Combined Liver Transplantation and Gastric Sleeve Resection for Patients With Medically Complicated Obesity and End-Stage Liver Disease


Corresponding author: Julie K. Heimbach,


Obesity is increasingly common before and after liver transplantation (LT), yet optimal management remains unclear. Our aim was to analyze the effectiveness of a multidisciplinary protocol for obese patients requiring LT, including a noninvasive pretransplant weight loss program, and a combined LT plus sleeve gastrectomy (SG) for obese patients who failed to lose weight prior to LT. Since 2006, all patients referred LT with a BMI > 35 were enrolled. There were 37 patients who achieved weight loss and underwent LT alone, and 7 who underwent LT combined with SG. In those who received LT alone, weight gain to BMI > 35 was seen in 21/34, post-LT diabetes (DM) in 12/34, steatosis in 7/34, with 3 deaths plus 3 grafts losses. In patients undergoing the combined procedure, there were no deaths or graft losses. One patient developed a leak from the gastric staple line, and one had excess weight loss. No patients developed post-LT DM or steatosis, and all had substantial weight loss (mean BMI = 29). Noninvasive pretransplant weight loss was achieved by a majority, though weight gain post-LT was common. Combined LT plus SG resulted in effective weight loss and was associated with fewer post-LT metabolic complications. Long-term follow-up is needed.


body mass index


hepatitis C virus


liver transplantation


nonalcoholic steatohepatitis


sleeve gastrectomy


The prevalence of obesity in the United States is currently estimated to be approximately 35% percent [1]. Anorexia and wasting accompanies chronic liver disease and traditionally has required pretransplant nutritional strategies aimed at maintaining or increasing body weight. However, the current epidemic of obesity has also impacted patients with end-stage liver disease, and may be either the primary cause of liver disease or at least a contributing factor in the patient's liver disease [2, 3]. Recent evidence has shown that nonalcoholic steatohepatitis (NASH) is the most rapidly rising indication for liver transplantation (LT) in the United States and in projected to become the most common indication [4]. While obese patients may be transplanted with medium-term outcomes similar to nonobese patients [3-6], the long-term impact of obesity on post-LT outcomes including recurrence of NASH and hepatitis C virus (HCV) is becoming increasingly evident [7-9]. Additionally, obesity is strongly associated with diabetes, heart disease and cancer which are leading causes of morbidity and mortality post-LT [10].

Options for weight reduction for obese patients in need of LT include a noninvasive approach of rigorous dietary and behavioral modification employed both pre and posttransplantation. However, the noninvasive approach may not be successful for all patients, particularly those with long-standing severe obesity. Bariatric surgery may be suitable for patients with early-stage liver disease [11], but it is not indicated for patients with decompensated liver disease [12]. Posttransplant bariatric surgery may be an option though requires two separate major operations and may be associated with increased technical difficulty due to adhesions and complications related to long-term immunosuppression [13-16].

Thus, we developed a program which offered a combined LT with sleeve gastrectomy (SG) for obese patients with end-stage liver disease with a MELD score high enough to proceed to LT who were unsuccessful in attaining weight loss in an aggressive noninvasive pretransplant weight loss protocol. Sleeve gastrectomy was chosen over gastric bypass due to decreased technical complexity and the lack of mal-absorption which may impact early posttransplant immunosuppression levels [17]. The gastric band has been previously reported in combination with LT [18] however due to the decreased efficacy of gastric band in achieving weight loss compared to gastric bypass, as well as the potential for complications from a foreign body in an immunosuppressed patient and challenges in accessing the biliary tract via an endoscopic route if needed post-LT, sleeve gastrectomy was selected.


All adult patients listed for LT since 2006 with BMI > 35 kg/m2 (corrected for ascites) and at least one follow-up visit were enrolled in an aggressive weight management protocol. The protocol consisted of dietary education provided by an experienced transplant dietician at the initial visit and at each subsequent visit until transplantation or achievement of goal weight. Patients were also evaluated by the same transplant surgeon at the initial and at each subsequent visit to assign a specific weight loss goal (generally BMI < 35 kg/m2), to further emphasize the long-term benefits of weight loss and to assess progress with weight loss. Patients were asked to:

  1. follow a calorie restricted diet, generally 1200–1400 kcal for women and 1400–1600 kcal for men
  2. keep a intake record including type, amount and calorie total and bring record to follow-up visits
  3. weigh and record at least once per week and
  4. gradually increase activity to 30–40 minutes/day, which could be achieved in multiple segments through the day, with walking being the primary recommendation, modified based upon specific patient limitations.

The frequency of visits for obese patients in the noninvasive arm coincided with the scheduled follow-up with their transplant hepatologist, and typically ranged from every 3 to 6 months. All patients with identified psychiatric issues were followed by a transplant center psychiatrist and further referred to psychology for counseling as appropriate. Activity recommendations were provided by the surgeon and further re-enforced by the dietician, and were tailored to meet individual activity restrictions and available resources. Patients with ascites had their weight measure after paracentesis. Those with edema greater than +1 on physical exam were considered to be at their goal weight when they were within 5 kg of target.

Patients with severe obesity who were unsuccessful with achievement of pretransplant target weight (BMI < 35 kg/m2) who also had a MELD score high enough to allow them to proceed to transplantation underwent a combined liver transplant (LT) plus sleeve gastrectomy (SG). These seven patients were assessed by the bariatric surgery team and completed the standard bariatric surgery education. All seven LTSG were performed using deceased donors, with a caval sparing hepatectomy and a duct to duct biliary anastomosis. The SG was performed by an experienced bariatric surgeon following completion of the LT utilizing a 9 mm endoscope placed along the lesser curve of the stomach. A combination of 45-mm 4.8 and 60-mm 3.5 Endo-GIA staple loads were used to resect the antrum and greater curve. Care was taken to leave a small amount of cardia in place in order to avoid injury to the esophagus. The staple line was oversewn using a running 3–0 PDS suture and directly visualized from the lumen of the stomach to ensure hemostasis upon completion, utilizing the previously placed endoscope.

Post-LT management included follow-up visits with transplant center dietician at 1, 4 and 12 months for the noninvasive weight loss program. Patients are counseled on a low-calorie diet plus exercise. Patients undergoing combined LT plus sleeve gastrectomy are seen at 1, 4, 6 and 12 months with transplant center dietician and with bariatric surgery team including surgery, endocrinology and psychology. The diet consisted of clear liquid diet advancing to full liquids by day 3, pureed diet × 3 weeks, soft diet × 6 weeks, advancing to general diet by approximately 3 months. Patients are given calcium, B12, multi-vitamin with iron and vitamin D supplementation.

Data were collected prospectively, and supplemented from medical record review, with approval from our Institutional Review Board. Analysis was done using Student's t-test.


Demographics and outcomes of those enrolled in the protocol who achieved their weight goal, had a MELD high enough to proceed to LT, and thus subsequently underwent LT (N = 37) are shown in Table 1. MELD shown is the calculated score, not an exception score. There were 17/37 (46%) with MELD exception score in LT alone group. Notably, while all patients in this group did achieve weight loss to their target prior to transplant, there were 12/37 patients who were found to have a higher weight upon being called in for transplantation, compared to the prior clinic visit when they were at their goal. There were seven patients with severe obesity who were unsuccessful with achievement of pretransplant target weight (BMI < 35) who also had a MELD score high enough to allow them to proceed to transplantation who underwent a combined LTSG (demographics and outcomes Tables 1 and 2.) There was one patient in this group with a MELD exception score.

Table 1. Characteristics of 37 patients enrolled in the noninvasive weight loss protocol, compared to those who underwent combined liver transplant plus sleeve gastrectomy. MELD at transplant = mean calculated value at transplant, not including exception scores. BMI at LT is corrected for ascites
CharacteristicTotal n = 37Total N = 7p-Value
  1. Diagnoses: NASH = nonalcoholic steatohepatitis; HCV = hepatitis C virus; HCC = hepatocellular carcinoma; CCA = neoadjuvant chemoradiotherapy + hilar cholangiocarcinoma alpha-1 anti-trypsin deficiency; MELD = Model for End-stage Liver Disease; O.R. = operating room; BMI = body mass index

Gender M:F16:214:30.25
diagnosisNASH = 12NASH = 4 
 HCV + HCC = 9HCV/NASH = 1 
 HCC + other = 3Alpha-1/NASH = 1 
 CCA = 3  
 All other = 10  
Age at transplant50 (range 31–67)48 (range 44–60)0.01
MELD at transplant19 (range 8–35)32 (range 11–40)<0.001
Postoperative deaths2/37 (5%)00.27
O.R. Time, mean4:21 (2:54–7:51)4:59 (4:16–7:39)0.59
Mean BMI at presentation40 (36–46)49 (46–51)<0.001
Mean BMI at transplantation33 (28–40)48 (39–52)<0.001
Mean BMI at last follow-up36 (25–45)29 (23–35)0.003
N with BMI > 35 posttransplant21/35 (60%)0/70.001
Diabetes posttransplant12/35 (34%)0/70.03
Hemoglobin A1C at 1 year5.6 (range 4.6–7)5.6 (4.5–7)0.26
Iothalmate clearance, mL/min/BSA at 1 year48.355.80.29
Number of BP meds, mean at 1 year100.05
Ultrasound (+) steatosis7/35 (20%)0/70.10
Mean follow-up35 months (range 8–61 months)17 months (range 8–33 months)0.001
Table 2. Specific patient details for those undergoing combined liver transplantation and sleeve gastrectomy. BMI at LT is corrected for ascites. MELD score is physiologic score and (exception score.)
PatientLT DateDiagnosisMELD at LTBMI at LTOR time (h)LOS (days) Post-LTBMI at F/UComplication
  1. Diagnoses: NASH = nonalcoholic steatohepatitis; HCV = hepatitis C virus; Alpha-1 = alpha-1 anti-trypsin deficiency

57 male9/09NASH26454:121623Late HAT, excess weight loss
55 female9/10HHT11 (25)484:18834Steroid-resistant rejection
53 female10/10NASH27394:16830None
54 female1/11NASH40515:04835None
48 male6/11Alpha-1, NASH40524:551328None
60 male8/11HCV, NASH40514:2913326Early graft dysfunction, leak from gastric staple line
44 male11/11NASH40487:391627None

There were two patients who had a MELD high enough to proceed to transplantation, who had not achieved weight loss, who were ultimately denied for transplantation rather than being considered for the combined operation. Both patients had significant psychosocial issues, which had become more evident in the pretransplant period and one also had a prior gastric bypass and an HCC, which progressed beyond criteria. An additional patient was taken to the operating room with the intent to perform both LT and SG, however the SG was aborted due to the intraoperative finding of what appeared to be metastatic HCC invading the diaphragm at site of prior radio-frequency ablation.

Outcomes for all patients enrolled in the obesity protocol who achieved their pretransplant weight loss goal and underwent transplantation are shown in Table 1. The average BMI at enrollment was 40 kg/m2 (range 36–46.) The mean BMI at transplant was 33 kg/m2 (range 30–37). BMI trends for the noninvasive obesity management group are demonstrated in Figure 1. There were three deaths in this group (one from severe porto-pulmonary hypertension which developed postoperatively, one from intraoperative bleeding, and one from metastatic adenocarcinoma of unknown primary at 2.5 months from transplantation) and three patients who required retransplant (two for early graft dysfunction and one for chronic rejection). Notably, for the remaining patients, 21 (60%) have a posttransplant BMI > 35 kg/m2, 35% of patients have posttransplant diabetes mellitus, and 20% demonstrating steatosis on ultrasound (Table 1).

Figure 1.

BMI trends for 37 patients managed in the noninvasive pre-LT weight loss program. Mean follow-up is 33 months.

Seven patients were unsuccessful in achieving their weight loss target prior to transplant and underwent a combined LTSG (Tables 1 and 2). BMI trends for the combined group are shown in Figure 2. All seven patients who underwent the combined procedure are currently alive with normal allograft function, significant weight loss (mean BMI = 28 kg/m2), and none of the patients currently require insulin or oral hypo-glycemic treatment. In addition, none of the patients has steatosis based on protocol ultrasound performed for all LT patients at 4 months and annually. There was minimal additional operative time for LTSG patients when compared to the noninvasive weight loss group (see Table 1). All transplanted patients regardless of whether they also had sleeve gastrectomy received standard immunosuppression consisting of tacrolimus, mycophenolate mofetil and prednisone, with the same target levels (8–12 ng/mL up to month 4, then 6–10 ng/mL during months 4–12, then 4–8 ng/mL after 12 months). Both mycophenolate mofetil and prednisone tapered to off by 3 months.

Figure 2.

BMI trends for those patients who underwent combined liver transplant plus sleeve gastrectomy (N = 7). Mean follow-up is 17 months.

Complications in the group undergoing combined sleeve gastrectomy and liver transplant included one patient with severe early graft dysfunction who subsequently developed a leak from his gastric staple line, resulting in multiple re-operations and a very prolonged hospital stay. He eventually recovered and is currently home with normal allograft function. A second patient had a hepatic artery, which could not be visualized with postoperative ultra-sound due to body habitus. He underwent an angiogram revealing normal arterial flow, but developed a large groin hematoma, which was managed conservatively. By 8 months from transplant this patient had developed excessive weight loss (weight = 60.2 kg, BMI = 20) and required dietary counseling. The patient reported he felt well but simply had no interest in eating as he never felt hungry. This patient subsequently developed a late hepatic artery thrombosis at approximately 11 months posttransplant. This resulted in multiple hepatic abscesses requiring drainage and prolonged antibiotics. These have resolved and the patient currently has normal allograft function, and a stable BMI of 23. A third complication involved a patient with one episode of steroid-resistant rejection, which required treatment with thymoglobulin. There were no other episodes of steroid resistant rejection, and tacrolimus dosing did not require additional monitoring or adjustment, compared to standard patients.


Although outcomes following LT for selected patients with obesity are broadly similar, posttransplant obesity and complications of obesity are common. The ideal approach to the management of the obese liver transplant patient remains unknown. Our current study has several important findings pertinent to management of obesity in liver transplant recipients. The first notable result of this study is that an intensive noninvasive pretransplant weight loss protocol was effective for the majority of our patients in achieving pretransplant weight loss to a BMI of <35 kg/m2, with 37 (84%) of patients able to attain a BMI at or very close to their goal weight at the time of transplant (corrected for ascites measured at time of transplant). Patients in the noninvasive pretransplant weight loss protocol who met the target weight goal at a clinic visit (with an estimation made for edema and ascites) prior to the time of transplant and therefore were not considered for the combined procedure. Among patients who underwent LT alone, weight gain following transplant was common, but not universal. This highlights both the value and limitations of noninvasive pretransplant weight loss and the need for continued monitoring of these patients to determine which patients are at greatest risk for weight gain and associated metabolic complications, and thus who may have the greatest benefit from either considering the combined LT/SG approach or other targeted noninvasive weight gain prevention strategies such as more intensive follow-up.

A novel aspect of our study is the observation that a sleeve gastrectomy at the time of LT in carefully selected patients who have failed a rigorous noninvasive weight loss program is effective and can be performed with minimal additional operative time. Weight loss in our series of patients undergoing combined LT/SG has been steady and gradual and there was no difficulty in maintaining adequate immunosuppression levels. Whether the weight loss will be maintained will require long-term follow-up, which is not yet available. In the nontransplant setting, availability of long-term data regarding durability of sleeve gastrectomy is limited. A recent report of 102 patients undergoing laparoscopic sleeve gastrectomy reported at a mean follow-up of 49 months, the % excess weight lost (EWL) was 72.3%, while this had fallen to 54.5% EBL in the 23 patients achieving 6 year follow-up [13].

Bariatric surgery performed between 6 months and 10 years after LT has been described in case reports [14-17]. All of these reported cases have had minimal complications and patient outcomes appear satisfactory but with limited follow-up detail. The reports noted that the operations (Roux-Y-bypass and bilio-pancreatic diversion) were technically demanding procedures due to the altered surgical field from the prior LT. In addition, gastric bypass surgery can also affect intestinal drug absorption impacting immunosuppression and altering the risk for rejection [17]. It is not known how many patients are referred for bariatric surgery posttransplant, but do not undergo the procedure. Patients may lack insurance coverage for bariatric procedures, they may elect not to proceed due to hesitation to undergo yet another invasive procedure, or the procedure may be deferred or denied due to ongoing issues related to the transplant such as rejection, infection, or renal insufficiency.

A case of an intraoperative placement of a gastric band at the time of liver retransplantation has been published [19]. This procedure resulted in 45% excess body weight loss and resolution of metabolic syndrome comorbidities. Gastric bands may limit access to the GI tract for postoperative complications such as biliary stricture, though gastric bands are usually adjustable so there may not be a limitation to endoscopic access. In addition, there may be a theoretical increased risk of erosion of the band through the gastric wall in a patient on long-term immunosuppression. A sleeve gastrectomy is a procedure that does not interfere with future access to the biliary system should posttransplant complications arise. The effect on immunosuppressant medication absorption is presumably less than the malabsorptive procedures, but this is not known. In addition, the weight loss experienced is generally more gradual, since it is only a restrictive procedure and not a malabsorptive procedure, thus avoiding the rapid weight loss that may be a risk for liver injury.

One patient in our series developed a leak from the gastric staple line with severe consequences, which raises the question of whether these two operations should be combined or done in a staged fashion. The benefits of combined surgery are that it involves a single operation and recovery for the patient, and thus avoids a potentially more hostile reoperative field, as well as avoiding other barriers to weight loss surgery such as insurance coverage, or delays due to complications such as rejection, infection or disease recurrence. The disadvantage is the potential for increased complications. The best way to determine the optimal approach would be a randomized trial. Another option may be to consider a staged operation over a short interval, which may avoid the difficulty associated with adhesions, though does not address the concerns of two separate operations with regard to cost and may or may not impact postoperative complications. A second important complication was excess weight loss, which occurred when our first patient's appetite became severely impaired. The excess weight loss stabilized and improved with dietary counseling. While all subsequent patients reported a dramatic decrease in appetite, none have developed excess weight loss to date. Still, this complication highlights the importance of very careful follow-up and involvement of dieticians in the follow-up visits. It is possible this loss of appetite is simply related to reduced ghrelin secretion from the gastrectomy itself [20] but any additional effect specific to liver transplant patients or immunosuppression is not known.

The role of bariatric surgery in the context of LT continues to evolve. Combined liver transplant plus sleeve gastrectomy appears technically feasible in select patients, though it is not without risk. While gastric leaks are rare in the nontransplant population, the incidence in the transplant population is unknown and the consequence of a leak in this population is severe. Careful follow-up to avoid malnutrition and excessive weight loss is necessary. Long-term outcomes regarding durability of weight loss, impact on metabolic syndrome and other unanticipated complications must also be assessed.


The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.