JOP. J Pancreas (Online) 2016 Jan 08; 17(1):1-10.

REVIEW ARTICLE

Does Obesity influence Perioperative Outcomes of Pancreatic Resections? A Systematic Review of Literature and Meta-Analysis

Rugved V Kulkarni1, Savio George Barreto2, Surya P Rao3, Shailesh V Shrikhande4

1Department of General Surgery, Byramjee Jeejeebhoy Medical College, Pune, India 2Department of Gastrointestinal Surgery, Gastrointestinal Oncology, and Bariatric Surgery, Medanta Institute of Digestive and Hepatobiliary Sciences, Medanta, The Medicity, Gurgaon, India 3Department of Preventive and Social Medicine, Byramjee Jeejeebhoy Medical College, Pune, India 4Department of Gastrointestinal and Hepato-Pancreato-Biliary Surgical Oncology, Tata Memorial Centre, Parel, Mumbai, India

ABSTRACT Background Obesity is assuming world-wide epidemic proportions and is an independent risk factor for pancreatic cancer and Methods A thorough literature search was conducted using PubMed, Medline, Cochrane and Embase databases, using the search terms “pancreatic surgery”,surgical morbidity “pancreatoduodenectomy”, in general. It is essential “pancreaticoduodenectomy”, to determine if obesity “obesity”, influences “body perioperative mass index”, pancreatic “body mass surgical index”, outcomes. “surgical outcome”, “postoperative complications”, “complications” from January 1990 to June 2013 without restrictions. All publications from past 23 years were considered for review. Results outcomes following pancreatoduodenectomy Of the 598 articles retrieved, 22 studies were identified describing impact of obesity on perioperative . Obese individuals (with body mass index≥25) are at 2.2 times (relative risk=2.169; increaseconfidence in bloodinterval=1.572-2.994) loss with increasing more body risk mass for developing index/visceral post-operative fat area. No pancreatic evidence fistula to suggest compared impact to of non-obese obesity on individuals. other parameters Obese wasindividuals noted. Conclusionshad more wound infections (relative risk =1.69; confidence interval=1.225-2.274). 10 of 22 studies reported a significant development of delayed gastric emptying, post-pancreatectomy haemorrhage and intra-abdominal collections. Obesity contributes to an increase in intra-operative Obesity blood loss.increases the risk of post-operative pancreatic fistula and wound infections but not operating time or with the risk of morbidity [14] and mortality. Important factors related to the development of the complications INTRODUCTION

The incidence of obesity is increasing the world over include the texture of the gland (firm or soft) [15], size of [1]. Obesity is a well-known risk factor for cardiovascular such as choice of suture-material [15] and methods of the pancreatic duct [16], and surgeon-determined factors disease, hypertension, diabetes and certain malignancies of the pancreas has been suggested to be a determinant of reconstruction [17], amongst others. While fatty infiltration Recent evidence has not only suggested obesity to play a perioperative pancreatic surgical outcomes [18], this is not rolelike colon,in the ovary, development endometrium, of pancreatic prostrate cancer and breast [2-6]. uniformly reported [19-21]. also adversely affect outcomes in patients who undergo Several studies have directly examined the impact surgery for the cancer [12, 13]. [7-11], but

of a high body mass index (BMI) on outcomes following pancreatic head and neck cancers is a procedure fraught Pancreatoduodenectomy (PD), performed for pancreatic resection. However, their small sample sizes and inconsistent findings have limited the interpretation Received July 13th, 2015-Accepted August 24th, 2015 appreciate that over the next few years the number of PDs Keywords beingof their performed findings. on Despite obese this, patients it is importantis likely to forincrease us to Abbreviations and hence we need to objectively determine if obesity Mortality; surgery BMI body mass index; CI confidence interval; DGE areadelayed gastric emptying; POPF post-operative pancreatic fistula; PPH such an association is developed, it may enable us to post-pancreatectomy haemorrhage; RR relative risk; VFA visceral fat Correspondence understandinfluences morbidity the factors following leading to pancreatic these altered surgery. outcomes Once GI & HPB Surgical Oncology Convener – GI Disease Shailesh Management V Shrikhande Group and this could potentially help us to develop strategies Tata Memorial Centre aimed at minimising these risks. Parel, Mumbai, India 400 012 Phone This systematic review and meta-analysis examines the E-mail [email protected] evidence in the literature with an aim to determining the +91-22-24177173

JOP. Journal of the Pancreas - http://pancreas.imedpub.com/ - Vol. 17 No. 1 – Jan 2016. [ISSN 1590-8577] 1 JOP. J Pancreas (Online) 2016 Jan 08; 17(1):1-10. impact of obesity on perioperative surgical outcomes in [23]. The study included 14 Non-randomised experimental patients undergoing pancreatic surgery for malignant as studies reporting primary outcome and 8 studies reporting well as benign conditions of the pancreas. one or more of the secondary outcome measures. All METHODS the patients included in these studies were hospitalised

A literature search was conducted using PubMed, the heterogeneity of the eligible studies. After satisfying Medline, Cochrane and Embase databases, using the search thepatients. heterogeneity, Initially, Funnelfurther plots analysis were was created attempted. to confirm The terms “pancreatic surgery”, “pancreatoduodenectomy”, “pancreaticoduodenectomy”, “obesity”, “BMI”, “body mass index”, “surgical outcome”, “postoperative complications”, restrictedmain outcome to overall measure complications was identified and wound as post-operative infection as “complications” from January 1990 to June 2013 without otherpancreatic complications fistula. The were secondary inconsistently outcome reported. measures were any restrictions. All publications from past 23 years were considered for review. The meta-analysis followed the Binar

Inclusion Criteria y Fixed-Effect The criteria for including manuscripts in this review - Mantel Haenszel model and was carried out with the were as follows OpenMetaeffects model. [Analyst] The analysis software. was Relative carried risk, out confidenceseparately consideringintervals, and BMI tests more of significance than 30 andwere BMI derived more fr omthan fixed 25 surgery, as obesity criteria and were compared to non-obese (1) Participants – Human subjects undergoing pancreatic malignant diseases (2) Intervention – Pancreatic surgery for benign as well as in order to obtain a single estimate for decision making. Thegroup estimated (BMI<25). effect These measure study was estimates Relative were Risk. combinedBetween- (3) Comparative variables – Pre-operative BMI/Visceral fat study heterogeneity assessed using Cochran Q statistic (VF) / Visceral Fat Area (VFA) / Abdominal Wall Fat (AWF) d out for / Total Fat Area (TFA) with peri-operative variables was significant. Subgroup analysis was carrie operative(4) Outcomes variables data – –A. Overall Perioperative postoperative variables complications – Blood loss, the Relative risk among obese individuals (BMI more than operating time, lymph node number, tumor size. B. Post- primary outcome only. The Forest plots made indicates and specific complications (as defined by the authors) such 30 and BMI more than 25). RR of 1 denotes(one) indicate decreased that risk the as, pancreatic fistula (POPF), Delayed gastric emptying ofrisk complications. is equal in both the groups and >1 (more than one) complications(DGE), intra-abdominal abscess, wound infection, bleeding indicate increased risk; less than 1 (PPH), re-laparotomy, in-hospital stay, non-operative RESULTS studies. In addition, the bibliography of each manuscript Using the above search strategy, a total of 598 was(5) Types examined of study and– Prospective cross-referenced studies and forretrospective relevant publications / articles were retrieved of which 22 studies publications. (Figure 1 – Quorum Chart) Exclusion Criteria impact of obesity on perioperative outcomes following PD [12, 13, 20, 21, 24-41]. Table were 1 lists identified the various describing studies the by Review articles, letters/comments and editorials were their level of evidence. excluded. search were assessed for eligibility for inclusion in this systematicAbstracts review. of citations The data identified was arranged from the in literature a pre- discussion between the authors. specified spread sheet. Any disagreement was resolved by Quality Analysis Since all the studies included in the analysis were non- randomised studies, in order to objectively assess the quality of the studies included we used the Methodological tool for the evaluation of non-randomised studies [22]. StudiesIndex for with Nonrandomized 12 or more points Studies were (MINORS) considered – a validated as high quality and were included in the analysis. Statistical Analysis This study was performed in line with the recommendations of the Preferred Reporting Items for Figure 1. Quorum Chart.

Systematic reviews and Meta-Analysis (PRISMA) Statement JOP. Journal of the Pancreas - http://pancreas.imedpub.com/ - Vol. 17 No. 1 – Jan 2016. [ISSN 1590-8577] 2 JOP. J Pancreas (Online) 2016 Jan 08; 17(1):1-10.

Quality Assessment between patients with a normal or low BMI and those Using the MINORS tool, all 22 studies had a score of (Normal BMI group: 9.5-19 vs. (Table 2) and were included in the analysis. with an elevated BMI above normal [27, 28, 31, 34, 35] Obese group: 11-20; p-not morePatient than and 16 Surgical Demography total number of harvested lymph nodes were reported. significant). In all 5 studies, no significant differences in the Nature of Pancreatic Disease: Of the 22 studies, 80% 10,984 patients derived from retrospective analysis of of pancreatic resections were performed for malignancy prospectivelyThe analysis maintained was based databases. on a pooled Table sample 3 lists size the of (Table 4). 10 studies considered only malignant pancreatic various surgical procedures performed. The indications for the surgical procedures are provided in (Table 4). included subjects exclusively with pancreatic ductal disease [20, 26-28, 30, 31, 34-37], of which 9 studies Peri-Operative Outcomes included patients with benign and malignant pancreatic Operating Time: diseaseadenocarcinoma [12, 13, 21, [20, 25, 27, 29, 28,32, 30,33, 38-41] 31, 34-37]. while 11in a studies single increase in operative time associated with increasing While 9 out of 22 studies reported no study,Of these,the nature 4 studies of disease compared was not post-operative specified [24]. surgical group: 210–443 min vs. outcome to the nature of the pancreatic disease [29, 33, BMI/VFA [12, 20, 25, 27, 29, 31, 34-36] (Normal BMI Obese group: 402–464 min; p-not increasing BMI [21, 24, 28, 32, 38] (Normal BMI group: [29, 39]. Park et al. significant}, 5 studies reported an increase in OT with vs. high34, 39] in cancerof which of 2non-pancreatic studies compared origin, occurrence namely Ampulla, of POPF 8 other studies, however, lacked data on the association [29] concluded that POPF incidence was 185-363 min Obese group: 438-723 min; p<0.05). occurrence was independent of the nature of primary 39-41]. pancreaticbile duct andpathology. duodenu)m Hashimoto (p=0.015). et al. [39], However, too, failed POPF to between obesity and operating time [13, 26, 30, 33, 37, Peri-Operative Blood Loss: 10 out of 22 studies reported note an increased incidence of POPF depending on the 2 pathologyTsai et al.(p=0.383). [34] found that obesity (BMI > 30kg/m 300–1020a significant ml increase vs. in blood loss with increasing BMI/ whileVFA [20, 10 21, studies 25, 27-29, lacked 31, data32, 34, about 35] (Normal the association BMI group: of ) was Obese group: 650–1255 ml; p<0.05) associated with better disease-specific survival rate (HR – Stage of the Pancreatic Malignancy: Of the 22 studies et al. [38] and House et al. 0.73, p < 0.01). anperioperative increase in blood blood loss loss with and increasing BMI[12, 13, BMI 24, (Normal 26, 30, BMI 33, 37, 39-41]. Su [36] did not find vs. included, 6 studies mentioned about the stage of the pancreatic malignancy [20, 27, 28, 31, 34, 37]. In all these group: 595–719 ml Obese group: 686–881 ml; p-not amongststudies, there different was predominantlyBMI groups and Stage found II disease no sign (75.8% significant).Total Lymph Nodes Harvested: 5 of the 22 studies reported on the differences in total lymph nodes harvested difference.- 97.3%). 3 studies [20, 28, 34] compared cancer stage ificant Table 1. Level of evidence of the various studies Article No. of patients Type of Study Level of Evidence Duration in years Cases / Year Pausch T et al. 2012 [20] 408 Prognostic-Retrospective study 2 9 45.3 Balentine et al. 2011 [24] 201 Prognostic-Retrospective study 2 8 25.1 Tranchart et al. 103 Prognostic-Retrospective study 2 1.5 Hwang et al. 2011 [25] 159 Prognostic-Retrospective study 2 2012 [26] 68.6 et al. 2009 [21] 240 Prognostic-Retrospective study 2 32 6 26.5 Noun et al. 2008 [12] 92 Prognostic-Retrospective study 2 13.1 Williams 7.5 et al. 285 Prognostic-Retrospective study 2 7 Khan et al. 2010 [28] Prognostic-Retrospective study 2 22.5 Fleming 2009 [27] 7.8 36.5 Park et al. 2011 [29] 181 Prognostic-Retrospective study 2 24.1 586 26 et al. 2010 [32] Prognostic-Prospective study 2 7.5 Gaujoux et al. 2011 [31] 328 Prognostic-Retrospective study 2 5 Shimizu 317 6.7 47.3 Dandona et al. 2011 [30] 355 Prognostic-Retrospective study 2 14 25.3 65.6 Tsai et al. 2010 [34] Prognostic-Retrospective study 2 10 Balentine et al. Prognostic-Retrospective study 2 9 795 79.5 Greenblatt et al. 2011 [33] 4945 Prognostic-Retrospective study 2 4 2010 [37] 61 6.7 House et al. Prognostic-Retrospective study 2 5 1236.2 Benns et al. 2009 [35] Prognostic-Retrospective study 2 2007 [36] 356 71.2 Su et al. 2009 [38] 101 Prognostic-Retrospective study 2 2.5 40.4 306 17.5 17.4 Rosso et al. 2009 [13] 111 Prognostic-Retrospective study 2 2 55.5 Hashimoto et al. 2010 [39] Prognostic-Retrospective study 2 20.5 et al. 2009 [41] 85 Prognostic-Retrospective study 2 4 21.2 507 24.7 et al. 2008 [40] Prognostic-Retrospective study 2 14 33 Akizuki Ferrone 462 JOP. Journal of the Pancreas - http://pancreas.imedpub.com/ - Vol. 17 No. 1 – Jan 2016. [ISSN 1590-8577] 3 JOP. J Pancreas (Online) 2016 Jan 08; 17(1):1-10. Score 19 19 18 17 20 19 16 20 18 19 19 21 19 17 19 16 17 17 19 19 18 18 Adequate Statistics 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Prospective Sample Size Calculation 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 Group Matching 2 2 2 1 2 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Contemporary Contemporary Groups 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Adequate Control Group 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Loss Of Follow Up Less 5 Than % 0 0 0 0 0 0 0 2 0 0 0 2 0 0 0 0 0 0 0 0 0 0 Follow Up Period Appropriate For Study 1 1 1 2 2 1 1 2 1 2 2 2 2 0 1 1 1 1 2 2 0 1 Unbiased Unbiased Outcome Evaluation 2 2 2 2 2 2 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Reported End-Points 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Prospective Data Collection 2 2 2 2 2 2 2 2 2 2 2 1 2 2 2 1 1 0 2 2 2 2 Consecutive Consecutive Patients 2 2 1 0 2 2 1 1 1 1 1 2 1 1 1 0 1 1 1 1 2 1 Clearly Stated Aim 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Year 2007 2008 2008 2009 2009 2009 2009 2009 2009 2010 2010 2010 2010 2011 2011 2011 2011 2011 2011 2011 2012 2012 [39] [33] [26] [37] [24] [30] [21] [31] [40] [27] [41] [32] [20] [25] [36] [35] [13] [12] [28] [29] [34] [38] Assessment of Articles as per MINORS criteria Assessment of Articles as per MINORS House et al. House Ferrone et al. et al. Noun Benns et al. Su et al. Rosso et al. Williams et al. Fleming et al. Akizuki et al. Tsai et al. Balentine et al. Khan et al. et al. Hashimoto Balentine et al. Greenblatt et al. et al. Hwang Park et al. Shimizu et al. Gaujoux et al. Dandona et al. Pausch et al. Tranchart et al. Article Table 2. Table

JOP. Journal of the Pancreas - http://pancreas.imedpub.com/ - Vol. 17 No. 1 – Jan 2016. [ISSN 1590-8577] 4 JOP. J Pancreas (Online) 2016 Jan 08; 17(1):1-10.

Table 3. Types of Pancreatic Resections and their Pooled Incidence Type of Surgery Poolede Incidence 9.4-54.3%; p<0.05) while in three studies [26, 29, 32] an A. Classic PD increasing VFA was found to correlate with the occurrence Pancreatoduodenectomy (PD) 10352/10984 (94.2%) B. Pylorus-preserving PD of clinically significant POPF. A BMI >25 was determined to 6030/10984 (54.9%) et al. [13] actually found that a BMI >25 Distal Pancreatectomy 4322/10984 (39.3%) be an independent risk factor for POPF occurrence [13, 25, Sub-total Pancreatectomy 601/10984 (6.4%) 26, 29, 32]. Rosso Total Pancreatectomy 2 17/10984 (0.15%) was linked to risk of clinically significant POPF. 14/10984 (0.13%) Indications for the Pancreatic Resections VF thickness, rather than BMI (>30kg/m ), was found Table 4. Conversely, 4 other studies, found no correlation Indicationt Proportion to be associated with the development of POPF [36]. Malignant disease 8832/10984 (80.4%) vs. Pancreatic adenocarcinoma between the development of POPF and either abdominal Ampullary adenocarcinoma 313 BMI [21, 38, 41] (Normal BMI group: 4-18.8% vs. Obese 3975 fat / VF (17% 11%, p-not significant) [24] or elevated Bile duct cholangiocarcinoma Intraductal papillary mucinous neoplasm 96 group:In 8 2-13%; studies, p-not no data significant). on the association of obesity and Neuroendocrine tumors 131 63 Duodenal carcinoma 25 Other malignant 4229 POPF occurrence could be determined [20, 27, 28, 30, 31, Benign Disease 1671/10984 (15.3%) 33, 35, 37]. Pancreatic tumors 192 obese individuals is shown in Figures 3-5 Biliary tumors 4 consideredRelative riskBMI of > developing30 as obese POPF while among others obese considered and non- Others BMI>25 as obesity criteria. Hence the data was. Few regrouped authors Others 481/10984 (4.3%) 1475 separately from those defining obesity as BMI > 25. The There was no direct comparison of stage of pancreatic plotand in the Figure studies 3 defining obesity as BMI >30 analysed malignancy to peri-operative surgical outcomes in any of these studies. indicates the relative risk of POPF among obese (BMI ≥30) and non-obese individuals (BMI < 25). Morbidity Rate: The overall relative risk rate was 1.61 (CI=1.259-2.062) was reported across studies. indicating that obese individuals are 1.6 times more at risk An overall morbidity rate of 5 - 67.6% The forest plot in Figure 4 indicates the relative risk of Benns et al. of developing POPF compared to non-obese individuals. orbidity in patients with a BMI [35] of reportedmore than a 30kg/m significantly2 (Severity increased score post-operative pancreatic fistula among obese (BMI ≥ 25) risk (p=0.01) and severity of post-operative etm al. [20], on and non-obese (BMI<25) individuals. Obese individuals the other hand, reported that pre-operative weight-loss (with BMI ≥ 25) are at a 2.2 times (relative risk=2.169; - obese: 1.6, non-obese: 1.2, p=0.02). Pausch was CI=1.572-2.994) more risk for developing POPF compared associated with an increased post-operative morbidity to non-obese individuals. It also emphasizes that the post- (> 10% of body weight), rather than obesity, per se, operative risk of POPF becomes more prominent when operative morbidity associated with an increase in BMI the cut-off BMI value is considered to be ≥ 25. Hence, a (p<0.03). 9 of the 22 studies reported no increase in post- individualscombined plot (Figure was prepared 5). The tooverall find outrisk the ratio change shows in the risk ratio of POPF among obese and non-obese 11-40.2% vs. while[12, 20, 11 21, out 24, of 27, 22 28, studies 31, 34, lacked 36] (Normal data regarding BMI group: the 1.808Delayed (CI=1.487-2.198). Gastric Emptying: An overall DGE incidence rate Obese group: 5-67.6 %; p-not significant) of 4- 53% was reported across studies. et al. [33], too, reportedinfluence that of BMIa BMI on of post-operative >25kg/m2 was morbiditypredictive [25,of post- 26, operative29, 30, 32, morbidity 36-38, 40, in 41,patients 43]. Greenblattundergoing PD (odds ratio otherOnly studies, one studyno correlation noted a was significant noted between association the developmentbetween increasing of DGE VFA and andeither DGE increasing [29] (p=0.014). BMI [12, In 25, 8 (OR)-1.27The overall 90%CI). post-operative complication rate among vs. Obese obese and non-obese individuals is shown in Figure 2. The overall complication rate among obese individuals is 1.194 34, 36, 39, 41] (Normal BMI group: 4 – 42% times more than that of non-obese individuals. However, group: 10.5–53%; p-not significant), increasing VFA 13 studies lacked data about association between DGE ant [24] or increasing VF/ intra-abdominal fat [32]. the risk was not found to be statistically signific Post-operative Pancreatic Fistula: occurrence and increasing BMI [13, 20, 21, 26-28, 30, 31, (RR=1.194; CI=1.069-1.334, p=0.647). 33,Post-Pancreatectomy 35, 37, 38, 40]. Haemorrhage: An overall PPH incidence rate of 2 - 54.3% was reported across studies. In incidence rate of 0 – 10.5% was reported across studies. An overall POPF was noted between a raised BMI and the occurrence of Data on the relation between PPH and obesity was 7 studies [12, 25, 26, 29, 34, 39, 40], a significantvs. Obese association group: available in only four of the 22 studies. No increased risk

POPF(NormalJOP. Journal of the Pancreas BMI group: - http://pancreas.imedpub.com/ 2.9-33.6% - Vol. 17 No. 1 – Jan 2016. [ISSN 1590-8577] 5 JOP. J Pancreas (Online) 2016 Jan 08; 17(1):1-10.

Figure 2.

Forest Plot of Relative Risk for Overall Complications among Obese Individuals.

Figure 3.

Forest Plot of Relative Risk for Developing Pancreatic Fistula in Studies Defining Obesity as BMI>30.

Figure 4.

Forest Plot of Relative Risk for Developing POPF in Studies Defining Obesity as BMI>25.

Figure 5. of PPH wasForest noted Plot of toRelative correlate Risk for with Pancreatic increasing Fistula CombiningBMI, intra- All Studies. et al. incidence of post-operative intra-abdominal abscess in the vs. Obese group: obeseWilliams group (14% vs. [21] reported a significantly higher abdominal fat / VF or VFA in all the four studies [12, 24, with a normal BMI. Similarly, Su et al. [38], too, found that 29, 32] (Normal BMI group: 0-6.8% a BMI >25 was an independent 7%; p=0.05) risk as factorcompared for developingto patients 1.2–10.5%;Post-Operative p-not Intra-Abdominalsignificant). Abscess: An overall post-operative intra-abdominal abscess incidence rate of post-opera vs. 3.9%,

tive infectious complications (7.9% 1.7 – 40% was reported across studies. OR=6.5, p=0.005). Four other studies, however, reported JOP. Journal of the Pancreas - http://pancreas.imedpub.com/ - Vol. 17 No. 1 – Jan 2016. [ISSN 1590-8577]

6 JOP. J Pancreas (Online) 2016 Jan 08; 17(1):1-10.

Figure 6. no associationForest Plot Showingbetween Relative occurrence Risk of Wound of Infectionpost-operative among Obese and usedNon-Obese data Group. from the American College of Surgeons National

vs. Obese database. In the 21 studies, the number of resections per intra-abdominal abscess and increasing BMI, VFA or VF Surgical Quality Improvement Program (ACS NSQIP) [12, 24, 32, 41] (Normal BMI group: 1.7–30% group:2–40%;Pausch et al. p-not [20], significant). instead reported that patients with yearDISCUSSION varied from 7 to 80. Obesity is on the rise the world over and has assumed vs. larger AWF had fewer post-operative intra-abdominal epidemic proportions [44]. Moreover, a pooled analysis of abscess compared to those with smaller AWF (7.1% 1.7%;In 15 p=0.047). studies no data comparing the association between an independent risk factor for pancreatic cancer [45]. A post-operative intra-abdominal abscess and increasing BMI previous7 prospective meta-analysis cohort studies on the indicated impact ofthat BMI a high on outcomes BMI was of pancreatic surgery indicated that obesity increased the couldPost-Operative be determined Wound [13, 25-31, Infection 33-37, :39, An 40]. overall post- operative wound infection incidence rate of 3.9 - 15.8 % complexity of surgery and the development of POPF but was reported across studies. perioperativedid not significantly outcomes. influence However, other studies parameters have shown [46]. This [29, 32]meta-analysis that it is the solely visceral looked fat area, at BMImore and than its the influence BMI, which on post-operative wound infection with an increasing BMI/ While nine studies reported no increased incidence of there may not be a direct correlation between BMI and 3.9-11% vs. visceralcan potentially fat [29, influence32]. Thus, outcomes.till clearer Moreevidence importantly, emerges VFA [12, 21, 24, 25, 29, 32, 34, 35, 41] (Normal BMI group: House et al. on which constitutes a better parameter for assessing the Obese group: 5–15.8%; p-not significant), impact of obesity on surgical variables, it would be prudent [36] reported that BMI>30 was an independent predictor of post-operative wound infection (hazards ratio: 1.1; p=0.03) with a significant correlation between Given an appreciation of the scenario that we will A higher incidence of wound infections was noted in to at least assess both BMI and VF/VFA. VF thickness and post-operative wound infection. be seeing an increase in the number of obese patients obese individuals as compared to those who are non-obese undergoing surgery for pancreatic tumours, it is essential (Figure 6 ).

Non-Operative; relative Complications risk=1.69 (CI=1.225-2.274)et al. [32] perioperative outcomes. This information can aid the to determine if, and how, obesity would influence : Shimizu reported a significant increase in post-operative development of strategies to minimize any such negative pulmonary complications in patients with a high VFA on influences of obesity on pancreatic surgical outcomes. multivariateIn-Hospital analysis Stay: (p=0.04, OR-4.246).et al. [29] reported no poor outcomes following pancreaticoenteric anastomosis Fatty infiltration of the pancreas as a risk factor for (28.2 vs. While Park increased length of stay amongst patients with a high VFA overallhas been complications well appreciated following [18]. Thispancreatic review surgery, confirms there that 26 days; p-NS), two out of vs.22 Obesestudies group: reported 9.5– a while obesity does not significantly increase the risk for significantly increased hospital stay in obese patients [12, 21](Normal BMI group: 8-17 days pital is certainly an increased risk of POPF in obese patients or 23.1 days; p<0.05). In 6 of the 22 studies, it was noted that patients with an elevated VFA [12, 25, 26, 29, 34, 39, 40]. 38](Normalthere existed BMI no significantgroup: 9–23 difference days vs. Obesein length group: of hos 10–25 In fact, a BMI ≥ 25 was determined to be an independent stay among different BMI groups [20, 25, 27, 28, 34, 35, bestrisk factormethod for for POPF reconstructing occurrence the [13, pancreaticoenteric 25, 26, 29, 32]. Future randomised studies are needed to determine the days;Centre p-not Surgical significant). Volume: 21 of the 22 articles were based anastomosis in PD depending on the texture of the on single institution data [12, 13, 20, 21, 24-32, 34-41] except for the publication by Greenblat et al. [33] which risk of wound infections following pancreatic surgery. pancreatic gland [17]. Obesity also results in an increased

JOP. Journal of the Pancreas - http://pancreas.imedpub.com/ - Vol. 17 No. 1 – Jan 2016. [ISSN 1590-8577]

7 JOP. J Pancreas (Online) 2016 Jan 08; 17(1):1-10.

Obesity is not associated with a prolonged duration of surgery but may contribute to increased intra-operative having recorded more than 5 pancreatic resections per blood loss and hence the need for blood transfusions [20, studies that can be defined as “high-volume” centres

CONCLUSIONSyear [49] (7-80 resections per year). 21, 25, 27-29, 31, 32, 34, 35]. On closer analysis, VFA rather [29,than 32] merely following an elevated pancreatic BMI, surgery. influences Thus, the meticulous technical currently available comparing perioperative outcomes techniquedifficulty ofmay the help procedure reduce the and blood the increasedloss in obese blood patients loss followingThis study pancreatic summarizes surgery the highestbetween quality individuals of evidence with possibly at the expense of an increased operative time. low to normal BMI and those with an elevated BMI. The Obesity certainly does not appear to interfere with the extent of dissection in patients undergoing pancreatic and wound infections but does not increase operating evidence suggests that obesity increases the risk of POPF the lymph nodes harvested in these patients (9.5-19 vs. 11- abdominal collections. Obesity contributes to an increase surgery as per the lack of significant difference between intime intra-operative or influence the blood development loss but ofdoes DGE, not PPH impair and intra- the performance of a standard lymphadenectomy. 20; p–not significant) [27, 28, 31, 34, 35]. for obese patients to develop pulmonary complications An important finding in this review is the potential Conflict of interest of clinical pathways prior to-, as well as, following pancreatic[32]. These surgery findings focusing stress the on importance the incorporation of development of chest physiotherapy to help reduce the development of such problems which may arise due to poor breathing efforts All the authors have no conflicts of interest associated with obesity and also post-operative pain. References Obesity or increased visceral fat was not found 1. S Gortmaker S. The global obesity pandemic: Shaped by global drivers and winburn B, Sacks G, Hall K, McPherson K, Finegood D, Moodie M, following pancreatic surgery. Considering that the data includedto influence in the the review development has emerged of other from complications high-volume local2. Carmichael environments. AR. LancetObesity 2011; and 378:804-814.prognosis of breast[PMID: cancer.21872749] Obes Rev pancreatic centres, it can be well appreciated that the lack of effect of obesity on the vast majority of peri- 2006;3. Dignam 7:333-340. JJ, Polite [PMID: BN, 17038127] Yothers G, Raich P, Colangelo L, O'Connell MJ,

Wolmark N. Body mass index and outcomes in patients who receive aoperative valuable variables lesson for may new, be as a reflectionwell as, low-volume of standardization centres adjuvant chemotherapy for colon cancer. J Natl Cancer Inst 2006; in surgical technique [47] pre-existent in these centres – 98:1647-1654.4. [PMID: 17105987] who perform pancreatic resectional surgery. The effect of with increased risks of prostate cancer metastasis and death after initial obesity on enhanced recovery protocols following PD is just Gong Z, Agalliu I, Lin DW, Stanford JL, Kristal AR. Obesity is associated beginning to be understood with a recent study suggesting cancer diagnosis in middle-aged men. Cancer 2007; 109:1192-1202. the need to tailor the existing enhanced recovery protocols [PMID:5. 17311344] following PD to patients with raised BMI and respiratory L, Hogdall CK, Blaakaer J, Kjaer SK. Smoking and overweight: Negative co-morbidities [48]. prognosticKjaerbye-Thygesen factors in stage A, Frederikseniii epithelial K,ovarian Hogdall cancer. EV, Glud Cancer E, Christensen Epidemiol One of the short-comings of the analysis is the inherent Biomarkers Prev 2006; 15:798-803. [PMID: 16614126] heterogeneity in the studies available. Studies included RR. Treatment effects, disease recurrence, and survival in obese women different types of pancreatic resections (PD, distal with6. von early Gruenigen endometrial VE, carcinomaTian C, Frasure : A gynecologic H, Waggoner oncology S, Keys group H, Barakat study.

Cancer 2006; 107:2786-2791. [PMID:ker-Thurmond 17096437] K, Thun MJ. Overweight, differentpancreatectomy, parameters total for andassessing subtotal the impact pancreatectomy), of body fat obesity, and mortality from cancer in a prospectively studied cohort of varied indications (malignant and benign disease) and 7. Calle EE, Rodriguez C, Wal selective reporting of data in literature. Any attempt to sub- u.S.8. Adults. N Engl J Med 2003; 348:1625-1638. [PMID: 12711737] dividecontent these (BMI, into VF, outcomes VFA). This, for however,each surgery, reflects indication the non- or anthropometric factors and risk of pancreatic cancer: Results from Hanley AJ, Johnson KC, Villeneuve PJ, Mao Y. Physical activity, parameter assessed would not have been possible. The strength of this study rests in an exhaustive the canadian enhanced cancer surveillance system. Int J Cancer 2001; 94:140-147.9. [PMID: 11668489] literature review to identify manuscripts that compared Overall obesity, abdominal adiposity, diabetes and cigarette smoking in outcomes following pancreatic surgery between patients relationLarsson to the SC, risk Permert of pancreatic J, Hakansson cancer N, in Naslund two swedish I, Bergkvist population-based L, Wolk A. with a normal BMI and those with an elevated BMI. Due attention to the heterogeneity of data was taken cohorts.10. Br J Cancer 2005; 93:1310-1315. [PMID: 16288300] into consideration when performing the meta-analysis. Moreover, manuscripts included were subjected to an Li D. Diabetes and pancreatic cancer. Mol Carcinog 2012; 51:64-74. [PMID:11. 22162232] termine the quality of the manuscripts. The data included is from Michaud DS, Giovannucci E, Willett WC, Colditz GA, Stampfer MJ, objective (MINORS) assessment tool to de Fuchs CS. Physical activity, obesity, height, and the risk of pancreatic cancer. JAMA 2001; 286:921-929. [PMID: 11509056] JOP. Journal of the Pancreas - http://pancreas.imedpub.com/ - Vol. 17 No. 1 – Jan 2016. [ISSN 1590-8577] 8 JOP. J Pancreas (Online) 2016 Jan 08; 17(1):1-10.

12. 29. Park CM, Park JS, Cho ES, Kim JK, Yu JS, Yoon DS. The effect of visceral Naderi S, et al. The impact of obesity on surgical outcome after Noun R, Riachy E, Ghorra C, Yazbeck T, Tohme C, Abboud B, fat mass on pancreatic fistula after pancreaticoduodenectomy. J Invest pancreaticoduodenectomy.13. JOP 2008; 9:468-476. [PMID: 18648138] Surg30. 2012; 25:169-173. [PMID: 22583013] M, Jaeck D, et al. The role of "fatty pancreas" and of bmi in the occurrence Rosso E, Casnedi S, Pessaux P, Oussoultzoglou E, Panaro F, Mahfud patientsDandona undergoing M, Linehan pancreaticoduodenectomy D, Hawkins W, Strasberg for S, Gaopancreatic F, Wang-Gillam cancer. A. Influence of obesity and other risk factors on survival outcomes in of pancreatic fistula after pancreaticoduodenectomy. J Gastrointest Surg 14. Shukla P, Barreto S, Shrikhande S. The evolution of 2009; 13:1845-1851. [PMID: 19639369] Pancreas31. 2011; 40:931-937. [PMID: 21747317] Klimstra DS, et al. Impact of obesity and body fat distribution on survival Gaujoux S, Torres J, Olson S, Winston C, Gonen M, Brennan MF, pancreatoduodenectomy. Hepatogastroenterology 2011; 58:1409-1412. after pancreaticoduodenectomy for pancreatic adenocarcinoma. Ann [PMID:15. 21937418] anastomoses after pancreaticoduodenectomy: Do we need further Shrikhande SV, Qureshi SS, Rajneesh N, Shukla PJ. Pancreatic Surg32. Oncol 2012; 19:2908-2916. [PMID: 22411205] studies? World J Surg 2005; 29:1642-1649. [PMID: 16311866] Shimizu A, Tani M, Kawai M, Hirono S, Miyazawa M, Uchiyama K, ductal stents after pancreatic resections improve outcomes? Pancreas Yamaue H. Influence of visceral obesity for postoperative pulmonary 16. Shukla PJ, Barreto SG, Fingerhut A. Do transanastomotic pancreatic complications after pancreaticoduodenectomy. J Gastrointest Surg 2011; 15:1401-1410.33. [PMID: 21567293] 2010; 39:561-566. [PMID: 20562577] morbidityGreenblatt and mortality DY, Kelly KJ,after Rajamanickam pancreaticoduodenectomy. V, Wan Y, Hanson Ann T, RettammelSurg Oncol 17. Shukla PJ, Barreto SG, Fingerhut A, Bassi C, Buchler MW, Dervenis C, R, Winslow ER, Cho CS, et al. Preoperative factors predict perioperative Gouma D, Izbicki JR, Neoptolemos J, Padbury R, Sarr MG, Traverso W, Yeo CJ, Wente MN. Toward improving uniformity and standardization in the 2011;34. Tsai 18:2126-2135. S, Choti MA, [PMID: Assumpcao 21336514] L, Cameron JL, Gleisner AL, Herman reporting of pancreatic anastomoses: A new classification system by the international study group of pancreatic surgery (isgps). Surgery 2010; 18. Mathur A, Pitt HA, Marine M, Saxena R, Schmidt CM, Howard TJ, survival following pancreaticoduodenectomy for pancreatic cancer: A 147:144-153. [PMID: 19879614] JM, Eckhauser F, et al. Impact of obesity on perioperative outcomes and

Nakeeb A, et al. Fatty pancreas: A factor in postoperative pancreatic large single-institution study. J Gastrointest Surg 2010; 14:1143-1150. fistula.19. Lermite Ann Surg E, Pessaux 2007; 246:1058-1064. P, Brehant O, Teyssedou [PMID: 18043111] C, Pelletier I, Etienne S, [PMID:35. 20431978] obesity on outcomes following pancreatectomy for malignancy. Ann Surg after pancreaticoduodenectomy with pancreaticogastrostomy. J Am Coll Benns M, Woodall C, Scoggins C, McMasters K, Martin R. The impact of Arnaud JP. Risk factors of pancreatic fistula and delayed gastric emptying Oncol 2009; 16:2565-2569. [PMID: 19557479] Surg20. 2007; 204:588-596. [PMID: 17382217] M, Gonen M, et al. Preoperative predictors for complications after Grenacher L, et al. Cachexia but not obesity worsens the postoperative pancreaticoduodenectomy:36. House MG, Fong Y, Arnaoutakis Impact of DJ, bmi Sharma and body R, Winston fat distribution. CB, Protic J outcomePausch after T, Hartwigpancreatoduodenectomy W, Hinz U, Swolana in pancreatic T, Bundy cancer. B, Hackert Surgery T,

Gastrointest Surg 2008; 12:270-278. [PMID: 18060467] 2012;21. 152: 81-88. [PMID: 22770957] Petersen N, Berger D. Intra-abdominal fat predicts survival in pancreatic T, Doria C, et al. Impact of obesity on perioperative morbidity and 37. Balentine C, Enriquez J, Fisher W, Hodges S, Bansal V, Sansgiry S, Williams TK, Rosato EL, Kennedy EP, Chojnacki KA, Andrel J, Hyslop cancer.38. J Gastrointest Surg 2010; 14:1832-1837. [PMID: 20725799] mortality after pancreaticoduodenectomy. J Am Coll Surg 2009; 208:210- 217.22. [PMID: 19228532] Su Z, Koga R, Saiura A, Natori T, Yamaguchi T, Yamamoto J. Factors Slim K, Nini E, Forestier D, Kwiatkowski F, Panis Y, Chipponi influencing infectious complications after pancreatoduodenectomy. J 39. J. Methodological index for non-randomized studies (minors): Hepatobiliary Pancreat Sci 2010; 17:174-179. [PMID: 19517055] failure and delayed gastric emptying after pancreatoduodenectomy Development and validation of a new instrument. ANZ J Surg 2003; Hashimoto Y, Traverso LW. Incidence of pancreatic anastomotic 73:712-716.23. [PMID: 12956787] Clarke M, et al. The prisma statement for reporting systematic reviews and Liberatimeta-analyses A, Altman of D,studies Tetzlaff that J, Mulrowevaluate C, health Gøtzsche care PC, interventions: Ioannidis JP, in 507 consecutive patients: Use of a web-based calculator to improve homogeneity40. of definition. Surgery 2010; 147:503-515. [PMID: 20018335] Ferrone CR, Warshaw AL, Rattner DW, Berger D, Zheng H, Explanation and elaboration. Ann Intern Med 2009; 151: 65-94. [PMID:24. 19622512] Rawal B, Rodriguez R, et al. Pancreatic fistula rates after 462 distal does not increase complications following pancreatic surgery. J Surg Res pancreatectomies: Staplers do not decrease fistula rates. J Gastrointest Balentine CJ, Enriquez J, Cruz G, Hodges S, Bansal V, Jo E, et al. Obesity Surg41. 2008; 12:1691-1697. [PMID: 18704597] T, Hirata K Reconsideration of postoperative oral intake tolerance after 2011;25. Hwang 170:220-225. HK, Park [PMID: JS, Park 21514600] CI, Kim JK, Yoon DS. The impact of body pancreaticoduodenectomy:Akizuki E, Kimura Y, Nobuoka Prospective T, Imamura consecutive M, Nagayama analysis ofM, delayedSonoda mass index on pancreatic fistula after pancreaticoduodenectomy in asian gastric emptying according to the isgps definition and the amount of patients on the basis of asia-pacific perspective of body mass index. JOP dietary42. intake. Ann Surg 2009; 249:986-994. [PMID: 19474680] 2011; 12:586-592. [PMID: 22072248] P, et al. Preoperative ct scan helps to predict the occurrence of severe Wright JG, Swiontkowski MF, Heckman JD. Introducing levels of 26. Tranchart H, Gaujoux S, Rebours V, Vullierme MP, Dokmak S, Levy evidence43. to the journal. J Bone Joint Surg Am 2003; 85:1-3. [PMID: 12533564] pancreatic fistula after pancreaticoduodenectomy. Ann Surg 2012; Sakuramoto S, Sasako M, Yamaguchi T, Kinoshita T, Fujii M, 256:139-145. [PMID: 22609844] Nashimoto A, Furukawa H, et al. Adjuvant chemotherapy for gastric 27. Fleming JB, Gonzalez RJ, Petzel MQ, Lin E, Morris JS, Gomez H, Lee JE, et cancer with s-1, an oral fluoropyrimidine. N Engl J Med 2007; 357:1810- al. Influence of obesity on cancer-related outcomes after pancreatectomy 1820.44. [PMID: 17978289] to treat pancreatic adenocarcinoma. Arch Surg 2009; 144:216-221. James WP. The epidemiology of obesity: The size of the problem. J [PMID:28. Khan 19289659] S, Sclabas G, Reid-Lombardo K, Sarr MG, Nagorney D, Kendrick Intern45. Med 2008; 263:336-352. [PMID: 18312311] outcome in patients who undergo pancreatoduodenectomy for ML, Que FG, et al. Does body mass index/morbid obesity influence of pancreaticJiao L, Berrington cancer: A pooled de Gonzalez study of A, seven Hartge prospective P, Pfeiffer cohorts. RM, ParkCancer Y, Freedman DM, Gail MH, et al. Body mass index, effect modifiers, and risk pancreatic adenocarcinoma? J Gastrointest Surg 2010; 14:1820-1825. [PMID: 20676790] Causes Control 2010; 21:1305-1314. [PMID: 20383573]

JOP. Journal of the Pancreas - http://pancreas.imedpub.com/ - Vol. 17 No. 1 – Jan 2016. [ISSN 1590-8577] 9 JOP. J Pancreas (Online) 2016 Jan 08; 17(1):1-10.

Ramsey AM, Martin RC. Body mass index and outcomes from 48. pancreatic resection: A review and meta-analysis. J Gastrointest Surg discharge after pancreatoduodenectomy - what helps and what prevents? 46. Chaudhary A, Barreto S, Talole S, Singh A, Perwaiz A, Singh T. Early

2011; 15:1633-1642. [PMID: 21484490] Pancreas49. 2015; 44:273-278. [PMID: 25479587] Talole S, Sirohi B, et al. Evolution of pancreatoduodenectomy in a tertiary AN. Relationship between hospital volume and late survival after 47. Shrikhande SV, Barreto SG, Somashekar BA, Suradkar K, Shetty GS, Birkmeyer JD, Warshaw AL, Finlayson SR, Grove MR, Tosteson 10455881] cancer center in india: Improved results from service reconfiguration. pancreaticoduodenectomy. Surgery 1999; 126:178-183. [PMID: Pancreatology 2013; 13:63-71. [PMID: 23395572]

JOP. Journal of the Pancreas - http://pancreas.imedpub.com/ - Vol. 17 No. 1 – Jan 2016. [ISSN 1590-8577] 10