Impact of Prior Bariatric Surgery on Perioperative Liver Transplant

ORIGINAL ARTICLE Idriss et al.

Impact of Prior Bariatric Surgery on Perioperative Liver Transplant Outcomes Rajab Idriss,1 Jeanette Hasse,1 Tiffany Wu,2 Fatima Khan,1 Giovanna Saracino,1 Greg McKenna,1 Giuliano Testa,1 James Trotter,1 Goran Klintmalm,1 and Sumeet K. Asrani1 1 Baylor University Medical Center, Dallas, TX; and 2 Department of Internal Medicine, Cedars-Sinai Medical Center, Los Angeles, CA

Bariatric surgery (BS) is effective in treating morbid obesity, but the impact of prior BS on candidacy for liver transplantation (LT ) is unclear. We examined 78 patients with cirrhosis with prior BS compared with a concurrent cohort of 156 patients matched by age, Model for End-Stage Liver Disease score, and underlying liver disease. We compared rates of transplant denial after evaluation, delisting on the waiting list, and survival after LT. The median time from BS to LT evaluation was 7 years. Roux-en-Y gastric bypass was the most common BS procedure performed (63% of cohort). Nonalcoholic fatty liver disease was the leading etiology for liver cirrhosis (47%). Delisting/death on the waiting list was higher among patients with BS (33.3% versus 10.1%; P = 0.002), and the transplantation rate was lower (48.9% versus 65.2%; P = 0.03). Intention-to-treat (ITT) survival from listing to 1 year after LT was lower in the BS cohort versus concurrent cohort (1-year survival, 84% versus 90%; P = 0.05). On adjusted analysis, a history of BS was associated with an increased risk of death on the waiting list (hazard ratio [HR], 5.7; 95% confidence interval [CI], 2.2-15.1), but this impact was attenuated (HR, 4.9; 95% CI, 1.8-13.4) by the presence of malnutrition. When limited to matched controls by sex, mortality attributed to BS was no longer significant for females (P = 0.37) but was significant for males (P = 0.046). Sarcopenia, as captured by skeletal muscle index, was calculated in a subset of patients (n = 49). The total skeletal surface area was lower in the BS group (127 [105-141] cm2 versus 153 [131-191] cm2; P = 0.005). Rates of sarcopenia were higher among patients delisted after listing (71.4% versus 16.7%; P = 0.04). In conclusion, a history of BS was associated with higher rates of delisting on the waiting list as well as lower survival from the time of listing on ITT analysis. Presence of malnutrition and sarcopenia among patients with BS may contribute to worse outcomes. Liver Transplantation 25 217‒227 2019 AASLD. Received June 8, 2018; accepted October 18, 2018.

SEE EDITORIAL ON PAGE 203 (HTN), cardiovascular disease, and nonalcoholic fatty liver disease (NAFLD).(3) Bariatric surgery (BS) is considered one of the most effective options to treat Obesity is a major health burden worldwide. More than (1,2) severe obesity and reduces obesity-related comorbidi- one-third of the US population is considered obese. ties, such as NAFLD,(4,5) induces the disappearance of Morbid obesity is associated with an increased prev- steatohepatitis, reduces fibrosis, and improves relevant alence of diabetes mellitus (DM), hypertension metabolic comorbidities.(6-8) Increasingly, patients with decompensated cirrhosis Abbreviations: ALD, alcoholic liver disease; BMI, body mass index; who underwent BS are encountered for consideration of BS, bariatric surgery, CC, cryptogenic cirrhosis; CI, confidence liver transplantation (LT). The short-term and longterm interval; CKD, chronic kidney disease; CT, computed tomography; surgical complications of BS are well documented, DM, diabetes mellitus; HCV, hepatitis C virus; HR, hazard ratio; (7,9) HTN, hypertension; IQR, interquartile range; ITT, intention- including infection and nutrient deficiencies. Over to-treat; LT, liver transplantation; MELD, Model for End-Stage the last decade, several case series have described liver

Liver Disease; NAFLD, nonalcoholic fatty liver disease; NASH, failure after BS.(10-13) In most of the reports, liver fail- nonalcoholic steatohepatitis; RYGB, Roux-en-Y gastric bypass; SGA, subjective global assessment; SMI, skeletal muscle index. ure occurred early postoperatively, presented with quick

deterioration of liver function, and may have been asso- Address reprint requests to Sumeet K. Asrani, M.D., M.Sc., Baylor (12,14) University Medical Center, 3410 Worth Street, Suite 860, Dallas, ciated with rapid weight loss and malnutrition. TX 75246. Telephone: 214-820-8500; FAX: 214-820-0993; However, the effect of BS on liver function and the nutri- E-mail: [email protected] tion consequences in patients with cirrhosis, regardless of etiology, years after BS are not completely clear.

Original Article | 217 Idriss et al. Liver Transplantation, February 2019

Malnutrition occurs frequently in patients undergo- (MELD) score at evaluation, and etiology of disease. ing LT and adversely affects transplant outcomes.(15) Patients undergoing retransplant as well as those with Sarcopenia defined as the loss of muscle mass and hepatocellular carcinoma were also excluded to create function(16,17) has also been recognized in recent years a cohort of patients with decompensated liver disease as a predictor of pre- and post-LT outcomes.(18,19) being considered for primary LT. Sarcopenic obesity, characterized by loss of muscle mass and preservation of fat, is a condition that can be DATA SOURCE present in patients with cirrhosis who have undergone BS, and this sarcopenia may play a role in the outcomes We examined center-specific data using a prospec- of this group of patients.(20-22) tively maintained locked database capturing inpatient, We hypothesized that compared with matched outpatient, and nontransplant-related encounters in all patients who had not undergone BS, patients with evaluated patients before and after LT with seamless cirrhosis who had undergone BS would have worse linkage to clinical data and longterm outcomes. We outcomes on the waiting list as well as increased compli- extracted data on characteristics of liver disease (eg, cations and reduced graft and patient survival post-LT decompensation), demographics, characteristics of partially driven by malnutrition and sarcopenia. weight loss (amount of weight loss, time of BS relative to evaluation for LT, and stage of fibrosis). In addition, the type of BS was noted: Rou x-en-Y gastric bypass (RYGB), sleeve gastrectomy, adjustable gastric band, Patients and Methods or biliopancreatic diversion with duodenal switch. The primary aim of this study was to identify the independent impact of prior BS on outcomes among pa- NUTRITION STATUS tients on the LT waiting list as well as after LT. The Given that patient outcomes may be impacted by secondary outcome was to identify the effect of mal- numerous factors, we examined the role of nutrition nutrition and sarcopenia on outcomes before and after status as well as the presence of sarcopenia. The pres- LT among patients with prior BS. ence of malnutrition was assessed using the validated subjective global assessment (SGA) tool, and patients PATIENTS were classified as well-nourished or moderately or severely malnourished by trained registered dietitians We examined adult patients who were evaluated for LT (23-25) specializing in transplantation. The moderately between 1985 and 2017 at Baylor Simmons Transplant Institute. We identified adult patients who had a his- and severely malnourished patients were combined into a single group and identified as malnourished. tory of BS prior to evaluation for consideration of primary LT. Patients were matched (2:1) to a concurrent SGA has been consistently used at our institution since (26) To assess sarcopenia, abdominal computed cohort on age, Model for End-Stage Liver Disease 1987. tomography (CT) images within the 12 months prior to the date of LT evaluation among patients in the BS Authors had access to all the study data, take responsibility for cohort and concurrent cohort were analyzed for body the accuracy of the analysis, and had authority over manuscript composition. For the 2 study groups, transverse CT preparation and the decision to submit the manuscript for publication. images at the third lumbar vertebra were downloaded All authors approve the manuscript. as a digital imaging and communications in medicine

Funding is provided by a Baylor foundation grant. file and were analyzed using a novel software, Slice- James Trotter is on the speaker’s bureau for Gilead. O-Matic, version 4.3 (Tomovision, Montreal, Canada)

Additional supporting information may be found in the online version by a trained investigator (R.I.) according to previously (27) of this article. published protocols. Data extracted included total abdominal muscle area, intramuscular adipose tissue, Copyright © 2018 by the American Association for the Study of Liver Diseases. subcutaneous adipose tissue, and visceral adipose tissue. Sarcopenia was defined according to the following View this article online at wileyonlinelibrary.com. equation for the skeletal muscle index (SMI): (total ab-

DOI 10.1002/lt.25368 dominal muscle area in cm2)/(height in m2). The cutoff points for sarcopenia were determined according to

218 | Original Article Liver Transplantation, Vol. 25, No. 2, 2019 Idriss et al. the previously reported values for patients with cirrho- with 63% maintaining their weight loss by self-report; sis: males, <50 cm2/m2 and females, <39 cm2/m2.(28) 83% of BS patients had a previous history of abdominal surgeries (other than the BS). The date of liver cirrhosis

STATISTICAL ANALYSIS diagnosis was available for 51% (40/78) of the patients from the BS cohort. Liver cirrhosis was diagnosed after The primary outcomes when comparing the BS co- BS in 27 patients, 9 were diagnosed at time of BS, and hort with the concurrent cohort included denial for 4 patients were diagnosed before BS. The median time LT, listing for LT transplantation, delisting due to between the BS and evaluation for LT was 7 years. The death or deterioration in condition on the waiting list, demographics and clinical characteristics of all the pa- intention-to-treat (ITT) survival (from listing), and tients are shown in Table 1. By design, median (IQR) age survival within 1 and 3 years after LT. Having shown (54 [46-61] years versus 54 [47-62] years) and MELD inferior outcomes among patients with BS, we exam- score (12 [8-20] versus 13 [9-18]) at evaluation were sim- ined factors associated with increased mortality on the ilar between the BS and concurrent cohorts. Patients in waiting list. This included disease severity as well as the BS cohort tended to be female (83.3% versus 44.2%; sex, type of surgery (RYGB versus non-RYGB), mal- P < 0.01). The distribution of metabolic factors (DM, nutrition, and SMI. obesity, HTN, and body mass index [BMI]) was not Continuous data are presented as median and different between the 2 groups. The percent of patients interquartile range (IQR; 25th-75th percentiles). who were malnourished at the time of LT evaluated as Dichotomous data are presented as percentages. Group judged by SGA was higher in the BS versus concurrent comparisons were performed using nonparametric cohort (64.1% versus 39%; P < 0.01). statistical tests. The χ2 test or Fisher’s exact test was applied for dichotomous and categorical data. The Wilcoxon rank sum test was used to compare numer- OUTCOMES ON THE WAITING ical data of the 2 groups. Survival time was measured LIST from the time of listing (ITT survival) as well as after Transplant candidacy denial was numerically lower in

LT and censored at last contact or end of study. In the the BS versus concurrent cohorts (42.3% versus 55.8%;

ITT survival, given that mortality either before or after P = 0.05; Fig. 1). However, when comparing the BS

LT was considered an event of interest, censoring was cohort with the concurrent cohort, rates of delisting not done at LT. Kaplan-Meier survival curves were used or death on the waiting list were higher (33.3% ver- to present patient survival from listing and from LT, sus 10.1%; P = 0.002), and the rate of LT was lower and log-rank tests were used to compare time to failure from the time of listing (48.9% versus 65.2%; P = 0.03). between groups. Univariate and adjusted logistic regres- The median time to LT was similar in patients with sion models were used to assess the association between BS (118 days; IQR, 7-551 days) as compared with history of BS and delisting from the waiting list. the concurrent cohort (91 days; IQR, 8-436 days;

All statistical analyses used the SAS Enterprise P = 0.87). ITT survival from time of listing was lower

Guide statistical package, version 9.4 (SAS Institute in patients who had undergone BS compared with the

Inc., Cary, NC) and R statistical software, version 3.3.2 concurrent cohort at 1 year (84% versus 90%; 95%

(R Foundation for Statistical Computing, Vienna, confidence interval [CI], 74%-96% versus 82%-97%, < Austria). Statistical significance was set at P 0.05. respectively) and 3 years (72% versus 82%; 95% CI,

The study was approved by the institutional review 59%-87% versus 73%-91.7%, respectively; P = 0.055; board. Fig. 2). ITT survival from time of listing was lower in patients who had undergone BS compared with the concurrent cohort even after the exclusion of patients Results with acute liver failure (P = 0.048; Supporting Fig. 1). BASELINE CHARACTERISTICS

REASONS FOR REJECTION AND The BS cohort included 78 patients who were matched DELISTING to 156 control patients. The most common BS procedure was RYGB (63%). The mean weight loss between the Reasons for rejection (n = 76) and delisting (n = 62) BS and the time for evaluation for LT was 130 pounds were available in a subset of patients. Rejection rates

Original Article | 219 Idriss et al. Liver Transplantation, February 2019

TABLE 1. Characteristics of Patients With Cirrhosis Undergoing Transplant Evaluation With and Without a History of BS

Characteristic at Time of Transplant Evaluation Patients With a History of BS (n = 78) Concurrent Cohort (n = 156) P Value

Age, years 54 (46-61) 54 (47-62) 0.88 Sex, female 65 (83.3) 69 (44.2) <0.01 Race 0.20 White 67 (85.9) 136 (87.2) Black 6 (7.69) 11 (7.1) Other/unknown 5 (6.41) 4 (2.56) Hispanic 4 (5.13) 28 (17.95) 0.01 Etiology of liver disease NASH/CC 37 (47.4) 74 (47.4) 1.00 HCV 19 (24.4) 49 (31.4) 0.26 ALD 15 (19.2) 28 (17.9) 0.81 Acute or fulminant liver failure 0 (0) 6 (3.8) 0.02 BMI, kg/m2 29.7 (25.3-35.6) 28.7 (25.1-32.9) 0.22 MELD score 12 (8.0-20.0) 13.0 (9.0-18.0) 0.54 Platelets (normal 140-440 109/L) 117.5 (89.0-172.0) 90.0 (59.5-134.5) <0.01 Albumin (normal 3.5-5 g/dL) 3.2 (2.8-3.7) 3.3 (2.8-3.7) 0.72 Marker of decompensation Ascites* 36 (69.2) 86 (66.2) 0.6 Encephalopathy 52 (69.2) 104 (66.7) 1.00 Gastrointestinal bleed 21 (26.9) 48 (30.8) 0.54 Anasarca 43 (55.1) 39 (25.0) <0.01 Comorbidities DM 25 (32.1) 62 (39.7) 0.24 HTN 30 (38.5) 51 (32.7) 0.38 CKD 12 (15.4) 12 (7.7) 0.07 Malnourished by SGA 25/39 (64.1) 54/140 (39) <0.01 Body composition (n = 49) Sarcopenic 30% 37% 0.68 2 2 SMI, cm /m 44.0 (39.0-53.0) 51.0 (42.0-57.5) 0.07 SMI, males 46.0 (45.0-50.0) 54.0 (45.0-59.0) 0.20 SMI, females 43.0 (37.5-53.2) 41.0 (38.5-51.0) 0.88 2 Skeletal muscle surface, cm 127.0 (104.2-141.2) 152.6 (130.3-193.8) 0.005 2 Intramuscular adipose, cm 8.51 (4.7-10.9) 6.07 (4.0-11.3) 0.33 2 Visceral adipose tissue, cm 235 (191.4-309.0) 249.5 (188.8-334.2) 0.26 2 Subcutaneous adipose, cm 236 (196-307) 250 (189-325) 0.65

NOTE: Data are given as n (%) and median (IQR) unless otherwise noted. *Data on ascites are missing for 26 patients.

for comorbidities were similar in patients with BS TYPE OF WEIGHT REDUCTION versus those without (40% versus 54%; P = 0.03). SURGERY Delisting rates for comorbidities (27.8% versus 20.5%; P = ns) and noncompliance (16.7% versus Rates of delisting or death after listing were higher 9.1%; P = ns) were similar in patients with BS versus for RYGB versus non-RYGB surgeries (44% versus those without. 16.7%; P = 0.04). The rate of transplantation was

220 | Original Article Liver Transplantation, Vol. 25, No. 2, 2019 Idriss et al.

FIG. 1. Outcomes from evaluation among patients with a history of BS compared with a matched concurrent cohort matched by age, MELD, and diagnosis.

FIG. 2. ITT survival from time of listing for LT in BS versus concurrent cohort matched by age, MELD, and diagnosis. lower (23.4% versus 61.1%; P < 0.01). The presence of Data on sarcopenia according to type of BS were moderate or severe malnutrition by SGA was 68% in limited. However, the overall SMI and skeletal mus- RYGB versus 57% in non-RYGB patients (P = 0.5). cle surface measures were lower in the RYGB group

Original Article | 221 Idriss et al. Liver Transplantation, February 2019 versus the non-RYGB group. Overall, the SMI (42.0 SARCOPENIA [36.0-45.0] cm2/m2 versus 52.0 [43.0-57.0] cm2/m2;

P < 0.01) was lower in RYGB as compared with the Given the association between nutrition status and non-RYGB group (Supporting Table 1). outcomes after BS observed in the study, we hypothesized that rates of delisting and mortality on the LT waiting list were driven by malnutrition and sarcope- ROLE OF SEX nia. We evaluated a convenience cohort of 49 patients

Given that a large number of patients in the BS surgery (25 from the BS group and 24 from concurrent con- were female, we further explored the role of sex. When trols who had abdominal CT scans). Sarcopenia was limited to matched controls by sex, mortality attributed associated with malnutrition (SMI below established to BS was no longer significant for females (Fig. 3B) cutoff for patients for cirrhosis: 50% moderate/severely = but was significant for males (Fig. 3A). To explore this malnourished versus 7.7% well nourished; P 0.02). further, we generated new controls matched on age, In this selected group, the SMI was lower in the BS diagnosis, sex, and MELD score. Lack of significance group versus the control group (P = 0.07). The total was seen for females, but the reduced mortality in the skeletal muscle surface area was lower in the BS group 2 BS persisted in males (data not shown). (median [IQR]; 127 [105-141] cm versus 153 [131- 191] cm2; P = 0.005). The percentage of patients with sarcopenia (SMI below sex-specific cutoffs for patients TRANSPLANTATION with cirrhosis) was higher among delisted patients The median time to LT was 54 months (IQR, 25- (5 [71.4%] versus 2 [16.7%]; P = 0.04; Tables 2 and 3). 122 months) in the BS group and 60 months (IQR, 25-82 months) in the concurrent control group. Comparing the BS with the concurrent cohort, median (IQR) BMI at LT was 28.6 kg/m2 (26.0-35.7 kg/ Discussion m2) versus 28.8 kg/m2 (26.1-32.3 kg/m2), and median As the prevalence of obesity and BS in the general pop- (IQR) MELD at LT was 22.0 (14.5-33.0) versus 18.0 ulation increases, the number of patients with cirrho- (13.0-24.0). Prevalence of malnutrition according to sis who have undergone BS also is likely to increase. SGA at time of LT was 78.6% in the BS group and In this matched study, we examined the association 33.3% in the concurrent cohort (P = 0.01). For those of prior BS on outcomes before and after LT. Patients who underwent LT, survival was similar (Fig. 4). The with cirrhosis who underwent BS had higher rates of 1- and 3-year survival rates for BS versus the concur- delisting and lower rates of LT. In addition, survival rent cohort was 85% (95% CI, 71%-100%) and 84% from time of listing was lower when compared with a (95% CI, 74%-96%) versus 85% (95% CI, 71%-100%) concurrent cohort who had not undergone BS. There and 82% (95% CI, 72%-94%; P = 0.57), respectively. was a higher predominance of females and presence of The length of stay for the initial hospitalization was malnutrition in the BS group; even after adjustment for similar in the BS group (median [IQR]; 14 [8-21] days relevant factors, BS was an independent predictor of re- versus 9 [6-17] days; P = 0.13). The number of patients moval from the waiting list. Conversely, survival in the with at least 1 readmission (52% versus 52%) was sim- BS group was similar to that of a matched concurrent ilar (P = 1.0). cohort in the post-LT period. This suggests that there may be a selection bias whereby those who will not do

well after LT have an increased risk of dying while ADJUSTED ANALYSIS waiting for LT but that once they undergo LT, their We examined predictors of delisting and death. On risk is the same as patients who have not undergone BS. univariate analysis, the presence of BS was associated The presence of malnutrition as well as sarcopenia with a higher risk of delisting (hazard ratio [HR], 5.7; may partly account for the worse outcomes in patients 95% CI, 2.2-15.1). MELD score and sex were not sig- with a history of previous BS on the LT waiting list. nificant predictors. On bivariate analysis after adjust- Despite similar BMIs between the 2 groups, the prev- ment for malnutrition (HR, 1.9; 95% CI, 0.69-5.3), alence of malnutrition was higher in the BS group at this impact was attenuated but still significant (HR, the time of LT evaluation with 64% being malnour- 4.9; 95% CI, 1.8-13.4). ished versus 39% of the concurrent cohort. In addition

222 | Original Article Liver Transplantation, Vol. 25, No. 2, 2019 Idriss et al.

FIG. 3. (A) ITT survival from time of listing for LT in BS versus concurrent cohort matched by age, MELD, and diagnosis among males. (B) ITT survival from time of listing for LT in BS versus concurrent cohort matched by age, MELD, and diagnosis among females.

to liver cirrhosis and an associated catabolic state due to altered anatomy and physiology leading to nutri- to multiple factors,(29) patients post-BS surgery are at ent malabsorption, decreased food intake attributed an increased risk for nutritional deficiencies secondary to reduction in hunger sensation as well as increased

Original Article | 223 Idriss et al. Liver Transplantation, February 2019

FIG. 4. Survival after LT in BS versus concurrent cohort matched by age, MELD, and diagnosis.

TABLE 2. Relationship Between Sarcopenia by Malnutrition Status by SGA in Patients With Cirrhosis

Malnourished Patients (n = 16) Well-Nourished Patients (n = 13) P Value

SMI below sarcopenia cutoff 50.0% 7.7% 0.02 SMI, cm2/m2 42.5 (38.0-45.5) 55.0 (51.0-60.0) <0.01 Skeletal muscle surface, cm2 123.5 (103.1-136.5) 164.1 (140.3-200.4) <0.01 Intramuscular adipose, cm2 6.60 (5.10-10.49) 5.09 (2.62-9.12) 0.26 Visceral adipose tissue, cm2 106.3 (65.9-125.8) 139.6 (83.9-176.2) 0.83 Subcutaneous adipose, cm2 225.4 (152.3-268.7) 304.3 (238.6-382.1) 0.01

NOTE: Data are given as percentage of patients or as median value (IQR).

TABLE 3. Relationship Between Sarcopenia by Listing Status in Patients With Cirrhosis

Delisted or Death After Listing (n = 7) Listed (n = 12) P Value

Malnourished 4 (57.1) 6 (50.0) 0.69 SMI below cutoff for sarcopenia 5 (71.4) 2 (16.7) 0.04 SMI, cm2/m2 40.0 (37.0-45.5) 42.5 (40.8-47.0) 0.35 Skeletal muscle surface, cm2 111.4 (98.4-151.0) 126.0 (108.4-155.5) 0.45 Intramuscular adipose, cm2 6.30 (4.49-11.0) 6.99 (5.80-12.6) 0.47 Visceral adipose tissue, cm2 75.2 (42.7-109.0) 129.0 (108.0-149.0) 0.03 Subcutaneous adipose, cm2 200 (178-232) 265 (223-320) 0.05

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

224 | Original Article Liver Transplantation, Vol. 25, No. 2, 2019 Idriss et al. satiety and food intolerance. This difference may fur- Patient selection may play a role in our study both ther be modified by type of surgery, namely, malab- at the time of BS as well as at the time of listing. It sorptive surgery (eg, RYGB) versus restrictive surgery is possible that some of the patients in the BS cohort (eg, sleeve gastrectomy). For the former, one would inappropriately underwent BS, which may have con- conjecture that because RYGB includes malabsorp- tributed to decompensation that could not be sal- tive and restrictive components, the rate of malnutri- vaged by listing for LT. Furthermore, after listing, tion and micronutrient deficiencies would be higher patients with cirrhosis with BS waited longer for a LT than that of those who underwent sleeve gastrectomy. and may have a longer time to become progressively Despite intestinal surface area remaining intact after debilitated. sleeve gastrectomy, micronutrient deficiencies still Sex may also play a role though the study might occur in those patients.(30) However, protein malnu- have been underpowered to show a difference. When trition and micronutrient deficiencies tend to be more stratified by sex, though numerically lower survival was prevalent among those undergoing malabsorptive pro- seen in females, survival rate was not statistically sig- cedures.(31) BS guidelines recommend longterm nutri- nificant. However, the pattern of increased mortality ent monitoring and supplementation for all patients associated with BS persisted in men. Our results are who have undergone BS.(32) For patients with previous consistent with other studies evaluating risk factors for BS undergoing transplantation, this may translate into mortality after BS whereby liver disease and male sex dedicated follow-up with a transplant nutritionist, cor- are independent predictors of mortality.(41-43) rection of nutrient deficiencies, and consideration of Our study has several strengths. We assembled a enteral feeding both before and after transplantation. large cohort and appropriately matched them to a con- Second, sarcopenia may play a role. In recent years, current group. We examined both pre- and post-LT frailty, functional decline, and sarcopenia have been outcomes. We explored reasons for the findings by an recognized as common complications in patients with additional analysis and exploration of sarcopenia as a cirrhosis, and they are associated with higher wait- putative cause. Formal assessment by dietitians and the list mortality.(15-19,27,33-35) Previous studies that have use of validated measures further complemented the assessed sarcopenia in LT candidates using CT scans study. show that reduced skeletal muscle mass is associated There are several limitations to our study. First, with increased mortality after LT. (19,36-38) Similar to this is a single-center study, which affects the gener- our findings, Tandon et al. also reported increased alizability of the study. Second, inherent differences mortality among sarcopenic patients while on the LT in patient selection may be incompletely captured waiting list.(27) In another related study,(39) sarcopenia despite the presence of a matched concurrent cohort. was not associated with post-LT survival, but sarcope- Third, assessment of sarcopenia was done based on nic patients had a prolonged hospital stay when com- the cross-sectional imaging available for analysis. pared with nonsarcopenic patients. Therefore, we had a small number in each group and Sarcopenic obesity, characterized by loss of muscle a risk for selection bias. There were no serial measure- and preservation of fat, is a condition that may be pres- ments of body composition to determine if the degree ent in patients with cirrhosis who have undergone BS of sarcopenia changed before and after LT. Data on and may play a role in the outcomes of this group of multiple assessments on the waiting list were limited. patients.(20-22) Patients with BS may potentially serve The cause of malnutrition in transplant patients is as a model for sarcopenic obesity, whereby patients multifactorial but includes altered metabolism, reduced may have lost muscle mass but retained fat mass.(20,40) intake, and increased malabsorption. These factors are Inflammation induced by adipose tissue may negatively enhanced in patients who have undergone BS and may impact muscle mass commencing the process and pro- have resulted in a higher rate of malnutrition at pre- gression of sarcopenic obesity.(22) Comparing 2 small sentation for LT evaluation in patients who had under- selected groups of patients in our study, SMI and total gone BS versus those without BS. Repeated nutrition surface area were lower in the BS group. However, the assessments were not completed in patients during the BMI and subcutaneous and visceral fat percentages pretransplant phase, so it was not possible to determine were similar between the groups, suggesting that the if further nutrition deterioration occurred to a greater BS group could be considered as having sarcopenic extent in the BS versus control group while on the LT obesity. waiting list.

Original Article | 225 Idriss et al. Liver Transplantation, February 2019

In conclusion, patients with cirrhosis and a history 12) Kruschitz R, Luger M, Kienbacher C, Trauner M, Klammer C, of BS had higher rates of delisting as well as lower sur- Schindler K, et al. The effect of Rou x-en-Y vs. omega-loop gas- vival from time of listing when compared with a con- tric bypass on liver, metabolic parameters, and weight loss. Obes Surg 2016;26:2204-2212. current cohort without BS. BS is an accepted option 13) Cazzo E, Pareja JC, Chaim EA. Liver failure following bil- for treating morbid obesity, but the implications on iopancreatic diversions: a narrative review. Sao Paulo Med J patients with cirrhosis years after the surgery are a con- 2017;135:66-70. 14) Geerts A, Darius T, Chapelle T, Roeyen G, Francque S, cern. Optimizing management should include proper Libbrecht L, et al. The multicenter Belgian survey on liver patient selection for LT and identifying high-risk transplantation for hepatocellular failure after bariatric surgery. patients, taking into consideration important factors, Transplant Proc 2010;42:4395-4398. such as malnutrition and sarcopenia. Further studies 15) Kalafateli M, Mantzoukis K, Choi Yau Y, Mohammad AO, Arora S, Rodrigues S, et al. Malnutrition and sarcopenia pre- are needed to assess whether suboptimal outcomes are dict post-liver transplantation outcomes independently of the driven by patient selection at the time of BS, advanced Model for End-Stage Liver Disease score. J Cachexia Sarcopenia liver disease at the time of BS, type of BS, or devel- Muscle 2017;8:113-121. 16) Cruz-Jentoft AJ, Baeyens JP, Bauer JM, Boirie Y, Cederholm T, opment of malnutrition and its complications on the Landi F, et al. for European Working Group on Sarcopenia in waiting list. Older People. Sarcopenia: European consensus on definition and diagnosis: report of the European Working Group on Sarcopenia in Older People. Age Ageing 2010;39:412-423. REFERENCES 17) Janssen I. Evolution of sarcopenia research. Appl Physiol Nutr Metab 2010;35:707-712. 1) Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of 18) Golse N, Bucur PO, Ciacio O, Pittau G, Sa Cunha A, Adam obesity in the United States, 2009–2010. NCHS Data Brief R, et al. A new definition of sarcopenia in patients with 2012;82:1-8. cirrhosis undergoing liver transplantation. Liver Transpl 2017;23: 2) Flegal KM, Carroll MD, Kit BK, Ogden CL. Prevalence of obe- 143-154. sity and trends in the distribution of body mass index among US 19) Masuda T, Shirabe K, Ikegami T, Harimoto N, Yoshizumi T, adults, 1999–2010. JAMA 2012;307:491-497. Soejima Y, et al. Sarcopenia is a prognostic factor in living donor 3) Abdelaal M, le Roux CW, Docherty NG. Morbidity and mortal- liver transplantation. Liver Transpl 2014;20:401-407. ity associated with obesity. Ann Transl Med 2017;5:161. 20) Carias S, Castellanos AL, Vilchez V, Nair R, Dela Cruz AC, 4) Caiazzo R, Lassailly G, Leteurtre E, Baud G, Verkindt H, Watkins J, et al. Nonalcoholic steatohepatitis is strongly as- Raverdy V, et al. Rou x-en-Y gastric bypass versus adjustable gas- sociated with sarcopenic obesity in patients with cirrhosis un- tric banding to reduce nonalcoholic fatty liver disease: a 5-year dergoing liver transplant evaluation. J Gastroenterol Hepatol controlled longitudinal study. Ann Surg 2014;260:893-898. 2016;31:628-633. 5) Courcoulas AP, Christian NJ, Belle SH, Berk PD, Flum DR, 21) Montano-Loza AJ, Angulo P, Meza-Junco J, Prado CM, Sawyer Garcia L, et al. for Longitudinal Assessment of Bariatric Surgery MB, Beaumont C, et al. Sarcopenic obesity and myosteatosis (LABS) Consortium. Weight change and health outcomes at 3 are associated with higher mortality in patients with cirrhosis. years after bariatric surgery among individuals with severe obe- J Cachexia Sarcopenia Muscle 2016;7:126-135. sity. JAMA 2013;310:2416-2425. 22) Schrager MA, Metter EJ, Simonsick E, Ble A, Bandinelli S, 6) Lassailly G, Caiazzo R, Buob D, Pigeyre M, Verkindt H, Lauretani F, Ferrucci L. Sarcopenic obesity and inflammation Labreuche J, et al. Bariatric surgery reduces features of nonalco- in the InCHIANTI study. J Appl Physiol (1985) 2007;102: holic steatohepatitis in morbidly obese patients. Gastroenterology 919-925. 2015;149:379-388. 23) Hasse J, Strong S, Gorman MA, Liepa G. Subjective global 7) Chang SH, Stoll CR, Song J, Varela JE, Eagon CJ, Colditz assessment: alternative nutrition-assessment technique for liv- GA. The effectiveness and risks of bariatric surgery: an updated er-transplant candidates. Nutrition 1993;9:339-343. systematic review and meta-analysis, 2003–2012. JAMA Surg 24) Yadav SK, Choudhary NS, Saraf N, Saigal S, Goja S, Rastogi 2014;149:275-287. A, et al. Nutritional status using subjective global assessment 8) Nostedt JJ, Switzer NJ, Gill RS, Dang J, Birch DW, de Gara C, independently predicts outcome of patients waiting for et al. The effect of bariatric surgery on the spectrum of fatty liver living donor liver transplant. Indian J Gastroenterol 2017;36: disease. Can J Gastroenterol Hepatol 2016;2016:2059245. 275-281. 9) Flum DR, Belle SH, King WC, Wahed AS, Berk P, Chapman 25) Marr KJ, Shaheen AA, Lam L, Stapleton M, Burak K, Raman W, et al. for Longitudinal Assessment of Bariatric Surgery M. Nutritional status and the performance of multiple bedside (LABS) Consortium. Perioperative safety in the longitudinal as- tools for nutrition assessment among patients waiting for liver sessment of bariatric surgery. N Engl J Med 2009;361:445-454. transplantation: a Canadian experience. Clin Nutr ESPEN 10) Mayo Ossorio MA, Pacheco Garcia JM, Pérez Gomar D. 2017;17:68-74. Bazán Hinojo Mdel C, Vilchez Lopez FJ, Aguilar Diosdad 26) Detsky AS, McLaughlin JR, Baker JP, Johnston N, Whittaker M, Fernández Serrano JL. Long-term fulminant hepatic fail- S, Mendelson RA, Jeejeebhoy KN. What is subjective global ure in patients undergoing gastric bypass for morbid obesity [in assessment of nutritional status? JPEN J Parenter Enteral Nutr Spanish]. Nutr Hosp 2015;32:430-434. 1987;11:8-13. 11) Motamedi MAK, Barzin M, Ebrahimi M, Ebrahimi R, Khalaj 27) Tandon P, Ney M, Irwin I, Ma MM, Gramlich L, Bain VG, A. Severe fatal protein malnutrition and liver failure in a mor- et al. Severe muscle depletion in patients on the liver transplant bidly obese patient after mini-gastric bypass surgery: case report. wait list: its prevalence and independent prognostic value. Liver Int J Surg Case Rep 2017;33:71-74. Transpl 2012;18:1209-1216.

226 | Original Article Liver Transplantation, Vol. 25, No. 2, 2019 Idriss et al.

28) Carey EJ, Lai JC, Wang CW, Dasarathy S, Lobach I, Montano- 36) Englesbe MJ, Patel SP, He K, Lynch RJ, Schaubel DE, Harbaugh Loza AJ, Dunn MA, for Fitness, Life Enhancement, and C, et al. Sarcopenia and mortality after liver transplantation. Exercise in Liver Transplantation Consortium. A multicenter J Am Coll Surg 2010;211:271-278. study to define sarcopenia in patients with end-stage liver dis- 37) Hamaguchi Y, Kaido T, Okumura S, Kobayashi A, Shirai H, ease. Liver Transpl 2017;23:625-633. Yagi S, et al. Impact of skeletal muscle mass index, intramuscu- 29) Kondrup J, Müller MJ. Energy and protein requirements lar adipose tissue content, and visceral to subcutaneous adipose of patients with chronic liver disease. J Hepatol 1997;27: tissue area ratio on early mortality of living donor liver transplan- 239-247. tation. Transplantation 2017;101:565-574. 30) Via MA, Mechanick JI. Nutritional and micronutrient care of 38) Montano-Loza AJ, Meza-Junco J, Prado CM, Lieffers JR, bariatric surgery patients: current evidence update. Curr Obes Baracos VE, Bain VG, Sawyer MB. Muscle wasting is associ- Rep 2017;6:286-296. ated with mortality in patients with cirrhosis. Clin Gastroenterol 31) Lupoli R, Lembo E, Saldalamacchia G, Avola CK, Angrisani Hepatol 2012;10:166-173. L, Capaldo B. Bariatric surgery and long-term nutritional issues. 39) Montano-Loza AJ, Meza-Junco J, Baracos VE, Prado CM, Ma World J Diabetes 2017;8:464-474. M, Meeberg G, et al. Severe muscle depletion predicts postoper- 32) Parrott J, Frank L, Rabena R, Craggs-Dino L, Isom KA, ative length of stay but is not associated with survival after liver Greiman L. American society for metabolic and bariatric surgery transplantation. Liver Transpl 2014;20:640-648. integrated health nutritional guidelines for the surgical weight 40) Baumgartner RN. Body composition in healthy aging. Ann N Y loss patient 2016 update: micronutrients. Surg Obes Relat Dis Acad Sci 2000;904:437-448. 2017;13:727-741. 41) Benotti P, Wood GC, Winegar DA, Petrick AT, Still CD, 33) Lai JC, Feng S, Terrault NA, Lizaola B, Hayssen H, Covinsky Argyropoulos G, Gerhard GS. Risk factors associated with K. Frailty predicts waitlist mortality in liver transplant candi- mortality after Rou x-en-Y gastric bypass surgery. Ann Surg dates. Am J Transplant 2014;14:1870-1879. 2014;259:123-130. 34) Lai JC, Dodge JL, Sen S, Covinsky K, Feng S. Functional de- 42) Inaba CS, Koh CY, Sujatha-Bhaskar S, Silva JP, Chen Y, Nguyen cline in patients with cirrhosis awaiting liver transplantation: DV, Nguyen NT. One-year mortality after contemporary lapa- Results from the functional assessment in liver transplantation roscopic bariatric surgery: an analysis of the bariatric outcomes (FrAILT) study. Hepatology 2016;63:574-580. longitudinal database. J Am Coll Surg 2018;226:1166-1174. 35) DiMartini A, Cruz RJ Jr, Dew MA, Myaskovsky L, Goodpaster 43) Quilliot D, Sirveaux MA, Nomine-Criqui C, Fouquet T, Reibel B, Fox K, et al. Muscle mass predicts outcomes following liver N, Brunaud L. Evaluation of risk factors for complications after transplantation. Liver Transpl 2013;19:1172-1180. bariatric surgery. J Visc Surg 2018;155:201-210.

Original Article | 227