoscopic fi ndings and opera- reconstruction (RNY) in 14 and 70 patients and esopha- (OR = 10.6 [95% CI = 1.2–97.6], p = 0.027). At multivariate Neoadjuvant Radiochemotherapy in Adenocarcinoma tive approaches over the years. gectomy was performed in 3 and 7 patients, respectively. analysis mucosal levels of D-5 defensin and the number of of the Esophagus: ERCC1 Gene Polymorphisms for A signifi cantly higher number of procedures were done Prediction of Response and Prognosis METHODS: CFU of mucosa-associated Clostridiaceae spp resulted to be A retrospective review of a prospectively laparoscopicaly in the second group (108 vs 80, p = 0.001). independent predictors of chronic/relapsing pouchitis (tau Ralf Metzger, Ute Warnecke-Eberz, Hakan Alakus, Jan Brabender, maintained database was performed of patients who under- A large hiatus hernia (>2 cm) with a slipped fundoplica- = 0.46 [0.18], p = 0.024 and tau = 0.44 [0.18], p = 0.027, Daniel Vallbohmer, Peter P. Grimminger, Arnulf H. Hölscher, went re-operative intervention after one or more previous tion was found in 29 and 38 patients in the two groups. respectively). Elfriede Bollschweiler anti-refl ux surgeries. After institutional board review the In this subset of patients a signifi cantly larger number of CONCLUSION: database and charts were reviewed and analyzed for pre- RNY reconstructions were performed in the second group Chronic/relapsing pouchitis is associated Department of General, Visceral and Cancer Surgery, University of Abstracts

senting symptoms, pre-operative endoscopic fi ndings and (3 vs 20, p = 0.0003), while gastroplasty was used more fre- to increased expression of mucosal D-5 defensin. High Cologne, Cologne, Germany Tuesday operative procedures. levels of D-5 defensin may be the response to challenge quently in the fi rst group (8 vs 0, p = 0.0001). INTRODUCTION: Neoadjuvant radiochemotherapy (RT/CTx) of Clostridiaceae spp antigens or, alternatively they may RESULTS: There was a signifi cant increase in the number DISCUSSION: regimens were primarily designed for treatment of squa- While there has been no signifi cant change favour Clostridiaceae spp decreasing other competing bac- of re-operative interventions over the years (fi rst 5 yrs: 8 mous cell carcinoma (SCC) of the esophagus. Own results in the presenting symptoms and anatomical derangements teria familiae. procedures, second 5 yrs: 63 operations, third 5 yrs: 105 encountered in patients undergoing re-operative interven- demonstrate that also patients with locally advanced ade- operations, last 4 yrs: 140 operations). The procedure was tion for failed fundoplication, a larger number of cases nocarcinoma (AC) of the esophagus may achieve a major the fi rst re-operation in 289 cases, second in 25 cases, have been completed laparoscopically with an increasing /886 response in 30% with a 3 year survival rate of 80%. To iden- third in 1 case and in 1 case it was the fourth re-operation. utilization of RNY construction. Allergic Predisposition, Histamine and Histamine tify these patients, ERCC1 (rs11615) gene polymorphisms Receptor Expression (H1R, H2R) Are Associated with were analyzed. ERCC1 is a key enzyme of the nucleotide Complicated Courses of Sigmoid Diverticulitis excision and repair (NER) complex to prevent DNA inter- TUESDAY, MAY 10, 2011 and intrastrand crosslinks. Burkhard H. von Rahden1, Martin Grimm1, Stefan Kircher2, METHODS: Maria Lazariotou 1, Matthias Jung1, Christoph-Thomas Germer1 Genomic DNA from 144 patients with adeno- 1Department of Surgery, University Hospital Wuerzburg, carcinoma of the esophagus was extracted from paraffi n- 8:00 AM – 9:30 AM embedded tissues. For analysis of ERCC1 single nucleotide Wuerzburg, Germany;2 Department of Pathology, University S501a polymorphisms (SNPs) allelic discrimination was per- Hospital Wuerzburg, Wuerzburg, Germany TRANSLATIONAL SCIENCE PLENARY formed by quantitative real-time PCR. Two allele-specifi c BACKGROUND: We aimed to evaluate the hypothesis TaqMan probes in competition were used for amplifi ca- (PLENARY SESSION VI) that allergic predisposition and expression of histamine tion of ERCC1 (rs11615). Allelic genotyping was correlated receptors might contribute to complicated courses of sig- with histomorphologic tumor regression and survival after moid diverticulitis. neoadjuvant RT/CTx. Major response (MaHR) was defi ned METHODS: Expression of histamine and histamine as ≤ 10% vital residual tumor cells. 885 BACKGROUND: Defensins are small cationic peptides receptors (H1R, H2R) were analyzed on protein (IHC/ IF) RESULTS: Analysis of tumor regression revealed a MaHR with antibacterial activity expressed only in Paneth cells as well as mRNA (RT-PCR) level in surgical specimen (n = in 49/144 (34%) patients with a 5-year survival rate (5-YSR) Innate Immune Environment in Ileal Pouch Mucosa: (D-defensins) or generally in intestinal epithelial cells 101) of patients with complicated (n = 57) and uncompli- of 72% (p = 0.0001). ERCC1 gene polymorphisms for all D-5 Defensin Up-Regulation as Predictor of Chronic/ (E-defensins) that have a profound effect on gut microbi- cated (chronically-recurrent, n = 44) sigmoid diverticulitis. patients showed the following expression pattern: ERCC1 Relapsing Pouchitis ota. We have recently shown that chronic pouchitis, that Results were correlated with clinical parameters. polymorphism (rs11615) CC: n = 13 (9%), TT: n = 74 (51%), 1 2 2 2 Marco Scarpa, Alessia R. Grillo, Melania Scarpa, Paola Brun, occurs in 5% of patients after restorative proctocolectomy RESULTS: The mean number of comorbid diseases per CT: n = 57 (40%). ERCC1 polymorphism CT was identifi ed Carlo Castoro 1, Anna Pozza3, Diego Faggian4, Cesare Ruffolo5, and can cause pouch failure, is associated to a signifi cant patient was 1.76 ± 1.25. Thirty-nine of 100 patients (39%) as a predictor for response to the neoadjuvant RT/CTx (p < Renata D’Incà 3, Romeo Bardini3, Ignazio Castagliuolo2, increase Clostridiaceae spp. The aim of this study was to exhibited allergic predisposition (grass poll, food, drug, 0.001). The 5-YSR for patients with CT genotype was 53%. Imerio Angriman 3 gain further insight in the pathogenesis of chronic/relaps- pets, etc.). Comorbid diseases were signifi cantly associated Contrary to this, the 5-YSR for the group of patients with a CC/TT-polymorphism decreased to 33%. 1Department of Oncological Surgery, Venetian Oncology ing pouchitis by exploring the interplay between the with complicated diverticulitis (p = 0.027). Complicated microbiota adherent to the ileal pouch mucosa and defen- CONCLUSION: Institute (IOV-IRCCS), Padova, Italy;2Department of Histology, sigmoid diverticulitis was signifi cantly associated with Analysis of ERCC1 (rs11615) gene poly- sin expression. Microbiology and Medical Biotechnologies, University of Padova, high H1R and H2R expression, in comparison to uncom- morphisms reveals a signifi cant (p < 0.001) correlation with PATIENTS AND METHODS: Padova, Italy; 3Department of Surgical and Gastroenterological Thirty-two consecutive patients plicated types (p = 0.0304). Furthermore, an association of response and survival in patients with adenocarcinoma of complicated diverticulitis with allergic preconditions was the esophagus treated with a neoadjuvant radiochemother- Sciences, University of Padova, Padova, Italy;4Department of coming for follow up endoscopy, were recruited. On pouch found (OR = 3.2; 95% confi dence interval: 1.3271–7.8239, apy (CDDP, 5-FU, 36Gy). Single nucleotide polymorphisms Diagnostic Sciences and Special Therapy, University of Padova, biopsies we determined bacteria adherent to the mucosa, by culture method, and mRNA for D-and E-defensins and Toll Like p = 0.0097). IF double-labelling experiments showed a of ERCC1 (rs11615) could therefore be applied to further Padova, Italy; 5 Department of Surgery (Chirurgia IV), Ospedale Receptor-4 and -2, by quantitative Real Time RT-PCR. Serum strong correlation of increased histamine expression on individualize therapy in esophageal cancer. “Ca’ Foncello”, Treviso, Italy H1R as well as H2R expressing intestinal enterocytes (his- 38 39 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

typed as Villin Cre (+); Rb (+/fl ox); p21 (–/–). (These mice /888 /889 /890 will be referred to as p21-null-Rb-hets). All genotypes were Rage Gene Deletion Inhibits the Development and Accumulation of Pro-Cancer Cytokines in the Plasma Regulation of Retinoblastoma Protein (Rb) by p21 confi rmed by PCR. Enterocyte crypt fractions were isolated Progression of Ductal Neoplasia and Prolongs Fraction of Stored Packed Red Cells Is Critical for Adaptation to Massive Small Bowel and Rb protein levels were evaluated by Western Blot and Survival in a Mouse Model of Pancreatic Cancer 1,2 3 3 Resection quantifi ed by optical density using Image-J software. Mice Douglas Benson, Adam W. Beck, Marie Schluterman, were subjected to either a 50% small bowel resection or 1 1 1 3 1,4 1,2 Joseph DiNorcia, Minna K. Lee, Dorota N. Moroziewicz, Rolf A. Brekken , Christopher Silliman, Carlton C. Barnett Jennifer A. Leinicke, Jun Guo, Derek Wakeman, Brad W. Warnersham (transection and reanastomosis) operation. The rem- Megan D. Winner 1, Paritosh Suman1, Fei Bao3, Helen Remotti3, 1Surgery, University of Colorado at Denver HSC, Aurora, CO; Department of Surgery, Washington University in St. Louis, nant ileum was harvested 7 days after operation, and adap- Yu Shan Zou 2, Shi Fang Yan2, Wanglong Qiu1, Gloria H. Su1, 2Surgery, Denver Health Medical Center, Denver, CO;3Surgery, St. Louis, MO tation was evaluated by measuring villus height and crypt 2 1 Ann Marie Schmidt , John D. Allendorf University of Texas at Southwestern Medical Center, Dallas, TX; PURPOSE: Adaptive growth of the intestinal mucosa in depth. 1 4 Surgery, Columbia University College of Physicians and Surgeons, Bonfi ls Blood Center, Denver, CO response to massive gut loss is critical for autonomy from RESULTS: Elevated crypt Rb levels in p21-null mice were 2 New York, NY; Medicine, New York University Langone Medical INTRODUCTION: Perioperative blood transfusion has parenteral nutrition. Normal intestinal adaptation is char- successfully reduced to levels comparable to wild-type ani- Center, New York, NY; 3Pathology, Columbia University Medical been linked to decreased survival in cancer; however the acterized by the structural changes of increases in villus mals by deleting one Rb allele. Unlike the p21-null mice, Center, New York, NY exact causal mechanism has not been elucidated. Alloge- height and crypt depth. Our laboratory has previously dem- the p21-null-Rb-het mice that underwent SBR did manifest onstrated that p21-null mice do not to adapt after small BACKGROUND: Activating Kras mutations and inactiva- neic transfusions are known to expose patients to foreign adaptive changes in crypt depth and villus height when bowel resection (SBR). Further studies reveal that these ani- tion of the p16 tumor suppressor gene are commonly pres- cells and lipid mediators. We hypothesize stored packed compared to sham animals. (crypts: Sham 59.6 + 2.3 um, mals have dramatically increased levels of Rb protein in the ent in pancreatic ductal adenocarcinoma. The receptor for red cells contain pro-cancer cytokines that augment tumor SBR 74.3 + 0.4 um, p < 0.001; villus: Sham 167.5 + 0.8 um, crypt cells. Alternatively, we have previously shown that advanced glycation end-products (RAGE) is a cell surface progression. SBR 256.5+¬17.5 um, p > 0.006). intestine-specifi c Rb defi ciency results in intestinal muco- receptor implicated in carcinogenesis. METHODS: CONCLUSION: By deleting one Rb allele in a p21-null Chemiarray analysis for pro-cancer cytokines sal hyperplasia manifested by increased villus height, crypt background, Rb levels were successfully decreased to lev- OBJECTIVE: The aim of this study was to evaluate the was performed on the acellular fraction of stored leuko- depth, and crypt cell proliferative rate. In this study, we els comparable to wild-type animals. Correspondingly, effect of Rage gene deletion on the development and pro- cyte reduced (LR) and non leukocyte reduced (NLR) packed tested the hypothesis that the magnitude of Rb expression the normal adaptive response to intestinal resection was gression of pancreatic ductal neoplasia in a conditional red blood cells (pRBCs) at days 1(D1) and 42 (D42 -out- in p21-null mice is crucial for the normal adaptive response Abstracts

G12D restored in these p21 null mice. These fi ndings suggest a Tuesday Kras , conventional p16 knockout mouse line. date) of storage. This analysis revealed elevated levels of to massive SBR. angiogenin, EGF, MCP-1, PDGF, RANTES, and TNF-D. Spe- crucial mechanistic role for Rb in p21-directed small intes- MATERIALS AND METHODS: METHODS: To evaluate the impact cifi c enzyme-linked immunosorbent assay was performed To reduce Rb protein levels in p21-null mice, tine adaptation responses to massive small bowel resection. of Rage deletion on pancreatic intraepithelial neoplasia for each of these cytokines to confi rm elevation. Blood one Rb allele was deleted by breeding villin Cre Rb (fl ox/ G12D –/– (PanIN), Pdx-1-Cre/Kras mice were crossed with Rage products from each donor were compared to evaluate dif- fl ox) mice with p21-null mice, resulting in offspring geno- mice in the C57BL/6 background to generate Pdx-1-Cre/ ferences in storage time and leukoreduction on cytokine G12D –/– Kras /Rage mice. To evaluate the impact on pancre- concentration. Data were analyzed by ANOVA; p ≤ 0.05 sig- G12D atic ductal adenocarcinoma (PDAC), Pdx-1-Cre/Kras / nifi cant; N ≥ 4 per group. –/– -/- 9:30 AM – 12:00 PM p16 mice were crossed with Rage mice to generate Pdx-1- RESULTS: Cre/KrasG12D/p16–/–/Rage–/– mice. Mice were sacrifi ced and Angiogenin levels were different between LR S501a pancreata were procured for scoring of ductal neoplasia and NLR blood, 0ng/ml (undetectable) vs. 44.2 ± 3.7 ng/ml compared to Pdx-1-Cre/KrasG12D or Pdx-1-Cre/KrasG12D/p16-/- con- (p < 0.001) at D1. Storage time had no effect on concen- PLENARY SESSION VII trols. To evaluate the impact on survival, mice were aged tration. EGF levels are similar at D1 in LR v NLR, 216.4 ± to their natural death, and Kaplan-Meier analysis was per- 3.8 pg/ml vs. 241.1 ± 13.1 pg/ml, and increased with stor- formed. Categorical variables were compared using Fisher’s age time in NLR blood only, 1436.4 ± 238.6 pg/ml at D42 exact test, and median survivals were compared with the (p = 0.001). MCP-1 levels were increased with storage time 955 junction at least 5 years after antirefl ux surgery as iden- log-rank test. in LR, 86.3 ± 6.3pg/ml at D1 vs. 121.2 ± 6.1 pg/ml at D42 tifi ed in the Swedish Cancer Register. The controls were (p = 0.007), and increased markedly in NLR, 78.2 ± 7.3 pg/ Risk Factors for Developing of Esophageal G12D -/- matched to the cases regarding age, sex, and calendar-year RESULTS: At 16 weeks of age, Pdx-1-Cre/Kras /Rage Adenocarcinoma After Antirefl ux Surgery ml at D1 vs. 647.8 ± 220.7pg/ml at D42 (p = 0.02). PDGF of the antirefl ux surgery. Information about the study vari- mice had signifi cantly fewer PanIN2 and PanIN3 lesions levels are reduced in LR compared to NLR, 6.8 ± 0.2 ng/ml compared to controls (5.7% vs. 14.9%, p = 0.0024). At 12 Hedvig E. Löfdahl, Yunxia Lu, Pernilla Lagergren, ables recurrent refl ux, type of antirefl ux surgery, body mass vs. 61.6 ± 6.0 ng/ml (p < 0.001), and increased with stor- Jesper Lagergren weeks of age, Pdx-1-Cre/KrasG12D/p16–/–/Rage–/– mice had sig- index (BMI), and tobacco smoking status was collected age time only in NLR, 34.7 ± 9.7 ng/ml at D1 vs. 76.5 ± 1.7 Department of Molecular Medicine and Surgery, Karolinska through review of medical records, where the reviewer nifi cantly fewer PanIN2 and PanIN3 lesions compared to ng/ml by D42 (p = 0.003). RANTES levels are reduced in Institutet, Stockholm, Sweden was kept blinded to the case-control status. Relative risk, controls (10.6% vs. 18.9%, p = 0.0033). Strikingly, none LR compared to NLR, 7.2 ± 1.6 ng/ml vs. 14.2 ± 0.8 ng/ml of the Pdx-1-Cre/KrasG12D/p16–/–/Rage–/– mice had invasive BACKGROUND: expressed as odds ratio (OR) with 95% confi dence inter- (p < 0.001), with no changes with storage time. TNF alpha The incidence of esophageal adenocarci- val (CI), was calculated through multivariable conditional PDAC compared to a 45.5% incidence of invasive PDAC levels were not different between LR and NLR with no stor- noma has increased more rapidly than that any other solid G12D –/– logistic regression, adjusted for the study variables. in controls (p = 0.0411). Finally, Pdx-1-Cre/Kras /p16 / age time effect between D1 and D42. tumor in several western populations. The main risk fac- –/– Rage mice had markedly longer median survival than tors are gastroesophageal refl ux and obesity, while tobacco RESULTS: Included were 39 cases and 133 controls. Recur- CONCLUSION: Multiple factors were identifi ed in pRBCs controls (20.9 weeks vs. 16.0 weeks, p = 0.0012). smoking is a moderate risk factor. No certain protective rent refl ux was followed by an adjusted 3-fold increased that can augment tumor progression. These data reveal a CONCLUSION: effect has been shown from antirefl ux surgery. We hypoth- the risk of esophageal adenocarcinoma compared to those RAGE signaling inhibition signifi cantly trend toward elevated pro-cancer cytokines in non leuko- esized that patients who despite antirefl ux surgery develop without refl ux symptoms after antirefl ux surgery (OR 3.1, reduced the development of pancreatic neoplasms and cyte reduced pRBCs and signifi cant accumulation during esophageal adenocarcinoma more than 5 years after such 95% CI 1.3-7.3). There were no statistically signifi cant dif- signifi cantly prolonged survival in this mouse model of storage. The negative affects of blood transfusion are likely surgery more often have recurrent refl ux or are more often ferences in risk comparing total with partial fundoplica- pancreatic cancer. Further work is needed to target the multi-factorial, and the cytokines identifi ed herein may exposed to the other risk factors, compared to patients who tio (OR 0.9, 95% CI 0.4-2.2), BMI over 30 with BMI lower ligand-RAGE axis for possible prophylaxis and treatment of represent possible therapeutic targets to offset negative pancreatic cancer. do not develop this cancer after antirefl ux surgery. than 25 (OR 1.5, CI 0.4-5.5) or current smoking with non- effects of transfusion. Further, routine leukocyte reduction smoking (OR 1.2, 95% CI 0.5-2.8). may diminish untoward affects of pRBCs transfusion. METHODS: A case-control study was nested within a cohort of all patients in Sweden who had undergone anti- CONCLUSION: Recurrence of refl ux after antirefl ux sur- refl ux surgery during 1965-2006 according to the Swedish gery seems to be a key factor in explaining the lack of any Patient Register. The cases were patients who developed clear protective effect of antirefl ux surgery regarding risk of adenocarcinoma of the esophagus or gastroesophageal esophageal adenocarcinoma.

40 41 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

956 957 Mortality Risk Prediction Comparison 2. Estimation of potential bowel loss with conven- tional tumor resection, based on segmental vascular A Novel Approach for Quantifi cation of Hepatitis C A Root-Cause Analysis of Mortality Following Major Risk Prediction Tool sacrifi ce. Virus in Liver Cirrhosis Using In-Situ Real Time ReversePancreatectomy (All mean values) Mortality N = 184 No Mortality N = 630 ASA Score 2.48 2.47 3. If residual small bowel length with tumor resection Transcriptase PCR Charles M. Vollmer, Norberto J. Sanchez, Tara S. Kent, POSSUM was estimated to be >180 cm, conventional resection Mark P. Callery, Pancreatic Mortality Study Group Kewal K. Maudar Physiologic score Operative 19.97 18.60 22.4% 19.14 15.69 13.2% was planned. Surgery, Bhopal Memorial Hospital and Research Centre, Bhopal, Beth Israel Deaconess Medical Center, Boston, MA score Predicted Mortality Madhya Pradesh, India INTRODUCTION: NSQIP (Whipple Resections only) 4. Where bowel salvage was not possible with conven- Although mortality rates from pan- Predicted Mortality 8.18% 2.92% tional resection, the tumor and the entire mid-gut INTRODUCTION: HCV (hepatitis C virus) infects nearly createctomy have decreased worldwide, death remains an Charlson was excised en-bloc. If auto-transplantation was a 3% of the population worldwide and has emerged as a infrequent but profound event at an individual practice Predicted Mortality Institutional 5.17% 1.47% 2.94% 0.86% technical option, the specimen was fl ushed with Adjusted Predicted Mortality major causative agent of liver disease, resulting in acute and level. Root-cause analysis is a retrospective method com- cold preservative solution with ex-vivo tumor resec- chronic infections that can lead to fi brosis, cirrhosis and monly employed to understand adverse events. We eval- tion, followed by vascular reconstruction and auto- hepato-cellular carcinoma. Current laboratory diagnosis of uate whether emerging mortality-risk tools suffi ciently transplantation of the intestine. HCV is based on specifi c antibody detection (anti-hepatitis predict and account for actual clinical events that are often CONCLUSION: Root-cause analysis suggests that risk- C virus-anti-HCV) in serum. Determination of HCV RNA identifi ed by root-cause analysis. prediction should include, if not emphasize, operative 5. In the absence of malignancy, patients with extreme short bowel syndrome, following macroscopically concentrations reduces the pre-seroconversion period in METHODS: We assembled a Pancreatic Mortality Study factors related to pancreatectomy. While risk models can the diagnosis of HCV infection and supports management distinguish between mortalities and non-mortalities in complete tumor resection, were evaluated for intesti- Group comprised of 32 pancreatic surgeons from 14 insti- nal allo-transplantation. of interferon alpha-based therapies. As HCV replicates in tutions in 4 countries. Mortalities after pancreatectomy a collective fashion, they vastly miscalculate the actual RESULTS: the liver cells of these patients, detection and localization (30 and 90-day) were accrued from 2000–2010. For root- chance of death on an individual basis. Eight patients met criteria for inclusion. Their of HCV RNA in liver tissue are vital for diagnostic purposes cause analysis, each surgeon “deconstructed” the clini- characteristics are in Table 1. and clinical management of such patients, as well as for the cal events preceding a death to determine cause. We next 958 Five of the 8 patients are currently alive; 7 of the 8 patients, elucidation of viro-pathological mechanisms. tested whether mortality-risk tools (ASA, POSSUM, Charl- A Transplant Perspective on “Non-Resectable” Soft including all 5 survivors have had complete tumor resec- Abstracts MATERIALS AND METHODS: A total of 10 biopsy son, NSQIP) could predict those patients who would die Tissue Tumors at the Root of the Mesentery: Procedures,tion with macroscopically clear margins. The three deaths Tuesday samples diagnosed for liver cirrhosis that were negative (n = 184), and compared their prognostic accuracy against Promise and Pitfalls! occurred after failed/aborted resection in 1, after success- for the presence of anti-HCV and serum HCV RNA were a cohort of resections in which no patient died (n = 630). 1 1 1 ful ex-vivo resection, due to possible cardiac causes in a undertaken for analyzing the presence of viral nucleic acid RESULTS: Kishore R. Iyer, Hiroshi Sogawa, Gonzalo Rodriguez-Laiz, second, and after “swine H1N1 infl uenza” 7 weeks after 184 deaths (151 Whipples, 18 Distals, 15 Totals) 2 2 2 in liver tissues. The diagnosis of liver cirrhosis was made Samuel Singer , Daniel G. Coit, Murray F. Brennan successful intestinal allo-transplantation in the third were identifi ed from 10,783 performed 1 on the grounds of histology and/or laparoscopy. Since no by surgeons whose experience averaged 13.5 years. Overall Transplant Surgery, Mount Sinai Medical Center, New York, NY; patient. One patient has full graft function after intestinal 2 defi ned cause of liver cirrhosis was identifi ed the etiology 30- and 90-day mortalities were 0.92% and 1.71%. Indi- Surgery, Memorial Sloan Kettering Cancer Center, New York, NY allo-transplantation and one is currently awaiting intesti- of liver cirrhosis was classifi ed as cryptogenic. Qualita- vidual institutional rates ranged from 0.3–4.7%. Only 5 BACKGROUND: Soft tissue tumors at the mesenteric root nal transplant. Two of the fi ve survivors require parenteral tive screening for HCV was done through ELISA while the patients died intraoperatively, while the other 179 suc- are often not resectable by conventional means because of nutrition. Three patients, including both the auto-and allo- nucleic acid analysis was performed through Light Cycler cumbed at a median of 27 days. Mean patient age was 70 major vascular involvement; if benign and symptomatic, transplant are on full oral feeds. 2.0 (minimum detection limit 10 copies/ml). The detection years old (39% were >75 y). 89% of cases were for malig- intestinal allo-transplantation is a potential option. CONCLUSIONS: Our preliminary experience is consistent of HCV RNA in liver tissue biopsies was performed by fol- nancy, mostly pancreatic cancer(54%). Median operative AIM: To describe initial experience with a planned surgical with the surgical challenge of large tumors at the mesen- lowing standard protocol of HCV detection kit (Shenzhen time was 370 min and EBL was 750 cc (100–16,000 cc). approach to “inoperable” tumors at the root of the mesen- teric root involving the superior mesenteric vessels. We PG Biotech) with modifi cations using Light Cycler 2.0. Vascular repair or multivisceral resections were required tery within the setting of an intestinal transplant program. believe ex-vivo resection with auto-transplantation may RESULTS: It was observed that all the 10 samples were for 14% and 16% respectively. 82% had a variety of major be an option in carefully selected cases. However, our data METHODS: negative for the anti-HCV and serum HCV RNA. While the complications before death. 84% required ICU care, 51% We undertook retrospective review of patients indicate the need for extreme caution. Rapid access to intes- quantitative detection in liver biopsies following the modi- were transfused, and 36% were reoperated upon. 52% died with “inoperable” tumors at the mesenteric root, operated tinal transplant is critical for success. While our aggressive fi ed method showed the presence of HCV RNA in 3 samples during the index admission while another 9% died after a upon over a 2-year period. The operative approach con- approach challenges conventional limits of resectability out of the 10 studied. readmission. Almost half (n = 85) expired between 31 and sisted of: for these diffi cult tumors, the potential morbidity of this 90 days. Only 12% had autopsies. Operation-related com- CONCLUSION: The observed results indicate that the Light 1. Early mesenteric vascular dissection and segmental approach must be carefully considered in the context of the plications contributed to 42% of deaths, with pancreatic Cycler 2.0 following the modifi ed technique described here control. often benign clinical course of mesenteric desmoid tumors. fi stula being the most evident (16%). Technical errors (23%) may constitute a reliable method of quantitative detection Table 1: Patient Characteristics and poor patient selection (16%) were cited by surgeons. and localization of HCV in tissue, since it allows identifi ca- 5.4% of deaths had associated cancer progression – all tion of specifi c regions of the viral genome with high con- Patient occurring between 31–90 d. Even after root-cause scrutiny, number Age Sex Diagnosis Procedure Outcome Comments fi dence. It may be an important investigative tool to study the ultimate cause of death could not be determined for 41 1 30 Male Desmoid Aborted resection Died Abdominal wall and para-spinal fi xity at “serosilent” HCV infection in cryptogenic cases of liver. patients – most often between 31–90 d. While assorted risk operation Further studies, comprising a larger sample are under devel- 2 54 Male Lipo-sarcoma Ex-vivo resection; auto-transplant Alive Full oral feeds models predicted mortality with variable discrimination opment in order to assess possible association between the 3 59 Male Fibromatosis Conventional resection Alive Full oral feeds from non-mortalities, they consistently underestimated presence of HCV-RNA in liver and clinical manifestations, 4 53 Female Leiomyosarcoma Ex-vivo resection; auto-transplant Died On full oral feeds at discharge; no the actual mortality events we report (Table). Analysis with histopathological status, viral genotypes, and even in autopsy POSSUM illustrates the impact of operative performance 5 37 Male Fibromatosis Ex-vivo resection Died Extreme short bowel; intestinal patients without evidence of circulating HCV. on determining outcome. allo-transplant; died of ‘swine H1N1’ infl uenza 6 28 Female Fibromatosis ex-vivo resection; failed auto- Alive Mesenteric ischemia after auto- transplant; allo-transplant transplant; full graft function after allo-transplant 7 43 Male Desmoid Conventional resection Alive Complete enterectomy; awaiting allo-transplant 8 31 Female Desmoid; FAP Conventional resection Alive Short bowel, on stable tpn

42 43 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

number (33.3%). Ten patients (12.5%) required ERCP for RESULTS: A positive SMA margin was identifi ed in 8 (4%) 961 /959 management of pancreatico- or choledocho-jejunal anas- specimens. Accurate histologic measurements were avail- Outcomes of Primary Surveillance for Intraductal tomotic strictures, 2 patients required Puestow procedures, able in 178 (96%) of 186 patients with a negative SMA mar- Long Term Outcome of Patients Undergoing Papillary Mucinous Neoplasm Pancreaticoduodenectomy for Non-Malignant and incisional hernias were repaired in 5 patients (6.25%). gin; of these, the free margin measured ≤1 mm in 40 (22%) 1 1 1 Disease Three patients (3.8%) died during the follow-up period; patients, between 1 and 10 mm in 72 (40%), and ≥10 mm Christy E. Cauley, Joshua A. Waters, Ryan P. Dumas, none were related to the PD. in 66 (37%). Neoadjuvant chemotherapy and/or chemora-Juliana E. Meyer 1, Mohammad A. Al-Haddad2, John M. DeWitt2, Nicholas T. Orfanidis1, David E. Loren1, Carmi Santos2, CONCLUSIONS: diation was delivered to 150 (77%) patients; a positive SMAKeith D. Lillemoe 1, C. Max Schmidt1 Eugene P. Kennedy 3 1 3 This study reports the long-term outcome , Ali A. Siddiqui, Harish Lavu, margin was identifi ed in 3 (2%) of these patients versus 1Surgery, Indiana University School of Medicine, Indianapolis, Charles J. Yeo 3 1 of PD performed for non-malignant diseases. A minority , Thomas E. Kowalski 5 (11%) of the 44 patients who received no preoperative IN; 2Gastroenterology, Indiana University School of Medicine, 1 (36.3%) of patients developed post-operative symptoms Gastroenterology and Hepatology, Thomas Jefferson University that negatively impacted their daily function including therapy (p = 0.02). Locoregional recurrence developed in Indianapolis, IN Hospital, Philadelphia, PA;2 Department of Medicine, Thomas 23% of patients with a negative SMA margin ≤1 mm and diabetes, chronic pain, and diarrhea. Post-operative endo- BACKGROUND: Jefferson University Hospital, Philadelphia, PA;3 Department of 21% of patients with a negative margin >1 mm (p = 0.84). Limited data is available regarding the scopic and/or surgical interventions were rare. Careful natural history of patients undergoing primary surveil- Surgery, Thomas Jefferson University Hospital, Philadelphia, PA consideration of post-operative quality of life should be The overall survival of patients with a negative SMA mar- gin ≤ 1mm did not differ from that of patients with a nega- lance for intraductal papillary mucinous neoplasm (IPMN). BACKGROUND: Due to improved detection of pancreatic considered when planning pancreatic resection for non- We examine the outcome of patients selected for primary malignant lesions. tive margin > 1 mm (median 33 [95% CI: 22, 44] months v. lesions, the resection of non-malignant tumors and indeter- 38 [95% CI: 34, 43] months, p = 0.12). Preoperative CT pre- surveillance in a multidisciplinary pancreas cyst clinic. We minate masses has increased. Prior studies have examined dicted a negative SMA margin (median 8.5 mm, range 3–18 hypothesize that symptoms, radiologic characteristics, and outcomes of pancreaticoduodenectomy (PD) performed for 960 mm) in 6 (75%) of 8 patients in whom it was positive. The cytology will predict risk of developing pancreatic cancer malignancy, but few have examined the long-term implica- Incidence and Impact of a Close Superior Mesenteric concordance coeffi cient between the pathologic (median in patients undergoing primary surveillance for IPMN. tions of this operation for non-malignant lesions. Artery Margin Following Pancreaticoduodenectomy 2 mm) and radiographic (median 8 mm) measurements of METHODS: Between January 2002 and March 2010, 522 AIMS: To assess the long-term outcomes, quality of life for Adenocarcinoma the SMA margin in patients with a margin <10 mm was patients were diagnosed with IPMN at a single, high vol- 0.07 (95% CI: 0.02, 0.13) (Figure 1). (QOL), and need for invasive procedures in patients under- Matthew Katz1, Huamin Wang2, Aparna Balachandran3, ume institution. Patients were prospectively stratifi ed going PD for non-malignant disease. 3 4 1 1 as low or high oncologic risk based on analysis of demo- Priya Bhosale , Xuemei Wang, Peter W.T. Pisters, Jeffrey E. Lee, Abstracts METHODS: graphic, clinical, radiologic, and cytopathologic data. Any Tuesday Patients undergoing PD for non-malignantDoug B. Evans 5, Chusilp Charnsangavej3, Jason B. Fleming1 disease between 2006–2010 were retrospectively identi- 1 patient who underwent primary operative management or Surgical Oncology, UT MD Anderson Cancer Center, Houston, less than 3 months of surveillance was excluded. fi ed and contacted to complete a symptom survey. Charts TX; 2Pathology, UT MD Anderson Cancer Center, Houston, TX; were reviewed for imaging studies, hospital admissions, RESULTS: 3Diagnostic Radiology, UT MD Anderson Cancer Center, Houston, 174 (33%) patients underwent primary opera- and invasive procedures. Subjects were surveyed regarding 4 tive management for IPMN. Of these, 46 (26%) were found disease specifi c QOL parameters and complications occur- TX; Biostatistics, UT MD Anderson Cancer Center, Houston, TX; 5 to have invasive cancer. Alternatively, 292 (56%) patients ring six months or more post surgery. Information from Surgery, The Medical College of Wisconsin, Milwaukee, TX underwent primary surveillance for IPMN. Of these, 244 the phone survey was verifi ed with review of outpatient BACKGROUND: The margin most commonly positive (84%) were classifi ed as low oncologic risk. Mean duration records. following pancreaticoduodenectomy lies adjacent to the of primary surveillance was 35 (3–99) months. 28 (11%) RESULTS: Eighty patients were identifi ed from the surgical superior mesenteric artery (SMA). In patients with rectal patients initially stratifi ed as low-risk ultimately underwent database (77 pylorus preserving, 3 classic), out of 413 PDs cancer, tumor cells within 1 mm of the analogous radial pancreatic resection with a mean preoperative surveillance performed over the same period of time, of whom 33 com- margin predict early treatment failure. The signifi cance of of 11 (4–42) months. Indications for resection were 8 (29%) pleted the phone survey. Median age was 67 yrs. Median cancer within 1mm of the SMA margin in patients with new main duct dilation or mural nodule, 6 (21%) new or follow up was 3.1 yrs. Diagnoses were IPMN (67.5%), pancreatic cancer is unknown. A recent analysis of national worsening symptoms, 5 (18%) increasing lesion size, 2 (7%) chronic pancreatitis (20%), SCA (11.3%), and MCN (1.3%). practice patterns documented a lack of attention to the concerning cytopathology, and 7 (25%) patient preference. New diabetes developed in 19.2% with 60% of those new SMA dissection (Katz MHG, Ann Surg Onc 2010, PMID: Of the 28 patients resected after surveillance, 27 (96%) diabetics requiring insulin. Of those with pre-operative dia- 20811779). We therefore sought to determine the inci- demonstrated low-grade IPMN and 1 (4%) high-grade dys- betes, blood sugars were more diffi cult to control in 42.3%, dence and signifi cance of a close SMA margin. plastic (HGD) IPMN. The patient with HGD had a family history of pancreatic cancer, but was asymptomatic, and insulin was initiated in 28.6%. Overall, 18.2% of patients METHODS: All 194 patients with pancreatic adenocarci- Figure 1: Correlation between SMA margin measured on preoperative no radiographic or cytologic indicators of malignancy. Of initiated insulin during the follow-up. Chronic abdominal noma who underwent pancreaticoduodenectomy between CT scan and upon histopathologic evaluation. pain occurred in 27.3% of patients, negatively impacting non-operated patients, 2 (1%) developed invasive cancer at 2004 and 2008 were evaluated. The SMA margin dissec- CONCLUSIONS: daily function in 6.1% and requiring chronic opiates in tion was performed directly along the periadventitial plane Despite the use of neoadjuvant ther- 18 and 51 months of surveillance. Neither of these patients 3.0%. Increased bowel frequency was seen in 39.4% with of the SMA in all cases. The distance between cancer and apy and a standardized approach to SMA dissection, the demonstrated increasing cyst size or new concerning radio- improvement in only half of those treated with pancreatic the SMA margin was measured and prospectively recorded SMA margin was positive or measured ≤1 mm in 26% of graphic features prior to the diagnosis of invasive cancer. 48 enzymes. Bowel frequency adversely affected patients on using a standard histopathologic protocol. The correspond- specimens following resection for adenocarcinoma of the (16%) patients stratifi ed as high oncologic risk were initially a daily basis in 9.1%. Overall, 36.3% of patients surveyed ing distance between tumor and the SMA was retrospec- pancreatic head. The proximity of cancer cells to the SMA managed non-operatively due to age, comorbidity or patient developed diabetes or a postoperative symptom that nega- tively measured on the preoperative computed tomogram. was routinely underestimated by standard preoperative preference. Five-year survival for this group was 62%. Of tively impacted their daily life. The exclusion of patients The concordance correlation coeffi cient between these imaging. Nonetheless, associated with optimal surgical these, 13 (27%) patients have died during follow-up: 2 with chronic pancreatitis did not signifi cantly change this measures was calculated. technique and the routine use of multimodality therapy, (15%) from pancreas cancer, 7 (54%) from other causes, and microscopic cancer approximating the SMA did not have a 4 (31%) unknown. signifi cant impact upon survival. CONCLUSIONS: For IPMN initially classifi ed as low-risk, progression to pancreas cancer during surveillance was rare. Current accepted indications for resection did not forecast malignancy in this group. More accurate markers are needed to better guide IPMN surveillance. For poor operative can- didates with high risk IPMN, progression to invasive cancer during surveillance was more common, though a substantial portion succumb to non-IPMN related death.

44 45 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

962 964 vation were taken at different time periods (6, 12, 18, 24 /963 and 36 months) after their operation and examined for the Updated Results for Dual Modality Versus Percutaneous Activation of the BMP4 Pathway and Early activation of the BMP pathway (BMP4/pSMAD 1/5/8) and Drainage for the Treatment of Symptomatic Walled Is It Worth Looking? Abdominal Imaging After Expression of CDX2 Characterize the Development expression of CDX2 and CDX1. Samples were prepared Off Pancreatic Necrosis Pancreatic Cancer Resection: A National Study of Nonspecialized Columnar Metaplasia in a Human from squamous esophageal mucosa just above the esopha- Elan R. Witkowski, Jillian K. Smith, Elizaveta Ragulin-Coyne, Model of Barrett Esophagus Michael Gluck, Andrew S. Ross, Shayan Irani, S. Ian Gan, gogastric anastomosis, from columnar-appearing mucosa Sing Chau Ng, Shimul A. Shah, Jennifer F. Tseng Mehran Fotoohi, Robert Crane, Justin Siegal, Ellen Hauptmann, Manuel R. Pera Roman1, Daniel Castillo1, Sonia Puig1, and from squamous esophageal mucosa 2 cm below the Richard A. Kozarek Surgical Outcomes Analysis & Research, Department of Surgery, Carme de Bolós 5, Mar Iglesias2, Agustin Seoane3, upper esophageal sphincter level and were assessed by University of Massachusetts Medical School, Worcester, MA Digestive Disease Institute, Virginia Mason Medical Center, Laura Comerma 2, Vicente Munitiz4, Pascual Parrilla4, immunohistochemistry, quantitative real-time PCR (qRT- Seattle, WA BACKGROUND: Sequential followup imaging is often per- Richard Poulsom 6, Luis Grande1 PCR) and Western blot (WB). formed after pancreatic resection for cancer. We attempted 1 RESULTS: BACKGROUND: Treatment of symptomatic walled off Section of Gastrointestinal Surgery, Hospital Universitario del Mar, Since June 2006, 18 patients (16 male) were to quantify the volume and cost of complex abdominal 2 pancreatic necrosis (WOPN) by dual modality-endoscopic Barcelona, Spain; Department of Pathology, Hospital Universitario del included in the study. Thirteen had an intrathoracic anas- imaging after pancreatic resection nationwide, and deter- and percutaneous-drainage (DMD) has been shown to Mar, Barcelona, Spain; 3Service of Gastroenterology, Hospital tomosis. A short segment (mean length: 15.6 mm; longest mine whether their frequent use confers benefi t. decrease length of hospitalization (LOH), use of CT scans Universitario del Mar, Barcelona, Spain;4 Service of General and 30 mm) of NSCM was detected in 10 (56%) patients, with and drain studies by radiology in comparison to percutane- METHODS: Patients with pancreatic adenocarcinoma Digestive Surgery, Hospital Virgen de la Arrixaca, Murcia, Spain; an increasing prevalence along the time periods (17%, 25%, 38%, 42% and 71% at 6, 12, 18, 24 and 36 months, ous drainage alone. who underwent surgical resection were identifi ed in SEER- 5Programa de Recerca en Cancer, Institut de Recerca Hospital del Medicare (1991–2005). Claims for abdominal imaging (CT/ respectively). All cases except one were detected in patients AIM: To demonstrate the durability of the initial conclu- Mar (IMIM), Barcelona, Spain;6Histopathology Unit, London CTA, MRI/MRA, PET) ≤5 years after resection were analyzed. having an intrathoracic anastomosis. Seven of 10 NSCM sions as the cohort of DMD patients expanded. Methods Research Institute, Cancer Research UK, London, United Kingdom After initial screen, CT/CTA was used for longitudinal anal- segments were fi rst detected at 12–18 months postopera- A prospective database of patients undergoing DMD was yses. Univariate and multivariate analyses were performed BACKGROUND: The cardia type epithelium, a nonspe- tively. Immunohistochemistry detected nuclear expression analyzed and compared to patients who had standard per- by Kruskal-Wallis, logistic regression, and Cox. CT utiliza- cialized columnar type of metaplasia (NSCM), has been of pSMAD 1/5/8 in the squamous epithelium close to the cutaneous drainage from 2006 to the present time. tion was calculated by dividing total scans by months of proposed as an intermediate stage in the transdifferentia- anastomosis with strong expression in all epithelial cells of

RESULTS: 41 patients had undergone DMD with 39 com- tion process from normal squamous epithelium into the the NSCM areas. These results were confi rmed by WB anal- Abstracts

available postoperative data. Routine annual CT scanning Tuesday pleting therapy, defi ned as removal of percutaneous drains, was defi ned as at least one CT/CTA performed within each specialized intestinal type of differentiation present in ysis. Scattered nuclear expression of CDX2 was observed defi nitive surgery, or death. 43 patients underwent percuta- 12-month block, excluding year of death/censoring. To Barrett esophagus. Recent studies suggest that the BMP4 in 44/59 biopsies (75%) with NSCM. Two cases showed neous drainage alone. Patient characteristics including age, assess frequency of annual CT scanning in patients with pathway is involved in the transition of squamous epithe- isolated glands at 18 and 36 months that fully expressed sex, etiology of pancreatitis, and severity of pancreatitis superior survival, the top decile were further analyzed. lium into nonspecialized epithelium, and that CDX2 and CDX2 and co-expressed CDX1. BMP4 mRNA and CDX2 based on computed tomography severity index were indis- CDX1 transcription factors seem critical for acquisition of mRNA expression were signifi cantly greater in NSCM com- RESULTS: Within 5 years of pancreatic resection, 39316 stud- tinguishable between the two groups. Initial endoscopic the intestinal phenotype. Using an in vivo human model pared with squamous epithelium. ies were performed on 2792 patients. The majority of these access to the necroma was obtained by endoscopic ultra- of pathologic refl ux disease, we assessed prospectively the CONCLUSIONS: were CT scans (36521, 92.9%), and the remainder MRI (2425, In this human model of columnar meta- sound in 30 of the 41 patients undergoing DMD. The DMD development of NSCM and the involvement of the BMP4 6.2%) and PET (370, 0.9%). A minority received no imaging plasia, we identifi ed early activation of BMP4 in squamous cohort had shorter mean LOH (27 vs 55 days), time until pathway and CDX2 expression in this intermediate pheno- after resection, both when analyzing all patients (168/2792, epithelium and NSCM and observed that CDX2 but not removal of percutaneous drains (78 vs 188 days), fewer CT typic change. 6.0%), as well as the subset with >5 year-survival (11/265, CDX1 expression was detected in NSCM before the appear- scans (8.0 vs 14.3), drain studies (6.0 vs 13), and number of 4.2%). Mean monthly CT utilization per patient increased METHODS: Biopsy samples from the remnant esophagus ance of the intestinal phenotype. drains per patient (1.29 vs 2.0), all statistically signifi cant. from 0.4 in 1991 to 1.1 in 2005 (p < 0.0001), including the of patients having an esophagectomy with gastric preser- The DMD cohort also had fewer total ERCP’s (2.0 vs 2.6, immediate postoperative period. Overall mean utilization was p < 0.026). There have been 3 total deaths in the DMD 0.6 CT scans per month, but only 0.2 scans per month for group: 1 from MRSA pneumonia during therapy for inci- patients in the top survival decile. Among all patients, scans dentally found esophageal cancer 6 months after removal 10:30 AM – 11:45 AM were not evenly distributed over the years of the study: many of the percutaneous drain; 1 from congestive heart failure 2 S501bcd of the scans were clustered. Patients received a mean of 2.8 weeks after discharge from hospitalization for WOPN and scans within 3 months of surgery, 4.7 scans between 3 months QUICK SHOTS SESSION III electing hospice; and 1 patient with multi-system organ and 1 year, and 5.6 scans between 1 year and study termina- failure during treatment for DMD who was found to have tion. Among 1127 patients with suffi cient survival to allow for an occult pancreatic adenocarcinoma at autopsy. Three analysis, 569 (50.5%) received annual CT scans as previously surgical resection. Together with the value of bilirubin, the patients in the standard drainage cohort died with drains defi ned. Interestingly, only 81 (28.9%) of the top-performing /965 presence/absence of ascites and esophageal varices, and the in place while in the hospital undergoing therapy. No DMD patients received annual CT scans. Among all patients, post- Cholinesterases Predict Outcome in Patients rate of residual liver volume, which are our current param- patient needed surgery or had a pancreatico-cutaneous fi s- operative complications, non-white race, advanced stage, Undergoing Hepatic Resection for Hepatocellular eters to measure functional liver reserve, we sought to tula (PCF). and receipt of chemotherapy or radiation were predictive of Carcinoma: Results from a Retrospective Analysis investigate the value of preoperative cholinesterases (CHE) CONCLUSIONS: Compared to standard percutaneous receiving routine annual CT scans (p < 0.05). Routine annual in predicting postoperative adverse outcomes after hepatic Matteo Donadon1, Matteo M. Cimino1, Fabio Procopio1, drainage, DMD of WOPN reduces LOH and the use of CT scans were associated with negative rather than positive resection for HCC. Emanuela Morenghi 2 1 1 radiological and endoscopic resources. Surgery and PCFs impact on survival (HR for death 1.2, p = 0.02). Based on cur- , Angela Palmisano, Daniele Del Fabbro, METHODS: 1 1 We reviewed the records of 181 consecutive were avoided in patients undergoing DMD while single rent Medicare line-item payments, the bare minimum mean Marco Montorsi , Guido Torzilli 1 patients who underwent hepatic resection for HCC in digit mortality was maintained. imaging cost incurred per patient would be $3736, or $5287 Liver Surgery Unit, Third Department of Surgery, University of our Unit between 2001–2009. The value of preoperative over 5 years of survival. Milan, IRCCSIstituto Clinico Humanitas, Rozzano, Milan, Italy; CHE was analyzed against the occurrence of postopera- 2 CONCLUSIONS: Most patients undergo abdominal imag- Biostatistics Unit, University Of Milan IRCCS Istituto Clinico tive adverse events. Receiver-operator characteristic curve ing, usually CT, after pancreatic cancer resection. CT utili- Humanitas, Rozzano, Milan, Italy analysis was used to identify cut-off values of CHE that zation has increased in recent years. Administrative data BACKGROUND: Estimation of functional liver reserve predicted adverse outcomes. Univariate and multivariate from a large national database suggests that performance in patients with hepatocellular carcinoma (HCC) is of analyses on clinically relevant variables were performed. of routine annual CT scans after resection does not confer paramount importance to properly select candidates for P < 0.05 was considered statistically signifi cant. a survival benefi t.

46 47 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

RESULTS: Seventy-eight patients were enrolled from the pancreatic lesion was 3.2cm with a median CA19-9 of 10/1/2007–9/1/2010: 53 had a PD (median age 68 years; 351 ng/dl. At surgery, 553 (95%) pts had PBP and 30 (5%) /968 range 46–85) and 25 were inoperable or declined PD. patients had laparotomy+biopsy without further surgical Management and Perioperative Morbidity Among Signifi cant preoperative vulnerability was identifi ed in intervention. Among the 553 pts who had PBP, most (65%) Patients with Surgically Managed Pancreatic resected patients; VES-13 >3 in 24%; SPPB <10 in 49%; underwent double bypass, while a minority had either Adenocarcinoma: A Population-Based Analysis Fried’s exhaustion in 40%. The perioperative morbidity gastrojejunostomy (28%) or hepaticojejunostomy (7%) Using SEER-Medicare Data rate was 70% and three patients died (8.6%). Twenty-three alone. While the number of pancreaticoduodenectomies patients (43%) had severe complications (Clavien grade ≥ remained relatively stable over time, there was a temporal Skye C. Mayo1, Marta M. Gilson1, John L. Cameron1, III). Abnormal VES-13 showed predictive promise for post- decrease in the utilization of PBP (Figure). UnanticipatedHari Nathan 1, Joseph M. Herman2, Martin A. Makary1, operative admission to the SICU rs (53) = .34, p < .05, lon- locally advanced disease vs. liver/peritoneal metastasis Fredericas E. Eckhauser 1, Kenzo Hirose1, Barish H. Edil1, ger SICU stays rs (53) = .42, p < .01, and total hospital days the indication for PBP also changed over time (1996–2001:Michael A. Choti 1, Richard D. Schulick1, Christopher L. Wolfgang1, rs (53) = .26, p = .09. Self-reported exhaustion predicted 46% vs. 54%; 2002–2007: 49% vs. 51%; 2008–2010: 20%Timothy M. Pawlik 1 complications and post-operative admission to the SICU, vs. 80%, respectively)(P = 0.009). Following PBP, median 1Surgery, The Johns Hopkins Medical Institutions, Baltimore, MD; F2 (1, N = 53) = 4.1, p < .05. Older age was predictive of length of stay was 8 d and 37% pts had a complication 2Radiation Oncology, The Johns Hopkins Hospital, Baltimore, MD not being discharged to home rs (53) = –.35, p < .05 and (major: 14%) with a readmission rate of 18%. Postoperative INTRODUCTION: perioperative death, rs (53) = .31, p < .05. 22 patients have mortality after PBP was 2%. Few pts developed late obstruc- Surgical resection remains the only complete QoL data that shows a wide variance at baseline, tion (enteric: 3%; biliary: 4%) or needed a postoperative potentially curative therapeutic option for patients (pts) a signifi cant drop-off at 1 month and return to baseline by stent (enteric: 2%; biliary: 5%). Overall survival following with pancreatic adenocarcinoma (PAC). Over the last sev- 6 months. PBP was 6 months; pts with unsuspected metastatic disease eral decades, advances in surgical technique and periop- erative care have reduced perioperative mortality. Changes [ClinicalTrials.gov ID=NCT00883454] CONCLUSION: We identifi ed signifi cant vulnerability in as the indication for PBP had a signifi cantly worse survival compared with pts who had locally unresectable disease in perioperative management and morbidity associated RESULTS: Forty-nine (27%) of 181 patients had complica- older patients undergoing PD. Early analyses suggest that with PAC surgery, however, remain poorly characterized. comprehensive perioperative geriatric measures can predict (median survival: 5 vs. 8 months, respectively; HR = 1.43, tions, of which 36 (20%) were liver-related. Major morbid- P = 0.001). We sought to defi ne the utilization patterns of periopera-

post-surgical outcomes. If sought for and identifi ed, these Abstracts

ity occurred in 11 (6%) patients. The 30-day postoperative tive and operative procedures for pts with PAC, as well as Tuesday mortality was 1% (2 patients). A value of CHE ≤ 5,900 UI/L issues could be managed expectantly, leading to more accu- evaluate population-based temporal trends in morbidity had a sensitivity of 73% and a specifi city of 67% in predict- rate preoperative counseling and in-hospital care. and mortality. ing liver-related postoperative complications (P = 0.001). METHODS: 967 Using Surveillance, Epidemiology and End The multivariate analysis revealed that blood transfusion, Results (SEER)-Medicare linked data, we identifi ed 2461 pts major resections, and a value of CHE ≤ 5,900 UI/L indepen- Trends in the Palliative Surgical Management of Patients with PAC who underwent pancreatic resection from 1991- dently predicted the risk of morbidity. with Unresectable Pancreatic Adenocarcinoma: Lessons 2005. We collected data to assess morbidity, mortality, and CONCLUSIONS: The value of CHE contributed important Learned from a Large, Single Institution Experience survival while adjusting with the Elixhauser comorbidity information in predicting postoperative adverse outcomes Peter J. Kneuertz, Steven Cunningham, Sergio Lopez, index. Trends in preoperative comorbidity indices, periopera- after hepatic resection for HCC. Thus, it should be included John L. Cameron, Joseph M. Herman, Martin A. Makary, Frederic tive management, type of surgical procedures performed, as in the selection process of candidates to surgery for such E. Eckhauser, Kenzo Hirose, Barish H. Edil, well as changes in morbidity and mortality were examined. disease. Michael A. Choti, Richard D. Schulick, Christopher L. Wolfgang, RESULTS: Preoperative evaluation included ERCP (59%), Timothy M. Pawlik CT (92%), MRI (14%) and PET (2%) with a temporal 966 Department of Surgery, The Johns Hopkins University School of increase in the use of all four diagnostic modalities (all P < 0.005) (Table). Use of diagnostic laparoscopy increased Comprehensive Perioperative Geriatric Assessments Medicine, Baltimore, MD May Predict Surgical Outcomes in a Prospective Study from 3% to 16% (P < 0.001) over time. The proportion of INTRODUCTION: of Older Patients Undergoing Pancreaticoduodenectomy Traditionally, routine palliative bypass pts who underwent total pancreatectomy (n = 28; 1%) or (PBP) has been advocated for palliation of patients (pts) pancreaticoduodenectomy (n = 1945; 79%) did not change 1 2 1 CONCLUSION: Kevin K. Roggin, Joshua A. Hemmerich, Jeffrey B. Matthews, with pancreatic adenocarcinoma (PAC) who are explored Utilization of PBP for inoperable pan- (both P > 0.05) whereas distal pancreatectomy (n = 333; Mitchell C. Posner 1, William Dale2 with curative intent but have inoperable disease discovered creatic cancer has signifi cantly decreased over time. The 14%) increased over the study period (P = 0.04). There was 1Surgery, University of Chicago, Chicago, IL;2Internal Medicine, at the time of surgery. We examined trends in the relative indication for PBP has also changed over time, with the a temporal increase in median pt age (1991–1996: 71 yo; Section of Geriatrics and Palliative Care, University of Chicago, use of PBP over time with a particular emphasis on identi- discovery of unsuspected metastatic disease now being 1997–2000: 72 yo; 2001–2002: 73 yo; 2003–2005: 74 yo; Chicago, IL fying changes in surgical indications, type of bypass per- the main indication for PBP. While PBP remains effective, the P < 0.05) and number of pts with multiple preoperative formed, as well as perioperative outcomes associated with morbidity can be signifi cant and survival following PBP comorbidities (Elixhauser comorbidities ≥3: 1991–1996, INTRODUCTION: Older patients with pancreatic cancer surgical palliation. especially among pts with metastatic disease is extremely 10%; 1997–2000, 17%; 2001-2002, 25%; 2003-2005, 26% are often not offered pancreaticoduodenectomy (PD) due short. P < 0.001). Despite the increase in pt age and comorbidities to potentially high perioperative risk and prolonged post- METHODS: Between Jan 1996 and Jun 2010, 1913 pts with over time, overall perioperative morbidity (53%) did not operative recovery. These patients may have undetected PAC in the head of the pancreas were surgically explored. change during the study period (P = 0.97). The most com- vulnerabilitities related to frailty that may adversely affect Data regarding preoperative symptoms, intraoperative mon postoperative complications were bleeding and need surgical outcomes. fi ndings, type of surgical procedure performed, as well as for re-exploration, both of which decreased over time (9% METHODS: perioperative and long-term outcomes were collected and PD-eligible patient over 50 were enrolled onto analyzed using univariate and multivariate analyses. to 4% and 11% to 7%, respectively;both P < 0.05). In con- this IRB-approved prospective observational study. Extensive trast, there was a temporal increase in the number of percu- RESULTS: perioperative clinical information including Vulnerable Of the 1913 pts, 1330 (70%) underwent a pan- taneous interventional procedures (8% to 12%; P = 0.005). Elder Survey (VES-13), short physical performance battery creaticoduodenectomy, while 583 (30%) did not. Among Perioperative mortality decreased by half over the study (SPPB), and Fried’s Criteria for frailty were collected. Com- the 583 pts who did not undergo curative intent surgery, period(1991–1996: 6% vs. 2003-2005: 3%; P = 0.04). The plications were prospectively recorded and graded accord- most presented with either preoperative nausea/vomiting ing to established guidelines (Clavien scores). Quality of (18%) or jaundice (72%). The majority of pts had been eval- life (QoL) measures were taken at baseline and at 1, 3, and uated by CT scan (97%), while a minority had an MRI (5%) 6-months post-PD. or PET scan (1%). On preoperative imaging, median size of

48 49 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

overall 1- and 5-yr survival was 53% and 13%. A mod- sum tests. Univariate and multivariable associations with RESULTS: There were 20 patients who underwent TP- CONCLUSIONS: These preliminary results show that TP- est improvement in median survival (12 to 16 months; PRD were evaluated using logistic regression models and AITD. The median age was 42 years old (17–73) and 8 (40%) AIT signifi cantly improves pain, depression, anxiety and P = 0.01) was noted over the time periods examined. summarized with odds ratios and 95% confi dence intervals. were women. CP etiology was: idiopathic in 10 (50%), QOL measures in appropriately selected patients with CP. RESULTS: Of 583 patients undergoing DP, 90 patients with familiar in 5, alcohol in 4, autoimmune and gallstones Trends in Perioperative Staging, Morbidity, and Complicationspre-existing type I (n = 24) or type II (n = 66) diabetes were in 1 each. Prior pancreatic resection was performed in 2 971 1991–2005 for Patients with Resected Pancreatic Adenocarcinoma (10%) patients. The median islet cell equivalents infused excluded. The remaining 493 patients represent the study Acute Portomesenteric Venous Thrombosis Following 2 was 299,508. There was no mortality. Complications were 1991–96 1997–2000 2001–02 2003–05population with a mean age of 57 years and BMI (kg/m ) present in 8 (40%): bleeding in 4 (20%) due to anticoagu- Abdominal Surgery: Observe, Anticoagulate or Number of Patients 568 539 500 854 of 27.4. Operative indications included benign lesions in Operate? *Median Patient Age (years) 71 72 73 74 253 patients (51%) and malignancy in 240 (49%). Extent lation 2 of theses required a reoperation; gastroparesia in *CT 88% 89% 94% 94% of pancreatic resection was limited in 141 patients (37%), 3 (15%). Median LOS was 12 days. The median follow-up Michael J. Leonardi, Lindsay L. Hollander, Henry A. Pitt, Michael *MRI 2% 8% 19% 24% standard in 87 (23%), extended in 128 (33%), and subtotal was 16 months. The mean pre HbA1c was 6.03 and post G. House, Nicholas J. Zyromski, C. Max Schmidt, Attila Nakeeb, *PET <2% <2% <2% 4% in 30 (8%). Postoperatively, 179 patients (36%) required no Hb A1c was 7.72. The median post-op C-peptide levels Thomas J. Howard, Keith D. Lillemoe *ERCP 50% 61% 65% 60% postoperative treatment, 269 (55%) required perioperative were 0.78. Dose of Novolog were 12U (4–25) per day, Lan- Surgery, Indiana University School of Medicine, Indianapolis, IN *Diagnostic Laparoscopy 3% 8% 10% 16% tus 9.71U (5–20) per day, Humolog 7.5U (2–30).The QOL Pancreaticoduodenectomy 76% 82% 81% 78% hyperglycemic control, and 45 (9%) developed diabetes BACKGROUND: Portomesenteric venous thrombosis (PMVT) including 34 requiring insulin, and 11 oral hypoglycemic data are summarize in the table 1.The patients graded their *Distal Pancreatectomy 13% 10% 13% 16% abdominal pain pre and postoperative are as follows: none is a rare, but serious, complication of abdominal surgery Total Pancreatectomy 2% <2% <2% <2% medication. Patients developing PRD had higher mean to mild in [0 (0%) vs 15 (76%)], moderate in [9 (45%) vs 3 with no agreed standard of care. Management options *•3 Elixhauser 10% 17% 25% 26% preoperative glucose levels (98.2 vs. 112.1, p < 0.001), BMI Comorbidities 2 (14%)] and severe in [11 (55%) vs 2 (10%)].The DASS was include observation, anticoagulation, and thrombectomy. (27.3 vs. 29.1 kg/m , p=0.019), estimated blood loss (594 vs. Our study aims to characterize a large series of patients Overall Peri-operative 55% 50% 55% 53% 845 ml, p = 0.001), longer operative time (238 vs. 269 min, completed pre and post operative: Depression was catego- Complications with PMVT after abdominal surgery with a focus on man- p = 0.003), and greater pancreatic specimen length (9.3 vs. rized as follows: mild in [3 (14%) vs 2 (10%)], moderate in *Peri-operative 30-day 6% 4% 5% 3% [3 (14%) vs 1 (4%)], severe [4 (19%) vs 0], extremely-severe agement and outcomes. Mortality 10.9 cm, p = 0.024). Incidence of PRD correlated with CT METHODS: *Median Survival (months) 12 15 12 16 extent of pancreatic resection, occurring after limited resec- [3 (14%) vs 2 (10%)] (p = 0.012). Anxiety was categorized We performed a retrospective analysis of more *Signifi cant at P < 0.05 by test of trend. <2%: Per NCI data usage agreement, no as mild in [2 (10%) vs 1 (4%)], moderate in [4 (19%) vs than 4000 patients having abdominal surgery at an aca-

tion in 4.2%, standard in 4.6%, extended in 13.2%, and Abstracts cells with totals less than 11 were reported. subtotal in 30%. On multivariate analysis, factors associ- 1 (4%)], severe in [1 (4%) vs 1 (4%)], extremely-severe [4 demic medical center between January 2007 and August Tuesday CT = computed tomography; MRI = magnetic resonance imaging; PET = positronated with increased risk of PRD included preoperative glu- (19%) vs 2 (10%)] (p = 0.062). PDI was completed both pre 2010. Patients with postoperative thrombosis of the portal, emission tomography; ERCP = endoscopic retrograde cholangiopancreatography. cose ≥126 (OR 10.06, p < 0.001), BMI 5 unit increase (OR and post operative: 1.) Family/home responsibilities: Severe superior mesenteric, and/or splenic veins were reviewed. 1.6, p = 0.011), extended (OR 3.37, p = 0.031) or subtotal in [12 (61%) vs 2 (10%)]. 2.) Recreation: Severe in [16 (80%) Transplant patients and those with preexisting PMVT were resection (OR 9.86, p < 0.001), and blood transfusion (OR vs 4 (20%)]. 3.) Social Activity: Severe in [13 (66%) vs 3 excluded. The diagnosis was established by computed 2.38, p + 0.037). (15%)]. 4.) Ocupation: Severe in [14 (70%) vs 3 (20%)]. 5.) tomography (CT), magnetic resonance imaging (MRI), CONCLUSIONS: Sexual Behavior: Severe in [11 (55%) vs 2 (10%)]. 6.) Self and/or duplex ultrasound. Mortality associated with pancreatic CONCLUSION: Postresection diabetes occurs in approxi- resection for PAC has decreased by one-half. Despite surgi- care: Severe in [6 (30%) vs 0 (0%)] patients. 7.) Life support RESULTS: mately 9% of patients after DP. Factors associated with activities: Severe in [9 (45%) vs 1 (10%)]. Alcohol and drug Forty-four patients had PMVT (23 isolated por- cal resection for PAC being offered to older pts with more increased risk include elevated preoperative glucose, BMI, tal, 19 combined portal and mesenteric and/or splenic, 1 preoperative comorbidities, the incidence of perioperative abuse was diagnosed in 3 patients post-operative. Passive extent of resection, and need for blood transfusion. Three suicidal ideation was present in 2 patients. isolated mesenteric, 1 isolated splenic). Average patient age complications remained stable. Resection for PAC in an of these factors are easily attained preoperatively by clini- was 59 years, and 55% were male. Four patients (9%) were aging population with more medical comorbidities can be on preoperative anticoagulation that was held for surgery, cal evaluation and expected pancreatic remnant based on Table 1: performed safely, however further progress is necessary to CT imaging. These fi ndings will help counsel patients pre and 25 patients (57%) had a malignancy. Operations per- decrease morbidity. and postoperatively about the risks of postresection diabe- PreOp PostOp p Value formed included pancreas resection (21), liver resection PDI (5), colorectal resection (5), pancreas debridement (4), sple- 969 tes after DP. Family Home responsabilities nectomy (3), major biliary surgery (2), combined pancreas Severe 12 (61%) 2 (10%) Postresection Diabetes After Distal Pancreatectomy: 970 and colon resection (2), palliative gastrojejunostomy (1), Incidence and Risk Factors Recreation gastrectomy (1), and abdominal wall reconstruction (1). Assessment of Quality of Life Following Total Severe 16 (80%) 4 (20%) Michael J. Ferrara, Christine Lohse, Suresh T. Chari, Median time from operation to PMVT diagnosis was 14 Pancreatectomy and Islet Cells Autotransplant for Social Activity Santhi Swaroop Vege, Florencia G. Que, Kaye M. Reid Lombardo, days, and 23 patients were diagnosed after discharge from Chronic Pancreatitis Severe 13 (66%) 3 (15%) John H. Donohue, David M. Nagorney, Michael B. Farnell, Ocupation the original hospitalization. Diagnostic modalities included Michael L. Kendrick Juan R. Aguilar-Saavedra, Greg Lentz, Judith Scheman, Severe 14 (70%) 3 (20%) computed tomography (35), duplex ultrasound (8), and Tyler Stevens, Matthew Walsh Sexual Behavior magnetic resonance imaging (1). Treatment included Mayo Clinic, Rochester, MN Severe 11 (55%) 2 (10%) observation in 15 patients, anticoagulation in 24 patients Surgery, Cleveland Clinic, Cleveland, OH BACKGROUND: Diabetes is a known, but poorly under- Self care (20 continued as outpatients), and operative thrombec- BACKGROUND: stood sequela of pancreatic resection in a subset of patients. The primary goal in TP-IAT is to improve Severe 6 (30%) 0% tomy in 5 patients. All patients who underwent operative Our aim was to defi ne the incidence and risk factors of QOL by alleviating pain and discontinue narcotics while Life support activity thrombectomy developed PMVT on postoperative day 1 Severe 9 (45%) 1 (10%) postresection diabetes (PRD) after distal pancreatectomy (DP). preventing or minimizing surgical diabetes. QOL outcomes after their initial operation, and four of these patients had following any surgery for CP are lacking. DASS a portomesenteric venous resection and reconstruction as METHODS: A retrospective review of all consecutive Depression p = 0.012 OBJECTIVE: part of their initial operation. PMVT-induced liver abscess patients undergoing DP at our institution from January Review the results of TP-AIT by assessing Severe 4 (19%) 0% occurred in 1 patient. Small bowel and/or colon resec- 2004 through July 2010. Data was obtained from medi- QOL and beta cell function. Anxiety p = 0.062 tion for ischemia occurred in 2 patients. Seventeen (40%) cal records, pathology reports, and postoperative imaging. METHODS: Severe 1 (4%) 1 (4%) A retrospective review of a prospective data- Pain Scale patients were readmitted within 30 days of discharge from Postoperative CT scans were evaluated to determine extent base for TP-IAT from November 2007 through October Severe 11 (55%) 2 (10%) the PMVT admission. Only one patient died fi ve days after of resection, defi ned by site of transection as limited (adre- 2010. A standardized pre and post-operative using the a liver resection complicated by portal vein thrombosis nal or left), standard (superior mesenteric artery or left to Depression Anxiety Stress Scale (DASS) and the Pain Dis- requiring operative thrombectomy and enterectomy. adrenal), extended (neck), subtotal (into head). Compari- ability Index (PDI). A visual analogue pain scale was used sons between patients with and without PRD were evalu- to asses global pain. Diabetes assessment by measuring Hb ated using chi-square, Fisher’s exact, and Wilcoxon rank 1Ac pre and post surgery, and C-Peptide postoperative.

50 51 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

CONCLUSIONS: Portomesenteric venous thrombosis (PMVT) CONCLUSIONS: Pancreatic CT enhancement characteris- The 5-year OS and DFS rates were 37% and 56%, respec- had reduction of their intussusception followed by resec- is an uncommon complication following abdominal sur- tics quantifi ed by the L/E ratio can accurately assess the tively, for patients with extra-ampullary adenocarcinomas, tion and eight (6.7%) patients underwent reduction only. gery which may present early postoperatively, but often histological grade of pancreatic fi brosis and are a power- and 38% and 55%, respectively, for patients with periam- There were 7 (5.8%) negative surgical explorations. Of the presents after discharge. While the morbidity associated ful tool to predict the risk of developing a clinically-rel- pullary tumors. Advanced T stage and pathologic grade 3 or 66 idiopathic cases, 36 patients had spontaneous reduction with PMVT is high, the mortality is low (2%). Management evant PAF following PD. L/E ratio and pancreatic fi brosis 4 were associated with decreased survival. Nodal status and of the intussusception. Twenty four of the remaining 30 should be tailored to individual patient characteristics with are further associated with pancreatic texture, pancreatic margin status had no signifi cant impact on overall survival. patients were successfully managed nonoperatively and six respect to timing and severity of presentation. duct size, and the presence of diabetes. This information There was no difference in survival between periampullary had a negative surgical exploration. may be helpful in guiding pre- and postoperative patient and extra-ampullary duodenal tumors. CONCLUSION: In the current era of modern imaging the 972 management. CONCLUSION: Duodenal adenocarcinoma is a rare and incidence of adult intussusception has increased 4-fold. Preoperative Assessment of Pancreatic Fibrosis and found in the second portion of the duodenum in 59% of While idiopathic intussusception has become more com- Risk of Pancreatic Anastomotic Failure Following /973 patients. Advanced T stage and pathologic grade is associ- mon the majority of adult intussusception cases are still Pancreatoduodenectomy by Dual-Phase Computed Fifteen-Year Single Institution Experience with Surgical ated with poor survival. Extra-ampullary and periampul- associated with a pathologic lead point which in many Tomography Treatment of Duodenal Carcinoma: A Comparison lary carcinomas have similar survival following resection. cases is malignant. Based on our experience all patients Aggressive surgical resection remains the mainstay of with obstructive symptoms, hematochezia or palpable 1 1 2 of Periampullary and Extra-Ampullary Duodenal Yasushi Hashimoto, Guido M. Sclabas, Naoki Takahashi, Adenocarcinomas treatment. mass associated with intussusception should undergo pri- Yujiro Kirihara 1, Thomas C. Smyrk3, Marianne Huebner4, mary surgical resection without prior reduction to avoid Michael B. Farnell 1 Edwin O. Onkendi, Sarah Y. Boostrom, Michael G. Sarr, 974 unnecessary manipulation of a potential neoplasm. 1Surgery, Mayo Clinic, Rochester, MN;2Radiology, Mayo Clinic, Michael B. Farnell, David M. Nagorney, John H. Donohue, Adult Intussusception in the Last 25 Years of Modern Rochester, MN; 3Pathology, Mayo Clinic, Rochester, MN;4Health Michael L. Kendrick, Kaye M. Reid Lombardo, 975 William S. Harmsen, Florencia G. Que Imaging: Is Surgery Still Indicated Sciences Research, Mayo Clinic, Rochester, MN 1 1 1 Under-Utilization of Surgical Resection for Gastric BACKGROUND: Surgery, Mayo Clinic, Rochester, MN John H. Donohue, Travis Grotz, Edwin O. Onkendi, Delayed or decreased enhancement char- 2 Cancer in the Era of Multi-Modality Therapy BACKGROUND: Joseph A. Murray acteristics on computed tomography (CT) in patients with Duodenal adenocarcinoma constitutes 1 2 1 1 1 Surgery, Mayo Clinic, Rochester, MN;Gastroenterology, Mayo Lee J. McGhan, Barbara A. Pockaj, Richard J. Gray, Abstracts 44-55% of small bowel malignancies. Despite resection, the Tuesday pancreatic fi brosis have been described. However, studies Sanjay P. Bagaria 2, Nabil Wasif1 comparing clinical outcomes following pancreatoduode- disease-free survival (DFS) and overall survival (OS) remains Clinic, Rochester, MN 1General Surgery, Mayo Clinic Arizona, Scottsdale, AZ;2General nectomy (PD) to preoperative CT enhancement character- poor and is estimated to range from 15–53% in node- posi- BACKGROUND: Intussusception in adults, unlike in chil- istics are lacking. tive and 38–83% in node-negative at 5 years. Our aim was dren, is rare and commonly reported to be due to malig-Surgery, Mayo Clinic Florida, Jacksonville, FL to describe the surgical management of patients with duo- INTRODUCTION: AIM: To study the ability of dual-phase CT to assess pan- nancy. Adult colonic intussusception is associated with Surgical resection of gastric adenocar- denal adenocarcinomas. creatic fi brosis and to predict the risk of developing a pan- primary carcinoma in 65–70% of cases while adult small cinoma is the cornerstone of potentially curative therapy. creatic anastomotic failure (PAF) following PD. METHODS: We reviewed, retrospectively, the medical bowel intussusceptions are associated with malignancy in We studied the recent utilization of surgical resection for records of 124 patients treated for duodenal carcinoma by only 30–35% of cases, 70% of these lesions are metastatic. patients diagnosed with gastric adenocarcinoma in the METHODS: Review of 157 consecutive patients with preop- surgical palliation or curative resection from January 1994 As a result, most authors recommend surgical exploration United States in the era of multi-modality therapy. erative dual-phase CT between 2004 and 2009. Pancreatic to January 2009 at our institution. for diagnosis and therapy. In light of the current availabil- METHODS: A retrospective review of all patients with gas- CT attenuation upstream from the tumor was measured in ity of CT scans, we questioned the role of mandatory surgi- the pancreatic and hepatic imaging phases. The ratio of the RESULTS: 124 consecutive patients (75 males/49 females) tric adenocarcinoma diagnosed between 2004–2007 in the cal exploration for all adult intussusceptions. Surveillance, Epidemiology and End Results (SEER) cancer mean CT attenuation value of the pancreas in the hepatic underwent surgical treatment for duodenal adenocarci- METHODS: (late) and pancreatic (early) phase was calculated; L/E ratio noma between 1/1994 and 1/2009. The mean age at diag- A retrospective review of all records of adults registry was conducted. Utilization and extent of surgery and histological grade of pancreatic fi brosis assessed by nosis was 65 years (range 33–87 years). 110 patients (89%) treated at a large tertiary referral center for intussusception was reviewed according to stage at presentation, and fac- Masson trichrome immunostaining were then correlated were diagnosed by endoscopic biopsy and 14 patients by from 1983 to 2008 was performed. A comparative analysis tors associated with receipt of surgery were identifi ed. with the development of a clinically-relevant PAF and CT or surgical exploration. There were 25 periampullary of the medical history, imaging, operative and pathological RESULTS: A total of 18,815 patients diagnosed with gas- other clinical parameters. tumors and 99 extrampullary tumors. Eight patients had reports was performed. tric adenocarcinoma were identifi ed, of whom 47% under- adenocarcinoma in the fi rst portion of the duodenum (D1), RESULTS: RESULTS: A clinically-relevant PAF was found in 21 A total of 200 adult patients had intussuscep- went a surgical resection (defi ned as any procedure more 73 in the second portion (D2), 23 in the third portion (D3), patients (13.4%) with a morbidity rate of 39.5%. No post- tion over the 25-year study period. A total of 137 (68.5%) extensive than a simple local excision). Excluding stage IV and 15 in the fourth portion (D4). Two patients had tumors operative death was observed. The PAF group showed patients had enteric intussusception, 18 (9%) had ileocecal, patients, 6651/8848 patients (75%) underwent a cancer- at the junction of D1 and D2, one patient had tumors at maximum enhancement in the pancreatic and washout 15 (7.5%) had ileocolic and 30 (15%) patients had coloco- directed surgical resection (SR group). Table 1 shows the the junction of D2 and D3 and two at the junction of D3 in the hepatic CT phase, while the control group showed lic intussusception. CT scan was the most common imag- distribution of SR by AJCC stage at presentation for all and D4. The average tumor size was 4.1 cm (range 0.2–10 a delayed enhancement pattern. Degree of pancreatic ing modality (58.5%). A discrete pathologic process was patients. Patients in the SR group were younger than those cm). Seventy nine (64%) patients had grade 3 tumors, fi brosis and L/E ratio were signifi cantly lower for the PAF present in 134 patients (67%) of all intussusceptions. The not undergoing surgical resection (NSR group): mean age 27 patients (22%) had grade 2 and fi fteen patients (12%) group than the control group (21.0 ± 17.9 vs. 40.4 ± 29.8; remaining 66 patients (33%) were idiopathic. Neoplasms 68 ± 0.16 vs. 74 ± 0.29, respectively (p < 0.001). Patients had grade 4 tumors. There were no grade 1 tumors. Three P < 0.0001 and 0.86 ± 0.14 vs. 1.09 ± 0.24; P < 0.0001); were the most common etiology of intussusception, there in the SR group had a higher distribution of distal tumors patients had metastatic disease that was confi rmed at oper- fewer PAF patients showed an atrophic histological pattern were 25 primary (12.5%), 20 metastatic (10%), and 48 vs. proximal tumors compared to the NSR group (84% vs. ation. Nine patients had stage 1, 26 patients had stage 2, (14% vs. 39%; P = 0.046). The L/E ratio was positively cor- (24%) benign neoplasms identifi ed. Adhesions from prior 69%, p < 0.001) and were less likely to have T4 tumors (2% 33 had stage 3, and 56 had stage 4 disease. Eighty percent related with pancreatic fi brosis. Pancreatic fi brosis and L/E surgery (12), celiac disease (8), infl ammatory bowel dis- vs. 11%, p < 0.001). The majority of the 2197 patients in of patients underwent curative resection. Thirty four (34%) ratio increased with larger duct size (P < 0.001), the pres- ease (7), infection (7), Meckel’s diverticulum (6), and a gall the NSR group (65%) were not recommended any surgical patients had pyloric preserving Whipple resection, 33 ence of diabetes (P < 0.05), and the surgeon’s assessment of stone (1) accounted for the remaining 41 cases. Obstructive resection and surgery was contra-indicated due to other co- (33%) patients underwent a standard Whipple resection, pancreas fi rmness (P < 0.001). In multivariate analyses L/E symptoms, hematochezia and palpable mass were clinical morbid conditions in only 9%. The overall disease-specifi c 28 (28%) patients had segmental resection, 25 (20%) had ratio and BMI were signifi cant predictors for the develop- indicators present in 87% of patients with a neoplastic lead survival (DSS) at three years (excluding stage IV patients) palliative bypass surgery and 3 (3%) patients underwent ment of a clinically-relevant PAF; a 0.1 unit increase of L/E point. One hundred and twenty patients underwent opera- was 39% in the NSR group compared with 62% in the SR Whipple with en bloc resection. Six patients had positive ratio decreased the odds of a PAF by 54%. tive treatment for intussusception. Overall, 99 (82.5%) group (p < 0.001). For stage IA patients only, the three-year margins, 93 had negative margins. Fifty-one patients had patients underwent primary resection of the intussuscep- DSS for the NSR group was 68% compared with 94% in the node-positive disease and 49 had node-negative disease. tion without prior reduction, thirteen (10.8%) patients SR group (p < 0.001).

52 53 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

TABLE 1: Distribution of Extent of Surgical Resection by Tumor StageMETHODS: Novice robotic trainees (n = 10, PGY3-5 resi- SSAT POSTER SESSION ABSTRACTS dents) from 3 departments (General Surgery n = 5, Urology n = 3, Gynecology n = 2) participated over 2 months in 1) Printed as submitted by the authors. online didactics, 2) hands-on tutorial, and 3) self-practice indicates a Poster of Distinction. (individually scheduled sessions) using 9 inanimate exer- cises until profi ciency levels were achieved. 5 exercises used FLS models with modifi cations (Peg, Clutch/Camera Peg, Sunday, May 8, 2011 Pattern Cut, and Interrupted and Running Suture) and 4 used custom made components (Rubber Band, Stair Rub- ber Band, Clutch/Camera, and Cutting Rubber Band). Each task was scored for time and errors using modifi ed FLS met- Authors available at their posters to answer questions 12:00 PM – 2:00 PM; posters on display 8:00 AM – 5:00 PM. rics; task scores were normalized to the profi ciency level and a composite score equaled the sum of the 9 normal- ized task scores. Each trainee completed a questionnaire 12:00 PM – 2:00 PM and mobilize myeloid-derived suppressor cells (MDSCs), (5-point Likert scales) and performed a single proctored Hall A which are known to directly suppress anti-tumor immu- repetition of each task before (pre-test) and after (post-test) nity. Recently, our lab has shown that deletion of 5LO pro- training. Training was supervised by a designated proctor POSTER SESSION I foundly attenuates polyp development in the APC∆468 and feedback was given upon request. Comparisons used (NON-CME) murine model of polyposis. Moreover, deletion of 5LO Mann-Whitney tests; p < 0.05 was considered signifi cant. in the bone marrow of mice mitigated the recruitment of RESULTS: MDSCs to the spleen, mesenteric lymph nodes, and tumor While trainees felt moderately to very com- site and reduced arginase-1 activity. In this study, we inves- fortable with their laparoscopic skills (3.5 ± 0.5), they had tigate the effect of the 5LO deletion on the tumor infl am- CONCLUSIONS: In the modern era of cancer care a signifi - minimal prior robotic experience and poor comfort with matory microenvironment and polyposis. robotic skills (1.9 ± 1.1) at baseline. All 10 trainees com- cant proportion of patients in the United States with early- Basic: Colon-Rectal METHODS: pleted all curricular components in 9.5 to 12.0 hours. All APC∆468/5LO+/+ mice were crossed with stage gastric adenocarcinoma are not being offered surgical trainees reached profi ciency for all 9 exercises after com- 5LO-/- mice to generate an APC∆468/5LO−/− model. All resection (37% of stage IA patients in our study), despite pleting an average of 69.3 ± 18.4 repetitions over 4.4 ± 0.9 Su1837 mice were of the C57BL/B6 genetic background. Offspring a clear survival advantage demonstrated following sur- hours; practice was completed during 3-5 self-practice ses- were genotyped by PCR and sacrifi ced at 4 months of age. gery. Further study is required to explain this discrepancy. sions with moderate to extensive (4.6 ± 0.7) proctor feed- 5-Lipoxygenase Defi ciency Diminishes Th2 Cell, APC∆468/5LO+/+ mice were compared to APC∆468/5LO-/- Patient and physician education, national guidelines and back provided, which was rated as highly appropriate (4.9 Regulatory T-Cell and Dendritic Cell Infi ltration of mice using immunofl ourescense for Th1 cells (CD8), Th2 perhaps regionalization of care may be needed to improve ± 0.4). Profi ciency levels were rated as moderately diffi cult Murine Intestinal Polyps cells (CD4), regulatory T-cells (CD4/Foxp3), macrophages (Mac3 and Gr1), and dendritic cells (CD11c and IAb). patient outcomes. (2.9 ± 0.4), highly appropriate (4.1 ± 0.2), and provided Joseph D. Phillips*1, Mohammad W. Khan3, Eric C. Cheon2,

RESULTS: Poster Abstracts excellent feedback (100% agreement amongst trainees). Paul Grippo 1, Khashayarsha Khazaie3, David J. Bentrem1,4 No statistically signifi cant difference was observed in Performance improved signifi cantly after training for all 9 1 976 Surgery, Northwestern University, Feinberg School of Medicine, the infi ltration of CD8+ cells between the APC∆468/5LO+/+

tasks and according to the composite scores (644.6 ± 205 Sunday Chicago, IL;2 Surgery, Mount Sinai Hospital, Chicago, IL; and APC∆468/5LO-/- groups. However, CD4+, CD4+FoxP3+, Profi ciency-Based Training for Robotic Surgery: vs. 942.5 ± 53, p < 0.001). 100% of trainees indicated that and CD11c+IAb+ cell infi ltration was signifi cantly reduced 3Gastroenterology, Northwestern University, Feinberg School of Feasibility and Inital Results their robotic skills improved and self-rated comfort level (p < 0.05) in the polyps of 5LO-defi cient mice. Medicine, Chicago, IL;4Surgery, Jesse Brown VA Medical Center, improved from 1.9 ± 1.1 to 3.3 ± 1.0 (p = 0.002). CONCLUSION: Genevieve Dulan, Robert V. Rege, Deborah C. Hogg, Chicago, IL Deletion of 5LO results in decreased Kristine M. Gilberg-Fisher, Seifu T. Tesfay, Daniel J. Scott CONCLUSION: These data indicate that initial imple- colorectal polyposis and reduced infi ltration of infl amma- BACKGROUND: Surgery, UT Southwestern, Dallas, TX mentation of the comprehensive robotic training curricu- 5-Lipoxygenase (5LO) is an essential tory cells, including mast cells, MDSCs, Th2 effector cells, lum was feasible with all participants completing training. enzyme in arachidonic acid metabolism. Overexpression of INTRODUCTION: We previously developed a compre- regulatory T-cells, and dendritic cells. These results provide Moreover, signifi cant performance improvement was docu- 5LO has been demonstrated in a variety of cancers, includ- further evidence for the infl ammatory and tumorigenic role hensive, profi ciency-based robotic training curriculum mented after a modest amount of practice with all trainees ing colon cancer, as well as adenomatous polyps. As such, (online didactics, ½ day tutorial, and 9 inanimate exer- of 5LO and indicate that inhibition of 5LO may provide an achieving profi ciency. Further adoption of this curriculum it is a potential therapeutic target. Direct inhibition of 5LO effective treatment for colorectal polyposis and cancer. cises)that aimed to address 23 unique skills identifi ed via is encouraged. resulted in decreased proliferation in colon cancer cell lines task deconstruction of robotic operations. Prior studies and inhibited tumor growth in a colon cancer xenograft supported construct, content, and face validity. The pur- model. Furthermore, mast cells in particular utilize 5LO pose of this study was to assess feasibility and educational to potentiate the proliferation of intestinal epithelial cells outcomes following initial implementation.

54 55 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1838 Su1839 Su1840 The Oncogenetic microRNA 21 and TumorsuppressiveIntra-Gastric Administration of a Japanese Herbal Searching for Keys: Unlocking Mesenchymal Stem microRNAs 143 and 145 as Response Predictors for Medicine, Dai-Kenchu-To, Stimulates Colonic Motility Cells Homing Potentials After Acute Anal Sphincter Multimodal Treatment of the Locally Advanced Rectalvia Transient Receptor Potential Cation Channel, Injury Carcinoma Subfamily V, Member 1 (TRPV1) Receptors in Massarat Zutshi*1,2, Levilester Salcedo1, Margot S. Damaser2, Conscious Dogs Peter P. Grimminger*1, Uta Drebber2, Elfriede Bollschweiler1, Marc Penn 3 1 Jan Brabender 1, Ralf Metzger1, Arnulf H. HöLscher1, Chikashi Shibata*, Daisuke Kikuchi, Fumie Ikezawa, Colorectal Surgery, Cleveland Clinic Foundation, Cleveland, OH; Hakan Alakus 1, Magarethe Odenthal2, Daniel Vallbohmer1 Hirofumi Imoto, Iwao Sasaki 2Biomedical Engineering, Cleveland Clinic Foundation, Cleveland, OH; 1 General-, Visceral- and Tumorsurgery, University Clinic Cologne, Tohoku University School of Medicine, Sendai, Japan 3Cell biology and Regenerative Medicine, Cleveland Clinic Foundation, Cologne, Germany;2 Cleveland, OH Pathology, University Clinic Cologne, Cologne, A Japanese herbal medicine, Dai-Kenchu-To (DKT) is Germany clinically effective in adhesive and paralytic intestinal BACKGROUND: MCP-3 and SDF-1 upregulation after INTRODUCTION: Multimodal treatment strategies obstruction. We previously reported that intra-gastric (IG) has been demonstrated in our prelim- have been developed to improve the prognosis of locally administration of capsaicin, a major pungent ingredient of inary study. MCP-3 was upregulated more than SDF-1. In advanced rectal carcinoma patients. Predominantly chili and known to bind the transient receptor potential order to progress towards studies to re-establish homing in patients with histopathological response seem to benefi t cation channel, subfamily V, member 1 (TRPV1), stimulated order to direct stem cells at a time remote from the injury, from this treatment. Therefore molecular markers which colonic motility and defecation in dogs. DKT contains sev- it is important to study if other cytokines are responsible help to identify responders prior treatment are needed to eral pungent ingredients including dried ginger rhizome for homing. We hypothesize that specifi c chemokines are individualize treatment strategies. In the presented study and zanthoxylum fruit. Aim of the present study was to expressed at the time of anal sphincter injury, which can be we have analyzed the oncogenetic microRNA 21 and investigate the effect of IG DKT on colonic motility and homing factors for mesenchymal stem cells (MSC). We aim tumorsupressive microRNA 143 and 145 prior and after role of TRPV1 in DKT-induced colonic contractions. to identify the cytokines that are upregulated that could be stem cell homing factors in the anal sphincter after acute neoadjuvant radiochemotherapy to evaluate the poten- METHODS: Five mongrel dogs were used equipped with injury. tial predictive strength of those markers for neoadjuvant strain gauge force transducers at the proximal, middle, and treatment. distal colon to measure circular muscle activity. Colonic METHOD: Fifteen rats were grouped into: control (no METHODS: motility was studied after IG administration of DKT at the injury; C = 3), 24 hrs sham, (S-24 n = 3), 10 days sham 40 patients (24 male, 16 female; median age CONCLUSION: Certain cytokines apart from MCP-3 and 61 years) with locally advanced rectal carcinoma were dose of 2.5, 5 g dissolved in 70 ml saline via a catheter (S-10d n = 3), 24 hrs partial sphincter excision group (PSE- SDF-1 are upregulated and are downregulated on the 10th included in this study. All patients were treated with radio- placed in the proximal stomach in the interdigestive state. 24 n = 3) and 10 days partial sphincter excision (PSE-10d) chemotherapy (50.4 Gy, 5-FU) prior curative surgery. The To study the role of cholinergic and serotonergic recep- group. Sham injury was done by pressing a cue tip on the day. These can be further pursued to isolate key homing regression grade was defi ned as follows: Major responders tors in DKT-induced contractions in the colon, saline as anal sphincter. Partial anal sphincter excision was done cytokines. These cytokines are being investigated by an in- <10% and minor responders > 10% vital tumor cells in the control, muscarinic antagonist atropine (0.1 mg/kg), nico- by excising 25% of the anal sphincter. The anal sphincter vitro migration assay to establish if they can induce hom- histopathological evaluation. The intratumoral miRNA tinic antagonist hexamthonium (5 mg/kg), or serotonin-3 was isolated 24 hours, and 10 days after injury. Chemo- ing of mesenchymal stem cells. This will facilitate a better Poster Abstracts was isolated from the pretreatment biopsies and the cor- antagonist ondansetron (1 mg/kg), was injected intrave- kine expression was assessed by gene array analysis. A list understanding to be able to target cellular therapy in the responding surgically resected tumor tissue. The microRNA nously 5 min before IG administration of DKT (5 g). To of chemokines that were upregulated after anal sphincter future. Sunday expression was measured using RT-PCR and correlated to study the role of TRPV1, 70 ml saline as control or cap- injury compared to the non injury and sham injury group. clinicopathological parameters including histopathologi- sazepine (10 mg) dissolved in 70 ml saline was adminis- Statistical analysis was performed. cal regression data. tered into the stomach 5 min before IG administration of RESULTS: 84 chemokines were investigated. Compar- DKT (5 g). Area under the contractile curve for 30 minutes RESULTS: The pre- and post-treatment microRNA 21 and ing, C vs. PSE-24 44% (37 of 84) genes were signifi cantly after IG DKT administration was measured and expressed 143 expression did not correlate with the clinicopatho- upregulated (P < 0.05) while comparison of S-24 vs. PSE-24 as motility index (MI). logical parameters or histopathological regression. The showed 19% signifi cant upregulation (P < 0.05). Ten days post-treatment expression of microRNA 145 showed a RESULTS: IG DKT induced colonic contractions immedi- after injury, comparison of C vs. PSE-10d showed 5% signif- signifi cant difference to the pre-treatment expression (p < ately after the administration, and MI increased in a dose- icant upregulation (5 of 84) while S-10 vs. PSE-24 showed 0.001). During the neoadjuvant treatment the intratumoral dependent manner along the entire colon. Duration and 4.7% signifi cant upregulation (4 of 84). The chemokines microRNA 145 expression was signifi cantly upregulated. A amplitude of IG DKT-induced colonic contractions did not which had the highest fold expression over control 24 low microRNA 145 expression in the post-treatment tissue differ from spontaneously occurring colonic contractions. hours after injury were CCl3, CxCl6, Trem1 CCl4 and CCr1 was associated with a worse histomorphological regression Pretreatment with atropine and hexamthonium abolished and CCr2. (Graph) (p = 0.04). these IG DKT-induced colonic contractions, and MI was decreased along the entire colon compared to control (p < CONCLUSION: The microRNA 145 expression seems 0.05). Ondansetron also inhibited IG DKT-induced colonic to be upregulated by the neoadjuvant radiochemo- contractions and MI was decreased in the proximal, middle, therapy, however a low micro RNA 145 expression after and distal colon compared to control (p < 0.05). Capsaz- neoadjuvant treatment was associated with a worse his- epine inhibited IG DKT-induced colonic contractions, and topathological response. Strategies which modulate the MI at the proximal, middle, and distal colon was decreased microRNA 145 expression or the microRNA specifi c effect compared to control (p < 0.05). may be able to improve the effectiveness of neoadjuvant radiochemotherapy. CONCLUSIONS: These results indicate IG administration of DKT stimulates colonic motility via TRPV1, cholinergic, and serotonergic receptors. These effects of DKT might be a mechanism to ameliorate intestinal obstruction.

56 57 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

RESULTS: Su1841 Rat % Adhesion to Rabbit % Sidewall Defect Human NOD2 protein was detected by Western only patients with a major histopathologic response to blot and PCR in the NOD2 and 3020InsC cell lines, but not neoadjuvant therapy seem to have a signifi cant survival A Sprayable Hyaluronate/Carboxymethylcellulose Application Buttons Covered by Adhesions Control (n = 20) 80 ± 4 27.5 ± 9.0 the parental cell line. There was a signifi cant increase in benefi t. We have shown in a recent pilot study using micro- Based Adhesion Barrier Reduces Remote IntraabdominalSF Direct (n = 10) 47 ± 7* 0.4 ± 0.3* proliferation in the NOD2 cell line compared to the paren- array-technique that the expression profi le of microRNAs Adhesion Formation and Does Not Impair IntestinalSF Remote (n = 10) 83 ± 7 23.1 ± 6.7 tal cell line (p < 0.05). There was no signifi cant change in depends signifi cantly on the histopathologic response of Healing SS Direct (n = 10) 30 ± 10* 4.4 ± 2.9* proliferation in the 3020InsC cell line compared to parental patients with locally advanced esophageal cancer undergo- SS Remote (n = 10) 43 ± 12* 8.1 ± 4.4 and NOD2 cell lines. The fi gure shows monolayer wound ing multimodality treatment. This study aimed to validate Holly K. Sheldon*1, Melanie L. Gainsbury1, Michael R. Cassidy1, SS: Sepraspray; SF: Seprafi lm. Data are mean ± SEM. *P < 0.05 compared with healing for the three cell lines at 24 hours. The wounds in these identifi ed single microRNAs by real-time PCR. M. Jude Colt 2 2 2 control , Rubina L. Corazzini, OLGA L. Syrkina, the NOD2 cell line healed the quickest. They were nearly 2 2 1 PATIENTS AND METHODS: Eighty-eight patients with Keith E. Greenawalt , Thomas H. Jozefi ak, Arthur F. Stucchi, closed at 24 hours and undectable at 48 hours. The paren- 1 locally advanced esophageal cancer (cT2-4, Nx, M0) were James M. Becker CONCLUSIONS: While both SS and SF have comparable tal cell line had intermediate healing. There was minimal 1 included in the study. All patients received neoadjuvant General Surgery, Boston Medical Center, Boston, 2MA;Genzyme adhesion reduction effi cacy where applied, SS was addition- to no wound healing in the 3020InsC cell line at 24 hours chemoradiation (cisplatin, 5-FU, 45 Gy) and subsequently Corporation, Cambridge, MA ally effective in reducing adhesion formation at remote sites or 48 hours. In addition, the 3020InsC cells seemed to underwent transthoracic en bloc esophagectomy. Histo- BACKGROUND: Intraabdominal adhesions are a signifi - of peritoneal injury to which SS was not applied directly. bunch up in a “ridge” at the edge of the wound suggesting morphologic regression was defi ned as major histopath- cant source of postoperative morbidity. While bioresorb- These data suggest that SS may have widespread effi cacy migration into the wound was impaired. This “ridge” was ological response when resected specimens contained able solid physical barriers such as modifi ed hyaluronate/ throughout the in reducing adhesion forma- not seen in the NOD2 or parental cell lines. less than 10% vital residual tumor cells (major response: carboxymethylcellulose (HA/CMC) fi lm (Seprafi lm – SF) are tion without compromising intestinal wound healing. 34 patients; minor response: 54 patients). Intratumoral effective in preventing adhesions, their effi cacy is limited microRNA was isolated from pretherapeutic tissue biop- to the site of application. The aim of this study was to com- Su1841a sies and corresponding surgical specimens. Based on the pare the effectiveness of modifi ed HA/CMC sprayable pow- microarray results, the amplifi cation profi le of dysregulated NOD2 Mutation Results in Altered Wound Healing in der (Sepraspray – SS) and SF in preventing adhesions not microRNAs was analyzed and the microRNAs 192, 194 and only to sites of direct application, but also to remote peri- Epithelial Cells 622 were selected for the further analysis of the validation toneal defects to which a barrier was not directly applied. Lisa S. Poritz*1,2, Leonard R. Harris1 population. 1 Left: IEC-18 parental, Middle: NOD2 Wildtype, Right: 3020InsC mutant METHODS: Adhesion reduction was assessed in a rat isch- Surgery, The Milton S. Hershey Medical Center, Hershey, PA; RESULTS: The expression of all three microRNAs was sig- 2 emic button model and also in a rabbit cecal-sidewall injury Cellular and Molecular Physiology, The Milton S. Hershey Medical CONCLUSIONS: 1.) Wound healing was markedly retarded nifi cantly reduced during neoadjuvant therapy, showing model. Intraabdominal adhesions were induced in 30 rats Center, Hershey, PA in the 3020InsC cell line compared to the parental cell line lower levels in post-therapeutic tumor samples (p < 0.001). by creating 3 peritoneal sidewall ischemic buttons on each INTRODUCTION: A mutation in the NOD2 gene has while the NOD2 cell line healed more quickly than the Furthermore, the pre-therapeutic intratumoral expression 2 side of a midline incision. SS (5 mg/button) or SF (1 cm / been linked to terminal ileal Crohn’s Disease (CD). CD is parental cell line. 2.) The increase in proliferation seen in of microRNA 192 was signifi cantly correlated with histo- button) was applied intraoperatively over the 3 ischemic characterized by repeated mucosal injury and healing. No the NOD2 cell line may contribute to the improved wound pathologic response: patients with a major response had buttons on one side of the midline only. Adhesions were work has been done on the effect of the NOD2 mutation healing in this cell line. However, there was no change in a signifi cantly higher intratumoral microRNA 192 amount induced in 50 rabbits by cecal abrasion with the concurrent on wound healing. We have stably transfected the IEC-18 proliferation in the 3020InsC cell line to explain the lack of compared with patients having a minor response (p = 0.01). Poster Abstracts creation of a 3 cm × 5 cm sidewall defect delineated with cell line (ileal cells) with wildtype and the c-insertion muta- wound healing in the 3020InsC cell line. 3.) The disruption Moreover, by using an expression cut-off value of 0.63, silk sutures and knots. Operated control animals received tion of the NOD2 gene. We have previously characterized in the tight junction complex we have previously described the sensitivity, specifi city and accuracy of pre-therapeutic no SS or SF. On day 7 adhesions were scored in rats as the these cells and have found marked disruption of the tight in the 3020InsC cell line may contribute to the delay in microRNA 192 for assessment of histopathologic response Sunday percent of buttons with attached adhesions and in rabbits junction proteins in the cells carrying the NOD2 mutation. wound healing. 4.) Lack of wound healing may have impli- was 96%, 82% and 88% respectively (p = 0.05). cations in the mucosal injury and repair seen in CD. as the % of the sidewall defect covered by adhesions. To The hypothesis for this study is that there will be altered CONCLUSION: Our data support the role of micro RNA assess the effect of either SS or SF on intestinal healing, wound healing in the cells with the NOD2 mutation. 192 as a predictive marker for therapy response in the an additional 27 rats underwent a colonic transection dis- METHODS: IEC-18 cells were stably transfected with wild- Basic: Esophageal multimodality therapy of patients with locally advanced tal to the cecum, which was repaired with an end-to-end esophageal cancer. In a multi-institutional trial we will now anastomosis. SS or SF was applied circumferentially to the type human NOD2 (NOD2 cell line) or the c-insertion mutation (3020InsC cell line). Western blot and PCR was try to confi rm these results. anastomosis site. The anastomosed colonic sections were Su1842 removed 7 days later and their integrity assessed by burst used to identify NOD2. Cells were then grown in 6 well pressure and tensile strength measurements. plates for 72 hours until confl uent. The monolayer was The Micro-RNA 192 as an Effective Response Prediction then wounded with a plastic pipette tip in an X formation. Factor in the Multimodality RESULTS: The direct application of both SS and SF sig- 24 or 48 hours later the cells were stained Hema 3 stain kit *1 1 1 nifi cantly (p < 0.04) reduced adhesion formation compared and viewed with a phase contrast microscope. Duplicate, Daniel Vallbohmer, Peter P. Grimminger, Christoph Wandhoefer, 1 2 1 with controls in both effi cacy models (Table). While SF had unwounded, monolayers were used to measure cell prolif- Jan Brabender , Uta Drebber, Elfriede Bollschweiler, Arnulf H. no remote effects on adhesion formation in either model, eration using a coulter counter, n = 8. HöLscher 1, Ralf Metzger1, Magarethe Odenthal2 SS signifi cantly (p < 0.01) reduced adhesion formation to 1 Department of General, Visceral and Cancer Surgery, University of ischemic buttons to which the powder was not directly Cologne, Cologne, Germany;2Institute of Pathology, University of applied. In rabbits, SS reduced remote adhesion formation Cologne, Cologne, Germany to the sidewall defect by 70% (p < 0.1). In the anastomosis BACKGROUND: healing model, neither SF nor SS affected intestinal anasto- Neoadjuvant multimodality treatment motic burst pressure (Control: 240 ± 8.2 vs. SS 215 ± 19.0 is frequently applied to improve the poor prognosis asso- vs. SF: 216 ± 27.2 mm Hg) or tensile strength (Control: 2.4 ciated with locally advanced esophageal cancer. However, ± 0.2 vs. SS: 2.3 ± 0.2 vs. SF: 2.2 ± 0.2 Newton).

58 59 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Basic: Hepatic Basic: Pancreas Su1845 Su1846 Systemic Infl ammation with Multiorgan DysfunctionA Promising Novel Target in Pancreatic Cancer: HuR Su1843 Su1844 Is the Cause of Death in Murine Pancreatic Duct Modulates Multiple Core Signaling Pathways Required Ligation-Induced Acute Pancreatitis for Pancreatic Tumorigenesis Diazoxide, a Opening mitoKATP, Reduces Liver Pterostilbene Induces Mitochondrially-Derived *2 2 1,2 *1 2 1 Damage Secondary to Ischemic/Reperfusion Injury Apoptosis in Pancreatic Cancer Cells by Increasing Zuobiao Yuan, David Meyerholz, Deborah Williard, Vanessa A. Talbott, Koree Ahn, David W. Rittenhouse, 1,2 2 1,2 1 1 MnSOD Activity and Release of Cytochrome C and Erik Twait , Kempuraj Duraisamy, Isaac Samuel Nathan G. Richards , Agnes Witkiewicz3, Eugene P. Kennedy, Mateus A. Nogueira, Ana Maria M. Coelho*, Sandra N. Sampietre, 1 VA Medical Center, Iowa City, IA;2University of Iowa Carver Myriam Gorospe 4, Charles J. Yeo1, Jonathan R. Brody1 Nilza A. Molan, Luiz Augusto C. D’Albuquerque, Marcel C. MachadoSmac/DIABLO College of Medicine, Iowa City, IA 1 Surgery, Thomas Jefferson University, Philadelphia, PA; Gastroenterology, University of Sao Paulo, Sao Paulo, Brazil *1,2 1 1 Denise E. Mccormack, Debbie E. Mcdonald, David W. Mcfadden 2Biochemistry & Molecular Biology, Jefferson College of Graduate 1 2 Existing animal models of acute pancreatitis (e.g., cerulein, BACKGROUND/AIM: 3 Previous studies have demonstrated Surgery, University of Vermont, Burlington, VT;Surgery, Danbury choline-defi cient ethionine-supplemented diet) do not Studies, Philadelphia, PA;Pathology, Thomas Jefferson University, that diazoxide protects heart from ischemia/reperfusionHospital, Danbury, CT resemble gallstone pancreatitis as the etiologies are not Philadelphia, PA; 4National Institutes of Health, Baltimore, MD injury however there are no prior studies of the role of BACKGROUND: We have previously shown that Pteros- analogous. Distal pancreatic duct ligation (PDL) more INTRODUCTION: diazoxide on liver ischemic reperfusion injury. In the pres- Twelve core signaling pathways with tilbene (3, 5- dimethoxy-4-hydroxystilbene), a compound closely resembles early events in gallstone pancreatitis. We 540 overexpressed individual genes have recently been ent study we evaluated the effect of diazoxide on local and found in blueberries, inhibits cell proliferation and pro- recently developed a novel murine model of PDL-induced systemic liver ischemia/reperfusion (I/R) process. identifi ed as critical for the development of pancreatic duc- motes apoptosis in pancreatic cancer in vitro by induction acute pancreatitis associated with substantial mortality. tal adenocarcinoma (PDA) (Science 2008, 321:1801-1806). METHODS: of mitochondrial membrane depolarization and caspase 3/7 Wistar male rats underwent partial liver isch- Using this model, we previously showed that specifi c inhi- The mechanism of overexpression for nearly all (99%) of activation. To further investigate the role of mitochondria emia performed by clamping the pedicle from medium bition of the stress kinase ERK with in vivo gene modula- the identifi ed up-regulated genes in pancreatic tumori- and left anterior lateral segments during an hour under in Pterostilbene-induced apoptosis in pancreatic cancer, we tion signifi cantly improves survival. In the present study examined its effects on manganese superoxide dismutase genesis is unknown. We explored the hypothesis that mechanical ventilation. They were divided into 2 groups: we determine the cause of death in our murine model by post-transcriptional gene regulation may be a powerful Control Group (n = 26): rats received saline and Diazoxide (MnSOD) activity, Cytochrome C and Smac/DIABLO release. serial examination of multiple parameters in three groups: The mitochondrial enzyme MnSOD plays a critical role in alternative process in which these up-regulated genes were Group (n = 26): rats received IV diazoxide (3.5 mg/kg) 15 a) Acute pancreatitis group had PDL; b) Hepatic obstruc- regulation of cancer cell proliferation through an unknown being disrupted. A key component of this regulatory pro- tion group had bile duct ligation (BDL) without PDL; c) minutes before liver reperfusion. Four and 24 hours after mechanism. Smac/DIABLO is a mitochondrial protein that cess is Human antigen R (HuR), which can modulate gene reperfusion, blood were collected for determinations of potentiates apoptosis. Both Smac/DIABLO and Cytochrome C Sham operation group. The mice were observed for 15 days expression by binding to mRNAs that encode for tumor- AST, ALT, TNF-D, IL-6, and IL-10. Liver tissues were assem- have been shown to exit mitochondria and enter the cytosol post-operatively. BDL and Sham controls had no mortality. promoting proteins in cancer cells. Previously, we discov- bled for histologic analysis, malondialdehyde (MDA) con- during apoptosis. We hypothesized that Pterostilbene would Close to 100% mortality was seen in PDL-induced acute ered that HuR is a key marker for poor pathologic features tent, and mitochondrial oxidation and phosphorylation. increase MnSOD activity and cytosolic levels of Cytochrome pancreatitis with most deaths occurring between day 2 in PDA and is a predictive marker for gemcitabine-based Pulmonary vascular permeability and myeloperoxidade C and Smac/DIABLO in a dose dependent manner. and day 4. Characteristics of mice with acute pancreatitis chemotherapy. (MPO) were also determined. METHODS: included the following (ANOVA; P < 0.05): ERK activation METHODS: MIA and PANC-1 cell lines were treated with in the pancreas and distant organs; pancreatic neutro- Using a bioinformatic approach, we identi- RESULTS: Four hours after reperfusion Diazoxide Group 25 micromolar and 50 micromolar concentrations of phil infi ltration and acinar cell necrosis maximal on day fi ed putative HuR targets from the 540 overexpressed genes presented elevation of AST, ALT, TNF-D, IL-6 and IL-10 serum Pterostilbene for 48 hrs and quantitative MnSOD activity Poster Abstracts 2; increased plasma IL-1E and TNF-D levels on day 2, that in PDA (Nucl Acids Res. 2001, 29:246-254, PNAS 2004, levels signifi cantly lower than Control Group (p < 0.05). A was measured by superoxide dismutase ELISA assay. In sep- peaked on day 3, in parallel with worsening hypotension 101:2987-92). HuR binding to mRNA-targets was vali- signifi cant reduction on liver MDA content and on mito- arate experiments, MIA and PANC-1 cell cells were treated Sunday and bradycardia; bronchoalveolar lavage (BAL) fl uid neu- dated by PCR-based analysis in ribonucleoprotein immu- chondrial dysfunction were observed in Diazoxide Group with 25 micromolar, 50 micromolar and 75 micromolar trophil count and IL-1E level, and plasma aspartate ami- noprecipitated HuR:RNA complexes. Further validation of compared to Control Group (p < 0.05). No differences in concentrations of Pterostilbene for 24 hrs and cytosolic notransferase (AST) level, also peaked on day 3; pulmonary expression of HuR target genes was performed by quantita- pulmonary vascular permeability and MPO activity were extracts were analyzed for Cytochrome C and Smac/DIA- neutrophil infi ltration and plasma creatinine level peaked tive PCR analysis and immunoblotting. observed between groups. Twenty four hours after reperfu- BLO utilizing ELISA protocols. One way ANOVA and Tukey on day 4. Liver injury evidenced by raised AST after hepatic RESULTS: sion Diazoxide Group showed a reduction of AST and ALT post-hoc analysis were used for statistical analysis. We identifi ed 60 putative targets (11.1%) for obstruction was exacerbated by PDL. Increased plasma serum levels when compared to Control group (p < 0.05). RESULTS: Pterostilbene increased enzymatic activity of HuR regulation among the overexpressed genes in PDA. In IL-1E and TNF-D on day 2 after BDL subsided thereafter. comparison, genetic and epigenetic alterations contribute CONCLUSION: Diazoxide attenuates liver ischemia/reper- MnSOD in both cell lines in a dose dependent manner (p < 0.01). Pterostilbene treatment increased cystosolic levels of Although BDL was also associated with pulmonary neutro- only 1% and 1.8% respectively to the proposed mecha- fusion injury probably by a mechanism related to mito- Cytochrome C in MIA cells at the 25 micromolar (p < 0.05) phil infi ltration it was not associated with increased IL-1E or nisms by which these 540 genes are disrupted in PDA. Ten chondrial function preservation during liver ischemia. and 75 micromolar concentrations in PANC-1 cells (p < neutrophils in BAL fl uid. BDL-induced renal tubular dam- of the 60 putative target genes which are a part of 5 of the 0.05). Cystosolic levels of Smac/DIABLO increased in both age was not associated with raised plasma creatinine. Our 12 core signaling pathways in PDA, including K-Ras, were MIA and PANC-1 cells with treatment at the 75 micromolar fi ndings indicate that the high mortality rate seen in PDL- experimentally validated as specifi c HuR targets. induced acute pancreatitis in mice is associated with pro- concentration of Pterostilbene (p < 0.01). CONCLUSION: HuR is an unprecedented regulatory pro- gressive systemic increase of infl ammatory cytokine levels, CONCLUSION: We have previously demonstrated that tein of at least 5 critical signaling pathways in PDA. Target- cardiovascular instability, acute lung injury, liver injury, Pterostilbene, a natural plant-derived stilbene, inhib- ing multiple core pathways in PDA through silencing HuR and renal dysfunction, suggesting that systemic infl am- its pancreatic cancer in vitro through activation of the expression may be a potent therapeutic strategy to treat mation with multiorgan failure is the proximate cause of mitochondrial apoptosis pathway. MnSOD, an inducible this disease. mitochondrial enzyme that converts superoxide anion to death in this experimental model. Lung and renal injury hydrogen peroxide, has an essential role in regulation of as observed by morphology after hepatic obstruction alone pancreatic cancer cell proliferation. The results of our cur- is not associated with systemic infl ammation or death. In rent study demonstrate for the fi rst time that Pterostilbene conclusion, systemic infl ammation with multiorgan dys- increases MnSOD activity in pancreatic cancer cells. In function causes death in pancreatic duct ligation-induced addition, Pterostilbene increases cytosolic levels of Cyto- acute pancreatitis in mice. This experimental model is a chrome C and Smac/DIABLO in both cell lines, confi rm- useful analogy of severe gallstone pancreatitis to investi- ing mitochondrially derived apoptotic cell death. Further gate disease pathogenesis and to evaluate novel therapeutic studies are ongoing to elucidate the intricate relationship strategies. between intrinsic apoptosis, MnSOD activity, and pancre- atic cancer cell death upon Pterostilbene treatment.

60 61 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

CONCLUSION: Contractile activity can be inhibited pre- Basic: Small Bowel Su1849 dominantly via muscular D- and E-receptors. However, dur- ing POI long lasting changes in balance of muscular and Role of Hydrogen Sulfi de in Contractile Activity in Su1847 neuronal D- and E-receptors occur that might participate Circular Muscle of Rat Jejunum in pathophysiology of POI. These changes are paralleled by Changes in Neurotransmission via Alpha- and Munenori Nagao*1,2, Judith A. Duenes1,2, Michael G. Sarr1,2 intramural infl ammation and impaired intestinal transit. 1 2 Beta-Receptors During Postoperative Ileus in Rat DFG KA 2329/5-1 Surgery, Mayo Clinic, Rochester, MN;Gastroenterology Research Circular Jejunal Muscle Unit, Mayo Clinic, Rochester, MN

*1,2 1,2 1,2 BACKGROUND: Michael S. Kasparek, Brigitte Goetz, Bernhard Stoklas, Su1848 Endogenous hydrogen sulfi de (H2S) appears Petra Benhaqi 1,2 1,2 1,2 to be a new gasotransmitter that modulates small intestinal , Mario H. Mueller, Martin E. Kreis Central Vagal Activation During Early Postoperative 1 motility, but mechanisms of action of H S are unknown in Department of Surgery, Ludwig-Maximilians-University Munich, 2 Ileus in the Mouse gut smooth muscle. Munich, Germany;2Walter Brendel Centre of Experimental Medicine, Mia Karpitschka*2, Mario H. Mueller1, Michael S. Kasparek1,2, AIM: LMU Munich, Munich, Germany To determine effects and mechanisms of action of H2S Jorg Glatzle 3, Martin E. Kreis1 BACKGROUND: on contractile activity in circular muscle of rat jejunum. Our aim was to: 1) investigate the role 1 Surgery, University of Munich, Munich, Germany;2Walter-Brendel of D and E-receptors in control of contractile activity in rat METHODS: Jejunal circular muscle strips from Lewis rats Institute, University of Munich, Munich, Germany; 3Surgery, University of circular jejunal muscle; 2) explore changes in adrenergic were mounted in temperature-controlled tissue cham- Tuebingen, Tuebingen, Germany neurotransmission via these receptors during postoperative bers to measure contractile activity. Sodium hydrosulfi de INTRODUCTION: ileus (POI); 3) determine if these changes are paralleled by Postoperative ileus (POI) involves refl ex (NaHS) was used as an exogenous donor of H2S yielding physiologic levels of H S of 10–5, 10-4, 2 × 10–4, and 5 × 10–4 intramural infl ammation and delayed intestinal transit. inhibition of intestinal motility and a subsequent intesti- 2 Figure 1 METHODS: nal infl ammatory response that is characterized by efferent M, resp. We evaluated effects of NaHS on spontaneous con- Muscle strips (n = 8/rat) from 6 male Sprague tractile activity and after precontraction with bethanechol Dawley rats per group were studied in organ chambers. vagal modulation. However, the role of central vagal affer- ents in the early phase of POI is unknown. We, therefore, (10–4 M). L-cysteine (10–4, 10–3 and 10–2 M), the substrate for Groups: Naïve controls (NC), sham controls (SC) 24 h after endogenous production of H S via H S-producing enzymes, laparotomy, rats 12h (P12h), 1 (P1d), 3 (P3d), and 7days aimed to explore central vagal afferent nerve activation 2 2 early during POI. was also evaluated. We used atropine (10-7 M), phentol- (P7d) after laparotomy and standardized small bowel amine (10–5 M), and propranolol (5 × 10–6 M) together to MATERIAL AND METHODS: C57BL/6 mice were vagot- manipulation to induce POI. After spontaneous contractile establish non-adrenergic/non-cholinergic (NANC) condi- 2 omized 3–4 days prior to ileus experiments, while control activity (g/mm /min) was recorded, dose-response curves tions, and in separate experiments tetrodotoxin (10–6 M), animals received a sham operation without vagotomy. for phenylephrine (Da; D-agonist; 10-8-3 × 10-6M) and iso- capsaicin (10–5 and 10–4 M), L-NG-nitro arginine (L-NNA; For ileus experiments, laparotomy was either followed by prenaline (Ea; E-agonist; 3 × 10-10-10-7M) were established. 10–4 and 10–3 M), glibenclamide (10–5 and 10–4 M), apamin standardized small bowel manipulation to induce POI or Responses were repeated with tetrodotoxin (TTX; blocking (10–6 and 5 × 10–6 M) , and calyculin A (10–6 M) to study sham treatment for control. Then, after 3h or 9h, the brain enteric nerves; 10-6M), after precontraction with bethan- involvement of the NANC and enteric nervous system, pri- echol (3 × 10-6M), and with phentolamine (D-anatagonist; was removed, fi xed and Fos-immunohistochemistry was + Poster Abstracts mary visceral afferent nerve fi bers, nitric oxide, K ATP chan- 10-5M), or propranolol (E-antagonist; 5 × 10-6M). Intesti- performed to determine neuronal activation in the vagal + nel, K Ca channel, and myosin light chain phosphatase

nucleus of the solitary tract (NTS) and the area postrema. Sunday nal transit was studied by charcoal transit (% of small bowel activity, respectively, in mediating the inhibitory action of Each subgroup contained an n of 6. Data were analyzed by passed). Histology of jejunal whole mounts was performed H S. Motility was quantitated as area under the contractile for myeloperoxidase positive cells (MPO), macrophages, two-way ANOVA. 2 2 curve and used to compare the effects. and mastcells (cells/mm ).Data: mean ± SEM. RESULTS: The number of Fos-positive neurons in the NTS RESULTS: NaHS exhibited a marked, dose-dependent RESULTS: was 15 ± 0.5 at 3 h following ileus which was not different Spontaneous contractile activity was increased inhibitory effect (Figure 1) on spontaneous contractile from 14 ± 0.4 in sham controls and uninfl uenced by vagot- in SC, P12h, and P7d (NC 1.3 ± 0.3; SC 3.7 ± 0.9; P12h 4.5 activity (p < 0.05); threshold dose 10–4 M, ID (50% inhibi- ± 1.5; P7d 3.6 ± 1.0; p < 0.05), but not in P1d (1.7 ± 0.4) omy (12 ± 2). At 9 h after induction of POI, the number of 50 tory dose) = 3 × 10–4 M. However, surprisingly, NaHS did Figure 2 Fos-positive neurons was increased in the NTS to 186 ± 6 and P3d (1.8 ± 0.5; p = NS). Da and Ea inhibited spontane- not inhibit bethanechol-stimulated contractile activity. compared to 60 ± 4 in sham controls (p < 0.05) which was ous contractile activity dose-dependently in all groups (p L-cysteine also exerted a dose-dependent inhibitory effect not different to 74 ± 4 neurons after vagotomy (n.s.). In the < 0.05). In NC, TTX reduced Da-induced inhibition (p < on contractile activity (Figure 2) (p < 0.05). Establishment Basic: Stomach area postrema, after performing a vagotomy, the number 0.05), but TTX had no effect on Da-responses in POI groups of NANC conditions, blockade of all neural activity by (p = NS). In contrast, TTX did not affect Ea-response in of Fos positive cells was higher during POI at 9 h compared CONCLUSIONS: HS at physiologic concentrations and tetrodotoxin, exposure to capsaicin (block visceral affer- 2 to sham controls (14 ± 2 in POI vs. 2 ± 1 in sham controls, NC (p = NS), but increased responses in POI groups (p < ents), L-NNA (block NO production), glibenclamide (block L-cysteine inhibit contractile activity of rat jejunal circu- 0.05). Da and Ea-induced inhibition was reduced in P12h, p < 0.05). + + lar muscle. The inhibitory effect of H S on jejunal circular K channels), or apamin (block K channels) had no 2 P3d, and P7d and in all POI groups, respectively (p < 0.05). ATP Ca muscle does not appear to be mediated via K+ channels as CONCLUSIONS: Central nerve activation in the NTS effect on (i.e., did not block) the inhibitory response to ATP Precontraction had no effect on Da and Ea-responses (p = occurs via vagal afferents 9 h after induction of POI. This 5 × 10–4 M NaHS. After exposure to calyculin A (block myo- in the vascular tree or by the more classic pathways via the NS). Effects of Da and Ea were blocked by phentolamine or suggests that vagal afferents projecting to the CNS were sin light chain phosphatase), the inhibitory effect of NaHS enteric nervous system, a nitric oxide pathway, activity of propranolol (p < 0.05). Intestinal transit was delayed in all visceral afferent nerve fi bers, or K+ channels but rather via sensitized, potentially by infl ammatory mediators released was prevented. Ca POI animals and recovered over time (NC 53 ± 3; P12h 22 in the intestinal wall. Central activation in the area pos- myosin light chain phosphatase. (Support RO1-DK39337 ± 2; P7d 44 ± 2%; p < 0.05), but was unaffected in SC (50 trema may be secondary to higher permeability of the [MGS]) ± 5%; p = NS). MPO positive cells and mastcells increased blood-brain-barrier in this area, allowing access of infl am- postoperatively and peaked in P1d (NC 14±2; P1d 763±48) matory mediators via the systemic circulation with subse- and P12h (NC 9 ± 1; P12h 700 ± 79; both p < 0.05), respec- quent neuronal activation. tively; no effect was observed in SC (56 ± 37 and 30 ± 12, respectively; p = NS). Macrophages peaked in P3d (NC 367 ± 41; P3d 1306 ± 178; p < 0.05); counts in SC and P12h were similar to NC (SC 395 ± 82; P12h 706 ± 19; p = NS).

62 63 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1850 Su1851 Clinical: Biliary Su1601 Molecular Determinants of Hyperthermic IntraperitonealEnergy Metabolism Following High-Fat Diet and High Expression of Telomerase Is an Independent Chemotherapy (HIPEC) in a Model of Peritoneal GastricBariatric Surgeries (Gastric Bypass, Sleeve Gastrectomy, Su1600 Prognostic Factor in Ampullary Carcinoma Cancer Carcinogenesis Duodenal Switch and Ileum Transposition) in Rats Cost Comparison of Single Incision Laparoscopic Ryutaro Sakabe*, Yoshiaki Murakami, Kenichiro Uemura, 2 1 1 *1 1 1 Luigina Graziosi, Andrea Mencarelli, Barbara Renga, Yosuke Kodama, Helene Johannessen, Marianne W. Furnes, Cholecystectomy to Standard Laparoscopic Takeshi Sudo, Yasushi Hashimoto, Akira Nakashima, 1 2 2 1 1,2 1 Claudio D’Amore , Emanuel Cavazzoni, Annibale Donini, Chun-Mei Zhao , Baard Kulseng, Duan Chen Cholecystectomy Taijiro Sueda, Eiso Hiyama Stefano Fiorucci *1 1 Department of Cancer Research and Molecular Medicine, Norwegian Surgery, Hiroshima University, Hiroshima, Japan 1 Medicina Clinica e Sperimentale, University of Perugia, Perugia, Italy;University of Science and Technology, Trondheim, Norway;2Department Meredith C. Mcmahon*, Robert C. Martin, Farid Kehdy Department of Surgery, University of Louisville, Louisville, KY BACKGROUND AND AIMS: Telomerase activity and 2Chirurgia Generale e ‘Urgenza, University of Perugia, Perugia,of ItalySurgery and Endocrinology, St. Olavs University Hospital, Trondheim, Norway BACKGROUND: human telomerase reverse transcriptase (TERT), one of the BACKGROUND: Peritoneal dissemination is a common Single incision laparoscopic cholecystec- major catalytic subunits of telomerase, have been reported feature in patient with locally advanced gastric cancer. BACKGROUND/AIM: The role of energy metabolism tomy (SILC) has fallen into favor by some surgeons as a as a marker of tumor aggressiveness and poor prognosis in Syncronous peritoneal metastases are found in 10–20% of in obesity aetiology as well as bariatric surgeries remains way of improving patient satisfaction for cosmesis. Mul- some kinds of malignancies. However, no previous reports patients and during the follow-up a further 60% of patients unclear. The aim of the present study was to examine the tiple reports have been published describing the feasibil- have shown a correlation between telomerase or TERT with T4 stage develop peritoneal cancer with a rapid dete- energy metabolism in rats fed with high-fat diet or under- ity and safety of this approach. However, few reports have expression and patient outcome in ampullary carcinoma. rioration of survival rates. Systemic chemotherapy has little went different surgical procedures that are currently used compared SILC to conventional laparoscopic cholecystec- In the present study, we examined telomerase activity and effect on peritoneal dissemination because the peritoneal- in obese patients. tomy (LC). We set out to compare our experience with both TERT expression in ampullary carcinoma to determine blood barrier hinders drug distribution throughout the approaches from the years of 2008–2010. METHODS: Male rats were fed with food containing either whether these parameters can be used as prognostic indica- peritoneal cavity. Cytoreductive surgery and hyperthermic METHODS: 60% or 10% fat (as controls) for 36 weeks, or subjected to Twenty one SILC were performed in the tors of patient outcome. intraperitoneal chemotherapy (HIPEC) has been proposed gastric bypass (GB), sleeve gastrectomy (SG), duodenal period of 10/2008 and 10/2010. We retrospectively com- METHODS: A total 46 ampullary carcinoma patients who for the treatment of peritoneal dissemination, thought the switch (DS), SG + DS, ileum transposition (IT), or sham pared our experience to 100 conventional LC over a similar underwent pancreatoduodenectomy with regional lymph biological effects of this treatment are largely unexplored. operation (SO) for 8–14 weeks follow-up.Energy metabolic time period performed at the same institution. Variables node dissection were enrolled in this study. Telomerase AIMS: To investigate the effects of HIPEC on an experi- parameters were measured by an open-circuit indirect calo- compared include: BMI, age, pre-operative diagnosis, oper- activity was analyzed by the telomeric repeat amplifi cation mental murine model of peritoneal carcinomatosis. rimeter composed in comprehensive laboratory animal ative time, and hospital charges. protocol assay using frozen tissue samples. TERT was exam- RESULTS: MATERIALS AND METHODS: MKN45 cells lines were monitoring system. Five of the 21 (23.8%) attempted SILC proce- ined by immunohistochemistry using paraffi n embedded injected into the peritoneal cavity of NOD-SCID mice. Mice RESULTS: During the course of high-fat diet-induced dures had to be converted to traditional LC for various samples. Relationships between survival and clinicopatho- were randomized as follow: HIPEC plus mytomycine (16.5 obesity, total energy expenditure (TEE kcal/h/rat), but not reasons pertaining to safe identifi cation of anatomy. SILC logical factors including tumor characteristics, telomerase mg/L of perfusate each mouse) and cisplatin (125 mg/L of energy expenditure relative to body weight (EE kcal/h/100 patients were overall younger with a mean age 35.6 (18–65) activity, and TERT expression were evaluated using univari- perfusate each mouse), normothermic intraperitoneal che- g b.w.), was elevated at 3 and 20 weeks but not at 36 weeks compared to 45.9 (18–80) for the LC group (P 0.0062). The ate and multivariate analysis. motherapy (NIPEC), normothermic intraperitoneal saline compared with age-matched controls. Respiratory exchange average BMI in the SILC group was 25.8 (18–33) compared RESULTS: Telomerase activity was detected in 42 (91.3%) (NIPES) and hyperthermic intraperitoneal saline (HIS); and to 32.3 (16.4–65.5) in the LC group (P 0.0024). Operative Poster Abstracts ratio (RER) was reduced, due to low production of CO2 of ampullary carcinoma and 27 (58.7%) showed high control group. After 8 days from the intraperitoneal injec- time in the SILC group was signifi cantly longer with an (VCO2 ml/kg/h), throughout the course. After GB, TEE was activity (>8.9 TPG units, which is the maximum level of tion mice were sacrifi ced and the extent of peritoneal car- unchanged, but EE was increased at both 3 and 14 weeks average of 97 minutes compared to LC that averaged 53 telomerase in normal duodenal tissue samples). In con- Sunday cinogenesis was evaluated. All peritoneal and mesenteric postoperatively compared with SO. RER was unchanged minutes (P < 0.001). Hospital charges for the two tech- trast, TERT was expressed in 35 (76.1%), including 21 nodules were examined, counted and measured. at 3 weeks but reduced at 14 weeks. Oxygen consump- niques were not statistically different and averaged $20,128 weak and 14 strong immunohistochemical staining. Uni- (9,782–58,205) in SILC patients versus $20,751 (11,449– RESULTS: tion (VO ml/kg/h) was elevated at both 3 and 14 weeks, variate analysis revealed that tumor depth (P < 0.001), 100% of mice injected with MKN45 cells devel- 2 86,417) in LC patients (P 0.8846). Finally, in SILC patients, whereas VCO was unchanged at 3 weeks but increased at nodal status (P = 0.013), UICC stage (P = 0.009), perineural oped diffuse colonization of peritoneal cavity, HIPEC, 2 pre-operative diagnosis favored biliary colic (60%) and 14 weeks. After SG, neither TEE nor EE was changed neither invasion (P < 0.001), telomerase activity (P = 0.031) were effectively reduced the extent and severity of peritoneal dis- biliary dyskinesia (40%). In LC patients, the pre-operative 2 nor 11 weeks compared with preoperative levels. RER was signifi cantly associated with disease specifi c survival. TERT semination as measured by assessing the total number of diagnosis also favored biliary colic (54%); however, a larger increased only at 11 weeks during nighttime. After DS, TEE, expression showed a tendency toward association with peritoneal nodules and mesenteric nodules as well as the percentage of patients had acute cholecystitis (25%). but not EE, was reduced during both daytime and night- prognosis (P = 0.054). In multivariate analysis, only telom- total volume of peritoneal nodules. However, both HIPEC CONCLUSIONS: and NIPEC treatment were associated with a slight dete- time at 2 and 8 weeks compared with preoperative levels. Single incision laparoscopic cholecys- erase activity remained independent predictors of progno- rioration clinical parameters including weight loss. HIPEC RER was reduced at 2 weeks but increased at 8 weeks during tectomy is favored in younger patients with a lower BMI sis (P = 0.043). The 5-year survival rates of the patients with treatment effectively downregulated the expression of genes daytime but unchanged during nighttime. After SG+DS, and non-acute pathology but the operative time is signifi - high or low telomerase activity were 66.2% and 94.4%, involved in the formation of peritoneal cancer nodules. TEE was reduced at 2 weeks (both daytime and nighttime) cantly longer. However, our data suggests that even though respectively. Thus by gene array we found that HIPEC caused a robust and 8 weeks (both daytime and nighttime), whereas EE was operative time is longer in the single incision laparoscopic CONCLUSIONS: Our results suggest that telomerase activ- downregulation in CXC chemokine Receptor-4 (CXCR4); increased only at 8 weeks (both daytime and nighttime). cholecystectomy, hospital charges are comparable to con- ity has a prognostic signifi cance for survival in patients SMAD2, a family of proteins that mediate TGF-E signalling RER was unchanged neither at 2 nor 8 weeks. After IT, nei- ventional laparoscopic cholecystectomy. In terms of cost with resected ampullary carcinoma. Telomerase activity ther TEE nor EE was changed at 2 weeks. Resting VO and analysis, SILC is a justifi able option in a carefully selected pathway; C-terminal binding protein (CTBP1); collagen IV 2 may provide a new marker for evaluating the progression VCO (during 7–8 a.m.) were increased. RER was increased group of patients, although it does subject patients to lon- (COL4A2); APC, a gene correlated with beta-catenin expres- 2 of patients with ampullary carcinoma, and may help to during both daytime and nighttime, due to higher VCO . ger anesthesia time. sion. HIPEC treatment ameliorates long-term survival (5 2 identify patients in need of adjuvant treatment. weeks) by 50% in comparison to control mice. CONCLUSIONS: In high-fat diet-induced obese rats, TEE CONCLUSIONS: HIPEC protects against peritoneal dissemi- was increased initially and energy was provided mainly nation in a mouse model of peritoneal gastric cancer carci- by fat metabolism. GB or SG+DS increased EE and the fat nogenesis with a positive citology. HIPEC might represent oxidation, whereas SG or IT was without effects on EE but a prophylactic approach to patients with locally advanced increased the carbohydrate metabolism. The hypermeta- gastric cancer at a high risk to develop peritoneal cacinoma- bolic state is likely to contribute to the metabolic benefi ts tosis. HIPEC acts at multiple check-points in the process of after IT procedure. attachment and adhesion of tumor cells in the peritoneum.

64 65 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1602 Clinical: Colon-Rectal controls (all p < 0.05). 33% of patients were sexually inac- years vs. 46 years; p = 0.01) and were more likely to be tive at follow-up with higher age, preoperative vs. no XRT, male (PCDC: 67% male; SCDC: 55% male; p < 0.0001). LOS Single Incision Laparoscopic Cholecystectomy: and not living in a relationship being associated with was signifi cantly longer for SCDC than PCDC (14.4 days A Combined Analysis of Resident and Attending Su1603 sexual inactivity in multivariable analysis (p < 0.05). In vs. 5.2 days; p < 0.0001). The major diagnostic code (MDC) Learning Curves at a Single Institution sexually inactive patients, younger age, female gender, and for SCDC admissions was related to digestive system dis- Does Pelvic Radiotherapy Affect Genitourinary Function preoperative vs. no XRT were associated with the feeling orders (25%); by comparison, for all in-patients included Mark Joseph*, Michael Phillips, Christopher C. Rupp in Patients After Abdominoperineal Resection for Department of Surgery, University of North Carolina, Chapel Hill, NC that surgery for rectal cancer caused sexual inactivity (p < in the database regardless of CDC status, only 9.4% had Distal Rectal Cancer? 0.05). Postoperative vs. no XRT had no such effect (NS). a primary admission MDC for digestive disorders. SCDC INTRODUCTION: Single Incision Laparoscopic Cholecys- Michael S. Kasparek*1,4, Imran Hassan2, Robert R. Cima1, CONCLUSIONS: patients had a slightly greater mean Charlson comorbidity tectomy (SILC) is a recent technical modifi cation on stan- Although urinary and sexual function Dirk R. Larson 3, Rachel E. Gullerud3, Bruce G. Wolff1 might be impaired after APR, effects of XRT appear to be score (primary: 0.76; secondary: 1.29; p < 0.0001). Mean dard laparoscopic cholecystectomy that has been shown 1to total inpatient hospital payments were signifi cantly higher Division of Colon and Rectal Surgery, Mayo Clinic Rochester, limited. Indication and timing of XRT should be based on be safe and feasible. Studies to date have focused on deter- for the SCDC group ($44,349 vs. $12,693; p < 0.0001). Re- Rochester, MN;2 oncological aspects rather than on QOL and functional mining the role of SILC in modern surgery. Recent stud- Department of Surgery, Southern Illinois University admission rates to treat CDC (Table 1) following PCDC 3 outcome considerations. ies suggest that experienced laparoscopic surgeons have a School of Medicine, Springfi eld, IL;Section of Biostatistics, Mayo were twice as high at 30-days and three times higher at 4 short learning curve to become profi cient in SILC. How-Clinic Rochester, Rochester, MN; Department of Surgery, Ludwig- 60-days compared to SCDC patients. Only 47.5% of PCDC ever, little is known about the interaction of the learningMaximilians-University Munich, Munich, Germany Su1604 patients received any antibiotic therapy (metronidazole curves of residents and attending surgeons at academic OBJECTIVE: Abdominoperineal resection (APR) as well Hospital Costs, Length of Stay and Readmission Ratesand/or vancomycin) at the time of their fi rst discharge, and teaching programs. as radiotherapy (XRT) have detrimental effects on urinary for C. Diffi cile Colitis: Comparing Outcomes Betweenonly 26.4% of SCDC were provided discharge antibiotics. METHOD: We prospectively evaluated various metrics of and sexual function, but little is known about effects of CDC as the Primary and Secondary Admission both attending and resident surgeons as they progressed XRT in APR patients specifi cally. Therefore, our aim was to Diagnosis Readmission Rates for Primary and Secondary C. diffi cile colitis in their experience with SILC. Patients were placed into investigate potential effects of XRT on genitourinary func- cohorts of 25 based on teaching surgeon experience. Data David B. Stewart*, Li Wang PCDC SCDC p value tion in APR patients. 30-day Readmission Rates for CDC 5.5% 2.3% < 0.0001 recorded included patient specifi c variables as well as oper- Surgery/Division of Colon and Rectal Surgery, Penn State Hershey METHODS: International Consultation on Incontinence 60-day Readmission Rates for CDC 6.2% 2.8% < 0.0001 ative variables, complications, conversions to standard lap- Medical Center, Hershey, PA Questionnaire (ICIQ), American Urological Association Patients Given Metronidazole and/or 47.5% 26.4% < 0.0001 aroscopic cholecystectomy, and outcomes. Vancomycin at First Hospital Discharge Symptom Index (AUASI), Brief Sexual Function Inventory PURPOSE: It is unknown how C. diffi cile colitis (CDC) (BSFI) for men, and sexual function module of the Can- patients differ in terms of hospital costs, length of hospi- Operative Times of Patient Cohorts Versus Teaching Surgeoncer Rehabilitation Evaluation System (CARES) for women tal stay (LOS) and readmission rates depending on whether CONCLUSIONS: Among the study population, SCDC was Experience were mailed to 219 patients who underwent APR between patients were admitted principally to treat CDC versus more common and associated with greater hospital costs Cohort 1 Cohort 2 Cohort 3 Cohort1994 4 and 2004. Male sexual function was compared to developing CDC while admitted for other reasons. and longer LOS than PCDC. Admission to the hospital for (Patients 1–24) (Patients 25–49) (Patients 50–74) (Patients 75–101) digestive system disorders is associated with developing previously published normative data (OLeary MP, Urology METHODS: After IRB approval, the MarketScan Research CDC. Readmission rates for subsequent bouts of CDC are Operative 85 67 75 751995). 143 patients responded (response rate 65%; 71% Database was used to study individual-level healthcare times two to three times as likely for PCDC. A surprisingly large Poster Abstracts male) of whom 53 (37%) received preoperative, 35 (25%) claims data for the 2007 year. Patients with employer-based (minutes) percentage of patients are not provided with antibiotics at postoperative XRT, and 55 (38%) were treated with surgery insurance for 12 continuous months and who were treated Each cohort represents 25 patients. discharge, which helps explain the observed re-admission Sunday alone. Data: median [range] or mean (SD). in the hospital for CDC were analyzed. A principle admis- rates and may indicate a widespread under-treatment of RESULTS: Patients treated without XRT were older com- sion diagnosis of CDC during hospitalization was defi ned RESULTS: 101 patients underwent SILC. 17% of patients the disease. required conversion to standard laparoscopic cholecys- pared to both other groups (66 [38–93] years vs. preopera- as primary CDC (PCDC), while secondary CDC (SCDC) was tectomy. No signifi cant difference was found in operative tive XRT 57 [25–92] and postoperative XRT 60 [37–88]; p < defi ned as contracting CDC while in the hospital for other times between that of standard laparoscopic cholecystec- 0.05) while patients who underwent postoperative XRT had reasons. Two-sample t-tests and z-tests were used for a com- tomy, as well as within the experience-based cohorts (p = the longest follow-up (91 [6–137] months vs. preoperative parison of means, Chi-square tests were used to test asso- 0.152). There was an overall increase in operative time and XRT 47 [1–104] and no XRT 56 [1–134]; p < 0.05). Gender ciations, and logistic regression was conducted to analyze the number of ports needed in patients with acute chole- distribution was comparable between groups (NS) while 30-day and 60-day readmission rates for CDC. cystitis and gallstone pancreatitis, but this did not reach XRT patients had higher tumor stages (p < 0.05). Urinary RESULTS: A total of 4754 CDC patients were identifi ed statistical signifi cance. A reduction in operative time was function was impaired after preoperative XRT compared to (PCDC: 2128 (45%); SCDC: 2626 (55%); p = 005). PCDC shown in residents who were profi cient in standard laparo- patients treated with surgery alone (p < 0.05), while post- patients were slightly younger than SCDC patients (45 scopic cholecystectomy in terms of a learning curve. Oper- operative XRT had no effect. Patients reporting that their ative times remained the same for the teaching surgeon “QOL was worse after APR” showed higher AUASI symp- regardless of experience of resident surgeon. Operative and tom scores (13.2 [8.6] vs. better 7.2 [5.7] and similar 8.8 outcome data were similar between SILC and standard lap- [6.4]; p = 0.02) and AUASI satisfaction scale scores (2.8 [1.2] aroscopic cholecystectomy. vs. better 1.8 [1.5] and similar 2.2 [1.2]; p = 0.01), demon- strating correlation of deteriorated urinary function with CONCLUSION: SILC has a short learning curve for resi- perception of impaired QOL after APR. Pre- and postopera- dent surgeons who are profi cient in standard laparoscopic tive XRT increased sexual dysfunction in female patients surgery. SILC can be effectively taught with few complica- (p < 0.05), while sexual function in male patients was unaf- tions and outcomes similar to standard laparoscopic cho- fected by pre- and postoperative XRT (NS). However, male lecystectomy. Attending surgeons, although in a teaching APR patients had impaired sexual function on all domains operative role, can preserve operative effi ciency and safety of the BSFI compared to previously published, age-stratifi ed while training resident surgeons in SILC.

66 67 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1605 Su1606 patients (23%). HIPEC was performed in 26 patients (87%). CONCLUSIONS: The grading of the primary tumor seems Eight patients (27%) required a stay in the intensive care to be signifi cantly associated to positive surgical resection Laparoscopic Ventral Rectopexy for Rectal Prolapse Safety of Cytoreduction Surgery and Hyperthermic unit (median duration 4 days, range 2–44), and the median following liver resection for colorectal metastases. A higher Using Biological Mesh Intraperitoneal Chemotherapy in Older Patients hospital stay was 8 days (range 4–63). There were 19 grade 3 grading probably determines the presence of micrometas- Pierpaolo Sileri1, Luana Franceschilli*1, Sara Lazzaro1, Susan B. Kesmodel*, Cherif Boutros, Ajay Jain, Daniel Zhu, or higher (National Cancer Institute criteria) complications tases and therefore a higher incidence of positive surgical Giulio P. Angelucci 1, Lodovico Patrizi3, Valeria Fiaschetti2, H. Richard Alexander, Nader Hanna in 14 patients (47%, Table). The most common complica- margins. Emilio Piccione 3, Achille Gaspari1 Surgery, Surgical Oncology, University of Maryland, Baltimore, MD tion was bowel perforation/fi stula (6 patients, 20%). Eight 1 Surgery, University of Rome Tor Vergata, Rome, Italy;2Radiology, BACKGROUND: patients (27%) required reoperation. The overall morbidity Su1608 Cytoreduction Surgery (CRS) with was 63%. Two patients (7%) died of complications directly University of Rome Tor Vergata, Rome, Italy; 3Gynaecology, University Hyperthermic Intraperitoneal Chemotherapy (HIPEC) related to surgery at 5 and 44 days postoperatively. Magnetic Resonance Enterography for Crohn’s Disease: of Rome Tor Vergata, Rome, Italy is being used with increasing frequency to treat patients CONCLUSIONS: What the Surgeon Can Take Home BACKGROUND: with peritoneal carcinomatosis from various malignancies. In this series of patients ≥65 years of age Laparoscopic Ventral Mesh Rectopexy (LVR) Anna Pozza3, Marco Scarpa*1, Carmelo Lacognata2, The safety of this procedure in older patients has not been who underwent CRS ± HIPEC, the morbidity and mortality is a novel procedure to correct internal and external rectal Francesco Corbetti 2 4 5 established. We sought to determine the morbidity and was comparable to that previously reported in the litera- , Claudia Mescoli, Cesare Ruffolo, prolapse. 3 4 3 mortality of CRS and HIPEC in patients ≥65 years of age. ture (major morbidity 12–66%, mortality 1–12%). GivenMauro G. Frego , Massimo Rugge, Romeo Bardini, Several authors have shown that this approach is safe and 3 METHODS: the potential for signifi cant morbidity associated with thisImerio Angriman improves obstructive defecation symptoms and faecal A retrospective review was performed of all 1 treatment, patient selection for this procedure is critical. Department of Oncological Surgery, Venetian Oncology Institute incontinence, without inducing new-onset constipation, patients who underwent CRS ± HIPEC at a single institu- (IOV-IRCCS), Padova, Italy;2 possible after posterior rectopexy. Over the last decade, as tion between 9/1/2004 and 9/1/2010. Data were collected Department of Radiology, Azienda Su1607 3 for other procedures, the use of biological mesh has been on patient and tumor characteristics, operative factors, Ospedaliera di Padova, Padova, Italy; Department of Surgical and extended to correct pelvic fl oor disorders. However litera- and postoperative morbidity and mortality. This study was Liver Resection for Colorectal Metastases: Does PrimaryGastroenterological Sciences, University of Padova, Padova, Italy; ture data is scant. In this study we present our experience approved by the Institutional Review Board. Tumor Grade Predict Positive Surgical Margins? 4Department of Medical Diagnostic Sciences & Special Therapies with this procedure using biological mesh. (Pathology section), University of Padova, Padova, Italy;5Department Marco Massani, Cesare Ruffolo*, Roberta Bonariol, Ezio Caratozzolo, METHODS: Complications of CRS ± HIPEC of Surgery (IV unit), Regional Hospital “Ca’ Foncello”, Treviso, Italy Prospectively collected data on LVR for inter- Luca Bonariol, Francesco Calia Di Pinto, Bruno Pauletti, nal rectal prolapse were analysed. All patients underwent Nicolò Bassi BACKGROUND: Crohn’s disease (CD) is a life long, chronic, preoperative evaluation with defaecating proctography Grade II Department of Surgery (IV Unit), Regional Hospital Cà Foncello, relapsing condition that involves the entire digestive tract and/or pelvic dynamic MRI, full colonoscopy, anal physiol- Complication Type –2 1 3 4 5 Treviso, Treviso, Italy requiring often morphological assessment. MR enterogra- ogy studies, and endo-anal ultrasound. Bowel perforation/fi stula 1 3 2 phy (MRE) offers advantages of not using ionizing radiation A four trocars technique has been used. Briefl y, a veryHemorrhage 1 BACKGROUND: The gold standard of treatment for hepatic and yielding intra luminal and intra abdominal informa- superfi cial peritoneal window was made from the rightNegative re-exploration 1 colorectal metastases is radical surgical resection with R0 tions. The aim of our study was to identify how MRE can sacral promontory, over the right outer border of the meso-Wound 2 1 margins. In fact positive surgical margins are associated to be useful in planning surgical procedures. rectum down towards the right and left side of the deep Partial small bowel obstruction 2 a higher incidence of recurrences and lower survival rates. PATIENTS AND METHODS: In this retrospective study

Douglas pouch. The longitudinal incision was terminated Poster Abstracts Intra-abdominal abscess 1 The primary aim of this prospective study was to evaluate 35 patients who underwent MRE and then surgery for CD as low as possible close to the pelvic fl oor creating behind the presence of predictors of residual neoplastic tissue in the vagina a 4–5 cm pouch. A biological mesh was thenPancreatic fi stula 1 were enrolled from 2006 to 2010. MRE fi ndings were com- Urinary tract infection 1 3 surgical resection margins in patients who underwent pre- pared to intraoperative fi ndings. Histology of operative Sunday sutured anteriorly on the rectum with two parallel rows operative chemotherapy. of interrupted non-absorbable stitches and secured proxi-Pneumonia 2 specimens, systemic infl ammatory parameters (white blood mally on the sacral promontory. Line infection 1 PATIENTS AND METHODS: 87 consecutive patients that cells count, platelets count, CRP, ESR, albumin, iron) and Arrhythmia 2 underwent liver resection for colorectal metastases in our faecal lactoferrin were also evaluated. Cohen’s kappa agree- End-points were to evaluate surgical complications and ment test, sensitivity and sensibility, uni/multivariate logistic functional results of this technique such as changes in bowelNon-infectious pulmonary 2 1 unit from January 2005 till December 2009 were enrolled. The patients were divided into 2 groups: group 1 with 48 regression and non parametric statistics were performed. function (Wexner Constipation Score and Faecal Inconti-Deep venous thrombosis 2 nence Severity Index) at 3 and 6 months. Analysis was per- patients (55%) that underwent liver resection after preop- RESULTS: MRE identifi ed bowel stenosis with a sensitiv- Fall 1 erative chemotherapy and group 2 with 39 patients (45%) formed using Mann-Whitney U-test for unpaired data and Total 11 11 6 2 ity of 0.95 (95% CI 0.76–0.99), a specifi city of 0.72 (95% Wilcoxon signed rank test for paired data (two-sided p-test). that underwent liver resection alone. The following param- CI 0.39–0.92). The concordance of MRE fi ndings with eters were taken into consideration: age, sex, primary RESULTS: Twenty-fi ve consecutive patients underwent intraoperative fi ndings was high (Cohen’s k = 0.72 (0.16). tumor site, type of operation, tumor stage, grading of pri- LVR (median age 61 range 46–78 years, median follow up Abscesses were detected at MRE with a sensitivity of 0.92 mary tumor, surgical margins classifi ed as R0, R1 and R2. (95% CI 0.62–0.99), a specifi city of 0.90 (95% CI 0.69–0.98) was 4 months). Nineteen patients (76%) had a constipation RESULTS: score >5, while nine (36%) a FISI score >10. Three patients Of 175 patients who underwent CRS ± HIPEC, Univariate and multivariate analyses were performed. with a Cohen’s k = 0.82 (0.16). MRE identifi ed bowel fi stu- (12%) had mixed OD and FI. One patient required conver- 30 patients (18 males, 12 females) were ≥65 years of age las with a sensitivity of 0.71 (95% CI 0.42–0.90), a specifi c- RESULTS: Patients that underwent preoperative che- sion to open (3%). Median length of stay was 2 days. We (median 67, range 65–86). Patients were treated for vari- ity of 0.76 (95% CI 0.52–0.90) and with Cohen’s k = 0.47 did not observe mortality or major morbidity. Two patients ous malignancies including appendiceal (9, 30%), meso- motherapy had a higher probability of positive surgical (0.17). The grade of proximal bowel dilatation resulted to experienced UTI (8%). Preoperative constipation (median thelioma (5, 17%), colorectal (5, 17%), ovarian (4, 13%), margins (R1 or R2) and subsequently a non radical opera- be a signifi cant predictor of the possibility of using strictu- Wexner score 14) and faecal incontinence (median FISI gastric (2, 7%), primary peritoneal (2, 7%), endometrial tion compared to patients that underwent surgery alone roplasty instead of/associated to bowel resection either at score 11) improved signifi cantly at 3 months (Wexner 3, (1, 3%), malignant solitary fi brous tumor (1, 3%), and an (OR = 2.05). This result did not reach statistical signifi cance univariate or at multivariate analysis. FISI 4, both p < 0.001). One patient (3%) required Stapled unknown gastrointestinal primary (1, 3%). The American because the incidence of R1 and R2 margins was low in CONCLUSION: Our study confi rmed that MRE fi ndings Transanal Rectal Resection (STARR) for persisting symp- Society of Anesthesiologists score was 3 or greater in 28 both groups (7 patients in group 1 versus 3 patients in correlate signifi cantly with disease activity. Once decided toms 6 months after surgery. No patients had function patients (93%). Median operative time was 426 minutes gruop 2). There was no association between radical resec- that the patient should undergo surgical treatment MRE worsening. No patients complained of sexual dysfunction. (range 150–666) and median blood loss was 800 millili- tion and type of operation (p = 0.25) or between radical can provide the surgeon useful and adequate information CONCLUSIONS: Laparoscopic ventral rectopexy using ters (range 150–3,000). Seventeen patients (57%) required resection and tumor stage. Positive surgical margins cor- about abscess, stenosis and fi stulae. Detailed information biological mesh for internal rectal prolapse is safe and intraoperative blood transfusion (median # units 3, range related signifi cantly with primary tumor grade (Kendall’s about abscess could suggest percutaneous drainage that effective in ameliorating symptoms of obstructed defeca- 1–9). Median number of organs resected was 4 (range 1–9). tau b 0.157; OR 4.8). could ease the following surgery or avoid emergency lapa- tion and faecal incontinence. Completeness of cytoreduction (CCR) was CCR-0 in 18 rotomy. Proximal bowel dilatation can suggest the possibil- patients (60%), CCR-1 in 5 patients (17%), and CCR-2 in 7 ity to perform bowel sparing surgery such as stricturoplasty.

68 69 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1609 surgery for diverticulitis and to examine subgroups of Adjusted Odds Ratios (95% CI) of Outcomes Among Medicare Benefi ciaries Undergoing Elective Diverticulitis Surgery patients with comorbid congestive heart failure (CHF) and Clinical Examination Versus Computed Tomographychronic obstructive pulmonary disease (COPD) to deter- In-Hospital Intestinal Thrombo- in the Diagnosis of Acute Appendicitis mine whether outcomes in these patients are worse. Death Diversion Hemorrhage Wound Pulmonary Renal Cardiac Shock/sepsis embolic Age (n = 329) (n = 3,006 (n = 470) (n = 1,052 (n = 1,782 (n = 614) (n = 594) (n = 495) (n = 259) Rachel Wilson*, Jack Sariego METHODS: Data were derived from the 100% Medicare 70–74 * 1.1 * * * * * * * Surgery, Temple University School of Medicine, Philadelphia, PA Provider Analysis and Review (MEDPAR) inpatient fi les (1.03–1.23) BACKGROUND: from 2004–2007. We used a retrospective cohort study 75–79 2.5 1.4 1.4 * 1.6 1.7 1.5 1.6 1.4 While the rate of negative appendec- (1.71–3.71) (1.28–1.60) (1.09–1.81) (1.37–1.84) (1.33–2.14) (1.14–1.87) (1.21–2.06) (1.01–1.98) tomy seems to be decreasing in the era of multi-slice helical design to identify patients age 65 years and older with a 80–84 3.6 2.2 * 0.72 1.9 1.7 1.8 2.1 * computed tomography (CT scans), controversy still exists primary diagnosis of diverticulitis and who underwent (2.38–5.33) (1.92–2.46) (0.58–0.911 (1.66–2.29) (1.29–2.22) (1.41–2.41) (1.60–2.85) elective left colon resection, , or ileostomy. regarding the necessity of CT scan in the diagnosis of acute 85+ 9.7 3.7 * * 2.8 2.5 2.4 2.9 * appendicitis. We sought to determine the role of CT scan in Patients were then grouped by comorbid CHF and COPD (6.49–14.45) (3.21–4.33) (2.32–3.04) (1.81–3.38) (1.73–3.25) (2.11–4.08) 1.4 2.1 accurately diagnosing acute appendicitis. status. The primary outcome variable was in-hospital mor- COPD * 1.1 * * * * * tality. Secondary outcome variables included intestinal (n = 4,116) (1.00–1.22) (1.17–1. 63) (1.89–2.42) METHODS: CHF 3.5 1.8 1.4 2.0 3.9 3.8 4.2 This retrospective, IRB-approved, HIPAA-com- diversion and predetermined postoperative complication pliant study was performed in a 746-bed tertiary care urban (n = 1,486) (2.71–4.61) (1.58–2.07) (1.02–2.07) (1.55–2.42) (3.39–4.54) (3.03–4.69) (3.40–5.23) rates. We evaluated preoperative clinical characteristics and 3.3 1.7 academic hospital with 65000 annual emergency depart- performed multivariate analysis, adjusting for CHF, COPD, (2.58–4.28) (1.07–2.46) ment visits. Informed consent was waived and the medical age, gender, race, and comorbidities as measured by the Comparison group age 65-69, * indicates not signifi cant records system was used to identify all patients age 18-80 Charlson Index. years who underwent a non-incidental appendectomy dur- ing the 6-year period after the introduction of multi-slice RESULTS: A total of 23,765 individuals were eligible helical CT scanners to our institution. These patients were for inclusion. Mean age was 73.9 years and 71.7% were Su1611 RESULTS: Preoperative risk by CR-POSSUM was rela- correlated with the ICD-9 code for acute appendicitis, and female. Among all patients, 17.3% had COPD and 6.2% tively stable over time, as was BMI and ASA (see table). CR- the radiology database was used to identify which patients had CHF. Overall mortality, morbidity and intestinal diver- Gastrointestinal Surgery Risk Assessment Must FactorPOSSUM scores for obese (BMI ≥ 30) and morbidly obese underwent CT scan within one week of admission. Chi sion rate were 1.4%, 21.6% and 12.6%, respectively. Mor- Into NSQIP, Medicare Public Reporting and Pay-For- (BMI ≥ 35) patients did not vary signifi cantly (P = 0.08 and squared analysis was used to assess for signifi cant differ- tality increased with increasing age in both unadjusted Performance Measures: Bring CR-POSSUM Scoring to0.57, respectively), indicating it is an independent factor. ences in proportions, and a p value of 0.05 was considered and adjusted analysis. On multivariate analysis, patients the U.S. CR-POSSUM scores varied signifi cantly by ASA score: ASA signifi cant. with CHF had signifi cantly increased odds of in-hospital 1–2 mean 4.1 (SD 8.2) versus mean 7.9 (SD 8.1) for ASA 3–4 Anjali S. Kumar*, Lana Bijelic, Kirthi Kolli, Kimberly N. Hoang, RESULTS: mortality (OR: 3.5, 95% CI 2.71–4.61), intestinal diversion patients (P = 0.03). From April 2004 to April 2010, 510 patients (OR: 1.8, 95% CI 1.58–2.07), and all other measured post- Shafi k Sidani, Deborah Schnipper, Thomas J. Stahl meeting our criteria underwent non-incidental appendec- Colon and Rectal Surgery, Washington Hospital Center, Washington, DC operative complications. Patients with COPD did not have Preoperative CR-POSSUM Scores, ASA, and BMI Over 5 Years of a tomies. 427 (83.7%) of these patients underwent CT scan increased odds of mortality, but had signifi cantly increased BACKGROUND: within one week of admission. The overall negative appen- If the future of hospital and physician Systematic Sample of Colon and Rectal Surgery Inpatients odds of intestinal diversion, wound infections, and pulmo- reimbursement of medicare is to be based on pay-for- dectomy rate was 11.6%. In patients who underwent CT nary complications. Poster Abstracts scan, the negative appendectomy rate was 5.6%, whereas performance and/or hospital safety and quality measures CR-POSSUM, Mean BMI, Mean (process and outcome measures that do not factor for risk), YEAR N (Range) ASA, mean (Range) the negative appendectomy rate was 42.2% in patients 2006 36 3.9 (0.3–18.6) 2.25 27.9 (20–49)Sunday who did not get a CT scan (p < 0.001). In males, the nega- gastrointestinal surgeons must demand an integration of patient and procedure-specifi c models into the reimburse- 2007 23 8.0 (0.7–57.6) 2.4 27.2 (19–45) tive appendectomy with and without CT scan was 2.3% 2008 18 5.6 (0.3–22.5) 2.4 28.6 (18–41) and 24.0% respectively (p < 0.001). In females, the negative ment equation. Current risk prediction models utilized 2009 15 4.6 (0.9–22.0) 2.4 27.1 (20–44) appendectomy rates were 9.2% and 50.0% with and with- by the U.S. involve cumbersome equations of a multitude 2010 20 7.4 (0.4–26.0) 2.4 25.5 (17–35) out CT scan respectively (p < 0.001). of variables. An assessment by National Surgical Quality TOTAL 112 5.7 (0.3–57.6) 2.4 27.3 (17–49) Improvement Program (NSQIP) organizers published in (no signifi cant variation over time) CONCLUSION: CT scan signifi cantly reduces the rate of 2008 (JACS. 207:777–782), suggest risk should be accounted negative appendectomy in both males and females. for by collecting 5–10 measures. We propose that the vali- dated British model, CR-POSSUM, for risk stratifi cation in CONCLUSIONS: From this preliminary work, we propose Su1610 colon and rectal surgery can be assessed, possibly in com- that Medicare policy-makers consider this simple, vali- bination with ASA (American Society of Anesthesiologists dated, European-developed risk stratifi cation for patients CHF and COPD Predict Poor Surgical Outcomes in score) and body mass index (BMI), to predict operative risk. undergoing lower gastrointestinal surgery. Obesity is an Older Adults Undergoing Elective Diverticulitis Surgery increasing epidemic in the U.S. and should be given its due METHODS: CR-POSSUM, Physiological and Operative *1,3 1 2,3 2 role in the equation, while ASA could be left out. We will Amy Sheer , Eric B. Schneider, Albert W. Wu, Jodi B. Segal, Severity Score for the enUmeration of Mortality and Mor- 1 present morbidity/mortality data on this group to relate Anne O. Lidor bidity uses 6 physiologic (age, cardiac, systolic pressure, 1 CR-POSSUM scoring to our surgical outcomes. Surgery, Johns Hopkins University School of Medicine, Baltimore, heart rate, hemoglobin and urea level) and 4 operative MD; 2Department of Medicine, Johns Hopkins University School Unadjusted mortality rates for all diverticulitis patients with and without (operation type, peritoneal contamination, malignancy of Medicine, Baltimore, MD;3Department of Health Policy and CHF or COPD stratifi ed by age category, † indicates signifi cance status and operative urgency) parameters to calculate risk Management, Johns Hopkins Bloomberg School of Public Health, of morbidity and mortality from colorectal surgery. From a Baltimore, MD CONCLUSIONS: Our fi nding of signifi cantly worse out- systematic sample of 112 inpatients on the colon and rec- tal surgery service at our institution over a 5-year period, BACKGROUND AND AIMS: Diverticulitis is a common comes for elective surgical treatment of diverticulitis in older adults, and especially among those with CHF and we calculated CR-POSSUM scores using the web-based medical condition which disproportionally affects older algorithm [www.riskprediction.org.uk]. We also noted ASA adults and which accounts for over 300,000 hospitaliza- COPD, suggests that a reappraisal of the proper role of elec- tive surgery in this population may be warranted. score and BMI. STATA9 was used to calculate differences in tions yearly in the United States. This study aims to inves- means using two-sided t-tests. tigate outcomes among older patients undergoing elective

70 71 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1612 Clinical: Esophageal Su1614 Su1615 Ileocecectomy for Crohn’s Disease (CD): Which Factors Effi cacy and Durability of Laparoscopic Heller Advanced Esophageal Carcinoma: Is It Still Worth to Augment Intraoperative Small Bowel Preservation? Su1613 Myotomy: Patient Symptoms and Satisfaction at Operate? Liliana Bordeianou*1, Richard A. Hodin1, Abdulmetin Dursun1, Long Term Follow Up Dean Bogoevski*1, Matthias Reeh1, Maximilian Bockhorn1, Laparoscopic Heller’s Myotomy and Fundoplication Joshua R. Korzenik 1 2 2 2 Alexandra M. Koenig 1 1 2 , Vanessa P. Ho, Toyooki Sonoda, Sang Lee, in Patients with Massive Dilated Megaesophagus John G. Linn*, Anthony N. Chan, Sarwat Ahmad, Peter Muscarella, , Asad R. Kutup, Thomas Roesch, Sharon L. Stein 3 W. S. Melvin, Kyle A. Perry Jakob Izbicki 1 1 *1 2 1 Surgery, Massachusetts General Hospital, Boston, 2MA;Colorectal Carlos Pantanali, Fernando A. Herbella, Maria A. Henry, Center for Minimally Invasive Surgery, The Ohio State University 1 General, Visceral and Thoracic Surgery, University Clinic Hamburg- Jose F. Farah 1 3 1 Surgery, Cornell University, New York, NY; 3Colorectal Surgery, Case , Marco G. Patti, Jose C. Del Grande Medical Center, Columbus, OH Eppendorf, Hamburg, Germany;2Clinic for Interdisciplinary 1 Western University, Cleveland, OH Department of Surgery, Federal University of Sao Paulo, Sao Paulo, Endoscopy, University Clinic Hamburg – Eppendorf, Hamburg, 2 INTRODUCTION: Laparoscopic Heller myotomy with Brazil; Department of Surgery, State University of Sao Paulo, Germany INTRODUCTION: Preservation of small bowel is a cru- partial gastric fundoplication has become the standard Botucatu, Brazil;3 Department of surgery, University of Chicago, cial element in surgery for patients with CD, yet it is not treatment for achalasia. While this procedure has demon- OBJECTIVE: Chicago, IL The role of surgical therapy in patients with known whether perioperative factors or preoperative thera- strated excellent short term outcomes, there is a paucity locally advanced esophageal cancer is still controversially pies infl uence specimen length. Our aim was to utilize a INTRODUCTION: Laparoscopic Heller’s myotomy and of data regarding long term patient symptoms and satis- discussed. There is also controversy about whether neo- large multicenter database of CD patients treated with ileo- fundoplication is considered the treatment of choice for faction after the operation. We report a single institution adjuvant chemo or radio-chemotherapy should be the cecectomies (IC) to determine predictors of length of small non-advanced achalasia. The optimal treatment for end- series of minimally invasive Heller myotomy with long standard management in patients with locally advanced bowel resection. stage achalasia with esophageal dilation is still debatable. term evaluation of gastroesophageal refl ux (GERD) symp- esophageal carcinoma. Furthermore, many gastroenterolo- METHODS: Retrospective analysis of prospectively col- AIMS: This study aims to evaluate in a multicenter and toms, dysphagia, and patient satisfaction. gists and oncologists believe that surgery should be avoided lected data on patients with CD who underwent IC between retrospective study the outcomes of patients with mas- METHODS: A retrospective review of a prospectively- in locally advanced esophageal cancer due to high mortal- 9/1993–10/2010 at two academic centers was performed. sive dilated esophagus submitted to laparoscopic Heller’s collected database was conducted for patients undergo- ity and morbidity rates related to the procedure and the T-tests were used to determine whether the mean length of myotomy. ing laparoscopic Heller myotomy from 1995–2006 under very low benefi t for the patient. MATERIAL AND METHODS: small bowel resection (length of colon excluded) was modi- METHODS: 11 patients (mean age 56 years, 6 men) with an institutional review board approved protocol. Long Retrospective analysis of fi ed by patient demographics, preoperative medical treat- massive dilated megaesophagus (maximum esophageal term follow-up evaluation was performed by mail or tele- prospectively collected data of 256 patients with locally ment, type of CD, or surgeon expertise. A Multiple Linear diameter >10 cm) underwent a laparoscopic Heller myot- phone questionnaire. Outcomes included operative data, advanced esophageal cancer (220 patients with pT3 and 36 Regression model was fi tted to account for confounders omy and Pinotti fundoplication between 2000 and 2009 treatment for recurrent dysphagia, GERD symptoms, and patients with pT4) that were not neoadjuvantly treated and and to identify predictors of length of resection. at 3 different institutions. Preoperative workup included patient satisfaction with their operation. Post-myotomy had surgical resection with curative intend. One-hundred- RESULTS: 269 CD patients (51% female, mean age 39 y) upper digestive endoscopy, esophagram and esophageal refl ux symptoms were assessed using the validated GERD- sixty-one patients underwent extensive Ivor-Lewis thoraco- were included. On univariate analysis, the mean length manometry in all patients. Symptoms were evaluated at HRQL instrument. A score greater than 20 was considered abdominal esophagectomy (TAE) whereas 95 had limited of small bowel resection (20.8 cm, SD 17.18 cm) was not the last follow-up. indicative of signifi cant refl ux symptoms, as this correlated transhiatal (TH) resection. Locally advanced esophageal with patient dissatisfaction. affected by age (p = 0.23), sex (p = 0.12), history of prior RESULTS: cancer was defi ned, based on the fi nal histological report

On follow-up (mean 29, range 3–81, months), Poster Abstracts resections (p = 0.12), emergency surgery (p = 0.23), pres- RESULTS: 56 patients underwent primary laparoscopic of the resection specimen, as a tumor infi ltrating the parae- postoperative complaints were mild and occasional dys- sophageal tissue or the adjacent structures, without respect ence of fi brostenotic (p = 0.51), penetrating (p = 0.78) or phagia to solid food in 4 (36%), severe dysphagia in 3 (27%) Heller myotomy during the study period. At long term active disease (p = 0.97). Patients with microscopically follow-up, 7 patients were deceased and follow-up was of the lymph node affection, the distant metastases or the Sunday and absence of dysphagia in 4 (36%) patients. All patients histological grading. positive margins were not spared bowel length (22.5 vs. gained weight except for the 3 patients with severe dys- obtained in 29. Median follow up interval was 6 years RESULTS: 19.8 cm, p = 0.34). Patients with suspected malignancy phagia. The 3 patients with severe dysphagia underwent (range 4–14 years). All operations were completed laparo- Complete resection (R0) was achieved in 74.5% had wider resection margins(54.7 vs. 19.6 cm, p = 0.01). esophageal dilatation (n = 2) or laparoscopic esophagec- scopically, and 18 utilized robotic assistance. All patients of patients that underwent TA esophagectomy and in 5-ASA (p = 0.54), steroids (p = 0.51), azathioprine (p = 0.59) tomy (n = 1) and currently complain of occasional dyspha- reported relief of dysphagia postoperatively. At long term 61.1% of the patients with TH resection (p = 0.016). The and TNF agents (p = 0.13) did not augment length of resec- gia and gained weight. follow up, 72% of patients had received no further treat- median lymph node yield in TAE was signifi cantly higher tion, however exposure to 6-MP within 3 months of sur- ment for dysphagia. Of those with recurrent dysphagia, 2 CONCLUSION: (25 lymph nodes, range 2-89) than in patients operated TH gery decreased resection length (17.1 vs. 22.4, p = 0.03). Heller’s myotomy and fundoplication patients underwent Botox injection, 5 underwent pneu- (14 LN, range 2–100; p = 0.008), although no benefi t for Surgeons with expertise in CD were more likely to salvage relieves dysphagia even in patients with massive dilated matic dilatation, and 1 required a second esophageal myot- overall survival was found for patients with radical lymph- bowel (19.9 vs. 29.5 cm; p = 0.0008). On multiple linear esophagus. omy. All patients reported adequate relief of dysphagia adenectomy (lymph node yield of 19 or more – median 9 regression history of prior resections (p = 0.0001), sus- after repeat intervention. 55% of patients use acid-reduc- months vs. lymph node yield of 18 or less – median 10.8 pected malignancy (p = 0.02) and surgery by a non-expert ing medications to control GERD symptoms; however, months; p = 0.480). Patients with locally advanced esoph- (p = 0.0004) were predictive of longer resections. only 10% of patients reported Velanovich scores indicative ageal cancer but without evidence of tumor rest disease CONCLUSIONS: Length of small bowel resection during of severe GERD that impacts patients’ overall satisfaction (pT3 & pM0 & R0, irrespective of the lymph node status) ileocolic resection for CD is most affected by surgical exper- with their operation. When asked to reconsider their acha- had similar overall survival (median 23.7 months, 5-Y tise, concerns for malignancy and history of prior resec- lasia treatment, 97% of patients would choose laparoscopic of 22.4%) as pT2 without evidence of tumor rest disease tions. Various preoperative medical regimes do not appear Heller myotomy again. (median 33 months, 5-Y of 27%; p = 0.152). The operative to have an effect on ultimate resection length. CONCLUSION: Laparoscopic Heller myotomy provides method had signifi cant infl uence on the disease free sur- durable, effective, long term dysphagia relief in the vast vival (TAE-median 12.1 months, 5-Y 22.9%; or TH - median majority of patients. At a median follow-up interval of 6 10 months, 5-Y 8.9%; p = 0.049, data not shown). years, patients remain highly satisfi ed with the operation. DISCUSSION: Our results in the treatment of the patients While many patients report mild refl ux symptoms, up to with locally advanced esophageal carcinoma (median 13.7 10% report symptoms indicative of signifi cant GERD that months, 5-Y of 14.4%) are comparable to the results of the negatively impacts overall satisfaction with the operation. patients neoadjuvantly treated with chemoradiotherapy (median 10 to 14 months; 5-Y of 19-23%).

72 73 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1616 Su1617 Su1618 Su1619 Dysphagia After Esophagectomy for Esophageal CancerThe Myotomy Length on the Gastric Side Doesn’t Prognostic Factors for Adenocarcinoma of Revisional Surgery After Failed Esophagogastric Alexandra Koenig*, Dean Bogoevski, Maximilian Bockhorn, Infl uence the Final Outcome of Laparoscopic Heller Esophagogastric Junction Myotomy for Achalasia: Successful Esophageal Matthias Reeh, Yogesh K. Vashist, Thomas Roesch, Emre F. Yekebas,Dor for Esophageal Achalasia Fion S. Chan*, Daniel K. Tong, Kam H. Wong, Simon Law Preservation Jakob Izbicki Valentina Caruso1, Renato Salvador*2, Mario Costantini2, Surgery, University of Hong Kong, Hong Kong, Hong Kong Ross F. Goldberg*, Steven P. Bowers, Michael Parker, John Stauffer, University Medical Centre of Hamburg-Eppendorf, Hamburg, Germany, Lisa Zanatta 2, Nicola Passuello2, Loredana Nicoletti2, Francesco INTRODUCTION: Ronald A. Hinder, Horacio J. Asbun, C. Daniel Smith Hamburg, Germany The incidence of Barrett’s esophagus is Cavallin 2, Ermanno Ancona2, Giovanni Zaninotto1 low in Chinese population. Most of our patients who have Department of Surgery, Mayo Clinic - Florida, Jacksonville, FL OBJECTIVE: 1 To assess the impact of site of anastomosis Department of General Surgery, SS Giovanni e Paolo Hospital, adenocarcinomas of the esophagogastric junction (AEG) INTRODUCTION: Treatment failure with recurrent dys- on dysphagia in esophageal carcinoma. ULSS, Venice, Italy;2Department of Surgical and Gastroenterological are Siewert type II or III. This study aims to evaluate the phagia after Heller myotomy occurs in fewer than 10% of outcome after surgical resection in this group of patients, SUMMARY BACKGROUND DATA: As overall survival Sciences, Clinica Chirurgica I, University of Padova, Padua, Italy patients, most of whom will seek reoperation. There exist comparing clinicopathological differences between type II after esophageal cancer surgery remains poor, postopera- BACKGROUND: only limited reports of reoperation with esophageal preser- The controversy about the myotomy and III cancers, and identify prognostic factors. tive quality of life has gained importance as an additional length on the gastric side for esophageal achalasia is still an vation in such patients. MATERIAL AND METHODS: outcome parameter. While the two most commonly used unexplored fi eld. The aim of this study was to investigate the Patients who underwent METHODS AND PROCEDURES: We retrospectively reconstructive methods, intrathoracic and collar anas- fi nal outcome after classic myotomy (CM) vs long myotomy resection for AEG tumors between 1995 and 2008 were reviewed the records of patients presenting for symptom- tomosis, have shown similar oncologic effectiveness, it on the gastric side (LM) in two cohort of achalasia patients. included. Those with Siewert type I cancers were excluded. atic treatment failure after Heller myotomy. From March remains undecided whether subsequent quality of life is Data were retrieved from a prospectively collected data- PATIENTS AND METHODS: We evaluated 44 achalasia 1998 to November 2010, 47 patients were evaluated. Seven different. The present analysis compares these reconstruc- base. Patient characteristics and clinicopathological data patients who underwent laparoscopic Heller Dor. Patients patients had undergone more than one prior myotomy. tion methods, focusing on dysphagia as the main postop- and outcome were evaluated. with sigmoid-shaped esophagus were excluded. Symptoms Three patients became symptom-free after endoscopic dila- erative symptom. RESULTS: were scored using a detailed questionnaire for dysphagia, There were 126 patients (99 men, 27 women). tion and/or Botox injection therapy, and seven patients METHODS: Between 2003 and 2007, 71 patients (mean regurgitation, and chest pain; barium swallow, endoscopy The median age was 70 yrs (range: 23–87). Type II tumors opted not to undergo reoperation. Thirty-fi ve of 37 age 61.5 years; 72% male, 28% female) with complete and esophageal manometry were performed, before and 6 were found in 65 patients and type III in 61. Thoracotomy patients underwent reoperation with the goal of esopha- resection of esophageal carcinomas (37 adenocarcinomas, months after surgical treatment. 24 hours pH-monitoring was required for tumor extirpation in 50 (76.9%) and 34 geal preservation. Two patients with sigmoid megaesopha- 34 squamous cell cancers) and with long-term survival of at were performed 6 months after the Heller Dor. (55.7%) patients, respectively (p = 0.014). The operative gus underwent minimally invasive esophagectomy as our least 18 months were contacted and completed a modifi ed blood loss was higher for type III tumors (median 300 ml initial reoperation. Our approach was to proceed with a CM was defi ned as gastric myotomy length between 1.5– quality of life (QoL) questionnaire (median 24.1 months vs. 400 ml, p = 0.005). There was no hospital mortality. takedown of their previous fundoplication, dissection of 2.5 cm and LM as 2.5–3 cm gastric myotomy length. The after surgery). Our analysis compared the reconstruction Type III AEG was associated with signifi cantly higher inci- the periesophageal fi brosis, and assessment of need for an surgical treatment (CM or LM) were performed in two con- groups using a gastric tube with either collar (group A; n dence of poorly differentiated cancer (44.6% vs. 73.8%, p = extension of the myotomy. secutive cohorts. Treatment failure was defi ned as a postop- = 36) or high intrathoracic anastomosis (group B; n = 35). 0.003), higher number of nodal metastasis (median 3 vs. 7, RESULTS: erative symptom score >10th percentile of the preoperative Thirty-fi ve patients underwent laparoscopic In order to quantify the given answers from the patients p = 0.031), and advanced TNM stage (stage IIIA–IV disease score (i.e., >7). reoperation for either presumed incomplete myotomy or and the extent of the dysphagia, we have devised a scoring in 64.6% vs. 86.9%, p = 0.004). Overall median survival Poster Abstracts RESULTS: anatomic distortion due to the fundoplication. Intraopera- system. For each question we have assigned a point alloca- 44 patients (M:F = 24:20) represented the study for the whole group was 17 months. Patients with type II tive fi ndings were incomplete myotomy in 28 patients and tion, depending on the impact of the symptoms or grade population: 20 patients underwent CM and 24 patients AEG had longer survival (37.6 vs. 10.4 months), (p < 0.01). fundoplication failure in 7 patients. Intraoperative esopha- Sunday of dysphagia. had LM. Demographic and clinical parameters (age, sex, Independent variables identifi ed by Cox regression model gogastric perforation occurred in six patients. Fundoplica- symptom score, duration of symptoms, esophageal diam- for better survival were Siewert type II (p = 0.021), earlier RESULTS: Postoperatively, the rate of surgical compli- tion was not reconstructed in 15 patients. Of the 31 patients eter and manometric pattern) were similar among the N-stage (p < 0.01) and R0 resection (p < 0.01). cations of our study population was 27.8% (anastomotic undergoing a fi rst-time reoperation, 24 achieved relief of two groups. Median follow-up was 18 months (IRQ 8–33). leakage 22.2%) in group A vs. 11.4% in group B (p = 0.075). CONCLUSION: Type III AEG tumors were more likely to symptoms without re-intervention (77%). Less than half Mucosal tears were nil in both groups. One patient per The long-term follow-up showed symptoms of dysphagia be poorly differentiated, more advanced and had worse of patients undergoing two or more reoperations had suc- each group was considered a failure. In the CM group the in 29 patients (group A, n = 20; group B, n = 9; p = 0.007). survival compared to type II AEG cancers. Other indepen- cessful relief of symptoms. However, esophageal preserva- median symptom score decreased from 19.5 (IQR 13.5– Signifi cantly more patients within group A had to undergo dent prognostic factors were N-stage and R-category of tion was possible in 32 of the 35 patients in whom it was 20.5) to 0 (IQR 0–4) (p < 0.0001). The median LES resting endoscopic bougienage (13 vs. 1, p < 0.0001). The scoring resection. attempted (91%). Three patients failed a strategy of esopha- pressure dropped from 23.5 mmHg (18–30) to 10 mmHg system showed signifi cantly severe symptoms in patients geal preservation and eventually required esophagectomy. (7–15) (p = 0.006) and the median LES residual pressure with collar anastomosis (median 17) compared to intratho- CONCLUSIONS: from 11 mmHg (8–17) to 1.5 mmHg (0.4–3) (p = 0.002). In Laparoscopic reoperation with esopha- racic anastomosis (median 8). the LM the preoperative symptom score median was 16.5 geal preservation is successful in the majority of patients CONCLUSION: High intrathoracic anastomosis appears to (11–21.5) vs 3 (0-5) (p < 0.0001) of the postoperative evalu- with recurrent dysphagia after Heller myotomy. Not sur- carry a lower risk for dysphagic symptoms compared with ation. The median LES resting pressure decreased from 28 prising, there is a decrement in the rate of success with collar anastomosis, and should therefore be the preferred mmHg (25.6–66.8) to 10.3 mmHg (7–18) (p < 0.0002) and each successive reoperation. method for reconstruction after surgical resection of esoph- the median LES residual pressure from 15 mmHg (7–39.5) ageal carcinoma. to 3.7 mmHg (1.3–7) (p = 0.0005). Post-operative symptom score, resting and residual LES pressure, total and abdominal LES length were not statis- tically different in the two groups. Positive post-operative 24 hours pH-monitoring were similar in the two groups. Seven patients changed the manometric pattern after sur- gery (2 CM and 5 LM). CONCLUSION: Extending the myotomy length into the gastric side over 2.5 cm doesn’t change the fi nal outcome of laparoscopic Heller Dor.

74 75 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1620 Su1621 Su1622 Su1623 Endoscopic Stapling System for Trans Oral TreatmentNeoadjuvant Chemo-Radiotherapy Modifi es the Incidence and Resolution of Anemia with ParaesophagealThe Hypertensive Upper Esophageal Sphincter Is of GERD: Three Years Follow Up Histologic Grade of Esophageal Cancer Hernia Repair Related to GERD and Is Improved by Antirefl ux Aviel Roy-Shapira*1,2, Amol Bapaye3 Renato Salvador1, Gianpietro Zanchettin*1, Mantoan Silvia1, Chady Haurani*, Arthur M. Carlin, Zane Hammoud, Kellie Mcfarlin,Surgery 1 Surgery A, Soroka Univesity Hospital, Beer Sheva, Israel;2Faculty Luca Faccio 1, Angela Pecchielan1, Antonio Rella1, Madhu Prasad, Vic Velanovich Konstantinos I. Makris*, Andrew S. Kastenmeier, Daniel Davila, of the Health Sciences, Ben Gurion University, Beer Sheva, Israel; Francesco Cavallin 2, Michele Valmasoni1, Carlo Castoro2, Surgery, Henry Ford Hospital, Detroit, MI Lee L. Swanstrom, Christy M. Dunst 3 2 2 1 1 Deenanath Mangeshkar Hospital and Research Center, Pune,Matteo India Cagol , Rita Alfi eri, Ermanno Ancona, Alberto Ruol BACKGROUND: Paraesophageal hernias may produce a Minimally Invasive Surgery, Legacy Health System, Portland, OR 1 OBJECTIVES: Long Term Effi cacy of an endoscopic stal- Department of Surgical and Gastroenterological Sciences, Clinica variety of clinical sequelae. One of these is anemia, which BACKGROUND: Hypertension of the upper esophageal 2 ing system for trans oral treatment of GERD Chirurgica I, University of Padova, Padua, Italy;Surgical Oncology, may or may not be associated with esophageal or gastric sphincter (UES) is frequently encountered on manometries Istituto Oncologico Veneto, IOV-IRCCS, Padova, Italy ulcers. The purpose of this study was to assess the incidence INTRODUCTION: Between May and October 2007, an IRB performed for evaluation of foregut symptoms. Although of anemia in patients with paraesophageal hernias and fre- approved, pilot study of a new endoscopic stapling device BACKGROUND: The introduction of the number of nodal the clinical signifi cance of a hypertensive UES is variable, quency of resolution of anemia with hernia repair. for the treatment of GERD was conducted on 13 subjects in metastasis, grading and tumor location in the 2010 TNM it has been suggested that it may be a result of esophageal Pune, India. Subjects with history of PPI use >2 y for GERD staging could redefi ne the management of esophageal can- METHODS: Patients undergoing all types of laparoscopic acid exposure in patients with gastroesophageal refl ux dis- and no co-morbidity were included. The device is a modi- cer. In early stages (T1a/T1b and T2a/T2b) histologic grade or open paraesophageal hernia repairs from 7/96 through ease (GERD). There is no evidence in the literature to sup- fi ed gastroscope, which includes a surgical stapler, that modulates stage grouping and prognosis. The aim of the 9/10 were eligible for the study. Data gathered included age, port this theory. We hypothesized that successful control fi res a staggered quintuplet of standard titanium B shaped study was to investigate the hypothesis that neoadjuvant gender, type of paraesophageal hernia, presence of anemia of acid refl ux with surgery in patients with GERD may lead 4.8 mm staples, and an ultrasonic range fi nder. All proce- Chemo or Chemo-RadioTherapy (nCRT) may modify the (defi ned as hemoglobin level <13.5 gm/dl or symptomatic to resolution of UES hypertension, if it is a refl ux-related dures were done under general anesthesia by a single oper- grading of esophageal cancer. drop in hemoglobin), presence of symptomatic anemia or phenomenon. METHODS: ator. Either 2 or 3 staple quintuplets were used to staple the PATIENTS AND METHODS: We evaluated 463 patients specifi c surgical referral for anemia from a paraesophageal A prospectively maintained database of fundus to the esophagus, creating a 90–180 degree anterior who underwent surgical resection for esophageal cancer or hernia, presence of esophageal or gastric ulcers or erosions patients undergoing antirefl ux surgery at our institution fundoplication over the distal 2–3 cm of the esophagus. esophago-gastric junction: 252 had surgery (SURG) as fi rst by endoscopy, type of repair, and resolution of anemia. was retrospectively reviewed for the years 2006 to 2009. This is a report of the results of a three year follow up on treatment and 211 had nCRT before surgery. Pathological RESULTS: 187 patients underwent paraesophageal hernia Only patients with hypertensive UES (>120 mmHg), preop- this group of subjects fi ndings were evaluated from the pre-treatment endoscopic repair, of these 66 (35%) were anemic. Of these anemic erative objective fi ndings of GERD and subsequent primary METHODS: The original Informed consent specifi ed that biopsies and the surgical specimen. Patients who had R1-R2 patients, 37 (56%) were symptomatic from their anemia or antirefl ux surgery were included in our analysis. Preopera- the subjects may be contacted annually for 5 years follow- resection or a pathological stage M1 were excluded. The specifi cally referred for anemia, and 20 (30%) had esopha- tive UES pressure was compared to postoperative UES pres- ing the study. Accordingly subjects were contacted for a change of histologic grade (i.e. shift from well-differentiated geal or gastric ulceration/erosions. All patients underwent sure, while resolution of refl ux was confi rmed with 24-hour telephone interview during the fi rst week of October 2010. to less well-differentiated and vice versa) were retrospec- paraesophageal hernia repair, with 56 having documented pH study. The following data were collected: Velanovich GERD-HRQL tively investigated analyzing prospectively collected data. follow-up. Overall, of these, 33 (59%) had resolution of RESULTS: Thirty-three patients met the inclusion criteria. scores, PPI use, symptoms, satisfaction with the procedure, RESULTS: The histological type was adenocarcinoma (AK) their anemia. 68% of symptomatic patients had resolu- Thirty-one of those had normalization of their postopera- and wllingness to repeat the procedure again. in 170 pts (123 SURG/47 nCRT) and squamous cell carci- tion of their anemia, compared to 48% of asymptomatic tive pH and two had reduction without complete normal- Poster Abstracts patients (p = 0.1). Of patients with esophagogastric ulcer- RESULTS: 11 of the 13 subjects could be reached by phone. noma (SCC) in 293 pts (129 SURG/164 nCRT). Median age ization. The mean preoperative UES pressure was 154.3 ation/erosion, 85% were symptomatic and 94% had reso- Sunday GERD-HRQL scores were less than 9 or less in 10 subjects was statistically different between SURG 60.4 years (IQR mmHg (range 121.7–250.3) (normal 30–120 mmHg) and lution of anemia, compared to 53% of patients without and 15 in one subject, The latter subject improved his score 53.9-66.3) and nCRT 65.6 years (IQR 58-71.2)(p < 0.0001), decreased to a mean postoperative UES pressure of 122.9 ulceration/erosion (p = 0.003). from 29 to 15, in all others score improved by more than while the sex distribution was similar in the SURG and mmHg (range 52.9–262.7) (p < 0.05). The statistical signifi - 50%. All subjects would have agreed to do the procedure nCRT. Histologic grade was unchanged in 172/252 (68.3%) CONCLUSIONS: Anemia is a common fi nding in patients cance remained after excluding extremes. Twenty-seven again. Mean satisfaction score was 7.7 (6–10) on a scale in SURG group and 63/211 (29.9%) in nCRT group (p < with paraesophageal hernia. Most of these patients were patients had a decrease of the UES pressure (of which 18 of 1–10. There was no dysphagia. Three subjects resumed 0.0001). The shift from less well-differentiated to well- symptomatic because of their anemia. Those patients with had complete normalization), whereas 6 patients were PPI intake, (compared to 2 at the 2y follow up) 3 subjects differentiated grade was statistically different between the esophageal or gastric ulceration or erosion were very likely to found to have increased UES pressure after surgery. Both require PPI at a reduced dose 1 subjects takes PPI only after two groups: 9.5% (24/252) for SURG group and 64% (135/ have symptomatic anemia, and, interestingly, these patients patients with abnormal postoperative pH studies had per- a large meal and 4 subjects remained completely off PPI. 211) for nCRT group (p < 0.0001). The shift from less well- were more likely to have their anemia resolve with paraesoph- sistent hypertension of their UES. The mean length of time differentiated to well-differentiated grade wasn’t correlated ageal hernia repair. Patients with asymptomatic anemia or no between surgery and postoperative manometry was 267 CONCLUSION: At 3 years, the procedure remained effec- with the histological type: 9/123 AK (7.3%) and 15/129 ulceration/erosion were less likely to have resolution of their days. tive in improving the quality of life in moderate to severe SCC (11.6%) for the SURG group and 28/47 AK (59.6%) anemia with paraesophageal hernia repair. CONCLUSION: GERD without causing dysphagia. PPI use was eliminated Patients with GERD and a hypertensive and 107/168 SCC (65.2%) for the nCRT group. or reduced in 73% of subjects. All subjects remain satisfi ed UES, who undergo anti-refl ux surgery, demonstrate a sig- with the procedure and would do it all over again. Further CONCLUSIONS: nCRT can modify the histologic grade of nifi cant reduction of their UES pressure. This may be studies are necessary to validate these data and determine both AK and SCC of the esophagus. This fi ndings, based considered supportive of the suggestion of an etiologic optimal staple placement. on the 2010 TNM classifi cation, affects stage grouping and link between GERD and UES hypertension. Although the prognosis. mechanism of this association is unclear, it appears to be reversible with anti-refl ux surgery. Our results warrant fur- ther investigation in a prospective fashion for confi rmation and assessment of the potential clinical impact.

76 77 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1624 PPI-responsive symptoms. In addition, antirefl ux surgery underwent LNF without perioperative complications by ity and mortality. Laparoscopic liver resection (LLR) is gain- (Nissen 6, esophagojejunostomy 1) was performed on 7 a single surgeon (ARDLM) in a standard fashion using 5 ing acceptance as a safe method of resection. LLR of the Positional Effects on the Gastroesophageal Junction upright refl uxers with objective evidence of GERD and a upper abdominal ports. A 2 cm loose fundoplication cali- cirrhotic liver is technically challenging, but may result in and Clinical Presentation: Isolated Upright Refl ux vs.negative DeMeester score, all of whom showed complete brated over a 36 French bougie was performed with exten- less post operative morbidity, shorter OR time, and shorter Combined/Supine Refl ux symptomatic relief postoperatively. sive transhiatal mobilization of the esophagus and division length of stay (LOS). Our center is now applying this tech- Toshitaka Hoppo*, James D. Luketich, Blair A. Jobe CONCLUSION: Isolated upright refl uxers have lower LESP of short gastric vessels. nique to larger HCC than generally reported. Cardiothoracic Surgery, University of Pittsburgh Medical Center, and shorter LES length in upright than in supine position. Successful symptom control refl ux after LNF was achieved OBJECTIVE: To assess our initial experience using the fi rst Pittsburgh, PA Symptomatic isolated upright refl uxers are highly associ- at 1 and 3 months for all but one patient (92%). Patients 11 laparoscopic cases of liver resection in cirrhotic patients ated with a pathological GERD and would benefi t from declared their symptoms improved in 8/14 (57%), nor- matched with controls of open liver resection (OLR). BACKGROUND AND AIM: The effects of body position on antirefl ux surgery even if pH testing is negative. malized in 4/14 (28%) and unchanged in 2/14 (14%) at the antirefl ux barrier are poorly understood. The purpose METHODS: We evaluated resections performed for HCC 6 months. The 24-h MII-pH confi rmed the postoperative of this study was to assess the effect of body position on in patients with cirrhosis within a larger laparoscopic series. reduction of acid and non-acid refl ux episodes at 1, 3 and lower esophageal sphincter structure and clinical presenta- Su1625 Each was matched to 3 open cases for number of segments 6 months (Table 1). tion in two different patterns of refl ux: isolated upright and The Effect of Laparoscopic Nissen Fundoplication removed, demographics, co-morbidity indices, and back- ground liver histology. Non-parametric statistical analyses combined/supine. (LNF) on Acid and Non Acid Refl ux: A prospective Baseline 1 Month 3 Months 6 Months were used to compare surgical outcomes. Model for end- METHODS: Patients with typical and/or atypical symp- evaluation at 1, 3, and 6 months using 24-h MII-pH Parameter (n = 14) (n = 13) (n = 13) (n = 12) stage liver disease (MELD) and the Charslon co-morbidity toms were referred for the evaluation of GERD and under- pH-Multichannel Intraluminal Impedance (MII-pH) % time pH < 4 9.05 ± 5 0.26 ± 0.33* 0.4 ± 0.8* 0.3 ± 1.2* Refl ux episodes 4.7 ± 3 0.2 ± 0.44* 0.44 ± 0.98* 0.74 index± .82* was applied retrospectively. Analyses were performed went high-resolution manometry (HRM), in which 10 Antonio Ramos-De La Medina*1, Maura Torres-Aguilera2, >5 min including and excluding converted cases. swallows with 5ml water were delivered in the supine posi- Longest refl ux 33.9 ± 28 2.26 ± 2.78* 3.16 ± 4.2* 2.85 ± 3.1* 1 2 tion; this was followed by 5 swallows in the upright posi-Maria Fernanda Gonzalez-Medina , Alberto Aviles-Calderon, episode (min) RESULTS: Between October 2007 and May 2010, 11 cir- tion. Subsequently, a 48-hour wireless pH test or 24-hourFederico B. Roesch 2, Jose M. Remes Troche2 Number refl ux 68 ± 13 2.22 ± 1.78* 1.5 ± 2.23* 1.8 ±rhotic 4.7* patients underwent LLR for HCC. This included 2 multichannel intraluminal impedance-pH (MII-pH) was1 Gastrointestinal Surgery, Veracruz Regional Hospital, Boca del Rio, episodes right hemihepatectomies, 1 left hemihepatectomy, 7 left DeMeester score 32 ± 17.8 1.51 ± 1.18* 2.4 ±3.08* 1.98 ± 2.1* performed. The pH probe was placed 5 cm proximal to the Mexico;2 Medical-Biological Research Institute, University of Veracruz, lateral sectionectomies and one trisegmentectomy (seg- % bolus exposure 5.3 ± 3.2 0.30 ± 0.39* 0.04 ± 0.8* 0.8 ±ments 0.51* 5, 6, 7). Demographic characteristics were simi- upper border of the lower esophageal sphincter in both Veracruz, Mexico All refl ux episodes 86 ± 42 44 ± 24* 34 ± 12* 30 ± 16* testing modalities. Isolated upright refl ux was considered total lar between groups. Intra-operatively, there were no INTRODUCTION: In well-selected cases laparoscopic present when the supine fraction of time pH <4 was 0; com- Acid refl ux 53 ± 33 4.18 ± 3.9* 2 ± 2.82* 2.4 ±differences 3.1* in estimated blood loss, OR time, or transfusion bined/supine refl ux was present when the supine fraction Nissen fundoplication (LNF) provides symptomatic con- episodes total requirement. 3 cases were converted to open (27%); 2 for of time pH <4 was equal or greater than upright fraction trol as high 90% in GERD. Recently, the use of combined Non-acid refl ux 33 ± 30 19 ± 22* 17 ± 4* 19 ±bleeding 21* and 1 for anatomical uncertainty. All cases were episodes total time. Patients in who both the supine and upright fraction 24-h pH-multichannel intraluminal impedance (MII-pH) performed for HCC with no positive margins. Median mar- has allowed studying non-acid refl ux (NAR) episodes, an % Symptom Index 72 ± 31 22 ± 18* 16 ± 6* 25 ± 14* time pH <4 were 0, were excluded. Clinical presentation of (heartburn) gin distance was similar between LLR and OLR groups at important factor in refractory GERD. However, the infor- patients was reviewed. *p < 0.05 compared to baseline values 15mm and 8mm respectively. On radiology, median tumor mation regarding the effects of LNF on NAR is limited.

size in the LLR group was larger at 5.5 cm vs. 4.1 cm (p = Poster Abstracts Isolated Upright Refl uxers Combined/Supine Refl uxersAIM: To prospectively evaluate acid and NAR episodes 0.02). There were no differences in degree of fi brosis or ste- (n = 36) (n = 27) atosis in the background liver. Post-operatively, the average supine upright p-value supine upright p-valuebefore and at 1, 3 and 6 months after LNF in patients with Sunday LOS in the LLR and OLR groups were 6.6 (range 4–12 days) LESP 19.3 (10.6) 13.6 (12.7) 0.003 14.7 (6.9) 11.6 (8.8) 0.03typical symptoms of GERD. CONCLUSION: MII-pH provides an objective means for and 11.1 (range 8–57 days) days respectively. There was no LES 2.4 (0.6) 2.2 (0.6) 0.018 2.0 (0.4) 2.0 (0.6) 0.94MATERIAL AND METHODS: We prospectively assessed evaluating the effectiveness of LNF in controlling acid length difference in frequency of ICU admission, post-operative and reviewed collected data from 14 consecutive patients (5 and nonacid refl ux episodes. Selection of patients before The valuables were reported as the mean and SD. transfusion, or number of complications. When converted male and 9 female; mean age 46 ± 12 years) who underwent LNF surgery is an important task and guarantees successful cases were excluded OR time was lower in the LLR group at anti-refl ux surgery (LNF) between August 2009 and August symptom control at 6 months. 195 min vs. 266 min (p = 0.01). RESULTS: Between 12/2009 and 11/2010, 128 symptom- 2010. All subjects had heartburn and/or regurgitation and evidence of hiatal hernia and a positive symptom response CONCLUSION: LLR appears to be a feasible approach for atic patients (male 42, female 86, mean age 54.4 years) Clinical: Hepatic underwent bi-positional HRM. Of 128, 58 underwent to proton pump inhibitor (PPI) therapy. Nine patients had HCC in a cirrhotic liver. Our series has shown shorter OR 48-hour pH testing and 70 underwent 24-hour MII-pH. 36 evidence of erosive GERD during endoscopy. Before surgery time with similar negative margin rates and post-oper- ative morbidity in a cohort of tumors larger than gener- patients (male 10, female 26, mean age 54.2 years, mean all subjects underwent a symptomatic evaluation using a 5 Su1626 BMI 29.1) were identifi ed as upright refl uxers and 27 (male point type Likert scale for esophageal symptoms, stationary ally reported for a laparoscopic series. Although limited 10, female 17, mean age 53.9 years, mean BMI 26.2) exhib- esophageal manometry using a solid state catheter and a Comparison of Laparoscopic and Open Liver Resectionsin power, our data suggest a clinically signifi cant trend ited combined/supine refl ux. Nineteen of 36 (52.8%) of 24-h MII-pH using a six impedance 1-pH catheter (Sand- in Cirrhotic Patients Using a Matched Pair Analysis towards shorter LOS and OR time with LLR. As radiofre- quency ablation is emerging as a curative therapy in the patients with upright refl ux and 17 of 27 (63%) with com- hill Scientifi c, Littleton, CO, USA). The design of the cath- Faizal D. Bhojani*1, Adrian M. Fox1, Kristen B. Pitzul2, eter allowed recording impedance data at 3, 5, 7, 9, 15 and smaller HCC, laparoscopic surgery may play an increasing bined/supine refl ux had predominantly atypical symptoms Alice C. Wei 1, Carol-Anne Moulton1, Allan Okrainec2, 17 cm above the LES. Patients were asked to discontinue role in the management of larger resectable lesions. such as cough or hoarseness. Although lower esophageal Sean Cleary 1 sphincter pressure (LESP) was signifi cantly lower in upright PPI and histamine blockers one week before the baseline 1 than in supine in both groups, upright refl uxers had signifi - evaluation. MII-pH and esophageal manometry was per- Toronto General Hospital, Division of General Surgery., University cantly shorter LES length in upright than in supine. In 7 of formed 1, 3 and 6 months after LNF. Refl ux episodes (acid Health Network. University of Toronto, Toronto, ON, Canada; 36 of upright (19.4%) and 4/27 of combined/supine refl uxer and NAR), % of time of pH <4, JDM score, proximal extent 2Toronto Western Hospital, Division of General Surgery., University (14.8%), a hiatal hernia was not present in supine, but episodes and symptoms index were calculated according to Health Network. University of Toronto, Toronto, ON, Canada appeared in upright position manometrically. DeMeester standard defi nitions. INTRODUCTION: Surgical resection for hepatocellular scores of upright refl uxers were signifi cantly lower than RESULTS: At baseline 7/14 (50%) of subjects had a hypo- carcinoma (HCC) in patients with cirrhosis is associated those of combined/supine refl uxers (5.03 ± 5.3 vs 26.2 ± tensive LES and 12/14 (85%) had normal esophageal peri- with increased intra-operative and post-operative morbid- 5.2, p < 0.0001). However, 22 of 36 (61.1%) upright refl ux- stalsis, 2 subjects had criteria for ineffective peristalsis. ers were found to have either esophagitis, hiatal hernia or The mean LES pressure was 10.9 ± 3.3 mmHg. All patients

78 79 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1627 and postoperative complications (minor and major vs no RESULTS: Twenty-three patients were identifi ed. Ten METHODS: One hundred and fi fty-three patients who complication) there was no signifi cant difference in sur- patients underwent PVE without segment IV while 13 underwent pancreaticoduodenectomy for invasive ductal Surgical Outcomes of Hepatic Resection in Elderly vival between both groups (HR 0.64 (0.37–1.08), p = 0.09). patients underwent PVE+IV embolization. Post emboliza- carcinoma of the pancreas between 2004 and 2008 were Patients with Colorectal Liver Metastases CONCLUSION: Liver resection in elderly patients is safe tion liver volumes and hypertrophy rates did not differ retrospectively examined. The clinical and histopathologi- *1 1 1 Javairiah Fatima, Crystal M. Kavanagh, David M. Nagorney, with acceptable morbidity rates comparable to those of between the groups (Table 1). The majority of patients in cal factors, including intrapancreatic nerve invasion, were Florencia G. Que 1, William S. Harmsen2, John H. Donohue1, younger patients. Therefore chronologic age alone should the PVE+IV group underwent extended right hepatectomy analyzed in these patients. Intrapancreatic nerve invasion Michael L. Kendrick 1, Michael B. Farnell1, Kaye M. Reid Lombardo1 not be a contraindication for aggressive treatment, and while right hepatectomy was more common in the PVE (ne) were classifi ed into 4 groups according to the follow- 1 Division of Gastroenterologic and General Surgery, Mayo Clinic, should be considered in carefully selected patients. group. Twenty-six percent of patients were resected for ing semi-quantitative scores: no invasion (ne0) seen; slight an indication other than colorectal cancer. Seventy eight invasion (ne1) observed (1–3 points of cancer cell inva- Rochester, MN;2Division of Biomedical Statistics and Informatics, percent of patients received preoperative chemotherapy. sions); moderate invasion (ne2) observed (4–8 points); Mayo Clinic, Rochester, MN Su1628 Three patients did not undergo resection: two for progres- marked invasion (ne3) observed (>8 points) in the most BACKGROUND: Liver resections (LR) for colorectal metas- Safety and Effi cacy of Preoperative Portal Vein sion of disease and one for cirrhosis discovered at opera- extensively involved area under low power magnifi cation tases have been established to prolong survival with accept- Embolization in Patients at Risk for Postoperative tion. There was no difference in length of stay, operative (X100). The relationships between the degree of intrapan- able morbidity; however, with limited health resources Liver Failure blood loss or blood products transfused between groups. creatic nerve invasion and disease-free survival, as well as candidacy of the elderly for such operations has been ques- Most surgical complications were minor and rates were various histopathologic factors, were investigated. Kristen Massimino*1, Kenneth J. Kolbeck2, C. Kristian Enestvedt1, tioned in terms of safety and effi cacy. similar between the PVE and PVE+IV groups at 50% and RESULTS: The median disease-free survival of the 153 AIM: Susan L. Orloff 1, Kevin G. Billingsley1 54%, respectively. There were no episodes of post-operative To determine 30-day mortality, overall survival (OS),1 patients was 9 months, with a disease-free 5-year survival of and morbidity after liver resection for colorectal liver Surgery, Oregon Health & Science University, Portland, OR; liver failure or death. Complication rates from PVE were 17.1%. The incidence of histological intrapancreatic nerve 2 metastasis (CRLM) in patients ≥75y old in comparison to Interventional Radiology, Oregon Health & Science University, also similar at 36% among those undergoing PVE vs. 31% invasion was as follows: ne0 = 7 (5%); ne1 = 38 (25%); ne2 patients 60–74 y of age. Portland, OR for PVE+IV. One patient in the PVE group underwent modi- = 85 (56%); and ne3 = 19 (12%). There were signifi cant dif- fi ed surgical resection due to a complication of portal vein METHODS: BACKGROUND: ferences between groups; ne0 vs ne1 (P = 0.3163), ne1 vs Patients ≥75y (n = 68) (Group 1) were matched Portal vein embolization (PVE) is utilized in embolization. preparation for major hepatectomy to induce hypertrophy ne2 (P = 0.0007), ne0 vs ne2 (P = 0.0234), ne1 vs ne3 (P = on ASA status with patients 60–74 y of age (n = 67) (Group CONCLUSIONS: 2) who underwent LR for CRLM between January 1996 and and prevent postoperative liver insuffi ciency. Since 2006 Portal vein embolization resulted in an 0.0000), and ne0 vs ne3 (P = 0.0009) In multivariate analy- December 2005. Retrospective review of medical records our multidisciplinary team has been using PVE in patients overall 38% increase in sFLR volume. Complications related sis, a lack of lymph node metastases (P = 0.001), incidence was performed to compare outcomes between both groups. with <30% predicted future liver remnant or underly- to portal vein embolization occurred but they did not pre- of intrapancreatic nerve invasion (P = 0.001), and nega- vent eventual resection. Following embolization, resection tive surgical margin (P = 0.011), signifi cantly increased the RESULTS: ing liver disease prior to major hepatectomy. This strat- Females constituted 49% and 28% of Group 1 egy includes embolization of segment IV portal venous was associated with a low incidence of complications. disease-free survival. The tumor stage was not associated and Group 2, respectively (p = 0.02). Eighty-one percent branches when an extended right hepatectomy is in the with intrapancreatic nerve invasion. (P = 0.255) However, and 19% of the Group 1 and 67% and 33% of Group 2 a larger tumor size (P = 0.024), a higher incidence of lym- preoperative plan. The aim of this study is to report the Clinical: Pancreas had colon cancer and rectal cancer, respectively. Sixty-four short term outcomes following this approach. phatic invasion (P = 0.036), and the presence of extrapan- percent of Group 1 and 71% of Group 2 had node positive creatic nerve plexus invasion (P < 0.001), were identifi ed as METHODS: Records of patients who underwent PVE

disease of the primary tumor. A major LR (lobectomy or Poster Abstracts Su1629 independent factors associated with a higher incidence of extended lobectomy) was performed in 30.9% of Group 1 during 2006–2010 were retrospectively reviewed. Patient intrapancreatic nerve invasion. and 41.8% of Group 2 patients, p = 0.21. One patient in demographics, indications for and extent of resection, Intrapancreatic Nerve Invasion as a Predictor for Sunday CONCLUSION: Intrapancreatic nerve invasion may be the older group underwent a portal vein embolization vs. operative blood loss, post operative blood product trans- Recurrence After Pancreaticoduodenectomy in Patients fusion, length of hospitalization and complications were useful as a predictor for recurrence after pancreaticoduo- two patients in the younger cohort. There was no differ- with Invasive Ductal Carcinoma of the Pancreas ence in the overall morbidity rate between Group 1 [14.5%] analyzed. CT based volumetrics were performed to deter- denectomy in patients with invasive ductal carcinoma of and Group 2 [19.7%], p = 0.50). There were no signifi cant mine future liver remnant (FLR) and standardized future Kazuaki Shimada*1, Satoshi Nara1, Minoru Esaki1, the pancreas. differences in age groups and Clavien Grade 1-V postopera- liver remnant (sFLR) volume pre and post-PVE as well as Yoshihiro Sakamoto 1, Tomoo Kosuge1, Nobuyoshi Hiraoka2 tive complications (p = 0.08). There was only one death at degree of hypertrophy. Patients were stratifi ed by segment 1 Hepatobiliary and Pancreatic Surgery Division, National Cancer 30-days and it occurred in Group 1. Median hospital stay IV embolization and compared. Signifi cance was reported Center Hospital, Tokyo, Japan;2Pathology Division, National Cancer for both groups was 7 days (p = 0.35). Fifteen percent of for p < 0.05. Center Research Institute, Tokyo, Japan patients in Group 1 required ICU stay compared to 10% BACKGROUND: in the younger cohort (p = 0.46). Adjuvant chemotherapy Table 1: Results of volume analysis Neural invasion is a distinct route for was administered more frequently in Group 2 (51%) com- the spread of pancreatic carcinoma. However, the clinico- pared to Group 1 (25%; p = 0.002). There was no signifi cant PVE (n = 10) PVE+IV (n = 13) p-value pathologic signifi cance of neural invasion, with particular FLR volume pre-PVE (mL) 513.5 440.9 0.305 association between the age groups for disease recurrence reference to intrapancreatic nerve invasion, remains to be FLR volume post-PVE (mL) 694.9 579.6 0.123 elucidated. (p = 0.44) with a 5-year recurrence-free survival of 53.9%Change in volume (mL) 144.3 140.9 0.869 in Group 1 and 50.2% in Group 2. There was no differ-sFLR pre-PVE (%) 29.8 24.2 0.283 ence in 1-, 3-, 5-year, OS between the two groups (p = 0.21)sFLR post-PVE (%) 42.8 33.6 0.113 with a 5- year survival of 22.3% and 37.2% respectively.Hypertrophy rate (%) 38.9 38.3 0.283 On multivariable analysis, evaluating age 60–74 vs ≥75, sex

80 81 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

RESULTS: The incidence of peripancreatic fat invasion were not signifi cantly different between surgical and non Su1630 Su1633 was as follows: grade 0, n = 8 (6%); grade 1, n = 54 (41%); surgical patients (11,753 ± 16,954 vs. 41,537 ± 71,294 U/L, Assessment of Peripancreatic Fat Infi ltration of grade 2, n = 49 (37%); grade 3, n = 21 (16%). Five-year sur- 1,524 ± 2,925 vs. 1,335 ± 5,141 ng/mL, 10,234 ± 10,450 vs. Surgical Site Infection After Pancreaticoduodenectomy: Pancreatic Ductal Adenocarcinoma by Multidetector-vival rate was 55.6% for grade 0, 38.7% for grade 1, 16.4% 8,522 ± 21,279 U/mL, respectively; p = NS). Four patients Single Center Experience of 356 Consecutive Patients Row Computed Tomography: Correlation to for grade 2, and 0% for grade 3, respectively. There were with negative FNA cytology had surgical resection for high Teiichi Sugiura*, Katsuhiko Uesaka, Hideyuki Kanemoto, Extrapancreatic Nerve Invasion and Surgical Outcomessignifi cant differences in the survival rates between grade clinical suspicion. Final pathology of the resected patients Takashi Mizuno 1 and 2 (p = 0.003), and grade 2 and 3 (p = 0.032), respec- (n = 12) revealed IPMN (7), Mucinous neoplasm (2), pan- Division of Hepato-Biliary-Pancreatic Surgery, Shizuoka Cancer *1 1 2 Yusuke Yamamoto, Kazuaki Shimada, Takeuchi Yoshito, tively. Multivariate survival analysis showed that lymph creatic cancer (1), NET(1) and pancreatitis (1). Center, Shizuoka, Japan Hata Syojiro 1 1 1 1 , Satoshi Nara, Minoru Esaki, Yoshihiro Sakamoto, node metastasis, surgical margin positive, and grade 2 and CONCLUSION: Surgical resection of pancreatic cyst is 1 3 OBJECTIVE: Although the surgical site infection after Tomoo Kosuge , Nobuyoshi Hiraoka 3 pancreatic far infi ltration were independently associated based on the cyst size, location and patient comorbidities 1 Hepatobiliary and Pancreatic Surgery, National Cancer Center with a poorer prognosis. Only extrapancreatic nerve plexus pancreaticoduodenectomy (SSI) is a critical and frequent rather than results of the EUS examination. Routine EUS complication, its risk factors have not been fully evaluated. Hospital, Tokyo, Japan;2Diagnostic Radiology Division, National invasion was signifi cantly associated with grade 2 and 3 for cystic pancreatic lesions did not dictate the surgical Determining the risk factors for SSI may provide informa- Cancer Center Central Hospital, Tokyo, Japan;3Pathology Division, pancreatic fat infi ltration. management in this series. EUS can justify a non surgical CONCLUSION: tion on reducing complications and improving the outcome National Cancer Center Research Institute, Tokyo, Japan The grade of peripancreatic fat infi ltration approach for small cysts and when surgical resection carries of the patients who undergo pancreaticoduodenectomy. determined by MDCT, which was signifi cantly related with a high risk. BACKGROUND: Retroperitoneal invasion including extra- PATIENTS AND METHODS: the incidence of histological extrapancreatic nerve plexus A retrospective review using pancreatic nerve plexus invasion of pancreatic ductal invasion, may be useful as a predictor for survival after pan- prospectively collected data base of 356 patients who adenocarcinoma, which has been demonstrated as peri- Su1632 creaticoduodenectomy in patients with invasive ductal car- underwent pancreaticoduodenectomy between January pancreatic fat infi ltration by multidetector-row computed cinoma of the pancreas. Incidence and Implications of Impaired Glycemic 2003 and April 2010 was conducted. The demographic and tomography (MDCT), is a distinct characteristic associated Control Following Distal Pancreatectomy clinical data of each patient were reviewed. The outcome with unresectability and impaired survival after surgery. of interest was SSI which was classifi ed into incisional or The aim of this study was to investigate whether the degree Su1631 R.M. Walsh* General Surgery, Cleveland Clinic, Cleveland, OH organ/space SSI according to the CDC’s NNIS defi nition. of peripancreatic fat infi ltration determined by MDCT is a Does Routine Endoscopic Ultrasound Alter Surgical Multivariate analysis was performed to assess the risk fac- useful predictor of surgical outcomes in patients with pan- Management of Patients with Pancreatic Cystic Lesions? PURPOSE: Impaired glycemic control is an important tors for SSI. Bacterial appraisal was also conducted. creatic ductal carcinoma. long-term consequence of distal pancreatectomy partially A Retrospective Analysis of 93 Consecutive Patients RESULTS: The patients comprised of 231 males (65%) and due to preferential E cell location in the body and tail, and *1 3 2 125 females (35%) with a median age of 68 years, consisting Cherif Boutros, Emilia Genova, Ponnandai Somasundar, nature of underlying pancreatic disease. Understanding the 2 of 179 patients (50%) with pancreatic cancer, 35 (10%) with N. Joseph Espat frequency of this outcome may alter the operative approach, 1 pancreatic tumor including intraductal papillary mucinous Division of Surgical Oncology, University of Maryland School of including consideration of auto-islet transplantation. neoplasm or neuroendocrine tumor, 75 (21%) with bile Medicine, Baltimore, MD;2Surgical Oncology, Roger williams Medical METHODS: A pathology database was interrogated to duct cancer, 38 (11%) with ampulla of Vater carcinoma, Center, Providence, RI; 3 Department of Surgery, Saint Raphael identify all patients who had undergone a distal pancre- and 29 (8%) with others. Of 356 patients, 49 (14%) patients

Hospital, New Haven, CT atectomy from 1990 through 2009. Demographic data, had incisional SSI and 169 (47%) had organ/space SSI. Poster Abstracts INTRODUCTION: medication needed for glycemic control, histopathology, Endoscopic ultrasound (EUS) has been Multivariate analysis identifi ed main pancreatic duct (MPD) and volume resected were collected. Sunday increasingly used to evaluate pancreatic cystic lesions. We size < 3 mm (OR: 3.56), concomitant colectomy (OR: 3.27), investigated the impact of EUS results on the surgical man- RESULTS: 137 patients underwent distal pancreatectomy operating time > 480 min (OR: 2.32), and blood loss > agement for patients with pancreatic cystic lesions. which included 26 (19%) that were diabetic preoperatively. 1,100 ml (OR: 2.17) as signifi cant risk factors for incisional METHODS: Using IRB approved prospectively maintained New-onset diabetes was identifi ed in 27 of remaining 111 SSI. Whereas, the presence of pancreatic fi stula (OR: 6.50), data base, all patients with pancreatic cystic lesions pre- patients (24.3%) with a mean time of onset of 6.4 months, semi-closed drainage system (Penrose drain with retrieval sented to a tertiary care center over 20 months were and a median follow up of 25 months. Diabetes was seen pouch) (OR: 2.81), MPD < 3 mm (OR: 2.71), operating time included. All patients had EUS as part their evaluation pro- more commonly in patients with chronic pancreatitis > 480 minutes (OR: 2.16), and body mass index (BMI) > cess. Patients’ demographics, EUS fi ndings and fi ne needle (5/13; 38.5%) and adenocarcinoma (7/22, 31.8%), but was 22.0 kg/m2 (OR: 1.87) were found to be signifi cantly associ- aspiration (FNA) laboratory and cytological results when not seen in patients undergoing resection for serous cystad- ated with organ/space SSI. enoma (0/8), or intraductal papillary mucinous neoplasm applied were analyzed to assess the impact of EUS data on Bacterial isolation revealed that gut-derived microorgan- [IPMN] (0/9). Patients developing diabetes were older (63.4 PATIENTS AND METHODS: The clinical records of 132 the consequent management. isms such as Enterococcus species, Enterobacter species, vs. 55.1, p = 0.01) but there was no difference in gender or patients that underwent macroscopic curative pancreati- RESULTS: and Klebsiella species were the leading bacteria at the focus 93 consecutive patients (median age 68 years, ethnic distributions. The development of diabetes was not coduodenectomy for pancreatic adenocarcinoma between M:F 1:1.5) with pancreatic cystic lesions undergoing EUS of SSI. January 2004 and December 2008 were reviewed retro- linked to the volume of parenchyma resected: lower vol- were included. Pancreatic cysts mean size was 2.3 ± 1.7 cm CONCLUSIONS: Pancreatic fi stula was the most pow- spectively. The MDCT scan acquired in all patients were umes of tissue were resected in those with diabetes (119.2 and they were located at the pancreatic head, body and tail 3 3 erful risk factor for organ/space SSI. Except for patients’ reviewed in consensus by two abdominal radiologists. Peri- cm vs 192.7 cm ; p = 0.314); or related to patient body in 29, 36 and 28 patients respectively. 42 patients (45%) own factors such as smaller MPD or greater BMI, efforts pancreatic fat infi ltration was determined and classifi ed by mass index [BMI] (29.3 KgM-2 vs 28.1 KgM-2; p = 0.51). had fi ne needle aspiration (FNA); the size of the cystic to reduce pancreatic fi stula, application of closed drainage MDCT into 4 groups to the following grade: grade 0, no Of the 26 patients with pre-resection diabetes, 9 were tak- lesion in these patients were signifi cantly larger (2.9 ± 1.6 system, shorter and less bleeding surgery would contribute peripancreatic fat infi ltration; grade 1, a strand or streak ing insulin. In the remaining 17, 8 (47%) deteriorated and cm vs. 1.8 ± 1.7 cm; p = 0.008). There was no correlation to decrease the incidence of SSI after pancreaticoduodenec- structure radiating from the tumor (<5 mm); grade 2, a required insulin postoperatively. The development of post- between the rate of FNA and the cyst location. After full tomy. When SSIs once occur and antibacterial treatment strand or streak structure radiating from the tumor with- operative impaired glycemic control was associated with evaluation 12 patients (12.9%) underwent surgical resec- prolonged hospital stay (12.6 days vs. 7.7 days; p = 0.001). is necessary, it should be conducted targeting intestinal out abutting the major artery (>5 mm); grade 3, a streak tion. Resected cystic lesions were larger (3.4 ± 1.1 vs. 2.1 ± CONCLUSION: bacteria. structure radiating from a tumor and abutting to a major 1.7 cm, p = 0.02) and cysts were more likely to be resected Impaired glycemic control is common fol- artery. The clinicopathologic factors including the grade of when located at the pancreatic tail (2% of pancreatic head, lowing distal pancreatectomy, especially in those resected peripancreatic fat infi ltration as potentially important pre- 10% of pancreatic body and 28% of pancreatic tail cystic for chronic pancreatitis or cancer. In patients requiring dictors for survival were analyzed. The relationship of the lesions were resected; p = 0.005). Surgical resection was not distal pancreatectomy for chronic pancreatitis, improved grade of peripancreatic fat infi ltration and various histo- signifi cantly associated with preoperative FNA (8/42 vs. outcome may include islet replacement. pathologic factors were also investigated. 4/51; p = NS). Cyst aspiration amylase, CEA, CA19-9 levels

82 83 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1634 creatic duct upstream to the stricture. Patient records were Oxaliplatin was escalated from 30 mg/m2 in 10 mg inter- of distal pancreatic resection evolved from cut-and-sew, to reviewed to determine prior history of acute pancreatitis, vals up to 60 mg/m2. Concurrent radiation therapy con- stapled technique with green and recently white (vascular) Prior Acute Pancreatitis Is the Most Common Cause ofand patients were aggregated into one of two groups: with sisted of 4,500 cGy in 25 fractions (180 cGy/fx/d) followed cartridge. The aim of our study was to evaluate the rate of Obstructive Pancreatitis in Patients Undergoing Distaland without a history of prior acute pancreatitis. Demo- by a conedown to the tumor and margin for an additional clinically signifi cant fi stulas (International Study Group on Pancreatectomy graphics, GI and non-GI related co-morbidities, social his- 540 cGy x3 (total dose 5040 cGy in 28 fractions). Following Pancreatic Fistula (ISGPF) Grade B or C) following distal Daniel Chan*1, Michael L. Kendrick2, Michael B. Farnell2, tory, symptoms, etiology, and surgical presentations were completion of CRT, patients deemed resectable underwent pancreatectomy, and to identify variables associated with a Santhi Swaroop Vege 3 statistically analyzed between the two groups using Chi- surgery; those who remained unresectable for cure but did low rate of fi stula development. 1 square and Fisher’s exact methods. not progress (PD) received mFOLFOX6 x6 cycles. Survival Mayo Medical School, Mayo Clinic, Rochester, 2MN;Gastroenterologic METHODS: Clinical records of all patients who underwent was calculated using Kaplan-Meier analysis. End points of and General Surgery, Mayo Clinic, Rochester, MN; 3 RESULTS: Forty nine of 946 patients undergoing distal distal pancreatic resections between February 1999 and Gastroenterology the phase II portion were resectability and survival. & Hepatology, Mayo Clinic, Rochester, MN pancreatectomy from 1990–2005 had OP. All 49 patients July 2010 by a single surgeon were retrospectively reviewed RESULTS: BACKGROUND: presented at the time of surgery with evidence of chronic Fifteen patients were initially enrolled in the focusing on the incidence and type of pancreatic fi stula as The natural history of obstructive pan- pancreatitis by operative or surgical pathology report. Phase I component of the study and all completed neoad- defi ned by ISGPF. Study variables included age, gender, sur- creatitis (OP) is currently poorly characterized. While it has 35/49 (71.4%) patients had a history of acute pancreatitis, juvant therapy. The highest dose (60 mg/m2) of oxaliplatin gical approach, extent of resection, ASA classifi cation, type been reported with a few retrospective cases of necrotizing and 14/49 (28.6%) patients had no history of acute pan- was well tolerated and this was carried forward in the phase II of stapler cartridge, use of Seamguard™, and ISGPF clas- acute pancreatitis, there is uncertainty as to whether it is creatitis. There were no differences in the demographics, portion of the study. Grade 4 toxicities were observed during sifi cation. Statistical analysis was performed using Fisher’s the initial presentation of chronic pancreatitis, or a sequel comorbidities, social history, or etiology between the two Phase I (n = 2, pulmonary embolism and lymphopenia) and exact test, univariate and multivariate logistic regression. from previous acute pancreatitis. Our study attempts to groups. Our data demonstrated that in individuals with a phase II (n = 3, fatigue, leucopenia and thrombocytopenia). RESULTS: elucidate the pathogenesis of OP. Sixty four patients (median age 60, range 21–85; history of acute pancreatitis, necrotizing pancreatitis was Additional 9 patients were treated in the phase II portion. 54% male) underwent distal pancreatic resection (laparos- a common fi nding. This group exhibited less weight loss Overall, 24 subjects (14 men and 10 women, mean age 65 copy 50% vs. open 50%). The most common indications Table 1: Main Characteristics in Patients with Chronic Obstructiveas a presenting symptom, and they were also less likely to years) were enrolled and received CRT; 12 of the 24 did Pancreatitis with and Without a History of Acute Pancreatitis were pancreatic adenocarcinoma (N = 15; 23%) and neuro- have had a prior cholecystectomy (See Table 1). not complete the treatment. Reasons for not completing endocrine neoplasms (N = 14; 22%). treatment included progression (7), withdrawal of consent History of Acute No History of AcuteCONCLUSIONS: Prior acute pancreatitis was the main (2), grade 4 toxicity (3). Following CRT, 8 (33%) patients Clinically signifi cant pancreatic fi stula developed in 23% Pancreatitis Pancreatitis cause for OP in patients undergoing distal pancreatectomy, (N = 15). The rate of fi stula with cut-and-sew technique DEMOGRAPHICS were deemed possibly resectable and were explored. Two and de novo chronic pancreatitis was a less common cause. was 36% (4/11), with stapled green cartridge 31% (9/29) Number of cases (N) 35 14 The explanation for this is most likely due to the necrosis additional patients were found to have PD (carcinomato- Gender M:F 24 (68.6%):11 (31.4%) 7 (50%):7 (50%) sis). Four had stable disease (SD) but remained unresectable and only 5% (1/21) with stapled vascular cartridge. Uni- of the pancreas during acute pancreatitis resulting in subse- variate logistic regression identifi ed vascular cartridge size Mean Age at Surgery (years) 46.8 ± 10.8 45.6 ± 15.7 quent stricture formation. A prospective longitudinal study and 2 (8% of all study subjects) were resected for cure with BMI (Mean) 27.1 25.9 (p = 0.04, OR 0.11) and open stapled technique (p = 0.05, of necrotizing pancreatitis would confi rm this hypothesis. negative margins. Follow up was available for 23 patients. CO-MORBIDITIES Median overall survival was 14 months (9 and 15 months OR 0.12) as variables signifi cantly associated with a low fi s- Diabetes 2 (5.7%) 3 (21.4%) respectively for PD and SD). Of the 2 resected patients, one tula rate. Both vascular cartridge size (p = 0.05, OR 0.10) and Gallstones 5 (14.3%) 3 (21.4%) Su1635 died of disease at 21 months and one is alive without dis- open stapled technique (p = 0.04, OR 0.08) remained sig- Steatorrhea 0 (0.0%) 1 (7.1%) Can We Downstage Regionally Advanced Pancreatic ease at 11 months from trial entry. nifi cant when analyzed by multivariate logistic regression. Poster Abstracts SOCIAL HISTORY Division of pancreatic parenchyma with vascular cartridges Alcohol Use 21 (60.0%) 12 (85.7%) Cancer to Resectable: A Phase I/II Study of Induction CONCLUSIONS: Combined modality treatment for region- resulted in signifi cantly (p = 0.03, OR 9.0) lower fi stula rate Tobacco Use 22 (62.9%) 11 (78.6%) Oxaliplatin and 5FU Chemo-Radiation ally advanced pancreatic cancer with oxaliplatin, 5FU and Sunday SYMPTOMS compared to green cartridges. The use of Seamguard™ did Marcovalerio Melis*1, Theresa Ryan2, Howard S. Hochster3, radiation was reasonably well tolerated. The majority of not affect fi stula rate (16% vs. 27%; p = 0.34), nor did the Abdominal Pain 34 (97.1%) 12 (85.7%) patients remained unresectable. Survival data with this reg- Nausea/Vomiting 3 (8.6%) 3 (21.4%)Deirdre Cohen 2, Antonio Masi2, Anurag Chandra2, performance of multi-visceral resection vs. distal pancre- imen are comparable to others for locally advanced pan- Weight Loss 3* (8.6%) 7* (50.0%)H. Leon Pachter 1 1 atectomy/splenectomy alone (21% vs. 23%, p = 1.0). , Elliot Newman creas cancer. SURGICAL PRESENTATION 1 Surgery, New York University, New York, 2NY;Medicine, New York CONCLUSION: Prior cholecystectomy 10* (28.6%) 9* (64.3%) The optimal technique of pancreatic divi- University, New York, NY; 3 Prior necrotizing pancreatitis 13* (37.1%) 0* (0.0%) Medicine, Yale School of Medicine, New Su1636 sion has not been conclusively established. Dividing the documented Haven, CT pancreas utilizing vascular (2.5 mm) staple cartridge signifi - Pseudocyst 12 (34.3%) 7 (50.0%) Infl uence of Staple Size on Fistula Formation Followingcantly decreased the rate of clinically signifi cant pancreatic BACKGROUND: The majority of patients with pancreatic Time between diagnosis of 892 days (2.4 years) 1587 days (4.35 Distal Pancreatectomy fi stula, and we continue to favor this technique in our prac- pancreatitis and obstructive years) adenocarcinoma (PC) present with regionally advanced *1 1 1 2 tice. A prospective randomized trial is necessary to validate pancreatitis disease. This includes borderline resectable and locally Boris Sepesi, Jacob Moalem, Eva Galka, Peter Salzman, these results. *Denotes statistical signifi cance with p < 0.05 advanced unresectable tumors as defi ned by current NCCN Luke O. Schoeniger 1 guidelines for resectability. Chemo-radiation (CRT) is used 1 Department of Surgery, University of Rochester Medical Center, in this setting in attempt to control regional disease, and Rochester, NY;2Biostatistics and Computational Biology, University possibly downstage to resectable disease. We report a phase of Rochester, Rochester, NY I/II trial of a novel combination of 5FU/Oxaliplatin with BACKGROUND: concurrent radiation in patients presenting with regionally Pancreatic fi stula continues to be a source METHODS: We retrospectively reviewed all patients of advanced disease. of signifi cant morbidity following distal pancreatic resec- OP undergoing distal pancreatectomy from 1990-2005 at a METHODS: tions. The technique of pancreatic division varies widely single institution. We defi ned OP as pancreatitis satisfying Patients with biopsy-proven borderline resect- among surgeons, and there is no evidence that identifi es three criteria on imaging (CT, ERCP, MRI, MRCP, or EUS): 1. able or locally advanced unresectable pancreatic adenocar- a single method as superior. In our practice, the technique Stricture of the pancreatic duct, 2. Sparing of the pancreas cinoma were eligible. Chemotherapy included continuous downstream to the stricture, and 3. Dilatation of the pan- infusion 5FU (200 mg/m2) and oxaliplatin weekly for 5 weeks.

84 85 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1637 Su1638 Su1639 Su1640 Distal Pancreatectomy: Clinically Relevant Leakage Novel Prediction of Pancreatic Anastomotic Failure Changes in Pancreatic Endocrine and Exocrine Clinical Practice as an Engine Driving a Translational from the Pancreatic Stump Closure May Be Due to After Pancreatoduodenectomy Using Preoperative CT Function After First Episode of Acute Alcoholic Research Program in Pancreatic Cancer Drain Failure or Backpressure Imaging with the Evaluation of Remnant Pancreatic Pancreatitis Eugene P. Kennedy*1, Jonathan R. Brody1, Agnes Witkiewicz2, Yasushi Hashimoto*1,2, L. William Traverso1 Volume and Body Composition Juhani Sand*, Riitta Lappalainen-Lehto, Satu JäRvinen, Harish Lavu 1, Patricia K. Sauter1, Dane Grenda1, Dianne Macrae1, 1 Florence Williams 1 1 1 Center for Pancreatic Disease, St. Luke’s Hospital System, Boise, ID; *1 2 1 Hanna SeppäNen, Sari Raty, Johanna Laukkarinen, Isto Nordback , Ernest L. Rosato, Charles J. Yeo Yujiro Kirihara, Naoki Takahashi, Yasushi Hashimoto, Guido 1 2Department of Surgery, Graduate School of Biomedical Science, Department of Gastroenterology and Alimentary Tract Surgery, TampereSurgery, Thomas Jefferson University, Philadelphia,2Pathology, PA; M. Sclabas 1, Saboor Khan1, Junichi Sakagami4, Hiroshima University, Hiroshima University, Hiroshima, Japan Unversity Hospital. Finland, Tampere, Finland Thomas Jefferson University, Philadelphia, PA Marianne Huebner 3, Michael G. Sarr1, Michael B. Farnell1 1 OBJECTIVE: BACKGROUND: The most common complication after Surgery, Mayo Clinic, Rochester, MN;2Radiology, Mayo Clinic, BACKGROUND: Acute pancreatitis may cause diabetes Traditionally, clinical and research pursuits distal pancreatectomy (DP) is leakage from the pancreaticRochester, MN; 3Health Sciences Research, Mayo Clinic, Rochester, and impaired exocrine function, but especially after single have occurred in parallel. In the current climate of reduced research funding and increased need for translation of dis- stump closure. Clues to the leakage may help in prevention. MN; 4Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN episode of pancreatitis the possible changes in pancreatic coveries, an integrated program is an appealing strategy. OBJECTIVE: We sought to determine the incidence, sever- function are poorly documented. A cohort of patients with INTRODUCTION: An increased body mass index (BMI) This study was designed to demonstrate research benefi ts ity and risk factors for leakage using a large number of DP acute alcohol pancreatitis (Gastroenterology 2009;136:848- and pancreatic duct size are known predictors of pancreatic 55.) were further followed for their pancreatic function. of a robust clinical practice. cases performed with traditional open surgery. anastomotic failure (PAF) after pancreatoduodenectomy METHODS: METHODS: Review of our database revealed a total of PATIENTS AND METHODS: Our prospectively main- (PD). However, the impact of anthropomorphic measure- Patients were followed up to 7 yrs. Patients 1,005 pancreatic resection for the period January 1, 1995 tained single-surgeon database identifi ed 223 consecutive ments (remnant pancreatic (parenchymal) volume (RPV) were interviewed for possible recurrent acute pancreatitis through December 31, 2009 (annualized). The period Janu- DPs between 1992 and 2008. The operation was the same in and body composition on PAF) are unknown. The aim was and fecal elastase tests and tests for glucose metabolism ary 1, 2006 onwards included 711 pancreatic resections all cases performed with a hand-sewn fi sh-mouth closure of to determine if pancreatic remnant volume, subcutaneous/ were performed. and was correlated with research activities. the pancreatic stump. Using daily drain amylase measure- visceral adipose tissue (SAT/VAT) area, and skeletal muscle RESULTS: 49 out of the initial 119 patients were still ments stump leakage (LEAK) was defi ned and graded as no (SM) area calculated from the preoperative computed continuing in the follow-up program March 2010. 33/119 RESULTS: The growth in pancreatic resections produced LEAK, Grade A, B, or C according to the severity classifi ca- tomography (CT) predict PAF after PD. patients (28%) had developed a recurrent pancreatitis. 7 professional fees averaging $3,500/case to Surgery and $3,000/case to other departments. Hospital contribution tion system of the International Study Group on Pancreatic METHODS: In173 patients undergoing preoperative CT patients (6%) had diabetes already prior to the fi rst episode margin averaged $13,500/case. The increased pancreatic Surgery (ISGPS). “Clinically relevant” LEAK was defi ned as and PD at a single institution between 2004 and 2009, SM of acute pancreatitis. New onset diabetes was observed in resection volume has contributed more than $3 million ISGPS Grade B/C (LEAK) leaving the no LEAK and Grade area and SAT/VAT cross-sectional area at the 3rd lumbar 26 patients and tended to be more common in patients annually to the positive hospital bottom line. A cases in a “Non-clinically relevant” group (No LEAK). A vertebra were quantitated using the preoperative CT. Mus- with recurrent attacks, compared to the patients with a public web-based calculator was used to standardize the cle and adipose tissue were identifi ed semi-automatically single episode of acute pancreatitis (44% vs. 23%; p = 0.06). ISGPS system. The incidence of LEAK and risk factors for using the CT Hounsfi eld threshold method and remnant Also the number of patients with impaired glucose metabo- clinically-relevant LEAK were assessed. pancreatic volume as a volumetric sum of pancreatic paren- lism increased throughout the follow up. In patients with- RESULTS: out recurrent pancreatitis, the mean fecal elastase activity

ISGPS grading of these 223 cases were: No LEAK chymal area to the left of the surgical margin (left border Poster Abstracts 53%, Grade A 32%, Grade B 13.9%, and Grade C 0.9% with of SMA) over multiple cuts. Pancreatic duct size and paren- increased up to 3 yrs and then slowly decreased. The pro- an overall mortality of zero. Therefore the clinical-relevant chymal hardness were assessed by surgeon. The defi nition portion of patients (6%) with abnormally low elastase Sunday LEAK rate was 14.8% (Grade B+C). Of these B/C cases 24% of PAF was the International Study Group of Pancreatic Fis- values (<100 ug/g) was at its lowest 3 yrs after the single were due to surgical drain failure – lack of patency and/or tula (ISGPF) Classifi cation System; Grades B and C PAF were episode of acute pancreatitis. In patients without recurrent misplaced from their original location. The following risk considered clinically-relevant PAF. Patient demographics of pancreatitis, 23% of those who developed diabetes, also factors were signifi cant by univariate analysis: obesity (BMI those with clinically relevant PAF were compared to those developed exocrine pancreatic insuffi ciency and 60% of > 30) in 52% of LEAK cases vs. 21% of No LEAK cases, neu- without PAF. Associations with PAF by univariate logistic those with low elastase activity also had diabetes. The pan- roendocrine tumors (NET) in 30% of B/C cases vs. 10% of regression models were summarized with odds ratios and creatic function during the follow up did not correlate with No LEAK cases, soft pancreatic texture (88% of LEAK vs. 95% confi dence intervals (CI). The predictive ability for the severity of the initial pancreatitis. 66% of No LEAK cases), and blood loss >1,000 ml (21% several models was described using a concordance index CONCLUSIONS: The risk for new onset diabetes increases of LEAK cases vs. 7% of No LEAK cases). In 46 cases with (c-index). during the follow up also after a single episode of acute preoperative endoscopic ablation or stenting of the major RESULTS: PAF occurred in 22 patients (13%); Grades B alcoholic pancreatitis. Pancreatic exocrine function fi rst pancreatic sphincter a signifi cant decrease in the incidence and C were present in 15 (9%) and 7 (4%) patients resp. In improves for the 2–3 yrs after a single episode of acute pan- of ablation/stenting was observed in the LEAK cases (3%) univariate logistic regression analysis, RPV, VAT, SM, BMI, creatitis, and impairs only thereafter. Most of the patients vs. No LEAK cases (24%). Multivariate analysis identifi ed SAT, pancreatic duct size, and pancreatic texture (soft or who develop exocrine insuffi ciency also have diabetes. the following to be signifi cantly associated with LEAK: BMI hard) were all predictors of PAF with P values of <0.001, These results indicate that even a single episode of acute Since 2006, 5 randomized clinical trials and 3 approved pro- > 30 (odds ratio [OR], 3.04), soft pancreatic texture (OR, <0.001, 0.001, 0.001, 0.02, 0.03, and 0.04, resp. A multi- alcoholic pancreatitis leads to endocrine and exocrine tocols have opened. Tissue and plasma on greater than 400 4.89), intraoperative blood loss > 1,000 ml (OR, 7.01), and variate model with the known predictors using BMI and insuffi ciency of the pancreas during the following years. patients has been banked. 37 peer reviewed publications those who did not undergo preoperative major pancreatic duct size had a c-index of 0.75 (BMI; Odds ratio (OR): 1.13, have resulted. The research team now numbers 23 inde- sphincter ablation/stenting (OR, 9.51). 95% CI: 1.04−1.24, P = 0.005, duct size; OR: 0.71, 95% CI: pendent researchers from our institution and 34 indepen- CONCLUSIONS: Clinically-relevant LEAK was observed 0.51−0.95, P = 0.036). A better multivariate model included dent researchers from 16 collaborating institutions. Patient after DP in 14.8% of the 223 cases, one quarter of these B/C preoperative CT factors VAT and SM with a c-index = 0.96 participation in protocols is high, from 95% on our tissue cases might have been just non-relevant leaks (Grade A) if (VAT; OR: 1.24, 95% CI: 1.16−1.37, P < .001, SM; OR: 0.02, banking protocol to 74% on our recent published pancre- the drain had not malfunctioned. The risk for a clinically 95% CI: 0.01−0.08, P < 0.001). atic anastomosis trial. relevant LEAK after DP was associated with factors that CONCLUSIONS: Compared to established risk factors for CONCLUSIONS: Building a robust clinical practice can retard healing and with an intact major pancreatic sphinc- PAF, RPV and VAT were better predictors of PAF after PD. drive a complementary research program. Clinicians in ter. The latter suggests that LEAK may be due to the many Prediction of the risk of PAF after PD may be best estimated leadership roles in the effort can drive productivity, exten- causes of sphincter spasm that might promote leakage from by including these anthropomorphic measures from the sive collaboration, and high patient participation in eligible pancreatic stump due to back pressure. preoperative CT as well as using intraoperative fi ndings. protocols.

86 87 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1641 Su1642 Su1643 Su1644 Prognostic Value of Lymph Node Ratio Increases withThe Utility of Positron Emission Tomography Scans One Hundred Forty Six Resections for Intraductal Long Term Complications in Patients Number of Lymph Nodes Examined: A Concomitantin the Diagnosis and Management of Pancreatic Papillary Mucinous Neoplasm of the Pancreas Undergoing Pancreaticoduodenectomy with Review of SEER and MGH Patients Adenocarcinoma Megan D. Winner*, Minna K. Lee, Joseph Dinorcia, James A. Lee,Pancreaticogastrostomy Nakul Valsangkar*1,2, Devon M. Bush2, James S. Michaelson2, Megan D. Winner*, Minna K. Lee, Joseph Dinorcia, James A. Lee, Beth Schrope, John A. Chabot, John D. Allendorf Eileen Bock*, Michael G. Hurtuk, Margo Shoup, Gerard V. Aranha Carlos Fernandez Del-Castillo 1, Andrew L. Warshaw1, Beth Schrope, John A. Chabot, John D. Allendorf Department of Surgery, Columbia University College of Physicians Surgery:and Division of Surgical Oncology, Loyola University Medical Sarah P. Thayer 1 Department of Surgery, Columbia University College of Physicians and Surgeons, New York, NY Center, Maywood, IL, IL 1 Surgeons, New York, NY Surgery- Warshaw Institute, Massachusetts General Hospital, Boston, BACKGROUND: Intraductal Papillary Mucinous Neo- INTRODUCTION: While perioperative complications 2 MA; Laboratory for Quantitative Medicine, Massachusetts GeneralBACKGROUND: Positron emission tomography (PET) plasm (IPMN) is increasingly diagnosed due to the ubiquity of pancreaticoduodenectomy(PD) have been well docu- Hospital, Boston, MA using 18F-fl uorodeoxyglucose (FDG) is increasingly used of cross-sectional abdominal imaging and a growing aware- mented, long-term complications of PD have not been well BACKGROUND: Lymph node ratio (LNR) has been shown in the preoperative diagnosis and management of pan- ness of the pathology. The treatment algorithm for IPMN characterized. In this study, we investigate complications to predict survival in patients with pancreatic cancer. How- creatic cancer. Its anatomic utility is limited however, and remains controversial and depends heavily on the malig- experienced by patients status post PD with pancreatico- ever, its role has not been evaluated in the context of the the quality of standard imaging continues to improve. We nant potential of the disease and anticipated outcomes. gastrostomy (PG) reconstruction more than 12 months number of lymph nodes (LNs) examined. evaluated the effect of PET on the management of patients METHODS: We evaluated all patients who underwent sur- after the procedure. referred for treatment of pancreatic cancer. METHODS: METHODS: A national population-based dataset (SEER) of gical resection for IPMN between January 1997 and Decem- We performed a retrospective review of METHODS: 10,254 patients and a prospective database of 827 patients We evaluated all patients referred to our cen- ber 2009. Adenomas and moderately dysplastic lesions patients who underwent PD with PG more than 12 months who underwent surgery at a single institution were reviewed ter between January 2006 and September 2010 with a fi nal were classifi ed as benign. High grade dysplastic lesions, car- prior the study period. Data was collected by chart review for tumor and patient characteristics. In each database, diagnosis of pancreatic ductal adenocarcinoma and exam- cinoma in situ (CIS), and invasive cancer were considered as well as a mailed survey. We performed a retrospective patients were divided into subsets based on the number ined the role of PET in their diagnostic workup and manage- malignant. Pathology reports were retrospectively reviewed analysis assessing the incidence of new diagnoses of dia- of examined LNs, with cut points putting approximately a ment. Patients who had PET following successful surgical to distinguish main-duct, mixed, and branch-duct types. betes mellitus (DM) and the use of pancreatic enzyme third of the patients into each subset. The SEER database was resection or to stage metastatic disease were excluded. PET Continuous variables were compared using Student’s t test replacement in patients more than one year after PD at a divided into: ≤5, 6–12, and ≥13 examined LNs. The num- “positivity” was determined by radiologist interpretation. and categorical variables were compared using Fisher’s single institution from 1990 to 2010. The main outcome exact tests. measures were a diagnosis of DM based on random and ber of examined LNs at the single institution was generally RESULTS: Between January 2006 and September 2010, 636 RESULTS: fasting blood glucose measurements, and patients’ require- higher and correspondingly the single institution’s patients patients were diagnosed with pancreatic ductal adenocarci- Between 1997 and 2009, 146 patients under- ments for pancreatic enzyme replacement for control of were divided into: ≤9, 10–16, and ≥17 lymph nodes. Univari- noma and 52% went on to resection. PET was utilized in went surgical resection for IPMN. During this period 88 steatorrhea and malabsorption. ate and multivariate analyses were done using Kaplan-Meier thirty with resectable disease and in thirty-four with locally patients were surveyed for suspected IPMN, and 278 under- curves and Cox regression modeling. advanced carcinoma. PET failed to identify three metastatic went pancreatectomy for other cystic neoplasms. The RESULTS: Our cohort included 92 patients who under- RESULTS: The different subgroups were uniform in terms lesions in two patients (two liver metastases, one perito- majority of patients with IPMN presented as an incidental went PD with PG more than 12 months prior to the begin- of patient’s age at presentation, sex, tumor size, stage neal metastasis) and one primary lesion. In fi ve patients, fi nding (42.6%), followed by abdominal pain (19%) and ning of the study. The median length of follow up was 4.2 and primary site (chi2, P < 0.05). In SEER, the mean LNR PET produced a false positive result leading to additional pancreatitis (17%). Patients with benign disease were on years after the procedure. Preoperative diagnoses included Poster Abstracts decreased in a stepwise manner as the number of examined imaging and procedures that delayed surgery by an aver- average younger than those with malignant disease (66.4 periampullary tumors in 59 (63%) of the patients, cystic lymph nodes increased (0.38 for ≤ 5LNs vs. 0.19 for ≥13 age of three weeks. PET identifi ed liver metastases in three vs. 70.6 years, p = 0.01). The majority were female (54%) tumors in 14 (15%) of the patients, neuroendocrine tumors Sunday LNs, P < 0.05). A similar trend was also seen in the single patients preoperatively. Two were confi rmed by diagnostic and white (86.2%). Most resections were partial pancreatec- in 9 (10%) of the patients, chronic pancreatitis in 6 (7%) of institution’s dataset (0.29 for ≤ 9LNs vs. 0.15 for ≥ 17LNs, laparoscopy, thus there was no deviation from standard tomies (82%), but 27 patients required a total pancreatec- the patients, and other pathology in 5 (5%) of the patients. P < 0.05). Overall, age at diagnosis (>65 yrs), absolute num- protocol. PET altered the surgical or medical management tomy. Main-duct type was identifi ed in 33%, branch-duct Of the 92 patients, 45 (48.9%) now require pharmacologic bers of positive LNs, and a higher LNR (>0.2) were asso- of ten (15%) patients. One patient with liver metastases did type in 23%, and mixed-type in 39% of patients. Of the therapy with pancrelipase for symptoms of steatorrhea and ciated with a worse survival. On multivariate analysis in not proceed to surgery. In two patients, the identifi cation main-duct and mixed lesions, 39% harbored malignancy, malnutrition. Twelve patients had been diagnosed with SEER, a Cox regression model showed that LNR > 0.2 has of liver metastases by PET altered chemotherapy goals from 50% of which were invasive carcinomas. Eight (24%) diabetes mellitus prior to undergoing the procedure and an adverse impact on survival in each sub-group and the neoadjuvant to palliative. In one patient, PET revealed a branch-duct lesions were high-grade dysplastic or CIS, but were excluded from the second part of the study; of the magnitude of this impact increased with the number of synchronous colon cancer which was resected at the time none were invasive carcinomas. The overall rate of malig- remaining 80 patients, 18 (23%) became diabetic after the examined LNs: HR (±95% CI); ≤5 LNs:1.52 (±0.20); 6–12 of the patient’s pancreaticoduodenectomy. In four cases of nancy for IPMN patients was 35%. IPMN was identifi ed on procedure. Seven of these patients (39%) now require insu- LNs: 2.95 (±0.55), and ≥13 LNs: 3.25 (±0.76). For the sin- diagnostic uncertainty, PET resulted in an earlier diagnosis the fi nal surgical margin in 25% of patients. Five patients lin therapy for diabetes management, while twelve (67%) gle institution, the Cox regression showed that a lymph of pancreatic adenocarcinoma and shorter time to surgi- have undergone completion pancreatectomy at an aver- are controlled on oral pharmacologic agents only. In total, node ratio >0.2 had an adverse effect on survival, except cal resection. In two patients, PET was performed to follow age of 17.8 months after their initial surgery (range 8.7– 20 of the 92 patients underwent adjuvant chemotherapy in the subset where ≤9 LNs were examined; HR, (±95% CI): lymphoma and resulted in a secondary diagnosis of pan- 27.5). Lesions requiring reoperation were either main-duct and 19 underwent adjuvant radiation therapy. Of the 1.37 (±0.69), P = 0.126. In the subgroups where LNR was a creatic adenocarcinoma. One patient proceeded to surgical or mixed and tended to be malignant (p = 0.05). None of 18 patients who developed diabetes, 5 (28%) underwent signifi cant predictor, the strength of this prediction corre- resection, and the other had metastatic disease. these patients had IPMN at the original surgical margin. adjuvant chemotherapy and 4 (22%) underwent adjuvant Three additional patients have recurred, two with locally radiation therapy. Of the 45 patients who now require lated positively with the numbers of examined LNs: HR, (± CONCLUSIONS: In cases of known pancreatic adenocarci- advanced disease. pancrelipase therapy, 11 (24%) underwent adjuvant che- 95% CI); 10–16 LNs: 1.61 (±0.49), P = 0.010; ≥17 LNs: 2.17 noma, PET rarely altered clinical management and resulted CONCLUSIONS: motherapy and 11 (24%) underwent adjuvant radiation (±0.86), P = 0.003. in additional tests and procedures. In our series, PET was Our series confi rms a low rate of malig- therapy. CONCLUSION: most useful in cases of diagnostic uncertainty. nancy in branch-duct pathology and supports the judi- Although an LNR of > 0.2 has a strong CONCLUSIONS: negative predictive value for survival, the accuracy of this cious surveillance of these patients. Risk of recurrence Symptomatic malabsorption requring prediction and the relative risk of death is higher when necessitates continued surveillance after surgery. Patients pancreatic enzyme therapy is a long-term complication of more lymph nodes are examined. This suggests that LNR with malignancy were older, suggesting a four year lag-time PD with PG. These patients should be monitored for mal- must always be interpreted in the context of the number of between adenoma and invasive disease. absorption. After PD with PG, patients are as likely as the examined lymph nodes. general population to develop diabetes mellitus.

88 89 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1645 Clinical: Small Bowel Su1647 Clinical: Stomach Prospective Comparison of Longterm Outcomes in Neoadjuvant Treatment of Duodenal Adenocarcinoma: Patients of Severe Acute Pancreatitis Managed by Su1646 A Rescue Strategy Su1648 Operative and Non-Operative Means Edwin O. Onkendi*1, Sarah Y. Boostrom1, David M. Nagorney1, Pre Resection Gastric Bypass Reduces Post Resection Laparoscopic Roux-en-Y Gastric Bypass vs. Laparoscopic 1 *1 1 John H. Donohue 1 1 1 Prasanna Chandrasekaran, Rajesh Gupta, Yalakanti R. Babu, BMI but Not Liver Disease in Short Bowel Syndrome , Michael L. Kendrick, Michael B. Farnell, Band for the Treatment of Morbid Obesity: Data from 3 2 2 Michael G. Sarr 1 1 2 Mandeep Kang , Surinder S. Rana, Deepak K. Bhasin, , Kaye M. Reid Lombardo, Michael G. Haddock, the Nationwide Inpatient Sample (NIS), 2006–2008 Rajinder Singh 1 Jon Thompson*, Rebecca A. Weseman, Fedja A. Rochling, Florencia G. Que 1 Wendy J. Grant, Jean Botha, Alan Langnas, David F. Mercer 1 1 Division of Surgical gastroenterology,Department of Surgery, Post Surgery, Mayo Clinic, Rochester, MN;2Radiation Oncology, Mayo Hossein Masoomi*, Kevin M. Reavis, Steven Mills, Michael J. Stamos, Surgery, University Nebraska Medical Center, Omaha, NE Graduate Institute Of Medical Education and Research(PGIMER), Clinic, Rochester, MN Ninh T. Nguyen 2 ABSTRACT: Surgery, University of California, Irvine- Medical Center, Orange, CA Chandigarh, India;Department of Gastroenterology, PGIMER, Patients who develop short bowel syndrome BACKGROUND: Recent advances in chemotherapy have Chandigarh, India; 3Department of Radiology, PGIMER, Chandigarh,(SBS) while morbidly obese maintain a higher BMI and are been shown to downsize initially unresectable colon can- INTRODUCTION: Laparoscopic Roux-en-y gastric bypass India more likely to develop hepatobiliary complications than cers. The role of neoadjuvant therapy in duodenal adeno- (L-GBP) and Laparoscopic gastric banding (Lap-Band) are non-obese patients. The mechanisms for these fi ndings carcinoma, especially its effect on resectability, disease-free the two most commonly performed bariatric operations in BACKGROUND: The present study reports the long term are unknown. Our aim was to determine the effect of pre survival (DFS) and overall survival (OS) is unknown. Our the United States. functional and morphological changes following severe resection gastric bypass (GBP) on postresection BMI and aim was to evaluate the long-term outcome in initially OBJECTIVE: To compare surgical outcomes of L-GBP ver- acute pancreatitis and compares the patients managed by the incidence of liver disease in SBS patients. unresectable patients with duodenal adenocarcinoma fol- operative and non-operative methods. sus Lap-band in morbidly obese patients. METHODS: We reviewed 100 adult patients with SBS: 55 lowing neoadjuvant chemotherapy and rescue surgery. METHODS: METHODS: Using the Nationwide Inpatient Sample (NIS) 30 patients who had completed one year of patients with initial BMI < 35 were controls; 28 patients METHODS: A retrospective review between 1/1994–1/2010 database, clinical data of morbidly obese patients who follow up after recovery from attack of acute pancreatitis with initial BMI > 35 were the obese group; and 17 patients of all patients who underwent rescue duodenectomy fol- underwent L-GBP or Lap-band from 2006–2008 was ana- were evaluated. 72 hours fecalfat analysis, glucose toler- had undergone GBP prior to SBS. lowing neoadjuvant therapy was performed. lyzed. Outcome measures included patient characteristics, ance test and pancreatic morphology (MRI) were examined RESULTS: RESULTS: and recurrences if any were noted. There were no differences in age, gender, or Ten patients received neoadjuvant chemother- comorbidities, postoperative complications, length of hos- intestinal anatomy. Obese patients were more likely to apy prior to surgical resection (7 men, 3 women) with a pital stay (LOS), in-hospital mortality and hospital charges. RESULTS: Etiology was alcohol (15), gall stones (10), and weaned off PN (44% vs25% control and 12% GBP, p < .05). mean age of 54 years (range 45–67 years). Reasons for unre- RESULTS: idiopathic (5). Twelve patients were managed non-opera- Overall 38% of patients have undergone ostomy closure, sectable disease were vascular encasement in 6 patients, Of the 254,760 patients, 183,452 patients under- tively, out of which fi ve patients needed pigtail drainage. intestinal lengthening and intestinal transplantation with retroperitoneal extension of tumor abutting on the aorta went L-GBP and 71,308 patients underwent Lap-band. The 18 patients required open necrosectomy. Mean follow up a similar occurrence in all groups 11 (68%) of the GBP or inferior venacava in 3 patients and bulky local disease majority of patients were female (L-GBP: 81.0% vs. Lap- was 28.2 months. 16 patients (62%) had readmissions. 11 patients had gastric continuity reestablished. Pre resection causing malignant bowel obstruction in 1 patient. Six were Band: 78.3%, p < 0.001) and Caucasian (L-GBP: 73.9% vs. patients (36.7%) had exocrine defi ciency and 14 patients BMI in controls was signifi cantly lower than obese and GBP primary presentations and 4 were local recurrences. Of the Lap-band: 75.5%, p < 0.001) in both groups. The mean age (46.7%) had endocrine defi ciency, 7 patients had both. groups (26 vs 43 and 37). BMI at 1, 2, and 5 years was simi- 6 primary presentations, 4 received neoadjuvant therapy in the L-GBP group was lower compared with the Lap-band group (43.4 vs. 45.8 years, p < 0.001). Most of comorbidi- 82.8% patients had morphological changes in pancreas. lar in control and GBP groups (23, 23, and 23 versus 23, 24, with FOLFOX, one with chemoradiation with 5-FU and Poster Abstracts ties were signifi cantly higher in the L-GBP group. The over- Patients undergoing necrosectomy had higher incidence of and 26). Obese patients had a persistently increased BMI one with CPT-11, oxaliplatin and capecitabine. Of the 4 patients with locally recurrent disease, one had radiother- all post operative complications was lower in the Lap-band endocrine dysfunction (61.1% of patients in surgical group (37, 32, and 32 at 1, 2, and 5 years). % IBW trends were Sunday group (1.39% vs. 4.18%; p < 0.001). Specifi c postoperative and 25% in non-operative group (p = 0.057).Exocrine similar. 7 (41%) of the GBP patients had a pre resection apy with 5-FU and capecitabine, one had FOLFIRI, one had complications were all lower in the Lap-band group (UTI, abnormality was also signifi cantly higher in surgical group BMI > 35. BMI and %IBW were similar at 1,2, and 5 years CPT-11, 5-FU and leucovorin, and one had FOLFOX. All 10 patients underwent R0 resection following neoadjuvant pneumonia, acute renal failure, respiratory failure, myocar- compare to non-operative group(55.6% versus 8.3%, p = in those GBP patients with BMI > 35 and those < 35 (26, therapy. Histologic evaluation revealed that 2 patients had dial infarction, venous thromboembolism, ileus, abscess, 0.010). Morphological changes in pancreas were also higher 26, and 26 vs 22, 20, and 26).Cholelithiasis and cirrhosis complete pathologic response (1 on FOLFOX and 1 on wound infection and bowel obstruction). Compared with in operated group but was not statistically signifi cant. occurred to a similar extent in all 3 groups. Radiographic fatty liver tended to be higher in the GBP group (41% vs chemoradiation with 5-FU) and one patient had only 10% L-GBP, patients who underwent Lap-band had a shorter CONCLUSION: Patients of severe acute pancreatitis on 16% controls and 32% obese groups). End stage liver dis- viable tumor remaining after FOLFOX. Two patients had mean LOS (1.16 vs. 2.41 days, p < 0.001), lower mortal- follow up had signifi cant functional, and morphological ease occurred only in patients on PN > 1 year and tended to >50% decrease in tumor size. The average tumor size was ity (0.03% vs.0.06%, p < 0.001), and lower mean hospital changes with 62% patients requiring readmission. Patients 3.4 cm (range <1 cm–4.6 cm). Eight patients had grade 3 be higher in obese and GBP patients compared to controls charges ($29,753 vs. $39,570, p < 0.001). who were managed non-operatively had lesser incidence of tumors and 2 patients had grade 4 tumors. Three patients (33% and 33% vs 17%). CONCLUSION: Compared to laparoscopic gastric bypass, exocrine and endocrine defi ciencies, though there was no had positive lymph nodes and 7 had negative lymph nodes. laparoscopic gastric banding is associated with better peri- signifi cant difference in pancreatic morphological changes. CONCLUSIONS: GBP prevents the nutritional benefi ts Five patients had T3 tumors, 3 had T4 tumors, one had of obesity in SBS patients. This occurs independent of pre T2, and one had no residual tumor identifi ed on patho- operative outcomes with a shorter length of stay, lower SBS BMI suggesting that GBP itself rather than surgically logic evaluation. Four patients with tumors in the fourth morbidity and lower mortality. Further studies are needed induced weight loss is the important factor. However, GBP portion (D4) and 4 patients with tumors in the third por- to examine the weight loss outcome between the two does not appear to eliminate the increased risk of hepatobi- tion (D3) underwent segmental resection, while 1 patient operations. liary disease observed in obese SBS patients. with tumor at the second portion (D2) underwent standard Whipple resection and 1 patient with tumor at D2/D3 junc- tion underwent standard Whipple with en bloc resection. On follow up, 5 patients are alive today (follow-up range 15–48 months) including 2 of the patients presenting with recurrent disease. Of these 5 patients, 2 had positive nodes and 3 had negative nodes. Since rescue surgery, all have no evidence of recurrent disease. CONCLUSION: Long-term survival can be achieved in select patients with initially unresectable duodenal adeno- carcinoma. From our experience, neoadjuvant chemother- apy may improve resectability of previously unresectable duodenal adenocarcinoma.

90 91 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Su1649 Su1650 Su1651 Su1652 Minimally Invasive Gastrectomy for Gastric Cancer: Pre-Operative Nomogram to Predict Risk of Laparoscopic Gastrectomy: A Single-Center ExperienceIs There a Golden Window for the LAP-BANDTM: A Single Institution Experience Peri-Operative Mortality Following Gastric Rebecca Kowalski*1,2, Jennifer Montes1,2, Tanuja Damani1,2, Greater Intra-Band Pressure Is Needed as Time Passes Maithao N. LE*, Joyce Ho, Evelyn L. Kachikwu, Vijay Trisal, Resections for Malignancy Paresh C. Shah 1,2 with LAGB? 1 2 Joseph Kim, Alessio Pigazzi, Joshua Ellenhorn Mashaal Dhir*1, Lynette M. Smith2, Fred Ullrich2, Quan P. Ly1, Surgery, Lenox Hill Hospital, New York, NY;Hofstra Medical School, David A. Nguyen*, Grace J. Kim, Carson D. Liu General & Oncologic Surgery, City of Hope, Duarte, CA Aaron R. Sasson 1, Chandrakanth Are1 Manhassett, NY Bariatric Surgery, Skylex Advanced Surgical, Inc, Los Angeles, CA 1 BACKGROUND: The application of minimally invasive Surgery, Division of Surgical Oncology, University of Nebraska BACKGROUND: Laparoscopic gastric resection has been INTRODUCTION: Lap band adjustments have tradition- techniques for gastric malignancy is in evolution. In Asia, Medical Center, Omaha, NE;2Biostatistics, University of Nebraska well established in Asia but has been slow to develop in the ally relied upon volume adjustment. However, individual laparoscopic assisted gastric resection is commonly per-Medical Center, Omaha, NE United States. Few centers have been able to replicate the anatomy, location of band placement, and time of implan- Asian experience with resection and outcomes. formed with open reconstruction of intestinal continuity. INTRODUCTION: Surgery remains one of the major tation between patients vary greatly, and the standard size In the United States, minimally invasive bariatric surgical treatment options available to patients with gastric cancer; METHODS: We present a single-institution retrospective of the lap band does not accommodate for these variations. Thus, band volume does not allow for a reliable explana- techniques have been adapted to enable gastric resection therefore operative mortality is an important consider- study of our recent experience with laparoscopic gastric tion to weight loss following an adjustment. Another factor with intracorporeal reconstruction. The aim of this study ation. There is currently no pre-operative, patient-specifi c resections from January 2007 to August 2010. All patients that may serve to explain the weight loss seen after adjust- is to access the short term outcomes of minimally invasive method to determine the likely peri-operative mortality undergoing planned resections for benign and/or malig- gastectromy and reconstruction for gastric malignancy at a nant disease were included in this study. Our technique ment is intra-band pressure. This study seeks to compare for each individual patient. The aim of this study was to correlations between intra-band pressure and band volume single institution. is similar to that described elsewhere in the literature. We develop a pre-operative nomogram based on the presence with duration of band placement and average weight loss. METHOD: Between November 2004 and October 2010, 68 of co-morbidities to predict risk of peri-operative mortality routinely perform a modifi ed D2 nodal dissection for all PARTICIPANTS: patients with gastric malignancies underwent MIG. Short following gastric resection for malignancy. adenocarcinoma including peri-pancreatic, hepatic and A random sample of 40 morbidly obese celiac nodes, sparing only the splenic hilum. All anasta- patients with the AP LAP-BANDTM implanted between 2007 term clinical and histopathologic results were analyzed. METHODS: The National Inpatient Sample database was moses were completed intra-corporeally. Wedge resections to 2010 was enrolled in the study. Participants must have had RESULTS: A total of 68 minimally invasive gastric resec- queried to identify adult patients that underwent gastric were performed with linear staplers. the band implanted for more than three months before the tions were performed on 31 men and 37 women with resection for malignancy. The pre-operative co-morbidities, RESULTS: time of study as well completing a 2nd follow up visit after the median age of 69 years (range 35–96). There were no con- identifi ed as predictors were used and a nomogram was A total of 39 consecutive patients with a mean adjustment visit within 90 days. Adjustments were conducted versions to open resection. Indications for the procedures created with multivariate regression using the Taylor age of 67.72 years of age were included in our analysis. The using a manometer to measure intra-band pressure. study included 16 men and 23 women. Demographics, ASA were gastric adenocarcinoma (56), gastrointestinal stromal expansion method in SAS Software, SURVEYLOGISTIC RESULTS: tumor (GIST) (7), high grade dysplasia (3), carcinoid (1) and Procedure. A Training set (Years 1993, 1996, 1997, 1999, status, pTNM stage, histologic type of the tumor, number Participants in this study loss an average of rabdomyosarcoma (1). There were 17 total gastrectomies, 2000, 2002, 2004 and 2005) was utilized to develop the of resected lymph nodes, and postoperative complications 7.6 lb ± 0.5 lb over a period of 30 days after their adjust- 40 distal, 5 proximal, and 6 wedge gastrectomies. Among model and a validation set (Years 1994–95, 1998, 2001 and were studied. The operations performed included total gas- ment. Furthermore, greater weight loss is correlated with more recent band implantation. Interestingly, there is also patients with gastric adenocarcinoma, 40 underwent a 2003) was utilized to validate this model. trectomy (8 pts), subtotal gastrectomy (13 pts), distal gas- a positive correlation between intra-band pressure and time D2 lymphadenectomy and 14 underwent a D1 lymphad- trectomy (6 pts), and wedge resection (10 pts). Indications RESULTS: A total of 14,235 and 9,404 patients were of band implantation. It takes more intra-band pressure to for resection were gastric adenocarcinoma in 22 patients Poster Abstracts enectomy. The median lymph node count was 27.5 (range included in the training set and validation set, respectively. (56.41%), GIST in 9 patients (23.08%), and ulcers or dyspla- obtain weight loss as time goes on. Data reveals that the 8–71) for D2 lymphadenectomy and 20.5 (range 8–29) for The overall actual observed peri-operative mortality rate for correlation between band volume and time of band place-

sia in 8 patients (20.51%). One total gastrectomy (2.56%) Sunday D1 lymphadenectomy. All patients were resected with neg- training set and validation set were 5.9% and 6.7%, respec- ment is very weak compared to intra-band pressure (R2 = was performed laparoscopic-assisted. The mean operating ative margins. The median operating time was 385 min- tively. Figure 1 depicts the constructed nomogram. The 0.07 and R2 = 0.66 respectively). time was 262 minutes (median 257 min, range 86–635 utes (range 76–579 minutes). Median estimated blood loss decile- based calibration plots for the training and valida- min). Average blood loss was 177 cc (median 150 cc, was 200 mL (range 10–1,000 mL). Median length of stay tion sets revealed a good agreement between the observed Intra-Band Pressure and Average Weight Loss range 25–600 cc). The average length of stay was 9.85 days for total gastrectomy was 7 days (range 5–19) and 6 days probabilities and nomogram predicted probabilities. The (median 8 days, range 2–86 days). The mean time to bowel Year of Band Average Intra-Band Average Weight Year of Band (range 1–39) for partial gastrectomy. There was one periop- accuracy of the nomogram was further reinforced by a con- erative mortality (1.5%), eight major (11.8%) and 12 minor function was 3.88 days (median 3.5 days, range 1–9 days). Implantation Pressure (psi) Loss (lbs) Implantation cordance index of 0.75 with 95% confi dence interval of 9/2007 3.70 –3.5 2007 (17.7%) complications. With a median follow-up of 12.5 The mean tumor size for GIST was 4.09 cm (range 1.4–13.3 0.73 to 0.77 calculated using the validation set. 1/2008 2.35 –6.4 2008 months (range 0.2–70.5), 82.4% of patients are alive with cm). Mean tumor size for adenocarcinoma was 4.34cm 4/2008 2.90 –6.5 2009 no evidence of disease. (range 1.3-16.3 cm) and the mean number of lymph nodes 8/2008 2.66 –9.8 2010 harvested was 16.39 (median 15, range 4–32 nodes). All 31 CONCLUSIONS: Minimally invasive gastrectomy provides 1/2009 2.27 patients with malignant disease achieved R0 resections. an oncologically sound alternative to open gastrectomy 4/2009 2.05 There was no mortality. There were no anastomotic leaks. 8/2009 1.62 and can be performed with a low mortality rate and an Major post-operative morbidities (defi ned as reoperation 1/2010 1.63 acceptable overall morbidity rate. for any cause, intra-abdominal abscess, further invasive Intra-band pressure data are divided into groups of three month interval. Average procedure, or duodenal stump complication) occurred 9 weight loss data are group by one year interval. times in 6 patients (15.38%). Minor post-operative morbid- ities occurred 14 times in 11 patients (28.21%). At the time of this report, one patient had died of unrelated causes, and CONCLUSION: This study reveals that intra-band pressure all remaining patients were alive and free of disease. is more highly correlated to weight loss after an adjustment than adjusted band volume. Furthermore, it seems that CONCLUSIONS: Laparoscopic gastric resection can be Pre-operative nomogram to predict peri-operative mortality following intra-band pressure increases as time goes on after the ini- performed with equivalent or improved peri-operative gastric resections for malignancy tial date of band implantation. As time passes, it takes more outcomes relative to open gastrectomy, while maintaining CONCLUSION: pressure to get the same response from the band. Perhaps This pre-operative nomogram has been oncologic principles. Long-term follow up is necessary to the capsule formation around the band and foreign body shown to accurately predict the risk of peri-operative mor- ensure oncologic equivalency. We are prospectively follow- reaction cause the band to need more pressure to infl ate. tality following gastric resections for malignancy. ing all gastrectomy patients to report short and long term Therefore, it is critical that patients take the band seriously oncologic survival. right after they have the procedure done as it takes more pressure to achieve the same weight loss as time passes.

92 93 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Combined Science 49656, and 49657. Patients were grouped based on func- MONDAY, MAY 9, 2011 tional status as listed in NSQIP: independent, partially dependent and totally dependent. Any patients with an Su2087 unknown functional health status were excluded. Pre- 12:00 PM – 2:00 PM operative and operative variables were recorded for all A Comparison of Short Term Outcomes Following patients. Short-term outcomes between the groups were Hall A Abdominal Wall Hernia Repair Based on Pre-Operativethen compared. POSTER SESSION I Functional Health Status RESULTS: 76,397 patients were identifi ed that underwent (NON-CME) Emily Albright*, John S. Roth, Daniel L. Davenport abdominal wall hernia repair. 74,785 were classifi ed as inde- University of Kentucky, Lexington, KY pendent (97.9%), 1,317 as partially dependent (1.7%) and INTRODUCTION: Abdominal wall hernias are a common 295 as totally dependent (0.4%). Classifi cation as totally problem facing surgeons today. As the patient population dependent was associated with a signifi cant increase risk respectively) after injury (Figure). However recovery was ages not only is there an increase in the number of co- in all short term outcomes examined: wound occurrence, Basic: Colon-Rectal complete and no signifi cant difference was seen pressures morbidities but a decrease in the ability to perform basic pneumonia, pulmonary embolism, urinary tract infection, when compared to NI group. daily activities. Within the American College of Surgeons myocardial infarction, deep venous thrombosis, sepsis, Mo1861 National Surgical Quality Improvement Program (ACS return to the operating room and death (p < 0.001 for all). At 5 weeks after IM treatment signifi cant peak pressures NSQIP) this is tracked as a patient’s pre-operative func- DISCUSSION: Given the signifi cant increase in compli- Sustained Anal Pressure Improvement After Anal (P ≤ 0.001) after MSC treatment were seen compared to NI tional health status. cations and mortality following ventral hernia repair in Sphincter Injury and Serial IV Infusions Suggests group. However, after IV treatment signifi cant resting (P = 0.01) and peak pressures (P ≤ 0.001) were seen compared METHODS: We reviewed all patients that underwent patients with limited functional health status, a period of Homing of Mesenchymal Stem Cells watchful waiting should be strongly advised, especially for to NI group Marked decrease in fi brosis and scar tissue was abdominal wall hernia repair from 2005–2010 in the ACS *1 1 2 asymptomatic hernias. Massarat Zutshi, Levilester Salcedo, Margot S. Damaser, seen in the MSC treated group (see fi gure). NSQIP database. Patients were identifi ed based on the fol- Marc Penn 3 lowing CPT codes: 49560, 49561, 49565, 49566, 49568, 1 Colorectal Surgery, Cleveland Clinic Foundation, Cleveland, OH; 49570, 49572, 49585, 49586, 49652, 29653, 49654, 49655, 2Biomedical Engineering, Cleveland Clinic Foundation, Cleveland, OH; 3Cell Biology and Regenerative Medicine, Cleveland Clinic Short Term Outcomes Based on Functional Health Status Foundation, Cleveland, OH

Independent % Partially Dependent % Totally Dependent % p Wound Occurrence 2,968 4.0 133 10.1 44 14.9 <0.001 We have previously shown that mesenchymal stem cells Pneumonia 479 0.6 69 5.2 44 14.9 <0.001 (MSC) improve anal sphincter pressures after injury. We Pulmonary Embolism 150 0.2 14 1.1 4 1.4 <0.001 aim to demonstrate MSC homing to the injured anal Urinary Tract Infection 648 0.9 46 3.5 25 8.5 < 0.001 sphincter by comparing anal sphincter pressures following Myocardial Infarction 57 0.1 6 0. 5 1.7 < 0.001 IM MSC injection and serial IV MSC infusion in a rat model Deep Venous Thrombosis 195 0.3 19 1.4 10 3.4 < 0.001 of acute anal sphincter injury. Sepsis 678 0.9 76 5.8 33 11.2 < 0.001 METHOD: Return to the OR within 30 days 1,615 2.2 105 8.0 45 15.3 < 0.001 45 virgin rats were divided into injury (n = 35) Death 178 0.2 58 4.4 44 14.9 < 0.001 and no injury (NI, n = 10) groups. The injury group was divided into saline(PBS) or MSC treatment and a control group (n = 5) which received no treatment. Each treat- ment group was further divided into IM and serial IV (n = 5) group . The MSC IM and IV (n = 5) and control groups were followed up for 5 weeks The injury was a partial anal

sphincter excision (PSE) of 25% of the anal sphincter. Anal Poster Abstracts pressures were recorded prior, 10 days and 5 weeks after

treatment with a balloon connected to a digital recorder. CONCLUSION: Although IM MSC treatment after injury Monday 24 hours after injury, the animals received 5 × 106 labeled causes increase in anal pressures it is not sustained at 5 MSC or 0.2 ml saline into the anal sphincter for IM treat- weeks . MSC home after serial IV infusion to the injured ment , while IV treatment group received the same dose anal sphincter causing increase in the anal sphincter pres- daily for 6 consecutive days via the tail vein. Anal sphinc- sures which are comparable to sham treatment and increase ters were harvested and submitted for Masson’s staining. till 5 weeks. Healing is by fi brosis in the control animals RESULTS: 10 days after IM treatment, signifi cant increase while the MSC treated group showed less scarring with IV in resting (P < 0.001) and peak pressures (P < 0.001) was infusion group showing the least scarring. seen after MSC treatment when compared with PBS after injury. (9.78 ± 0.84, 13.13 ± 1.2, respectively) vs. (6.23 ± 0.48, 8.32 ± 0.64, respectively).When compared with the NI group and MSC treatment, recovery of the anal pres- sures was not complete (resting (P = 0.04) and peak pres- sures (P = 0.02). The IV infusion group showed signifi cantly increased resting (P < 0.001) and peak pressures (P < 0.001) in MSC treated animals compared with PBS (11.03 ± 0.71, 16.68 ± 1.33, respectively) vs. (6.94 ± 0.28, 8.56 ± 0.34,

94 95 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

fi brous sheath; it can bind calcium after being phosphory- Mo1862 Mo1864 Mo1865 lated during sperm capacitation. Six transcript variants of Gene Chip Analysis for Detection of Potential Tumor The Impact of Citrulline on L-Arginine/Adma Ratio in CABYR encoding fi ve protein isoforms have been identi- HPP1 Mediates Tumor Suppression by Upregulation of Suppressor Genes in Colorectal Cancer Cell Lines Sepsis fi ed. Normally, CABYR is expressed in the human germ JAK1-STAT Signaling Pathways in Colorectal Cancer *1 2 1 1 line but not in adult human tissues, thus, it is considered *1,2 2 3 2 Michael C. Gock, Dirk Koczan, Ernst Klar, Michael Linnebacher Abul Elahi*, Jonathan M. Hernandez, Whalen Clark, Theodor Asgeirsson, SEN Zhang, Greg Cavey, Robert Nunoo, 1 a cancer testis (CT) protein. A high frequency of expres- 4 Department of General Surgery, University of Rostock, Rostock, Leigh Ann Humphries, Jian Wang, David Shibata Anthony Senagore 2 sion in tumor tissues and restricted expression in normal 1 2 Germany; Institute of Immunology, University of Rostock, Rostock, H. Lee Moffi tt Cancer Center and Research Institute, Tampa, FL Surgery/Colorectal, Spectrum Health, Grand Rapids,Research, MI; Germany tissues make CT proteins attractive cancer vaccine tar- Spectrum Health, Grand Rapids, MI; 3Research, Southwest Michigan gets. Increased expression of CABYR a/b and c isoforms in INTRODUCTION: The novel tumor suppressor gene, HPP1 BACKGROUND: Inovation Center, Kalamazoo, MI; 4Surgery/ Colorectal, University of Anomalies within the chromosomal lung cancers and isoforms c and d in brain tumors have is epigenetically silenced in over 80% of colorectal cancers.Southern California, Los Angeles, CA structure are fundamentally involved in tumorigenesis of been reported. The expression patterns of CABYR genes in Despite growing interest in HPP1 as a potential serum and colorectal cancer (CRC). Actual generations of gene chips colorectal cancer (CRC) are not well characterized. The aim stool-based biomarker, very little is known about its bio- INTRODUCTION: Sepsis leads to a complex systemic provide very high resolution to study phenomena like loss of this study is to evaluate the expression of CABYR a/b and logic mechanisms of action. We have previously reported response of cytokines (pro and anti- infl ammatory) which of heterozygoty (LOH) or total losses that are known to c in CRC tumors and to determine if they hold promise as that the growth suppressive effects of HPP1 require the may be further modulated by endothelial nitric oxide cause inactivation of tumor suppressor genes (TSG). In this vaccine targets. activation of STAT1 and STAT2 and are associated with the (NO). Pharmacologic manipulation of endothelial NO study we use the latest generation of gene chips to reveal METHOD: Consenting CRC patients (pts) who underwent concomitant downregulation of oncogenic STAT3. Janus via L-arginine supplemention or inhibitors has provided recurrent LOH and total losses in coding regions of CRC Associated Kinases (JAK) are the upstream regulators of inconsistent clinical data, referred to as the “arginine para- cell genome and thereby to reveal unknown TSGs. elective resection for whom tumor samples were avail- able were included in the study. Demographic, clinical, STAT proteins and can demonstrate cross reactivity. In this dox”. Bioavailability of L-arginine is negatively impacted METHODS: The genome of 7 established human CRC and pathologic data were collected prospectively. Tissues study, we sought to elucidate the roles of the different JAKs by metabolism of L-arginine in the liver, toxicity of paren- cell lines was analyzed using latest gene chip technology were OCT embedded and stored at –80° C until analysis. in mediating HPP1’s biologic functions. teral L-arginine, and asymmetric dimethylarginine (ADMA) inhibition. The ratio of concentrations of L-arginine and (1.800k Genome Wide SNP 6.0 chip by Affymetrix). Results Total purifi ed RNA was isolated from tissue samples and METHODS: Full-length HPP1 was amplifi ed and cloned ADMA correlates with improved endothelial function and were compared in selected cases with conventional cytoge- cDNA synthesized. CABYR a/b and c expression was ana- into the pcDNA3.1 expression vector which was then sta- is an indicator of improved NO synthase activity. L-citrul- netic (mFISH analysis). After selecting potential TSGs their lyzed by quantitative PCR (QPCR) using the SYBR Green bly transfected into the HCT116 colorectal cancer cell line. line offers an indirect and better tolerated means of supple- expression profi le on mRNA level was analyzed using end platform. Tumor and normal tissue expression levels were We have previously demonstrated that this results in sub- menting arginine and therefore improved NO availability. point and real time PCR. cDNA cloning with subsequent determined and compared as were tumor and testis CABYR stantial in vivo and in-vitro growth suppression. Western The goal of this study was to determine whether citrulline sequencing was performed to reveal possible mutations as expression levels. Tumors with expression levels 0.1% or Blot analyses were performed to assess both expression and (CIT) supplementation altered the L-arginine/ADMA ratio well as alternative splice variants. more than the testis were considered positive. Expression phosphotyrosine activation of JAK1, JAK2, JAK3, and TYK2 in a rat sepsis model with varying degrees of severity. RESULTS: Gene chip results were basically congruent to of CABYR a/b and c in a set of normal human tissues was in HPP1 transfectants and empty vector controls. Transient METHODS: those of mFISH analysis but provided a major improve- analyzed by RT-PCR. knockdown was achieved using RNA interference with sub- Sepsis was induced with 2 cecal ligation (CLP) ment in terms of resolution. In this initial analysis we lim- RESULTS: A total of 47 paired CRC and normal tissue sequent analysis of effects on STAT1 and STAT2 as well as models in Wistar rats. Citrulline supplementation was pro- ited us to examine a microsatellite unstable cell line. We specimens (18 M/29 F, age 65 ± 16.8) were studied (85% on cell behavior by MTT and soft agar assays. vided to half the animals as 8% citrulline supplemented feed for 3 weeks. Baseline L-arginine, ADMA, and citrul- found four genes as candidate TSGs, two of them turned colon, 15% rectal). Stage breakdown was as follows: Stage RESULTS: We have demonstrated that overexpression of line were assessed followed by comparison of the following out to be described as TSG for CRC (FHIT und WWOX). 2, 26; Stage 3, 19; and Stage 4, 2 pts. The percent of pts HPP1 results in a substantial increase in both the expression groups at day 0, 1, and 3: Sham operated; CLP 8 mm (local- CONCLUSION: with a relative expression ratio of malignant to normal tis- and activation of JAK1 with corresponding reductions for We discovered two genes as potential, ized); and CLP 12 mm (extensive). Blood was obtained via sues (MN ratio) over 1 was 70% for CABYR a/b and 72% TYK2 (the dominant kinase for STAT3). JAK1 siRNA treat- so far unknown, TSGs for CRC. These genes need further cardiac stick at the time of euthanasia. Measurements were for CABYR-c. The percent with both a MN ratio over 1 and ment of HPP1 transfectants resulted in a 90% reduction of investigation. Additionally gene chip analysis proved to be performed with liquid chromatography and mass-spec- expression levels over 0.1% of testis was 23.4% for CABYR- JAK1 expression. The inhibition of JAK1 resulted in a dra- a simple and useful technique detection of possible TSGs in trometry (LC/MS). Parametric 2 sample t-test and Mann a/b and 25.5% for CABYR-c. Except for very low expres- matic reduction in the levels of activated STAT1 and STAT2. tumor cell lines. Whitney U was used for statistical analysis with signifi - sion of CABYR a/b in the brain, CABYR a/b or c was not JAK1 knockdown in HPP1 transfectants also resulted in a cance set at p < 0.05. expressed in normal tissues. signifi cant 2-fold increase in proliferation (p < 0.04) and an Mo1863 RESULTS: CONCLUSION: RT-PCR analysis confi rmed the frequent 8-fold increase in anchorage-independent colony forma- 98 Wistar rats were evaluated and survival

The Cancer Testis Antigens CABYR a/b and CABY C Are Poster Abstracts overexpression of CABYR a/b and c in most CRC tumors tion (p < 0.01) as compared to controls. was similar (85%) in both sepsis models with and with- Expressed in a Subset of Colorectal Cancers and Hold out citrulline. The baseline L-arginine/ADMA ratio did not compared to adjacent normal tissues. In 23-26% of tumors CONCLUSIONS: Suppression of JAK1 activity in the set- change signifi cantly by the administration of CIT (113 vs. Monday Promise as Targets for Specifi c Immunotherapy expression was more than 0.1% of the expression level of ting of HPP1 overexpression results in an attenuation of 164 µmol/l, p = 0.078). The ratio improved signifi cantly on *1 3 2 testis. A larger and more diverse group of tumors (Stage STAT1/STAT2 activity and a restoration of proliferiative C.M. Shantha Kumara H, Otavia L. Caballero, Tao Su, post-op day 1 and 3 in the CIT/Sham compared to non- 1 2 1 1 1–4) needs to be assessed to determine if CABYR expres- potential. JAK1 appears to be the critical upstream regula- Xiaohong Yan , Aqeel Ahmed, Sonali a C. Herath, Vesna Cekic, citrulline group (104 vs. 222 µmol/l , p = 0.042 and 120 vs. 1 4 2 sion correlates with T, N, or fi nal tumor stage. Evaluation tor of HPP1-mediated growth suppression via activation of Daniel D. Kirchoff , Andrew J. Simpson, Carlos Cordon-Cardo, 241 µmol/l, p = 0.004) and the same was seen in the CIT/ of blood for anti-CABYR antibodies is also needed. These STAT1/STAT2 pathways. Richard L. Whelan 1 CLP8 group (115 vs. 186 µmol/l, p = 0.048 and 72 vs.153, 1 2 results support CABYR as a potential therapeutic vaccine Surgery, St. Luke Roosevelt Hospital, New York, HerbertNY; Irving candidate in CRC. p = 0.0180). There was no statistical difference in ratio Comprehensive Cancer Center, Columbia University, New York, NY; between CIT/CLP12 and non-citrulline group on post-op 3Ludwig Collaborative Laboratory for Cancer Biology and Therapy day 1 and 3 (68 vs. 97 µmol/l , p = 0.055 and 95 vs.112 Department of Neurosurgery, Johns Hopkins University School of µmol/l, p = 0.568). Medicine, Baltimore, MD;4Ludwig Institute for Cancer Research, CONCLUSION: The data suggest that 3 week supplemen- New York, NY tation with citrulline does not alter the baseline L-arginine/ ADMA ratios but does appear to improve both the response INTRODUCTION: Human CABYR is a testis specifi c poly- to sham laparotomy and localized sepsis. Citrulline may morphic calcium binding tyrosine phosphorylation regu- provide a safe means of modulating pathologic endothelial lative protein isolated from spermatozoa. This protein is dysfunction. Further work is needed to confi rm the results localized to the sperm fl agellum in association with the and mechanism of this effect.

96 97 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Basic: Esophageal Basic: Hepatic Basic: Pancreas Mo1869 Pancreatic Consistency and Hounsfi eld Unit on CT Mo1866 Mo1867 Mo1868 Scan in Mature (65+) Subjects Yuichi Kitagawa* The Prognostic Signifi cance of Multiple Molecular Replicative Stress-Induced Midkine Expression Pentoxifylline Inhibits Tumor Necrosis Factor Surgery, National Center for Geriatrics and Gerontology, Obu, Japan Markers in Blood of Patients with Multimodal Synthesis and Improves Liver Regeneration After and Secretion Activates Notch Signaling and Is INTRODUCTION: Treatment of Esophageal Cancer Partial Hepatectomy in Rats by a Mechanism Related Linked to Epithelial-Mesenchymal Transition and Consistency of the pancreas is one of to Inhibition of TGF Beta 1 Expression Chemoresistance in Pancreatic Cancer the most important prognostic factors for pancreatic anas- Jan Brabender*, Peter P. Grimminger, Arnulf H. HöLscher, tomosis. However, pancreatic stiffness had been evaluated, Daniel Vallbohmer, Elfriede Bollschweiler, Ralf Metzger Rodrigo B. Martino, Ana Maria M. Coelho, Marcia S. Kubrusly, Cenap GüNgor, Hilke Zander, Florian Gebauer, Michael Tachezy,subjectively. We introduce a tactile sensor for digitize the Department of Surgery, University of Cologne, Cologne, Germany Sandra N. Sampietre, Marcel C. Machado*, Luiz Augusto C. Emre F. Yekebas, Jakob R. Izbicki, Maximilian Bockhorn* consistency of the pancreas. BACKGROUND: Aim of this study was to evaluate the D’Albuquerque General, Visceral and Thoracic Surgery, University Hospital, MATERIALS AND METHODS: Gastroenterology, University of Sao Paulo, Sao Paulo, Brazil Hamburg, Germany A material has its own prognostic signifi cance of ERCC1 (Excision Repair Cross Com- resonance frequency. If material touches an oscillated plementing gene), COX2 (cyclooxygenase-2), TS (thymidylate BACKGROUND/AIM: Liver regeneration in small for Pancreatic ductal adenocarcinoma (PDAC) still remains an object, a shift of the resonance frequency will be observed. synthase), DPD (Dihydropyrimidine-Dehydrogenase)-RNA sizes liver grafts and in subtotal may be sup- exceptional case among solid tumors, since its incidence The difference between the frequencies under pre and post expression and the single-nucleotide-polymorphisms (SNP) of pressed increasing the mortality. TNF-D and IL-6 are strong nearly equals mortality rate. Despite formidable improve- oscillate conditions depends on the stiffness of the object. ERCC1, GNAS, and XRCC1 in blood of patients with neoadju- promoters of hepatic regeneration after hepatectomy and ment of understanding the molecular mechanisms contrib- So, the consistency of material can be measured by moni- vant treated cancer of the esophagus. recognized as initial phase cytokines in infl ammatory uting to cancer progression and metastasis, no real progress toring the shift in the frequency. The tactile sensor system MATERIAL AND METHODS: A total of 29 patients with response to systemic infection or injury, and in liver dam- has been done in catching cellular mechanisms governing is composed of a sensor probe, an amplifi er and a fi lter; locally advanced cancer (cT3-T4, Nx, M0) of the esopha- age following hepatectomy or ischemia/ reperfusion injury. chemotherapy-resistance in cancer cells, although almost the probe is connected to a piezoelectric transducer with a gus were included in this study. Blood samples from each TGF-beta 1 is an anti-infl ammatory cytokine and a potent all PDACs are intrinsically chemoresistant. It was recently resonance frequency of 57 kHz. Measurements were made patient were drawn prior to neoadjuvant radio-chemo- growth inhibitor. There are evidences that TNF-D induces shown that chemoresistant PDAC cells overexpress the 200 times per second, and the magnitude of frequency therapy (cis-Platin, 5-FU, 36Gy). Transthoracic en-bloc TGF-beta 1 expression in many cell types. We hypothesized Notch-2 receptor and acquire an epithelial-mesenchymal change was processed by connected computer with origi- esophagectomy was performed in all patients following that pentoxyfi line (PTX) an inhibitor of NF-kappa-B could transitioned (EMT) phenotype. Cellular mechanisms that nal software. 85 patients with mean age 70.1 ± 9.5 were completion of neadjuvant therapy. After extraction of cel- improve liver regeneration by reducing TGF-beta 1 expres- assist an EMT phenotype are not completely understood. registered with this study. These patients did not have any lular tumor-RNA from sera samples, quantitative molecular sion in the liver and the deleterious effects of high levels However, the present study identifi ed Midkine (MK) as fre- clinical sign of pancreatitis. 35 patients were excluded from analysis was done by real-time RT-PCR (Taqman©). Histo- of TNF-D after partial hepatectomy. The aim of the present quently overexpressed in chemoresistant PDAC through this study, because of high serum amylase level. And mean morphological regression was defi ned as major response study was to evaluate the effect of pentoxifylline on liver gene-expression profi ling and real-time PCR. We also found serum amylase level was 92.9 ± 21.5 IU/L. Patients were when resected specimen contained <10% of residual vital regeneration after partial hepatectomy in rats. that MK expression is inducible by the chemotherapeutic divided two groups into over and under 65 years old. Con- sistency of the pancreas was measured with a tactile sen- tumor cells, and minor response with >10% of vital resid- METHODS: Wistar rats were randomized into 2 groups: gemcitabine in a dose-dependent manner in chemoresistant sor system (Venous Handy Biosensor system TM, AXIOM, ual tumor cells. Group C (n = 26) animals submitted to 70% hepatectomy PDAC cells whereas no induction was observed in chemo- sensitive PDAC cells confi rmed by real-time PCR. In addi- Fukushima, Japan). Pancreas head and body (neck) were RESULTS: Nineteen of 29 (65.5%) of patients showed a and saline administration, and Group PTX (n = 26) animals tion, depletion of MK by RNAi correlated with decreased measured 2 or 3 times on the same point. If the patients minor histopathological response and 10 of 29 (34.5%) submitted to 70% hepatectomy receiving pentoxifylline proliferation rates and more interestingly with a strong had pancreatic tumor, the measurement was performed showed a major-response to neadjuvant radio-chemother- (25 mg/kg). AST, ALT, TNF-D, and IL-6 serum levels were increase in sensitization towards gemcitabine. Main attrac- far from the tumor. Hounsfi eld Unit (HU) on CT scan apy. The median survival of patients was 2.08 years (0.15– measured at 2 and 6 h after hepatectomy. At 48 h after tion is the identifi cation of Notch-2 receptor as a new MK was measured in head and neck of the pancreas. HU was 4.53). Blood TS RNA expression and the ERCC1 C118T hepatectomy residual liver were removed and stained with interactor in PDAC cells. Surprisingly, our results indicate measured on SOMATOME Sensation Cardiac-64 (SIEMENS SNP were signifi cantly associated with patient’s survival. hematoxylin-eosin. Hepatocyte proliferation was assessed that the MK-Notch-2 interaction activates Notch signaling Japan, Tokyo, Japan) CT system with SIEMENS image server Patients with a high TS expression had a median survival by mitotic index and by proliferating cell nuclear antigen in PDAC cells leading to induction of EMT, upregulation using the original software. of 1.1 years (0.21–3.16) compared to 3.36 years for patient’s (PCNA) staining. Total RNA was isolated from liver tissues of NF-.B and increased chemoresistance. Combined, these RESULTS: with a low TS expression (p = 0.031, Logrank test). Patient’s and qRT-PCR analysis for TGF-beta 1. The consistency of the pancreas head was 670.0 results suggest that gemcitabine induced MK upregulation ± 172.9 Hz in mature and 708.0 ± 239.3 Hz in younger, with the ERCC1 genotype GG had a median survival of 0.3 RESULTS: AST and ALT levels were not signifi cantly Poster Abstracts is strikingly involved in maintaining a drug resistance phe- respectively. The consistency of the pancreas neck was years (0.1–0.5) compared to 2.8 years (0.4–1.9) for the AG reduced in rats treated by PTX. TNF-D and IL-6 blood lev- notype through binding and activation of Notch signaling. 737.1 ± 178.4 Hz in mature and 684.2 ± 174.7 Hz in genotype and 3.0 years (0.4–2.2) for the AA genotype (p < els increased at 2 and 6 hours after partial hepatectomy. Monday Thus, inhibition of this interaction may represent a novel younger, respectively. The HU of the pancreas head was 0.001). There were no associations between patient’s clini- Administration of PTX signifi cantly decreased serum strategy to attain the battle of chemoresistance in PDAC. 42.9 ± 4.9 HU in mature and 43.0 ± 4.3 HU in younger, cal variables and gene expression status. levels of TNF-D and IL-6 at 2 hours after resection. A sig- respectively. The HU of the pancreas neck was 43.7 ± 5.4 CONCLUSIONS: nifi cant reduction in TGF-beta 1 expression in liver tis- TS RNA expression and the ERCC1 C118T HU in mature and 44.4 ± 5.7 HU in younger, respectively. sue was observed in rats receiving PTX. The pathological SNP in blood are associated with the prognosis of patients There was no correlation between age and hardness of the results obtained 48 hours after the hepatectomy showed an with neoadjuvant treated esophageal cancer and are poten- pancreas. Also, there was no relation with hardness and HU increased mitotic index and PCNA labeling index in ani- tial non-invasive prognostic markers in this disease. Future in each age group. studies are warranted to determine the clinical potential of mals treated with PTX compared to saline treated rats CONCLUSION: This method is easily maintained the con- this molecular approach. CONCLUSION: Pentoxifylline decreases the systemic sistency of pancreas non-invasively. Further investigation infl ammatory response reduces liver TGF-beta 1 expres- will be needed for elucidate the relationship between con- sion, enhancing liver regeneration and may be useful to sistency and Hounsfi eld Unit of pancreas. improve the function of small for size liver grafts.

98 99 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1870 Basic: Small Bowel Mo1872 B Lymphocytes Inhibit Proliferation of Pancreatic Eating Behavior in Rats Subjected to Ileum Cancer Mo1871 Transposition and Sleeve Gastrectomy Patrick B. White*, Sue Wang, Deborah A. Swartz-Basile, Helene Johannessen*1, Yosuke Kodama1, Hailong Cao1, TNF-D Induced Vectorial Secretion of IL-8 Corresponds Henry A. Pitt, Keith D. Lillemoe, Nicholas J. Zyromski Baard Kulseng 1,2, Chun-Mei Zhao1, Duan Chen1 to Development of Transepithelial Electrical Resistance 1 Surgery, Indiana University School of Medicine, Indianapolis, IN Department of Cancer Research and Molecular Medicine, Norwegian in Caco-2 Cells INTRODUCTION: Tumor immunology is important in University of Science and Technology, Trondheim, Norway; cancer progression, but poorly understood. Few data evalu- Dennis I. Sonnier*, Stephanie R. Bailey, Timothy A. Pritts, 2Departments of Surgery and Endocrinology, St. Olav’s University Figure 2: Development of TNF-Į induced vectorial secretion of IL-8 ate the effects of B lymphocytes on cancer. In our model Alex B. Lentsch Hospital, Trondheim, Norway Department of Surgery, University of Cincinnati, Cincinnati, OH in Caco-2 cells. IL-8 production was measured after interval apical or of murine pancreatic cancer in obesity, we have observed basolateral TNF-Į treatment (100 ng/ml). Control cells were treatedBACKGROUND/AIM: with We have reported that eating fewer circulating B lymphocytes, reduced B lymphocyte The intestinal epithelium, a vital part of the innate immune serum free media. *p < 0.05 vs. other groups by ANOVA and SNKbehavior post differed in rats that underwent gastric bypass, gene expression, and fewer tumor infi ltrating B lympho- system, provides a physical barrier between gut lumen and hoc test. sleeve gastrectomy, and/or duodenal switch procedures. cytes in the tumor microenvironment of obese mice. These lamina propria and plays an active role in the infl amma- Ileum transposition has been recently suggested as a meta- mice also had larger tumors than control mice. These tory response. Previously, we found that polarized intes- METHODS: The Caco-2 cell, a transformed colon cancer bolic surgery. The aim of the present study was to examine observations suggest that decreased B lymphocytes, as seen tinal epithelial cells respond to an infl ammatory stimulus cell line, exhibits characteristics of intestinal epithelial cells the eating behavior in rats that underwent ileum transposi- in obesity, permit accelerated pancreatic cancer growth. We in a directional, vectorial manner, with secretion of IL-8 in culture. Cells were plated on Transwell fi lters for up to tion (IT) and sleeve gastrectomy (SG). therefore hypothesized that B lymphocyte activity inhibits directed toward the infl ammatory stimulus. 32 days. At intervals, groups of cells (n = 3) were treated METHODS: the growth of murine pancreatic cancer. Male rats were subjected to laparotomy or IT, Under specifi c conditions, cellular differentiation in vitro with TNF-D (100 ng/ml) or serum free media in the apical and 6 weeks later both groups underwent SG. Body weight mirrors the in vivo process, where the cell matures into a or basolateral chamber. After 24 hours, production of the was recorded weekly. Calorie intake, body composition, columnar, polarized epithelial cell with distinct apical and chemokine IL-8 was determined in apical and basolateral eating behavior, energy expenditure and fecal energy den- basolateral regions. Transepithelial electrical resistance media by ELISA. Electrical resistance was measured imme- sity were measured by comprehensive laboratory animal (TEER) is used to detect formation of tight junctions and as diately prior to treatment and TEER was calculated by mul- monitoring system, dual x-ray absorptiometry, and bomb a measure of apical to basolateral polarity. The relationship tiplying the resistance times the area of the monolayer. calorimeter before and after surgery. of TEER to cellular processes such as brush border enzyme RESULTS: TEER increased slowly from day of plating until RESULTS: Body weight was lower in rats undergoing IT transport, nutrient receptors and barrier function is well day 13. TEER increased markedly between days 13 and 18, than those subjected to laparotomy (at 1 week, p < 0.05) described, the relationship between TEER and the intesti- 2 to a peak level above 2,000 Ω cm and maintained a level of until 6 weeks (p > 0.05) postoperatively. Fat compartment nal epithelial cell’s infl ammatory polarity is unknown. 2 1,500 Ω cm or greater for the remainder of the experiment. was reduced at 2 weeks after IT compared to laparotomy We hypothesized that increased TEER corresponds with the Prior to the increase in resistance, treatment of apical or (p = 0.02). After SG, body weight was reduced in both intestinal epithelial cell’s ability to respond to an infl am- basolateral chambers with TNF-D resulted in equal apical groups but more so in IT-rats (pre-operation 466 ± 14 g matory signal in a vectorial manner. and basolateral secretion of IL-8. After the TEER increase, as vs. post-operation 378 ± 21 g, p < 0.01), which was due to on day 20, apical TNF-D treatment resulted in only apical reduced fat compartment. Satiety ratio was higher during IL-8 secretion while basolateral stimulation results in both daytime than nighttime regardless of surgery procedures. apical and basolateral IL-8 secretion. Subsequent apical or At both 2nd and 6th week after IT, satiety ratio was reduced basolateral treatments, as on day 27, maintained this vecto- during daytime but not nighttime. Calorie intake per 24 rial secretion pattern, but with a signifi cant increase in the h per rat as well as per 100 g body weight was increased, magnitude of basolateral secretion. which was on account of increases in number of meals, METHODS: Seven-month old C57 Bl/6J mice were studied. CONCLUSIONS: In the present study, the ability of intes- meal size, and meal duration, particularly during daytime. Splenic lymphocytes were extracted by pressure injection tinal epithelial cells to respond to infl ammatory stimuli Rate of eating was unchanged neither during daytime nor and B lymphocytes isolated by nylon wool microfi ltration. in a vectorial manner correlated with an increase in TEER. nighttime. Energy expenditure (kcal/h/100 g body weight)

The B lymphocytes were activated with 10 μg/ml LPS. For These fi ndings are important because they suggest the was unchanged. The fecal energy density was unchanged. Poster Abstracts time course evaluation, B cells were incubated for 18, 24, 42, importance of cell polarity in the development of the intes- In laparotomy rats following SG, calorie intake relative to and 48 hours before collection of the supernatant media. For tinal epithelial cell’s mature infl ammatory phenotype. body weight was increased due to an increased appetite Monday during daytime 2 weeks after SG. In IT-rats following SG, concentration evaluation, B cells were seeded at 1.5 × 106, however, satiety ratio, calorie intake and energy expendi- 3 × 106, and 6 × 106 cells/ml and incubated for 24 hours. ture were unchanged neither during daytime nor night- 50μl of the conditioned media was collected and added to time, but eating behavior was altered as characterized by 5000 PAN02 murine pancreatic cancer cells. After 20 hours reduced rate of eating during both daytime and nighttime, of growth, a standard MTT assay was performed to evalu- reduced meal size during nighttime, and increased meal ate proliferation. Student’s t-test and ANOVA were used to duration during both daytime and nighttime. compare results, with a p value < 0.05 considered signifi cant. Figure 1: Time course of transepithelial electrical resistance in Caco-2 CONCLUSION: After IT procedure, satiety ratio was cells. Resistance values were measured twice weekly after plating. reduced and calorie intake increased, but these changes RESULTS: B lymphocyte conditioned media signifi cantly TEER was calculated by multiplying the resistance ( ) by the area of the Ÿ disappeared after additional SG. Instead, the altered eating inhibited PAN02 proliferation in both a time- and concen- membrane (4.5 cm2 tration-dependent fashion relative to control (see fi gure). ). Note the marked increase in TEER between days behavior took place as manifested by eating slowly with 13 and 18. small meal size, which may have the metabolic benefi ts. CONCLUSION: These data show that B lymphocyte con- ditioned media inhibited PAN02 proliferation in both a time- and concentration dependent fashion. We conclude that B lymphocytes play an important role in the tumor immunology of pancreatic cancer.

100 101 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

vs. 55.1 ± 19, p = 0.017) under hypoxic conditions while Mo1873 Basic: Stomach no signifi cant reduction was seen in normoxia compared Hypoxia Upregulates Neurokinin-1 Receptor (NK-1R)to controls. Signaling in Rat Peritoneal Mesothelial Cells (RPMCs):CONCLUSIONS: Ischemia, or hypoxia, contributes to Mo1875 A Key Early Event in Intraabdominal Adhesion adhesion formation through the upregulation of NK-1R Overexpression of GLO1 Promotes Tumor Invasion in Formation signaling suggesting that early hypoxia-induced signaling Gastric Cancer pathways may be novel targets for adhesion prevention. Melanie L. Gainsbury*, Holly K. Sheldon, Somdutta Mitra, Chia-Siu Wang*1, Kwang-Huei Lin2 Rami Abdou, Daniel I. Chu, Karen L. Reed, Arthur F. Stucchi, 1 Department of General Surgery, Chang Gung Memorial Hospital, Mo1874 James M. Becker Chia-Yi, Pu-Tz City, Taiwan;2Department of Biochemistry, Chang Surgery, Boston University Medical Center, Boston, MA Topographic Expression of Intestinal Sugar Sensors Gung University, Taoyuan, Taiwan and Transporters in Obesity: Tissue Analysis of Patients INTRODUCTION: Intraabdominal adhesions are an inevi- Complementary DNA (cDNA) microarrays were applied to table consequence of surgery. Despite good surgical tech- Undergoing identify aberrantly expressed genes in patients with gas- nique, unavoidable ischemic and hypoxic events can Hina Y. Bhutta*1,2, David B. Rhoads2,3, Stanley W. Ashley1,2, tric cancer. Among the most exuberantly expressed genes exacerbate adhesiogenesis. Our laboratory has shownAli Tavakkolizadeh 1,2 Glyoxalase I (GLO1) was one that encodes. We assessed that the NK-1R, its ligand substance P (SP) and decreased 1 General Surgery, Brigham and Women’s Hospital, Boston, MA; whether overexpression of GLO1 was a potential risk fac- peritoneal tissue plasminogen activator (tPA) secretion 2 3 tor for gastric cancer. GLO1 mRNA transcripts and protein are correlated with adhesiogenesis in vivo. However, the Harvard Medical School, Boston, MA;Pediatric Endocrinology, Massachusetts General Hospital, Boston, MA expression levels were determined by Western blot test and mechanism by which NK-1R expression is regulated during real time quantitative PCR (qPCR). These tests verifi ed that adhesiogenesis remains unclear. We hypothesized that sur- INTRODUCTION: Obesity and associated type 2 diabe- Figure 1: Topographic mRNA expression of SGLT1 and T1R2 in the jejunum. Proximal jejunum compared to distal jejunum.*P = 0.05 the expression of GLO1 was higher in gastric cancer tissues gically-induced hypoxia, as a result of tying or clamping off tes mellitus (T2DM) have reached epidemic proportions compared to adjacent noncancerous tissues. Furthermore, vessels or electrocautery, may be a critical event modulat- around the world. Roux-en-Y Gastric Bypass surgery (RYGB) Immunohistochemistry were studied in paraffi n-fi xed sec- ing NK-1R expression and adhesion formation. The aim of remains the gold standard therapy for obesity that also tions of gastrectomized specimens of 114 patients. The IHC this study was to evaluate the effect of hypoxia on NK-1R leads to a rapid resolution of T2DM. Our previous rodent scores of GLO1 in patients with serosa invasion (T3 and T4) signaling and tPA secretion in vitro. studies have suggested a role for SGLT1 in mediating these were signifi cantly higher than those without serosa inva- METHODS: Primary RPMCs were isolated 24-hrs after metabolic benefi ts. These experiments had shown that sion (T1 and T2). Also, the IHC scores of GLO1 were signifi - our previously described adhesion-inducing ischemic but- RYGB alters SGLT1 function by blunting intestinal taste cantly greater in patients with lymph node metastasis than ton surgery. Confl uent cells were incubated under hypoxic receptor stimulation, leading to decreased SGLT1 func- those without metastasis, and were signifi cantly higher in tion. There is a paucity of data on intestinal expression of stages III and IV than in stages I and II. Five-year survival (1% O2) or normoxic (21% O2) conditions for 24-hrs and exposed to SP (10 and 100 nM), a NK-1R antagonist (NK- these transporters in humans. We aimed to investigate the of patients with lower expression of GLO1 gene was sig- 1RA; CJ-12,255, Pfi zer, Inc.), or SP+NK-1RA. Cells were topography of intestinal expression of glucose transporters nifi cantly better than among patients with a higher expres- harvested to assess the upregulation of NK-1R signaling along with sweet taste receptors amongst an obese cohort sion. To investigate the functional signifi cance of GLO1 by Western blot for phosphorylated-ERK1/2 proteins, key of patients undergoing RYGB. expression, GLO1 short hairpin RNA (shRNA) expression downstream intracellular signaling molecules in the NK-1R METHODS: 10 normoglycemic patients (4 male, 6 female) plasmids were transfected into gastric cancer cell lines. pathway. Media was harvested to measure tPA secretion via undergoing RYGB for obesity were enrolled in this study Cancer cells depletion of GLO1 signifi cantly reduced the ELISA. In addition, rat peritoneal tissue was collected to over a 2 month period (mean age = 51 years, mean BMI migration and invasion ability of each cell line. The results assess NK-1R expression by immunofl uorescence and tissue = 44.7). Whilst undergoing RYGB surgery, jejunal biopsies point to GLO1 as prospective prognostic marker for gastric hypoxia by pimonidazole staining. were taken 40 cm (proximal) and 140 cm (distal) from the cancer and its role in gastric cancer progression. RESULTS: Pimonidazole staining was increased in peri- ligament of treitz. Following RNA extraction from the tis- Key words: gastric cancer, molecular biomarker, prognostic toneal tissue after ischemic button surgery compared to sues and reverse transcription, mRNA expression of SGLT1, factor, GLO1 non-operative controls. Immunofl uorescence showed GLUT2, and T1R2 (a protein involved in the sweet taste Figure 2: Correlation between jejunal SGLT1 mRNA expression and Poster Abstracts upregulation of the NK-1R in vivo, primarily localized to receptor heterodimer) was measured using qRT-PCR. Sta- BMI. R = 0.29 for distal jejunum the mesothelial cell layer on the surface of ischemic but- tistical analysis was performed using paired students’ t-test and Pearson’s correlation coeffi cient. Monday tons within 6 hrs of surgery compared to non-operative CONCLUSION: Transporters and sensors that we believe controls. In vitro, phospho-ERK1 levels were increased in RESULTS: Intestinal expression of SGLT1 was higher in are involved in achieving the metabolic benefi t of RYGB are hypoxic RPMCs by 1.6 fold with both 10 nM SP (1.0 vs. the proximal jejunum than in the distal jejunum (p = present in human tissue and show topographic variation 1.66 ± 0.12; p < 0.05) and 100 nM SP (1.0 vs. 1.61 ± 0.43; p 0.05) (Figure 1). There was also a trend towards an inverse within the jejunum. Importantly, the glucose transporter < 0.05) compared to normoxic controls. Phospho-ERK2 lev- correlation between SGLT1 expression in distal jejunum SGLT1, which accounts for almost all active transport of els were also increased in hypoxic RPMCs by 1.6 fold (1.0 and BMI (r = 0.29) (Figure 2). GLUT2 mRNA had a higher glucose across the apical border of the intestine, shows an vs. 1.64 ± 0.22; p < 0.05) with 10 nM SP and 1.8 fold (1.0 expression pattern in the proximal jejunum compared to inverse correlation with BMI in the obese range. Further vs. 1.86 ± 0.24; p < 0.05) with 100 nM SP versus normoxic distal, although this was not signifi cant (p = 0.16). There work is needed to elucidate the role of this transporter in controls. NK-1RA attenuated this effect by reducing phos- was no association between GLUT2 levels and BMI. There the etiology of obesity. pho-ERK1 levels by 54% (1.0 vs. 0.46 ± 0.13; p < 0.01) and was no topographic difference or correlation with BMI in phospho-ERK2 levels by 51% (1.0 vs. 0.49 ± 0.16; p < 0.01). T1R2 expression. SP-mediated tPA secretion was decreased by 45% (100 ± 19

102 103 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Clinical: Biliary Mo1585 CONCLUSION: These data suggest that using the MAGS Mo1587 device for SSL cholecystectomy results in equivalent or Magnetically Anchored Cautery Dissector Improves better operative outcomes and less workload compared to Gangrenous Cholecystitis: A Diffi cult Diagnosis Mo1584 Triangulation and Depth Perception During Single Site using only laparoscopic instrumentation, since better tri- Jacqueline J. Choi*1, Brian A. Coakley1, Kai B. Dallas1, Laparoscopic Cholecystectomy angulation and depth perception are afforded. AdditionalSimon Buttrick 1, Irini A. Scordi-Bello2, Scott Q. Nguyen1, Cholecystectomy in the Very Elderly: Is Ninety the 1 Nabeel Arain*1, Erin Webb1, Deborah C. Hogg1, Richard Bergs3, investigations are encouraged. Celia M. Divino New Seventy? 1 Jeffrey A. Cadeddu 2, Raul Fernandez3, Daniel J. Scott1 Surgery, The Mount Sinai School of Medicine, New York, NY; Attila Dubecz*, Hubert J. Stein, Miriam Langer, Michael Schweigert,1 2 Department of Surgery, University of Texas Southwestern Medical Mo1586 Anatomic and Clinical Pathology, The Mount Sinai School of Rudolf J. Stadlhuber Center, Dallas, CA;2Department of Urology, University of Texas Medicine, New York, NY Surgery, Klinikum Nurnberg, Nurnberg, Germany Prognostic Impact of Thymidylate Synthase Expression Southwestern Medical Center, Dallas, TX;3Texas Manufacturing INTRODUCTION: Gangrenous cholecystitis (GC) rep- BACKGROUND: in Adjuvant Gemcitabine Plus S-1 Chemotherapy After Nonagenarians are the fastest grow-Assistance Center/Automation and Robotics Research Institute, Surgical Resection for Bile Duct Cancer resents an advanced variant of acute disease ing sector of population across Western Europe, AlthoughUniversity of Texas Arlington, Arlington, TX which carries a signifi cant risk for increased morbidity and Hironori Kobayashi*, Yoshiaki Murakami, Kenichiro Uemura, prevalence of gall bladder disease is high, elective cholecys- INTRODUCTION: mortality. There are no diagnostic criteria for GC, making it tectomy is still controversial in this age group. The purpose of this study was to evalu- Takeshi Sudo, Yasushi Hashimoto, Akira Nakashima, Naru Kondo,a challenging diagnosis to make pre-operatively. We set out ate operative outcomes and workload during Single Site METHODS: Hiroki Ohge, Taijiro Sueda to determine which factors were predictive for GC in order A retrospective chart review was conducted Laparoscopy (SSL) using a Magnetic Anchoring and Guid- Graduate School of Biomedical Sciences, Hiroshima University, to better identify which patients might benefi t from early of done in patients over 90 years of age ance System (MAGS) cautery dissector compared to a con- Hiroshima, Japan surgical management. at our institution between 2004 and December 2009. Dur- ventional laparoscopic hook cautery dissector (LAP). METHODS: ing this period, a total of 3,007 cholecystectomies were OBJECTIVE: Although the prognosis in patients with bile The medical records of 200 patients (76 with METHODS: Each dissector was used to perform SSL cho- gangrenous and 124 with acute, non-gangrenous cholecys- performed on patients of all ages. Data collected included duct cancer remains poor, adjuvant gemcitabine plus S-1 titis) treated at The Mount Sinai Medical Center from March demographics, patient co-morbidities, indications for sur- lecystectomies in non-survival porcine models (n = 6, 3 chemotherapy (GEM + S-1) after surgical resection for bile gery, type of surgery performed, intraoperative fi ndings, MAGS, 3 LAP). A single surgeon with a standardized team 2003 to September 2009 were retrospectively reviewed. duct cancer has been shown to improve survival. S-1 is a Specifi cally, presenting symptoms, physical examination, histology, perioperative morbidity and mortality. performed all procedures in a 2-day period. The device used novel oral fl uoropyrimidine combination including tega- in the fi rst operation was randomized and devices were laboratory values, radiographic fi ndings (ultrasonogra- RESULTS: Twenty-two nonagenarians (19 females) under- fur (a prodrug of 5-fl uorouracil; 5-FU), dihydropyrimidine phy and computed tomography), perioperative data and alternated thereafter. For both MAGS and LAP procedures, dehydrogenase (DPD) inhibitor (5-chloro-2,4-dihydroxy- went cholecystectomy during the study period. Of these an umbilical multiport access device with two 5mm and pathological fi ndings were recorded. Univariate analysis patients, 8 (33%) had diabetes, 12 (50%) had hyperten- pyrimidine), and orotate phosphoribosyltransferase (OPRT) was carried out with a two-tailed chi-square test for each of one 12 mm ports was used. For MAGS cases, the cautery inhibitor (potassium oxonate). To clarify the relationship sion, and 10 (42%) had coronary artery disease. Nineteen the categorical variables, and a two-sample t-test with the device was inserted through the umbilical fascial defect, between expression of intratumoral enzymes related to patients (86%) underwent an emergency procedure. In Welch correction for the continuous variables. Multivariate coupled magnetically, and the operating arm was deployed; the metabolism of 5-FU and its derivatives and response one patient cholecystectomy was indicated for non-resolv- analysis was then performed using a binary logistical linear 2 graspers (1 straight and 1 articulating) and a laparoscope to adjuvant chemotherapy with GEM + S-1 for bile duct ing pain after attempted conservative therapy, only two regression model and the model of best fi t was determined. were used for retraction and visualization, respectively. For cancer, we evaluated thymidylate synthase (TS), DPD, and P-values of 0.05 or less were considered to indicate statisti- patients were operated electively. Laparoscopic cholecys- LAP cases, 2 percutaneous retraction sutures, 1 articulating OPRT expression immunohistochemically in resected bile tectomy was attempted in 13 patients (59%), 3 patients cal signifi cance. SPSS for Windows (Version 18.0.2, 2010 grasper, a hook cautery dissector, and a laparoscope were duct cancer tissues. Chicago: SPSS Inc.) was used for all analysis. needed a conversion, and 8 patients (36%) considered used. Operative outcomes and surgeon ratings (scale 1–5, 1 METHODS: Polyclonal antibodies were used to immu- RESULTS: unfi t to undergo a laparoscopic approach had an open pro- = superior rating) were recorded. Workload was assessed by On univariate analysis, multiple comorbidi- nostain sections of 106 formalin-fi xed paraffi n-embedded cedure. Mean operation time was 71 minutes. Histology the surgeon and the assistant using a previously validated ties were associated with a signifi cantly increased risk of specimens of bile duct cancer resected between 1989 and showed gangrenous cholecystitis in 14 (64%) patients. The NASA-TLX rating tool which assessed mental, physical, GC, including age over 50 years (p = 0.002), diabetes mel- 2010. The relationship between intratumoral TS, DPD and litus (p = 0.002), coronary artery disease (CAD) (p < 0.001), mean length of stay was 10 days (4-23d). Two patients (9%) temporal, performance, effort, and frustration levels (scale OPRT expression and prognosis was evaluated statistically. hypertension (p < 0.001), hyperlipidemia (p = 0.002) and required intensive care following surgery. There were no 1–10, 1 = superior rating). Comparisons used Mann-Whit- common bile duct injuries, one patient had a cystic stump RESULTS: Out of 106 patients, 41 (38.7%) received adju- steroid use (p = 0.008). In addition, gangrenous pathology ney tests; p < 0.05 was considered signifi cant. presented more frequently with fever > 38 C (p = 0.010), leak. One patient died in the postoperative period (4.5%). RESULTS: vant GEM + S-1 chemotherapy. High intratumoral TS, DPD All patients with an emergency operation were classifi ed as Pig weight was 54 ± 1.9 kg and compressed and OPRT expression was present in 52 (49.1%), 39 (36.8%) elevated heart rate > 100 (p = 0.025), WBC > 13,000 (p = 0.002), and neutrophil shift >75% (p < 0.001). Multivari- at least ASA III. Conversion rate, percentage of open pro- abdominal wall thickness for MAGS was 1.8 ± 0.1 cm. and 57 (53.8%) cases, respectively. Comparison of overall ate analysis confi rmed that CAD, alcohol use, steroid use, Poster Abstracts Operative outcomes and surgeon ratings for MAGS and survival between High and Low intratumoral DPD or OPRT cedures, percentage of advanced histology, ASA-score and nausea, leukocytosis, neutrophil shift and bilirubin levels hospital stay were signifi cantly higher when compared to LAP (respectively) were: operative time 25.3 ± 1.5 vs. 28.7 expression revealed no signifi cant difference regardless the displayed the highest predictive capacity for GC. Using a Monday all patients. ± 6.4 minutes (n.s.), EBL 5.0 ± 0 vs. 28.3 ± 40.4 cc (n.s.), application of adjuvant GEM + S-1 chemotherapy. In the critical view achievement 1.0 ± 0 vs. 2.0 ± 1.0 (n.s.), ease of logistical regression model of best fi t, we demonstrated that CONCLUSION: GEM + S-1 (+) group, overall survival was signifi cantly lon- a scoring system utilizing the above factors could diagnose Our study demonstrates that in unselected dissection 2.0 ± 0 vs. 3.7 ± 0.6 (n.s.), ergonomics 1.0 ± 0 vs. ger in the Low TS subgroup than in the High TS subgroup nonagenarians,cholecystectomy is safe with acceptable 4.0 ± 0 (n.s.), technical challenges 2.0 ± 0 vs. 3.7 ± 0.6 (n.s.). GC with a sensitivity of 81%, specifi city of 77%, PPV 70% (P = 0.015), whereas in the GEM + S-1 (-) group, there was and NPV of 90%. perioperative morbidity and mortality even as an emer- For both MAGS and LAP cases, complete cholecystectomies no signifi cant difference between the High TS and Low TS gency procedure. However, our data also suggests that cho- were achieved and bile spillage was similar. For MAGS, subgroups. Moreover, in the High TS group, there was no GC was associated with a signifi cantly higher rate of post- lecystitis appears to be a neglected condition in this age depth perception and instrument triangulation were excel- signifi cant difference in overall survival between the GEM operative complications when compared to acute cholecys- group. Elective cholecystectomy should be considered even lent and there were no complications. For LAP, the parallel + S-1 (+) and GEM + S-1 (-) subgroups, whereas in the Low titis (OR = 1.99, p = 0.05). in the very elderly. arrangement of the laparoscope and the dissector hindered TS group overall survival was signifi cantly higher in the DISCUSSION: GC represents a complex clinical problem depth perception and caused numerous instrument con- GEM + S-1 (+) subgroup by univariate analysis (P < 0.0001). which is frequently diffi cult to diagnose preoperatively and fl icts; complications included a superfi cial liver laceration CONCLUSION: Low intratumoral TS expression was asso- often results in poor clinical outcomes. This study shows from a Keith needle and an inadvertent burn to the dia- ciated with increased overall survival in patients with bile that a host of risk factors are correlated with GC and, phragm during gallbladder dissection. For MAGS and LAP duct cancer who received adjuvant GEM + S-1 chemother- therefore, may potentially be used for diagnostic purposes. The high sensitivity of our prediction model shows that (respectively), surgeon workload ratings were 2.6 ± 0.2 vs. apy. TS is a relevant predictive marker of benefi t from adju- relatively few clinical variables may be used to objectively 5.6 ± 1.1 (p < 0.001) and assistant ratings were 4.8 ± 0.8 vs. vant GEM + S-1 chemotherapy in patients with resected 5.8 ± 1.9 (n.s.). stratify patients at risk for GC and, thus, determine which bile duct cancer. individuals may benefi t from prompt surgical intervention.

104 105 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Clinical: Colon-Rectal abdominal colectomy with ileoproctostomy provides a sat- CONCLUSIONS: Colorectal surgery results in signifi cant Mo1591 isfactory outcome for patients with colonic inertia, how- ongoing risk of SBO according to the colorectal type of pro- ever, this should be offered only after thorough evaluation cedure. This risk seems to be similar between laparoscopic Total Abdominal Colectomy for Refractory Ulcerative Mo1588 of whole gut transit time. and open approach, higher after elective surgery and for Colitis—Evolution in Surgical Treatment patients with previous surgery. Number of readmissions for Alessandro Fichera*1,3, Marco Zoccali1, Carla Felice2, Can Whole Gut Scintigraphy Optimize Patient Mo1589 SBO predicts the need of surgery. David T. Rubin 2,3 Selection and Outcomes in Slow Transit Constipation? 1 Surgery, University of Chicago, Chicago,2 Medicine,IL; University of *1 2 1 Adhesional Small Bowel Obstruction After Open and Deborah Keller, Murali Pathikonda, Amit Khanna, Mo1590 Chicago, Chicago, IL; 3Infl ammatory Bowel Disease Center, University 2 Laparoscopic Colorectal Surgery: A Prospective Henry P. Parkman of Chicago Medical Center, Chicago, IL 1 A Critical Analysis of 28 Patients with Metachronous Surgery, Temple University Hospital, Philadelphia,2Medicine, PA; Longer-Term Study Peritoneal Dissemination from Colorectal Cancer PURPOSE: Total abdominal colectomy (TAC) with end Temple University Hospital, Philadelphia, PA Pierpaolo Sileri1, Stefano D’ Ugo*1, Luana Franceschilli1, ileostomy is the procedure of choice for patients with Giulio P. Angelucci 1 1 1 Ashraf Haddad*, Jesus Esquivel INTRODUCTION: Whole gut transit scintigraphy (WGTS) , Mara Capperucci, Emanuele Picone, medically refractory ulcerative colitis (UC). A laparoscopic 1 1 2 General Surgery, St. Agnes Hospital, Baltimore, MD distinguishes isolated colonic from generalized GI motilityPaolo Gentileschi , Nicola Di Lorenzo, Vincenzo Formica, approach has been shown to be safe and effective and has disorders. Including WGTS in the pre-operative work-upMario Roselli 2, Achille Gaspari1 BACKGROUND: Peritoneal dissemination occurs approx- become the preferred strategy in our practice. We report allows appropriate diagnosis and management, and should1 Surgery, University of Rome Tor Vergata, Rome, Italy;2Oncology, imately in 8% of patients with colorectal cancer at the time the evolved experience of a single colorectal surgeon (AF) be the standard of care for evaluation of colonic inertia.University of Rome Tor Vergata, Rome, Italy of diagnosis and in about 25% more at the time of disease and compare results of laparoscopic-assisted (LA) TAC, hand- Our primary objective was to evaluate functional outcomes BACKGROUND: recurrence. It has been traditionally seen as a universally assisted (HA) TAC and single incision laparoscopy (SIL) TAC. following colectomy for slow transit constipation in a sin- Open colorectal surgery (CRS) leads to lethal condition with no reasonable treatment options as METHODS: Since May 2010 ten consecutive patients with gle institution using preoperative scintigraphy. Secondary high rates of adhesive small bowel obstruction (SBO) and most patients will present in an advanced stage. medically refractory UC underwent SIL TAC and were case objectives were to describe predictive factors of outcome incisional hernia development with large clinical impact METHODS: We reviewed the records of all patients in our matched by age, gender, BMI and smoking history to 10 after surgery. and fi nancial burden. We evaluated the cumulative inci- dence of access related complications in a cohort of patients colorectal cancer with peritoneal dissemination database LA TAC (from Feb 2003 to Jan 2007) and 10 HA TAC (from METHODS: After obtaining Institutional Review Board who underwent open and laparoscopic CRS. from 2005 to 2009. A retrospective analysis of key clinical Feb 2006 to Apr 2010). Patient, disease and surgery-related approval, we identifi ed all patients who had surgery for and histopathological features including their TNM clas- METHODS: We reviewed cases of elective or emergency factors were analysed and short-term outcomes were colonic inertia from January 2003 to August 2010 using sifi cation at the time of diagnosis of their primary tumor compared. surgical billing codes and a GI departmental database. All CRS patients kept prospectively on a database and exam- and their Peritoneal Surface Disease Severity Score (PSDSS) RESULTS: There were no statistical differences in age, patients who had WGTS prior to surgery were included ined annually. Case notes were studied for SBO episodes at the time of diagnosis of their peritoneal dissemination gender, BMI and smoking history between the groups per in the analysis. Demographics, defecation, constipation requiring admission or reintervention. Development of was carried out. the case match design. Disease duration, histologic disease details, pharmacologic treatment, symptoms, psychiatric incisional hernia with or without repair was also recorded. RESULTS: Twenty eight patients with metachronous peri- activity, Mayo score for ulcerative colitis, nutritional and medications, operation performed, and outcomes were The diagnosis of SBO was defi ned by a combination of clin- toneal dissemination were identifi ed. There were 17 (61%) infl ammatory parameters, steroid and anti-TNF therapies extracted from a retrospective chart review and follow-up ical criteria and imaging. Time interval of SBO, surgery type male patients. Mean age was 54 years. There were 7 patients did not differ between groups. SIL TAC patients were more telephone interviews. and setting, readmission length and fi ndings at reinterven- tion were recorded. Patients undergoing CRS for infl amma- with stage II (A or B) and 21 patients with stage III (A, B or likely to have received immunosuppressive therapy within RESULTS: Fourteen patients had surgery for colonic iner- tory bowel disease, patients with peritoneal carcinosis, or C). Mean overall time to developing carcinomatosis was 30 days of the surgery (p = 0.016) than the other groups. tia during the study period. Patients were mainly young patients with SBO secondary to local or peritoneal recur- 27.8 months amongst all 28 patients. There were 6 patients There were no conversion or intraoperative complications females (86%, mean age 44.1 years) with a high coexistence rence during the follow-up were excluded. Patients who with stage IIIC. Mean time to developing carcinomatosis noted. The estimated blood loss did not differ between of psychiatric disease (79%). The most common surgical underwent other abdominal surgery during the follow-up was 16.8 months in this group and 5 out of 6 (83%), pre- groups. The length of surgery was signifi cantly shorter for intervention performed was total abdominal colectomy were also excluded. Data were analyzed using Mann-Whit- sented with a PSDSS of 4, making them not candidates for SIL TAC (139.0 + 23.7 minutes) when compared with LA with ileoproctostomy (79%); 3 patients had an end ileos- ney U test and chi-square test. The Kaplan Meier method cytoreductive surgery and HIPEC. (270.9 + 45.4) and HA (182.8 + 31.6) (p < 0.001). SIL TAC tomy. WGTS results were available for 11 of 14 patients; 3 was used for cumulative probability of developing SBO. CONCLUSION: It appears from these data that patients patients tolerated solid diet intake sooner (3.0 + 0.5 days) patients had their studies at an outside facility. All surgical RESULTS: with stage IIIC colorectal cancer develop carcinomatosis at than the other groups (3.7 + 0.5) (p = 0.019). Postoperative patients had delayed 72-hour transit limited to the colon From 01/03 to 11/10, 911 patients satisfi ed our criteria and underwent elective (52.6%) or emer- a faster rate and in a more virulent form when compared length of stay and short-term complication rates did not

without coexistent gastric or small bowel delay on WGTS. to other stages. Trials at preventing carcinomatosis in this differ between groups. Poster Abstracts Seven patients were available for follow-up interview at a gency (47.4%) CRS (68.7% open and 31.3% laparoscopic). Median follow-up was 46.2 months (range 0.2 -115.0). group of patients are needed. CONCLUSIONS: In this experience of a single surgeon, SIL

median interval of 23 months. Post-operatively, 86% had Monday TAC was associated with shorter procedure time, faster time complete resolution of pain and bloating, while 14% had Sixty-three patients (6.9%) experienced 83 SBO episodes to solid diet and was equally safe to traditional LA TAC and considerable relief. Post-operative stool consistency was and 22 required surgery (2.4%). There was a large varia- HA TAC. Additional study and experience will solidify this semi-solid in all patients. At follow-up patients reported tion in the time of fi rst SBO occurrence, 43.9% occurred approach for future patients and surgeons. between 2–5 bowel movements daily, while pre-operative within 3 months, 29% between 3 and 12 months and 27% frequency averaged one per week. No patients reported after 1 year. The risk of surgery at fi rst admission for SBO fecal incontinence or required anti-diarrheal medication 24% and the number of readmissions predicted the need after surgery. All patients contacted for follow-up reported of surgery. The risk of reoperation was greatest during the that they were overall satisfi ed with the outcome of their fi rst year after CRS and steadily rised every year thereafter. surgery, claimed to feel better than before surgery, and SBO was higher after pelvic surgery or extensive resections would make the decision to have the procedure again. compared to minor procedures (14% vs 3%; p < 0.0001; HR 7.33). Likewise, SBO risk was higher after elective compared CONCLUSION: The diagnosis of colonic inertia mandates to emergency surgery (11.1% vs 6.9%; p = 0.03; HR 2.0, but careful patient selection and may be improved with WGTS. similar after open compared to laparoscopic surgery (9.9% We believe appropriate patient selection can optimize surgi- vs 7.3%; p > 0.05; HR 0.8). Any previous or additional sur- cal outcomes. A pre-operative diagnostic regimen of WGTS gery raised the overall risk of SBO from 5.4% to 16.4%. and anal manometry combined with surgical intervention Incisional hernia development was slightly superior, after was associated with high patient satisfaction and improve- open surgery. ment in constipation, abdominal pain, and bloating. Total

106 107 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1592 gestive heart failure (OR: 1.72), chronic pulmonary disease Mo1594 Mo1595 (OR: 1.40), metastasis (OR: 1.34), male gender (OR: 1.23) Outcomes of Right Versus Left Colectomy for Colon and left colectomy (OR: 1.12) all independently increased Early Postoperative Outcomes After Laparoscopic Laparoscopic Colorectal Surgery for Bowel Endometriosis Cancer the risk of abscess/leak. There was no effect of age, diabetes, Segmental Colorectal Resection for Endometriosis: with Transvaginal Resection and Specimen Extraction: Hossein Masoomi*, Brian Buchberg, Phat T. Dang, hypertension, liver disease, obesity, anemia, peripheral vas- The Impact of Surgical Experience Perioperative Results in 40 Consecutive Patients Joseph C. Carmichael, Steven Mills, Michael J. Stamos cular disorder and surgical technique (laparoscopic versus Giacomo Ruffo1, Stefano Partelli*1,2, Filippo Scopelliti1,2, Giacomo Ruffo1, Stefano Crippa*1,2, Stefano Partelli1,2, Surgery (colorectal division), University of California, Irvine-Medical open) on abscess/leak in these patients. Stefano Crippa 1,2, Alberto Sartori1, Giuliano Barugola1,2, Alberto Manzoni 2, Marcello Ceccaroni3, Paride Mainardi4, Center, Orange, CA CONCLUSION: Patients in the right colectomy group were Lorenzo Rettore 1, Massimo Falconi2 Massimo Falconi 1,2, Luca Minelli3 1 1 INTRODUCTION: Colocolonic or colorectal anastomosis older and had more comorbidities and postoperative com- Department of Surgery, Ospedale “Sacro Cuore-Don Calabria”, Negrar,Department of Surgery, Ospedale Sacro Cuore Don Calabria, Negrar, perceived as more technically challenging than ileocolic plications compared to the left colectomy group. In-hospi- Italy; 2Department of Surgery, University of Verona, Verona, Italy Italy; 2Department of Surgery, University of Verona, Verona, Italy; anastomosis. Therefore, right colectomy (RC) is generally tal mortality and length of hospital stay were comparable BACKGROUND: 3Department of Gynaecology, Ospedale Sacro Cuore Don Calabria, between groups. Left colectomy had higher intraoperative The aim was to determine the predictors believed to be a simpler operation with better outcomes Negrar, Italy;4 Department of Radiology, Ospedale Sacro Cuore Don complications and total hospital charges than right colec- of postoperative complications in women undergoing laparo- than left colectomy (LC). Calabria, Negrar, Italy tomy. Patient characteristics and comorbidities were more scopic colorectal resection (LCR) for deep bowel endometriosis. OBJECTIVES: Our study was intended to compare out- important in predicting abscess/leak than anastomotic types. METHODS: Patients undergoing LCR from 2002 to 2009 BACKGROUND: Bowel resection represents a safe and fea- comes between right and left colectomy in patients with were identifi ed (n = 720). Demographic and clinical data sible option in treating severe bowel endometriosis and is colon cancer and identify factors that increase the risk of Mo1593 as well as perioperative details were assessed. A complica- associated with several well-known benefi ts. An abdominal developing postoperative abdominal abscess or anasto- tion was defi ned by the presence of any deviation from the incision is however needed for specimen extraction. Trans- motic leakage (abscess/leak) in these patients. Sacral Nerve Stimulation for Fecal Incontinence: normal postoperative course. Records were analyzed retro- vaginal approach can be an interesting approach for colorec- METHODS: Using the 2007 Nationwide Inpatient Sample Predictors of Long-Term Success spectively using logistic regression. tal resections in order to avoid mini-laparotomy. Aim of this (NIS) database, patients who underwent elective RC and Tracy L. Hull*2, Steven D. Wexner1, Anders Mellgren3, RESULTS: study is to evaluate our experience with laparoscopic trans- The median patient age was 35 years (IQR 29; vaginal colorectal resection for bowel endometriosis. LC (left hemicolectomy or sigmoidectomy) for colon can-John A. Coller 5, Robert D. Madoff3, Ghislain Devroede4 36). Signifi cant dischezia (visual analogue scale [VAS] > 5) 1 METHODS: cer were examined. Patient characteristics, comorbidities, Cleveland Clinic Florida, Weston, FL;2Cleveland Clinic Ohio, Ohio, was the most common symptom (59%) and 360 (50%) 40 patients (median age 30 years, range: 26–43) perioperative complications, length of stay (LOS), totalFL; 3University of Minnesota, Minneanapolis, MN;4Centre Hospitalier women presented a radiological stenosis > 30%. Overall, underwent laparoscopic colorectal resection for symptom- atic deep infi ltrating endometriosis of the rectum or rec- hospital charges and in-hospital mortality were evaluated.Universitaire de Sherbrooke, Sherbrooke, QC, Canada; 5Lahey Clinic, 468 patients (65%) had previous surgery. In 558 (77%) tosigmoid at a referral center. Laparoscopic intracorporeal Regression analysis was performed to identify independentBerlington, MA patients a recto-sigmoid resection was carried out whereas predictors of abscess/leakage. a rectal resection and an ileocolic resection were performed division of the distal rectum and exteriorization of the sur- PURPOSE: Sacral nerve stimulation (SNS) is effective in gical specimen via a colpotomy incision were performed RESULTS: A total of 50,799 patients underwent elective in 19% and 3% of cases, respectively. The median opera- the treatment of urinary incontinence and is currently in order to complete the resection. Transanal mechanical RC and LC for colon cancer during 2007 (RC: 63.54%; tive time was 300 minutes (IQR 240; 360) with a median under review by the FDA for the treatment of fecal inconti- colorectal anastomosis was then carried out. All colorectal LC: 36.46%). Overall, 9.63% were performed laparoscopi- blood loss of 200 ml (IQR 150; 400). A temporary ileostomy nence (FI) in subjects who have failed or are not candidates resections were performed by the same surgeon. Periopera- cally (RC: 9.70% vs. LC: 9.47%, P = 0.39). RC patients had was performed in 128 cases. Postoperative mortality was for other treatments. The aim of this study was to analyze tive outcomes were analyzed. more comorbidities. However, similar numbers of RC and nil whereas 71 (10%) women experienced a complication. the results of patients with FI treated with SNS in a large RESULTS: LC patients had distant metastases at the time of surgery The patients were discharged on median day 8 (IQR 7; 10). 19 patients (47.5%) had previous laparotomic prospective clinical trial to identify any potential predic- (30.7% vs. 30.8%; p = 0.77). Table below shows outcome The most frequent complication was anastomotic stricture (n = 3) or laparoscopic (n = 16) surgery for endometriosis. tors of therapeutic success 3 years after implant. measures of RC and LC. (3%) followed by hemorrhage (2.5%), rectal vaginal fi stula All patients underwent laparoscopic removal of multiple METHODS: Patients with ≥50% improvement in FI epi- (1%) and anastomotic leakage (1%). At univariate analysis, implants of deep infi ltrating pelvic endometriosis. No hys- terectomy was performed and there was no conversion to Outcome Measures of Right and Left Colectomy for Colon Cancersodes and/or days per week during a 10-14 day test stimu- the presence of a temporary ileostomy, an ultra-low anas- lation period using the chronic lead were implanted with tomosis and the number of consecutive operations (<300 laparotomy. At preoperative imaging 30 patients (75%) had Right Colectomy Left Colectomy a Medtronic InterStim neurostimulator. This analysis com- cases), correlated signifi cantly with postoperative compli- bowel stenosis (median stenosis rate 55%; median length Outcome Measures N = 32,276 N = 18,523 P-valuepared patients who had ≥50% improvement in FI episodes cations (odds ratio: 2.68, 4.13 and 1.93, respectively, P < of stenosis 2.7 cm). All patients underwent a laparoscopic Age (mean, year) 70.4 65.8 < 0.01per week (success) to patients who had <50% improvement 0.05). A number of consecutive operations <300 proved resection of the rectum or of the rectosigmoid, and low Female (%) 54.2 46.5 < 0.01 colorectal anastomoses was performed in 77.5% of patients. at 3 years. Baseline demographics were compared between to be the only independent predictor of complications, Poster Abstracts Caucasian (%) 80.0 75.7 < 0.01the groups including age, gender, BMI, duration, severity, shown through multivariate analysis (odds ratio: 2.25, CI 12 patients (30%) had a temporary ileostomy. Median opera- Patients with •1 0.30 1.32 < 0.01 tive time (including both gynaecologic and colorectal sur- type, and etiology of FI, number of pregnancies and deliv- 95% 1.33-3.80, P = 0.003). Monday Intraoperative gery) was 330 minutes, and median blood loss was 280 mL. Complications (%) eries, irritable bowel syndrome, internal and external anal CONCLUSIONS: LCR for endometriosis is a safe procedure Median ileus was 3 days, median length of stay was 7 days. Overall Frequency 39.18 37.0 < 0.01sphincter defects, prior surgeries, and quality of life. of Postoperative having a 10% rate of complications with no mortality. The Mortality was nil, morbidity was 15%. Three patients devel- Complications (%) RESULTS: 66 of the 77 subjects (86%) achieved therapeu- surgical experience is the only independent predictor of oped recto-vaginal fi stula and another one an anastomotic Patients with •1 28.43 26.75 < 0.01tic success with a median reduction of 96% in episodes complications after LCR for intestinal deep endometriosis. leakage. Overall, four patients (10%) required reoperation, complication (%) of FI; 31 (40%) patients achieved complete continence. A learning curve exists for LCR since a signifi cant improve- and fi ve required blood transfusions (12.5%). All major com- Anastomotic Leakage (%) 1.63 1.50 0.252Only 11 patients (14%) did not meet the ≥50% criterion ment of postoperative outcomes was observed after the fi rst plications including recto-vaginal fi stulas and anastomotic Abdominal Abscess (%) 3.19 3.79 < 0.01 Mean Length of Hospital 7.37 7.38 0.93for success with a median reduction of 30% in FI episodes. 300 cases. Women affected by deep intestinal endometrio- failure occurred in the fi rst half of the group. stay (days) Improvement in FI episodes during test stimulation was sis should be referred to experienced laparoscopic surgeons. CONCLUSIONS: In the setting of deep infi ltrating pelvic In-hospital Mortality (%) 1.37 1.49 0.29greater in subjects with therapeutic success (88%) than in endometriosis, colorectal resection with a combined lapa- Mean Total Hospital 44,183 48,700 < 0.01those without success (75%) (p = 0.02). No other differ- Charges (\$) roscopic/transvaginal approach, avoiding a mini-laparot- ences were noted between the groups. omy or the extension of port-site incisions, represent a CONCLUSION: The present analysis indicates that the viable option. All major complications occurred among the therapy was effective but that the only factor which cor- fi rst half of our patients. A prospective randomized trial is related with success at 3 years was improvement during test necessary in order to properly evaluate the potential bene- Using multivariate regression analysis, Native American stimulation. fi ts of transvaginal versus laparotomic specimen extraction race [odds ratio (OR): 2.02], renal failure (OR: 1.97), con- evaluating both perioperative and long-term outcomes.

108 109 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

CONCLUSIONS: Discoid resection is a safe procedure for included intractable symptoms, gastrointestinal bleeding, of hand-assisted laparoscopy, the volume of laparoscopic- Mo1596 the treatment of single endometriosic nodule <2.5 cm and malignancy. Complete resection of malignant CDF assisted colorectal surgeries was 10%. For the year follow- Discoid Resection with Combined Laparoscopic/ in size in the rectum. Rate of overall morbidity and spe- with negative margins was achieved in half of patients after ing implementation of hand-assisted laparoscopy, the rate Endoscopic Approach for Rectal Endometriosis: cifi c bowel complications was acceptable, with a decrease en bloc resection. Palliative bypass was performed in the of minimally invasive cases increased to 39%, followed Perioperative and Long-Term Outcomes over time refl ecting a learning curve process. Long-term remaining patients with unresectable disease. Thirteen by 89% and 90% for the subsequent years (p < 0.0005, outcomes are satisfactory with signifi cant improvements patients with benign CDF had resection of the fi stula which see Table 1). Rates of complications in minimally invasive 1 *1,2 3 Giacomo Ruffo, Stefano Crippa , Claudia Bonardi, in symptoms and a very low rate of bowel endometriosis required a duodenal bypass procedure in 2. There were no cases, including rates of wound infection and anastomotic Marco Scioscia 3, Anna Pesci4, Stefano Partelli1,2, Marco Benini5, recurrence. perioperative deaths, and the morbidity rate was 38%. leak, did not change with the increased use of minimally Massimo Falconi 1,2, Luca Minelli3 Median survival for patients with malignant CDF was 20 invasive techniques (p = 0.076 and 0.618, respectively). 1 Department of Surgery, Ospedale Sacro Cuore Don Calabria, Negrar,Mo1597 months (range 1–150 months). Two patients with malig- Conversion rates decreased dramatically, from 42% to 12% Italy; 2Department of Surgery, Università di Verona, Verona, Italy; nant CDF had >5 yr survival. All patients with benign CDF (p < 0.0005). Similar patterns were noted when individual 3Department of Gynaecology, Ospedale Sacro Cuore Don Calabria, Management and Outcomes of Primary Coloduodenal who underwent fi stula resection had resolution of fi stula- surgeons were analyzed independently. In terms of opera- Fistulas Negrar, Italy;4 Department of Pathology, Ospedale Sacro Cuore Don related symptoms with one recurrence. tive effi ciency, prior to the use of hand-assisted laparoscopy, Calabria, Negrar, Italy;5 Division of GI endoscopy, Ospedale Sacro Ashwin S. Kamath*1, Corey W. Iqbal1, Tuan H. Pham1, CONCLUSION: Benign CDF is amenable to operative ther- the operative time for laparoscopic-assisted resections was Cuore Don Calabria, Negrar, Italy Bruce G. Wolff 2, Heidi Chua2, John H. Donohue1, apy. Complete resection of malignant CDF can impart sur- signifi cantly higher than for open resections (p = 0.001, see 2 2 Table 2). After implementation of hand-assisted laparos- BACKGROUND: Robert R. Cima , Richard Devine vival benefi t and should be offered to patients. Bowel endometriosis involves the rec-1 copy, the operative times for minimally invasive resections General Surgery, Mayo Clinic, Rochester MN, Rochester, MN; tum in most cases. In this setting, laparoscopic rectal and for open resections equalized (p = 0.519). resections with low colo-rectal anastomosis are frequently 2Colon and Rectal Surgery, Mayo Clinic, Rochester, MN Mo1598 required and are associated with good symptomatic relief PURPOSE: Primary coloduodenal fi stula (CDF) is a rare Use of Hand-Assisted Laparoscopy in a Colorectal Table 1: Learning Curve and Safety. Values Reported are but with the risk for major complications. Discoid resec- entity. We reviewed our experience with the management Cancer Practice Leads to a Rapid and Safe Adoption Percentages. tions have been proposed in the treatment of small (<2.5 and outcomes of CDF. cm) single endometriosic nodule of the rectum without of Minimally Invasive Techniques METHODS: Prior to HAL bowel stenosis as an alternative to segmental resection. This is a retrospective study from 1975–2005 Sonia T. Orcutt*, Christy L. Marshall, Courtney J. Balentine, Implementation 2008 2009 2010 p-Value METHODS: of patients with primary CDF. Patients were followed Celia Robinson, Daniel A. Anaya, Avo Artinyan, Samir S. Awad,MIS surgeries 10.2% 38.5% 89.0% 89.7% <0.0005 Between 2004 and 2009, 81 patients (median through clinic visits and mail correspondence with a mean (% total cases) age 35 years) underwent laparoscopic excisions of endo- David H. Berger, Daniel Albo follow-up of 56 ± 14 mos. Complications metriosis with full thickness discoid resections of bowel Surgery, Baylor College of Medicine, Houston, TX Wound 19.4% 19.4% 18.5% 7.7% 0.076 endometriosis using a circular stapler inserted transanally. INTRODUCTION: infection rate 32 procedures were performed between 2004 and 2006 Hand-assisted laparoscopy has beenAnastomotic 0% 3.3% 4.6% 0% 0.618 (early period) and 49 between 2007 and 2009 (late period). shown to maintain the short-term advantages of standardleak rate Perioperative and long-term outcomes were analyzed. laparoscopy while providing surgeons with stereotacticConversion rate 41.9% 3.3% 7.8% 11.5% <0.0005 feedback. In addition, it has been proposed to reduce theHAL = hand-assisted laparoscopy; MIS = minimally invasive. RESULTS: Dysmenorrhea, chronic pelvic pain, dysche- length of the training curve for minimally invasive colorec- sia, dyspareunia, rectal bleeding were present preop- tal surgery. We hypothesize that the widespread adoption eratively in 100%, 75%, 78%, 80%, and 2.5% of cases, of hand-assisted laparoscopy in colorectal cancer resections respectively. No laparoconversion was required and over- would allow for a rapid increase in the use of minimally Table 2: Learning Curve and Operative Effi ciency. Values Reported all median operative time was 180 minutes with median invasive techniques without compromising patient safety Are Median (Range). blood loss of 100 ml. Median time of discoid resection or operative effi ciency. was 15 minutes. Temporary ileostomy was required in one Prior to HAL patient (1.5%). There was no mortality, overall morbidity METHODS: A retrospective analysis of a prospectively Implementation 2008 2009 2010 was 16%. Median length of stay was 5 days. Rates of intra- maintained colorectal cancer database containing cases Open cases 198 200 337 259 Operative (38–545) (103–711) (241–662) (205–705) abdominal haemorrhages, rectal bleeding, enteric fi stula, prior and subsequent to the adoption of hand-assistedtime (minutes) MIS cases 238 201 227 251 blood transfusions and reoperation were 2.5%, 6%, 1.5%, laparoscopy was completed. There was no signifi cant varia- (138–450) (113–457) (96–639) (107–446) 6%, and 6%, respectively. Rectal bleeding from the stapler tion in the composition of surgeons in the practice duringp-value 0.001 0.996 0.008 0.519Poster Abstracts line were successfully managed using conservative endo- the study period. The change in distribution of minimallyMIS = minimally invasive. scopic management. At pathology, three patients (3.5%) invasive cases over time quantifi ed the rapidity of imple- Monday showed a microscopic involvement of the surgical margin mentation of minimally invasive techniques. Surrogate of the discoid resection by endometriosic tissue. markers of the learning curve, including complication CONCLUSION: Implementation of hand-assisted laparos- copy into a colorectal cancer practice leads to an abbre- There was a decrease of median operative time (220 vs 150 rates, conversion rates, and operative times, were similarly min, p < 0.0001), blood loss (200 vs 130 ml, p = 0.105), evaluated. The Fisher’s exact and Mann-Whitney U tests viation of the laparoscopic learning curve with a rapid length of stay (6 vs 5 days, p = 0.07), overall morbidity were used, where appropriate, for analysis. adoption of minimally invasive techniques, while main- taining patient safety and operative effi ciency. (28% vs 8%, p = 0.017), haemorrhages (6% vs 0, p = 0.07), RESULTS: 115 hand-assisted laparoscopic cases were per- rectal bleeding (12.5% vs 2%, p = 0.06), blood transfu- formed during the study period, of which 44% were rectal sions (12.5% vs 2%, p = 0.06) from early to late period resections, 44% were right-sided, 11% were left-sided, and refl ecting an improved learning curve. 2% were subtotal or total resections. Prior to the initiation Long-term follow-up data were available for 67 patients (83%) with a median follow-up was 41 months. Postop- erative rates of dysmenorrhea, chronic pelvic pain, dys- RESULTS: Twenty-two (22) patients at a mean age of 54 ± chesia and dyspareunia decreased to 30%, 24%, 22%, and 3 yrs were diagnosed with primary CDF: benign (n = 14) or 13.5%, respectively (p < 0.0001). In nine patients (13.5%) malignant (n = 8). Benign CDF were due to Crohn’s disease there was a diagnosis of endometriosis-recurrence, but (n = 9) or peptic ulcer disease (n = 5), while malignant CDF only four (6%) underwent reoperation. The rate of bowel was primarily due to colon cancer (n = 7) plus one patient endometriosis recurrence was 1.5%. with lymphoma. Indications for operative intervention

110 111 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1599 RESULTS: Between 2003 and 2008, 210 newly diagnosed RESULTS: 685 patients underwent surgical treatment of adenectomy are done as the open method. Recently, VATS- young-onset CRC patients undergoing surgery had MSI test- rectal prolapse. Most patients underwent open SR+R (247 E in prone position is remarkable in Japan because the lung Impact of Preoperative Microsatellite Instability ing. Results of testing were available preoperatively in 103 open SR+R, 193 open R, 161 laparoscopic SR+R, 84 laparo- moves below by the gravity, and a good operative fi eld is Testing on Surgical Management in Young-Onset patients and postoperatively, secondary to the refl ex pro- scopic R). Patients undergoing all procedures had similar obtained. We introduce our methods and devices perfomed Colorectal Cancer Patients: Results from a Refl ex tocol, in 107 patients. MSI-H tumors were found in 16/103 comorbidity profi les, but the patients who had a laparo- in prone position at our hospital. Testing Protocol (16%) in the preoperatively tested group and 12/107 (11%) scopic R were signifi cantly older patients (mean age 61.4 PATIENTS: Twenty patients in left lateral position and in the postoperatively tested group. Comparison of clini- years) than the other three groups (p = 0.04). Operating Jennifer Holder-Murray*1, Rajesh Pendlimari1, Lisa Boardman4, 15 patients in prone position, with esophageal squamous copathologic features and surgical intervention between time ranged from 128 minutes (open R) to 185 minutes Thomas C. Smyrk 2 1 3 cell carcinomas underwent VATS-E since 2005 and 2009, , David W. Larson, Noralane M. Lindor, groups are listed in Table 1. MSI-H status known preopera- (laparoscopic SR +R; p < 0.001). Overall, open SR+R and 1 respectively. Eric J. Dozois tively signifi cantly infl uenced surgical recommendations of open R were associated with signifi cantly more morbidity 1 2 METHODS: Colon & Rectal Surgery, Mayo Clinic, Rochester, MN;Anatomic total colectomy compared to patients where status was not than laparoscopic SR+R and R (OR 0.44, 95% CI 0.23-0.84, At fi rst the patients are fi xed at semi-prone Pathology, Mayo Clinic, Rochester, MN; 3Medical Genetics, Mayo known until after surgery (94% vs. 8%, p <0.0001). Hys- p = 0.01). Comparing all four procedure types, there was a position because both prone and left lateral positions can rd th Clinic, Rochester, MN; 4Gastroenterology & Hepatology, Mayo Clinic, terectomy was performed in 8/10 women when preopera- trend to decreased overall morbidity with laparoscopic R, be set by rotating. Four 10 mm ports are used at the 3 , 5 , th th Rochester, MN tive MSI-H status was known (one had metastatic disease, but it did not achieve statistical signifi cance (OR 0.31, 95% 7 and 9 , and mini-thoracotomy incision of 3.5 cm is one was of childbearing age). There was only one female CI 0.07-1.40, p = 0.13). Length of stay and SSI rates were made at the 5th intercostal space (ICS). The pneumothorax PURPOSE: In an effort to capture patients in our surgical patient in the postoperative group and she did not undergo signifi cantly lower with laparoscopic R as compared to the by maintaining CO2 insuffl ation pressure of 6 mmHg is practice considered high-risk for Lynch Syndrome because hysterectomy. Germ-line mutations were positive in 10/16 other three procedure types (see table). made, and esophagectomy is performed in prone position. of young age, refl ex microsatellite instability (MSI) testing (63%) MSI-H patients tested. Two patients had germ-line The lymph nodes around the trachea and bronchus, above was initiated on all colorectal cancers resected from young testing available preoperatively. Surgical Site Overallthe diaphragm and along the bilateral recurrent laryngeal patients not tested preoperatively. This protocol provides CONCLUSION: Length of Infections, Complicationnerves are dissected. The sponge spacer is used to mobi- a unique opportunity to retrospectively compare surgical MSI-H status was found in 13% of young- Procedure Type N Stay, Days n (%) Rate, lizen (%) the esophagus, rotate the trachea and aspirate blood. management in high-risk patients who were tested either onset colorectal cancer patients operated at our institution, Open rectopexy (R) 193 5.3 16 (8.3) 14 (7.3)Working space at the lower mediastinal area is obtained preoperatively or postoperatively for MSI. We aimed to and 63% of those tested, had germ-line mutations. Knowl- Open sigmoid resection 247 7.9 35 (14.2) 34 (13.7) and rectopexy (SR+R) by retracting the esophagus using the retractor or sponge determine if MSI status altered surgical management when edge of MSI status preoperatively signifi cantly infl uenced spacer , and at the left upper mediastinal area is obtained is surgical management with an increase in total colectomy Laparoscopic rectopexy 84 3.2 0 (0.0) 2 (2.4) the result was known preoperatively. (R) obtained by stripping the residual esophagus and retracting and hysterectomy compared to patients whose MSI-H sta- METHODS: Starting in 2003, the pathologist handling the Laparoscopic sigmoid 161 5.4 11 (6.8) 11 (6.8)the trachea using the retractor. tus was discovered postoperatively. The absence of germ- resection specimens ordered MSI testing on all newly diag- resection and rectopexy RESULTS: line testing in MSI-H patients did not appear to infl uence (SR+R) 1.) Mean estimated blood loss was 174 ml of nosed young-onset CRC (≤50 years old) cases not tested surgical decision making. *p < 0.001 *p < 0.001 *p < 0.001chest procedure in prone position. 2.) The rate of recur- prior to surgical resection. Patients with infl ammatory rent laryngeal nerve paralysis was 13.3% (2/15), and anas- bowel disease and polyposis syndromes were excluded. We Mo1600 tomotic leak was 8.9% (1/15). 3.) There was no incidence of categorized tumors as microsatellite stable (MSS), which CONCLUSIONS: Patients who undergo laparoscopic rec- conversion to open method, and of postoperative pneumo- included MSI-low tumors, and MSI-high (MSI-H), with or Laparoscopic Rectopexy: The Procedure of Choice for topexy have a shorter length of stay and lower SSI rate nia. 4.) Lymphadenectomy along the left recurrent laryn- without germ-line mutation. Clinicopathologic features Rectal Prolapse to Reduce Surgical Site Infections and as compared to patients who undergo other abdominal geal nerve was nearly done in prone position in the late and surgical procedures performed were reviewed. Length of Stay procedures for repair of rectal prolapse. Further study is term. 5.) The sponge spacer is available for VATS-E. necessary to determine the long-term outcomes from CONCLUSION: 1.) VATS-E in prone position is a simple Table 1: Clinicopathologic Features and Surgical InterventionTrent of Magruder*, Elizabeth C. Wick, Susan Gearhart, laparoscopic rectopexy, but in high-risk patients, the lapa- MSI-H Patients in Preoperatively Tested and Postoperatively JonathanTested E. Efron and feasible method. 2.) Safe VATS-E is facilitated by put- roscopic approach should be considered to decrease peri- ting the patient in the semi-prone position, using the mini- Groups Department of Surgery, Johns Hopkins University School of Medicine, operative risk. Baltimore, MD thoracotomy procedure and using a sponge-spacer. MSI-H Preop MSI-H Postop PURPOSE: Characteristics N = 16 N = 12 p-Value Rectal prolapse is commonly seen in patients Clinical: Esophageal Age, years (mean) 39.3 ± 8.2 41 ± 10 with 0.42 signifi cant comorbities. Multiple approaches have been described, including the use of laparoscopy. The BMI, kg/m2 28.7 ± 9.9 26.2 ± 5.8 0.85 Poster Abstracts Female 10 (62.5%) 1 (8.3%) 0.005purpose of this study was to determine if laparoscopic Mo1601

Location, Right colon 9 (56.3%)† approaches for repair of rectal prolapse are associated with Video Assisted Thoracoscopic Surgery of Esophagus (VATS-E) Monday 6 (50%) 0.52 less short-term morbidity than open approaches. Transverse colon 1 (6.3%) 0 (0%) 0.57 in Left Lateral and Prone Position: Our Methods and Devices at Left colon 1 (6.3%) 1 (8.3%) 0.68 METHODS: The ACS NSQIP database was queried for Our Hospital Sigmoid colon 0 (0%) 1 (8.3%) 0.43 patients who underwent laparoscopic or open rectopexy Hiroshi Makino*1,2, Tsutomu Nomura1, Masao Miyashita1, Rectum 6 (37.5%) 1 1 1 † 4 (33.3%) 0.57 (R) or sigmoid resection and rectopexy (SR+R) between Nobutoshi Hagiwara , Satoshi Matsumoto, Yoshiharu Nakamura, Stage I 5 (31.3%) 3 (25%) 0.53 2005 and 2008. Comorbidities analyzed included diabe- Nobuyuki Sakurazawa 1, Eiji Uchida1 tes, COPD, hypertension, cardiac (history of congestive 1 Stage II 6 (37.5%) 4 (33.3%) 0.57 Gastroenterological Surgery, Nippon Medical School, Tokyo, Japan; Stage III 3 (18.8%) 2 (16.7%) 0.64 heart failure, myocardial infarction, previous percutaneous 2Surgery, Nippon Medical School, Chiba-Hokusoh Hospital, Inzai-City, Stage IV 2 (12.5%) 3 (25%) 0.36 cardiac intervention or surgery), and neurologic (history Chiba-Prefecture, Japan Family history: 1° relative CRC 7 (43.8%) 2 (16.7%) 0.11 of transient ischemic attack or cerebrovascular accident). Any family history CRC 13 (81.3%) 5 (41.7%) 0.04 Postoperative complications analyzed included surgical INTRODUCTION: Video assisted thoracoscopic surgery Total colectomy recommended 15 (94%) 1 (8.3%) < 0.0001 site infections (SSI), pneumonia, reintubation, pulmonary of esophagus (VATS-E) is generally performed in left lat- Total colectomy performed 11 (68.8%) 1 (8.3%) 0.002 embolus, stroke, myocardial infarction, and sepsis. Chi- eral position in Japan because esophagectomy and lymph Hysterectomy (females only) 8 (80%) 0 (0%) 0.27 Genetic testing performed 12 (75%) 4 (33.3%) 0.09 square or t-test/ANOVAs were used to assess signifi cance for Positive germline mutation 9 (56.3%) 1 (8.3%) 0.01 categorical and continuous variables, respectively. Logistic regression analysis was used to determine risk factors for morbidity after rectal prolapse repair.

112 113 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1602 Comparison of the Pre-Stretta and Post-Stretta Means of the CRT to PLO for curative intent is lacking. Although laryn- Outcome Variables geal preservation in CRT is more preferred, whether CRT Radiofrequency Energy Delivery to the Lower can produce comparable disease control and survival over Esophageal Sphincter Reduces Esophageal Acid Mean PLO is yet to be elucidated. Exposure and Improves GERD Symptoms: A Studies Patients Follow-up Pre- Post- METHOD: This study aims to compare the recurrence rate Meta-Analysis Outcome Variable (n) (n) (mo) Stretta Stretta P-Value GERD-HRQL 9 433 19.8 26.11 9.25 0.0001 and overall survival after CRT and PLO for cervical esopha- Ambar Banerjee*1, W. S. Melvin1, Steven J. Naber2, Kyle A. Perry1 score geal cancer. A retrospective comparison was conducted on 1 Center of Minimally Invasive Surgery, The Ohio State University, QOLRAD score 4 250 25.2 3.30 4.97 0.0010 consecutive patients receiving either defi nitive CRT or PLO Heartburn score 9 525 24.1 3.55 1.19 0.0001 Columbus, OH;2Statistical Consulting Service, The Ohio State in Prince of Wales Hospital. Satisfaction score 5 366 21.9 1.43 4.07 0.0006 RESULTS: University, Columbus, OH Esophageal 11 364 11.9 10.29 6.51 0.0003 Between January 1998 and June 2010 (12.5 INTRODUCTION: Acid Exposure years inclusive), 57 patients (47 males and 10 females) with Several endolumenal approaches to (% Ph <4) the management of gastroesophageal refl ux disease (GERD) mean age of 61.4 ± 8.0 years received curative treatment for Demeester score 7 267 13.1 44.37 28.53 0.0074 cervical esophageal cancer. Forty patients (70.1%) received have been developed as an intermediate therapy between LES pressure 7 263 8.7 16.54 20.24 0.0302 conservative medical management and the more invasive defi nitive CRT while 17 patients (30.9%) underwent PLO surgical fundoplication. One such intervention, the Stretta with either gastric (94.1%) or colonic transposition (5.9%). Clinical AJCC stage I, II, and III diseases were confi rmed in Procedure (Mederi Therapeutics Inc, Greenwich, CT), uti- CONCLUSION: Radiofrequency ablation of the lower 0 (0%), 9 (22.5%) and 31 (77.5%) patients in CRT group lizes application of radiofrequency energy to the lower esophageal sphincter produces signifi cant improvement RESULTS: esophageal sphincter to control gastroesophageal refl ux. A RBP was present in 97% of normal subjects while those in PLO group were 0 (0%), 4 (23.5%) and 13 in both subjective and objective refl ux indices. These fi nd- and 100% of patients. The mean (SD) amplitude of the RBP (76.5%) patients respectively. Both groups were compa- Multiple studies have debated its short and long-term effec- ings suggest that the Stretta procedure provides signifi cant tiveness. The objective of this study is to review the pub- in normal subjects was 6.8 (3.7) mmHg. The RBP in refl ux rable for age, gender, ASA class and clinical stage. In CRT symptomatic relief for many patients and represents an patients was signifi cantly lower than that in normal subjects group, all patients completed two courses of chemotherapy lished data and perform an in-depth statistical analysis of alternative to medical treatment and surgical fundoplica- the effi cacy of this procedure in the management of GERD. [3.6 (7.0) mmHg, p < 0.003]. After Nissen fundoplication, (infusional 5-fl uorouracil plus cisplatin) and radiotherapy tion in select patients. the RBP increased to 12.0 (3.2) mmHg and was signifi cantly (dosage 50–60 Gy). Complete tumor response was observed METHODS: A systematic search of the literature, pub- greater than the preoperative values (p < 0.0001). The 95th in 21 patients (52.5%). Of the remaining 19 treatment fail- lished to date in English and indexed in MEDLINE and Mo1603 percentile value for RBP in normal subjects was 10.4 mmHg ure patients, 5 (12.5%) underwent salvage PLO, 1 (2.5%) PubMed, was carried out in November 2010 with device refused salvage surgery, and 13 (32.5%) had distant metas- Ramp Bolus Pressure: A Determinant for Tailoring the and after Nissen fundoplication was 20.0 mmHg. and condition specifi c keywords. Studies were selected on tasis upon re-staging. In PLO group, the operative morbid- CONCLUSION: The ramp bolus pressure is a frequent the basis of availability of data on at least two of the fol- Degree of the Fundoplication ity and mortality rates were 70.5% and 11.7% respectively. manometric fi nding that indicates the degree of outfl ow lowing parameters: esophageal manometry, pH study, qual- Shahin Ayazi*, Abraham Nisim, Joerg Zehetner, Kimbely S. Grant, Three patients (17.6%) had positive margin involve- resistance in the LES region. This amplitude increases after ity of life indices and medication usage. Exclusion criteria Gaurav Sharma, Arzu Oezcelik, Daniel S. OH, John C. Lipham, ment requiring postoperative adjuvant radiotherapy. In a a Nissen Fundoplication. These fi ndings suggest that a con- included studies involving children or those with a follow- Steven R. Demeester, Jeffrey A. Hagen, Tom R. Demeester median follow-up of 14.4 (range 1.2–109.9) months, there traction amplitude above 20mmHg is necessary to over- up less than 3 months. Data were presented as weighted Surgery, University of Southern California, Los Angeles, CA was no difference in local recurrence rate (42.5% vs. 52.9%, come the resistance of a Nissen fundoplication and this mean values based on the number of patients in each P = 0.469) and distant recurrence rate (32.5% vs. 29.4%, P = INTRODUCTION: threshold may be a better determinant of the need to tailor included study so that the analysis represented the mean Traditionally, a contraction amplitude 0.819) between CRT and PLO groups. The overall survival the type of fundoplication performed. change in each outcome variable over all patients studied. above 30 mmHg in the distal esophagus was used as a in CRT group was 17.1 (95% C.I. 11.2–22.1) months while The meta-analysis consisted of weighted t-tests on the dif- threshold for performing a 360° fundoplication. Recently, that of PLO group was 14.4 (95% C.I. 7.8–22.9) months. ference between mean pre-Stretta and post-Stretta values the relevance of esophageal body contraction amplitudes Mo1604 Their difference was not statistically signifi cant (P = 0.943, in selecting patients for a Nissen fundoplication has been of various outcome variables. Analysis was restricted to the Chemoradiotherapy or Pharyngo-Laryngo- log rank test). reported mean values of outcomes for pre-operative and questioned. We hypothesize that the ramp bolus pressure Esophagectomy for Cervical Esophageal Squamous CONCLUSION: latest post-operative observations for single Stretta proce- (RBP) may be a better determinant of the need to tailor the Despite a relatively high local treatment dures and for results summarized across all patients. type of fundoplication. The aim of this study was to deter- Cancer failure risk, the tumor recurrence rate and overall survival mine normal values for the RBP in asymptomatic subjects of CRT were comparable to PLO for patients with cervical RESULTS: 20 studies, containing 1441 patients, published Shirley Y. Liu*, Philip W. Chiu, Anthony Y. Teoh, Man Yee Yung, Poster Abstracts and in patients before and after a Nissen fundoplication. esophageal cancer. between 2001 and 2010 met the inclusion criteria. The Candice C. LAM, Simon K. Wong, Enders K. NG mean follow-up was 17.1 months (range 4–53 months). METHODS: The ramp bolus pressure (RBP) was deter- Department of Surgery, The Chinese University of Hong Kong, Hong Monday Outcome data are outlined in the table. GERD health mined by measuring the mean pressure preceding the Kong, China upstroke of the contraction wave 5 cm above the lower related quality of life (HQRL), quality of life refl ux and INTRODUCTION: Cervical esophageal squamous cell esophageal sphincter (LES) (Figure). We measured the RBP dyspepsia (QOLRAD), heartburn, and patient satisfaction carcinoma (SCC) is primarily treated by either concurrent in 53 asymptomatic volunteers and 37 patients with refl ux scores were all signifi cantly improved following the Stretta chemoradiotherapy (CRT) or pharyngo-laryngo-esopha- symptoms before and after a Nissen fundoplication. All of procedure. Esophageal acid exposure and DeMeester score gectomy (PLO). Literature evidence directly comparing were both signifi cantly reduced following Stretta treat- the refl ux patients had an excellent outcome and none had ment. Lower esophageal sphincter pressure showed a trend dysphagia at the time of postoperative evaluation. toward increased post-treatment pressures, but failed to reach statistical signifi cance in this analysis.

114 115 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1605 Mo1606 Clinical Outcomes After Esophagectomy Stratifi ed by BMI > 30

Trans-Thoracic Approaches to Esophagectomy Obesity Does Not Affect Outcomes in Gastro- Normal BMI N = 117 Obese (BMI > 30) N = 40 P value 95% Confi dence Interval Signifi cant? Associated with Higher Morbidity Esophageal Cancer Factor Leak 24 (20.5%) 10 (25%) .745 .677-.813 No Ross F. Goldberg*, Steven P. Bowers, Michael Parker, John Stauffer,Russell Farmer*, Kelly M. Mcmasters, Charles R. Scoggins, Complication 73 (62.4%) 29 (72.5%) .268 .198-.377 No Michael G. Heckman, Colleen S. Thomas, Horacio J. Asbun, Robert C. Martin Clavien Complication Grade 3+ 49 (41.9%) 20 (50%) .471 .393-.549 No John A. Odell, C. Daniel Smith Department of Surgery, University of Louisville, Louisville, KY EBL: mean, (SEM) 486.9 (43.1) 485 (58.2) .985 (-)165 - (+) 168 No Department of Surgery, Mayo Clinic, Jacksonville, FL BACKGROUND: Transfusion given 45 (38.5%) 16 (40%) 1.00 .961-1.00 No Obesity is a growing epidemic in the Units transfused: mean, median (SEM) 2.47, 2, (.18) 2.75, 2, (.536) .522 -1.16 - .599 No INTRODUCTION: Esophagectomy is performed through United States. The CDC estimates that 25.6% of all adults Margin Positivity 14 (12.0%) 1 (2.5%) .079 .075 - .180 No various approaches, both open and minimally invasive in the US have a Body Mass Index (BMI) >30. Obesity and Nodes removed: mean, (SEM) 18.9 (1.26) 20.9 (2.10) .420 -6.89 - 2.89 No (MIE). This study’s aim is to review the perioperative out- its associated comorbidities have been shown to increase Positive nodes: mean, (SEM) 4.15 (.76) 5.35 (1.51) .449 -4.30 - 1.91 No comes after all types of esophagectomy while quantifying risk for perioperative complications. The purpose of this patients’ preoperative comorbidities. study was to compare the rate of complication in normal METHODS AND PROCEDURES: We retrospectively vs. obese patients undergoing esophageal anastomosis fol- reviewed the charts of 97 patients who underwent either lowing oncologic resection. MIE (N = 56) or open esophagectomy (N = 41) between METHODS: We reviewed our prospective 350 patient January 2007 and August 2010. Indications for esophagec- esophageal-gastric database and found 166 esophageal tomy included adenocarcinoma (N = 73), squamous cell anastomoses performed for esophageal and gastric cancer carcinoma (N = 11), Barrett’s with high-grade dysplasia (N = from 1994 to 2009, 157 with BMI data. With informed con- 8), end-stage achalasia (N = 3), mesh erosion (N = 1) and sent and IRB approval, we compared clinical, pathologic, leiomyoma (N = 1). In both groups males predominated and outcomes data in the perioperative period between (MIE:86%, open:83%) with a similar median age of 66 patients with BMI ≥30 and BMI <30. We then evaluated and 64 years, respectively. Neoadjuvant chemo/radiother- the presence of complication, severity of complication as apy was administered to 20 (36%) and 19 (46%) patients indicated by the 5 point Clavien grading scale, and overall respectively. All MIE were performed either in a 3-fi eld, survival. Statistical correlations were calculated with Chi- prone VATS technique (N = 48), or a 2-fi eld, laparoscopic Square (Monte-Carlo), independent T-test, Kaplan-Meier, extra-pleural technique (N = 8). Open esophagectomy and log-rank analyses as appropriate. included trans-thoracic techniques: Ivor-Lewis (N = 10), thoracoabdominal approach with cervical incision (N = RESULTS: Of the 157 patients reviewed with a median 11), and 3-fi eld esophagectomy (N = 8) and extra-pleural age of 62 (range 17–84);120 (76.4%) were men, 37 women, transhiatal technique (N = 12). Using the Charlson Comor- 40 were Obese (25.5%) and 117 were not Obese (74.5%). bidity Index, the MIE group had 3 patients categorized as Of these 157 patients, 62 (39.4%) had neoadjuvant che- Overall Survival Following Esophageal Anastomosis: BMI <30 vs. BMI >30 low-risk, 41 as moderate-risk and 12 as high-risk while the motherapy with 59 (37.5%) adjuvant treatment. No statis- open group had 33 moderate-risk and 8 high-risk patients. tically signifi cant difference was seen between the obese Postoperative complications were stratifi ed using the Cla- and non-obese patients regarding development of peri- Mo1607 trocars decreased in successive quartiles (p < 0.05). Esoph- vien Classifi cation Scale, with minor complications classi- operative complication (72.5% vs. 62.4%, p = .268, 95% agotomy/gastrotomy occurred in 3 patients. Postoperative fi ed as Class 1–2 and major complications as Class 3–5. CI .198–.377), grade of complication (p = .471, 95% CI .393 Defi ning the Learning Curve of Laparo-Endoscopic complications occurred in 9%, none specifi c to Heller myot- RESULTS: Average operative time for the MIE trans-tho- –.549), or anastomotic leak (25% vs. 20.5%, p = .745, 95% Single Site (LESS) Heller Myotomy omy (Table). No conversions to open operations occurred. racic was 333 minutes; for the MIE extra-pleural was 291 CI .687-.813). Intraoperative variables also showed no sig- Sharona B. Ross*, Tony J. Kurian, Kenneth Luberice, Length of stay was 1 day. Postoperative frequency/severity minutes, for the open trans-thoracic was 215 minutes and nifi cant difference including: estimated blood loss (EBL) Natalie Donn, Harold Paul, Melissa Rosas, Chinyere P. Okpaleke,scores were: dysphagia = 2/0, regurgitation = 0/0, heartburn open extra-pleural was 192 minutes. Postoperative compli- (p = .985), need for transfusion (P = .863), number of units Alexander S. Rosemurgy = 0/0; scores were similar and less than before myotomy transfused (p = .522), number of nodes collected on lymph- cations occurred in 42 MIE patients and 21 open patients; Surgery, University of South Florida, Tampa, FL across all quartiles, p < 0.001(Figure). There were no appar- Poster Abstracts these were major in 19 (34%) MIE and 17 (41%) open (p = adenectomy (p = .420). There was a decreased incidence ent scars, except where an additional trocar was placed. INTRODUCTION: 0.53). The predominant in-hospital complications were of positive margin after resection in the Obese (2.5%) Initial outcomes suggest Laparo- Monday cardiac, mainly arrhythmias (32% vs. 41%, P = 0.40), pul- than non-Obese (12.0%) (p = .079, 95% CI = .075–.180). Endoscopic Single Site (LESS) Heller myotomy with anterior 25-Patient Quartiles monary (34% vs. 44%, P = 0.40), and leaks (11% vs. 10%, In the superobese (BMI >35), there was a signifi cant trend fundoplication provides safe, effi cacious, and cosmetically P = 1.00). Postoperative 30-day mortalities were 4% in the toward greater intraoperative transfusion (4.5 vs. 2.4 units, superior outcomes relative to conventional laparoscopy. 1–25 26–50 51–75 76–100 MIE group and 7% in the open group (P = 0.65). Median p = .006) and increased mean number of positive lymph This study was undertaken to defi ne the learning curve ofLength of Operation 115 102 125 141 length of stay was 8 days (range 5–51 days) for the MIE, nodes on (8.9 vs. 3.9, p = .024). There was no difference in LESS Heller myotomy with anterior fundoplication. (minutes) (113 ± 33.2) (116 ± 41.9) (126 ± 42.7) (145 ± 39.8) overall survival in the Obese with median of 16 months Additional Trocar 6 3 2 0 9 days (range 6–38 days) for the open technique. When METHODS: 100 patients underwent LESS Heller myotomyPlacement (range 0–180 months) and non-Obese 16.5 months (range looking at trans-thoracic versus extra-pleural techniques, with anterior fundoplication. Symptom frequency andLength of Stay 1 1 1 1 there were signifi cant differences in major complications = 0–180 months), (p = .968). (days) (1.4 ± 0.76) (1.7 ± 0.25) (1.7 ± 0.25) (2.7 ± 0.83) severity were scored utilizing a Likert scale (0 = never/not (43% vs. 11%, P = 0.013), pulmonary complications (43% Conversion to 0 0 0 0 CONCLUSIONS: Obesity does not affect rate of complica- bothersome to 10 = always/very bothersome). Symptom vs. 17%, P = 0.032), hoarseness (4% vs. 22%, P = 0.027) and “Open” tion, grade of complication, or rate of anastomotic leak, resolution, additional trocar placements, and complica- wound infections (4% vs. 22%, P = 0.027). Postoperative Atrial Edema, Left Atrial fi brillation, Atrial or overall survival in patients undergoing esophageal anas- tions were compared among patient quartiles. Median dataComplications fi brillation, bundle- branch Capnothorax, fi brillation CONCLUSIONS: This series shows no differences in major tomosis for oncologic resection. Further investigation is Myocardial block Edema, are presented. Infarct Myocardial complications and mortality between MIE and open esoph- needed to determine the effect of obesity on margin posi- RESULTS: Preoperative frequency/severity scores were: dys- Infarct agectomies in similar patient groups. If the pleural cavity tive resection as a function of procedure-limiting body phagia = 10/8, regurgitation = 8/6, heartburn = 2/2. Addi- is violated, regardless of approach, there are higher risks habitus. of pulmonary complications. Avoiding the pleural cavity tional trocars were placed in 11 patients (11%), of whom however may increase the risk of hoarseness and wound 81% were in the fi rst two quartiles; placement of additional infections.

116 117 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

abnormal gastric emptying study (n = 22) or retained food [LTA]: Trans-abdominal [TA]). Operative approach was tai- diet, 7% a liquid diet, and 7% NPO. Inpatient enteral nutri- in the stomach on endoscopy after an overnight fast (n = 9) lored to pt and tumor (EGD+CT) characteristics with a view tion (EN) was initiated in 68.9%, but continued on dis- (DGES: Abnormal study), and 49 had no gastric emptying to maximize complete resection whilst minimizing mor- charge in only 54% secondary to failure to thrive (57.7%), study performed (DGES: No study). The comparison group bidity. Data presented as median (range). Mann Whitney dysphagia (20%), anastomic leak (15%), chyle leak (2.5%), (No DGES) consisted of 140 patients. A Nissen fundoplica- U or Fishers Exact test determined signifi cance (*p < 0.05). esophagostomy (2.5%), and duodenal stump leak (2.5%). The mean time to discontinuance of EN and removal of tion was performed in all patients, and in 17 of 31 patients RESULTS: Of 241 pts in the database, 60 patients under- the JT was 1.83 and 3.4 months respectively. Although in the DGES: Abnormal study group a gastric drainage went resection of EGJ tumors (I/II/III = 12/22/26). IL was only 17.8% (n = 5) of patients requiring adjuvant therapy procedure was added. Demographic data, preoperative pH performed in 31 (EGJ I/II/III = 10/17/4), LTA in 13 (EGJ were utilizing their JT at the commencement of therapy, and manometry values, and mean follow up were similar I/II/III = 1/1/11), and TA in 16 (EGJ I/II/III = 1/4/11). TA 75% (n = 21) required EN during its course. The median between groups. Delayed gastric emptying symptoms were pts were older (IL 63 y [24–82]: LTA 62 [27–80]: TA 73 y time to adjuvant therapy was found to be slightly longer relieved in 91% of patients, but a favorable surgical out- [38-82])*. TA (7.2 cm [1–11]) and LTA (9.8 cm [4–14]) had in those who required outpatient EN versus those who did come was signifi cantly less likely in patients with DGES larger tumors than IL (4.1 cm [0.4–8.3])*. Post-op pulmo- not (61 vs 90 days, p = 0.08). However, the median time to (see table). The majority of unsatisfactory results in DGES nary complications(IL 10/31 LTA 1/13 TA 4/16), anasto- adjuvant therapy did not differ between those who were patients were due to persistent symptoms of delayed gastric motic leak (IL 4/31 LTA 1/13 TA 1/16), LOS (IL 11d: LTA and weren’t receiving EN at the time of adjuvant therapy emptying. Relief of refl ux symptoms was similar. The addi- 10d: TA 10d), and mortality (IL 2/31: LTA 0/13: LTA 0/16) commencement (80 vs 92 days, p = 0.27). Age (p = 0.42), tion of a drainage procedure did not increase the likelihood did not differ between groups. Lymph node retrieval was number of co-morbidities (p = 0.21), pre-operative percent of a favorable surgical outcome (70.6% vs. 71.4%, p = 1). higher in IL (IL 37 [8–66]: LTA 25 [7–64]: TA 25 [4–40])*, body weight loss (p = 0.88), and clinical stage (p = 0.78) but complete resection was achieved in 57/60 (95%) and CONCLUSIONS: No DGES DGES, DGES, were not signifi cantly associated with outpatient JT use. LESS Heller myotomy with anterior fun- (Comparison DGES, Normal Abnormal did not differ between groups. Patients who suffered a post-operative complication were doplication well palliates symptoms of achalasia with no Group) No Study Study Study p CONCLUSIONS: apparent scar, without inducing refl ux. Placement of addi- (n = 140) (n = 49) (n = 60) (n = 31) Value A tailored approach to cancers of the EGJ most likely to require EN (p = 0.006), an association that tional trocars primarily occurred early in the experience. Favorable 127 37 45 22 0.004 based on patient and tumor characteristics is feasible, is strengthened as the number of complications increased For surgeons profi cient with the conventional laparoscopic symptomatic (90.7%) (75.6%) (75%) (70.1%) associated with a high complete resection rate, and does (p = 00006). outcome (%) not increase morbidity. approach, the learning curve of LESS Heller myotomy with Refl ux symptom 132 44 54 15 0.57 anterior fundoplication is short and safe, as profi ciency is control (%) (94.3%) (89.8%) (94.7%) (88.2%) Table 1: Jejunostomy Tube Associated Complications quickly attained. Mo1610 In-Patient Out-Patient Utilization and Morbidity Associated with Routine Complications Major Minor Major Minor Mo1608 CONCLUSIONS: Patients with GERD and delayed gas- Placement of a Feeding Jejunostomy at the Time of Dislodgment 2 5 The Impact of Delayed Gastric Emptying on Outcometric emptying symptoms are less likely to have a favorable Gastro-Esophageal Resection Leak 5 Small bowel 2 2 of Nissen Fundoplication in GERD Patients symptomatic outcome after Nissen fundoplication com- pared to GERD patients without delayed gastric emptying Omar H. Llaguna*, Hong Jin Kim, Karyn B. Stitzenberg, obstruction Weisheng Chen, Steven R. Demeester, Shahin Ayazi*, symptoms, despite similar relief of refl ux symptoms. The Benjamin F. Calvo, Michael O. Meyers Unable to unclog 1 Site infection 2 Gaurav Sharma, Joerg Zehetner, Kimbely S. Grant, addition of a drainage procedure in patients with proven Division of Surgical Oncology and Endocrine Surgery, University of North Carolina School of Medicine, Chapel Hill, NC Pain requiring 1 Florian Augustin, Daniel S. OH, John C. Lipham, delayed gastric emptying did not improve the symptomatic removal Jeffrey A. Hagen, Tom R. Demeester outcome. BACKGROUND: Broken catheter 1 Surgery, University of Southern California, Los Angeles, CA The purpose of the study was to evaluate the utilization and morbidity associated with feeding jeju- Non-functional 1 Mo1609 Major = Required operative intervention, Minor = Non-operative intervention or no BACKGROUND: Symptoms of delayed gastric emptying nostomy tubes (JT) placed at the time of gastro-esophageal intervention required; no prolongation of hospital stay. are common in patients with gastroesophageal refl ux dis- A Tailored Surgical Approach to Esophago-Gastric (GE) resection. ease (GERD). The aim of this study was to compare symp- Junction Cancers: Can We Maximize Complete METHODS: Under IRB approval a prospective database of tomatic outcome after Nissen fundoplication in patients Resection Without Increasing Morbidity? patients undergoing GE resection from January 2004 to Sep- CONCLUSIONS: Routine JT placement carries a low mor- with and without delayed gastric emptying symptoms tember 2010 was reviewed. Data analyzed included patient (DGES). Frank Schwenter*, Sara Najmeh, Lorenzo E. Ferri bidity in patients undergoing GE resection. Because it is Poster Abstracts demographics, post-operative complications, JT use, and JT diffi cult to pre-operative ascertain who will need prolonged PATIENTS AND METHODS: Surgery, McGill University, Montreal, QC, Canada specifi c complications. Fisher’s Exact tests explored associa- Records from 650 patients EN, the routine placement of a JT is recommended, particu- Monday BACKGROUND: who had a primary Nissen fundoplication for proven GERD The primary goal in surgery for esoph- tions with utilization of a JT following resection. larly in those who will likely require adjuvant therapy or from 2002–2008 were reviewed, and patients with symp- ago-gastric junction (EGJ) cancers is to obtain a complete RESULTS: 74 patients (51 males, 23 females, mean age are at high risk for post-operative complications. Despite toms of delayed gastric emptying (nausea, early or pro- resection (R0), as this is the strongest predictor of survival. 60.3±14.1 years) underwent placement of a JT at the time patient desires for early removal of unused JT, caution longed satiety) were identifi ed. A comparison group of age The published rate of complete resection is only approx. of resection (total gastrectomy = 29, Ivor-Lewis = 28, subto- should be taken if adjuvant therapy is being considered. and sex matched patients who had primary antirefl ux sur- 70% and the optimal surgical approach to achieve R0 is tal gastrectomy = 8, proximal gastrectomy = 6, transhiatal gery during the same period for proven GERD but without controversial. We sought to determine the infl uence of an esophagectomy = 3) of both malignant (98.6%) and benign symptoms of delayed gastric emptying was also identifi ed. approach tailored to patient and tumor characteristics on (1.4%) disease processes. 22 JT specifi c complications (2 Outcome after surgery was compared between groups, and complete resection and post-operative outcomes. major, 19 minor) were identifi ed. (Table 1) Re-operation a favorable symptomatic outcome was defi ned as either METHODS: A prospectively entered database on all upper was required in the management of 2 complications (small complete relief of all preoperative symptoms without trou- GI cancer resections at a single institution (2005–2010) bowel obstructions), while all others were easily managed blesome new symptoms, or residual symptoms occurring was accessed for EGJ tumors (Type I, II, III). Patient demo- by an interventional radiologist (n = 8), bedside physi- less than once a month. graphics, operative and tumor characteristics, short term cian (n = 5), or did not require intervention (n = 6). 86% RESULTS: Preoperative symptoms of delayed gastric emp- outcomes, and margin status (proximal/distal/radial) were of patients were discharged tolerating a post-gastrectomy tying were present in 140 patients: 60 had a normal gastric compared between patients undergoing one of three opera- emptying study (DGES: Normal study), 31 had either an tive approaches (Ivor-Lewis [IL]: Left Thoraco-abdominal

118 119 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1611 CONCLUSIONS: The more common prevalence of pha- Lap (n = 17) Open (n = 22) Total (n = 39) P Value ryngeal refl ux, as compared to proximal esophageal refl ux, Male 10 (59%) 9 (41%) 19 (49%) 0.267 Pharyngeal pH Monitoring May Be Superior to particularly in subjects with extraesophageal symptoms, Age (years) 64 (55–70) 58 (46-69) 61 (51–69) 0.256 Proximal pH Monitoring in the Detection of suggests that pharyngeal pH monitoring may be a more Tumor Size (cm) 3.2 (1.9–5.3) 5.5 (4.0–10.0) 4.05 (3.0–8.0) 0.016 Laryngopharyngeal Refl ux sensitive diagnostic tool for LPR than proximal pH moni- Margins negative 15 (88%) 17 (77%) 32 (82%) 0.425 EBL (mL) 200 (100–212) 300 (200–500) 200 (150–350) 0.027 Candice L. Wilshire*, Kelly M. Galey, Thomas J. Watson, toring. Symptom relief and pharyngeal pH normalization post fundoplication provides further evidence of the utility LOS (days) 4 (3-6) 6 (5-7) 6 (4–7) 0.041 Carolyn E. Jones, Daniel Raymond, Virginia R. Litle, ASA Class 1 2 3 0 4 (25%) 12 (75%) 2 (9%) 6 (27%) 14 (64%) 2 (5%) 10 (26%) 26 (68%) 0.439 Jeffrey H. Peters of ambulatory pharyngeal pH monitoring. BMI 27.7 (25.2–30.3) 30.7 (26.3–33.7) 28.4 (25.4–33.4) 0.335 Thoracic and Foregut Surgery, University of Rochester Medical Center, Major complications 2 (12%) 4 (18%) 6 (15%) 0.583 Rochester, NY Clinical: Hepatic Malignant 14 (82%) 10 (46%) 24 (62%) 0.778 Disease CRC Metastases HCC 11 (65%) 3 (18%) 0 1 (6%) 2 1 (5%) 5 (23%) 2 (9%) 2 (9%) 4 (18%) 21 (55%) 8 (21%) 2 (5%) 3 (5%) 6 0.012 BACKGROUND: Determining a causal relationship between Intrahepatic Cholangiocarcinoma (12%) 0 3 (14%) (11%) 3 (8%) Adenoma FNH Hemangioma abnormal refl ux into the proximal esophagus/pharynx and Mo1612 ()Medians with inter-quartile range or % EBL = estimated blood loss, LOS = length of stay, ASA = American Society of Anesthesiologists BMI = body mass index extraesophageal manifestations of gastroesophageal refl ux CRC = colorectal cancer, HCC = hepatocellular carcinoma, FNH = focal nodular hyperplasia disease (GERD) remains a diagnostic challenge. In this Laparoscopic Hepatic Resection: Do Favorable Short- study we aim to determine whether pharyngeal pH moni- Term Outcomes Make It the Procedure of Choice for toring provides superior sensitivity over dual-channel pH Lesions of the Left Hemiliver? testing in detecting laryngopharyngeal refl ux (LPR). C. Kristian Enestvedt*1, Brian S. Diggs1, Brett C. Sheppard1, METHODS: 7 control subjects and 17 symptomaticSusan L. Orloff 2, Kevin G. Billingsley3 CONCLUSIONS: In this select group of patients undergo- who only underwent liver resection and group 1 with 48 patients, 4 with typical GERD and 13 with primary respi-1 Surgery, Oregon Health & Science University, Portland,2Abdominal OR; ing resection of the left hemi-liver, short-term outcomes patients (55%) who received preoperative chemotherapy. ratory symptoms, underwent 24-hour ambulatory esopha-Organ Transplantation, Oregon Health & Science University, Portland, favored the laparoscopic approach. The open approach was All variables, which could infl uence surgical outcome were geal multichannel intraluminal impedance (MII)-dual pH OR; 3Surgical Oncology, Oregon Health & Science University, Portland, OR used more often for larger, benign tumors. Long-term data analyzed: hepatic injury found on histology, liver function simultaneously with pharyngeal pH monitoring. The dis- BACKGROUND: regarding disease recurrence and survival are needed in this tests ( total bilirubin, ALT, AST, Albumin, INR) before sur- tal pH sensor was placed 5 cm above the manometrically Recent reports demonstrate the feasibil- group of patients. These results suggest that laparoscopic gery till hospital discharge, chemotherapy treatment (5FU, determined upper border of the lower esophageal sphincter ity of laparoscopic liver resection for patients with benign resection is safe and for appropriately selected patients may Oxaliplatin, Irinotecan, monoclonal antibodies, intraarte- (LES) and the proximal 15 cm above. Pharyngeal pH was and malignant conditions. We hypothesized that patients be the optimal technical approach. rial or systemic), number of infusions, interval between monitored concomitantly using a separate pH probe posi- who underwent laparoscopic resections for lesions of the chemotherapy and surgery, need of transfusion, postop- tioned 1cm below the uvula. Data collection was synchro- left liver would have short term outcomes similar to those Mo1613 erative complications, length of hospital stay, comorbidi- nized between the devices. Esophageal refl ux was considered who had open surgery. ties. Univariate, bivariate and multivariate analyses were present if pH dropped to < 4 in either pH sensor, and/or a METHODS: A retrospective review compared laparoscopic Liver Resection for Colorectal Metastases Is Safe Afterperformed. drop occurred ≥50% from baseline in impedance 3, 5, 7, or Preoperative Chemotherapy: A Prospective left liver resections for both benign and malignant diseases RESULTS: The probability of developing severe steatosis 9 cm above LES (distal) or 15 and 17 cm above LES (proxi- Non-Randomized Controlled Study to open controls for the years 2007–2010. Demographic (2,58 times) and necrosis (4 times) was higher in patients mal). Separate pH thresholds of <5.5, 5.0, 4.5, and 4.0 were information, pathology, tumor characteristics, procedure Marco Massani, Roberta Bonariol, Ezio Caratozzolo, that received preoperative chemotherapy; furthermore, the defi ned for refl ux episodes detected in the pharyngeal probe. related factors, complications and mortality were assessed. RESULTS: Cesare Ruffolo*, Francesco Calia Di Pinto, Nicolò Bassi risk of steatohepatitis in these patients was higher. The val- At a threshold of pH < 5.5, an average of 1 (±4) Univariate analysis was performed with signifi cance of II Department of Surgery (IV unit), Regional Hospital Cà Foncello, ues of liver function tests of group 1 compared to group 0, pharyngeal refl ux event over 24 hours was seen in control p < 0.05. Treviso, Treviso, Italy as expected, were worse (p = 0.007). The number of infu- subjects. Symptomatic patients had greater pharyngeal RESULTS: Seventeen patients had laparoscopic resection BACKGROUND: sions did not imply a higher number of surgery complica- pH exposure than controls, averaging 7 (±14) episodes/24 during the study period, with 1 case converted to open Preoperative chemotherapy before liver tions (p = 0.075), but it correlated directly with bilirubin hours in those with typical GERD symptoms and 46 (±76) (5.9%), and there were 22 open cases. Three anatomic resection can determine hepatic parenchyma damage values and inversely with albumin values. No correlation in those with respiratory symptoms. Total pharyngeal refl ux left hepatectomies, 11 left lateral segmentectomies, and 3 with a negative infl uence on the patients’ postoperative was found between time interval from the last infusion and events (603) were markedly more common in patients with lesser segmentectomies were performed in the laparoscopic outcome. surgery and surgical outcome. With regard to transfusion respiratory symptoms than either control (10) or typical group versus 15, 7, and 0 in the open group, respectively. AIM: The aim of this prospective study was to assess the need and length of hospital stay no difference between the Poster Abstracts GERD symptoms (28). Further, the highest number of pha- There were no differences in demographics, margin sta- effects of preoperative chemotherapy on hepatic paren- two groups was found. At univariate analysis, most differ- ryngeal refl ux episodes recorded across all pH thresholds tus, or major complications (see table). The laparoscopy chyma in a consecutive and homogeneous series of patients ences resulted to be signifi cant but were not confi rmed in Monday was observed in subjects presenting with primary respira- group had signifi cantly lower intraoperative blood loss and evaluating the outcome in terms of postoperative morbid- the multivariate analysis. tory symptoms: 603, 91, 38, and 40 events at pH < 5.5, shorter length of stay compared to the open group. Median ity at 30 days after surgery, liver function and the need of CONCLUSIONS: 5.0, 4.5, and 4.0, respectively. 6 of the 11 patients with Preoperative chemotherapy can cause tumor size was signifi cantly larger in the open group. There transfusion. liver damage without a negative infl uence on surgi- abnormal distal pH results had corresponding abnormal was one death in the open group and none in the laparo- MATERIALS AND METHODS: Eighty-seven consecutive cal outcome; therefore liver resection is safe also after pharyngeal acid exposure; however, only 3 had concomi- scopic group. tant positive proximal esophageal pH results. Pharyngeal patients that underwent liver resection for colorectal metas- chemotherapy. pH also appears superior to the proximal esophageal pH tases in our surgical unit were included: the patients were in differentiating GERD related respiratory symptoms, as divided into two groups: group 0 with 39 patients (45%) compared to gastrointestinal. Fundoplication normalized pharyngeal pH and markedly relieved symptoms in a sin- gle patient with severe respiratory symptoms and normal proximal esophageal acid exposure.

120 121 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1614 This case suggests that optimizing conditioning of the signifi cance (p = .027, OR 2.2). When AFP- patients were Mo1616 future liver remnant can render patients eligible for for- excluded, a clear trend towards better survival was seen Increasing Safety of Liver Resection in Cirrhotic :mal liver resections that are otherwise only candidates for in AFPmod patients compared to AFPhigh (73% vs. 51%, Feasibility Demonstration of Single Incision Multiple An Initial Experience Combining Autologous CD 133+palliative treatment. Since an R0 resection is still the only p = .11). In univariate analysis inclusive only of AFP+ Port Laparoscopic Liver Resection (SIMPLLR) in Three Stem Cell Application with In-Situ Splitting and Twopotentially curative treatment, and liver transplants are an patients, tumor size >5 cm was associated with high AFP Patients Stage Hemihepatectomy option for only a minority of patients, a combination of levels (p = .036, OR 2.1), but poor differentiation and vas- Francis Cannizzo, John R. Romanelli, Ramanathan Seshadri*, this innovative approach of a two stage liver resection with cular invasion were not. Andrea Alexander*, Claus F. Eisenberger, Alexander Rehders, Richard B. Arenas liver augmentation by CD133+ bone marrow derived stem Stefan A. Topp, Matthias C. Schauer, Jan Schulte Am Esch, Department of Surgery, Baystate Medical Center, Springfi eld, MA cells may change perspectives for some cirrhotic patients Table 1: Predictive Value of High Serum AFP (>1000 ng/ml) to Wolfram T. Knoefel with hepatocellular cancer. Histologic Features Indicative of Aggressive Tumor Biology INTRODUCTION: Laparoscopic liver resection has reduced Surgery, Heinrich-Heine-University, Duesseldorf, Germany morbidity and length of stay for hepatic surgery. Introduc- Major liver resections carry a signifi cant mortality due Mo1615 Vascular Invasion Poorly-Differentiated Histologytion of single incision laparoscopic surgery exploits the to liver insuffi ciency in patients with liver cirrhosis. To Sensitivity 42% 49% specimen extraction incision and may improve patient Serum AFP Elevation in Patients with Hepatocellular Specifi city 71% 65% increase safety of major liver resections we aimed to com- satisfaction. Carcinoma Is Associated with Poor Survival and Tumor Positive Predictive Value 87% 47% bine the regenerative potential of bone marrow derived Negative Predictive Value 21% 67% METHODS: Three patients were selected based on the stem cell application and the safety of a novel two stage Size, but Not with Histologic Features Indicative of anatomy of anticipated resections to undergo single inci- procedure in a patient with centrally located hepatocellular Aggressive Tumor Biology sion laparoscopic liver resections. Pt 1 is a 59 yo female with cancer and Child B cirrhosis. Kunal R. Parikh*, Ghalib A. Jibara, Fotini Manizate, a 5.5 cm left lateral segment hepatocellular carcinoma. Pt 2 A 76-year-old male patient with a cryptogenic liver cirrhosis Daniel Labow, Spiros P. Hiotis is a 69 yo female with a 4.0 cm HCC in Reidel’s lobe. Pt 3 is presented with a mass of 5 cm in Segments IV a/b. A biopsy Surgical Oncology, Mount Sinai School of Medicine, New York, NY a 21 yo female with a 2.5 cm symptomatic, indeterminate left lateral segment lesion. All patients were counseled as to showed a moderately differentiated (G2) hepatocellular INTRODUCTION: carcinoma. Since a suffi cient distance to the portal pedicle Serum alpha-Fetoprotein (AFP) is the risks of this procedure and the possibility of conversion and the hepatic veins was suggested by preoperative imag- widely utilized in screening patients at risk for hepatocel- to standard laparoscopic or open technique in the event ing and no contraindication for an intentionally curative lular carcinoma (HCC). AFP levels also provide prognostic of intraoperative complications. All patients gave informed resection was seen, the patient was explored. In the opera- information, as several reports have associated high AFP consent for participation. levels (>1000 ng/ml) with poor survival. Although clearly tion the limits of the carcinoma were diffi cult to discern RESULTS: Pt 1 underwent a single incision multiport left associated with transformation to HCC and with poor within the cirrhotic liver and a non-anatomic Segment lateral segmentectomy (EBL: 500 cc, LOS: 4 days). Pathol- outcomes, the etiologic process leading to AFP elevation IVa/b resection resulted in an R1 situation. The patient was ogy showed a moderately differentiated HCC resected with remains debatable. Some experts interpret high AFP as an not considered eligible for a formal central or left resec- clear margins. She is NED at 15.5 months. Pt 2 underwent indication of de-differentiation and vascular invasion. We tion due to the advanced liver cirrhosis. He recovered from adhesiolysis and resection of Reidel’s lobe using a single- investigate here the predictive value of AFP to survival surgery without complications and left the hospital on pod port SILS™ device (Covidien, Inc) (EBL: 1700 cc, LOS: 5 and poor histologic features following liver resection in a 15 with persisting ascites. Postoperative staging was pT2b, days). IOUS-guided parenchymal transection wa s com- single-institution series. pN1, L0,V0,Pn1 G2, cM0, R1. pleted laparoscopically. Minilaparotomy was required for METHODS: After further recovery, medical suppression of his ascites A retrospective analysis was performed on specimen inspection and delivery. She received 1 unit and extensive counselling of the patient it was decided to 238 patients with HCC maintained in a prospective data- PRBCs intraoperatively. Pathology revealed a 4 cm poorly attempt a curative resection. To improve hepatic reserve an base. All patients had underlying chronic hepatitis B, and differentiated HCC abutting Glisson’s capsule. She required in-situ splitting of the left and right liver (in-situ left hemi- underwent liver resection as defi nitive therapy for HCC. readmission for pain control. She is NED at 13.25 months Kaplan-Meier, chi square, univariate, and multivariate followup. Pt 3 had a single incision multiport left lateral hepatectomy) leaving only the artery, the bile duct and the CONCLUSION: HCCs that do not produce serum AFP are regression analysis were performed using SPSS software segmentectomy (EBL: 150 cc, LOS: 2 days). Pathology middle and left hepatic veins intact, was performed. This associated with better survival following liver resection, and (SPSS, Chicago, IL). showed a 2.8 cm FNH abutting the capsule. She had com- ischemic injury to the left liver was combined with condi- tumors associated with high levels of AFP (>1000 ng/ml) have RESULTS: plete resolution of her symptoms at followup. tioning of the intact right liver with autologous CD133+ Seventy six patients (32%) were diagnosed worse survival outcomes overall. High serum AFP levels are bone marrow derived stem cells via the right portal vein. with tumors that did not produce AFP (AFP-, <9 ng/ml). associated with tumor size, but not with poorly-differenti- CONCLUSION: SIMPLLR can be a viable adjunct to lapa- Poster Abstracts The patient tolerated this procedure well. However, the cir- Of the 143 patients (60%) whose tumors did produce AFP ated histology or vascular invasion on regression analysis, roscopic hepatic resection for selected lesions especially in rhotic right liver did not increase in size. On pod 12 the (AFP+), 80 (56%) were diagnosed with moderate serum and the predictive value of high AFP for these features is the left lateral segment and some marginal lesions. Further Monday left liver was removed. The patient developed a moderate AFP elevation (AFPmod, 10–999 ng/ml), and 63 (44%) had poor (Table 1). The association of high serum AFP to poor study and experience is needed to defi ne the limits of this hepatic insuffi ciency with bilirubin levels up to 4 g/dl and high AFP (AFPhigh, >1000 ng/ml). AFP- patients had bet- survival and large tumor size is presumably due to delay in technique and its optimum applications. Prior to adoption, occasional need to substitute his coagulation factors. The ter survival following resection compared to AFP+ patients diagnosis, rather than more aggressive tumor biology. standard laparoscopic solid organ surgery should be mas- patient did not develop renal insuffi ciency, encephalopa- (72% vs. 61% 5-yr, p = .013). When analyzed inclusive of tered. A low threshold to convert to standard laparoscopic thy or hepato-pulmonary syndrome and was discharged all 238 patients, AFPhigh was associated to tumor size >5 or open approach is necessary to assure patient safety and on pod 15 after the conditioning operation in excellent cm (p = .021, OR 2.1), poor differentiation (p = .006), and outcomes. general condition. He is tumor-free at his fi rst oncologic vascular invasion (p = .028, OR 4.8) by univariate analy- follow-up. sis. In multivariate analysis only tumor size > 5cm retained

122 123 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Clinical: Pancreas Mo1618 CONCLUSIONS: Pancreatic leak remains a frequent com- Mo1620 plication after distal pancreatectomy, although severe Staple and Non-Staple Closure of Pancreatic Remnant complications (≥grade III) rarely occur. Staple closure of the Are Body/Tail Pancreatic Cancers Really Worse Per Se? Mo1617 After Distal Pancreatectomy: A Multicenter pancreatic remnant is associated with a signifi cantly lower Debunking a Myth Retrospective Analysis of 388 Patients clinical fi stula rate. Elizaveta Ragulin-Coyne*1, Elan R. Witkowski1, Jillian K. Smith1, Usefulness of Modifi ed Possum for Major Pancreatic *1 2 3 Joshua S. Hill 3 4 1 Resection Daisuke Ban, Kazuaki Shimada, Masaru Konishi, , James T. Mcphee, Sing Chau NG, Katsuhiko Uesaka 4, Akio Saiura5, Masaji Hashimoto6 Mo1619 Waddah B. AL-Refaie 2, Shimul A. Shah1, Jennifer F. Tseng1 Takeshi Nishi*, Tsuneo Tanaka, Yasunari Kawabata, Seiji S.1 Yano 1 Department of Hepato-Biliary-Pancreatic Surgery, Tokyo Medical Surgical Outcomes Analysis & Research (SOAR) Department of Digestive and General Surgery, Shimane University Graduate School of Perioperative Blood Transfusions May Infl uence and Dental University, Tokyo, Japan;2Hepatobiliary and Pancreatic Prognosis After Surgery for Pancreatic Cancer Surgery, University of Massachusetts Medical School, Worcester, MA; Medicine, Izumoshi, Japan 2 Surgery Division, National Cancer Center Hospital, Tokyo, Japan; Independent of Complications or Body Mass Index: Division of Surgical Oncology, Department of Surgery, University AIM: 3 3 We examined the usefulness of three modifi ed ver- Department of Surgery, National Cancer Center Hospital East, Multivariate Analysis of 270 Resected Patients of Minnesota, Minneapolis, MN;Department of Surgical Oncology, sions of POSSUM in pancreatectomy. We also explored Kashiwa, Japan;4Division of Hepato-Biliary-Pancreatic Surgery, Division of Surgery, University of Texas MD Anderson Center, wheter a high-risku group can be defi ned. Shizuoka Cancer Center Hospital, Shizuoka, Japan;5Department of Tobias Keck*, Ulrich F. Wellner, Ulrich T. Hopt, Frank MakowiecHouston, TX; 4Division of Vascular Surgery, Department of Surgery, PATIENTS AND METHODS: The data of 431 consecutiveGastrointestinal Surgery, Cancer Institute Hospital, Tokyo, Japan; Department of Surgery, University of Freiburg, Freiburg, GermanyBrigham and Women’s Hospital, Boston, MA patients who underwent 316 pancreatoduodenectomy and 6Department of Digestive Surgery, Toranomon Hospital, Tokyo, Japan Survival after surgery for pancreatic cancer (PaCa) remains BACKGROUND: Pancreatic cancers of the body and tail 115 distal pancreatectomy between June 1996 to Decem- poor, even after curative resection. Factors like nodal dis- BACKGROUND: Distal pancreatectomy is a simple sur- have been described to have poorer survival compared with ber 2007 were evaluated retrospectively. The morbidity and gical procedure with a low mortality rate, but morbidity ease (lymph node ratio), resection margin, grading and that of the head. In order to eliminate the possible con- mortality equations of Portsmouth POSSUM (P-POSSUM), associated with pancreatic fi stula remains a problem. This tumor size have been identifi ed as prognostic factors in founding of later stage at diagnosis for body/tail lesions, elderly POSSUM (E-POSSUM), and Hiroshima POSSUM study sought to identify the risk factors for clinical pancre- many series. Overweight/adipositas has recently been sug- we examined the impact of cancer location on overall sur- (H-POSSUM), were estimated. The subjects were divided atic fi stula, as defi ned by the International Study Group of gested as a further (negative) prognostic factor. Periopera- vival in a cohort of early stage pancreatic adenocarcinoma into two groups by physiological score (PS); 1) low-risk grop Pancreatic Surgery (ISGPS). tive complications and blood transfusions (blood-Tx) have patients. is less than 20 of PS, 2) high-risk group is more than 21. been suggested to worsen prognosis in various cancers. We METHODS: METHODS: RESULTS: The medical records of all patients who analyzed our current experience after resection of PaCa SEER-Medicare 1991–2005 was used to iden- Postoperaive complications were seen in 170 underwent distal pancreatectomy at fi ve Japanese insti- tify patients with localized, node negative (Stage 1) pan- cases (39.4%). Operative death within 30 days of surgery with additional consideration of the above mentioned tutions between January 2001 and June 2009 were retro- parameters. creatic adenocarcinoma as defi ned by SEER historic stage was only 4 (0.9%), and 11 deaths during hospital stay spectively reviewed. All relevant anonymized data from and abstracted histological data. Chi square analyses were METHODS: (2.6%). The mean PS in the group with postoperative patients entered into electronic case report forms were syn- Long term outcome could be assessed in 270 performed to defi ne and describe the head and body/tail complication was signifi cantly higher than that in the thesized into a common database Data obtained for each patients after resection of PaCa (81% head, 13% distal, cohorts. Kaplan-Meier survival analysis was performed group without complication. The observed and expected patient included the following: age, sex, body mass index, 6% total pancreatectomy) since 1995. Perioperative blood comparing head vs. body/tail groups in resected, unre- (O/E) ratio of postoperative morbidity was 0.72 (range; diabetes mellitus, American Society of Anesthesiologists transfusions were given in 46%. One third of the patients sected and overall cohorts. Finally, Cox proportional haz- 0.57–0.90) by original POSSUM, 0.95 (range; 0.50–1.28) by classifi cation, previous laparotomy, primary disease surgi- underwent additional mesenterico-portal vein resection. ards modeling was used to control for the effects of age, E-POSSUM, 0.94 range; 0.68–1.17) by H-POSSUM. The O/E cal procedures, intraoperative bleeding, operation time, Free margins were achieved in 71%. 70% had nodal disease, sex, race, resection, chemotherapy, and radiation on head ratio of hospital death was 0.17 (range;0-0.24) by original hospital stay, postoperative complications, and mortality. and 45% had more than one positive node. Postoperative and body/tail groups. POSSUM, 0.61 (range; 0–1.50) by P-POSSUM, 0.65 (range; Pancreatic fi stula was assigned according to the ISGPS clas- morbidity was 49% (any), 31% (surgical) or 10% (severe; RESULTS: 0–1.00) by E-POSSUM, 1.00 (range; 0–2.00) by H-POSSUM. sifi cation into 4 categories: no fi stula, Grade A, B, and C requiring relaparotomy and/or mechanical ventilation), 1,320 patients with localized, node-negative The morbidity and mortality in high-risk group were sig- fi stula. The primary endpoint was the occurrence of clini- respectively. Survival was analyzed by a Kaplan-Meier- pancreatic adenocarcinoma were identifi ed; 309 of them nifi cantly higher than those in low-risk group. cal pancreatic fi stula Grade B or C. method and the Cox-regression model. underwent resection. In all, 1,098 head and 222 body/tail cancers were identifi ed. Within resected group, median sur- CONCLUSION: RESULTS: RESULTS: The E-POSSUM and H-POSSUM were bet- Of the 388 patients studied, 226 were male and Overall fi ve year survival was 16% (16 true fi ve vival from diagnosis were similar for head and body/tail ter than the original POSSUM in terms of O/E ratio, an 162 were women, with a median age of 65 years (range year survivors). In univariate analysis positive margins (p lesions (18 mo for head vs. 18.5 mo for body/tail, p = 0.66). index of accuracy. We also found that a high-risk group 13–85 years). Pancreatic ductal carcinoma was present in = 0.002), more than one involved node (p < 0.001), poor Within unresected group, head patients demonstrated a can be defi ned by PS scores, which can be determined 169 patients (51.0%). Management of the pancreatic rem- grading (G3/G4; p < 0.02) and blood-Tx (p = 0.004) were trend toward longer survival (6.3 mo vs. 5.0 mo, p = 0.07).

nant was varied: stapler suture was used in 224 patients, Poster Abstracts preoperatively. associated with poorer survival. Other parameters like BMI, For all patients (resected and unresected), the survival was hand-sewn closure was employed in 43 patients (11.1%), tumor size, postoperative complications (all above defi ni- equivalent (head, 7.5 mo vs. body/tail 7.2 mo, p = 0.64).

and no suture with duct ligation was performed in 118 Monday tions), vein resection, gender, location of PaCa/resection Cox analysis was used to assess independent factors impact- patients (30.4%). (head/distal) or time period of surgery did not infl uence ing survival. In the total cohort, age >85 was a negative Overall incidence of any pancreatic fi stula (clinical or bio- survival. In multivariate (Cox) survival analysis, again, the predictive factor HR 1.22 p = 0.03; chemotherapy (HR 0.58 chemical) was 58% (n = 225). Grade A, B and C pancre- resection margin (p < 0.01; RR 1.5), metastatic nodes (> p < 0.01) and resection (HR 0.32 p < 0.01) were positive atic fi stula occurred in 95 patients (24.5%), 128 patients one; p = 0.002; RR 1.6), blood-Tx (p = 0.03, RR 1.3) and predictive factors. After adjustment, body/tail lesions did (33.0%), and 2 patients (0.5%), respectively. There was (borderline) grading (p = 0.09) independently infl uenced not demonstrate a signifi cantly different risk of death com- no operative mortality and no in-hospital deaths, but 4 prognosis. pared to head lesions (HR 1.1 p = 0.24). patients (1.0%) required reoperation. The distribution of CONCLUSIONS: In our study long term prognosis after CONCLUSION: grades according to the Clavien-Dindo classifi cation was as Using a large national database, we have resection of pancreatic cancer was infl uenced not only by demonstrated that localized, node-negative pancreatic follows: 100 patients (25.8%) grade 0; 125 (32.2%) grade 1; ‘established’ tumor-related parameters but also by periop- 156 (40.2%) grade 2; 5 (0.1%) grade 3a; 1 (0.3%) grade 3b; cancers in the body and tail of the pancreas have equiva- erative blood transfusions. This effect seems to be inde- and 1 (0.3%) grade 4a. There was no grade 4b or 5 compli- lent survival compared to head lesions, after adjustment pendent of perioperative complications or type/extent of cations. In multivariate analysis, diabetes mellitus, method of patient and treatment factors. Patients with resectable resection. In contrast to other recently published results we of stump closure, and duration of operation were found to tumors of the body and tail deserve prompt and compa- be independently associated with clinical pancreatic fi stula, could not demonstrate an impact of body mass index on rable surgical evaluation. with respective hazard ratios (95% confi dence interval) of prognosis. 3.55 (1.83-6.85: p < 0.001), 0.32(0.199-0.53: p < 0.001), and 0.32 (0.19-0.52: p < 0.001).

124 125 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

METHODS: A retrospective review of a prospective data- where laparoscopy clearly offers improved visibility and Mo1621 Mo1622 base for Pancreatic Cyst Neoplasms Database (PCND) July smaller incisions, the benefi ts of laparoscopy during com- Usefulness of the International Study Group for Changes in Incretin Responses and Glucose 2000 through September 2010. plex phases of PD are less clear. We have utilized a tech- Pancreatic Fistula (ISGPF) Classifi cation for Clinical Metabolism After Pancreatectomy RESULTS: There were 800 patients identifi ed in the nique of laparoscopic hand assist (LHA) during PD to Decision Making perform the early phases of dissection and mobilization, Yasuhisa Mori*, Takao Ohtsuka, Kosuke Tsutsumi, Takaharu PCND. Sixty one (7.6%) patients (40 non operated and 21 which in turn allows the remainder of the procedure to be Yasui, Junji Ueda, Shunichi Takahata, Masafumi Nakamura, Florian Gebauer, Michael Tachezy, Yogesh K. Vashist, operated) were identifi ed as serous cystic neoplasms. Of performed through a handport incision in the right subcos- Masao Tanaka Guellue Cataldegirmen, Jakob R. Izbicki, Maximilian Bockhorn* the non operated patients the median age was 71 years old Surgery and Oncology, Kyushu university, Fukuoka, Japan General- Visceral and Thoracic Surgery, University Medical-Center and included 30 women (75%). At the time of diagnosis 38 tal area (see Figure 1). This report examines our experience Hamburg-Eppendorf, Hamburg, Germany (97%) of the patients were asymptomatic and median cyst with LHA-PD compared to a matched control group under- OBJECTIVE: Despite the reduced number of E-cells, glucose going standard PD through a bilateral subcostal incision. BACKGROUND: size was 2.77 cm. A CT scan was performed as diagnostic metabolism often improves after major pancreatic resec- Due to its retrospective character, the test in 37 (95%). An FNA was performed in 32 (80%) of the tion. To investigate the postoperative pancreatic E-cell func- classifi cation system of the International Study Group of patients, cellularity was benign in 38 (96%) and atypical tion and hormonal responses of glucagon-like peptide-1 Pancreatic Fistula (ISGPF) lacks any prognostic capacity cells were present on 2 (4%) of the patients. Median fol- (GLP-1) and glucose-dependent insulinotropic polypeptide regarding which patients will develop further fi stula-related low up was 3.05 years. After fi rst year follow-up 1 patient (GIP) using the two different types of pancreatectomy; pan- complications. This study aimed to critically evaluate the became symptomatic and was resected. Cystic size ≥4 cm creatoduodenectomy (PD) and distal pancreatectomy (DP). usefulness of the ISGPF classifi cation system with respect were present in 10 (25%) patients and <4 cm in 30 (75%). to clinical decision making. SUMMARY BACKGROUND DATA: Incretins have been The growth rate of the group was 0.35 cm per year. For cysts recently identifi ed to have important roles in the regula- METHODS: Between 1992 and 2009, 1966 patients under- ≥4 cm the growth rate was 0.24 cm per year and in cysts <4 tion of the pancreatic E-cell function by stimulating insulin went surgery of the pancreas. The defi nition of postop- cm the growth rate was 0.35 cm per year (p = 0.22). Twenty secretion, and become a topic in the fi eld of the treatment erative pancreatic fi stula (POPF) was based on the ISGPF one patients underwent surgery. The median age was 68 of diabetes mellitus. Roles of incretins after pancreatec- consensus. All patient data were entered into a prospective years and 17 patients (80%) were female. Indications for tomy which causes the decreased number of E-cells has not clinical data management system. surgery included presence of symptoms in 7 (34%), mucin been well documented. RESULTS: 276 patients (14%) developed POPF. ISGPF type positive in 6 (28%), increase in size 6 (28%) and FNA with atypical cells in 2 (10%). Surgical procedures performed METHODS: Oral glucose tolerance tests were performed A fi stula was seen in 65 patients (24%), type B in 110 (39%) and type C in 101 (37%). While 101 patients (37%) required included; distal pancreatectomy and splenectomy in 11 in 34 patients who underwent pancreatectomy (20 PD and (50%). whipple in 8 (38%) and central pancreatectomy in 1 14 DP) before, 1 month and 6 months after operation. reoperation, 175 (63%) were managed conservatively. Uni- variate analysis identifi ed underlying disease, type of oper- (5%) and 1 (5%) diagnostic laparotomy. A CT scan was per- Figure 1: Pancreaticoduodenectomy performed through a handport There were 14 patients with malignant diseases (70%) in 20 formed as a diagnostic test in 19 (90%), median cystic size patients undergoing PD and 5 (36%) in 14 DP. The changes ation, and high levels of serum amylase or bilirubin on the incision day of onset to be prognostic parameters for reoperation. was 3.87cm. An FNA was performed in all of the patients, in the serum glucose and insulin concentrations, homeo- METHODS: 31 patients underwent LHA-PD between Multivariate analysis found elevated serum C-reactive pro- cellularity was benign in 19 (90%) and atypical cells were stasis model assessment of insulin resistance (HOMA- 2007–2010 for benign lesions or malignant tumors <3 cm tein to be an independent factor for increased in-hospital present on 2 (8%). Seventeen (80%) of the patients were IR) and homeostasis model assessment of E-cell function and without vascular invasion. A group of 31 consecutive mortality. Solely due to their death, 20 patients had to be follow for 1 year, 7 (30%) for 2 years and 4 (20%) for 3 years (HOMA-E) were analyzed. The plasma concentrations of patients undergoing standard PD and matched for clini- classifi ed as having a type C fi stula, even though they suf- before they were referred to the CCF. Cystic size ≥4 cm was GLP-1 and GIP were also measured. copathological features was used as the control group. All fered only ‘only type A or B. present in 5 (20%) patients and <4 cm was present in 16 RESULTS: Plasma GLP-1 concentration 1 month after DP (80%) of the patients. The growth rate as a group was 0.29 procedures were performed by a single surgeon. CONCLUSIONS: The ISGPF classifi cation system is of no transiently increased (P < 0.05), while such a change was cm per year. In cysts ≥4 cm the growth rate was 0.78 cm per RESULTS: LHA-PD and standard PD groups were similar use in clinical decision making, since it does not adequately not observed in PD group. On the other hand, GIP concen- year and in cysts <4 cm the growth rate was 0.45 cm per with respect to operative time (385 vs 379 minutes) and describe a large subgroup of patients. To improve clinical trations after PD continued to be lower compared with the year (p = 0.42). The growth rate in both surgical and non need for transfusion (4 vs 7 patients). Pathological measures decision making about management of patients, it is cru- preoperative values (P < 0.01), although there was no sig- surgical patients were not statistical different (p = 0.23) and of the adequacy of resection were similar in both groups, cial that the existing ISGPF classifi cation system is merged nifi cant change in DP group. Postoperative glucose concen- in patients with cysts <4 cm (p = 0.23). For cysts ≥4 cm, with a positive posterior margin in 3 patients in each group with newer clinical data. trations were signifi cantly lower than preoperative values patients operated had a statistically faster rate of growth and lymph node recovery slightly greater in the LHA-PD in PD, while those values were the same in DP. Postopera- compared to unoperated patients undergoing observation group than in the standard PD group (19 vs 14, p = 0.02). Poster Abstracts tive insulin concentrations were signifi cantly lower than Mo1623 (p = 0.011). Frequency of post-operative complications including pan- preoperative values in both PD and DP groups. Although CONCLUSIONS: Size of SCT alone is not an independent creatic leak, wound infection and gastroparesis was similar Monday HOMA-IR improved transiently 1 month after PD (P < Pancreatic Serous Cystic Neoplasms: Is Size an Indication for Surgery? indication for resection. The indications for surgery are in each group and there was no perioperative mortality. 0.05), such improvement was not observed after DP. The symptoms, and growth rate of approximate 1 cm per year Overall length of stay was lower in the LHA-PD group (10 value of HOMA-E after PD was signifi cantly lower than that Juan R Aguilar-Saavedra*, Greg Lentz, Sricharan Chalikonda, in cysts ≥4 cm. vs 14 days), although this did not reach statistical signifi - before operation (P < 0.05), while that of DP was not sig- David Vogt, Matthew Walsh cance (p = 0.1). When adjusting for extreme outliers (two nifi cantly different from the preoperative value. Surgery, Cleveland Clinic, Cleveland, OH Mo1624 standard deviations above the mean), the LHA-PD group CONCLUSIONS: The glucose metabolism after PD was dif- BACKGROUND: Symptomatic SCTs should be resected; had signifi cantly shorter length of stay than the standard A Novel Approach to Pancreaticoduodenectomy: ferent from that after DP in terms of postoperative E-cell but is unclear whether growth rate dictates that resection PD group (8.2 vs 10.8 days, p = 0.02). Utilization of a Laparoscopic Hand Assisted Technique function and hormonal responses of GLP-1 and GIP. Incre- should be performed for size ≥4 cm. CONCLUSION: LHA-PD is safe and feasible and can be tins seem to have roles to regulate the postoperative glu- OBJECTIVE: Nicholas N. Nissen*, Vijay Menon accomplished with operative time and morbidity similar to cose metabolism after pancreatectomy. To determine if the SCTs size and growth rate are indicators for resection. Hepatobiliary Surgery and Liver Transplantation, Cedars-Sinai standard PD and without compromising oncologic princi- Medical Center, Los Angeles, CA ples. This laparoscopic approach has the benefi t of smaller abdominal incisions, improved visualization during early INTRODUCTION: The application of laparoscopic tech- phases of surgery and decreased length of stay. niques to pancreaticoduodenectomy (PD) has not received widespread acceptance. Unlike in distal pancreatic surgery,

126 127 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1625 CONCLUSIONS: Adjuvant therapy should be considered CONCLUSION: Systematic review and meta-analysis, con- symptoms following surgery were abdominal pain 100%, for AC patients with node positive cancer. Identifi cation fi rmed central pancreatectomy as a safe surgical procedure nausea/vomiting 91%, and anorexia 88%. 19 patients Is Postoperative Adjuvant Therapy Indicated for of risk factors for recurrence in the node negative patient with good long term functional reserve although with a required hospitalization with a total of 55 admissions due Ampullary Adenocarcinoma? who may benefi t from adjuvant therapy should be further low increasing of morbidity (pancreatic fi stula) when com- to pain exacerbations. 9% of the patients required enteral Veeraiah Siripurapu*1, Paul P. Koffer2, Fang Zhu1, Yun Shin Chun1, investigated. pared to distal pancreatectomy. tube feeds and 12% of patients were TPN dependent. James C. Watson 1, John P. Hoffman1 The daily PO morphine equivalent was 414.12 mg (SD = 1 Fox Chase Cancer Center, Philadelphia, 2PA;Temple University, Mo1626 Mo1627 366.31). Quality of life scores are presented in Table 1. CONCLUSION: Philadelphia, PA Systematic Review of Central Pancreatectomy “The What Do You Do When Surgery for Chronic Opioid analgesic long-term therapy can BACKGROUND: Ampullary carcinoma (AC) is a rare Dagradi-Serio-Iacono Operation” and Meta-Analysis Pancreatitis Pain Fails: Is Long-Term Opioid Therapylikely be safely administered to patients with intractable pain associated with failed operative therapy of chronic pan- entity accounting for only 6% of all periampullary tumors. Versus Distal Pancreatectomy Effective? creatitis. General health and physical vitality/energy scores These often present earlier due to proximity to the com- *1 2 1 mon bile duct or pancreatic duct. Despite this, rates of Calogero Iacono, Giuseppe Verlato, Andrea Ruzzenente, William M. Leppard*, Stefanie M. Owczarski, Alok Madan, were above standard norms. However, despite utilization of 1 1 local recurrence have been documented as high as 50%,Tommaso Campagnaro , Alessandro Valdegamberi, Jeffrey J. Borckardt, Katherine A. Morgan, David B. Adams high dose narcotic analgesics, social functioning and patient even with an R0 resection. No randomized trial addressesLuca Bortolasi 1, Chiara Tezza1, Alfredo Guglielmi1 Surgery, Medical University of South Carolina, Charleston, SC physical and emotional role limits scores were low. 1 the role of chemoradiation, as refl ected in current NCCN Department of Surgery, University of Verona Medical School, Verona, INTRODUCTION: Surgical management of chronic pan- guidelines. The purpose of this study is to assess the role of Italy; 2Unit of Epidemiology & Medical Statistics Department of creatitis is associated with poor outcomes due to intractable Mo1628 chemoradiation for AC. Medicine & Public Health, Verona, Italy pain and concomitant physical and emotional dysfunction. Detailed Analysis of Learning Curve in Pancreatic AIM: A retrospective analysis of patients with AC, identify- BACKGROUND: Central Pancreatectomy (CP), fi rst described Conventional thinking discounts the effi cacy and safety of Surgery: Surgeon and Hospital Volume Are Equally ing and analyzing demographics, clinical stage at presen- by Dagradi and Serio in 1984 and popularized worldwide long-term opioid treatment in patients with chronic pain Important tation, surgical procedures, pathological stage (7th Edition by Iacono, is a parenchyma-sparing surgical procedure that unassociated with a terminal disease. The purpose of this Ulrich F. Wellner*, Frank Makowiec, Ulrich T. Hopt, Tobias Keck AJCC) , neo-adjuvant & adjuvant treatment and complica- allows removing benign and/or low grade malignant lesion study is to report quality of life outcomes measured utiliz- Surgery, University Hospital Freiburg, Freiburg, Germany tions of surgery. from the neck and proximal body of pancreas. ing a computerized clinical tracking system in a chronic RESULTS: pancreatitis patient cohort treated with long-term opioid INTRODUCTION: Basic operation techniques can be Forty fi ve patients were identifi ed. Seventy per- The aim of the study was to evaluate, from all published analgesia by a surgeon-directed multi-disciplinary team. cent were female with 48% staged 2b and higher. Twenty learned in a relatively short time period even in the case of studies, short and long term surgical results of CP and to METHODS: eight patients had N0 disease (Group A) compared to 17 evaluate results of comparative studies versus distal pancre- A retrospective review and analysis of patients relatively complex operations. However a “learning curve” N1 patients (Group B). Seven patients received adjuvant atectomy (DP). treated for failed surgical management of chronic pan- is to be expected during which the surgeon will gain expe- chemotherapy only, 3 adjuvant radiotherapy only, with 12 creatitis from 1995 and 2009 was undertaken with IRB rience and eventually be able to refi ne his technique and METHODS: All published studies between 1988 and octo- patients receiving both therapies. Twenty-fi ve per cent of approval. Surgical failure was defi ned as persistent post- reduce his complication rates. The aim of this study was ber 2010 were systematically reviewed. Results of com- the N0 group compared to 82% of the N1 group had adju- operative pancreatic pain requiring daily narcotic analgesia to evaluate the learning curve effect for pancreatic surgery, parative studies, comparing CP versus DP, were pooled vant therapy (p = 0.004). Overall survival and recurrence managed by a multidisciplinary surgeon-led team. Patient which can only be studied at a high-volume center. by standard meta-analytic techniques using the random free survival by N0 status and adjuvant therapy demon- demographics, index operation, BMI, gastrointestinal METHODS: effects model. Over period of ten years, outcome of pan- strated no difference (p = 0.664, p = 0.610). Survival of N1 symptoms, frequency of hospital readmission, daily mor- creatic operations performed by two “senior” pancreatic RESULTS: and adjuvant therapy was better for those receiving post- Eighty-nine studies with 939 cases of CP were phine equivalent requirements, tube feed requirement, and surgeons (SPS) and one specializing “junior” pancreatic sur- operative chemoradiotherapy (p = 0.035). Overall median recognized and included in the systematic review. Nine- TPN dependence. Quality of Life data was collected utiliz- geon (JPS) were evaluated relative to increasing experience. survival for N0 and N1 was 58 and 53 months. Median sur- hundred and nine open resection were performed, 30 ing SF-36v2 implemented by the Element System, a com- Three equally sized blocks of consecutive operations were vival stratifi ed by nodal status and adjuvant therapy was laparoscopic, 11 of whom were robotic assisted. Postopera- puterized clinical tracking system. analyzed for JPS versus SPS. Statistical testing was done 58 months for either group by N0 status. Those N1 patients tive morbidity rate was 39.37%; pancreatic fi stula rate was with SPSS Ver 17.0 at a signifi cance level of p = 0.05. receiving adjuvant therapy had a median survival of 78 30%. Endocrine and exocrine pancreatic insuffi ciency were Table 1: Quality of Life Scores RESULTS: From 2001 to 2010, n = 583 pancreatic opera- months, compared to the 21 months for those who did not reported in 4.5% and 8% of patients, respectively. Overall General Health 57.35 SD = 3.93tions were performed at our institution. Of these, n = 245 (p = 0.035). Median follow up was 27.5 months. mortality rate was 0.92%. Ten comparative studies, includ- ing 309 patients submitted to central pancreatectomy and Physical Functioning 33.19 SD = 11.80were performed by two SPS, n = 212 by the JPS and n = Social Functioning 36.05 SD = 12.24 Poster Abstracts 430 to distal pancreatectomy, were analyzed for meta- 126 by other surgeons. For the JPS, signifi cant postopera- Mental Health 42.91 SD = 5.66tive morbidity rate decreased signifi cantly (from 25% to

analysis. Surgical operation time, blood loss, length of stay, Physical Quality of Life 39.42 SD = 7.60 Monday morbidity and exocrine failure presented a signifi cant het- 9%, p = 0.022) with increasing case load to reach a level Psychological Quality of Life 42.74 SD = 8.14at the average SPS level (15%) after around 70 pancreatic erogeneity across studies, with the I-squared ranging from Role Limits (Physical) 33.54 SD = 8.09 49.4% to 93.8%; re-operation, endocrine failure and pan- Role Limits (Emotional) 35.39 SD = 14.77operations. This was due to a decreasing rate of reopera- creatic fi stula did not presented signifi cant heterogeneity. Bodily pain 41.33 SD = 9.76tions (from 21% to 12%, p = n.s.), postoperative bleeding Central pancreatectomy had a higher post-operative mor- Vitality/Energy 55.22 SD = 8.68(from 16% to 0% p = 0.001) as well as mortality (from 4% bidity (pooled RR = 1.56, 95% CI 1.09–2.21), with higher to 0%, p = n.s.). Decreasing complication rates were accom- incidence of pancreatic fi stula (pooled RR = 1.70, 95% CI panied by a rise in technically demanding procedures and 1.31–2.19) compared to distal pancreatectomy. However, RESULTS: 34 patients were indexed in the study. 51 oper- oncologic radicality, as demonstrated by an increasing rate the odds ratio for post-operative endocrine insuffi ciency ations were performed and the average time from opera- of portal venous resections (from 14% to 23%, p = n.s.) and was 0.26 (95% CI 0.16–0.41), revealing a statistically sig- tion was 5.1 years (SD = 2.94). 65% of the patients were laparoscopic or laparoscopically assisted procedures (from nifi cant benefi t to CP (p < 0.001). female. The average age of the patients was 41 years old 0% to 20%, p < 0.001). (SD = 10). The BMI was 24.5 (SD = 5.23). 29% underwent CONCLUSION: The odds ratio for exocrine failure was 0.59 (95% CI 0.33– With increasing experience, the pancreatic lateral pancreaticojejunostomy; 21% a distal pancreatec- 1.07), but it was not signifi cant (p = 0.084) because of the surgeon can minimize his complication rate while simultane- tomy, 20% underwent a Whipple procedure and 14% large heterogeneity among studies (I-squared 64.4%). ously increasing technically demanding procedures. The learn- Survival of node positive ampullary carcinoma by adjuvant therapy underwent sphincteroplasty. The other operations were a ing curve in this fi eld of surgery requires a relatively high case Beger procedure (4%), revision of gastrojejunostomy (4%), load even for the setting of a high-volume center and refl ects pancreatic neck resection (2%), pancreatic necrosectomy the importance of individual surgeon volume. This constitutes (2%), choledochojejunostomy (2%) and Frey (2%). Patient a strong argument for centralization of pancreatic surgery.

128 129 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1629 very low mortality rate at HVCs despite a major morbidity Mo1630 Mo1631 rate comparable to low- and medium-volume centers is an The Impact of Regionalization of Pancreaticoduo- intriguing fi nding that likely indicates higher case com- Early Experience with Minimally Invasive Surgical Pancreatic Enucleation:A Safe and Potentially denectomy for Pancreatic Cancer in North Carolina plexity with early complication recognition and manage- Pancreatic Débridement Underutilized Operation Since 2004 ment to “rescue” these patients. Nicholas J. Zyromski*, Michael G. House, Attila Nakeeb, Purvi Parikh*1, Henry A. Pitt2, Ashar Ata1, Ankesh Nigam1 1 Ryan Z. Swan*, David Sindram, Amanda Walters, John B. Martinie, Thomas J. Howard, C. Max Schmidt, Michael J. Leonardi, HenryDepartment of Surgery, Albany Medical Center, Albany, NY; David A. Iannitti A. Pitt, Keith D. Lillemoe 2Department of Surgery, Indiana University, Indianapolis, IN Surgery, Indiana University, Indianapolis, IN Surgery, Carolinas Medical Center, Charlotte, NC BACKGROUND: The most common management of small INTRODUCTION: Pancreaticoduodenectomy (PD) for BACKGROUND: Minimally invasive surgical approaches pancreatic cysts and neuroendocrine tumors is an extensive pancreatic cancer carries a signifi cant morbidity and mor- are increasingly applied to patients with necrotizing pan- parenchymal resection. Standard resections for benign and tality. Current evidence suggests that hospitals with a high creatitis. Early experience with these approaches at our borderline neoplasms of the pancreas, however, are asso- annual PD volume provide improved outcomes after PD high volume pancreatic referral center was reviewed. ciated with a signifi cant risk of early morbidity and long- term functional impairment. In comparison,enucleation for cancer and regionalization to specialty centers is advo- METHODS: With IRB approval, medical records of patients is a low-risk procedure that preserves healthy parenchyma cated. While better outcomes at high-volume centers undergoing minimally invasive surgical pancreatic debride- and pancreatic function. The aim of this study was to eval- (HVCs) are accepted, a rapid regionalization process could ment between 2007 and 2010 were reviewed. Data were uate the outcomes of pancreatic enucleations at a national also have detrimental effects. For example, HVCs must be collected for descriptive analysis. equipped to accommodate increased volume without com- level. RESULTS: Fourteen patients were approached with mini- promising outcomes. North Carolina (NC) is unique in that METHODS: The American College of Surgeons-National mally invasive surgical techniques: 3 retroperitoneal (VARD); there are multiple HVCs within the state to accommodate Quality Improvement Program (ACS-NSQIP) Participant 3 laparoscopic transabdominal; and 9 laparoscopic trans- regionalization, including the establishment of our HVC in Use File was queried for all patients undergoing a pancre- gastric. Pancreatitis etiologies were: biliary (n = 8), pan- late 2006. Upon noting an increase in referral for surgical atic enucleation (CPT code 48120) from 2005–2009. Patient creas divisum (2), alcohol, IPMN, and idiopathic (1 each). evaluation since 2007, we began investigation into region- demographics, preoperative variables, and post-operative Median time from initial pancreatitis to intervention was alization of PD and its effects upon outcomes during this morbidity and mortality were evaluated. Observed (O) and 10 weeks (range 6–32 weeks). Five patients had preopera- transition period within NC. expected (E) morbidity and mortality as well as indices tive infected necrosis; 5 had preoperative percutaneous METHODS: (O/E) were evaluated. An O/E ratio of less than 1.0 suggests The NC Hospital Based And Freestanding drains. Nine patients were admitted from home for elective that outcomes were better than expected. Ambulatory Surgery Facility Database was queried by ICD-9 débridement. Four patients (all laparoscopic transgastric) code for all PD performed in NC during two time periods: were converted to open operations. Four patients required RESULTS: Over the fi ve year period, 347 patients had a pan- 2004–06 and 2007–09. Hospitals were categorized by over- early (<30 day) reoperation: 3 for recurrent peripancre- creatic enucleation, which represents 3.1% of 11,026 pancre- all PD volume into three groups: Low (1–9 PD/yr), medium atic collections, 1 for cholecystectomy/jejunostomy. Two atic operations in the database. Fifty-seven percent of patients (10–19 PD/yr), and high (≥20 PD/yr) volume. Regionaliza- patients required late (>1 year) reoperation for recurrent were female. The mean age and median BMI were 54.8 years tion and operative mortality and major morbidity of PD pancreatitis and a disconnected left pancreatic remnant. and 27.9 kg/m2. Eighty-one percent of the enucleations were performed for pancreatic cancer were assessed by compar- Eleven patients had infected necrosis. In the 10 patients done for benign pancreatic diagnoses. After excision of a pan- ing volume groups across time periods. Statistical analysis completed using a minimally invasive technique, the creatic lesion, 9.5% of patients had sepsis, and 8.9% developed was performed using Chi-Square and Fisher’s Exact Test on median length of hospital stay was 5 days (range 1–103 organ space infection. Less than 3% of patients had any pul- SAS Software. days); 8 were discharged to home and 2 to nursing homes. monary complications and less than 2% had renal failure. No RESULTS: patients had any postoperative cardiac problems. The mean The number of PD for pancreatic cancer CONCLUSIONS: Good outcomes following minimally inva- length of stay was 8 days. Only two patients died postoper- increased by 91% (129 to 246 cases) at HVCs while decreas- sive pancreatic débridement can be achieved in select patients. atively. Overall morbidity and 30-day mortality were 19.8% ing at low-volume (62 to 58 cases) and medium-volume (80 Disease heterogeneity infl uences patient selection, and long- and 0.58%, respectively. Morbidity and mortality indices were to 46 cases) centers. The percentage of PD for pancreatic term follow-up is essential in these complex patients. cancer performed at HVCs increased signifi cantly (47.6% 1.21 and 0.58, respectively. to 70.3%), while decreasing for low- and medium-volume CONCLUSION: Pancreatic enucleation is an uncommon Poster Abstracts centers (p < 0.001) (Figure 1). Mortality was signifi cantly but very safe procedure that preserves pancreatic function. less at HVCs (2.8%) compared to low-volume centers For benign and premalignant pancreatic lesions, enucle- Monday (10.3%) for the 2007–09 timeframe (p = 0.038), and was not ation may be underutilized. different across periods for any group (Figure 2a). Mortality for all PD performed for pancreatic cancer in NC decreased from 6.6% to 4.6% across time periods (p = 0.31). Major morbidity was not signifi cantly different between volume groups within either time period; however, there was a sig- nifi cant increase in major morbidity at low-volume centers (p = 0.018) (Figure 2b). CONCLUSIONS: Regionalization of PD for pancreatic can- cer is occurring in NC, with a near doubling of PD per- formed at HVCs across these time periods. Mortality was signifi cantly lower at HVCs during the most recent period, and importantly, this rapid and substantial regionaliza- tion has only served to enhance outcomes at HVCs. The

130 131 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1632 Mo1633 Mo1634 severity of CD recurrence after ileo-colonic resection (IC) are unknown. Aim of this study was to compare 2 groups Does Stenting Decrease the Rate of Postoperative PancreaticComplications of Endoscopic Preoperative Biliary Predictive Clinical Factor for Clinically of CD patients undergoing either laparoscopic or open IC Fistula Following Pancreaticoduodenectomy? Drainage (PBD) for Pancreatic Cancer: A Tertiary Relevant Postoperative Pancreatic Fistula After to verify whether the surgical approach might infl uence Toshiyuki Moriya*, Clancy J. Clark, Yujiro Kirihara, Hospital Experience Pancreaticoduodenectomy the recurrence rate of the population of study. Patient’s Michael L. Kendrick, Kaye M. Reid Lombardo, Damien M. Tan*1, Sanjeev Singh1, Joshua A. Waters2, Kenichiro Uemura*, Yoshiaki Murakami, Takeshi Sudo, satisfaction with regard to the treatment received was also Florencia G. Que, Michael B. Farnell Stuart Sherman 1, Lee Mchenry1, Glen A. Lehman1, Yasushi Hashimoto, Akira Nakashima, Taijiro Sueda analysed. 80 consecutive patients undergoing elective IC General Surgery, Mayo Clinic, Rochester, MN by either laparoscopic approach (LAP = 30; 38%) or con- Evan L. Fogel 1, C. Max Schmidt2, Keith D. Lillemoe2, Surgery, Hiroshima University, Hiroshima, Japan ventional open surgery (OPEN = 50; 62%) were enrolled in BACKGROUND: Michael G. House 2 1 1 Pancreatic surgeons have employed , James L. Watkins, Gregory A. Cote BACKGROUND: Recent reports suggested that the drain a prospective longitudinal study. Study protocol includes 1 2 numerous interventions, medical and surgical, in an effort Gastroenterology, Indiana University, Indianapolis,Surgery, IN; amylase value in drain on the postoperative day one is the 3-years follow up (FU). Recurrence was investigated by CD to reduce the incidence of postoperative pancreatic fi stulaIndiana University, Indianapolis, IN useful predictive factor for postoperative pancreatic fi stula activity index and clinical examination every 6 months (POPF) following pancreaticoduodenectomy (PD). Stent- BACKGROUND: PBD for pancreatic cancer is widely (POPF) after pancreatic resection. However, the predictive and colonoscopy at 12 and 36 months in all patients. At 12 ing the pancreatic anastomosis has been thought to reduce practiced despite a recent randomized study demonstrat- clinical data for clinically relevant POPF after pancreatico- and 36 months patients were also asked to fi ll out a specifi c the incidence of POPF. In our experience, 0.9% (4/449) of ing increased morbidity in patients undergoing pancreato- duodenectomy have not been clearly established. Treatment Satisfaction Questionnaire. We have already patients have required endoscopic retrieval of the anasto- duodenectomy (PD) after PBD. The role for PBD remains METHODS: From Dec. 2003 to Oct. 2010, prospectively reported short term results and results of the 1st year of FU motic stent. Although uncommon, these late complica- unclear, particularly in those patients who present with collected data from 175 consecutive patients who under- in a previous comparative study that included 28 patients tions have led us to investigate the effi cacy of anastomotic deep jaundice. The goal of this study was to analyze the went pancreaticoduodenectomy with two-layered duct to from this series. In this study we have analysed the results stenting in the reduction of POPF. clinical characteristics and outcomes of patients with mucosa pancreaticogastrostomy were evaluated. The pre- from the 3rd year of FU. Fisher’s exact test and t-test were METHODS: Between January 1999 and September 2010, resectable pancreatic cancer who underwent endoscopic dictive clinical data (WBC, serum amylase, serum albumin, used to statistically compare qualitative and quantitative 553 patients underwent PD by a single surgeon (MBF). PBD, with particular emphasis on patients who present C-reactive protein, drain amylase, drain fl uid volume, etc.) data respectively. Internal trans-anastomotic stenting was routine from Janu- with deep jaundice. for clinically relevant POPF (ISGPF Grade B and C) were 62 patients completed 3-years follow up (FU) (RANGE 3–6 ary 1999 to May 2008 and, in the more recent years, a stent METHODS: A retrospective cohort study of patients who analyzed by logistic regression analysis. years) and of these, 60 (97%) underwent a colonoscopy was only used temporarily to facilitate construction of the underwent PBD via ERCP prior to PD for pancreatic can- RESULTS: Of 175 patients, 31 (18%) developed pancre- with ileoscopy. For 18 patients (23%) the 3-years FU is still pancreaticojejunostomy. We have retrospectively reviewed cer between 3/2008–5/2010 was performed. Patient char- atic fi stula by ISGPF criteria; Grade A in 21 patients (12%), in progress. Clinical recurrence at 3 years was 2 (9%) in the patient records and compared clinicopathologic factors acteristics at clinical presentation as well as PBD-specifi c Grade B in 8 (5%), and Grade C in 3 (2%). By univariate LAP and 3 (8%) in the OPEN group. Endoscopic recurrence between stent (n = 449) and no stent (n = 104) groups to and postoperative complications were examined. The inci- analysis, drain amylase on postoperative day (POD) 2, 3, 4, according to Ruttgers score was seen in 14 (63%) patients elucidate the effectiveness of trans-anastomotic stenting dence of complications among patients with or without 5, C-reactive protein POD 3 and 4 were found to be signifi - in the LAP group and in 27(69%) in the OPEN group. The on the incidence of POPF. deep jaundice (total bilirubin >14.6 mg/dL) at the time of cantly associated with clinically relevant POPF (p < 0.05). frequency, severity and pattern of recurrence did not differ RESULTS: The clinically relevant fi stula (ISGPS grade B/C) PBD was analyzed. Two-way statistical comparisons were By multivariate analysis, the independent predictive factor between the 2 groups (p = 0.6). Three patients (1 in the LAP rates in stent and no stent groups were similar, at 10% and performed. for clinically relevant POPF was C-reactive protein on POD group and 2 in the OPEN group; tot. 4%) underwent re- resection during the FU period and 1 patient in the OPEN 13% (p = 0.385), respectively. The clinically relevant fi stula RESULTS: Eighty-eight patients underwent PBD for 4 (p < 0.018). Based on the receiver operating characteristic rate in patients with a small pancreatic duct (≤3 mm, n = curve analysis, C-reactive protein >18 mg/dl on POD 4 dis- group underwent incisional hernia repair. No signifi cant potentially resectable pancreatic cancer. Sixty-two patients differences in terms of readmissions or need for medication 167) were also similar, at 17% (23/130) and 25% (9/37) (p = (70%) went on to operative exploration after 26 patients played the optimal sensitivity (64%) and specifi city (93%). 0.376), respectively. In the subgroup of patients with soft were seen during the FU period. However patients’ satisfac- (30%) refused surgery or were not medically fi t. Complete CONCLUSIONS: C-reactive protein >18 mg/dl on POD pancreatic gland (n = 64), grade B and C fi stulae were 31% tion was still signifi cantly in favour of laparoscopy also in operative resection was accomplished in 46 patients, 14 of 4 is the predictive factor for clinically relevant POPF after (13/41) for stent and 17% (4/23) for no stent groups (p = the long-term assessment. whom (30%) had deep jaundice prior to PBD. Median age pancreaticoduodenectomy when diagnose POPF by ISGPF 0.203), respectively. was 70 years (range, 48–87). PBD was successful in reliev- criteria. No differences were observed in terms of frequency, time- CONCLUSION: Internal trans-anastomotic pancreatic ing jaundice in 98% of patients. PBD-specifi c complica- of-onset and severity of recurrence in a 3-years follow up in duct stenting does not decrease the frequency or severity of tions, namely cholangitis, were recorded in 2 patients (6%), this consecutive series of patients undergoing laparoscopic postoperative pancreatic fi stulae. with no PBD-specifi c mortality. The majority of stents were Clinical: Small Bowel vs open IC. Hence long-term outcome for laparoscopy are

≥10Fr plastic (73.9%) or metallic (8.7%). comparable to standard procedure. However treatment’s Poster Abstracts satisfaction assessment showed a strongly signifi cant pref- The median time from PBD to operation was 4 weeks (range, Mo1635 erence of patients for the mini-invasive approach due to Monday 0–30). Operative complications were recorded in 56% of Recurrence and Long Term Results of Laparoscopic better cosmetic results, hospital experience and surgical patients: Clavien class I (23%), II (39%), III (19%), and IV treatment acceptance. Laparoscopic IC should be offered to (8%). There were 3 postoperative deaths. Median length of vs. Open Ileo-Colonic Resection in Crohn’s Disease: A Prospetive Longitudinal Study CD patients referred for surgery even though doesn’t seems postoperative stay was 8 days (range, 5–32). There were no to infl uence disease recurrence statistical differences in PBD-related (7.1% vs. 12.9%, p = Giuseppe S. Sica*1, Edoardo Iaculli1, Sara Di Carlo1, 1.0) or operative (50% vs. 58%, p = 0.74) complications Rosa Scaramuzzo 1, Livia Biancone2, Sara Onali2, Cristina Fiorani1, for patients with or without deep jaundice, respectively. Alessandro Sturiale 1, Achille Gaspari1 Among patients who did not undergo neoadjuvant therapy 1 2 (n = 39), the median time to surgery from PBD was similar General Surgery, University of Tor Vergata, Rome, Italy;Gastroenterology, for deep (18 days, range 7–43) and non-deep jaundice (27 University of Tor Vergata, Rome, Italy days, range 5–175) patients (p = 0.09). Modifi cations in host immune response have been dem- CONCLUSION: Patients with pancreatic cancer and onstrated in both the pathogenesis of Crohn’s disease obstructive jaundice often present with deep jaundice. PBD- (CD) and after laparoscopic resections for cancer. Rela- related complications are uncommon at an expert endo- tionships between surgical approaches and the rates and scopic referral center, irrespective of the depth of jaundice at presentation. The benefi ts of PBD prior to PD in patients who present with deep jaundice are not established, and the optimal timing of PD after PBD merits further study.

132 133 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Clinical: Stomach into neurologically impaired or elderly patients, the PEGs RESULTS: There were 1608 LAGB (27%), 3770 LRYGB Mo1638 that dislodge months and years later require expensive (64%), and 504 ORYGB (9%) surgeries. AIEs occurred in transportation, emergency room visits, surgical consulta- 5% of the overall sample and were most frequent dur- Comorbidities Remission After Roux-en-Y Gastric Mo1636 tions, and fl uoroscopic confi rmation of replacement. The ing ORYGB (7.3%), followed by LRYGB (5.5%) and LAGB Bypass for Morbid Obesity Is Sustained in a late removal complication, and its associated costs, are (3%), with no signifi cant difference between the ORYGB Long-Term Follow-Up Late Accidental Dislodgement of the Percutaneous overlooked and underestimated. These data also suggest a and LRYGB groups (p = 0.13). The most common AIEs were 1 *1 1 Endoscopic Gastrostomy: An Underestimated Burden Rafael M. Laurino Neto, Fernando A. Herbella, Renato M. Tauil, need for improvement in the design of the soft inner bum- organ injuries (1.0%), followed by anesthesia events (0.9%) 1 2 on Patients and the Healthcare System Fabricio S. Silva , Marco G. Patti per or a novel mechanism to secure a PEG in light of this and equipment failure (0.8%). The rate of composite end-1 Department of Surgery, Federal University of Sao Paulo, Sao Paulo, signifi cant burden to the healthcare system. point was 8.8% in the AIE group compared to 3.9% among Laura H. Rosenberger*, Timothy Newhook, Robert G. Sawyer, Brazil; 2Department of Surgery, University of Chicago, Chicago, IL Bruce D. Schirmer those without a AIE (p < 0.001). While incidence of death INTRODUCTION: General Surgery, University of Virginia Health System, Charlottesville, VAMo1637 (0.3%) and DVT/PE (0.4%) were similar (p > 0.05) across Roux-en-Y gastric bypass (RYGB) is those with or without an AIE, abdominal re-operation considered an effective therapy for weight loss although INTRODUCTION: Since its introduction in 1980, the Per- Prevalence of Adverse Intraoperative Events During (4.8% vs. 2.4%; p = 0.01), percutaneous drain placement weight regain may be observed in a long-term follow-up. cutaneous Endoscopic Gastrostomy (PEG) has become an Obesity Surgery and Their Sequelae (1.0% vs. 0.3%; p = 0.02) and endoscopic intervention Obesity-related comorbities are also well treated by RYGB effi cient means of providing long-term enteral access for Alexander J. Greenstein*4, Abdus S. Wahed2, Abidemi Adeniji2, (2.4% vs. 1.1%; p = 0.04) were more common among those due to weight loss and intestinal hormone changes. Few nutrition. Conveniently, the soft inner bumper allows PEG Anita P. Courcoulas 3, Gregory Dakin8, David Flum5, with an AIE. Multivariable analysis revealed that patients studies reported long-term status of comorbities especially removal with relatively minimal external traction. Conse- Vincent L. Harrison 1 7 1 with an AIE were at 90% greater risk of composite com- if weight regain is present. quently, a major complication is early accidental dislodge- , James E. Mitchell, Robert W. O’Rourke, John R. Pender 6 3 1 plication than those without an event (RR = 1.90, 95% CI: AIMS: This study aims to analyze: (1) the resolution of ment, from which signifi cant morbidity may occur. We , Ramesh Ramanathan, Bruce M. Wolfe 1 1.26–2.88; p = 0.002). obesity-related comorbities after RYGB in a long-term fol- have perhaps underestimated and underappreciated the Surgery, Oregon Health & Science University, Portland, OR; 2 CONCLUSION: low-up and (2) its relationship to weight regain. burden to the healthcare system due to this issue—not only Data Coordinating Center, University of Pittsburgh, Pittsburgh, There is no signifi cant difference in the 3 rate of AIEs in patients undergoing ORYGB versus LRYGB. in the acute setting but over the lifetime of the PEG. PA; Surgery, University of Pittsburgh Medical Center, Pittsburgh, METHODS: 143 patients (mean age 41, 18 male) were While associations between specifi c AIEs and post-opera- PA; 4Surgery, Mount Sinai Medical Center, New York,5 Surgery,NY; followed-up after RYGB for morbid obesity for at least 5 METHODS: A retrospective analysis of PEG placements tive complications could not be assessed due to the rarity of University of Washington, Seattle, WA; 6 years (mean follow-up 90, range 60–155 months). Mean was conducted at our institution, identifying all PEG tubes Surgery, East Carolina both events, the occurrence of an AIE is not insignifi cant. University, Greenville, NC; 7Neuropsychiatric Research Institute, body mass index before operation was 52 (range 39–82) placed between July 1, 2007 and July 1, 2010 by one faculty Patients with AIEs are at nearly double the risk of future 2 8 Kg/m . Diabetes, cardiopaty, arterial hypertension, dyslip- surgeon. Patient charts were reviewed for 30-day mortal-University of North Dakota, Fargo, ND; Surgery, Weill College of complication and thus merit close follow-up. idemia, sleep apnea, arthropathy, and infertility were pres- ity, major and minor complications, including early dis-Medicine, New York, NY ent in 26 (18%), 13 (9%), 89 (62%), 18 (12,5%), 85 (60%), lodgement, and subsequent management. Patients were BACKGROUND: Adverse intraoperative events (AIEs) dur- 88 (61,5%), 7 (8%) patients, respectively. reviewed until intentional removal of the PEG, cessation of ing surgery are a well known entity. A better understand- records, or patient mortality. RESULTS: Mean body mass index at last follow-up was ing of AIEs and their relationship with outcomes is helpful 33 (range 19–47) Kg/m2. Comorbidities status is depicted RESULTS: A total of 563 PEGs were placed during our for surgeon preparation and preoperative patient counsel- in Table 1. Comorbidities resolution was not related to defi ned time period. The 30-day mortality rate was 7.8% ing. The goals of this study are to test the hypotheses that the % of weight loss for cardiopathy (p = 0.7), hyperten- (44/563), 7-day early accidental dislodgement was 4.1% the laparoscopic approach to bariatric surgery results in sion (p = 0.3), dyslipidemia (p = 1), sleep apnea (p = 0.1), (23/563), and the total lifetime accidental PEG dislodge- fewer AIEs than the open approach and that patients who and infertility (p = 0.2) but it was related to arthropathy ment rate was 12.8% (72/563). Of those early dislodge- suffer an AIE are at greater risk of 30 day post-operative (p < 0.001). ments, 11 were replaced directly with a replacement complications. gastrostomy tube, 6 were replaced with a second endo- CONCLUSION: Our results show that comorbidities remis- METHODS: The study included 5,882 patients from the scopic PEG following several days of gastric decompression sion after RYGB is sustained in the majority of patients in Longitudinal Assessment of Bariatric Surgery (LABS) study and antimicrobials, 5 by an open gastrostomy, and 1 was a long-term follow-up. Weight regain is linked to worse undergoing one of three types of primary bariatric sur- not replaced at all. An additional 49 PEGs dislodged follow- results for arthropathy. geries between March 2005 and April 2009–laparoscopic ing discharge while at rehabilitation facilities and nursing adjusted gastric banding (LAGB), laparoscopic Roux-en-Y

homes. The vast majority required an emergency depart- Poster Abstracts gastric bypass (LRYGB), or open Roux-en-Y gastric bypass ment visit, level 1 surgical consultation, replacement gas- (ORYGB). AIEs included organ injuries, anesthesia related Table 1: trostomy tube, and radiographic confi rmation of position, Monday events, anastomotic revisions and equipment failure. Rates totaling an average of $3,535 in hospital charges. Diabetes Cardiopathy Hypertension Dyslipidemia Sleep apnea Arthropathy Infertility of AIE were compared between LRYGB and ORYGB groups resolved 77% 31% 57% 72% 50% 19% 28% CONCLUSION: Many large PEG reviews report an early using Fisher’s exact test. The relationship between AIEs and improved 23% 46% 38% 22% 46% 52% 28% accidental dislodgement rate between 0.6% and 4.0%. The a composite end point of 30-day major adverse complica- unchanged 0 15% 55% 0% 2% 25% 43% most clinically signifi cant accidental removals occur in tions (death, venous thromboembolism, percutaneous, the fi rst 7 days following placement, in which the stom- endoscopic, or operative reintervention and failure to be ach may fall away from the abdominal wall and open discharged from the hospital within 30 days from surgery) gastrostomy may cause obvious morbidity. Our early dis- was evaluated using a multivariable relative risk model lodgement rate (4.1%) is consistent with current reports. adjusting for DVT/PE, obstructive sleep apnea, body mass However, if followed longitudinally, a signifi cantly higher index, procedure type, and inability to walk 200 ft, as well rate of late dislodgement is seen (12.8%). Frequently placed as for clustering due to surgeon and site.

134 135 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Mo1639 CONCLUSIONS: There is a paucity of research on the CONCLUSIONS: WL after RYGB is affected by gender, with a history of H. pylori (18.8 vs 14.0 vs 16.7; p = 0.991); quality of EGD reporting in preoperative management of weight, and age suggesting patient specifi c or ‘host’ related operative time (184.8 m vs 202.7 m vs 200.1 m; p = 0.28); Esophagogastroduodenoscopy (EGD) Reporting gastric cancer. Experienced practitioners felt that the most factors which are independent of behavior or DC. Further or LOS (3.3 d vs 2.7 d vs 3.2 d; 0.81). There was a signifi cant for Preoperative Management of Gastric Cancer: important aspects of an EGD report for preoperative man- research to identify why some patients have poor WL after difference, however, in the number of preoperative comor- Evaluation of Quality agement of gastric cancer were location, size and descrip- RYBG could shed light on the causes of obesity itself. bidities with which the patients presented (non s/u- 3.96; Nikila C. Ravindran*1, Alyson L. Mahar3, Calvin H. Law2,3, tion of neoplasm. Evaluation of EGD reports by these s- 6.20; u- 5.47; p < 0.01). There was a trending difference Natalie G. Coburn 2,3, Jill M. Tinmouth2,3 criteria reveals need for standardization of EGD reporting Mo1641 between the preoperative waist circumference of those 1 to improve gastric cancer care. without strictures/ulcers and those with strictures (132.8 University of Calgary, Calgary, AB, Canada;2University of Toronto, Anastomotic Complications Following RYGB: Can cm vs 150.1 cm; p = 0.097). The average time of diagno- Toronto, ON, Canada; 3 Sunnybrook Health Science Centre, Toronto, ON, Ulcers/Strictures Be Predicted? sis after surgery was 2.86 months for strictures and 12.6 Canada Mo1640 Shushmita Ahmed*, Alex Taylor, Chhavi Bajaj, Dylan Gwaltney,months for ulcers (p < 0.01). Post operatively, there was a The Weight Loss Response to Roux-en-Y Gastric Bypass AIM: To identify important features of the EGD report for Kate Kiely, John M. Morton signifi cant difference in the percent of patients who had preoperative management of gastric cancer and to develop Is Host Mediated Surgery, Stanford University, Stanford, CA at least one complication (excluding strictures and ulcers) and validate a tool for the same. *1 2 3 among the three groups (non s/u-11%; s-10%; u-47%). Joseph A. Caruana, Shah Faisal, Katherine L. Hess, BACKGROUND: 4,5 1 1 Two of the most common complications There was a signifi cant difference in the percent of patients METHODS: There were 3 parts to the study. Part 1: WeScott V. Monte , Mark Cavaretta, Dang Tuan Pham 1 following Roux-en-Y gastric bypass (RYGB) are anastomotic who were readmitted after surgery (non s/u-7.9%; s-60%; conducted a systematic literature review of Medline, Surgery, Sisters of Charity Hospital and State University of New strictures and marginal ulcers. Our study aims to fi nd pre- u-69%) and were reoperated upon (non-s/u- 4.0%; s-22%; Embase and the Cochrane Databases using the search York at Buffalo, Buffalo, NY;2Medicine, Sisters of Charity Hospital dictors of stricture and marginal ulcer formation to better u-44%). Finally, there is a trending difference in 12 month terms “gastric,” “stomach” & “cancer,” “carcinoma,” “neo-and State University of New York at Buffalo, Buffalo, NY;3 Health prevent these complications. % excess weight loss among the three groups (non s/u- plasm,” or “tumour.” 2 independent evaluators reviewedInformation, D’Youville College, Buffalo, NY; 4Pharmacy, State METHODS: At a single academic institution, between 83.9%; s-97.8%; u-97%; p = 0.082). the abstracts; only those that addressed: “What are theUniversity of New York at Buffalo, Buffalo, NY; 5 CPL Associates, 2004 and 2010, we measured the preoperative and 12 CONCLUSION: important features of an EGD for the preoperative man-Buffalo, NY Patients with strictures and ulcers tend to agement of gastric cancer?” were retained. Part 2: A conve- month weights, laboratory values, preoperative comorbidi- have more preoperative comorbidities than patients with- nience sample comprising 5 gastroenterologists (GIs) and INTRODUCTION: Following Roux-en-Y gastric bypass ties and postoperative complications in 992 patients under- out strictures/ulcers. Patients with strictures or ulcers have a 5 general surgeons (GSxs) from 2 academic and 2 commu- (RYGB) suboptimal weight loss (SOWL), defi ned as <40% going RYGB. Of these, 20 patients developed anastomotic greater likelihood of readmittance and reoperation. Finally, nity hospitals was identifi ed. Semi-structured interviews excess weight loss (EWL) at 18 months, occurs in ~10% or strictures and 17 patients developed marginal ulcers. Lab patients with strictures or ulcers have a greater %EWL at 12 focused on important elements of an EGD report pertain- >15,000 patients yearly in the US alone. Re operative tech- values included platelet counts, C-reactive protein levels, months than patients without these complications. ing to gastric cancer and graded these on a 4 point Likert niques to accelerate weight loss (WL) in these “failures” hemoglobin A1C levels, as well as lipid and complete meta- scale. Part 3: Two abstractors separately examined 224 EGD have been disappointing. bolic panels. Data were analyzed using Students’ T tests and ANOVAs. reports at diagnosis from all patients diagnosed with gastric METHODS: To learn more about the WL response we adenocarcinoma (2005–2008) at a community hospital, an studied 2,427 consecutive patients with open RYGB from RESULTS: Among patients without strictures/ulcers (non academic hospital and a regional Cancer Centre for report 2000 to 2008. First, longitudinal patterns of WL were plot- s/u), patients with strictures (s), and patients with ulcers elements based on Part 2 results. Kappa statistic was used to ted from %EWL measured at follow up offi ce visits with (u), there was no signifi cant difference in: age at surgery compare interobserver reliability for each report element, determination of the effects of gender, body mass index (44.2 vs 44.0 vs 43.7, p = 0.98); preoperative BMI (46.7 vs overall report quality and adequacy for surgical planning. (BMI), and age. Secondly, 21 patients with SOWL were 47.0 vs 47.4, p = 0.91); preoperative CRP levels (10.3 mg/ RESULTS: Part 1: The literature review yielded 7117 compared retrospectively with 84 matched case-controls dl vs 13.2 mg/dl vs 8.8 mg/dl, p = 0.49); preoperative hA1C abstracts, none of which addressed quality of EGD in pre- having >40% EWL for demographic, anthropometric, levels (6.3% vs 6.1% vs 6.0%, p = 0.42); percent of patients operative planning for gastric cancer. Part 2: Study sample social, and operative factors as well as co-morbidities and consisted of 80% males who perform a median of 275 dietary compliance (DC). Finally, a prospective group with (80–1,000) EGDs/year. All respondents agreed that size and SOWL underwent aggressive dietary and behavior counsel- distance of gastric neoplasm from GEJ should be included ing over a 3 month period. Quantile regression techniques, in the EGD report. Additionally, 90% felt that appearance multivariate analysis, or paired t-test was used to analyze

of mass, video and photos were important. Tattooing of outcomes. Poster Abstracts a neoplasm was important if it was small (30%) or to be RESULTS: %EWL over time was signifi cantly less in males, treated laparoscopically (40%). All GSxs indicated they patients with preoperative BMI > 55, and those >42 years. Monday would repeat EGD themselves to confi rm location of the After controlling for those variables in patients with SOWL tumour for surgical planning, regardless of quality and con- only medication usage and DC were different. In the pro- tent of the EGD report. Part 3: Interobserver agreement was spective SOWL group, patients reported improved protein excellent (K > 0.7) in abstraction of tumour distance from intake, decreased hunger and stress, and better meal fre- GEJ, Siewart Type, tattooing and description of tumour quency yet had no signifi cant WL after the intervention. appearance. These were documented in 31 vs 33% (K 0.8 [0.7–0.9]), 0% (K 1) and 3% (K 1) of reports respectively. Ulceration was used as a descriptor for tumour appearance in 56 vs 57% (K 0.8 [0.7–0.9]). Agreement was fair in report adequacy for surgical planning (K 0.3 [0.2–0.4]) with 30 vs 33% of reports being inadequate.

136 137 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

TUESDAY, MAY 10, 2011 Tu1871 Tu1872 Colorectal Cancer Is Associated with Elevated PlasmaMesenchymal Stem Cells in a Scaffold: Do They 12:00 PM – 2:00 PM Levels of Soluble Galectin-3 Survive? *1,2 1 2 Hall A C. M. Shantha Kumara H*, Joon H. Jang, Sajith A. Herath, Massarat Zutshi , Levilester Salcedo, Anthony Calabro Daniel D. Kirchoff, Xiaohong Yan, Vesna Cekic, 1 Colorectal Surgery, Cleveland Clinic Foundation, Cleveland, OH; POSTER SESSION I Michael J. Grieco, Richard L. Whelan 2Biomedical Engineering, Cleveland Clinic Foundation, Cleveland, OH Surgery, St Luke Roosevelt Hospital, New York, NY (NON-CME) Mesenchymal stem cells (MSC) have been shown to aid tis- INTRODUCTION: Galectin-3 is a beta-galactoside-binding sue healing by virtue of the factors they secrete that affect protein expressed in many types of human cells including the surrounding cells. Direct or IV injection of these cells epithelial and immune cells. Gal-3 is synthesized in the is the most feasible way of administration, however, this is cytoplasm as a cytosolic protein but is found in a variety of not always possible especially when the target area is hol- Basic: Colon-Rectal any tumor within the fi rst 10 days and 1 with cecal tumor at 4 weeks. Of the 10 mice surviving to study completion, locations within the cell. Cytoplasmic Gal-3 functions as an low. Scaffolds using collagen,chondroitin, and hyaluron 8 harbored tumors within their cecal pouches (Figure 2) apoptosis inhibitor while nuclear Gal-3 has pro-apoptotic have been studied for tissue engineering in orthopedics The basis of this pilot study is to evaluate the progress of Tu1870 and 2 had generalized carcinomatosis secondary to perfo- activity. Loss of Gal-3 in the nucleus and its accumulation rated cecal pouches. Successful generation of orthotopic in the cytoplasm is commonly seen in a variety of human MSC in a scaffold. Our aim is to evaluate hydrogel as a scaf- A Novel Method to Generate Colon Cancer Orthotopictumor was seen in 85% (11/13) of mice. Liver metastases cancers including colorectal and prostate carcinoma; these fold and study if MSC survive, proliferate and migrate into Tumors in Mice: Implantation Using the Cecal Pouchwere detected in 3 mice without carcinomatosis (Figure 2a). changes also correlate with angiogenesis as well as tumor the surrounding. Technique Using histological analysis, all non-perforated cecal tumors invasion and progression. Gal-3 is believed to be a media- METHOD: Bone marrow from femurs of rats was har- *1,2 2 2 (8/8) involved entire bowel wall (mucosa, submucosa and tor of VEGF’s and bFGF’s pro-angiogenic effects. Gal-3 is vested and cultured according to our laboratory protocol. Carlos H. Chan , Anne-Laure Nouvion, Luisa Demarte, also found in the blood in a soluble form where it acts as a ValéRie Breton 2, Claire Turbide2, Nicole Beauchemin2, muscularis propria), but were still enclosed by the outer Cultured cells were passage every 3 days until passage 4 was pouch layer, representing T3 tumors. ligand. Increased blood levels of soluble Gal-3 (sGal-3) have reached. Cells were sorted by fl ow cytometry by positive Lorenzo E. Ferri 1 been noted in patients (pts) with breast, lung, and head 1 selection using intracellular adhesion molecule (ICAM-1) Department of Surgery, McGill University, Montreal, QC, Canada; and neck cancers but its levels in colorectal cancer (CRC) (PE+ve). A hydrogel pellet was created by mixing 300 ul 2Goodman Cancer Research Centre, McGill University, Montreal, QC, pts have not been well studied. This study’s purpose was to of hyaluronic acid and 8 ul of hydrogen peroxide (H202) Canada compare preoperative (PreOp) plasma sGal-3 levels in CRC which was used for crosslinking. One million MSC were and benign colon disease pts. BACKGROUND: The ectopic subcutaneous tumor model then suspended in a 1ml solution an introduced into the is the most commonly used pre-clinical model for cancer METHODS: Patients undergoing colorectal resection for hydrogel pellet.and plated using a basic stem cell media. drug development, but disregards the importance of pri- CRC or benign problems that had been prospectively Cells were maintained and observed every 2–3 days for 3 mary tumor microenvironment. Studies have shown that enrolled in an IRB approved tissue and data bank for weeks. At 3 weeks the pellets were digested using a 5% col- cancer progression and responses to chemotherapy are whom adequate Preop plasma samples were available were lagenase solution to free the MSC and resubmitted for sort- different between ectopic and orthotopic tumors. Cecal studied. Demographic, clinical, operative, and pathologic ing by fl ow cytometry. A separate passage 4 of MSC without microinjection of cancer cells is the standard orthotopic data were collected. The Gal-3 levels in the preoperative hydrogel was submitted for fl ow cytometry as controls. The colon tumor model, but major concerns have been cited plasma samples were determined via ELISA in duplicate percentage of positively sorted MSC cells from hydrogel Figure 1: Cecal pouch technique for tumor implantation: A) regarding reliability and technical diffi culties with this and reported as mean ± SD. The t-test was used to compare grown was then compared to the MSC without the hydro- Exteriorized cecum through midline abdominal incision; B) Creation of gel matrix. method. We here describe a novel and highly reliable tech- cecal pouch by invaginating the bowel wall; C) Closure of cecal pouch the results of the 2 groups (signifi cance p < 0.05). nique of developing orthotopic colon cancer in mice. by 3-0 silk tie. RESULTS: A total of 83 CRC (77% colon, 23% rectal) and METHODS: Using C57BL6 mice, midline abdominal inci- 71 benign disease pts (adenoma 39%, diverticulitis 61%) sions were made and small pouches were created at the tip were studied. The male:female ratio’s were similar but the of the cecums by invaginating the bowel wall (Figure 1b). CRC pts were older (p = 0.001). In the CRC group the stage Intact tumor fragments (approximately 2 mm × 2 mm × distribution was: stage 1, 22 (31%); stage 2, 27 (36%); stage 2 mm) derived from liver metastases of MC-38 cells (mouse 3, 24 (23%); stage 4, 10 (10%). The mean PreOp plasma colon cancer cell line) were enveloped within the cecal sGal-3 level was signifi cantly higher in the CRC group than pouches, which were then closed with 3-0 silk ties (Figure in the benign disease group (13.6 ± 7.8 ng/ml vs. 8.3 ± 3.4 1c). Mice were monitored over a 6-week period. At experi- ng/ml, p < 0.0001). No correlation was found between mental or clinical endpoints, the abdominal cavities were sGal-3 levels and tumor stage. examined for the presence of cecal tumors, gross liver and CONCLUSION: Figure 2: Mouse with a large tumor enclosed in the cecal pouch The mean Preop plasma sGal-3 level was lymph node metastasis, and carcinomatosis. The size of (black arrow) and gross liver metastases (white arrows) at 6-week 63.8% higher in the CRC vs. the benign disease group cecal tumors and number of liver surface metastatic nod- post-implantation. (p < 0.0001), which is notable. Although not proven, the ules were also determined. Tumors and metastatic nodules source of the added Gal-3 in the plasma of CRC pts may be were examined by routine histology. CONCLUSION: the tumor cells, stromal cells, and infl ammatory cells sur- Poster Abstracts RESULTS: The average operating time was approximately Our novel cecal pouch approach may rep- rounding the cancer. Higher sGal-3 levels may be related resent an effi cient method to generate clinically relevant

10 minutes per mouse using this novel cecal pouch tech- to neovascularization and infl ammation-induced tissue Tuesday nique. The 6-week survival was 77% (10/13); 2 died without orthotopic cecal tumors for quantitative analysis and pre- remodeling at tumor sites. Further studies with a larger clinical studies. population of healthy control and CRC pts are needed to better determine if there is a correlation between plasma sGal-3 levels and cancer stage or progression. sGal-3 holds some promise as a prognostic marker.

138 139 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

RESULTS: Sorted MSC from the bone marrow showed mutated adenomas when compared to conventional KRAS- The presented results indicate YB-1 as a novel and auspi- Basic: Pancreas 54.1% (494,567 of 914,167) MSC-PE (+) prior to hydrogel mutated adenomas (p = 0.006). The gene SYK, which codes cious therapeutic target for esophageal tumor suppression crosslink. One week later MSC were seen leaking through for the protein spleen tyrosine kinase, was upregulated 8.63 aimed at overcoming tumor aggressiveness and multi-drug the hydrogel into the culture media (see fi gure). These free fold in serrated KRAS-mutated adenomas when compared resistance. Future research activities should especially focus Tu1876 cells were sorted, which showed 18.8% (58,068 of 748,336) with the traditional adenoma group (p = 0.004). The gene on the molecular mechanisms of YB-1-mediated antipro- MSC-PE (+) cells. At three weeks, MSC that were freed from TNFRSF25, which encodes tumor necrosis factor receptor liferative effects and its interaction with different growth Albumin Administration in Acute Pancreatitis Increases the Hydrogel showed 13.7% (110,423 of 802,464) MSC- superfamily member 25, was upregulated 5.59 fold in the factor receptors. Lung and Pancreatic Damage Reversed by Nitric Oxide PE (+), in comparison with the control MSC that showed serrated KRAS-mutated group (p = 0.03). Synthase Inhibitor 37.2% (523,189 of 1,476,723). 1 1 1 CONCLUSIONS: Differential expression of mRNA may Basic: Hepatic Emilio E. Abdo, Ana Maria M. Coelho, Sandra N. Sampietre, CONCLUSION: The pilot study demonstrates that Hydro- distinguish serrated KRAS-mutated adenomas from con- Nilza A. Molan 1, Rosely A. Patzina2, José Eduardo M. Cunha1, gel can provide a scaffold and provide a temporary milieu ventional KRAS-mutated adenomas. These data provide Luiz Augusto C. D’Albuquerque 1, Marcel C. Machado*1 where MSC can survive and traverse into the surrounding evidence of distinct molecular genetic profi les for the two Tu1875 1 Gastroenterology, University of Sao Paulo, Sao Paulo, Brazil; areas. This scaffold can contain MSC in an area for a few adenoma types and may contribute to the identifi cation of 2 Ischemic Preconditioning-Like Effect of Pathology, University of São Paulo School of Medicine, Clinical weeks and thus is a feasible way of delivering cells. It can be specifi c genetic markers and therapeutic targets for their Hospital, Sao Paulo, Brazil a valuable in future cellular therapies that target areas that respective endpoint carcinomas. Polyunsaturated Fatty Acid-Rich Diet on Hepatic BACKGROUN/AIM: need to be fi lled. Ischemia/Reperfusion Injury Colloid resuscitation in acute pan- creatitis (AP) is a matter of controversy due to the possible Ana Maria M. Coelho*1, Hilton K. Takahashi2, Sandra N. Sampietre1, Basic: Esophageal deleterious effect on lung function. Previous study demon- Tu1873 José T. Stefano 1 1 2 , Luiz Augusto C. D’Albuquerque, Rui Curi, strated that albumin administration increases lung damage 1 Differential mRNA Expression Distinguishes Serrated Marcel C. Machado 1 in burns and that this effect can be reversed by inhibition KRAS-Mutated Adenomas from Conventional Tu1874 Gastroenterology, University of Sao Paulo, Sao Paulo, Brazil; of nitric oxide synthase (iNOS) .We hypothesized that albu- 2 KRAS-Mutated Adenomas of the Colon Y-Box Binding Protein-1-Mediated Antiproliferative Physiology and Biophysics, University of Sao Paulo, Sao Paulo, Brazilmin administration in AP may be deleterious not only to BACKGROUND/AIM: the lungs but also to the pancreas and that those effects can Michael R. Cassidy*1, Michael J. O’Brien2, Shi Yang2, Earl Gillespie1, Effects in Esophageal Cancer Most of the studies related to liver protection from ischemia/reperfusion injury are related to be reversed by inhibition of iNOS. The aim of this study Arthur F. Stucchi 1, James M. Becker1 Sabrina Thieltges*, Tatyana Kalinina, Yogesh K. Vashist, was to evaluate whether inhibition of iNOS reverses the 1 ischemic preconditioning. This effect is associated with Surgery, Boston University Medical Center, Boston, 2MA;Pathology, Emre F. Yekebas, Jakob R. Izbicki effect of albumin on lung and pancreatic damage in AP. Department of General-, Visceral- and Thoracic surgery, uncoupling of mitochondrial respiratory chain. Previous Boston University Medical Center, Boston, MA METHODS: Universitätsklinikum Hamburg-Eppendorf, Hamburg, Germany studies demonstrated that lipid infusion increased the pro- AP was induced in male Wistar rats by intra- BACKGROUND: While the BRAF-mutated serrated path- duction of uncoupling protein in hepatocytes. We hypoth- ductal 5% taurocholate injection. To evaluate the effect way to colorectal cancer has been well described in terms In a recent study we identifi ed the impact of the onco- esized that diets enriched with polyunsaturated fatty acid of albumin on lung damage in AP, animals received IV of clinicopathological and molecular characteristics of genic transcription factor Y-box binding protein-1 (YB-1) (PUFAs) may have a preconditioning effect on liver isch- saline (Group I) or human albumin (Group II) immediately precursor lesions and CpG island methylator phenotype in esophageal cancer and its infl uence on different protein emia/reperfusion injury. The aim of this study was to eval- after AP. To evaluate the effect of iNOS inhibition on lung (CIMP-High) endpoint carcinomas, an alternative serrated tyrosine kinase receptor expression. Expression of HER-2, uate the effects of PUFAs rich diet on hepatic ischemia/ damage in AP, iNOS specifi c inhibitor S-methylisothiourea pathway distinguished by KRAS mutations is less well Epidermal Growth Factor Receptor (EGFR) and Insulin- reperfusion injury (SMT) was given to animals immediately after AP. The ani- defi ned, in part because of overlap of its clinicopathological like-Growth-Factor 1 Receptor (IGF-1R) is associated with METHODS: Wistar male rats were fed a standard diet (SD) mals were divided into groups: Group III: saline was given and molecular characteristics with those of KRAS-mutated tumor growth and poor survival. Having demonstrated or a high-fat diet (HFD) enriched with Z-6 and Z-3 PUFA after AP and SMT, and Group IV: albumin was given after conventional adenomas and carcinomas. We postulated that YB-1 was simultaneously expressed with these growth or only Z-6 for 4 weeks and evaluated in pre and post- AP and SMT. After 12 hours serum amylase levels, lung that differential mRNA expression profi les could determine factor receptors in esophageal cancer, the aim of this study ischemia/reperfusion (I/R) conditions. In pre-I/R condition myeloperoxidase (MPO) activity, pulmonary vascular per- whether there were specifi c genetic differences between was to evaluate the effect of inhibition of YB-1. livers were collected for determination of fatty acid compo- meability, and histological analysis in pulmonary and pan- KRAS-mutated serrated adenomas and KRAS-mutated con- For this reason, three esophageal cancer cell lines were sta- sition, liver mitochondrial function, expression of peroxi- creatic tissue were determined. ventional adenomas. bly transfected with YB-1 specifi c siRNA. Effi cacy of YB-1 some proliferator-activated receptor alpha (PPAR-D) gene, RESULTS: Serum amylase levels, lung MPO activities, vas- METHODS: Laser capture microdissection was employed knock-down was assessed by quantitative reverse transcrip- malondialdehyde (MDA) content and histological analy- cular permeability and infl ammatory infi ltration, and pan- to selectively analyze the epithelial component of archived tion PCR and immunoblotting. Depletion of YB-1 lead to sis. Rats were submitted to partial liver ischemia during an creatic edema were signifi cantly increased after AP. hour. Four hours after liver reperfusion serum activities of formalin fi xed serrated and conventional adenomas known a decrease of HER-2 mRNA expression by >90% in com- Albumin administrated after AP increased lung permeabil- AST and ALT, serum levels of TNF-D, IL-6, IL-10, and PGE2, to exhibit KRAS mutations (n = 4 in each group). RNA was parison to control clones. YB-1 knockdown suppressed ity and infl ammatory infi ltration, pancreatic edema, and liver mitochondrial function, MDA content, and histology extracted and purifi ed using the Paradise Plus RNA extrac- mRNA expression of IGF-1R by 71 to 75%. In immunoblot serum levels of amylase compared to saline administration were evaluated. tion kit (Applied Biosystems). cDNA was created from the analysis, reduction of YB-1 expression resulted in a simulta- (p < 0.05). However, albumin administration with SMT RNA template and prepared for real-time PCR (RT-PCR) neous decrease of EGFR and HER-2 expression. Antiprolif- RESULTS: The HFD groups showed a signifi cant liver reduced lung permeability and infl ammatory infi ltration, using a cDNA synthesis and pre-amplifi cation kit (SA Bio- erative effects of YB-1 siRNA in esophageal cancer cells was mitochondrial uncoupling and increase in MDA content and pancreatic edema compared to albumin administra- sciences). A human cancer pathway fi nder PCR array (SA detected by MTT assay. Downregulation of YB-1 in esopha- compared to the SD groups. Four hours after I/R injury a tion without SMT (p < 0.05). There were no signifi cant dif- Biosciences), which probes for 84 cancer-related genes of geal cancer cell cultures affected antibody-induced cell reduction of liver damage with decrease in serum activities ferences in lung MPO activities among groups. interest, was used in RT-PCR. We compared mRNA expres- growth inhibition and sensibilized cells to drug treatment. of AST and ALT, in liver mitochondrial dysfunction and in sion for the 84 genes of interest among the adenoma Cisplatin treatment resulted in lowest proliferation rates. CONCLUSION: Restoration of extracellular fl uid in AP histological analysis, and a marked reduction on systemic Poster Abstracts with albumin increased the lung and pancreatic damage. groups. Data were analyzed using SA Biosciences PCR array In YB-1 knockdown cells, inhibition of cell proliferation to infl ammation were found in the HFD groups when com- Inhibition of iNOS before albumin administration reduced data analysis software. 50% (IC50) was achieved with 5 μM cisplatin in compari- pared to the group SD. Tuesday RESULTS: son to > 20 μM in control cells. Even the antiproliferative albumin induced damaging effects in AP. Of 84 cancer-related genes analyzed, three CONCLUSION: High-fat diets enriched with PUFAs have a genes were differentially expressed between conventional effects of herceptin are intensifi ed by use of YB-1-specifi c siRNA, showing up to 60% lower cell proliferation rates in preconditioning effect protecting the liver from ischemia/ KRAS-mutated adenomas and serrated KRAS-mutated reperfusion injury and should be object of future studies. adenomas. The gene IFNA1, which codes for the protein comparison to control cells. interferon D1, was upregulated 3.81 fold in serrated KRAS-

140 141 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1877 Tu1878 RESULTS: Using Kaplan Meier analysis, male sex strongly 5x10-6M). Responses to bethanechol (3x10-6M) were stud- correlated with decreased survival (p < 0.001). All males ied without and with E1a (10-6M), E2a (3x10-6M), E3a Heme Oxygenase-1 Gene Promoter Polymorphism andMale Sex Predicts Adverse Events and Decreased Survival had more metastatic events compared to females, 4.15 ± (10-8M), TTX, or propranolol. Electrical fi eld stimulation Acute Pancreatitis in Murine Pancreatic Adenocarcinoma 0.29 vs. 2.58 ± 0.28 (p = 0.0002), and adverse events, 40.3% (EFS; 30Hz) was applied to release neurotransmitters from Aiste Pupine*3, Barry J. Evans2, Jean-Marc Navenot2, Douglas Benson*1,2, Angela Sauaia1, Christopher Silliman1,3, vs. 17.9% (p = 0.0057). Adverse events were increased in enteric nerves. Dominant inhibitory effects of endog- Zi-Xuan Wang 2, Juozas Pundzius3, Giedrius Barauskas3, Peter K. Moore 1, Marguerite Kelher1,3, Carlton C. Barnett1,2 transfused males, 46.8% vs. 15.6% (p = 0.0007), and this enously released nitric oxide were blocked with l-nitro-l- Zilvinas Dambrauskas 3, Hwyda A. Arafat1,2 1 Surgery, University of Colorado at Denver HSC, Aurora,2Surgery, CO; difference was exaggerated with receipt of female blood, arginine (L-NNA; 10-4M) and the effect of E-agonists on 1 Surgery, Thomas Jefferson University, Philadelphia,2Pathology, PA; Denver Health Medical Center, Denver, CO; 3Bonfi ls Blood Center, 58.1% vs. 17.4% (p = 0.003). Moreover, transfused male EFS responses were studied without and with propranolol mice had reduced weight gain from baseline compared to or atropine (10-7M). Data is mean±SEM change of baseline Thomas Jefferson University, Philadelphia, PA; 3Surgery, LithuanianDenver, CO saline control, 3.88% ± 0.97 vs. 6.78% ± 0.58 (p = 0.03), this contractile activity [%] (negative value: inhibition). University of Health Sciences, Kaunas, Lithuania INTRODUCTION: Males with pancreatic cancer have been association was not seen in females. RESULTS: Beta-agonists caused a dose-dependent inhi- INTRODUCTION: shown to have more adverse events and decreased survival Acute pancreatitis is a severe and fre- CONCLUSION: Collectively, our analysis suggests that bition of spontaneous and stimulated contractile activ- compared to females. Moreover, perioperative blood trans- quently a life-threatening disease, which can lead to pan- male sex modifi es survival and the association of blood ity (all p < 0.01; ANOVA) independent of precontraction fusion appears to be linked to decreased survival in patients creatic necrosis, acute lung injury, SIRS and MODS. The transfusions and blood donor gender on adverse events in and blockade of the enteric nervous system by TTX (NS; with pancreatic cancer. Our lab has demonstrated, using inducible enzyme heme oxygenase-1 (HO-1) is an anti- pancreas cancer. These data demonstrate sex dimorphism ANOVA). Propranolol prevented this inhibition in large an immunocompetent murine model of pancreatic cancer, oxidative, anti-infl ammatory, and cytoprotective enzyme that might infl uence treatment options in patients. Fur- part (responses at maximum concentration of E-agonist: male mice have more metastatic events than female, which that is induced in response to cellular stress. The HO-1 pro- ther study to delineate the mechanism of these fi ndings is E1a: –60 ± 2 vs. –23 ± 5% (with propranolol); E2a: –97 is accentuated by receipt of blood product from female moter contains (GT)n dinucleotide repeats and is highly warranted. ± 1% vs. –27 ± 12%; E3a: –80 ± 5 vs. –13 ± 14%; all p < donors. Using this model, we hypothesize male mice have polymorphic in the population. The presence of longer 0.05; ANOVA). E1a increased bethanechol-induced excita- more adverse events and decreased survival, which is wors- repeats have been shown to be associated with lower levels tion more than 3-fold (1,073 ± 891 vs. 3,919 ± 891%; p < ened by blood transfusion. of HO-1 expression in vitro and is associated with many Basic: Small Bowel 0.05; ANOVA) while E2a, E3a, TTX, and propranolol had diseases in vivo. In this study, we hypothesized that the METHODS: 133 C57/BL6 mice, age 8–9 weeks, under- no effect on bethanechol-induced excitation (NS; ANOVA). number of GT repeats in HO-1 promoter can infl uence the went splenic inoculation of 2.5 × 105 PanO murine pan- 2 Tu1879 EFS with L-NNA caused excitation (263 ± 124%; p < 0.02; occurrence of acute pancreatitis due to its protective func- creatic adenocarcinoma cells. 58% were male. 80% were ANOVA) that was increased in presence of E2a (705 ± 241%; tion. Patients with acute pancreatitis are more likely to transfused at day 7 with 1 ml/kg supernatant from day-42 Role of Different Beta-Receptor Subtypes in Control 10-7M; p < 0.05; ANOVA). This effect of E2a was prevented have long repeats than controls. red cells from male (50%) or female donors (50%), while of Contractile Activity in Rat Jejunal Circular Muscleby propranolol and atropine (293 ± 156% and 65 ± 29%, METHODS: control mice received saline. Mice were followed clinically Acute pancreatitis (n = 131) patients and age- Brigitte Goetz*1,2, Bernhard Stoklas1,2, Petra Benhaqi1,2, respectively; both p < 0.05 vs. with E2a). E1a and E3a had and sex-matched healthy controls (n = 33) were studied. and weighed three times per week with planned necropsy no effect on EFS responses. Mario H. Mueller 1,2, Martin E. Kreis1,2, Michael S. Kasparek1,2 Peripheral blood samples from pancreatitis patients were at 5 weeks, gross metastatic events were recorded. Adverse 1 CONCLUSION: In jejunal circular muscle of naïve rats collected on admission. Genomic DNA was extracted from events were determined as >5 metastases, presence of bowel Department of Surgery, Ludwig-Maximilians-University Munich, 2 spontaneous and stimulated contractile activity can be obstruction, or death prior to study endpoint. Survival Munich, Germany;Walter Brendel Centre of Experimental Medicine, the blood samples of patient and control groups. The HO-1 inhibited via beta-1, 2, and 3-adrenergic mechanisms. This analysis was performed using Kaplan-Meier and logistic LMU Munich, Munich, Germany promoter region with the GT repeats was PCR amplifi ed inhibition is mediated by effects on receptors located on regression used to evaluate the interaction between mouse BACKGROUND: with fl uorescent tagged primers. The PCR products were Beta-receptors participate in control of the smooth muscle and not on the enteric nervous sys- sex and transfusion. analyzed by ABI 3130 genetic analyzer and the exact size gastrointestinal contractile activity and might be involved tem. Activation of beta-1 receptors increases the response of the PCR products was determined by GeneMapper soft- in pathophysiology of motility disorders such as postopera- to exogenous bethanechol while the endogenous release of ware. The short allele was defi ned to contain 27 GT repeats tive ileus. Our aim was to determine the mechanisms of acetylcholine can be increased by beta-2-adrenergic mech- or fewer. The long allele was more than 27 repeats. action of three beta-receptor subtypes in rat jejunal circular anisms. These potentially pro-contractile effects might be RESULTS: The subjects were categorized into 3 groups muscle. mediated by presynaptic beta-1 and 2 receptors on the based on the genotype results one short and one long alleles METHODS: Muscle strips (n = 8 per rat) were obtained enteric nervous system and need further investigations. (S/L), two short alleles (S/S) and two long alleles (L/L). The from 6 naïve, male Sprague Dawley rats and studied in organ DFG KA 2329/5-1 presence of S/L was similar between the patient group chambers. Dose-responses to exogenous beta-1- (E1a; xamoterol; (41.2%) and the controls (39.4%). Interestingly, 46.6% of 10-8-3x10-6M), beta-2- (E2a; fenoterol; 3x10-9-10-6M), and patients were carriers of two long repeats (L/L) vs 24.2% of beta-3-agonists (E3a; BRL37344; 10-10-3x10-8M) were control subjects, whereas 12.2% of patients were carriers of studied without and with precontraction with bethanechol two short repeats vs 36.4% of control population. (3x10-6M), tetrodotoxin (TTX; blocking enteric nervous CONCLUSION: Our data demonstrate a strong bias toward system; 10-6M), or propranolol (non-selective E-antagonist; longer alleles among patients with acute pancreatitis. Thus, polymorphism of the GT repeats in the HO-1 promoter region may be a risk factor for developing acute pancreati- tis. Further studies are now underway to analyze the pan- creatic levels of HO-1 protein in acute pancreatitis patients

and controls and to determine whether the presence of the Poster Abstracts short alleles facilitate HO-1 upregulation and consequently

promote its protective anti-infl ammatory function in acute Tuesday pancreatitis.

142 143 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1880 Tu1881 Basic: Stomach Tu1883 Mesenteric Afferent Nerve Sensitivity in the Small SPHK1 Regulates c-Myc Translation Through Mechanistic Role of p38 MAPK in Gastric Cancer Intestine During Mechanical Ileus in Mice Chk2-Dependent HuR Phosphorylation Tu1882 Dissemination in a Rodent Model Peritoneal Metastasis Mario H. Mueller*1,2, Xue Zhao2, Sarah Mittler2, Ping Jiang*2, Alexis D. Smith2, Ruiyun Li2, Lan Xiao2, Andrea Mencarelli1, Luigina Graziosi2, Barbara Renga1, Failure of Diabetes Remission After Roux-en-Y Gastric Michael S. Kasparek 1,2 1 Tingxi Yu 2 2 1,2 Chiara Santorelli 2 1 1 , Martin E. Kreis , Jian-Ying Wang, Douglas J. Turner Bypass , Giuseppe Palladino, Sabrina Cipriani, 1 Surgery, University of Munich, Munich, Germany;2Walter-Brendel 1 University of Maryland, Columbia, MD;2BVAMC and University Emanuel Cavazzoni 2, Annibale Donini2, Stefano Fiorucci*1 Institute, University of Munich, Munich, Germany of Maryland, Baltimore, MD John R. Pender*, Melissa Reed, William H. Chapman, 1 Medicina Clinica e Sperimentale, University of Perugia, Perugia, Italy; Walter J. Pories, G. L. Dohm INTRODUCTION: INTRODUCTION: 2Chirurgia Generale e D’Urgenza, University of Perugia, Perugia, Italy Mechanical ileus is a frequent disorder Intestinal epithelial barrier dysfunc- Surgery, East Carolina University, Greenville, NC seen in general surgery. While standard therapy is surgical tion results from a wide variety of pathologic conditions BACKGROUND: High levels of activated p38 MAPK are BACKGROUND: removal of the obstacle, little is known on alterations in such as infection, trauma, infl ammation and malignancy. There is an 80% improvement glycemic detected in poorly differentiated gastric cancers, associates gut physiology during intestinal dilatation. We aimed to At the gastrointestinal mucosal layer cells must be capable control after gastric bypass in diabetic patients. Why some with a poor clinical outcome and are thought to promote investigate afferent nerve sensitivity and leucocyte num- of maintaining their integrity, and do this through the diabetic patients do not respond to surgery has yet to be the local invasion and metastastic dissemination. Aims. In bers in the intestinal wall during mechanical ileus. interplay of multiple active processes including cell prolif- explained. this study we have evaluated the effects of a new p38 MAPK METHODS: C57Bl6 mice were anesthetized by isofl urane eration, migration, differentiation, and apoptosis. Previous METHODS: Seventeen obese Caucasian women (eight inhibitor, ML3403, on metastatic dissemination of gastric inhalation. After a mini-laparotomy, the small intestine reports from our lab have shown that Sphingosine-1- diabetic and nine euglycemic) were tested along with nine cancer cells and sensitization to chemotherapy. phophate (S1P) promotes intestinal epithelial barrier func- was ligated approximately 5 cm distal to the ligament of lean euglycemic control subjects. Tests were performed at METHODS: Peritoneal metastasis was induced, by intra- tion through regulation of these processes. Sphingosine Treitz, while controls received a mini-laparotomy only. three time points relative to surgery: 2–4 weeks before sur- peritoneal (I.P.) implantation of 1 × 107 of gastric cancer kinase-1 (SphK-1) critically regulates S1P production, as Intestinal afferent sensitivity and leucocyte infi ltration gery, the week of surgery, and 3 months post-surgery. At cells (MKN74 and MKN45) in 8-week-old NOD-SCID mice. it phosphorylates sphingosine to form S1P, which in turn into the intestinal wall were investigated in ileus animals each time point relative to surgery, an IV glucose tolerance In the fi rst set of experiments we have compared the peri- can function as a second messenger or be secreted extracel- or controls after 3, 9 and 24 hours in different subgroups test with minimal model analysis was performed, followed toneal metastasis of MKN45 and MKN74, that express lularly for autocrine and paracrine effects through signal- (each n = 6). A dilated segment of small intestine located by a 150 kcal Mixed Meal Tolerance Test (MMTT). There different levels of activated p38 MAPK. In a second set of ing via S1P receptors. In the current study we hypothesized 5 mm proximal to the ligature was prepared for multi-unit were 8 “responders” and 3 “non-responders” whose data experiments we have investigated whether ML3403 (13 mg/kg that SphK-1 overexpression would increase S1P produc- mesenteric afferent nerve and motility recordings in vitro. was subsequently compared. I.P. for 7 days) protects against peritoneal dissemination. In tion and lead to augmented cellular proliferation through Histological assessment by myeloperoxidase (MPO) stains RESULTS: addition, to invetsigate the role of p38 inhibition in cancer enhancement of c-Myc mRNA by Chk2-dependent HuR Insulin secretion in response to a meal chal- was performed on segments proximal and distal to the liga- cells, MKN45 cells were incubated with ML3403 (10 µM) for phosphorylation. lenge was markedly less in non-responders than in respond- ture in ileus animals. ers before surgery. 3 months post-surgery non-responders 16 hours before inoculation. Finally, mice inoculated with METHODS: RESULTS: Small intestinal motility in vitro was reduced SphK-1 overexpression stable cell lines were secreted as much insulin as responders. Post-surgery, GLP-1 vcvancer cells and treated with ML 3403 (6.5 mg/kg, I.P. in ileus segments compared to controls at all 3 time points selected in rat intestinal epithelial cells (IECs). SphK-1 secretion in response to a meal demonstrated no difference daily) were randomized to receive cisplatin (9.08 mg/kg; I.P. after induction of ileus (p < 0.05). Maximum afferent fi r- activity and S1P production were measured by radioac- between responders and non-responders. Insulin sensitiv- single dose) or 5-Fluorouracil (300 mg/kg, orally for 4 days) ing to serosal 5-HT (500 μM) peaked at 30 ± 3 imp sec-1 tive isotope assay. pGL3-Luc-c-Myc 3’UTR was generated ity (ISI) improved in responders but not in non-responders. alone or in combination with ML3403. Assessment of peri- to determine translational effi ciency of c-Myc 3’UTR by toneal diffusion was made 10 days later. 3h after the beginning of ileus compared to 10 ± 1 imp CONCLUSION: Pancreatic beta cells failure can not explain -1 reporter system. c-Myc protein synthesis was measured sec in controls (mean ± SEM; p < 0.001). No differences RESULTS: In the present study we have found that two by L-[35S] methionine and L-[35S] cysteine incorporation the results since insulin secretion is at least as robust in the in 5-HT sensitivity were observed between segments from gastric cell lines - the moderately differentiated MKN74 cell assays. The interaction of HuR and c-Myc mRNA was con- non-responders as it was in responders. Our data support ileus and control animals at the 9h and 24h time point. line and the poorly differentiated MKN45 cell line—pres- fi rmed by biotin labeled c-Myc 3’UTR pull-down assays. the hypothesis that “failure” of diabetes reversal is likely Ileus segments were less sensitive to bradykinin (0.5 μM) due to an inability to recover insulin sensitivity. ent different levels of activated p38 MAPK and that high compared to control segments at all time points. This dif- RESULTS: SphK-1 activity and S1P production was levels of expression of this kinase associates to an increased ference was 32 ± 4 imp sec-1 following ileus compared to 42 increased ~5 fold compared to control vector cells. Cell capacity to induce cancer dissemination in a rodent model ± 2 imp sec-1 in controls at 24h after the beginning of ileus cycle analysis showed that cell proliferation in SphK1 over- of peritoneal carcinomatosis. One important observation (p < 0.05). Continuous mechanical ramp distension of the expression cells was enhanced, as G1 to S phase transition we made was that administration of ML3403 caused a intestinal loop was followed by a pressure dependent rise in was increased versus vector cells. C-myc protein levels were robust reduction of peritoneal diffusion of MNK45 cells. afferent nerve discharge that was reduced in ileus segments increased by ~10 fold. This effect was due to increased Moreover pre-implant exposure to ML3403, caused a robust compared to controls (p < 0.05). MPO stains demonstrated translation, as there was no observed change in levels of attenuation of the metastatic potential of MNK45 cells. By a rise in leucocyte infi ltration in the intestine distal to the c-myc mRNA, and silencing c-myc mRNA reversed the gene array analysis we found that such a protective effect ligature which was 191 ± 32 cells per mm2 compared to effects. Overexpression of SphK-1 promoted the transloca- correlates with a robust downregulation in the expres- 3±0.6 cells in the dilated proximal intestine 24 h after the tion of HuR from nucleus to cytoplasm. HuR phosphoryla- sion of CXC chemokine Receptor-4 (CXCR4), Fms-related beginning of ileus (p < 0.01). tion was increased by ~12 fold, and this was accompanied tyrosine kinase 4 (FLT4), the non-receptor spleen tyrosine CONCLUSIONS: Afferent nerve sensitivity is altered sec- similarly by increase (~5 fold) in Chk2 phosphorylation, kinase (SYK) and the collagen D2 (IV) in neoplasic foci. ondary to mechanical ileus. Reduced afferent sensitivity for Chk2 silencing reduced this HuR phosphorylation. Finally the inhibition of p38 MAPK in vivo increased the bradykinin and painful distension stimuli may hint at acti- CONCLUSIONS: Our fi ndings demonstrate that SphK-1 sensitivity of tumor cells to cisplatin and associated with a robust downregulation in the expression of the multidrugs

vation of anti-nociceptive mechanisms during mechanical regulates cell proliferation by enhancing c-Myc translation Poster Abstracts ileus, while the pattern of leucocyte infi ltration is unclear in IECs, the enhanced c-Myc expression was modulated resistance (MDR)-1, a well defi ned marker of resistance to chemotherapy. Finally, P38 inhibition increases longterm

and warrants further investigation. though HuR phosphorylation by Chk2 and provided new Tuesday insight into the molecular functions of SphK1. survival (5 weeks). CONCLUSIONS: p38 MAPK inhibition is benefi cial in pre- venting the peritoneal dissemination of poorly differen- tiated gastric cancer cells by modulating attachment and diffusion of tumor cells in the peritoneum. In addition p38 inhibition increases susceptibility to chemotherapy.

144 145 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Clinical: Biliary Tu1624 Tu1625 OBJECTIVES: Our study was intended to evaluate retro- spectively outcomes of laparoscopic appendectomy (LA) Optimal Timing of Elective Laparoscopic Risk Factors of Development of Gangrenous versus open appendectomy (OA) in perforated and non- Tu1623 Cholecystectomy After Endoscopic Clearance Cholecystitis and Its Treatment Outcomes perforated appendicitis in adults (between age 18 to 65 of Choledocholithiasis in the Patients with Mehrdad Nikfarjam, Vachara Niumsawatt*, Arun H. Sethu, years old). Overuse of Computed Tomography in Patients Choledochocystolithiasis Vijayaragavan Muralidharan, Michael A. Fink, Graham Starkey,METHODS: Using the Nationwide Inpatient Sample data- with Complicated Gallstone Disease Robert Jones, Christopher Christophi Sang-Woo Cha*1, Seung Yeon Na1, Anna Kim1, Sae Hee Kim1, base (NIS), clinical data of adults who underwent LA and Jaime Benarroch-Gampel*, Casey A. Boyd, Kristin Sheffi eld, Surgery, University of Melbourne, Melbourne, VIC, Australia Hyang Ie Lee 1, Yun Jung Lee1, Hyeon Woong Yang1, OA for suspected acute appendicitis (perforated and non- Samantha H. Dallefeld, Courtney M. Townsend, Taylor S. Riall Sung Hee Jung 1 2 BACKGROUND: perforated) was evaluated from 2006 to 2008. Incidental Surgery, University of Texas Medical Branch, Galveston, TX , Joo Seung Park Gangrenous cholecystitis (GC) is con- 1 Division of Gastroenterology, Department of Internal Medicine, Eulji sidered a more severe form of acute cholecystitis. The risk and elective appendectomies were excluded. Outcome BACKGROUND: Ultrasound (US) is the preferred initial University College of Medicine, Eulji University Hospital, Daejeon, factors associated with this condition and its impact on measures included patient characteristics (age, sex and race), comorbidities, postoperative complications, length diagnostic image test for evaluating patients with suspected Republic of Korea;2 morbidity and mortality compared to non-gangrenous Department of Surgery, Eulji University College of of hospital stay (LOS), total hospital charges and in-hospital gallstone disease. Compared to US, computed tomography Medicine, Eulji University Hospital, Deajeon, Republic of Korea acute cholecystitis (NGAC) is poorly defi ned and based largely scans (CT) are more expensive, associated with signifi cant on older studies. mortality. radiation exposure, and have lower sensitivity, positive and BACKGROUND/AIM: Elective laparoscopic cholecystec- RESULTS: METHODS: Patients with histologically confi rmed acute A total of 573,244 adult underwent urgent negative predictive value for gallstone disease. tomy (LC) is common after endoscopic clearance of cho- cholecystitis treated in a specialized tertiary hospital unit appendectomy during these three years (67.31% of all METHODS: ledocholithiasis for patients with choledochocystolithiasis appendectomies). 45.9% of patients were female and This is a retrospective review of a prospec- accompanied by acute cholecystitis. However, the optimal between 2005 and 2010 were identifi ed from prospectively tively collected database of patients emergently admitted maintained database. Those with GC were then compared 69.5% Caucasian. The mean age was 37 years. In overall, timing in those patients remains uncertain. The aim of this 65.21% of all appendectomies were performed laparoscopi- with acute biliary disease to our institution between Janu- study was to identify whether the time interval between to those with NGAC. ary 2005 and May 2010. The use of CT and US imaging on cally. LA was performed more in female than male (67.64% two procedures can affect the course of LC in terms of con- RESULTS: 184 patients with NGAC and 106 patients with admission was described, and multivariate logistic regres- vs. 63.15% p < 0.01). The rate of perforated appendicitis version rate or complications. GC were identifi ed. The risk factors associated with GC sion was used to evaluate factors predicting receipt of CT. was lower in female (21.96% vs. 26.09%, p < 0.01). The PATIENTS/METHODS: included older age (P 0.001), diabetes (P 0.049), delay in RESULTS: Among the patients with cho- table shows outcomes of OA and LA in non-perforated and From January 2005 to May 2010, 562 consecu- ledochocystolithiasis accompanied by acute cholecystitis operation (P < 0.001), temperature of >38° C (P < 0.001), perforated appendicitis in adults. tive patients presented to the emergency department with from January 2005 to June 2010, those who underwent LC tachycardia (P 0.002), detection of muscle rigidity on complicated gallstone disease. Patients had a mean age of after endoscopic bile duct stone removal by endoscopic ret- examination (P 0.01), greater elevations in white-cell count Outcomes of Laparoscopic and Open Appendectomy in 45 + 20 years. Seventy percent of patients were female, 46% rograde cholangiopancreatography (ERCP) (n = 358) were (WCC) (P < 0.001), C-reactive protein (CRP) (P 0.001), bili- Non-Perforated and Perforated Appendicitis were white, and 41% were Hispanic. Sixty-fi ve percent of studied retrospectively, comparing the course of LC in four rubin (P 0.029) GGT (P < 0.001), and elevations in urea and patients had acute cholecystitis, 24% had gallstone pancre- time interval groups; LC ≤7 days, 8–14 days, 15–21 days creatinine. (P < 0.05). There were no overall differences in Non-perforated Perforated atitis, and 11% had common bile duct stones as the admit- and ≥22 days after ERCP. Primary outcomes are postopera- complications between the two groups. There was a lower LA OA P-value LA OA P-value ting diagnosis. A total of 493 patients underwent imaging incidence of common bile duct stones in the GC groupOverall Complication 4.60 7.59 <0.001 22.30 35.33 <0.001 tive complication rates, conversion rates into open surgery (%) during admission. Overall, 42% of patients (n = 234) under- and occurrence rates of preoperative biliary events. (13% versus 25%% P = 0.017). GC was associated with went CT. Approximately 25% of patients (n = 141) under- increased mortality (P 0.017), but this was not an indepen-Wound Infection (%) 0.15 0.42 <0.001 0.58 2.09 <0.001 RESULTS: Three hundred fi fty eight patients (mean age, Intra-abdominal 0.26 0.76 <0.001 1.65 3.57 <0.001 went both CT and US, while 47% (n = 259) and 16% (n = dent risk factor for mortality on multivariate analysis. Abscess (%) 93) respectively underwent US or CT only. In those who 60 years) were analyzed (group 1, n = 42; group 2, n = 112; CONCLUSION: Gangrenous cholecystitis has certain clini-Mean Lenght of 1.66 2.40 <0.001 4.04 5.99 <0.001 underwent both imaging procedures, CT was performed group 3, n = 105; group 4, n = 99). Conversion rates into Hospital Stay (days) cal features and associated laboratory fi ndings that may prior to ultrasound in 67% of patients. Patients undergo- open surgery, postoperative complications of LC and pre- Mean Hospital 22,830 20,849 <0.001 32,313 38,319 <0.001 help differentiate it from NGAC. It is not associated with ing imaging in the evening (7 pm–7 am) were nearly four operative biliary events were 5% (18 of 358), 3% (11 of Charges (\$) overall increased complications when treatment is taken inIn-hospital Mortality 0.03 0.05 <0.001 0.06 0.31 <0.001 times more likely to get a CT (OR 3.87, 95% CI 2.63–5.69). 358), and 2% (8 of 358) and there was no signifi cant differ- a specialized unit. (%) In addition, increasing age (by 5 years increments, OR 1.14, ence between four groups. LA: Laparoscopic Appendectomy OA: Open Appendectomy 95% CI 1.09–1.21) was associated with increased odds of Multivariate analysis showed both age older than 80 years receiving CT on admission. Weekend admission, gender, (95% confi dence interval, 1.03–17.29; P = .045) and history Clinical: Colon-Rectal race, previous ER visits, and diagnosis (acute cholecystitis of previous operation (95% confi dence interval, 1.82–14.27; CONCLUSION: vs. gallstone pancreatitis vs. common bile duct stones) did P = .002) to be independent risk factors for conversion to LA is safe and associated with lower mor- not predict use of CT scanning. open sugery. The endoscopic sphincterotomy may be a fac- Tu1626 bidity, lower mortality and shorter hospital stay with acute perforated and non-perforated appendicitis. Also, in per- CONCLUSIONS: Though US is the initial diagnostic image tor in the lower conversion rates (95% confi dence interval, Comparison of Outcomes of Laparoscopic Versus Openforated cases LA had an advantage over OA in hospital test and CT is of little additional benefi t, our study dem- 0.01–0.34; P = .003). Among the patients who underwent Appendectomy in Adults: Data from the Nationwidecharges. Laparoscopic appendectomy should be considered onstrates the overuse of CT has in the evaluation of acute, percutaneous transhepatic gallbladder drainage (PTGBD) Inpatient Sample (NIS), 2006–2008 the procedure of choice for perforated and non-perforated complicated gallstone disease. A large proportion of our had signifi cantly more postoperative complications of LC appendicitis in adults. sample (41%) underwent CT either alone or in conjunction (95% confi dence interval, 1.07–9.77; P = .008). There were Hossein Masoomi*, Steven Mills, Matthew O. Dolich, with US and evening imaging was the biggest predictor of no independent risk factors for the occurrence of preopera- Joseph C. Carmichael, Ninh T. Nguyen, Michael J. Stamos CT use. CT was frequently obtained overnight when US tive biliary events. Surgery, University of California, Irvine- Medical Center, Orange, CA Poster Abstracts was not available. The results of our multivariate analysis CONCLUSIONS: The time interval between endoscopic INTRODUCTION: Although laparoscopic appendectomy suggest that CT is performed not to clarify the diagnosis, clearance of choledocholithiasis and LC did not affect the is being performed with increased frequency, the utiliza- Tuesday but rather a surrogate for the ideal study. In patients with latter procedure in terms of complications or conversion tion of laparoscopy in the management of acute appen- strong clinical suspicion of complicated gallstone disease, to open surgery. However, the conversion rate was lower dicitis remains controversial and it continues to be used US should be encourage as the initial and only test for in patients who underwent endoscopic sphinctertomy and selectively. evaluation if no other complications are suspected. In case postoperative complication rate was higher in patients who of limited resources, surgeons and emergency physicians received PTGBD. should be able to perform good quality bedside right upper quadrant ultrasounds to avoid receipt of unnecessary addi- tional studies.

146 147 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1627 Tu1628 Post-Operative Post-Operative of surgery, and pathology fi ndings inconsistent with acute Complications Complications appendicitis. A total of 66 patients underwent open surgery, Survival Rates Following Radical Resection for Short Term Outcomes After Colorectal Surgery in In >80 In <80 P Value211 patients underwent multiport laparoscopic surgery, Persistent and Recurrent Anal Cancer Are Poor Octogenarians. Open surgery 28% 32% NSand 9 patients underwent LESS surgery. Data was collected George Roxin*1, Abdel Aziz Shaheem1, Anthony R. Maclean1, Pierpaolo Sileri, Giulio P. Angelucci*, Federico Perrone, LPS 25% 13% NSon postoperative narcotic consumption, postoperative hos- ASA • III 42% 37.5% NS Corinne M. Doll 2 1 Luana Franceschilli, Sara Lazzaro, Achille Gaspari pital length of stay (LOS), and readmissions related to the , William D. Buie Comorbility index v • 2 33% 22% NS 1 Department of General Surgery- Colorectal Surgery, University of Surgery, University of Rome Tor Vergata, Rome, Italy Gender M 27% F 28.5% M 30% F 22% NSoriginal procedure. Statistical signifi cance was defi ned as 2 p < 0.05. Calgary, Calgary, AB, Canada;Department of Oncology, University INTRODUCTION: Colorectal Surgery (CRS) for elderly BMI • 25 50% 33% NS Surgery duration • 180 min 54% 30% NSRESULTS: of Calgary, Calgary, AB, Canada patients represents a challenge. The aim of this study was Patients ranged in age from 24 to 52 years (mean Elective surgery 32% 23% NS= 34 years) in the LESS group and 18 to 70 years (mean = 18 PURPOSE: Sphincter-sparing chemoradiotherapy (CRT) is to assess the impact of age on short-term outcomes after Emergency surgery 22% 29% NSyears) in the open surgery group. The mean dose of intra- the standard of care for patients with cancer. CRS. Previous surgery 44% 26% NS venous narcotics consumed postoperatively was 2.0 mg of Despite good response rates, some patients require radical METHODS: LPS converted patients 50% 0% NS One-hundred consecutive octogenarians morphine (SD = 2.8) in the LESS surgery group, and 3.4 mg surgery (abdominoperineal resection or exenteration)for Malignant disease 34% 27.5% NS (group A) underwent CRS between 06/03 and 09/10. Data (SD = 7.4) in the open surgery group (p = 0.58). The mean persistent or recurrent disease. The purpose of this study Benign disease 20% 24% NS were prospectively collected and entered in a database. amount of oral narcotics (Percocet, Vicodin, or Tylenol #3) was to evaluate the outcomes of patients following radical Comorbidities were quantifi ed using the Charlson Comor- consumed postoperatively was 4.4 tablets (SD = 3.6 tablets) resection for persistent or recurrent disease. bidity Index and ASA classifi cation. Outcome measures in the LESS surgery group and 3.9 tablets (SD = 3.3) in the METHODS: All patients treated with CRT or RT for anal were postoperative complications (within 30 days after CONCLUSIONS: open surgery group (p = 0.62). Postoperative LOS was 36.1 cancer between 1990 and 2008 were identifi ed using the surgery) and 30-day mortality rates. These results were The relation between age and outcomes hours (SD = 16.6 hours) in the LESS surgery group and 33.9 regional cancer registry. Inclusion criteria: pathologically compared to a cohort of 100 patients <80 years (group B) after CRS is complex and may be confounded by differ- hours (SD = 26.3 hours) in the open surgery group (p = confi rmed squamous cell carcinoma of the anal canal, well-matched for ASA score, Charlson Comorbidity Index, ences in pre-existing comorbidities, different kind of dis- 0.81). 3 patients required readmission in the open surgery localized disease, receiving ≥45 Gy of RT. Statistical analysis and type of surgery. eases, urgency of surgery and type of treatment received. group; one patient developed a wound infection requiring was conducted to calculate survival rates and identify risk RESULTS: The mean age in group A was 85 years (range: According to our results there was no difference in the out- intravenous antibiotics, one patient developed an intraab- factors for poor outcomes. 80–104; 63 F; 37 M) and 55 years in group B (range 13–79; come in the two groups. Age is not correlated with postop- dominal abscess requiring percutaneous drainage, and one RESULTS: 105 patients (25 male) were identifi ed. Median 46 F; 54 M). Elective surgery was similar between the two erative complications and it is not an independent predictor patient developed fascial dehiscence secondary to a broken age was 57 years (range 33–87 yr). Median follow up was 40 groups (63% vs. 52%) as well as the choice of laparoscopic of morbidity and mortality in CRS. Octagenarians undergo- suture requiring reoperation. No patients required readmis- months (range 3–185). Median T stage and size at presen- approach (12% vs. 31%). Surgery for malignant disease was ing CRS have an acceptable perioperative morbidity and sion in the LESS surgery group (Z = –0.253, p = 0.80). tation were 2 and 35 mm (range 6–260 mm) respectively. 56% in group A and 58% in group B. The mean Charlson mortality rate and survival rate, and should not be denied Median dose to the primary tumor was 54 Gy (45–76 Gy) co-morbidity index rate was 0.6% in octogenarians and surgery based on age alone. Comorbidity index scores and Postoperative Variables at a median 28 fractions (range 20–37). 92 patients (88.5%) 0.3% in control group (p = 0.03). ASA > III was similar ASA scores are useful tools to identify poor risk patients. between the two groups (24% vs. 16%, NS). Overall com- Nevertheless this surgery in elderly patients should be per- LESS Group Open Group p-Value received concurrent chemotherapy. At 3 month follow-up Intravenous narcotics 2.0 (SD = 2.8) 3.4 (SD = 7.4) 0.58 85 (83.3%) patients had complete clinical response, one plication rate was 27%, being 28% for group A and 26% formed by experienced surgeons in specialized centers to (mg morphine) patient had insuffi cient follow-up. 19 (18.1%) patients had group B (NS). keep postoperative risk to a minimum. Oral narcotics (# tabs) 4.4 (SD = 3.6) 3.9 (SD = 3.3) 0.62 persistent disease, of whom 12 underwent radical resection Twenty-eight octegenarians patients between group expe- Postoperative LOS (hours) 36.1 (SD = 16.6) 33.9 (SD = 26.3) 0.81 for cure; 5 (41.7%) are alive at a median of 106 months rienced short-term complications after surgery, 4 (14%) Tu1629 from diagnosis, while the remaining 7 died at a median 15 of them requiring additional surgery; in this group 18% Equivalent Outcomes of Laparoendoscopic Single-SiteDISCUSSION: months from diagnosis. Recurrence occurred in 17 (20.0%) While LESS appendectomy for acute appen- had Charlson comorbidity index ≥2 and 36% were ASA (LESS) Surgery and Open Surgery for Appendectomydicitis has been demonstrated to be safe and feasible in patients at a median time of 9 months (4–53 m). 8 patients III or more. Twenty-six patients in group B experienced underwent radical resection for recurrence; 2 (25.0%) Koji Park*, Deva Boone, John Afthinos, James Mcginty, select patients, this study shows equivalent outcomes of short term complications, 5 of them (19%) requiring addi- both approaches in regards to postoperative narcotic con- patients are alive at 127 and 21 months, the remaining tional surgery; in this group 8% had Charlson comorbidity Ninan Koshy, Julio Teixeira 6 died at a median time from recurrence of 24.5 months. General Surgery, St. Luke’s Roosevelt Hospital Center, New York, NYsumption, postoperative LOS, and postoperative readmis- index ≥2 and 23% were classifi ed ASA III and above. No sions for procedure-related complications. While more Overall, disease specifi c and disease free survival, were 66%, differences were observed in terms of postoperative com- INTRODUCTION: 71% and 67% respectively for all anal canal carcinoma Laparoendoscopic single-site(LESS) patients who underwent open surgery were readmitted plications between emergency and elective surgery,and surgery for appendectomy has not been shown to have postoperatively, our modest sample size may have prohib- patients at 5 years. Multivariate analysis showed tumor laparoscopic versus open. size to be a signifi cant risk factor for disease free (aOR 1.02, any clear clinical benefi t compared to traditional multiport ited detection of statistical signifi cance. Further investiga- 95% CI: 1.01–1.04) and overall survival (aOR 1.03, 95% CI: Postoperative (30-day) mortality rate was 2% in octogenar- laparoscopic appendectomy. Likewise, studies indicate that tion with a larger sample size is necessary to clarify the 1.01–1.04) for all patients. ians vs 0% in control group (NS). The average length of multiport laparoscopic appendectomy and open appendec- short-term and long-term outcomes of LESS appendectomy hospital stay was 13 days (range: 5–41 days) in group A vs tomy have equivalent outcomes. In this study, we directly compared to open appendectomy. CONCLUSION: Despite good response rates to sphincter- 13 days (range: 4–31 days) in group B (NS). compare the outcomes of open and LESS surgery for appen- sparing therapy, 20% still required radical resection for per- dectomy in uncomplicated cases of acute appendicitis. sistent or recurrent disease. Survival rates following radical METHODS: resection are poor. We reviewed outcomes of all LESS and open appendectomies for acute appendicitis at a single institu- tion between January and December of 2008. Exclusion Poster Abstracts criteria included the following: perforated appendicitis documented on preoperative radiography or at the time Tuesday

148 149 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1630 Tu1631 Tu1632 Tu1633 Robotic-Assisted Proctectomy for Infl ammatory Gastrointestinal Cancer Surgery in Patients with a Colorectal Resection in Transplant Recipients: Is It Safe?Population-Based Assessment of Prognostic Factors Bowel Disease: A Case Matched Comparative Study Prior Ventriculoperitoneal Shunt Avraham Reshef*, Luca Stocchi, Pokala R. Kiran Associated with Neuroendocrine Tumors of the Colon of Laparoscopic and Robotic-Assisted Restorative andShilpi Wadhwa*1, Andrew Barina1, Katherine S. Virgo1, Colorectal surgery, Cleveland clinic, Cleveland, OH Ryaz Chagpar*, Yan Xing, Barry W. Feig, Yi-Ju Chiang, Completion Proctectomy 2 1 2 Edel M. Doorley , Anil Bahadursingh, Riccardo A. Audisio, INTRODUCTION: Major abdominal procedures in trans- Y. Nancy You, Janice N. Cormier 1 Surgical Oncology, MD Anderson Cancer Center, Houston, TX Aaron T. Miller, Julia R. Berian, Michele A. Rubin, Roger D.Frank Hurst, E. Johnson plant recipients are considered high-risk. The aim of this 1 2 Alessandro Fichera, Konstantin Umanskiy* Surgery, St Louis University, St Louis, MO;University of Liverpool, study is to evaluate the safety of colorectal resection in INTRODUCTION: The incidence of neuroendocrine University of Chicago, Chicago, IL Liverpool, United Kingdom solid organ transplant recipients. tumors (NETs) of the colon has been increasing over the BACKGROUND: METHODS: PURPOSE: Robotic-assisted proctectomy (RP) is increas- The estimated prevalence of hydro- Solid organ transplant recipients who under- past decade, although little is known about associated ingly being performed at select centers, however, data are cephalus in all age groups is between 1–1.5%. Placement went elective and urgent colorectal procedures from prognostic factors. We sought to determine clinicopatho- still lacking regarding the outcomes of RP in infl amma- of a ventriculoperitoneal (VP) shunt in such patients offers 1994–2010 were identifi ed from prospectively maintained logic factors infl uencing overall survival (OS). tory bowel disease (IBD). Rectal dissection in IBD patients them relatively normal lives. There is minimal data con- databases. Demographics, indications, surgical procedures, METHODS: The National Cancer Database (1998–2002) is often technically diffi cult due to persistent infl amma- cerning the risk of postoperative complications in shunted graft survival and perioperative outcomes were assessed. was used to identify patients diagnosed with colonic NETs. tion, resulting in distorted anatomical planes and friable patients undergoing subsequent major visceral operations. RESULTS: Out of 7,937 consecutive patients undergoing 5-year OS was determined using the methods of Kaplan- mesorectum. The purpose of this study was to determine We hypothesized that healthy adults who had VP shunts transplant since 1975, 90 patients who underwent colorec- Meier and a Cox proportional hazards model was used to whether the robotic-assisted approach, which provides placed for acquired conditions and later underwent gastric tal resection after heart (23), lung (17), kidney (29) and assess clinicopathologic factors associated with OS. technical advantages over laparoscopic proctectomy (LP), or colon cancer surgery would frequently have dense shunt- liver (21) transplant (63 men, 27 women, mean age: 57 ± RESULTS: related adhesions and high rates of adverse outcomes. A total of 2,188 patients were identifi ed with a including improved 3-dimensional vision, increased dex- 10 years) were identifi ed. The most frequent indication for median age of 62.0 years and a 5 yr OS of 56.8% (median METHODS: terity, and precise tissue handling, would be benefi cial in We assumed that all veterans were healthy surgery was diverticulitis (58%) followed by cancer (25%) follow up = 4.2 years). The majority of colonic NETS were IBD patients. on entry into military service. We searched national VA and IBD (15%). The mean interval time from transplanta- localized to the cecum (41.1%), followed by the sigmoid METHODS: 17 consecutive RP (10 restorative with ileal databases from 1994–2003 to identify all VA patients with tion to surgery was 6.3 years (ranges between 1 week and 33 (16.3%), ascending colon (11.7%), rectosigmoid (11.6%), pouch anal anastomosis (IPAA), 7 completion proctecto- shunts for acquired conditions and a curative-intent opera- years). Elective procedures were performed in 46 patients transverse (6.4%) and descending colon (1.7%). Distal colonic mies) performed using da Vinci S system between 2009 and tion for stomach or colon cancer. We conducted a chart (sigmoidectomy in 17, right colectomy in 11, subtotal col- NETs (sigmoid or rectosigmoid) had the most favorable 2010, were case-matched to LP (10 IPAA, 7 completion) by review to determine their clinical courses. ectomy in 9, proctectomy in 8) and urgent procedures were prognosis (69.1%) compared to proximal NETS located performed in 44 (Hartmann’s procedure in 29, sigmoid- age, gender, diagnosis (UC vs. Crohn’s disease), body-mass RESULTS: Five patients had codes for VP shunt, gastric in the cecum (57.3%), ascending (39.3%), and transverse/ ectomy in 9 and right colectomy in 2). Only 22 patients index (BMI), type of proctectomy, and preoperative use of cancer, and gastrectomy; 3 met our inclusion criteria. descending colon (31.1%) (p < 0.001). Distal colonic NETs (24%) underwent primary anastomosis without diverting anti-TNF medications. Retrospective analysis compared Fourteen had codes for VP shunt, colon cancer, and colec- were also more likely to present in younger patients with ileostomy. Six patients (7%) were left with a permanent peri- and post-operative outcomes. A subset of patients tomy; 4 met our criteria. One of the evaluable gastrectomy tumors that were of low histologic grade, ≤2 cm in size, and stoma. When compared to elective surgeries, urgent proce- receiving IPAA was evaluated for postoperative functional patients had dense shunt-related adhesions. There were no confi ned to the muscularis propria. On multivariate analy- dures had signifi cantly increased post-operative mortality outcomes and quality of life (QOL). post-operative complications in any of the 7 patients. VP sis, age (>75 years), high histologic grade, positive nodal (18% vs. 0%, p = 0.002), longer length of stay (16.5 ± 13 vs. RESULTS: shunts were managed by administering prophylactic anti- ratio (≥0.3), tumor depth, and distant metastases were all Operative time was longer in the RP group 8.5 ± 5 days, p < 0.001) and non-signifi cantly higher over- biotics preoperatively and isolating the shunt with surgical independently associated with poorer OS. (p = .01), while estimated blood loss was similar (344 mL in all morbidity (40% vs. 28%, p = 0.27). Kidney transplant sponges intra-operatively. CONCLUSION: RP and 189 mL in LP, p = .07). Even though time to return recipients were the only group without any postoperative Distal colonic NETS are associated with of bowel function was similar, 3.4 days in RP and 2.2 days CONCLUSIONS: We believe this is the fi rst report analyz- deaths (p = 0.05). All patients except one (kidney) retained improved 5-year OS compared to more proximally located in LP (p = .06), the length of hospital stay was longer in RP ing the clinical course of adults with VP shunts who later their graft function. tumors, likely due to earlier presentation of disease. Tumor vs. LP (7.7 vs. 5.3 days, p = .02). There was no difference have major abdominal cancer surgery. The presence of a location itself, however, was not independently associated in post-operative complications between the two groups, shunt was associated with dense adhesions in one of the Short Term (30 Days) Outcomes with OS after adjusting for other clinicopathologic factors. including anastomotic leak, wound infection, postopera- 7 patients in this series (14%) but not with increased risk of tive ileus, and bowel obstruction. Patients receiving IPAA (6 post-operative complications. Elective Urgent P RP, 8 LP) reported similar functional outcomes with regard Number 46 44 to mean number of daily bowel movements (6 RP vs. 7.3 Age 59.5 ± 10 54.3 ± 10 0.01 LP). There was a trend towards decreased frequency of peri- Mortality 0 8 (18%) 0.002 anal pruritis in the RP group compared to LP. The IPAA Length of stay 8.5 ± 5 16.5 ± 13 <0.001 groups reported similar improvement in QOL, satisfaction overall morbidity 13 (28%) 17 (40%) 0.3 Leakage 2 (4%) 1 (2%) 0.6 with procedure and recommendation of the procedure to Wound infection 4 (8%) 3 (6%) 0.7 others. Abscess 2 (4%) 2 (4%) 1 CONCLUSION: This study reports the results of our initial DVT 2 (4%) 3 (6%) 1 series of consecutive RP compared to LP in IBD patients. While RP operative times and length of hospital stay were CONCLUSIONS: longer than LP, short-term outcomes including complica- Elective colorectal procedures in trans- Poster Abstracts tions were similar between the groups. In a subset of RP plant recipients are safe and often associated with proxi- patients undergoing IPAA, pouch functional outcomes mal diversion. Urgent surgery is associated with substantial Tuesday were similar or slightly better compared to LP IPAA group. mortality except in kidney transplant recipients. As experience accumulates we expect RP operative times and length of stay to approach those of LP. Larger stud- ies will be required to further evaluate the role of RP for patients with IBD.

150 151 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1634 RESULTS: Of 150 patients (41% females, mean age 64 Tu1635 pooled proportion of patients achieving fi stula closure in years and 59% males, mean age 65 years) undergoing radi- non-Crohns fi stula-in-ano patients was 0.56 (95% CI 0.52- Does Tumor Location in Colon and Rectum Correlatecal resection for T1 colorectal cancer, 18 (12%) had nodal A Systematic Review of the Anal Fistula Plug for 0.60). The proportion achieving closure in Crohns patients with the Risk of Nodal Metastasis in T1 Colorectal metastasis. The risk of nodal metastasis at individual tumor Patients with Crohns and Non-Crohns Related was similar (0.56, 95% CI 0.40–0.72). Cancer? Fistula-In-Ano sites was 20.7% upper rectum, 18.2% lower rectum, 18.2% CONCLUSIONS: This systematic review suggests that fi s- Supakij Khomvilai*1, Pokala R. Kiran1, Madhusudhan R. Sanaka2, ascending colon, 12.5% sigmoid, 10.5% mid rectum, 4.3% James O’Riordan*, Indraneel Datta, Nancy N. Baxter tula closure is achieved using the AFP in close to 60% of Ian C. Lavery 1 cecum and 0% for hepatic fl exure, transverse, splenic Surgery, St Michael’s Hospital, Toronto, ON, Canada patients without Crohns disease, a lower success rate than 1 fl exure and descending colon. Age of patients (P = 0.35), Colorectal Surgery, Digestive Disease Institute, Cleveland, OH; INTRODUCTION: initially reported with this technique. Although patients gender (P = 0.82), size of lesion (P = 0.58), tumor differenti- The management of fi stula-in-ano is 2Gastroenterology and Hepatology, Digestive Disease Institute, a balance between fi stula cure and preservation of conti- with Crohns disease appear to have similar success with ation (P = 0.33) and lymphovascular invasion (P = 0.1) were the AFP, only 39 patients were evaluated in studies meet- Cleveland, OH not signifi cantly associated with nodal metastasis.T1 rectal nence. The anal fi stula plug (AFP) is an increasingly used PURPOSE: method to treat fi stula-in-ano. Comparison of studies on ing our inclusion criteria; the AFP has not been adequately For T1 colorectal cancer resected endoscopi- cancer had higher risk of nodal metastasis (16.9%) com- evaluated in this population. cally, the risk of lymph nodal involvement impacts further pared to colon cancer (8.8%) although this did not reach the AFP is limited by diverse fi stula types, heterogeneity management. Whether this risk may vary depending upon statistical signifi cance (P = 0.14, OR = 2.12). of fi stula characteristics and limited follow-up. The aim of this systematic review was to summarize the AFP literature the location in the colon and rectum has been poorly char- CONCLUSION: The risk of lymph node metastasis var- Clinical: Esophageal acterized. We evaluate the risk of lymph node metastasis in for patients with Crohns and non-Crohns related fi stula- ies depending upon location of the T1 colorectal cancer. in-ano in as homogenous a group of patients as possible. T1 colorectal cancer depending upon the location of the These fi ndings support the incorporation of the location of METHODS: Tu1636 primary tumor and evaluate factors that may predict the the primary into the management algorhithm when faced Pubmed, Medline, Embase and the Cochrane presence of metastasis in each part of colon and rectum. with the decision whether to offer radical surgery or colo- medical databases were searched from the period 1995 to High Resolution Manometry Findings in Patients with METHOD: Data of all patients who underwent radical noscopic surveillance for patients with T1 cancers detected 2010 as well as abstracts from The American Society of Esophageal Epiphrenic Diverticula Colon and Rectal Surgeons (ASCRS), The Society for Sur- resection for T1 colorectal cancer from January 1997–March at polypectomy. 1 *1 1 gery of the Alimentary Tract (SSAT), The European Society Fernando P. Vicentine, Fernando A. Herbella, Luciana C. Silva, 2010 were evaluated. Patient and tumor factors (location, 2 1 of Coloproctology (ESCP) and The Association of Colo-Marco G. Patti , Jose C. Del Grande grade, presence of lymphovascular invasion) associated 1 Department of Surgery, Federal University of Sao Paulo, Sao Paulo, with the fi nding of lymph nodal involvement for patients proctology of Great Britain and Ireland (ACPGBI) meetings Brazil; 2 with a radical resection specimen were evaluated. between 2007 and 2010. Case reports, letters, comments, Department of Surgery, University of Chicago, Chicago, IL conference proceedings and non-English language articles INTRODUCTION: The pathophysiology of esophageal were excluded. Patients with rectovaginal, anovaginal, rec- epiphrenic diverticula is still uncertain even though a Risk Factor Associated with Nodal Metastasis for T1 Colorectal Cancer tourethral or ileal-pouch vaginal fi stulas were also excluded concomitant motility disorder is found in the majority of from the review as were studies where the mean or median patients in different series. High resolution manometry Lymph node positive follow-up was less than 3 months. Studies were only may allow detection of motor abnormalities in a higher Variable Overall N = 150 (total = 18 (12%)) P-value OR (95% CI) included if results for patients with and without Crohns number of patients with esophageal epiphrenic diverticula Age (average) 64.92 ± 13.01 62.22 ± 13.01 0.35 0.92 (0.76 ; 1.10) disease could be differentiated. compared to conventional manometry. Age <65 69 (46.0%) 10 (14.5%) 0.39 0.65 (0.24,1.74) >65 81 (54.0%) 8 (9.9%) AIMS: This study aims to evaluate the high resolution Gender Female 62 (41.3%) 7 (11.3%) 0.82 1.12 (0.41,3.08) manometry fi ndings in patients with esophageal epiphrenic Male 88 (58.7%) 11 (12.5%) diverticula. TumorSite Cecum 23 (15.3%) 1 (4.3%) 0.78 METHODS: Ascending 11 (7.3%) 2 (18.2%) Nine individuals (mean age 63 ± 10 years, Hepatic Flexure 6 (4.0%) 0 (0%) 4 females) with esophageal epiphrenic diverticula were Transverse 2 (1.3%) 0 (0%) studied. All patients underwent an upper digestive endos- Splenic Flexure 3 (2.0%) 0 (0%) copy and barium esophagram. A single diverticulum was Descending 6 (4.0%) 0 (0%) observed in 8 patients and multiple diverticula in 1. All Sigmoid 40 (26.7%) 5 (12.5%) patients underwent a high resolution manometry. Pres- UpperRectum 29 (19.3%) 6 (20.7%) sure and relaxation of the upper and lower sphincters were Success by Study in non-Crohns Patients MidRectum 19 (12.7%) 2 (10.5%) recorded. Esophageal body function was measured at 3 and LowRectum 11 (7.3%) 2 (18.2%) RESULTS: From the initial search results, 71 articles or 7 cm above the upper border of the LES. Visual analysis of Tumor Differentiation Well 28 (18.7%) 1 (3.6%) 0.33 Moderate 103 (68.7%) 15 (14.6%) abstracts were identifi ed from the title as being of relevance color pressure plots for identifi cation of segmental abnormali- Poorly 19 (12.7%) 2 (10.5%) to the review. Two researchers (J.M.O’R and N.D.) then ties was performed by 2 researchers experienced in HRM. Lymphovascular Negative 140 (93.3%) 15 (10.7%) 0.1 3.57 (0.83,15.30) independently read the selected abstracts and selected the RESULTS: Upper esophageal sphincter was normal in all invasion Postive 10 (6.7%) 3 (30.0%) studies which matched the inclusion criteria. Any discor- patients. Esophageal body was abnormal in 8 patients, Size (average) 2.43 ± 1.88 2.66 +/- 1.74 0.58 1.07 (0.85 ; 1.35) dance was resolved by discussion (N.B.) so that a fi nal total lower esophageal sphincter was abnormal in 4 patients. Size <2 cm. 78 (52.0%) 7 (9.0%) 0.24 1.83 (0.67,5.01) of 18 studies (2 abstracts, 16 papers) were included in the >2 cm. 72 (48.0%) 11 (15.3%) Named esophageal motility disorders were found in 7 review. Study sample size ranged from 4 to 60 patients; a patients: achalasia in 6, diffuse esophageal spasm in 1. In total of 487 patients were included in all studies (448 non- 1 patient, a segmental hypercontractile zone was noticed Poster Abstracts crohns and 39 crohns patients). The plug was extruded in with pressure of 196 mmHg. the perioperative period in 36 patients (7.4%). The pro- Tuesday portion of patients achieving fi stula closure varied widely CONCLUSION: High resolution manometry demonstrated between studies for non-Crohns patients, ranging from 0.2 motor abnormalities in all patients with esophageal epi- (95% CI 0.04 -0.48) to 0.86 (95% CI 0.64-0.97) (Figure). The phrenic diverticula.

152 153 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1637 Tu1639 Tu1640 Tu1641 Is the Number of Lymph Nodes Removed and the Results of the Introduction of a Minimally Invasive The Outcomes of Curative Treatment for Advanced Long Term Follow Up of Type of Resection Associated with Postoperative Esophagectomy Program in a Tertiary Referral Center Carcinoma of the Cervical Esophagus Luis Durand*, Roberto De AntóN, Miguel Caracoche, Complications After Esophagectomy for Esophageal Rachel L. Blom*1, Markus W. Hollmann2, Jean H. Klinkenbijl1, Hiroshi Sato* Mariano Gimenez, Pedro Ferraina Cancer? Thomas M. Van Gulik 1, Olivier R. Busch1, Mark I. Van Berge Esophageal Surgery, Shizuoka Cancer Center, Sunto-gun, Japan Surgery, Hospital de Clínicas Universidad de Buenos Aires, Buenos 1 Aires, Argentina Arzu Oezcelik*1,2, Steven R. Demeester1, Jeffrey A. Hagen1, Henegouwen BACKGROUND: The characteristics and prognosis of 1 Shahin Ayazi 1, Joerg Zehetner1, John C. Lipham1, Surgery, Academic Medical Center, Amsterdam, Netherlands; carcinoma of the cervical esophagus differ from those of BACKGROUND: Short esophagus (SE) is defi ned as the Tom R. Demeester 1 2Anesthesiology, Academic Medical Center, Amsterdam, Netherlands carcinoma of the hypopharynx and thoracic esophagus. inability of the gastro-esophageal junction to lay tension 1 Surgery, Keck School of Medicine, University of Southern California,OBJECTIVE: Conventional open esophagectomies are A surgical resection (OP) is widely accepted as the stan- free 3 cm or more below the hiatus, after a high mediastinal dissection. If it is not recognized and treated a subsequent Los Angeles/, CA;2Surgery, University of Essen, Essen, Germany accompanied by a high rate of postoperative complica- dard treatment for advanced carcinoma of the cervical posibility of re-herniation and failure of the antirefl ux pro- BACKGROUND: tions. Although not yet proven by randomized clinical tri- esophagus. On the other hand, oncologists have recently Several studies have shown that the type advocated that a nonsurgical approach with defi nitive cedure increases, due to a repair under tension. Since the of resection and the number of removed lymph nodes are als, minimally invasive esophageal surgery appears to be a outcome of the Collis has had mixed reviews we studied promising technique that could be associated with a lower chemoradiotherapy (CRT) should be the standard treat- independent prognostic factor for an improved survival ment for this disease. It is uncertain whether defi nitive the course of our patients undergoing the procedure. after esophagectomy for esophageal cancer. The aim of this morbidity rate. The objective of this study was to compare PATIENTS AND METHODS: the results of minimally invasive esophagectomies to con- CRT achieves treatment outcomes comparable to surgery, Eleven operated patients study was to evaluate whether the type of resection and (2 Collis-Belsey and 9 Collis-Nissen), between 2001 and ventional open esophagectomies in a non-randomized because conducting a clinical randomized trial is quite dif- the number of removed lymph nodes have an infl uence on 2009, were followed up for a mean period of 43 months. patient series. fi cult because of the differing treatment characteristics. major postoperative complications. OBJECTIVES: Preop. Status: Chronic GERD symptoms partially controlled METHODS: METHODS: Preoperative characteristics and the postop- The purpose of this retrospective study was The records of all patients who underwent to evaluate the outcomes of patients who received curative by PPI (permanent), with persistence of important regurgi- erative course of patients that underwent a transthoracic tation as the main symptom. Barium Swallow consistently an esophagectomy for cancer between 2002 and 2007 treatment, of advanced stage squamous cell carcinoma of were reviewed. The stage, intraoperative data, pathology esophagectomy for esophageal carcinoma were registered showed a fi xed hiatal hernia bigger than 5 cm, with easy in a prospectively monitored database. The results of the cervical esophagus. refl ux. At endoscopy 1 case showed long and 4 cases short reports, postoperative complications and the outcome MATERIAL AND METHODS: were reported. Postoperative complications were graded patients that underwent a minimally invasive esophagec- A total of 50 patients with Barrett’s esophagus, 4 patients presented erosive disease, and using the Clavien Classifi cation. Major complications were tomy and a conventional open resection were compared. squamous cell carcinoma of the cervical esophagus were 2 cases normal mucosa. Manometry and 24 hs pHmetry defi ned as complications ≥ grade IIIb. RESULTS: From October 2009 until November 2010 a total treated from September 2002 to December 2009. Among were pathologic in all cases. RESULTS: of 73 esophageal cancer patients underwent a transthoracic them, 30 patients were resectable cases. This population Follow up: every patient had a clinical, radiological and The study population consisted of 365 patients included of 6 patients with clinical Stage -II disease, 40 with a median age of 63 years. En bloc esophagectomy was resection of whom 34 by means of a minimally invasive endoscopic evaluation, both during the fi rst year and after resection. Preoperative characteristics were comparable for with Stage III, and 4 with Stage IV (the 6th UICC-TNM three years. Satisfaction with the procedure was surveyed. performed in 229 patients (63%) and transhiatal in 136 Classifi cation) squamous cell carcinoma of the thoracic both groups. There was a trend towards a shorter hospi- RESULTS: patients were followed up for a mean period (37%). The mean number of removed lymph nodes was 54 esophagus. The CRT regimen was 5-fl uorouracil combined tal stay in the minimally invasive group (10.5 versus 13.5 of 43 months (6 patients with more than 5 years of follow after en bloc and 22 after transhiatal esophagectomy. Major with cisplatin plus 30 Gy of radiation. This treatment was complications after an enbloc esophagectomy were seen in days, p = 0.19). The overall complication rate was 63% in up). One patient died after four years of surgery for unre- both groups; pulmonary complications were present in repeated twice every 5 weeks, followed by chemotherapy, lated causes. 42 patients (18%) and after a tranhiatal esophagectomy in which was repeated every 4 weeks. The average observa- 23 (16%). The median ICU and hospital stay was 4 and 17 32% of patients in the open TTOCR group versus 31% in Clinical evaluation: All the patients controlled regurgi- the MIE group. tion period was 777 days. The treatment outcomes were days respectively. On multivariate analysis, type of resec- reviewed retrospectively. tation. Two needed low dosis of PPI for mild heartburn. CONCLUSION: Nine complained of very mild dysphagia. All the patients tion and number of lymph nodes removed were not asso- The morbidity after a minimally invasive RESULTS: ciated with major postoperative complications. However resection or an open transthoracic esophageal resection is The initial treatment was OP in 9 patients and improved their symptoms and would choose surgery again. survival was improved after enbloc esophagectomy and comparable. Furthermore, there is a trend towards a shorter CRT in 41 patients (including 21 resectable cases). The com- Radiological study: showed no GE refl ux, and no re-herni- with increasing number of removed lymph nodes. hospital stay after minimally invasive surgery. Therefore, plete response (CR) and partial response rate after defi nitive ation in any patient. CRT was 44% (57% in resectable cases), 56%, respectively. CONCLUSION: The study shows that the type of resection minimally invasive esophagectomies appear to be a safe Endoscopy: no erosive lesions. The long Barrett’s desapeared, and the number of lymph nodes removed are not associ- technique for patients with potentially curable esophageal Eleven (27%) patients received TPN and seven (17%) under- and the 4 short Barrett’s patients remained unchanged. ated with major postoperative complications. Further it carcinoma. went gastrostomy during CRT period. Seven patients (17%) Three patients had postop functional studies, with normal confi rms previous studies that the survival is improved underwent salvage surgery. Sixteen patients underwent OP values of 24 hs pHmetry. including salvage surgery. The mortality and morbidity rate after enbloc esophagectomy with increased number of Similar results (clinical evaluation, radiological study and were zero and 38%, respectively. The overall 1-, 2-, and 3- lymph nodes removed. endoscopy) were obteined after three years. year survival rates in CRT group were 76%, 51% and 42%, respectively, and 65%, 49%, 49%, respectively in the OP During the same period of time, 15 other patients were group. Postoperative oral intake was good in all patients. operated with preop suspicion of SE, but in these cases it was possible to reduce the cardia into the abdomen with- CONCLUSIONS: These results suggest that the patients out tension, after a high mediastinal dissection. There were responded as well to CRT as to OP. Defi nitive CRT provides four (27%) postop re-herniations in this group of patients. at least comparable survival as surgery. The frequency of Another 145 patients were operated without preop suspi- salvage treatment after defi nitive CRT was higher in the cion of SE, and 11 patients (8%) presented hernia recur- Poster Abstracts CRT group. Oral intake was good in the OP group. These rence in the postop studies. Tuesday results should be presented to the patients before determin- CONCLUSIONS: The Collis procedure plus a fundupli- ing the initial treatment. OP is recommended for patients cation turned out to be effective to restore the antirefl ux with severe stenosis and CRT for preserving the pharynxs. barrier controlling regurgitation, and healing peptic esoph- The effectiveness of CRT cannot be predicted, therefore agitis. It was also adequated for the anatomical repair, the optimal strategy is the preferred treatment chosen by avoiding re-herniation. All patients improved their symp- the patients after they have been fully informed of their toms, and they would choose the surgical option again, options. despite some mild postoperative discomfort. The good results were mantained over long-term follow up.

154 155 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1642 Tu1643 Tu1644 Tu1645 Implementation of an Enhanced Recovery Program inThe Application of Automatic Document Classifi cation Prior Chemoradiation and Retrosternal Placement ofGastric Electrical Stimulation: Surgical Complications Esophageal Surgery to Cancer Staging for Esophageal Pathological Reports Conduit Were Risk Factors for Development of Benignand Impact on Supplemental Nutrition Rachel L. Blom*1, Mark Van Heijl1, Willem A. Bemelman1, Yung-Han Sun*1,3, Chih-Cheng Hsieh1,2, Chun-Hsien Chen3 Anastomotic Stricture After Esophagectomy Using a Andrew S. Kastenmeier*, Konstantinos I. Makris, Lee L. Swanstrom, Markus W. Hollmann 2, Robert Tepaske2, Thomas M. Van Gulik11, Department of Surgery, Taipei Veterans General Hospital, Taipei, Hand-Sewn Technique Christy M. Dunst 1 1 2 Minimally Invasive Surgery, Legacy Health System, Portland, OR Jean H. Klinkenbijl , Olivier R. Busch, Mark I. Van Berge Taiwan; School of Medicine, National Yang-Ming University, Taipei, Daniel K. Tong*, Simon Law, Fion S. Chan, Kam H. Wong 1 3 Henegouwen Taiwan; Department of Information Management, Chang Gung Surgery, The University of Hong Kong, Hong Kong, Hong Kong BACKGROUND: Gastroparesis that is refractory to medi- 1 Surgery, Academic Medical Center, Amsterdam, Netherlands; University, Taoyuan, Taiwan BACKGROUND: cal management is a debilitating disease. Patients battle 2 Benign anastomotic stricture is a com- Anesthesiology, Academic Medical Center, Amsterdam, NetherlandsBACKGROUNDS: More than 40,000 Taiwan residents died plication of esophagectomy that affects quality of life. The chronic nausea, vomiting, abdominal pain and early satiety. BACKGROUND: A recent development in gastrointestinal of cancers in 2009 according to the statistics of Department purpose of this study is to identify risk factors for develop- In severe cases, these symptoms result in weight loss, mal- surgery is the implementation of enhanced recovery after of Health, Taiwan. Due to the advance in medical experi- ment of such strictures after a hand-sewn anastomosis. nutrition, and dependence on supplemental nutrition. The ence and knowledge over the last decade, the prognosis of limited therapeutic options have resulted in a renewed surgery (ERAS) programs. Evidence regarding the benefi t METHODS: A retrospective study was performed on a cancer patients has been signifi cantly improved and there interest in exploring surgical therapies for palliation of of these programs in patients undergoing esophageal sur- prospectively collected database for patients who had are more drugs as well as alternative treatments to help refractory gastroparesis. Gastric electrical stimulation (GES) gery is scarce. Therefore we investigated the feasibility and undergone curative esophagectomy from 1994–2008 for patients at relatively late stage of cancers. Cancer staging has gained popularity due to initial reports of safety and possible benefi t of a perioperative enhanced recovery pro- esophageal cancer. Patients who had the anastomosis con- is an important indicator for assessing the effects of cancer effi cacy. This study adds to the growing body of literature gram in patients undergoing esophagectomy for malignant structed by a one-layer continuous technique using a fi ne treatment and prognosis. Its effectiveness may be affected by describing our patients’ experiences and surgical out- disease. absorbable monofi lament suture were included. Stricture by the interpretation profi ciency of cancer registration comes with this GES. METHODS: From January 2009 until March 2010 all was defi ned as dysphagia promoting endoscopic dilata- staff who read the pathological reports of cancer patients. METHODS: esophageal cancer patients undergoing surgical resection tion. Risk factors for development of stricture were iden- Patients with gastroparesis were considered However, the manual interpretation process is somewhat who were treated according to the ERAS were included in tifi ed using univariate and multivariate logistic regression for GES after documented failure of medical management ineffi cient and time consuming. The aim of this study was this study. ERAS items included preoperative counselling, analyses. and documented delayed gastric emptying. All patients to explore the effectiveness of computationally convert- preoperative nutrition, early removal of nasogastric tube who underwent GES from March 2003 to September 2010 ing pathological reports of esophageal cancer into cancer RESULTS: Esophagectomy was performed in 526 patients. and early mobilization. Primary outcome parameters were were included. Patient data was reviewed retrospectively. staging reports by using effi cient document classifi cation The median age was 66 yrs (21–89) and 423 (80.4%) were overall hospital stay and the incidence of postoperative RESULTS: techniques. male. Benign strictures developed in 125 (23.4%) patients Thirty-nine patients had 44 gastric stimulators complications. Outcome measures in the ERAS cohort were at a median of 55 days (range 18–2,230) after surgery, laparoscopically implanted during the study period. There MATERIALS AND METHODS: Pathological reports of compared to a cohort of patients who underwent surgical requiring a median of 2 dilatations (range: 1–25); 113 were 12 patients with diabetic gastroparesis and 27 patients 234 patients undergone esophagectomy from year 2000 resection in the three years prior to implementation of the (85%) needed less than 4 dilatations. Of the 125 patients, with idiopathic gastroparesis. Mean follow-up time was 21 to 2008 in Division of Thoracic Surgery, Taipei-Veterans ERAS protocol. 75 (60%) developed stricture within 60 days after esopha- months. Preoperatively, 27% of patients were dependent General Hospital, Taiwan were collected in this study. The RESULTS: A total of 89 esophageal cancer patients under- gectomy. Prior chemoradiation (OR 2.602, 95% CI [1.635– on supplemental nutrition (2 TPN, 8 tube feeds). After GES, reports were computationally converted into weighted went surgical resection between January 2009 and March 4.141], p < 0.001) and retrosternal placement of conduit 60% were able to eliminate all supplemental nutrition in a frequency vectors of keywords by using text mining tech- 2010. 72 patients were included in the ERAS protocol (ERAS (OR 2.806, 95% CI [1.349–5.838], p = 0.006) were inde- mean time of 5 months. Neither patient on TPN was able niques to analyze cancer staging related keywords in the + group) and compared to 296 patients who underwent an pendent predictive factors on multivariate analysis. Organ to eliminate the need for supplemental nutrition; however, reports. Then, J48 decision tree induction algorithm, a esophagectomy between 2005 and 2008 (ERAS – group). used for esophageal replacement, anastomtoic leakage, the both were able to transition from TPN to tube feeds. There supervised learning algorithm, was used to evaluate the Patient characteristics were comparable with the excep- site of anastomosis, and medical complications were not. were 13 complications that required operative interven- performance of our document classifi cation model for tion of a history of cardiovascular disease and the number When refractory stricture was defi ned by those requiring 4 tion in 11 patients (30%). Complications included stimula- automatic cancer staging based on the 234 vectors. of patients who underwent neoadjuvant therapy (50% in or more dilatations, the only predictive factor was delayed tor pack malposition (4/11), stimulator lead malfunction ERAS + versus 36% in ERAS –, p = 0.03 and 60% in ERAS RESULTS: The average prediction accuracy rate for cell appearance of stricture at 60 days or more after operation (2/11), stimulator pack malfunction (2/11), ventral hernia + versus 22% in ERAS –, p < 0.001 respectively). Overall type could reach 95.3%, and those for T, N and M status (OR 2.562, 95% CI [1.082–6.067], p = 0.032). (2/11), stimulator infection (1/11), and bowel obstruc- reach 84.47%, 92.72% and 94.87% respectively. tion due to stimulator leads (1/11). Nine stimulators were hospital stay was 14 days in ERAS + versus 15 days in ERAS CONCLUSIONS: Neoadjuvant chemoradiation and retroster- – (p0.04). There were no signifi cant differences in the inci- explanted (4/9 battery failure, 2/9 malfunction, 2/9 fore- nal placement of conduit were independent predictors for gut surgery requiring stimulator removal, 1/9 infection). dence of postoperative complications in both groups. Results of the Automatic Document Classifi cations to Different development of benign anastomotic strictures after esoph- CONCLUSION: Catagories of Esophageal Cancer Staging Seven patients requiring stimulator explant elected to have The implementation of an ERAS program agectomy. Most patients required less than 4 dilatations. their stimulator replaced. Two patients did not have ade- in esophageal surgery resulted in a small but signifi cant Delayed in appearance of stricture (more than 60 days after 5-Fold Cross 10-Fold Cross quate palliation with GES and elected to undergo subtotal reduction of overall hospital stay, whereas overall morbid- Hold-Out Method Validation Validation surgery) predicted the need of more dilatations. gastrectomy. ity was not affected. Cell type 95.30% 95.73% 95.74% CONCLUSIONS: GES is an effective surgical interven- T status 84.47% 85.04% 86.14% N status 92.74% 92.75% 92.72% tion for eliminating reliance on supplemental nutrition M status 94.87% 94.88% 94.89% in patients with refractory gastroparesis. Additionally, GES The shown data were averaged by 10 times replications using J48 decision tree appears to facilitate the transition to enteral nutrition in induction algorithm patients who are TPN dependent. The incidence of surgi- Poster Abstracts cal complications requiring operative intervention is sig-

nifi cant and should be carefully considered prior to GES. Tuesday CONCLUSIONS: In esophageal cancer, using the J48 deci- All patients who required stimulator explant for a compli- sion tree induction algorithm, the average prediction accu- cation or battery failure elected to have their stimulator racy rate is high, the model may effi ciently and effectively replaced suggesting that they felt stimulator benefi ts out- assist the physicians or cancer registration staffs to improve weighed potential risks. the accuracy rate of cancer pathological stage and reduce the time-consuming stage in the large number of data pro- cessing in studies.

156 157 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1646 Tu1647 Tu1648 Clinical: Hepatic Are All Hiatal Hernia Recurrences the Same? A LargeHigh Resolution Impedance Manometry Findings in Lymph Node Ratio Is a Signifi cant Predictor of Retrospective Review of Laparoscopic Hiatal Hernia Nutcracker Esophagus Disease Specifi c Mortality in Patients Undergoing Tu1649 and Paraesophageal Hernia Repair with Nissen Masato Hoshino*, Abhishek Sundaram, Arpad Juhasz, Esophagectomy for Cancer Safety and Effi cacy of Hepatectomy for Colorectal Fundoplication Fumiaki Yano, Kazuto Tsuboi, Tommy H. Lee, Sumeet K. Mittal *1 1,2 Matthew P. Fox, Robert C. Martin Metastases in the Elderly Surgery, Creighton University Medical Center, Omaha, NE 1 Elizabeth Schmidt, Vishal Kothari, Jason F. Reynoso, Department of Surgery, University of Louisville, Louisville, KY; Robert M. Cannon*, Robert C. Martin, Glenda Callender, Dmitry Oleynikov* OBJECTIVE: 2Division of Surgical Oncology, University of Louisville, Louisville, KY The objective of this study was to correlate Kelly M. Mcmasters, Charles R. Scoggins Surgery, UNMC, Omaha, NE, NE OBJECTIVE: th high resolution impedance manometry (HRIM) fi ndings The 7 edition of the AJCC staging system Surgery, University of Louisville, Louisville, KY BACKGROUND: with symptoms in patients diagnosed with nutcracker classifi es nodal stage by the number of malignant nodes Laparoscopic hiatal hernia repair is asso- INTRODUCTION: ciated with recurrence rates ranging from 10% to 35%. esophagus based on conventional manometric criteria. found. This method may be confounded by variations Metastasectomy is the most effective Adequate mediastinal dissection, tension free closure of METHODS: A retrospective review of a prospectively in lymphadenectomy and specimen review. The ratio of therapy for hepatic colorectal metastases, yet the safety the hiatus, and crural reinforcement of large hernias with maintained database identifi ed patients with a nutcracker lymph nodes containing metastases to the total nodes and effi cacy of liver resection in elderly patients is not fully excised (LNR) has been suggested an alternative. Few stud- bioprosthestic mesh has been shown to reduce hiatal her- esophagus (distal esophageal amplitude > 180 mmHg). understood. The aim of this study was to determine the ies of LNR have included patients undergoing preoperative nia recurrence rates. Data is scarce concerning treatment Patients with a history of previous foregut surgery or safety and effi cacy of hepatic metastasectomy in elderly chemoradiation. In this study, we seek to verify the validity of asymptomatic hiatal and paraesophageal hernia recur- with type I, II, or III achalasia were excluded. We used patients with colorectal liver metastases (CLM). of LNR for staging, especially in those treated with neoad- METHODS: With IRB approval, a retrospective review of rences found on upper endoscopy or barium swallow. This two different stratifi cation schemes based on the previ- juvant therapy, compare it to the current AJCC system, and a hepatobiliary database was performed on consecutive large retrospective review evaluates incidence and manage- ously described Chicago classifi cation of normal integrated identify other variables affecting outcome. patients treated with metastasectomy for CLM. Patients ment of symptomatic and asymptomatic recurrent of hia- relaxation pressure (IRP) and distal contractile integral METHODS: tal and paraesophageal hernias. (DCI) values. One stratifi cation scheme was based on IRP: A review of our prospective esophageal database were stratifi ed by age (<70 years old vs. ≥70). Age 70 was indentifi ed 92 patients who underwent esophagectomy at out METHODS: A single institution retrospective review of all group A (IRP < 15 mmHg) and group B (IRP ≥ 15 mmHg). chosen as a breakpoint to allow for comparison to current The other stratifi cation scheme was based on DCI: group institution from 1988 until 2010 for esophageal cancer. Uni- literature. Patient and tumor characteristics as well as oper- laparoscopic hiatal and paraesophageal hernias with nissen variate and multivariate analysis were performed. fundoplication from February 2002 to November 2010 was X (DCI ≥ 8000 mmHg/s/cm) and group Y (DCI > 8000 ative factors were compared, as were morbidity, mortality, RESULTS: The mean age at diagnosis of our patients was performed. All patients underwent reduction of the hernia mmHg/s/cm). Patients rated their symptoms namely chest and length of stay. Data analysis was performed using chi 60.4. 79% were male. 47.8% had neoadjuvant therapy. Trans- sac, mediastinal dissection, primary tension free closure of pain, heartburn, dysphagia and regurgitation on a scale of squared, Fisher’s exact test, student’s t, and Wilcoxon rank 0–3. Chi square test was used to compare categorical vari- thoracic esophagectomy was performed in all but 3 instances. sum where appropriate. Survival was analyzed according to the hiatus, and placement of a bioprosthetic mesh if the 76% of the tumors were adenocarcinoma. 80.4% were located hiatus was greater than 5 centimeters. Patients were seen ables. Mann-Whitney test was used to compare continuous Kaplan-Meier and Cox proportional hazards and compared variables. in the lower 3rd of the esophagus or the GE junction. Overall using the log rank test. at two weeks, six months, and one year. All patients under- survival at 2 and 5 years was 52.5% and 19.7%. went a barium swallow at one year to evaluate their repairs. RESULTS: Thirty patients satisfi ed the inclusion and exclu- RESULTS: There were 59 elderly (median age 75) and 220 Patients with a hiatal hernia recurrence were identifi ed and sion criteria. There were 21 (70%) females. Fifteen patients Univariate analysis identifi ed a statistically non-signifi cant non-elderly (median age 58) patients analyzed over a 13 then classifi ed as symptomatic or asymptomatic. (50%) presented with dysphagia while 15 patients (50%) worse disease specifi c survival in patients with adenocarci- year period. The Charlson Comorbidity Index was higher noma (p = .207). Analysis was then carried forward in only RESULTS: 243 patients with hiatal and paraesophageal presented with chest pain. Twenty-six patients (74%) had in the elderly group (mean 4.7 vs. 4.4 p = 0.015). Postopera- spastic waves (DCI > 5,000 mmHg/s/cm and normal con- the adeno cohort. Upon stratifi cation by LNR, no signifi - tive morbidity and mortality were 52.5 and 0% vs. 48.2 and hernias underwent laparoscopic repair with nissen fundo- cant difference was found (p = .373). Conversely, the N-1 tractile front velocity). Of these 22 (85%) had impaired 0.9% in the elderly and non elderly groups, respectively plication from 2002 to 2010. 194 of 243 had been seen at group had worse DS survival then N-0 and N-2 (p = .011 relaxation i.e. IRP ≥ 15 mmHg. There was no difference in (p = .56 and p = 1.0). There was no difference in the sever- one year and underwent a barium swallow. There were 17 and p = .002). Further stratifi cation by neoadjuvant ther- clinical symptoms or manometric variables betwen groups ity of complications between the two groups (p = 0.136). (7%) recurrences seen over an average follow up period of apy showed worse survival in N-1 and 3 vs N-2 (p = .004 Length of stay was also similar between the two groups 3.5 years. 4 were symptomatic and 13 were asymptomatic. A (7/30) and B (23/30). Group X comprised 7 (24%) patients and .028) but not N-0 (p = .092 and .403) in the untreated (p = 0.23), however, elderly pts more likely to be discharged All asymptomatic patients were followed closely with either and Group Y comprised 22 (76%) patients. The number group. No minimum number of nodes harvested produced no intervention or reinstitution of proton pump inhibi- of patients with chest pain was signifi cantly (p = 0.023) additional relationships between survival, N-stage, and LNR. to a nursing facility (17.1% vs. 1.4% p = 0.0002). There tors. All of the symptomatic patients underwent revisional higher in group Y (14 patients) when compared to group were no differences in median DFS (21.8 months vs. 28.9 X (1 patient). The mean chest pain score in group Y (mean Multivariate analysis with a Cox regression including LNR, months; p = 0.07) or OS (36 months vs. 47.7 months; p = laparoscopic nissen fundoplications. Short term follow up total nodes harvested, age, gender, race, T-Stage, N-Stage, of these patients have shown no recurrence. score = 1.5) was signifi cantly (p = 0.027) higher than group 0.16) between the elderly and non-elderly groups, respec- X (mean score = 0.3). There was a positive correlation (R = tumor histology, tumor location, neoadjuvant therapy, tively. When analyzed as a continuous variable, age had CONCLUSION: Laparoscopic hiatal hernia repair with nis- 0.399) between chest pain and DCI which was statistically adjuvant therapy, preop weight loss, and serum albumin no signifi cant effect on OS or DFS (hazard ratio 1.012 95% sen fundoplication is an effective technique to repair hiatal signifi cant (p = 0.032). showed LNR to be a signifi cant predictor of DS mortal- CI 0.997–1.028 p = .1295 and 1.005 95% CI 0.991–1.019 and paraesophageal hernias. The use of adequate mediasti- ity (p = .012). N1 stage was also associated with mortality CONCLUSION: In nutcracker patients chest pain is associ- p = 0.470 respectively) There was no difference in pat- nal dissection, primary tension free closure of the hiatus, (HR = 6.8, p = .001), but N2 and N3 stages conferred a non- terns of recurrence (lung only, liver and extrahepatic, liver ated with increasing DCI. signifi cant survival benefi t (HR = .198, p = .062 and HR = and bioprosthetic mesh placement for a hiatus greater than only, liver and lung, extrahepatic only) (p = 0.59, p = 0.97, .830, p = .798). Other predictors of mortality included high 5 centimeters signifi cantly decreases hiatal hernia recur- p = 0.16, p = 0.29, p = 0.51). grade (HR = 20.6, p = .009), alcohol use (HR = 4.29, p = rences over a 3.5 year average time period. Patients who CONCLUSIONS: are asymptomatic can be followed closely without inter- .004) and tobacco use (HR = 5.84. p = .006). Hepatic metastasectomy is safe and effective for elderly patients with colorectal metastases. vention and those with symptomatic recurrences can be CONCLUSION: LNR was found to be a signifi cant predictor of Poster Abstracts repaired again safely and adequately. DS mortality in a cohort containing large numbers of patients Age alone should not be considered a contraindication to hepatic resection, including major hepatectomy. treated with neoadjuvant therapy. However, neither a staging Tuesday system based on LNR, nor its effi cacy compared to the current staging system could be determined from this data.

Synchronous Morbidity Mortality EBL (ml, LOS (days, Fong Score Transfused Major Hepa- Colectomy R0 Resection Age (%) (%) median) median) (median) (%) tectomy (%) Pringle (%) (%) (%) •70 52.5 0 250 7 2 54.2 52.5 37.5 15.3 81.4 <70 48.2 p = 0.56 0.91 p = 1.0 300 p = 0.44 7 p = 0.23 2 p = 0.45 50.7 p = 0.66 53.1 p = 0.93 27.8 p = 0.11 10.1 p = 0.36 88.7 p = 0.14

158 159 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1650 Tu1651 Tu1652 CONCLUSIONS: A multi-disciplinary approach allows some patients with complicated diffuse porto-mesenteric Outcome After Resection of Colorectal Liver Repeat Hepatic Resection for Recurrent Colorectal Re-Visiting Surgical Options for Diffuse Porto- venous venous thrombosis to avoid or at least delay MVT. Metastases: No Prognostic Infl uence of Complications,Liver Metastases: Impact of Neoadjuvant Therapy Mesenteric Thrombosis in the Era of Multi-Visceral Our approach is not new, but simply uses MVT as an abso- Blood Transfusions or Body Mass Index and Long Term Outcome Transplantation: A Case for Aggressive Conservatismlute last resort. We believe our initial results, with a sur- Frank Makowiec*, Hannes P. Neeff, Andrea Klock, Oliver Drognitz,Frank Makowiec*, Hannes P. Neeff, Oliver Drognitz, Andrea Klock, Kishore R. Iyer*1, Riccardo A. Superina3, Hiroshi Sogawa1, gical philosophy of “aggressive conservatism”, even in Ulrich T. Hopt Ulrich T. Hopt Lauren K. Schwartz 1,2, Thomas Schiano1,2 widespread porto-mesenteric thrombosis, appear to show 1 promise. MVT remains an option if the other strategies fail. Department of Surgery, University of Freiburg, Freiburg, Germany Department of Surgery, University of Freiburg, Freiburg, Germany Transplant Surgery, Mount Sinai Medical Center, New York, NY; 2Medicine, Mount Sinai Medical Center, New York,3 Surgery,NY; Because of low mortality and steadily improving outcomes liver Although advances in surgical and multimodal treat- Tu1653 resection for colorectal metastases (CRC-LM) is increasingly ment have lead to prolongation of survival in patients Children’s Memorial Hospital, Chicago, IL used as a part of multimodal therapy of metastasized CRC. with colorectal liver metastases (CRC-LM) many patients BACKGROUND: Patients with diffuse porto-mesenteric Multicenter Results of Stereotactic Body Radiotherapy In addition to established oncological parameters (e.g.; develop recurrence that is often isolated to the liver. Repeat thrombosis (PMT) are often not candidates for shunt sur- (SBRT) for Secondary Tumors of the Liver Fong criteria) factors like BMI, perioperative blood trans- resection may, again, prolong survival or even heal some gery and are candidates for multi-visceral transplantation Daniel Rojas1, Laura Snyder1, Rafael A. Ibarra1, Deepjot Singh2, fusions and postoperative complications were discussed patients. We evaluated the perioperative (under special (MVT), despite normal liver and intestinal function. WhileMin Yao 3, Charles Kunos3, Federico N. Aucejo4, for an infl uence on long term prognosis. We, therefore, consideration of neoadjuvant chemotherapy; neoCTx) and results of intestinal and multi-visceral transplantationChristopher Siegel 1, Michael T. Milano5, Alan Katz5, evaluated our survival data for prognostic factors including long-term outcomes after repeat hepatic resection for CRC- steadily improve, with 5-year survival following MVT Karynof A. Goodman 6, Kevin Stephans6, Galal El-Gazzaz4, above mentioned parameters. LM in a relatively large single institution series. the order of 50%, waiting-list mortality upwards of 25% Charles M. Miller 4, John J. Fung4, Juan Sanabria*1 METHODS: METHODS: and need for immunosuppression, alternatives to MVT Long term outcome could be assessed in 267 Since 1999, 70 repeat hepatectomies (64% 1 Surgery & Nutrition, UH Case Medical Center, Cleveland, OH; patients (76% male, 50% synchronous CRC-LM) after fi rst wedge/segmental, 36% hemihepatectomy) for recurrent need to be considered when possible. 2Medicine, UH Case Medical Center, Cleveland, 3OH;Radiation liver resection for CRC-LM since 1998. Perioperative blood CRC-LM were performed in 61 patients (8 pts. 2, one 3 re- AIM: To describe initial experience with a conservative,Oncology, UH Case Medical Center, Cleveland, OH; 4Surgery, transfusions were given in 29%. Postoperative morbidity resections). Median interval from fi rst to fi rst repeat liver step-wise surgical approach in patients with diffuse PMT Cleveland Clinic, Cleveland, OH; 5Radiation Oncology, MSKCC, (any complication) was 52%. Survival was analyzed by a resection was 15 months. Before repeat hepatectomy any referred for MVT. Kaplan-Meier-method and Cox-regression model. kind of CTx had been given in 63 cases (90%). NeoCTx New York, NY; 6Radiation Oncology, University of Rochester Medical METHODS: Retrospective review of patients with diffuse RESULTS: (CTx within 6 months before surgery) was given in 26 cases Center, Rochester, NY Free hepatic margins were achieved in 88%. PMT referred to a single surgeon, for consideration of MVT. Overall fi ve year survival was 45%. In univariate analysis (37%), in most cases with Oxaliplatin (additional antibod- BACKGROUND: More than 250,000 patients are diag- ies in n = 11). PMT was defi ned as confl uent thrombosis of portal, splenic, positive hepatic margins (p < 0.01), CRC-LM size ≥3cm superior and inferior mesenteric veins. Surgical options nosed with liver metastases every year in USA. Less than RESULTS: (p = 0.005), increased tumor markers (CA 19-9 > 100 and/ After neoCTx the number and size of metas- graded along a risk-continuum are traditional shunts fol- 20% of those lesions are amenable to defi nitive surgical or CEA > 200; p = 0.01) and node positive primary CRC tases was slightly smaller (but n.s.). Free hepatic margins lowed progressively by ‘make-shift’ shunts, ablative pro- management due to advance local disease or a medical con- (p = 0.05) were associated with poorer survival. In multivar- were achieved in 77% (independent of neoCTx). Mortality cedures (the complete, single-stage ‘modifi ed’ Sugiura dition. Non-surgical therapies, i.e., TACE, RFA have limited iate analysis CRC-LM size (p < 0.01), node positive primary was 2/70 (2.9%) No mortality was observed after neoCTx. procedure), isolated liver transplant followed lastly by MVT. response and no signifi cant impact on patient survival. CRC (p < 0.05) and increased tumor markers (p < 0.05) Complication rates were 52% (any), 6% (hepatic failure/ SBRT has emerged as an alternative therapy. RESULTS: There were 6 patients referred for MVT with independently infl uenced prognosis. The hepatic resec- bilirubine > 6 mg/dl), 17% (infectious) or 13% (relaparot- OBJECTIVE: To determine the response of liver metasta- tion status showed a trend but no signifi cant prognostic omy), all independent of neoCTx. Overall fi ve year survival diffuse PMT and life-threatening gastric and/or esophageal variceal bleeding, as the proximate indication for trans- ses to SBRT and if SBRT treatment may confer survival ben- infl uence. Synchronous and metachronous metastases (5 ySurv; n = 59) was 54%. The only signifi cant prognostic efi t to patients with non-resectable liver metastases. had almost identical outcomes. The occurrence of postop- factor was the number of metastases (5 ySurv 71% with one plant. Three of the 6 had normal liver function and 5 of the METHODS: erative complications (any, as well as hepatic or infectious vs. 46% with > one metastasis; p = 0.05). Patients with > 6 had normal intestinal function. Patients with secondary liver tumors treated complications), BMI and perioperative blood transfusions one metastasis had more positive margins (33% vs. 17% Following evaluation for MVT, the 3 patients with normal with SBRT from four Academic Medical Centers were did not show any infl uence on long term prognosis. in 1 metastasis). The size of metastasis, time interval after liver function underwent single stage modifi ed Sugiura Pro- entered into a common database. Descriptive statistics and survival curves were performed using SPSS. CONCLUSIONS: In our study long term prognosis after fi rst liver surgery, nodal status of the primary CRC, preop- cedures (esophago-gastric devascularization with splenec- RESULTS: fi rst resection of colorectal LM was mainly infl uenced by erative CTx, gender and (surprisingly) the resection margin tomy and esophageal transaction). One patient underwent 155 patients underwent SBRT for liver metas- established tumor-related parameters. As recently shown did not signifi cantly predict survival. a meso-atrial shunt using a PTFE graft with infl ow from tases. 52% of tumors originated in the GI tract while 35% by other groups the resection margin had only a minor CONCLUSIONS: Repeat hepatectomy for recurrent CRC- a mesenteric collateral vein. One patient with hepatitis C, were from the thorax including breast. 89% of treated infl uence (maybe due to further multimodal or repeat ther- LM can be performed with low morbidity and acceptable who had had had an aborted isolated liver transplant prior neoplasms responded to SBRT at a median dose of 32 Gy apy). Further examined factors like perioperative complica- morbidity. Modern NeoCTx regimen do not increase com- to referral for MVT, was successfully transplanted with an in 3 consecutive fractions (median). A Grade III/IV local tions, perioperative blood transfusions or body mass index plication rates. Long term survival after repeat hepatec- isolated liver graft after retrieving an MV graft, when trial response was observed with a mean decrease in maximum did not infl uence the outcome in our experience. tomy in this selected group is rather high and comparable dissection identifi ed a large venous collateral deemed sat- diameter from 4.3 ± 1.9 cm to 2.5 ± 1.3 cm and a calculated to data after fi rst liver resection. Patients with only one isfactory for portal venous infl ow to a transplanted liver. mean total tumor volume reduction of 39%. Recurrences CRC-LM are best candidates for surgery. The resection mar- Only one of the 6 patients with sclerosing cholangitis and in the radiated fi eld were observed in 2% of treated cases. gin had only a minor infl uence, maybe due to multimodal refractory tpn-dependence underwent uneventful MVT. SBRT did not conferred a survival advantage in patients or repeat therapy. All patients are alive with normal liver and intestine func- with secondary liver tumors when compared with matched controls (p > 0.05). Systemic recurrences were common. tion, including the 2 transplant patients (1 isolated liver, 1 Poster Abstracts MVT), with minimum follow-up of 19 months. The patient 42% of patients were readmitted to the hospital for medi- cal complications. No complications attributable to fi du- with the meso-atrial shunt required angioplasty of the Tuesday shunt infl ow for a single re-bleed and has done well on cial placement or SBRT treatment were observed. continued follow-up. CONCLUSION: SBRT is a safe and effective treatment modality for the local control of secondary liver neoplasms. Further analyses are undergoing to determine grade of response to SBRT according to tumor type.

160 161 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Clinical: Pancreas Tu1655 Tu1656 CONCLUSION: Analysis of intratumoral DPD and hENT1 expression enables the stratifi cation of pancreatic adeno- Symptomatic Change and Gastrointestinal Quality of Prognostic Impact of Dihydropyrimidine Dehydrogenase carcinoma patients treated with adjuvant GEM+S-1 che- Tu1654 Life After Pancreatectomy and Human Equilibrative Nucleoside Transporter 1 motherapy based on their likelihood of survival, and Lewis G. Rashid*, Vic Velanovich Expression in Adjuvant Gemcitabine Plus S-1 may have a potential to optimize adjuvant chemotherapy Is Biological Behavior of Combined Type IPMN of Surgery, Henry Ford Hospital, Detroit, MI Chemotherapy After Surgical Resection for Pancreaticincluding gemcitabine and S-1 for resected pancreatic the Pancreas Similar to Main Duct Type or Branch Adenocarcinoma adenocarcinoma. Duct Type? BACKGROUND: Pancreatectomy is a standard operation for many benign and malignant diseases. Although pan- Naru Kondo*, Yoshiaki Murakami, Kenichiro Uemura, Hiroshi Kono*, Takao Ohtsuka, Yoshihiko Sadakari, Tu1657 createctomy is a high-risk surgical procedure for short- Takeshi Sudo, Yasushi Hashimoto, Akira Nakashima, Yousuke Nagayoshi, Yasuhisa Mori, Kosuke Tsutsumi, term complications, it may also have long reaching effects Hiroki Ohge, Taijiro Sueda Prognostic Relevance of the Lymph Node Ratio and Takaharu Yasui, Shunichi Takahata, Masafumi Nakamura, in gastrointestinal-related symptoms. The purpose of this Surgery, Division of Clinical Medical Science, Graduate School of Masao Tanaka Variability of Lymph Nodes Evaluated Pathologically study was to assess the gastrointestinal quality of life of Biomedical Sciences, Hiroshima University,, Hiroshima, Japan for Pancreatic Cancer Survival Surgery and Oncology, Kyushu university, Fukuoka, Japan patients who have undergone pancreatectomy. BACKGROUND & AIM: Although intratumoral dihydro- Marianne Huebner*1,2, Michael L. Kendrick1, Kaye M. Reid OBJECTIVE: Intraductal papillary mucinous neoplasms METHODS: Patients who have undergone pancreatec- pyrimidine dehydrogenase (DPD) and human equilibra-Lombardo 1, Florencia G. Que1, Terry M. Therneau2, (IPMNs) of the pancreas are mainly divided into branch tomy were contacted by telephone and asked a series of tive nucleoside transporter 1 (hENT1) expression has beenJohn H. Donohue 1 1 1 duct type and main duct type, and malignant potential qualitative open-ended questions about their preoperative , Michael G. Sarr, Michael B. Farnell reported to be associated with chemosensitivity to each1 2 symptoms, change in preoperative symptoms, and new Surgery, Mayo Clinic, Rochester, MN;Biomedical Statistics and of main duct type is considered to be higher than that of fl uoropyrimidine and gemcitabine, their impact on com- postoperative symptoms. In addition, they completed the Informatics, Mayo Clinic, Rochester, MN branch duct type. International guidelines recommend bined chemotherapy is unclear. The aim of this study is to Gastrointestinal Symptom Rating Scale (GSRS), which assess BACKGROUND: resection of all main duct IPMNs and branch duct type investigate whether intratumoral DPD and hENT1 expres- Lymph node (LN) metastasis is associ- 5 domains: Refl ux Syndrome (RS), Acute Pain Syndrome having cyst size greater than 30 mm, mural nodule, dila- sion can predict survival of pancreatic carcinoma patients ated with poor survival in pancreatic cancer. The variabil- (APS), Indigestion Syndrome (IS), Diarrhea Syndrome (DS), tion of main pancreatic duct, or symptoms such as pancre- treated with adjuvant gemcitabine plus S-1 (GEM+S-1) ity in number of LN evaluated pathologically (leading to and Constipation Syndrome (CS). Higher scores indicated atitis. Combined type IPMNs are also indication for surgical chemotherapy. the pN status) and its impact on the LN ratio (LNR) is not resection because this type is considered to have the same more severe symptoms. In addition, age, gender, pancreatic known. METHODS: Intratumoral DPD and hENT1 expression malignant potential as the main duct type; however, bio- disease, type of pancreatectomy, and date of pancreatec- AIM: logical behavior of combined type IPMNs has not been well tomy were recorded. was investigated by immunohistochemistry in 86 pan- To determine the minimum number of LN that needs creatic adenocarcinoma patients who received adjuvant to be evaluated to allow accurate stratifi cation for progno- understood. Moreover, defi nition of combined type IPMNs RESULTS: 52 patients were both able and agreeable to GEM+S-1 chemotherapy after surgical resection. Associa- sis and therapy, and to determine prognostic signifi cance seems to be different from institution to institution, and participate in the study. 26 had periampullary malignancy, tions between clinicopathological factors including DPD of LNR. whether combined type might be derived from branch 17 cystic lesions, 5 neuroendocrine tumors, and 4 benign and hENT1 expression and disease free survival (DFS) and METHODS: duct or main duct IPMN has not been well elucidated. The pathology. 30 patients had Whipple pancreatectomy, 16 499 patients underwent pancreatectomy for overall survival (OS) were evaluated by univariate and mul- aim of this study was to investigate the malignant poten- distal pancreatectomy, 3 central pancreatectomy, 2 enucle- pancreatic cancer from 1997–2004. LNR, number of nodes tivariate analyses. Furthermore, we classifi ed all 86 patients tial and prognosis of combined type IPMNs defi ned by our ation, and 1 a total pancreatectomy. Follow-up from time examined, and number of metastatic nodes identifi ed into the three groups according to the number of favorable criteria using preoperative imaging study. of operation was from 1 to 13 years. 69% reported change pathologically were used to identify risk survival groups factors related to DPD and hENT1 expression. (2 favorable METHODS: from preoperative symptoms, of these 69% reporting symp- using a re-scaled log rank statistic validated with bootstrap We considered that combined type IPMNs factors: low DPD and high hENT1, 1 favorable factor: low tomatic improvement, 31% no change, and 0% worse. 50% resampling methods. The prognostic value was assessed result from tumor spread of branch duct IPMNs to main DPD and hENT1 or high DPD and hENT1, 0 favorable fac- of patients did experience new, different symptoms com- with Cox regression models. The probability that a patient pancreatic duct, and therefore, defi ned the tumor with tor: high DPD and low hENT1). Association between this pared to their preoperative state. The median GSRS scores was staged correctly as N0 was estimated with a beta-bino- cystically dilated branch(es) and main pancreatic duct dil- combined classifi cation and survival was also evaluated. atation (more than 5 mm) as combined type. The study (with interquartile ranges) were RS 0 (0-1), APS 0 (0–1), IS 2 mial model. population included 177 IPMNs resected at our institution. (1–4), DS 2 (0.5–4.5), and CS 0 (0–1). Patients with Whipple RESULTS: High DPD and hENT1 expressions were RESULTS: The median number of LN examined was 10 There were 34 main duct type (23 carcinomas), 62 branch operation tended toward worse scores compared to other observed in 35 (41%) and 63 (72%) patients, respectively. (range 2–74). To evaluate the minimum number of LNs duct type (13 carcinomas), and 81 combined type (26 carci- types of pancreatectomy in the RS domain (0.5 vs. 0, p = According to the combined classifi cation with DPD and needed to reliably stratify patients as pN0, cutoff values of nomas). The frequency of malignant IPMNs and prognosis 0.08) and IS domain (3.5 vs. 1.5, p = 0.06). 68% of patients hENT1, 37 (43%), 41 (48%), and 8 (9%) patients had 2, <11 vs ≥11 nodes examined in these pN0 patients classi- were compared between each type. with Whipple operation had new symptoms compared to 1, and 0 favorable factors, respectively. Univariate analysis fi ed patients into low or high risk for survival. Overall sur- 32% of patients with another type of pancreatectomy (p = revealed that patients with low DPD expression had signifi - RESULTS: The incidence of malignant IPMNs in main vival for pathologically node negative (pN0) patients with 0.002). However, patients <2 yrs from the operation had no cantly longer DFS (P = 0.006) and OS (P = 0.011) than those duct type (68%) was signifi cantly higher than that of <11 nodes examined was worse than for pN0 patients with different scores than those >2 yrs from the operation. with high DPD, and patients with high hENT1 expression ≥11 LNs (Figure 1) with a hazard ratio of 1.67 (95% CI:1.2, branch duct (21%), or combined type (32%) (p < 0.05). On had signifi cantly longer DFS (P = 0.003) and OS (P = 0.003) CONCLUSIONS: Patients undergoing pancreatectomy have 2.3, p < 0.003) and 3-year survivals of 32% vs 50% respec- the other hand, 10-year disease-specifi c survival rates of than those with low hENT1. The combined classifi cation frequently had their preoperative symptoms improved, tively. The probability of missing a positive node when branch duct type, main duct type, and combined IPMNs with DPD and hENT1 was also signifi cantly associated with but also have new postoperative symptoms. This is more <11 nodes were evaluated was 20%. The 3-year survival for were 64%, 61%, and 66%, respectively. Even when focus- DFS (P < 0.001) and OS (P < 0.001). Sub-analysis between common with Whipple operations. However, the severity pN1 patients with <11 nodes evaluated was similar to pN1 ing on invasive IPMNs, the 10-year disease-specifi c survival each two of the three groups revealed that patients with 2 of these symptoms are relatively mild, as measured by the patients with ≥11 nodes evaluated. When LNR was evalu- rate were not different in branch-duct (22%), main duct favorable factors had signifi cantly longer DFS and OS than type (39%), and combined type (23%) (p = 0.4). GSRS, except in the IS and DS domains. These mild symp- ated, LNR ≥0.17 indicated worse survival with adjusted those with 1 (DFS: P = 0.017, OS: P = 0.016) and than those Poster Abstracts toms seem to persist over time. This information will help hazard ratio of 1.72 (95% CI:1.3, 2.3, p < 0.001) than LNR CONCLUSION: Biological behavior of combined type with 0 (DFS: P < 0.001, OS: P < 0.001). Patients with 1 favor- with patient counseling. < 0.17 with 3-year survivals of 37% vs. 19% respectively. IPMNs seems to be similar to branch duct type IPMNs; able factor had signifi cantly longer DFS (P = 0.004) and OS Tuesday CONCLUSION: Patients with N0 disease who have less however, such a result might be caused by our defi nition (P = 0.002) than those with 0. In multivariate analysis, each than 11 LN evaluated have worse survival outcomes sug- criteria and underestimation of the malignant change of DPD and hENT1expression were independently associated gesting that metastatic nodes were missed by sampling too branch duct IPMN based on the selection of patients for with DFS (DPD: P = 0.003, hENT1: P = 0.004) and OS (DPD: few nodes (<11). For pN1 patients, LNR more accurately resection. Further investigation using other defi nition cri- P = 0.032, hENT1: P = 0.026). The combined classifi cation stratifi es patient cohorts with respect to pancreatic cancer teria of combined type IPMNs might be necessary. with DPD and hENT1 was also independently associated survival. with DFS (P < 0.001) and OS (P = 0.004).

162 163 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1658 Tu1659 Tu1660 Step Up Approach in Emphysematous Pancreatitis Pancreatic Enucleation: Improved Outcomes ComparedGlycemic Control in Non-Diabetic Patients Is Associated Rajesh Gupta*1, Yalakanti R. Babu1, Prasanna Chandrasekaran1, to Resection with Better Outcomes Following Pancreatectomy Mandeep Kang 2, Deepak K. Bhasin3, Surinder S. Rana3, Kathryn M. Dalbec*, Christy E. Cauley, Henry A. Pitt, Minna K. Lee*, Joseph Dinorcia, Megan D. Winner, James A. Lee, Rajinder Singh 1 Attila Nakeeb, C. Max Schmidt, Nicholas J. Zyromski, Beth Schrope, John A. Chabot, John D. Allendorf 1 Division of Surgical Gastroenterology, Department of General Michael G. House, Keith D. Lillemoe Department of Surgery, Columbia University College of Physicians and Surgery, Postgraduate Institute of Medical Education and Research, Surgery, IN Univ School of Medicine, Indianapolis, IN Surgeons, New York, NY Chandigarh, India;2 Department of Radiodiagnosis, Postgraduate BACKGROUND: Pancreatic enucleation is associated with BACKGROUND: Hyperglycemia is a risk factor for peri- Institute of Medical Education and Research, Chandigarh, India; a lower postoperative mortality than pancreatic resection. operative morbidity and mortality. Recent studies evaluate 3Department of Gastroenterology, Postgraduate Institute of Medical Enucleation also has the potential advantages of reduced the role of hyperglycemia on postoperative outcomes in Education and Research, Chandigarh, India morbidity and preservation of pancreatic parenchyma. patients undergoing cardiac, vascular, and colorectal sur- BACKGROUND: Emphysematous pancreatitis (EP) is However, enucleation is an uncommon operation, and gery, yet none specifi cally evaluate patients undergoing considered a virulent form of infected pancreatic necrosis good comparative data with resection are lacking. There- pancreatic surgery. (IPN). The objective of this study was to evaluate the effec- fore, the aim of this analysis was to compare the outcomes OBJECTIVE: The aim of this study was to examine the tiveness of step up approach in the management of EP. of pancreatic enucleation and resection. impact of postoperative hyperglycemia on patient out- METHODS: METHODS: Patients of IPN were analysed for presence of From 1998 through 2010, 45 patients with comes after pancreatectomy. gas in the pancreatic or peripancreatic area.prospectively small (mean 2.3 cm, range 0.5–5.5 cm) neuroendocrine METHODS: We reviewed the medical records of all between April 2008 and December 2009. tumors (49%), mucinous cystic tumors (22%), serous/sim- patients who underwent pancreaticoduodenectomy (PD) ple cysts (20%) and other benign lesions (9%) underwent END POINTS: Effectiveness of step up approach in EP and and distal pancreatectomy (DP) between January 1, 2008 pancreatic enucleation. Since 2005, 16 of 31 enucleations and December 31, 2009. We gathered data on demograph- compare result with non emphysematous IPN. (52%) were performed laparoscopically or robotically. These Kaplan-Meier survival curves for LN counts in pN0 patients. RESULTS: ics, blood glucose during the fi rst 72 postoperative hours, There were 53 patients of IPN enrolled during 45 patients were matched by age, gender, pathology, lesion and perioperative complications for all preoperative non- this study period. 16 of these patients (30.1%) had EP. All size and location as well as by year of surgery and approach diabetic patients. Three-day mean blood glucose measure- patients in EP group required percutaneous catheter drain- (open vs. laparoscopic or robotic) with 45 patients undergo- ments and episodes of hyper- (blood glucose greater than age (PCD), 8 (50%) were managed with PCD alone, while ing pancreatoduodenectomy (n = 19) or distal pancreatec- 200 mg/dL) and hypo- (blood glucose less than 70 mg/dL) the remaining 8 (50%) required necrosectomy after ini- tomy (n = 26). Serious morbidity included Grade B and C glycemia were recorded. Continuous variables were com- tial PCD. In non emphysematous IPN group (N = 37), one pancreatic fi stuls and was otherwise defi ned in accordance pared using the Wilcoxon rank-sum test. Categorical vari- patient (2.7%) was managed conservatively, 17 (45.9%) with the American College of Surgeons- National Surgical ables were compared using chi-square or Fisher’s exact test. Quality Improvement Program. Outcomes were compared managed with PCD alone, and 19 (51.3%) underwent RESULTS: necrosectomy after initial PCD. No statistically signifi cant with standard statistical analyses. A total of 201 patients underwent either PD (118) or DP (84). The mean age of patients was 62.9 years difference was present in the two groups as regards etiol- RESULTS: Operative time was nearly two hours shorter ogy (p = 0.15), percentage of necrosis (p = 0.74), APACHE with the majority being female (56%) and white (76%). for enucleation (183 vs 279 mins, p < 0.05). Operative 2 II score (p = 0.19), CTSI score (p = 0.78), bacterial isolates The mean body mass index was 25.7 kg/m . Ninety- blood loss was signifi cantly lower with enucleation (160 vs nine (49%) patients received an insulin infusion postop- (p = 0.26), presence of fungemia (p = 0.12), therapeutic 788 ml, p < 0.05). Fewer patients undergoing enucleation intervention (p = 0.3), multiorgan failure (p = 0.36), out- eratively. Eleven (6%) patients experienced an episode of required monitoring in an Intensive Care Unit (18 vs 44%, hypoglycemia without serious sequelae. All patients had come (p = 0.58). p < 0.05). Serious morbidity was signifi cantly less among CONCLUSION: an average of 5.6 blood glucose measurements daily. Dur- Step up approach could be employed in EP patients who underwent enucleation compared to those ing the fi rst 72 postoperative hours, 81 (40%) patients had with similar results as in non-emphysematous IPN. who had a resection (13 vs 36%, p < 0.05). None of the mean blood glucoses greater than 140 mg/dL. Patients who 45 enucleation patients died within 30 days whereas one had three-day mean blood glucoses greater than 140 mg/ resection patient (2.2%) died postoperatively. Median post- dL experienced a signifi cantly greater rate of overall com- operative length of stay was two days less for enucleation plications (57% vs. 41%, p = 0.03), including more frequent patients (6 vs 8 days). Only three enucleation patients (7%), infectious complications (37% vs. 17%; p < 0.01), major all with insulinomas, developed new postoperative diabe- complications (31% vs. 10%, p < 0.01), and pancreatic fi s- tes, and none had new exocrine insuffi ciency after surgery. tulae (17% vs. 8%, p = 0.04) than patients who had three- Kaplan-Meier survival curves for lymph node ratios in pN1 patients. CONCLUSION: Compared to resection, pancreatic enu- day means less than or equal to 140 mg/dL. Patients with cleation is associated with improved operative and post- higher mean blood glucose measurements also had signifi - operative outcomes. For small benign and premalignant cantly longer median lengths of stay (8 days vs. 7 days, p < pancreatic lesions, enucleation should be considered the 0.01) compared to patients with lower mean blood glucose procedure of choice when technically appropriate. measurements. During the fi rst 24 postoperative hours, 72

(36%) patients had a blood glucose measurement greater Poster Abstracts than 200 mg/dL. Patients who had an episode of hypergly-

cemia experienced a signifi cantly greater rate of pancreatic Tuesday fi stulae (19% vs. 7%, p = 0.01) than patients who did not. CONCLUSION: Among non-diabetic patients, even mod- est postoperative hyperglycemia following PD or DP is associated with increased morbidity and longer lengths of stay. Careful attention to adequate glycemic control during the early postoperative period may lead to better outcomes in patients undergoing pancreatectomy.

164 165 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1661 CONCLUSION: This study indicates that there is a modest Tu1663 increase in the overall incidence of pancreatic cancer which National Trends in the Incidence and Stage of Diseaseis largely due to notable increases in the incidence of lesions Fate of the Pancreatic Remnant After Resection for at Initial Presentation for Patients with Pancreatic involving the body and tail of the pancreas. The data also Intraductal Papillary-Mucinous Neoplasm Adenocarcinoma demonstrated an increase in the incidence of localized disease Toshiyuki Moriya*1,2, L. William Traverso2,3 1 Gautam K. Malhotra*1, Lynette M. Smith2, Quan P. Ly3, at the time of initial presentation which was again more pro- General Surgery, Mayo Clinic, Rochester, MN;2General Surgery, Aaron R. Sasson 3, Chandrakanth Are3 nounced for lesions located in the body and tail. Virginia Mason Medical Center, Seattle, WA; 3Center for Pancreatic 1 College of Medicine, University of Nebraska Medical Center, Omaha, Disease, St. Lukes Hospital System, Boise, ID NE; 2Department of Biostatistics, College of Public Health, UniversityTu1662 BACKGROUND: Long term follow-up of the pancreatic of Nebraska Medical Center, Omaha, NE;3Division of Surgical Assessing Quality of Life in Pancreatic Cancer: A remnant after resection for intraductal papillary mucinous Oncology, Department of Surgery, University of Nebraska Medical Shifting Landscape neoplasm (IPMN) is required to determine an understand- Center, Omaha, NE ing of this disease’s recurrence. Seven studies have followed Jillian K. Smith*, Jessica P. Simons, Elan R. Witkowski, 486 resected IPMN cases with a mean follow-up time of 26 BACKGROUND: There is a paucity of data on trends in Andreea Bodnari, Mary Beth Curnen, Sing Chau NG, to 53 months showing a 2% recurrence of disease. the incidence and stage of disease (SOD) at initial presenta- Mary E. Sullivan, Shauneen Valliere, Shimul A. Shah, OBJECTIVE: To examine a large single institution resection tion for patients with pancreatic cancer (PC). The aims of Jennifer F. Tseng database for IPMN maintained for annual life-long follow-up. Surgical Outcomes Analysis & Research, Department of Surgery, the study were to analyze the national trends in the inci- RESULTS: METHODS: dence and SOD at initial presentation for patients with PC University of Massachusetts Medical School, Worcester, MA Stakeholders (n = 65) included nurses (n = 24), Using a prospective IRB-approved database we reviewed 203 pancreatic resections for IPMN between 1989 as stratifi ed by location. family members of patients with pancreatic cancer (n = BACKGROUND: 13), pancreatic cancer survivors (n = 5), physicians (n = 5), and 2010. Annual pancreas protocol CT was the method METHODS: Pancreatic cancer is a devastating dis- The Surveillance, Epidemiology, and End ease with fi ve-year survival of <5%; thus, additional out- and other, including scientists (n = 18). Regarding general of imaging. Results (SEER) database (SEER 17, 2000–2007) was used come measures accounting for quality of life (QOL) and stakeholder perceptions of living with pancreatic cancer, RESULTS: Of the 203 resected specimens the IPMN lesions to identify patients with pancreatic cancer located in the functional status become increasingly important. Little is the mean rank for overall QOL was 4.8/10 (scale 0–10, 10 = were – adenoma (38%), borderline (25%), carcinoma in situ head (C25.0), body (C25.1), or tail (C25.2). SEER*Stat 6.6.2 understood, however, about QOL in living with pancre- perfect health); for immediately post-operative was 4.1/10; (CIS,16%), and invasive cancer (21%). Surgical margins was used to calculate age-adjusted incidence data which is atic cancer and undergoing treatment, including surgery. and for 6 months post-operative was 5.7/10 (Figure). Indi- were benign in all but one case with CIS (0.5%). At median presented as number of new cases per 100,000. The SOD Decision-making for patients is frequently infl uenced by vidual pancreatic cancer patients (n = 12), in contrast, dem- follow-up of 40 months, 8% (17/203) patients were found (localized, regional and distant) at initial presentation was stakeholders such as family members, healthcare profes- onstrated a mean global QOL score of 58% preoperatively, to have new disease at a median interval of 38 months (range 4–127 mo) from initial resection. All of these had analyzed to detect trends. sionals, or even patients themselves who are not currently 72% immediately post-operatively, and 63% > 3 months negative surgical margins for dysplasia except one with RESULTS: Overall, we found a modest increase in the from surgery (see fi gure). Intriguingly, the individual func- experiencing the health state of interest, e.g., post-operative adenoma. No treatment was required for 12 cases with side- incidence of pancreatic cancer for all locations from 7.5 recovery. We sought to determine how educated stakehold- tional scales demonstrated improvement after the imme- diate post-operative period over the time studied, which branch disease by imaging while surgical treatment was in 2000 to 8.1 per 100,000 in 2007. This increase in the ers evaluated QOL with pancreatic cancer, and to compare required in 5 (2%)—2 adenomas, 1 carcinoma in situ, and 2 correlated with the stakeholders’ survey; however, the incidence was predominantly due to signifi cant increases these measured perceptions with directly assessed QOL of invasive ductal carcinoma (one with liver mets). There was in the number of patients with lesions located in the body patients throughout diagnosis and treatment. overall individual patient-assessed QOL showed a paradox- no relationship between the 17 cases with new lesions and ical decline after this period. (0.86 to 1.2 cases per 100,000; 43% increase, p < 0.05) and METHODS: First, stakeholder opinions on pancreatic can- the 186 without lesions in demographics, procedure type, tail (1.0 to 1.4 cases per 100,000; 36% increase, p < 0.05) cer diagnosis and treatment were gathered via anonymous CONCLUSIONS: Overall quality of life of pancreatic cancer duct location, histology, or original margin status. of the pancreas. There were no signifi cant changes in the simultaneous polling using TurningPoint 4.2.3 software in patients appears to initially improve following surgery, but CONCLUSIONS: Possibly because of life-long annual incidence of lesions located in the head of the pancreas. November 2010 during a pancreatic cancer symposium. then declines, despite improvement in functional status imaging we found a higher incidence of recurrence than We also noted a statistically signifi cant increase in the inci- Respondents included patients, family, and healthcare pro- seen across multiple scales. Our data suggest that the opin- the literature’s 2%. Following partial pancreatic resection dence of patients with localized disease involving all loca- fessionals. Second, individual patient assessments of QOL ions of stakeholders, including patients further removed for IPMN and a 40 month follow-up with annual imag- tions (0.54 to 0.63 cases per 100,000; increase of 16%, p < were conducted by trained physicians using the EORTC from diagnosis and surgery, do not refl ect the perceptions ing we found 8% of cases developed a new IPMN lesion in 0.05). The increase in the incidence of localized disease was QLQ-C30 and PAN26 standardized surveys administered to of patients actually living through the surgical treatment the pancreatic remnant and 2% required a second resec- more pronounced for lesions located in the body (0.06 to patients at initial clinic appointment with surgical oncolo- of pancreatic cancer. As QOL and functional status become tion. We suspect, as our follow-up time increases, that new 0.09 cases per 100,000; 39% increase, p < 0.05) and tail of gist and follow-up appointments from November 2007 increasingly important in assessing outcomes for cancer lesions will constantly appear regardless if a negative surgi- the pancreas (0.06 to 0.12; 98% increase, p < 0.05). through May 2009. patients, additional investigation into the quantifi cation cal margins at initial resection. and defi nition of QOL is warranted. Poster Abstracts Tuesday

166 167 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1664 RESULTS: Of the 70 consecutive patients with SAP, 14 Tu1665 Tu1666 patients were managed conservatively, 29 underwent PCD Maximum Extent of Pancreatic Necrosis and Not Sitealone and the remaining 27 underwent initial PCD fol- Routine Peritoneal Drainage Following Elective DistalSurgical Apgar Score (SAS) Predicts Perioperative Morbidity of Necrosis Guide the Therapeutic Modality in Severelowed by surgery. <30% necrosis was present in 12 patients. Pancreatectomy: Is It Necessary? in Patients Undergoing Pancreaticoduodenectomy (PD) Acute Pancreatitis There were signifi cantly higher number of patients man- Elena M. Paulus*, Stephen W. Behrman, Ben L. Zarzaur at a High-Volume Center Yalakanti R. Babu*1, Rajesh Gupta1, Prasanna Chandrasekaran1, aged conservatively compared to those undergoing PCD Surgery, University of TN Health Science Center-Memphis, Memphis, TNAdam Berger*1, John Lindenmeyer1, Benjamin E. Leiby3, Mandeep Kang 2, Deepak K. Bhasin3, Surinder S. Rana3, (see Table 1). 30–50% necrosis was noted in 17 patients. Zvi Grunwald 2, Ernest L. Rosato1, Eugene P. Kennedy1, Signifi cantly higher number of patients underwent man- Peritoneal drainage (PD) of the operative bed following Rajinder Singh 1 elective pancreatectomy has traditionally been utilized Charlesto J. Yeo 1, Mura Assifi1 1 2 agement with PCD alone in this group compared to those 1 2 General Surgery, PGIMER, Chandigarh, India;Radiodiagnosis, PGIMER, also requiring necrosectomy (see table). More than 50% hypothetically allow early recognition and potential thera- Surgery, Thomas Jefferson University, Philadelphia,Anesthesiology, PA; 3 3 Chandigarh, India; Gastroenterology, PGIMER, Chandigarh, India necrosis was present in 41 patients. Proportion of patients peutic treatment of a pancreatic leak following reconstruc-Thomas Jefferson University, Philadelphia, PA; Department of tion or ligation of the remnant gland. Recent literature BACKGROUND: Site and extent of pancreatic necrosis are requiring open necrosectomy as step up approach were sig- Clinical Pharmacology, Thomas Jefferson University, Philadelphia, PA suggests drainage is not helpful in those following pancre- evaluated as tools in assessment of severity and prognosis nifi cantly higher in this group compared to those managed OBJECTIVES: atectomy and in fact may be detrimental. Data specifi c to PD remains a procedure that carries con- in acute pancreatitis in previous studies. The aim of the with PCD alone (see table). those having elective distal pancreatectomy (DP) has not siderable morbidity even when performed at high-volume present study was to assess whether the site and maxi- Pancreatic necrosis of head of pancreas with or without received prior evaluation. We hypothesized that PD does centers. Numerous studies have evaluated pre-, peri-, and mum extent of pancreatic necrosis during the stay has involvement of body or tail was noted in 36 patients (see not mitigate the morbidity or the need for post-operative post-operative risk factors to predict patients at higher risk. role in guiding the therapeutic modality in Severe Acute table) and pancreatic necrosis involving only the body therapeutic intervention in those having DP and further The aim of this study was to determine if SAS predicts peri- Pancreatitis and or tail was noted in 34 patients (see table). No signifi - may not be an effective means toward preventing or recog- operative morbidity and mortality. METHODS: Retrospective analysis of prospectively main- cant difference was noted in management based on site of nizing intra-abdominal complications when they do occur. METHODS: We examined 405 patients who underwent tained database of 70 consecutive patients with SAP from necrosis (Table 1). METHODS: We retrospectively reviewed 69 patients hav- successful PD between January 2000 and July 2009 with April 2008 to December 2009. All the patients underwent CONCLUSIONS: Site of pancreatic necrosis has not infl u- ing elective DP at a university hospital from 1997–2010. complete data. Records were reviewed for the following CT abdomen after 4 days after the onset of acute pancreati- enced the management modality. However signifi cantly Factors examined included the development of post-oper- intraoperative SAS variables: lowest heart rate, lowest mean tis and CT scans were repeated as and when required. As a higher number of patients with >50% necrosis required ative intra-abdominal complications (fi stula, pseudocyst, arterial pressure, and blood loss. Postoperative complica- step up approach, patients were initially managed conser- surgery compared to PCD alone group while Signifi cantly bleeding, abscess) particularly those that required thera- tions were graded using the Clavien scale (Grades 0 to 5), vatively and as per the indication underwent Percutane- higher percentage of patients with <30% necrosis could be peutic intervention (radiologic drainage or re-operative and SAS (range 0 to 10) was determined (Guwande AA, ous Catheter Drainage (PCD) and Surgery. The maximum managed conservatively. surgery). PD was utilized at the discretion of the attending J Am Coll Surg. 2007). The Cochran-Armitage test for trend extent of pancreatic necrosis was categorized into three physician. Criteria for drain removal include output <50 cc/ was used to determine the association between grouped groups, i.e., less than 30% necrosis, 30–50% necrosis and day and/or amylase value <3X serum level. The comparison scores (0–2, 3–4, 5–6, 7–8, 9–10) and each of the outcomes. more than 50% necrosis. The site of pancreatic necrosis between those with and without drains was made utilizing RESULTS: The average age was 64 years old (range, was also categorized into two groups, i.e pancreatic necro- chi-square analysis or Fischer’s exact test where appropriate 18–91) and there were equal numbers of males and females sis involving head region and necrosis sparing the head with signifi cance assessed at the 95th percentile. (199:206). There were nine perioperative deaths (2%), 146 region. The Maximum Extent and Site of Pancreatic Necro- grade two or higher complications (26%), and 64 major sis in patients managed with different modalities were RESULTS: Sixty-nine patients had DP during the study complications (grades 3–5, 16%). Additionally, 39 patients compared. period, 30 of who did not have PD. Sixty percent were female, 75% had surgery for malignancy and 23% had con- developed pancreatic fi stulae (9.6%). Statistical analysis current extra-pancreatic organ removal with no difference determined that SAS was a signifi cant predictor of grade 2 or higher complications (p < 0.0001), major morbidity Table 1: CT Scan in Acute Pancreatitis between groups. Thirty-four patients (49%) suffered 45 complications post-operatively. The majority of the mor- (p = 0.02), and pancreatic fi stula (p = 0.01), but not mortal- ity (p = 0.9). PCD Alone PCD Followed by Conservative Vs PCD PCD Vs PCD Followed by Surgery bidity was intra-abdominal in nature including 15 with CT Abdomen Conservative (n = 14) (n = 29) Surgery (n = 27) P [OR] P [OR] abscess, 6 with pancreatic fi stulae and 11 with pseudocysts. CONCLUSIONS: We demonstrate that the SAS is a sig- Maximum extent of necrosis during the 5 (42) 3 (18) 6 (15) 3 (25) 12 (71) 4 (33) 2(11) 21(51) 0.04[0.2] 0.2[2.5] 0. 7[1.2] 0.6[1.5] 0.003[0.1] 0.02[3.7] Twelve and 19 patients respectively required radiologic nifi cant predictor of perioperative morbidity for patients stay <30% (n = 12) 30–50% (n = 17) 14 (34) >50% (n = 41) drainage, re-operation or both post-operatively. There was undergoing PD. This score should be used to identify Site of necrosis Head Region ± Body, 5 (14) 9 (26) 18(50) 11(32) 13(36) 14(42) 0.1[2.9] 0.3[0.5] no difference between those with drains and those without patients at higher and lower risk in order to prioritize use of tail* (n = 36) Only Body ± tail (n = 34) with respect to overall or intra-abdominal complications, critical care beds and hospital resources. Disease specifi c Mortality 1 (7) 2 (7) 11(40) 0.9[0.9] 0.003[9.3] the need for radiologic drainage or re-operation. Of the * complete necrosis - conservative group (n = 1); PCD alone group (n = 3); PCD followed by surgery (n = 6) 39 patients having PD, 19 had post-operative abdominal morbidity. The drain was useful in preventing, identifying, and/or treating the complication in only 3. CONCLUSIONS: 1.) PD following elective DP does not confer a reduction in overall morbidity, post-operative intra-abdominal complications or the need for therapeutic intervention versus those with no drains. 2.) The presence of a drain, in general, was not helpful in preventing, detect- Poster Abstracts ing or treating a post-operative intra-abdominal complica- Tuesday tion. 3.) A multi-institutional, prospective, randomized study would be helpful to further investigate these prelimi- nary fi ndings.

168 169 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1667 CONCLUSIONS: Elderly patients tolerate major pancreatic Tu1669 a Braun entero-enterostomy 3 to 7 years out from their resection for adenocarcinoma; however, major morbidity Whipple procedure (4 had a pylorus sparing Whipple and Pancreatic Resection in the Octogenarian and mortality are higher. The majority of elderly patients Results of Braun Entero-Enterostomy for Suspected one a Whipple with a distal gastrectomy and Billroth II Joseph A. Blansfi eld*, Mohsen M. Shabahang, Nicole L. Woll, require rehabilitation after discharge which may hinder Afferent Loop Syndrome in Patients Who Had a anastomosis). 4 of the Braun entero-enterostomies were Erica L. Case, William E. Strodel, Angela Huttenstine, their overall quality of life. Elderly patients should be coun- Whipple for Chronic Pancreatic Pain done with laparoscopy. The fi fth had an open Puestow Kathy J. Gorton, Marie A. Hunsinger seled regarding outcomes prior to pancreatic resection. Michael Lalla*, Maurice E. Arregui type revision of the pancreatico-jejunostomy along with Surgical Oncology, Geisinger Medical Center, Danville, PA More studies are needed to further stratify elderly patient’s Surgery, St. Vincent’s Hospital, Indianapolis, IN a Braun entero-enterostomy. 4 reported an improvement- risk from pancreatectomy. 3 with a Braun entero-enterostomy and one with a Braun BACKGROUND: INTRODUCTION: This study compares outcomes of elderly Patients suffering from chronic pan- combined with a Puestow. One patient did not have any patients (80 years old or greater) undergoing pancreatic Tu1668 creatic pain may undergo a Whipple procedure to help improvement. This patient had a standard Whipple with resection to a control group of younger patients. relieve their pain. However, after months or years of relief, gastrectomy and Billroth II. METHODS: This is a retrospective study of all patients Targeting Early Deaths Following Pancreatico- some develop symptoms of nausea, vomiting and recur- RESULTS: duodenectomy to Improve Survival rent abdominal pain. We present a retrospective review The table represents the subjective classifi cation undergoing pancreatic resection for a neoplasm at our of the intensity of the patient’s symptoms before and after institution between January 2004 and December 2010. Whalen Clark*1,2, Natalie Donn1,2, Farhaad C. Golkar1,2, of patients who had a Whipple for chronic pancreatic pain who developed clinical symptoms suspicious for Braun entero-enterostomy. With the exception of patient RESULTS: Melissa Silva 1,2, Harold Paul1,2, Kenneth Luberice1,2, One hundred fi fty patients underwent pan- afferent loop syndrome and their outcomes after a Braun E, all patients described some measure of improvement in Jonathan M. Hernandez 1,2 1,2 creatic resections for a neoplasm during the study period. , Sharona B. Ross, entero-enterostomy. their symptoms. Twenty nine patients were in the octogenarian group; 24 ofLeigh Ann Humphries 1,2 1,2 , Alexander S. Rosemurgy METHODS: CONCLUSIONS: The development of nausea, vomiting these patients had an operation for an adenocarcinoma. 1A Surgery, University of South Florida, Tampa,2 TampaFL; General We reviewed our patient database over the last 5 years and identifi ed patients who underwent a Whip- and abdominal pain after a latency period in patients with group of 31 patients (<80 years old) were included as a com-Hospital, Tampa, FL pylorus sparing Whipple for chronic pain syndrome may parison group. The groups were evenly matched for type of ple for chronic pancreatic pain. 16 patients were identifi ed INTRODUCTION: Increasing focus on outcomes with and 7 had suspected afferent loop syndrome based on their be due to poor drainage of the afferent loop which may be operation, duration of operation and estimated blood loss. relieved by diversion through a Braun entero-enterostomy. Table 1 shows demographic data for the two groups. disease specifi c therapies encourages referrals to centers clinical symptoms, labs, and multiple imaging modalities with optimal outcomes. Strategies to improve survival after including 2-hour quantitative HIDA scans. 5 underwent Final pathology was comparable. Three patients had posi- pancreaticoduodenectomy must address early mortality to tive margins in the elderly population (12.5%) versus 5 in improve long-term survival. This study was undertaken to Symptom Comparison Pre and Post Braun Entero-Enterostomy the younger population (16%). Eight patients were node analyze early mortality after pancreaticoduodenectomy as positive in the elderly population compared with nine part of a strategy to improve long-term outcome. Patient/Symptom Belching/ Bloating Nausea Vomiting RUQ Pain LUQ Pain Epigastric Pain Overall patients in the younger population. Median tumor size was A — — 9 4 4 0 5 0 —— —— 8 0 METHODS: comparable at 3cm in each group. 1,031 patients who underwent pancreati- B 10 2 10 5 —— 3 3 —— — — 10 6 coduodenectomy at a university affi liated hospital from C 4 0 9 5 4 0 10 8 10 7 10 5 10 7 Post-operatively, elderly patients had a statistically signifi - 1992–2010 were studied. Medium data are reported. D with Puestow 6 4 6 4 5 3 6 4 5 4 6 4 6 4 cant increase in length of stay compared to controls (13 RESULTS: E 8 8 8 8 8 8 8 8 7 7 8 8 no change versus 8 days, p < 0.02). Length of intensive care stay was 58 (5.6%) patients, 60% male and age 72 years, Patients reported a subjective numerical value to their symptoms with 10 being severe and 0 resolved not signifi cantly different between the two groups. Two died within 90 days after pancreaticoduodenectomy. All patients died within thirty days in the elderly group com- patients had a least one signifi cant comorbidity; signifi cant pared with no deaths in the control population. Major comorbidities were cardiorespiratory (N = 86), endocrine (N = 26), gastrointestinal (N = 47), hematologic (N = 8), or morbidity in the elderly population was 45% compared Tu1670 fusion in 4 patients (9.5%); length pancreas resected 7.04 with 32% in the younger population. other (N = 14). 60% of patients had depressed serum albu- ± 0.4 cm; tumor diameter 2.6 ± 0.2 cm; margins negative min levels and 43% were jaundiced. ASA Class was: 17% Laparoscopic-Assisted Spleen Preserving Left-Sided in all patients; nodes harvested 4.24 ± 1.3/patient;cancer Eldely Control Class II, 72% Class III, and 10% Class IV. 74% had malig- Pancreatectomy (DP-SP): A Safe Effective Option for Total 24 31 0/42 (0%). Liquid diet tolerated median postoperative day nant disease, 17% had pre-malignant diseases, and 10% Benign Tumors of the Pancreas 1 (range 0–2), solid diet day 2 (range 2–4). Mean length Men 8 18 had benign disease. 22% of patients underwent a major *1 1 2 1 of hospitalization 2.93 ± 0.16d, median 3d (range 1–6d). BMI,average 24 31 vascular resection or colon resection at the time of pancre- Kaylene Barrera, Dilip Parekh, William Boswell, Rick Selby, 1 Clinically signifi cant pancreatic fi stulae (ISGPF grades 2–3) Median Time to OR (days) 29 20 aticoduodenectomy. Causes of death were cancer progres- Sujit Kulkarni Procedure 1 in 6 patients (14%). Signifi cant overall morbidity: Clavien Surgery, University of Southern California, Los Angeles, CA; Whipple 20 22 sion/cachexia (9%), cardiorespiratrory failure (17%), leak/ 2 grade 2 in 16%, Clavien 3 in 2% and Clavien 4-5 in 0%. Left pancreatectomy 3 7 perforation (10%), infection (7%), vascular/bleeding (26%), Radiology, University of Southern California, Los Angeles, CA Pulmonary morbidity in 2 patients(4.5%); mortality was Total pancreatectomy 1 2 or other (5%). Of all deaths, early deaths (i.e., < 30 days) Open DP-SP is an established safe procedure for benign 0%. One patient developed a partial splenic infarct, there Average EBL 400 350 are more likely due to cardiorespiratory disease (21%), leak/ pancreatic tumors. Experience with laparoscopic DP-SP is were no pseudoaneurysms or splenic vein thrombosis in perforation (17%), and vascular/bleeding complications limited to small studies and anecdotal reports, safety of standard DP-SP. (31%), and later deaths (i.e., 31–90 days) are most likely preserving the splenic vessels is controversial and the War- CONCLUSIONS: In this largest single institution, single At time of discharge, 15 elderly patients (62.5%) required associated with progression of cancer/cachexia (10%), car- shaw technique advocated. diorespiratory disease (13%), or vascular/bleeding (21%). surgeon experience the standard laparoscopic-assisted rehabilitation stays compared with 6 patients (19%) in the METHODS: control group (p < 0.0005). Further, in terms of post-oper- CONCLUSIONS: 120 patients underwent laparoscopic-assisted DP-SP is a safe operation. The Warshaw technique should Death within 90 days after pancreatico- left-sided pancreatectomy; in 42 patients the spleen was pre- be limited to salvaging DP-SP if disruption of the splenic ative chemotherapy, 8 elderly patients underwent chemo- duodenectomy is uncommon, occurs in relatively older Poster Abstracts therapy (33%) compared with 24 of the control group (77%) served (35%) and are analysed here. Standard laparoscopic- vein is encountered during standard laparoscopic DP-SP. deconditioned patients, and is generally not causally related assisted DP-SP (splenic vessels preserved) was performed Due to limited nodal dissection DP-SP should be limited (p < 0.004). Six elderly patients underwent post-operative Tuesday to underlying malignancy. Early death is generally associ- in 38 and Warshaw technique (splenic vessels ligated) in to benign tumors only. Prognosis for long term splenic sal- radiation compared with 23 patients in the control group ated with vascular or bleeding complications. Strategies to (p < 0.0009). Most elderly patients did not complete post- 4 patients.: Age 51 ± 2.7y; ASA 2.2 ± 0.1; BMI 26.8 ± 0.9; vage is excellent and splenic thrombosis was not seen in minimize early death should focus on careful patient selec- OR time 169.7 ± 9.9 min; blood loss 223.9 ± 42.2; trans- the standard laparoscopic-assisted DP-SP. operative chemotherapy or radiation therapy regimens. tion and prompt recognition and management of herald bleeding or vascular thrombosis, as it can often result in perioperative death following pancreaticoduodenectomy.

170 171 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1671 Clinical: Small Bowel Tu1674 Clinical: Stomach Operative Complications of Pancreatoduodenectomy Risk of Small Intestinal Bacterial Overgrowth in in Patients with Elevated Serum Bilirubin Tu1673 Roux-en-Y Gastric Bypass Tu1675 Ahmed E. Mukhtar*1, Gregory A. Cote1, Joshua A. Waters2, Walter W. Chan*1, Christopher C. Thompson1, David B. Lautz2, Laparoscopic Lysis of Adhesions for Small Bowel Predictive Factors of Myocardial Infarction in Bariatric Keith D. Lillemoe 2 2 2 Robert Burakoff 1 , C. Max Schmidt, Michael G. House Obstruction Is an Uncommon but Safe Procedure Surgery 1 Gastroenterology, Indiana University, Indianapolis,2Surgery, IN; Indiana 1 Division of Gastroenterology, Hepatology and Endoscopy, Brigham *1 2 1 1 University, Indianapolis, IN Anton Simorov, Jason F. Reynoso, Oleg Dolghi, Dmitry Oleynikov* and Women’s Hospital, Boston, MA;2Department of Surgery, Nassim Naderi, Hossein Masoomi, Shaista Malik, Narula Jagat, 2 Surgery, UNMC, Omaha, NE, NE Brigham and Women’s Hospital, Boston, MA Ninh T. Nguyen BACKGROUND: Hyperbilirubinemia is associated with 1 INTRODUCTION: Medicine- Cardiology, University of California, Irvine-Medical Center, immune inhibition and coagulopathy, thus preoperative Laparoscopic lysis of adhesions (LLOA) BACKGROUND: Small intestinal bacterial overgrowth Orange, CA;2 Surgery, University of California, Irvine-Medical Center, biliary drainage (PBD) has become standard practice at for small bowel obstruction (SBO) has not increased in (SIBO) has been associated with small intestinal and Orange, CA many institutions for patients with obstructive jaundice prevalence as other minimally invasive procedures such as colonic surgeries. However, with few large clinical trials who are being considered for pancreatoduodenectomy cholecystectomy, herniorrhaphy, and appendectomy have. and potential confounding by factors such as PPI use, the BACKGROUND: Morbid obesity is associated with multi- (PD). This study examines the peri- and postoperative Loss of laparoscopic workspace and friability of distended role of upper gastrointestinal surgeries in clinically signifi - ple comorbidities, particularly hypertension, diabetes mel- complications in patients with an elevated serum bilirubin bowel increase both the diffi culty and risk of this proce- cant SIBO remains unclear. Patients with prior Roux-en-Y litus, and hyperlipidemia which are the known risk factors undergoing PD at a high volume center. dure. This study examines the prevalence, safety and effi - gastric bypass (RYGB) frequently present with symptoms for Myocardial Infarction (MI). cacy of LLOA and open lysis of adhesions (OLOA). similar to SIBO, such as nausea, vomiting, bloating, gas, OBJECTIVE: METHODS: We conducted a retrospective cohort study of Our aim was to identify factors associated METHODS: University HealthSystem Consortium (UHC) diarrhea, and abdominal pain that may lead to reconstruc- with an increased likelihood of developing MI after bariat- all patients undergoing PD with a preoperative serum total tive surgery. bilirubin ≥10 mg/dL between 1998 and 2008. We excluded is an alliance of more than 100 academic medical centers ric surgery. AIM: patients who had previously undergone PBD or did not and affi liate hospitals. UHC’s Clinical DataBase/Resource To identify the risk of SIBO among RYGB patients METHODS: Using the Nationwide Inpatient Sample (NIS) have available follow-up data for review. Patients were Manager (CDB/RM) allows comparison of patient-level using lactulose breath test (LBT). database, clinical data of patients who underwent bariatric identifi ed from a prospective pancreatic surgical database risk-adjusted outcomes for performance improvement. METHODS: This was a retrospective cohort study of 313 surgery from 2006–2008 was examined. Regression analysis which included peri- and postoperative complications. This study is a multi-center, retrospective outcome anal- symptomatic patients referred to a tertiary care center for was performed to evaluate effect of patient characteristics (age, ysis of discharge data of adult patients with small bowel sex and race), comorbidities, payer-type and surgical tech- RESULTS: Forty-nine deeply jaundiced patients with a LBT. Conditions and use of medications thought to be asso- obstruction secondary to adhesions who undergo LLOA or ciated with SIBO (bowel surgery, diabetes, irritable bowel niques (open vs. laparoscopy and gastric bypass vs. non-gastric mean age of 64 ± 12.5 years were identifi ed. PD was per- OLOA. Main outcome measures analyzed were mortality, bypass) on post-operative MI in bariatric surgery. formed for pancreatic adenocarcinoma (36), ampullary syndrome, pancreatic insuffi ciency, thyroid disorder, pro- morbidity, 30-day readmission, intensive care unit (ICU) ton pump inhibitor [PPI] use) and presenting symptoms RESULT: A total of 304,515 patients underwent bariatric adenoma/adenocarcinoma (7), cholangiocarcinoma (4), admission, length of stay (LOS) and cost. and chronic pancreatitis (2). For these patients, the mean were reviewed. Patients with prior RYGB were identifi ed. surgery during these three years. The majority of patients RESULTS: total serum bilirubin level was 16.78 ± 5.61 mg/dl. Other 20,514 patients with SBO underwent LLOA Positive LBT required a rise > 20 ppm in breath hydrogen or were female (80%) and Caucasian (74%). The mean age was preoperative clinical fi ndings included: mean albumin (n = 2940) or OLOA (n = 17,574) over a 45 month period methane concentration within 60 minutes. Univariate and 44.1 and 31.7% were older than 50 years. The incidence of multivariate analyses were performed using Chi-square or comorbidities were high as expected: hypertension (HTN, 2.79 ± 0.59 g/dl; mean International Normalized Ratio between October 2007 and November 2010. LLAO repre- Fisher-exact test and logistic regression, respectively. Statis- 54.5%), diabetes mellitus (DM, 29.7%), hyperlipidemia 1.21 ± 0.40; and median carcinoantigen 19-9 level 270 u/ sented 14.3% of all operative treatments. Comparing post- operative outcomes LLOA showed lower mortality (1.12% tical signifi cance was established using p < 0.05. (20.1%), chronic lung disease (18.3%), sleep apnea (15.7%), ml (range: 66 – 794). Postoperative complications were smoking (13.2%), liver disease (8.9%), congestive heart fail- recorded in 22 patients (45%), and 2 patients died within LLOA vs. 4.36% OLOA; p < 0.0001), lower morbidity (5.65% RESULTS: 105 (35.5%) patients had positive LBT. Overall, LLOA vs. 15.77% OLOA; p < 0.0001), reduced length of stay ure (CHF: 1.3%), renal failure (0.8%), peripheral vascular 30 days of operation. Three patients (6%) required reopera- 25 patients had prior RYGB. Univariate analysis revealed an disease (PVD, 0.5%) and alcohol abuse (0.2%) respectively. (7.99 ± 10.84 days LLOA vs. 15.77 ± 17.56 days OLOA; p < tion for postoperative complications. Major postoperative increased rate of positive LBT among RYGB patients com- The most common payer-type was private including HMO complications requiring intervention included bleeding 0.0001), reduced costs ($21,053 ± 1 3,665 LLOA vs. $39,453 pared to controls (70.0% vs 34.0%, OR = 4.54, p = 0.02). On (73.5%). 52.75% of procedures were performed in the aca- (3 patients), pancreatic fi stula (4), cardiac (5), pulmonary ± 26,564 OLOA; p < 0.0001), lower 30-day readmission multivariate analysis, RYGB remained independently asso- demic medical centers. (5), and surgical site infection (5). Minor complications rate (6.21% LLOA vs. 8.51% OLOA; p < 0.001), and lower ciated with positive LBT when other risk factors for SIBO Overall prevalence of MI was 0.20% (612 patients). MI was sig- included self-limited gastroparesis (5 pts), uncomplicated ICU admission rate (11.97% LLOA vs. 36.27% OLOA; p < were controlled (OR = 4.35, p = 0.04). Secondary analysis 0.0001). Higher risk patients with major/extreme severity by adding an interaction term for PPI use and RYGB to the nifi cantly higher in open procedures compare to laparoscopic anastomotic leaks (6), and superfi cial wound infection (3). procedures (0.32% vs. 0.18%; P < 0.001). The prevalence of The mean length of postoperative stay was 13 ± 11 days. of illness scores represented a signifi cantly higher portion multivariate analysis demonstrated no effect modifi cation of the OLOA group (23.7% LLOA vs. 46.6% OLOA; p < by PPI on RYGB’s risk for SIBO. A separate multivariate MI was signifi cantly lower in Non-Gastric bypass compare to CONCLUSION: Gastric bypass surgery (0.17% vs. 0.21%; P = 0.044). Pancreatoduodenectomy in patients with 0.001). analysis was performed using RYGB and presenting symp- deep jaundice is associated with substantial morbidity and CONCLUSION: toms as predictors. RYGB remained an independent risk Using multivariate regression analysis, age older than 50 mortality and often requires prolonged hospitalization. A LLOA represents a small portion of opera- factor for positive LBT when presenting symptoms were (odds ratio [OR]: 2.27), male sex (OR:1.62), congestive randomized clinical trial evaluating the potential benefi t of tive treatment for SBO with the vast majority of operative controlled. heart failure (OR:5.38), hypertension (OR: 2.34), chronic PBD at a high-volume pancreatic center is justifi ed. treatments utilizing open technique. OLOA patients were lung disease (OR: 2.23), smoking (OR:1.61), hyperlipidemia CONCLUSIONS: RYGB increased the risk for symptom- higher risk than LLOA patients. However, when LLOA is (OR: 1.43), sleep apnea (1.74%), gastric bypass (OR:1.37) atic SIBO. When other factors associated with SIBO were performed it is a safe and effi cacious option. LLOA should and academic medical center (OR:1.28) were associated be considered as treatment of SBO when the surgeon is controlled, RYGB remained an independent risk factor for with higher MI rate. There was no effect of race, diabetes, comfortable with the technique and the patient is a good SIBO. In particular, PPI use, which was previously identi- liver disease, renal failure, peripheral vascular disease, alco- candidate (no evidence of sepsis, laparoscopic workspace fi ed as a strong SIBO risk factor and is widely used in RYGB hol abuse and payer-type on MI. Poster Abstracts can be created, and adhesions amenable to laparoscopic patients, did not signifi cantly confound or modify the CONCLUSION: Age over 50, male sex, smoking, CHF, Tuesday lysis). effect of RYGB on SIBO. When evaluating RYGB patients with post-operative abdominal symptoms, SIBO should hypertention, chronic lung disease, hyperlipidemia and be considered prior to pursuing more invasive diagnostic gastric bypass surgery all increase the risk of postopera- and therapeutic procedures. Further prospective studies are tive myocardial infarction in patients undergoing weight needed to better identify the RYGB patients most at risk for loss surgery. Surgeons should be aware of these risk factors SIBO and formulate the optimal treatment strategy in this when selecting patients for surgery and further preopera- patient group. tive screening and pre-treatment may be needed in this subset of high-risk patients

172 173 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT 52ND ANNUAL MEETING • MAY 6–10, 2011 • CHICAGO, IL

Tu1676 Tu1677 postop. Medicare national average for CT imaging was 391 RESULTS: Patients that are diabetic lose signifi cantly per 1,000 patients, and for non-CT imaging, 3,556 per 1000 less %EWL (63.8% ± 2.9%, 60.7 ± 3.6 lbs) on average as Treatment of Anastomotic Strictures and Chronic Gastrointestinal Resource Utilization After Gastric patients. Surgical technique was not associated with any compared to non-diabetic patients (70.6% ± 4.9%, 69.8 ± Leaks After Bariatric Surgery with Endoscopic Stent Bypass pattern of re-intervention or complication, except for inci- 1.4 lbs), P = 0.037. Further analysis reveals that both pre- Placement Ross F. Goldberg*1, Michael Parker1, John Stauffer1, sional hernias which were greater after open RYGB. operative %LBM and three year post-operative %LBM for Jessica M. Gutierrez*, Michael G. Sarr, James M. Swain Mauricia Buchanan 1, Jacob Sylvia1, Salman Moti1, Susanne CONCLUSIONS: Even though the benefi ts of RYGB are diabetic patients (52.4% ± 0.8%, 67.5% ± 0.9%) were signif- Surgery, Mayo Clinic, Rochester, MN Preissler 1, James M. Swain2, Michael G. Sarr2, Horacio J. Asbun1, well established in the literature, patients after RYGB icantly lower than the non-diabetic group (57.0% ± 0.3%, 1 1 72.9% ± 0.4%), P = 0.005 and P = 0.01 respectively, but BACKGROUND: C. Daniel Smith , Steven P. Bowers appear to have a greater requirement for gastrointestinal Fistula, leakage, and stricture formation1 both groups increased their %LBM. There were no signifi - Department of Surgery, Mayo Clinic – Florida, Jacksonville, FL; procedures than the average American population. This after bariatric surgery are associated with substantial early cant differences between the change in %LBM for diabetic and late morbidity. We evaluated the effi cacy of endolumi- 2Department of Surgery, Mayo Clinic – Minnesota, Rochester, MN would not be unexpected in patients having undergone a major gastrointestinal operation. patients (15.1% ± 0.7%) and non-diabetic patients (16.2% nal stent therapy as defi nitive management. INTRODUCTION: Bariatric surgery decreases weight and ± 0.4%), P = 0.34. METHODS: its associated comorbidities, but it is unclear whether A retrospective review was performed from Tu1678 2001 to October 2010 from a single institution including health care resources are over-utilized in the post-gastric all patients treated with stent placement after bariatric sur- bypass state. There are few, long-term studies documenting Diabetic Patients Have Less Lean Body Mass Which gery. We collected age, sex, body mass index (BMI), time the utilization of health care resources after gastric bypass Is Correlated with Less Excess Weight Loss in from operation to onset of symptoms, number of endos- in a mixed American population. Our aim was to review Laparoscopic Adjustable Gastric Banding (LAGB) copies needed, stent-related complications, and long term all interventions related to gastrointestinal complications Over Three Years, N = 601 follow up. after gastric bypass surgery. RESULTS: METHODS AND PROCEDURES: We retrospectively David A. Nguyen*, Grace J. Kim, Carson D. Liu 22 patients (17 females/5 males, mean age 51; Bariatric Surgery, Skylex Advanced Surgical, Inc, Los Angeles, CA range 36–71 years) were treated with stents; 16 had gastric reviewed the medical records of 261 subjects in Olmsted bypass, 5 duodenal switch, and one gastric band. Eleven County (thus captured by the Rochester Epidemiology BACKGROUND: LAGB has been shown to be an effective of these operations were revisional bariatric surgery. 20 Project) who underwent a Roux-en-Y gastric bypass (RYGB) treatment of weight loss for the morbidly obese. However, patients presented from their primary bariatric operation to from Jan. 1997 through Dec. 2004. Patients were predomi- it has been found that diabetics experience lower percent- our institution with a mean of 386 days (range of 11–1875 nantly female (84%), with a mean age of 45 years ± 10, age excess weight loss when compare to non-diabetics after days). The main indications were: Fistulas- 4 gastrocutane- and a mean preoperative BMI of 49 ± 9; 72 patients (28%) LAGB. This study seeks to analyze the difference in weight ous and 1 esophagopleural, and Strictures- 16 gastrojejunal underwent laparoscopic operation. Retrocolic technique loss between diabetics and their counter parts by looking and 1 duodeno-ileal (s/p duodenal switch). Eight patients was used in 188 patients (72%). All subsequent gastrointes- at excess percentage weight loss and change in lean body had undergone multiple endoscopic dilations at outside tinal surgical and endoscopic interventions and abdomi- mass. nal imaging studies were reviewed. Imaging studies were institutions and one stent placement prior to our evalua- METHODS: A retrospective analysis was performed for 91 CONCLUSION: Diabetic patients have less lean body mass categorized as Abdominal CT (CTAP) and non-CT imaging tion. A total of 89 were performed (range: 2 diabetic patients (BMI = 43.1 ± 0.77 kg/m2) and 510 non- before surgery and consistently have less lean body mass (including ultrasonography and fl uoroscopy). Resource uti- to 9 per patient). Mean duration of endoscopic treatment diabetic patients (BMI = 42.5 ± 0.36 kg/m2) that underwent after weight loss surgery with LAGB as compared to non- lization rates for 2006 for Americans aged 55–64 years, as was 125 days (range: 17–367 days). All patients had cov- LAGB within a span of three years using an electronic medi- diabetics. The less weight loss (%EWL) can be attributed to published by the CDC, and published Medicare benefi ciary ered metal stents placed. Stent migration occurred in 10 cal record system. Percentage excess weight loss (%EWL) and less lean body mass, but both groups increased their % lean utilization rates for 2001 were used as a reference. patients, 3 of whom required operative removal, from the percentage lean body mass (%LBM) were analyzed for patients body mass in similar amounts with signifi cant preservation colon, small bowel and Roux limb. Two patients died at RESULTS: Median patient follow-up was 7 years (range from 2007–2010. Bioimpedance was measured via two limb of lean body mass and reversal of disease processes. home with ongoing symptoms after stent removal. Of the 6–13) after RYGB, for a total number of 2,039 patient-years technique using one arm and one leg. Data was analyzed with remaining 20 patients, 13 patients required revisional sur- (pt-yrs). A total of 162 abdominal operations were per- ANOVA and * P < 0.05 is considered signifi cant. gery. 2 patients at last hospital visit had return of symptoms formed in 95 patients (36%); 58 operations in 48 patients and will undergo revisional surgery. In 5 patients stent were repairs of abdominal wall hernias and 34 were re-oper- placement seems to have been effective defi nitive treat- ations for wound complications in 21 patients. Sixty-nine ment with a mean follow up of 280 days (range 127–475 gastrointestinal operations, 46 in 32 patients (12%) were days); 2 had closure of a chronic gastrocutaneous fi stula directly attributed to the post-RYGB state. There were 132 (success 2 of 5), but only 3 of 17 patients had resolution of upper endoscopies performed in 78 patients (30%), a rate a chronic anastomotic stricture. of 65 per 1,000 pt-yrs, compared to 28 upper endoscopies CONCLUSIONS: Endoscopic therapy provides a less inva- per 1,000 pt-yrs for Americans aged 55–64. Eighteen study sive approach and in selected patients may warrant attempt patients had 32 therapeutic endoscopies for complications as defi nitive therapy. Although successful in only 5 of 22 of RYGB. One-hundred and two patients underwent a total patients (22%), stent placement relieves symptoms and of 348 CTAP; a rate of 171 per 1,000 pt-yrs, which were allows resumption of adequate oral nutrition. Stents appear evenly distributed throughout patient-years of follow-up. more successful for fi stulas rather than anastomotic stric- For non-CT imaging, 119 patients underwent 432 studies, a rate of 212 per 1,000 pt-yrs; half were in the fi rst three years tures. If unsuccessful, revisionary surgery will be necessary. Poster Abstracts Tuesday

174 175 THE SOCIETY FOR SURGERY OF THE ALIMENTARY TRACT

Combined Science RESULTS: CSP patients included whites (n = 20,809), blacks (n = 5,025), Hispanics (n = 5,455), and Asians (n = 2012 ANNUAL MEETING 3,813). By Kaplan-Meier method, Asians had the longest Tu2077 survival (Table 1; log-rank P < 0.001). Multivariate Cox regression analysis showed that race/ethnicity indepen- Assessment of Molecular-Genetic Alterations and dently predicted survival even when controlling for disease Be sure to join us for next year’s Annual Meeting— Racial/Ethnic Disparities in Outcomes in Patients extent, treatment, and SES. Analysis of colon cancer tissues mark your calendars now! with Colon Cancer from white, black, Hispanic, and Asian patients (n = 19 for V600E Wendy Lee*1, Brian Mailey1, Avo Artinyan2, Xiaoming Shen1, each group) showed highest frequency of BRAF muta- Jianming Lu 1, Marjun Philip N. Duldulao1, Julio Garcia-Aguilar1, tion in white and Hispanic tissues. KRAS codon 12 muta- Joseph Kim 1 tion was most frequent in Asian tissues. In addition, we May 18–22, 2012, San Diego, CA observed that MSI was lowest in tissues from Asians. 1 City of Hope National Medical Center, Duarte, 2CA;Michael E. Debakey Veterans Affairs Medical Center, Baylor College of Medicine, Table 1: Survival Data and Molecular Analysis for Colon Cancer Houston, TX Patients BACKGROUND: We previously observed racial/ethnic disparities in survival for patients with GI cancers in Los CSP data Frequency of genetic alterations Race/ethnicity Median survival BRAFV600E KRAS codon MSI Angeles County (LAC). However, we could not account for (months) these disparities by standard clinicopathologic factors or mutation 12 mutation socioeconomic status (SES) and access to care. Our objec- White 71 21% 26% 11% Black 52 5% 21% 26% tive was to examine potential racial/ethnic disparities in Hispanic 81 16% 32% 7% tumor biology in colon cancer patients to determine if Asian 107 0% 53% 0% molecular-genetic differences are associated with patient outcomes. CONCLUSIONS: METHODS: From the LAC Cancer Surveillance Program We identifi ed racial/ethnic disparities in (CSP), we identifi ed colon cancer patients from 1988 to survival for colon cancer patients, which parallel best the 2006. We grouped patients by race/ethnicity and compared detection of MSI in the corresponding racial/ethnic group. clinicopathologic factors; and constructed Kaplan-Meier This suggests that disparities in survival could be secondary survival curves. We then obtained archived colon cancer to differences in tumor biology between major racial/eth- specimens (n = 76) from surgical patients at our institu- nic groups. We are coordinating a multicenter study in LAC tion. DNA was extracted and evaluated for BRAF, KRAS, and to further investigate the potential effect of these genetic TP53 mutations and for microsatellite instability (MSI). alterations on clinical outcomes.

SSAT 900 Cummings Center, Suite 221-U Beverly, MA 01915 Telephone: (978) 927-8330 Facsimile: (978) 524-8890 E-Mail: [email protected] Web Site: www.ssat.com

176 SSAT 52nd Annual Meeting SCHEDULE-AT-A-GLANCE

ADMIT

FRIDAY, 5/6/2011 ADMIT 8:00 AM – 2:30 PM RESIDENTS & FELLOWS RESEAR (by invitation only) Monroe Ballro SATURDAY, 5/7/2011 8:00 AM – 5:00 PM ADMIT MAINTENANCE OF CER THE SURGEON IN THE MANAG AND ESOPHAGEAL DISEASES 8:30 AM – 10:00 AM DDW COMBINED CLINICAL SYMP (AASLD-accredited) Bariatric Surgery, Obesity, an SUNDAY, 5/8/2011 7:45 AM – 8:15 AM OPENING SESSION 8:15 AM – 9:15 AM PRESIDENTIAL PLENARY A (PLEN 9:15 AM – 10:00 AM PRESIDENTIAL ADDRESS Happy Mother’s Day 10:30 AM – 11:15 AM PRESIDENTIAL PLENARY B (PLEN 11:15 AM – 12:00 PM DORIS AND JOHN L. CAMERON G Bioengineering and Clinical A Tumor Cell Microchip 12:00 PM – 2:00 PM POSTER SESSION I (non-CME) 12:30 PM – 1:45 PM ADMIT MEET-THE-PROFESSOR Avoiding and Repairing Chole Modern Treatment of Barrett’ Metaplasia to Cancer Single Port Surgery: Beyond C 7KH 'LIÀFXOW 9HQWUDO +HUQLD  30 ²  30 &217529(56,(6 ,1 *, 685*(5< Debate 1: Nissen Fundoplicat Treatment of GERD Debate 2: Robotics Applicatio TUESDAY, 5/10/2011  6XUJHU\ ,QGXVWU\ 6SRQVRUH 5HDO $GYDQFH 7KDW :LOO %HQ 2:15 PM – 3:45 PM DDW COMBINED CLINICAL SYMP 'HÀQLWLRQ DQG 0DQDJHPHQW RI (SSAT, AGA)  30 ²  30 9,'(2 6(66,21 , 63(&,$/7< 9,' 2:15 PM – 4:45 PM STATE-OF-THE-ART CONFERENCE Personalized Medicine in Gas Potential Applications in Cl 2:15 PM – 5:15 PM PLENARY SESSION III 4:00 PM – 5:00 PM CLINICAL WARD ROUNDS: DELAYE  2) $33(1',&,7,6 ,6 685*(5 MONDAY, 5/9/2011  $0 ²  $0 9,'(2 6(66,21 ,, %5($.)$67 $  $0 ²  $0 &/,1,&$/ :$5' 5281'6 /$3$5 5(6(&7,21 :+,&+ 3$7,(17 8:30 AM – 10:00 AM SSAT PUBLIC POLICY COMMITTE Staying Alive: Strategies for Acc 9:30 AM – 11:00 AM SSAT/AHPBA JOINT SYMPOSIUM ADMIT Pancreatic Neuroendocrine Tu

May 6 – 10, 2011 0F&RUPLFN3ODFH‡&KLFDJR,OOLQRLV