Delayed Sleeve Gastrectomy Following Liver Transplantation: a 5-Year Experience Mackenzie C

ORIGINAL ARTICLE Morris et al.

Delayed Sleeve Gastrectomy Following Liver Transplantation: A 5-Year Experience Mackenzie C. Morris,1,2 Andrew D. Jung,1,2 Young Kim,1,2 Tiffany C. Lee ,1,2 Tiffany E. Kaiser,1,2 Jonathan R. Thompson,1,2 Khurram Bari,2,3 Shimul A. Shah,1,2 Robert M. Cohen,2,3 Daniel P. Schauer,2,3 Eric P. Smith,2,3 and Tayyab S. Diwan1,2 1 Department of Surgery, Section of Transplantation, University of Cincinnati School of Medicine, Cincinnati, OH; 2 Cincinnati Collaborative for Obesity Research, University of Cincinnati, Cincinnati, OH; and 3 Department of Internal Medicine, University of Cincinnati School of Medicine, Cincinnati, OH

Obesity has become an epidemic in the United States over the past decade, and recent studies have shown this trend in the liver transplantation (LT ) population. These patients may be candidates for laparoscopic sleeve gastrectomy (LSG) to promote significant and sustained weight loss to prevent recurrence of nonalcoholic steatohepatitis. However, safety remains a concern, and efficacy in this setting is uncertain. A single-institution database from 2014 to 2018 was queried for patients undergoing LSG following LT. The selection criteria for surgery were consistent with National Institutes of Health guidelines, and patients were at least 6 months after LT. A total of 15 patients (median age, 59.0 years; Caucasian, 86.7%; and female, 60%) underwent LSG following LT. Median time from LT to LSG was 2.2 years with a median follow-up period of 2.6 years. The median hospital length of stay (LOS) was 2 days after LSG. Mortality and rate of liver allograft rejection was 0, and there was 1 postoperative complication (a surgical site infection). Following LSG, body mass index (BMI) decreased from 42.7 to 35.9 kg/m2 (P < 0.01), and in 12 patients with at least 1 year of follow-up, the total body weight loss was 20.6%. Following LSG in patients with diabetes, the median daily insulin requirements decreased from 98 (49-118) to 0 (0-29) units/day (P = 0.02), and 60% discontinued insulin. Post-LT patients had a similar decrease in BMI and reduction in comorbidities at 1 year compared with a matched non-LT patient cohort. In the largest patient series to date, we show that LSG following LT is safe, effective, and does not increase the incidence of liver allograft rejection. Larger longer-term studies are needed to confirm underlying metabolic changes following LSG. Liver Transplantation 25 1673‒1681 2019 AASLD. Received May 2, 2019; accepted August 2, 2019.

The prevalence of individuals who are overweight or for 2016, 29.6% and 35.2% of the adult US population obese is increasing at an alarming rate, and obesity is were obese and overweight, respectively.(2) Obesity is a now considered a worldwide epidemic.(1) According to risk factor for increased mortality and is associated with Centers for Disease Control and Prevention estimates comorbidities, such as hyperlipidemia, diabetes mellitus (DM), hypertension (HTN), and atherosclerosis.(3) A host of medical and surgical weight loss options Abbreviations: %EBMIL, percentage of estimated body mass index loss; %EWL, percentage of excess weight loss; %TWL, percentage of total have been studied in response to the growing obesity weight loss; AIH, autoimmune hepatitis; ALP, alkaline phosphatase; epidemic. Medical weight loss options, such as diet ALT, alanine aminotransferase; AST, aspartate aminotransferase; and exercise, have demonstrated at best a modest 10% BMI, body mass index; CAD, coronary artery disease; CVD, sustained total body weight loss in the general pop- cerebrovascular disease; DM, diabetes mellitus; EtOH, ethanol; HBV, (4) hepatitis B virus; HCV, hepatitis C virus; HTN, hypertension; ICU, ulation. In comparison, surgical options, such as intensive care unit; IQR, interquartile range; IS, immunosuppression; Roux-en-Y gastric bypass (RYGB) and laparoscopic LSG, laparoscopic sleeve gastrectomy; LOS, length of stay; LT, liver sleeve gastrectomy (LSG), have demonstrated a 5-year transplantation; MMF, mycophenolate mofetil; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; OSA, percentage of total weight loss (%TWL) of 25.5% and (5) obstructive sleep apnea; PBC, primary biliary cirrhosis; RYGB, Roux- 18.8%, respectively. Bariatric surgery has proven en-Y gastric bypass; SSI, surgical site infection. to be safe and results not only in effective and sustained weight loss, but also in complete resolution or

Original Article | 1673 Morris et al. Liver Transplantation, November 2019 significant improvement in comorbidities even before following LT for LSG. The selection criteria for weight significant weight loss.(6) As a result of the documented loss surgery was consistent with current National safety and efficacy of bariatric surgery, the possibility of Institutes of Health guidelines (body mass index [BMI] employing bariatric surgery in more complex patient ≥40 kg/m2 alone or BMI ≥35 kg/m2 with an obesi- populations is increasingly being considered. ty-related comorbidity), and patients were a minimum A major manifestation of the obesity epidemic is the of 6 months after their LT. After institutional review increased incidence of nonalcoholic fatty liver disease board approval, the prospectively maintained bariatric (NAFLD). Strikingly, as much as 30% of Americans surgery database from 2014 to 2018 was retrospectively are affected by NAFLD. NAFLD involves a spectrum queried. From the initial data set (n = 302), 15 patients of disease including steatosis alone to nonalcoholic ste- who underwent LSG following LT were identified. atohepatitis (NASH), which is an inflammatory sub- Patient demographics, including age, sex, race, comor- type of the disease. Cirrhosis is observed in up to 20% bidities, etiologies of liver disease, and maintenance of patients with NASH in a 10-year period.(6) In 2009, immunosuppression (IS), were reviewed. Operative NASH was the third most common indication for liver and weight loss–specific variables, including length of transplantation (LT). This has not only increased 8-fold surgery, postoperative complications, hospital length since 2001 but also is the only indication that increased of stay (LOS), incidence of liver allograft rejection, in frequency over that time period.(7) NASH-related preoperative and postoperative weight, BMI, and ideal cirrhosis will likely become the leading indication for body weight, were reviewed and included for analysis. LT in the next decade.(6) DM was defined as either insulin dependence, hemo- In patients with end-stage liver disease, however, globin A1c >6.5%, or the use of metformin plus an ad- there is limited literature regarding outcomes after bar- ditional oral antihyperglycemic medication. Metabolic iatric surgery. Small patient series have demonstrated syndrome was defined as per the National Cholesterol the utility of bariatric surgery in patients with end- Education Program Adult Treatment Panel III crite- stage liver disease, both simultaneously during LT and ria.(13) Because of the retrospective nature of our study following LT. (8-12) The optimal technique and timing and inability to ascertain waist circumference, we used of bariatric surgery remains undefined, and the safety BMI ≥35 kg/m2 as a surrogate for waist circumference. and efficacy are uncertain. The aim of our study is to As part of our multidisciplinary approach, a sur- assess, in a larger cohort of patients, the efficacy and geon, psychologist, nutritionist, and exercise therapist safety of LSG following LT. evaluated all patients referred to our bariatric program. During the initial assessment, the surgeon evaluated and provided education on the LSG procedure, out- Patients and Methods comes, and required lifestyle modifications. Dietary education was delivered, individualized exercise rec- PATIENT SELECTION ommendations were specified, and each patient under-

In 2014, a bariatric program at the University of went a psychological evaluation. For patients choosing

Cincinnati was implemented to assess obese patients to enroll in our bariatric program, subsequent visits were scheduled at regular intervals and included ongo- ing assessments by the transplant surgeon and dieti- Address reprint requests to Tayyab S. Diwan, M.D., Department of cian (and other members of the multidisciplinary team Surgery, Section of Transplantation, University of Cincinnati College of as indicated). Because of a lack of clinical trials and Medicine, 231 Albert Sabin Way, Suite 1556, Cincinnati, OH 45267- 0519. Telephone: 513-558-3892; E-mail: [email protected] concern for potential drug-drug interactions with IS therapies, antiobesity medications were not provided Shimul A. Shah has grants from Organ Recovery Systems and advises for Sirtex. Jonathan R. Thompson owns stock, is employed, and has as a treatment option for this unique population. Once intellectual property rights for Standard Bariatrics. the patients were deemed appropriate for weight loss

Copyright © 2019 by the American Association for the Study of Liver surgery, they were referred for perioperative care for Diseases. risk assessment and stratification. Further cardiopul- monary workup was performed as indicated. IS medi- View this article online at wileyonlinelibrary.com. cations were not adjusted prior to LSG. DOI 10.1002/lt.25637 LSG was performed as previously described by Freeman et al.(14) In brief, all operations were performed

1674 | Original Article Liver Transplantation, Vol. 25, No. 11, 2019 Morris et al. by a single fellowship-trained surgeon (T.S.D.). Gastric a dietician. Micronutrient and vitamin supplements volume was reduced by approximately 60%-80%, and were recommended as per Academy of Nutrition and residual capacity was estimated to be <300 mL. All Dietetics guidelines.(16) sleeve gastrectomies were performed laparoscopically, and none required conversion to an open procedure. STATISTICAL ANALYSIS Intraoperative liver biopsy was performed at the dis- cretion of the operating surgeon and was not routinely Ideal body weight was defined as the weight at which performed at the time of LSG. the patient’s BMI was equal to 25 kg/m2. All data are presented as the median (interquartile range [IQR]).

POSTOPERATIVE CARE LT patients were propensity matched by age, sex, and BMI to non-LT patients undergoing LSG. Fisher’s All patients were cared for by a multidisciplinary team exact test was used to compare categorical variables. that included a transplant hepatologist, transplant sur- Wilcoxon rank sum test was used to compare continu- geon, and dietician during their hospital stay. All pa- ous variables between 2 groups. A P value of <0.05 was tients remained nil per os on the day of surgery, and a considered significant. JMP Pro, version 14 and SAS, liquid diet was slowly initiated on the morning of post- version 9.4 (SAS Institute, Cary, NC) were used for all operative day 1. Perioperative deep vein thrombosis statistical analyses. prophylaxis included chemoprophylaxis with subcuta- neous heparin or enoxaparin, sequential compression devices, and early ambulation. As previously demonstrated via pharmacokinetic Results and pharmacogenetic analyses, dose modifications of A total of 15 patients who underwent LSG following IS agents after LSG are likely unnecessary.(15) However, LT were identified. The patients were predominantly during the early post-LSG period when the patient is Caucasian (86.7%), female (60%), with a median age only consuming liquids and may experience symptoms of 59 (51-62) years old and on tacrolimus plus myco- (ie, nausea, vomiting, and diarrhea) that could impact phenolate mofetil (MMF; 66.7%) IS maintenance reg- drug absorption, temporary adjustments to their IS imens. All patients except 1 had a diagnosis of NASH regimens are employed. Prior to LSG surgery, a trans- prior to LT. Median time from LT to LSG was 2.2 plant pharmacist evaluates each patient’s IS regimen (1.4-5.4) years, and median follow-up was 2.6 (0.6- and recommends the appropriate formulation and dose 3.9) years. Demographics, comorbidities, etiology of adjustments based on the patient’s specific medica- liver disease, and IS regimens are detailed in Table 2. tions (Table 1). Following LSG, these recommended Following LSG, BMI decreased from 42.7 (40.1- IS modifications are implemented. Once patients are 44.8) kg/m2 to 35.9 (30.2-37.3) kg/m2 (P < 0.01). The transitioned to a soft diet and can swallow pills safely, percentage of excess body mass index loss (%EBMIL) their pre-LSG IS regimen drug doses and formula- was 45.0% (32.2%-61.0%). The %TWL and percent- tions are resumed. Therapeutic drug monitoring of IS age of excess weight loss (%EWL) were 18.5% (13.5%- is performed per transplant center standard guidelines. 22.8%) and 44.7% (32.3%-59.3%), respectively. In All patients were followed regularly in the outpa- patients with a minimum of 1 year of follow-up tient setting and evaluated by both the surgeon and (n = 12), BMI decreased from 42.3 (40.1-44.6) kg/m2

TABLE 1. IS Drug Dose and Administration Modifications Performed During the Early Post-LSG Period in LT Patients at Our Center

IS Medication Dose Administration

Azathioprine No change Oral suspension Cyclosporine No change Oral suspension MMF No change Intravenous immediately postoperatively and then transition to oral suspension when able Prednisone No change Oral suspension postoperatively while an inpatient. For outpatients, crush tablets, place contents in a cup, add a small amount of water, stir, and drink; then, rinse cup with water and drink again. Tacrolimus Decrease by 50% Sublingually

Original Article | 1675 Morris et al. Liver Transplantation, November 2019

TABLE 2. Patient Demographics

Posttransplant LSG (n = 15)

Sex, male 6 (40.0) Age, years 59 (51-62) Race Caucasian 13 (86.7) African American 2 (13.3) OSA 3 (20.0) HTN 11 (73.3) DM 8 (53.3) CAD 0 (0) CVD 0 (0) Etiology of liver failure FIG. 1. The %TWL for each individual patient following LSG. NASH 14 (93.3) Associated etiologies AIH 2 (13.3) PBC 1 (6.7) EtOH 3 (20.0) HBV 1 (6.7) HCV 1 (6.7) Maintenance IS regimens Tacrolimus monotherapy 1 (6.7) Tacrolimus + MMF 10 (66.7) Tacrolimus + MMF + prednisone 2 (13.3) Cyclosporine + MMF 1 (6.7) Cyclosporine + azathioprine 1 (6.7)

NOTE: Data are given as median (IQR) or n (%).

FIG. 2. The daily insulin requirement for patients on insulin to 33.5 (30.0-37.8) kg/m2 (P < 0.01). The %EBMIL, therapy preoperatively and in the postoperative period following LSG. %TWL, and %EWL after 1 year were 55.0% (35.0%- 66.7%), 20.6% (14.9%-26.4%), and 51.5% (33.8%- 62.5%), respectively. The %TWL for each individual patient following LSG is depicted in Fig. 1. All patients hemoglobin A1c, creatinine, and alanine aminotrans- except 1 individual were able to keep their weight off ferase (ALT), aspartate aminotransferase (AST), total and continue to lose weight at 1 year. This patient did bilirubin, and alkaline phosphatase (ALP) levels were not follow the prescribed diet plan and was not doing similar before and after LSG. The number of antihy- any exercise at the time of last follow-up. She was sub- pertensive medications went from 1 (0-2) before LSG sequently counseled and referred back to her psycholo- to 1 (0-1) following LSG. The number of patients gist, nutritionist, and exercise therapist. with obstructive sleep apnea (OSA) following LSG The number of patients on insulin decreased from went from 3 to 1, the number of patients with hyper- 5 to 2 patients following LSG (P > 0.05). Among lipidemia went from 9 to 7 patients, and the number of patients on insulin prior to LSG, the median insulin patients with metabolic syndrome went from 10 to 7 requirement per day decreased from 98 (49-118) units/ patients (each P > 0.05). Additionally, no patients have day to 0 (0-29) units/day (P = 0.02). The 2 patients experienced liver allograft rejection after LSG during who remained on insulin following LSG had a reduc- the follow-up period. Outcomes following LSG in tion of over 50%: from 120 and 116 units of insulin per post-LT patients are further described in Table 3. day to 8 and 50 units of insulin per day, respectively. Eventually, all patients resumed their pre-LSG The daily insulin requirement of each patient with maintenance IS drug regimens after LSG. However, insulin-dependent diabetes is depicted in Fig. 2. The the IS drug dose and formulation varied during the

1676 | Original Article Liver Transplantation, Vol. 25, No. 11, 2019 Morris et al.

TABLE 3. Outcomes Following Posttransplant LSG

Pre-LSG (n = 15) Post-LSG (n = 15) P Value

BMI, kg/m2 42.7 (40.1-44.8) 35.9 (30.2-37.3) <0.01 Antihypertensive medications 1 (0-2) 1 (0-1) 0.17 Insulin dependent 5 (33.3) 2 (13.3) 0.40 Insulin, units/day 98 (49-118) 0 (0-29) 0.02 Hemoglobin A1c* 6.3 (5.4-9.2) 6.4 (5.7-7.2) 0.69 Creatinine, mg/dL† 1.2 (1-1.4) 1.2 (1.0-1.3) 0.82 ALT, units/L† 22 (13.5-39.8) 15.5 (10.3-30.3) 0.24 AST, units/L† 21 (15.3-33) 17 (14.5-26.3) 0.49 Total bilirubin, mg/dL† 0.5 (0.3-0.8) 0.6 (0.4-0.8) 0.49 ALP, units/L† 83 (67-109.8) 108 (77.5-124.3) 0.25 OSA 3 (20) 1 (7.1) 0.60 Hyperlipidemia 9 (60) 7 (46.7) 0.72 Metabolic syndrome 10 (66.7) 7 (46.7) 0.27

NOTE: Data are given as median (IQR). *n = 8. †n = 12.

FIG. 3. Tacrolimus trough concentrations (ng/mL) for each individual patient on a tacrolimus-containing maintenance IS drug regimen before and after LSG and at the time of last follow-up. early postoperative period (approximately 2 weeks), LSG to initial outpatient IS drug level (ie, tacrolimus as described previously. Therapeutic drug monitor- or cyclosporine) assessment was 14.0 (5-200) days. ing frequency and target concentrations of IS varied Figure 3 illustrates the individual tacrolimus trough according to the elapsed time after LT per our trans- concentrations (ng/mL) for each of the 13 patients plant center’s standard of care. Median time from IS maintained on a tacrolimus-based regimen at 3 time drug level (ie, tacrolimus or cyclosporine) assessment points: before LSG, after LSG, and at time of last to LSG was 52.0 (6-302) days. Median time from follow-up.

Original Article | 1677 Morris et al. Liver Transplantation, November 2019

Intraoperative liver biopsy was selectively performed In our matched analysis, operative time, conversion at the time of LSG in 5 patients. The pathology ranged to an open procedure, LOS, intensive care admis- from mild to moderate steatosis and mild or no fibrosis sions, and 30-day complications were similar between (Table 4). None of these patients have had any indica- post-LT and non-LT patients undergoing LSG. The tion for a subsequent liver biopsy since their LSG. post-LT patients had a longer period of follow-up and higher estimated blood loss. The BMI between post-LT patients and non-LT patients had a simi- TABLE 4. Results of Intraoperative Liver Biopsy Performed lar decrease at 1 year. Obesity-associated comorbid- During LSG ities did not change following LSG in the post-LT or the non-LT LSG groups. All outcomes compar- Patient Biopsy Results ing post-LT and non-LT patients following LSG are 1 Moderate steatosis (macrosteatosis 40%, microsteatosis 10%) detailed in Table 5. No fibrosis 2 Minimal macrosteatosis (<5%) Mild portal fibrosis (stage 1) Discussion 3 Mild microsteatosis (5%) Mild portal fibrosis (stage 1) In this study, we examined the efficacy and safety of 4 Moderate steatosis (macrosteatosis 25%, microsteatosis 15%) bariatric surgery among a high-risk post-LT popu- No fibrosis lation. We found that patient BMI significantly de- 2 5 Mild steatosis (macrosteatosis 10%, microsteatosis 15%) creased from 42.7 to 35.9 kg/m following LSG. No fibrosis Insulin requirements markedly decreased in all patients with 3 of the 5 having insulin discontinued. The safety of LSG was demonstrated through a low complication TABLE 5. Matched Analysis of Post-LT and Non-LT rate, which consisted of a single surgical site infec- = Patients (n 14) tion (SSI), no rejection episodes or change in allograft Post-LT LSG Non-LT LSG P Value function as evidenced by normal liver enzymes, and no postoperative deaths. Follow-up, years 2.6 (0.6-3.9) 0.8 (0.3-1.2) 0.01 < In the present study, we had a %EWL of 47.5% and LT to LSG interval, years 2.2 (1.4-5.4) — — a significant decrease in BMI following LSG. These Operative time, minutes 117 (93-232) 114 (90-136) 0.59 findings are comparable to %EWL seen in the non- Estimated blood loss 50 (25-100) 5 (5-5) <0.01 (EBL), mL transplant population that range from 53.3% in LSG (17) Conversion to an open 0 (0) 0 (0) — and 62.6% in RYGB. In the post-LT population, procedure %EWL ranges from 43% to 76% in studies with a (11,18-21) LOS, days 2 (1-2) 1 (1-1) 0.52 minimum of 12-month follow-up. In a recent ICU admissions 1 (6.7) 1 (7.0) 1.00 series of 12 patients who underwent LSG following 30-day complications 1 (6.7) 2 (13.3) 0.56 LT, Tsamalaidze et al. demonstrated a change in BMI BMI, kg/m2 >0.99 similar to a matched non-LT cohort and a %EWL of Pre-LSG 42.5 (40-44) 42.5 (39-45) 50% at 1 year.(11) All other data are limited to smaller Post-LSG 33.5 (30.3-37.6) 33.8 (28.2-37.1) patient series or single patient reports. Our study Comorbid conditions further supports this previous literature with a larger Insulin-dependent DM 0.08/1.00* number of patients, and it demonstrates that effec- Pre-LSG 5 (35.7) 3 (21.4) tive weight loss can be seen in the post-LT population Post-LSG 2 (14.3) 3 (21.4) undergoing LSG. We also show similar decreases in HTN 0.18/0.99* BMI at 1 year in matched-control patients without Pre-LSG 10 (71.4) 10 (71.4) Post-LSG 7 (50.0) 9 (64.3) previous LT following LSG. Additionally, we demon-

Hyperlipidemia 0.32/0.99* strate good efficacy combined with a low complication Pre-LSG 3 (33.3) 7 (50.0) rate.

Post-LSG 2 (22.2) 6 (42.9) DM, hyperlipidemia, cardiac disease, and OSA are conditions frequently associated with obe- NOTE: Data are given as median (IQR) or n (%). sity. Significant improvement or resolution of these *P values are given as post-LT LSG/non-LT LSG. obesity-associated conditions has been previously

1678 | Original Article Liver Transplantation, Vol. 25, No. 11, 2019 Morris et al. demonstrated in the general population following bar- patient remained hemodynamically stable and did not iatric surgery.(22-24) In the current study, we demon- require a blood transfusion. Our operative time and strated complete resolution of insulin dependence in hospital LOS were similar to previously reported liter- 3 of 5 patients and decreased daily insulin requirement ature and our non-LT matched controls.(11) Although by an average of 62.2 units per day among patients EBL was increased in post-LT patients, this finding is with DM requiring insulin. This is consistent with, if likely not clinically significant. not higher than, the diabetic remission rate of 25%- RYGB has been the standard of care in bariatric sur- 50% previously described in the literature.(25,26) We gery due to its impressive weight loss effects. However, did not find any differences in hemoglobin A1c levels because LSG results in similar weight loss and meta- when comparing patients before and after LSG. This bolic improvement but with less risk of complications, likely reflects the close monitoring and tight control it is increasingly the preferred choice, particularly in a of DM that occurs at our center after LT. Conversely, complex patient population such as LT recipients.(29) we did not observe any significant decrease in the Concerns with altered drug absorption and the need number of hypertensive medications needed or in the for increased dosages of IS medications, in addition to incidence of OSA, hyperlipidemia, or metabolic syn- the hypovitaminoses associated with RYGB, also sup- drome. Similarly, no decrease was seen in DM, hyper- port the use of LSG.(15,29) A final benefit of LSG com- lipidemia, or HTN in our non-LT matched controls pared with RYGB is preserved access to the biliary tree at 1 year. for a potential future need for endoscopic evaluation of The other significant comorbidity in this population the allograft liver. For these reasons, we prefer LSG for is the recurrence of NASH cirrhosis. NASH is pre- bariatric surgery in this particular patient population. dicted to become the primary indication for LT, and Although we believe that LSG provides the best these patients remain at risk for recurrence in their balance of safety and efficacy for this population, the liver allograft. Recurrence of NAFLD is estimated to optimal timing of bariatric surgery in this unique pop- develop in 30%-60% of these patients without LSG.(27) ulation remains undefined. Bariatric surgery has been However, previous studies have demonstrated a signif- described in patients with liver disease prior to LT, icant improvement in liver fibrosis and steatohepati- simultaneously during LT, and in a delayed fashion tis following bariatric surgery.(28) This is an important following LT. Several small series have been published benefit of bariatric surgery in this specific patient pop- describing bariatric surgery in patients with liver dis- ulation and will be of the utmost importance in pre- ease.(8,30-32) The efficacy in weight loss and resolution serving allograft function and improving longterm of comorbidities is comparable to patients without liver outcomes. Although this study did not directly exam- disease, but it comes with a higher complication risk. ine the recurrence of NASH in allografts, the weight Weight loss prior to LT along with the reduction or loss, metabolic improvements, and stability of liver resolution of medical comorbidities may help improve enzymes suggests that LSG will likely reduce the prob- perioperative outcomes. This may further translate to ability of recurrent NASH. better outcomes following LT, but this result remains Operating on this complex patient population is not unknown.(33) The largest series on simultaneous LSG without risks, and the complication rate has been cited and LT by Zamora-Valdes et al.(10) describes a %TWL as up to 40%. These complications include infection, of 33% at 12 months. They reported an overall com- staple line leaks, reoperation, and even death.(9,11,19- plication rate nearing 50% and 2 postoperative deaths. 21) In the previously largest study of LSG after LT, However, most were likely associated with the LT and Tsamalaidze et al. observed a major complication rate not the LSG. The complications ranged from infec- of 25% and no reoperations or deaths.(11) Another tion to hepatic artery thrombosis and allograft failure. study in the post-LT population demonstrated a com- Complications related to LSG, namely, gastric staple plication rate >40%, which included 2 deaths; however, line leak and severe reflux, were observed in 1 out of 29 this series evaluated RYGB following LT. The other patients and 3 out of 29 patients, respectively. Delayed patient series reported complication rates between 17% LSG has been described extensively previously, but and 40%. In comparison, our study demonstrated a low overall, the literature demonstrates both simultane- complication rate of 6.7%, consisting of a single SSI. ous and delayed techniques demonstrating effective Intensive care unit (ICU) monitoring was indicated weight loss, and complication rates are difficult to for 1 patient due to intraoperative blood loss, but the compare due to the significant morbidity associated

Original Article | 1679 Morris et al. Liver Transplantation, November 2019 with undergoing LT. In our study, the delayed wait- and longer follow-up to identify the optimal approach ing period of at least 6 months following LT may add to this complex population. an element of selection bias. However, it likely con- In summary, in the largest series to date, we demon- tributes to the low complication rate and findings of strate the safety and weight loss efficacy of LSG in stable allograft function with no episodes of rejection. post-LT recipients along with a pronounced effect on Furthermore, the optimal timing to perform LSG in the resolution of DM. Our experience highlights the patients with liver disease remains unclear, and differ- relatively low morbidity and mortality in the delayed ent approaches may be appropriate for select individ- approach for LSG following LT. uals. Further studies to help predict the patients who benefit from each approach are warranted. Additionally, our experience demonstrates that our REFERENCES patients experienced a reduction in their IS drug levels 1) Abelson P, Kennedy D. The obesity epidemic. Science during the early post-LSG period (Fig. 3). Although 2004;304:1413. the study was not designed to analyze reasons for such 2) Winer LK, Dhar VK, Wima K, Morris MC, Lee TC, Shah SA, et al. The impact of tumor location on resection and survival for observations, we believe it to be multifactorial, such pancreatic ductal adenocarcinoma. J Surg Res 2019;239:60-66. that the commonly experienced gastrointestinal symp- 3) Lowenfels AB, Maisonneuve P, DiMagno EP, Elitsur Y, Gates LK toms (nausea, vomiting, and diarrhea) along with the Jr, Perrault J, Whitcomb DC. Hereditary pancreatitis and the risk IS dose and formulation modifications in the early of pancreatic cancer. International Hereditary Pancreatitis Study Group. J Natl Cancer Inst 1997;89:442-446. postoperative period likely impacted drug absorp- 4) Romero-Gómez M, Zelber-Sagi S, Trenell M. Treatment of tion, resulting in reduced IS concentrations. Although NAFLD with diet, physical activity and exercise. J Hepatol our patients did not experience any episodes of liver 2017;67:829-846. 5) Arterburn D, Wellman R, Emiliano A, Smith SR, Odegaard allograft rejection in our series, this is a possibility in AO, Murali S, et al.; for PCORnet Bariatric Study Collaborative. the setting of subtherapeutic IS drug levels. Therefore, Comparative effectiveness and safety of bariatric procedures we recommend for bariatric and transplant programs for weight loss: a PCORnet Cohort Study. Ann Intern Med 2018;169:741-750. to collaborate in the care of these patients to ensure 6) Rinella ME. Nonalcoholic fatty liver disease: a systematic review. vigilant therapeutic drug monitoring so that adequate JAMA 2015;313:2263-2273. IS coverage is maintained before, during, and after 7) Charlton MR, Burns JM, Pedersen RA, Watt KD, Heimbach LSG in the LT population. JK, Dierkhising RA. Frequency and outcomes of liver transplantation for nonalcoholic steatohepatitis in the United States. Our study has several limitations that need to be Gastroenterology 2011;141:1249-1253. discussed. The first is the retrospective nature of the 8) Shimizu H, Phuong V, Maia M, Kroh M, Chand B, Schauer PR, study and the potential for selection bias. However, Brethauer SA. Bariatric surgery in patients with liver cirrhosis. Surg Obes Relat Dis 2013;9:1-6. we included all of our patients in the database at the 9) Lin MY, Tavakol MM, Sarin A, Amirkiai SM, Rogers SJ, Carter time of our query to mitigate this limitation. Second, J T, Posselt AM. Safety and feasibility of sleeve gastrectomy in we do not have any control or comparative groups, morbidly obese patients following liver transplantation. Surg Endosc 2013;27:81-85. and therefore, we cannot make any direct comparisons 10) Zamora-Valdes D, Watt KD, Kellogg TA, Poterucha JJ, Di Cecco to other techniques, but there are currently no trials SR, Francisco-Ziller NM, et al. Long-term outcomes of patients comparing these groups. Although this is the largest undergoing simultaneous liver transplantation and sleeve gastrec- study evaluating this population, the relatively small tomy. Hepatology 2018;68:485-495. 11) Tsamalaidze L, Stauffer JA, Arasi LC, Villacreses DE, Franco numbers make it hard to discern significant findings, JSS, Bowers S, Elli EF. Laparoscopic sleeve gastrectomy for mor- most importantly the morbidity associated with LSG. bid obesity in patients after orthotopic liver transplant: a matched Third, the true number of patients referred to our clinic case-control study. Obes Surg 2018;28:444-450. 12) Diwan TS, Rice TC, Heimbach JK, Schauer D P. Liver transplan- is unknown. Therefore, we are unable to describe the tation and bariatric surgery: timing and outcomes. Liver Transpl patients who were evaluated in our clinic, but they did 2018;24:1280-1287. not undergo LSG for various reasons. This restriction 13) Huang PL. A comprehensive definition for metabolic syndrome. Dis Model Mech 2009;2:231-237. limits our ability to compare delayed LSG to simulta- 14) Freeman CM, Woodle ES, Shi J, Alexander JW, Leggett PL, Shah neous LT and LSG. In delayed LSG, some patients SA, et al. Addressing morbid obesity as a barrier to renal trans- may not have access to surgery for various reasons fol- plantation with laparoscopic sleeve gastrectomy. Am J Transplant lowing LT compared with simultaneous LT and LSG 2015;15:1360-1368. 15) Diwan TS, Lichvar AB, Leino AD, Vinks AA, Christians U, where all at-risk patients are treated. These limitations Shields AR, et al. Pharmacokinetic and pharmacogenetic analysis make it necessary for further studies with more patients

1680 | Original Article Liver Transplantation, Vol. 25, No. 11, 2019 Morris et al.

of immunosuppressive agents after laparoscopic sleeve gastrec- the medically complicated obese: new challenges on the horizon. tomy. Clin Transplant 2017;31(6). World J Hepatol 2015;7:2315-2318. 16) Snyder-Marlow G, Taylor D, Lenhard MJ. Nutrition care for 26) Courcoulas A P, Yanovski SZ, Bonds D, Eggerman TL, Horlick patients undergoing laparoscopic sleeve gastrectomy for weight M, Staten MA, Arterburn DE. Long-term outcomes of bariatric loss. J Am Diet Assoc 2010;110:600-607. surgery: a National Institutes of Health symposium. JAMA Surg 17) Golzarand M, Toolabi K, Farid R. The bariatric surgery and 2014;149:1323-1329. weight losing: a meta-analysis in the long- and very long-term 27) Yalamanchili K, Saadeh S, Klintmalm GB, Jennings LW, Davis effects of laparoscopic adjustable gastric banding, laparoscopic GL. Nonalcoholic fatty liver disease after liver transplantation Roux-en-Y gastric bypass and laparoscopic sleeve gastrectomy on for cryptogenic cirrhosis or nonalcoholic fatty liver disease. Liver weight loss in adults. Surg Endosc 2017;31:4331-4345. Transpl 2010;16:431-439. 18) Gentileschi P, Venza M, Benavoli D, Lirosi F, Camperchioli I, 28) Lassailly G, Caiazzo R, Buob D, Pigeyre M, Verkindt H, D’Eletto M, et al. Intragastric balloon followed by biliopancreatic Labreuche J, et al. Bariatric surgery reduces features of nonalco- diversion in a liver transplant recipient: a case report. Obes Surg holic steatohepatitis in morbidly obese patients. Gastroenterology 2009;19:1460-1463. 2015;149:379-388. 19) Khoraki J, Katz MG, Funk LM, Greenberg JA, Fernandez LA, Campos 29) Lager CJ, Esfandiari NH, Subauste AR, Kraftson AT, Brown MB, GM. Feasibility and outcomes of laparoscopic sleeve gastrectomy Cassidy RB, et al. Roux-en-Y gastric bypass vs. sleeve gastrectomy: after solid organ transplantation. Surg Obes Relat Dis 2016;12:75-83. balancing the risks of surgery with the benefits of weight loss. 20) Osseis M, Lazzati A, Salloum C, Gavara CG, Compagnon P, Obes Surg 2017;27:154-161. Feray C, et al. Sleeve gastrectomy after liver transplantation: feasi- 30) Lin MY, Tavakol MM, Sarin A, Amirkiai SM, Rogers SJ, Carter bility and outcomes. Obes Surg 2018;28:242-248. J T, Posselt AM. Laparoscopic sleeve gastrectomy is safe and ef- 21) Al-Nowaylati AR, Al-Haddad BJ, Dorman RB, Alsaied OA, ficacious for pretransplant candidates. Surg Obes Relat Dis Lake JR, Chinnakotla S, et al. Gastric bypass after liver transplan- 2013;9:653-658. tation. Liver Transpl 2013;19:1324-1329. 31) Takata MC, Campos GM, Ciovica R, Rabl C, Rogers SJ, Cello J P, 22) Nguyen NT, Varela JE. Bariatric surgery for obesity and meta- et al. Laparoscopic bariatric surgery improves candidacy in mor- bolic disorders: state of the art. Nat Rev Gastroenterol Hepatol bidly obese patients awaiting transplantation. Surg Obes Relat Dis 2017;14:160-169. 2008;4:159-164. 23) Sarkhosh K, Switzer NJ, El-Hadi M, Birch DW, Shi X, Karmali 32) Safwan M, Collins KM, Abouljoud MS, Salgia R. Outcome of S. The impact of bariatric surgery on obstructive sleep apnea: a liver transplantation in patients with prior bariatric surgery. Liver systematic review. Obes Surg 2013;23:414-423. Transpl 2017;23:1415-1421. 24) Owen JG, Yazdi F, Reisin E. Bariatric surgery and hypertension. 33) Tevar AD, Clarke C, Wang J, Rudich SM, Woodle ES, Lentsch Am J Hypertens 2017;31:11-17. AB, Edwards ML. Clinical review of nonalcoholic steato- 25) Reino DC, Weigle KE, Dutson EP, Bodzin AS, Lunsford KE, hepatitis in liver surgery and transplantation. J Am Coll Surg Busuttil RW. Liver transplantation and sleeve gastrectomy in 2010;210:515-526.

Original Article | 1681