- Email: [email protected]
Bariatric Surgery in Patients With Cirrhosis With and Without Portal Hypertension: A Single-Center Experience
ORIGINAL ARTICLE
Bariatric Surgery in Patients With Cirrhosis With and Without Portal Hypertension: A Single-Center Experience Laura Pestana, DO; James Swain, MD; Ross Dierkhising, MS; Michael L. Kendrick, MD; Patrick S. Kamath, MD; and Kymberly D. Watt, MD Abstract Objective: To assess safety and outcomes (metabolic and liver) of bariatric surgery in patients with cirrhosis with or without portal hypertension. Patients and Methods: This study is a retrospective review of 14 patients with Child’s A cirrhosis with or without portal hypertension who were prospectively enrolled from February 23, 2009, through November 9, 2011, with 6- to 24-month follow-up after bariatric surgery (11 patients underwent sleeve gastrectomy [78.6%] and 3 gastric bypass [21.4%]). Four patients had portal hypertension detected by esophagogastroduodenoscopy. Results: The mean patient age was 55.5 years, and 10 of 14 patients were women. The mean weight decreased from 12518 to 9417 at 1 year (P<.001) and 9317 kg at 2 years (P<.001) postsurgery. The prevalence of diabetes decreased from 10 of 14 patients to 4 of 12 (P¼.01) and 1 of 6 (P¼.02) at 1 and 2 years postsurgery. The frequency of dyslipidemia and hypertension decreased but was not statistically significant; however, the number of medications required to control them decreased. Hepatic steatosis was detected by perioperative liver biopsy in 13 of 14 patients (5%-30% steatosis in 6 patients, 31%-60% in 6, and >60% in 1). At 1 year postsurgery, only 1 of 8 patients who underwent follow-up ultrasound imaging showed evidence of steatosis. The bilirubin level was above 2 mg/dL in 1 patient at 1 year postsurgery. One patient had encephalopathy at 2 years postsurgery. None of the patients developed peri- or postoperative bleeding or surgical complications. Conclusion: Bariatric surgery in patients with compensated cirrhosis even with mild portal hypertension is well tolerated and safe with minimal risk of postoperative complications if performed in a large referral center. This population can experience the beneficial effects of weight loss and improved metabolic syndrome, as well as reduced hepatic steatosis. ª 2015 Mayo Foundation for Medical Education and Research
O
besity is common in patients with cirrhosis and is associated with an increased risk of hepatic decompensation. Weight loss is a goal in all cases, but it is extremely challenging to achieve. Bariatric surgery is an approved surgical intervention for the treatment of morbid obesity (body mass index [calculated as the weight in kilograms divided by the height in meters squared] 40 kg/m2) or moderate obesity (body mass index 35 kg/ m2) with considerable obesity-related comorbidities.1,2 This operation can provide substantial weight loss with improvement or complete resolution of obesity-related conditions, including diabetes, hypertension, hyperlipidemia, and obstructive sleep apnea.1,2 For patients with advanced fibrosis and cirrhosis, historically,
n
Mayo Clin Proc. 2015;90(2):209-215
bariatric surgery was not advised or offered. Complications of bariatric surgery, including bleeding, gastrointestinal symptoms, nutritional or electrolyte abnormalities, and stomal stenosis, can be seen in 10% to 17% of patients without cirrhosis.3 Patients with cirrhosis who undergo surgery of any kind are at increased risk of mortality from liver failure, renal failure, or even postoperative bleeding due to impaired coagulation. This risk depends on the degree of liver dysfunction or model for end-stage liver disease score.4 According to a National Inpatient Sample study,5 the mortality risk of bariatric surgery in patients without cirrhosis is 2 to 3 times that in patients with well-compensated cirrhosis and 32 times that in patients with decompensated cirrhosis. However, mortality associated with bariatric
Mayo Clin Proc. n February 2015;90(2):209-215 n http://dx.doi.org/10.1016/j.mayocp.2014.11.012 www.mayoclinicproceedings.org n ª 2015 Mayo Foundation for Medical Education and Research
From the Department of Internal Medicine (L.P.), Department of Surgery (J.S., M.L.K.), Division of Biomedical Statistics and Informatics (R.D.), and Division of Gastroenterology and Hepatology (P.S.K., K.D.W.), Mayo Clinic, Rochester, MN.
209
MAYO CLINIC PROCEEDINGS
surgery is considered to be generally low (perioperative mortality ranging from 0.08% to 0.22% and postoperative mortality ranging from 0.31% to 0.35%)1,3; thus, a more relevant issue may be the risk/benefit ratio. The benefits of bariatric surgery in this patient population are not well described in the literature, thus making risk/benefit assessment challenging. Most data on outcomes of bariatric surgery in patients with cirrhosis come from small numbers of patients who are incidentally found to have cirrhosis at the time of surgery,6,7 which limits outcome measures. A large National Inpatient Sample database study5 of patients with cirrhosis undergoing bariatric surgery was limited to survival and length of stay outcomes. It is known that nonalcoholic steatohepatitis (NASH), a common etiology of cirrhosis, is strongly associated with obesity and metabolic syndrome and may improve after bariatric surgery, benefiting the patient beyond weight loss and improved metabolic syndrome.8 Whether any benefit can be derived in a patient with NASH-related cirrhosis is unknown. The aim of this study was to assess the safety and benefits of bariatric surgery in patients with known cirrhosis with and without portal hypertension. This study evaluated surgical and medical outcomes, including effects on metabolic syndrome and liver function, over 6- to 24-month follow-up. PATIENTS AND METHODS This study is a retrospective review of 14 patients with known Child’s A cirrhosis with or without mild portal hypertension who were prospectively enrolled from February 23, 2009, through November 9, 2011, from hepatology or bariatic clinics and referred for bariatric surgery at Mayo Clinic, Rochester, Minnesota. The patients underwent either laparoscopic sleeve gastrectomy (the majority) or laparoscopic Roux-en-Y gastric bypass as determined by the surgeon involved (J.S. or M.L.K.). All patients with portal hypertension underwent laparoscopic sleeve gastrectomy. Preoperative evidence of metabolic syndrome was assessed, including the presence or absence of diabetes, hypertension, dyslipidemia, sleep apnea, and cardiovascular disease. Diabetes, hypertension, and hyperlipidemia were defined by clinical documentation of the endocrine team or on the basis of the 210
Mayo Clin Proc.
n
following standard definitions: the use of insulin or oral hypoglycemic agents, hemoglobin A1c level of greater than 6.5%, or fasting blood glucose level of greater than 126 mg/dL for diabetes. Hypertension was defined as a sustained blood pressure of greater than 140/90 or greater than 130/80 mm Hg for patients with diabetes or the use of antihypertensive medications. Dyslipidemia was defined as the use of lipidlowering therapy or on the basis of laboratory findings of low-density lipoprotein level of 130 mg/dL or more (3.3 mmol/L), triglyceride level of 150 mg/mL or more (1.7 mmol/L), or highdensity lipoprotein level of less than 40 mg/dL (1.0 mmol/L) in men and less than 50 mg/dL (1.3 mmol/L) in women. Improvement in these parameters was defined as discontinuation of medications required to maintain values below the above-mentioned values (while maintaining the required values without using medications). Values for liver enzymes and function tests before bariatric surgery were obtained. The presence or absence of features suggestive of portal hypertension (gastric and esophageal varices, portal hypertensive gastropathy, ascites, and encephalopathy) was assessed before bariatric surgery. The results of liver biopsy at the time of surgery were recorded. Postoperative measures over 2 years included liver function and evidence of decompensation, in addition to surgical and metabolic outcomes. Follow-up included data from 6 months to 2 years after bariatric surgery. This study was approved by the Mayo Clinic Institutional Review Board. STATISTICAL ANALYSES Variables were expressed as either mean SD or prevalence (%), as appropriate. For continuous variables, repeated-measures linear regression models were fit using time (baseline, 6 months, 1 year, and 2 years) as a factor to determine whether mean values changed over time. For binary variables, logistic regression models using generalized estimating equations were fit using time (baseline, 6 months, 1 year, and 2 years) as a factor to determine whether the odds of the outcome changed over time. In both modeling situations, the correlation between all pairs of times was assumed to be equal. If the logistic regression model using generalized estimating equations could not be fit because of the small
February 2015;90(2):209-215
n
http://dx.doi.org/10.1016/j.mayocp.2014.11.012 www.mayoclinicproceedings.org
BARIATRIC SURGERY IN PATIENTS WITH CIRRHOSIS
sample size, then the Wald test of marginal homogeneity was used to compare pairs of time points. RESULTS Demographic characteristics of the patient population are summarized in Table 1. The mean patient age was 55.514.5 years, and 10 of 14 patients were women. Model for end-stage liver disease scores ranged from 6 to 9 for all patients before surgery. Eleven patients underwent laparoscopic sleeve gastrectomy, and 3 patients underwent laparoscopic Roux-en-Y gastric bypass. Eleven patients had NASH, 2 had hepatitis C
TABLE 1. Demographic Characteristics of the Patient Population (N¼14)a,b Age (y) Sex: female Procedure Roux-en-Y gastric bypass Sleeve gastrectomy Underlying disease NASH HCV and NASH Cryptogenic Portal hypertension detected by EGD Portal hypertensive gastropathy Small esophageal varices Presurgery Mean weight (kg) Diabetes Insulin use Dyslipidemia Hypertension Hypertension medication 1 drug 2 drugs 3 drugs Steatosisd None Mild (5%-30%) Moderate (30%-60%) Severe (>60%) Inflammatory graded None Mild (grade 1) Moderate (grade 2)
55.514.5 10 (71.4) 3 (21.4) 11 (78.6) 11 2 1 4c 4 1
(78.6) (14.3) (7.1) (28.6) (28.6) (7.1)
12518 10 (71.4) 9 (90.0) 5 (35.7) 9 (64.3) 2 (22.2) 4 (44.4) 3 (33.3) 1 6 6 1
(7.1) (42.9) (42.9) (7.1)
1 (7.1) 10 (71.4) 3 (33.3)
EGD ¼ esophagogastroduodenoscopy; HCV ¼ hepatitis C virus; NASH ¼ nonalcoholic steatohepatitis. b Values are presented as mean SD or as No. (percentage). c One patient with both portal gastropathy and small varices. d Detected by perioperative liver biopsy: grade 1 inflammation ¼ scattered lobular, minimal portal infiltrate; grade 2 inflammation ¼ mild to moderate lobular, portal and periportal infiltrate. a
Mayo Clin Proc. n February 2015;90(2):209-215 www.mayoclinicproceedings.org
n
and NASH, and 1 had cryptogenic disease. Four patients had portal hypertension detected by esophagogastroduodenoscopy, 3 had only portal hypertensive gastropathy, and 1 had portal gastropathy and small varices. None of the patients had ascites or encephalopathy. Most patients had mild to moderate steatosis and mild to moderate inflammatory infiltrate as detected by perioperative liver biopsy despite the presence of cirrhosis.
Metabolic Outcomes Metabolic outcomes are listed in Table 2. The mean weight decreased from 12518 kg presurgery to 10216 kg at 6 months postsurgery (P<.001 vs presurgery), 9417 kg at 1 year postsurgery (P<.001 vs presurgery), and 9317 kg at 2 years postsurgery (P<.001 vs presurgery). Before surgery, diabetes was present in 10 (71.4%) patients, and this value decreased to 6 (50%) (P¼.09 vs presurgery), 4 (33.3%) (P¼.01 vs presurgery), and 1 (16.7%) (P¼.02 vs presurgery) at 6 months, 1 year, and 2 years postsurgery, respectively. Nine patients (90%) with diabetes were using insulin before surgery. This value decreased to 3 (50%) at 6 months postsurgery (P¼.07) and 1 (25%) at 1 year postsurgery (P¼.03). At 2 years postsurgery, diabetes was present in only 1 of 6 patients, and this patient was using insulin (P¼.14). Dyslipidemia was present in 5 patients (35.7%) presurgery, and these patients were using statins; dyslipidemia remained in these 5 patients at 6 months postsurgery (P¼1.0 vs presurgery), and 3 patients (21%) continued to have dyslipidemia at 1 year postsurgery (P¼.65 vs presurgery), but none had dyslipidemia at 2 years postsurgery (P¼.32 vs presurgery). Although the prevalence of hypertension increased (Table 2) from 9 patients (64.3%) presurgery to 2 patients (33.3%) at 2 years postsurgery, it was not statistically significant (P¼.08); however, the number of medications required to control them decreased. Before surgery, 6 patients (50%) were taking 2 or more antihypertensive medications. Only 4 (33.3%) and 3 (25%) (P¼.39 and P¼.11, respectively) patients required 2 or more medications at 6 months and 1 year postsurgery, respectively; 4 (66.7%) patients required no blood
http://dx.doi.org/10.1016/j.mayocp.2014.11.012
211
MAYO CLINIC PROCEEDINGS
TABLE 2. Metabolic Outcomesa,b Postsurgery Variable
Presurgery (n¼14)
6 mo (n¼12)
1 y (n¼12)
2 y (n¼6)
Weight (kg) P value vs presurgery % total body weight change Diabetes P value vs presurgery Insulin use P value vs presurgery Dyslipidemia P value vs presurgery Hypertension P value vs presurgery Hypertension medication 0 drugs (no HTN) 1 drug 2 drugs P value vs presurgery
12518
10216 <.001 18.4 6 (50.0) .09 3 (50.0) .07 5 (41.7) 1.0 8 (66.7) .90
9417 <.001 24.8 4 (33.3) .01 1 (25.0) .03 3/11c (27.3) .65 6 (50.0) .15
9317 <.001 25.6 1 (16.7) .02 1 (100) .14 0 .32 2 (33.3) .08
4 (33.3) 4 (33.3) 4 (33.3) .39
6 (50.0) 3 (25.0) 3 (25.0) .11
4 (66.7) 2 (33.3) 0 .39
10 (71.4) 9 (90.0) 5 (35.7) 9 (64.3)
5 (35.7) 2 (14.3) 7 (50.0)
HTN ¼ hypertension. Values are presented as mean SD or as No. (percentage). c Only 11 of 12 patients had lipid studies available. a
b
pressure medication at 2 years, but statistical significance was not found (P¼.39). Liver-related Outcomes Hepatic steatosis was detected by perioperative liver biopsy in 92.9% of patients (5%-30% steatosis in 6 patients, 30%-60% in 6, and >60% in 1). At 1 year postsurgery, only 12.5% (1 of 8) of patients who underwent follow-up ultrasound imaging showed evidence of steatosis (4 patients did not undergo ultrasound imaging for various reasons). The bilirubin level was above 2 mg/dL in 1 patient at 1 year postsurgery (bilirubin level 2.3 mg/ dL from 1.0 mg/dL). The bilirubin level of this patient remained stable (2.2 mg/dL) at 2 years postsurgery. The mean alanine aminotransferase (ALT) level decreased from 5131 U/L (n¼13) presurgery to 3521 U/ L at 6 months postsurgery (n¼9; P¼.26 vs presurgery), 3613 U/L at 1 year postsurgery (n¼9; P¼.03 vs presurgery), and 2811 U/L at 2 years postsurgery (n¼5; P¼.02 vs presurgery). The Figure depicts the change in mean weight (kg) and ALT level over time after bariatric surgery. The ALT level was elevated (>45 U/L) in 3 patients postsurgery: in 1 patient, the ALT level was 88 U/L at 6 months postsurgery and increased to 32 U/L at 1 212
Mayo Clin Proc.
n
year postsurgery. The ALT level slightly increased in the second patient at 2 years postsurgery (40 U/L at 1 year postsurgery and 47 U/L at 2 years postsurgery), and the ALT level of the third patient was 68 U/L at 1 year postsurgery, with no value recorded at 2 years postsurgery. None of these 3 patients had elevated bilirubin levels postsurgery. In the entire sample, 1 patient had encephalopathy at 2 years postsurgery, with comorbid mania at that time. This patient had a preexisting transjugular intrahepatic portosystemic shunt and was using sedative medications; thus, her encephalopathy was not considered to be related to bariatric surgery performed 2 years ago. None of the 4 patients with mild endoscopic portal hypertension developed perioperative or postoperative bleeding or surgical complications. This cohort had no surgical complications. DISCUSSION This study helps to better define the risk/ benefit ratio of bariatric surgery in the population with cirrhosis. This weight loss surgery can be safely performed in patients with well-compensated cirrhosis with or without mild portal hypertension in a large bariatric referral center with careful postoperative
February 2015;90(2):209-215
n
http://dx.doi.org/10.1016/j.mayocp.2014.11.012 www.mayoclinicproceedings.org
BARIATRIC SURGERY IN PATIENTS WITH CIRRHOSIS
Mayo Clin Proc. n February 2015;90(2):209-215 www.mayoclinicproceedings.org
n
150 Weight (kg)
ALT (U/L)
125
100 Mean ± SD
follow-up. Importantly, this study also found that bariatric surgery (predominantly sleeve gastrectomy) can result in significant weight loss and improvement in metabolic syndrome in patients with cirrhosis. The long-term benefits of these improvements are not defined in this 2-year follow-up study, but extrapolation from the medical literature would suggest long-term benefits of metabolic syndrome control with possible benefits in cardiovascular disease risk reduction. Liver function in patients with cirrhosis appears to remain stable up to 2 years postsurgery, with a potential improvement in hepatic steatosis, which suggests that possible stabilization or even delayed progression of decompensated cirrhosis could be anticipated. Clearly, further studies are warranted to further assess this. A large National Inpatient Sample database study5 of bariatric surgical outcomes in patients with and without cirrhosis found that patients with compensated and decompensated cirrhosis had higher mortality rates (0.9% [3 times] and 16.3% [32 times], respectively) than did patients without cirrhosis (0.3%). Unfortunately, this database could not provide the cause of death or specific liver-related or metabolic outcomes. It was suggested that centers performing large numbers of bariatric surgery procedures had a lower risk of death in the population with cirrhosis. Our study followed patients up to 2 years postsurgery and observed 100% survival, with only 1 patient showing evidence of hepatic decompensation 2 years after bariatric surgery. This patient had a preexisting transjugular intrahepatic portosystemic shunt, was using sedative medications, and had received a diagnosis of mania confounding her diagnosis of hepatic encephalopathy; thus, her encephalopathy was not considered to be related to bariatric surgery performed 2 years ago. Notably, 4 of our patients had mild portal hypertension detected by esophagogastroduodenoscopy before surgery. None of these patients developed postoperative bleeding complications despite portal hypertensive gastropathy changes in the operative site. This is the first study that purposely assessed bariatric surgery in patients with known portal hypertensive changes. Although more research remains to be done on the safety of bariatric surgery in patients with
75
50
25
0 Baseline
6 months
1 year
2 years
Time
FIGURE. Change in patient’s weight (kg) and ALT level (U/mL) over time after bariatric surgery. ALT ¼ alanine aminotransferase.
advanced liver disease, the limited findings suggest that laparoscopic sleeve gastrectomy may at least be well tolerated in patients with well-compensated cirrhosis with mild portal hypertensive changes. Although studies have consistently found a reduction in metabolic syndrome after bariatric surgery, little data exist on such metabolic outcomes in patients with advanced liver disease. A recent study by Shimizu et al9 considered 23 patients with cirrhosis (12 known and 11 unknown at the time of surgery) who underwent multiple bariatric surgery procedures. They noted improved glucose control in 87% of patients and improvement in hypertension and dyslipidemia in 69% and 67% of patients, respectively. Our study supports these findings, with a decrease in dyslipidemia from 35.7% presurgery to 0% at 2 years postsurgery (P¼.32) and a decrease in diabetes from 71.4% presurgery to 16.7% at 2 years postsurgery (P¼.02). Although there was no statistically significant reduction in the incidence of hypertension in this small sample, there may be clinical relevance because fewer patients with hypertension and fewer drugs to control it (if present) were noted. In our study, as in the study of Shimuzi et al,9 patients with cirrhosis achieve sustained weight loss and
http://dx.doi.org/10.1016/j.mayocp.2014.11.012
213
MAYO CLINIC PROCEEDINGS
had improvement in obesity-related comorbidities after bariatric surgery. Liver function appears to remain stable in our patient cohort after bariatric surgery. We found laboratory evidence of potentially worse liver function (defined as a bilirubin level of >2 mg/dL) in only 1 patient postsurgery (bilirubin level of 2.3 mg/dL at 1 year postsurgery and the level remained stable at 2 years postsurgery). In addition, the mean ALT level decreased to 3521 U/L (within the normal range, 7-45 U/L) at 6 months postsurgery (P¼.26) and decreased further to 2811 U/L at 2 years postsurgery (P¼.02). This improvement in liver transaminases likely reflects improved hepatic steatosis because improvement in NASH after significant weight loss with bariatric surgery has been suggested in patients without cirrhosis.8,10 Improvement in liver steatosis and/or inflammation may benefit these patients by delayed progression of liver disease, but further studies are needed to confirm this conclusion. The type of bariatric surgery procedure performed warrants comment. In patients with cirrhosis, we only selectively perform the Roux-en-Y gastric bypass procedure. In this procedure, more than 95% of the stomach is bypassed. Should the patient have variceal bleeding in the future, the esophagus and the gastric pouch can be visualized by endoscopy, but not the bypassed stomach. In rare cases, bleeding gastric varices may not be amenable to endoscopic treatment. Moreover, endoscopic retrograde cholangiopancreatography is challenging to perform in this group of patients because of the altered anatomy. In contrast, in the gastric sleeve procedure, more than 80% of the stomach is excised, but it does not result in discontinuity or bypass of any portion of the gastrointestinal tract. The entire esophagus and remnant stomach can be visualized by endoscopy should there be any bleeding in the gastrointestinal tract, and ability to perform endoscopic retrograde cholangiopancreatography is not affected. We thus prefer the gastric sleeve procedure as the bariatric surgery procedure in patients with cirrhosis. Our study had several limitations. First, the small sample size of patients with cirrhosis undergoing bariatric surgery limits the statistical power. Second, a 2-year follow-up period limits the extrapolation of long-term benefits 214
Mayo Clin Proc.
n
in this population with respect to delayed disease progression. The lack of postoperative histological data limits the interpretation of overall liver-related outcomes. The purpose of this study, however, was to assess peri- and postoperative safety, which has been established in this small sample. It should be noted that these procedures were performed (and results obtained) in a tertiary referral center with both a large bariatric surgical and a liver transplant program and potentially cannot be extrapolated to other settings. CONCLUSION Bariatric surgery in patients with compensated cirrhosis with or without mild portal hypertension is well tolerated and safe with minimal risk of postoperative complications if performed in a large referral center. This population can experience the beneficial effects of weight loss and improved metabolic syndrome, as well as reduced hepatic steatosis. Careful follow-up of these patients is required because any surgical procedure in individuals with cirrhosis is associated with an increased risk of death or decompensation. The longterm benefits of bariatric surgery in this population remain to be determined. Larger studies are needed with long-term follow-up to determine the true risk/benefit ratio. Abbreviations and Acronyms: ALT = alanine aminotransferase; NASH = nonalcoholic steatohepatitis Correspondence: Address to Kymberly D. Watt, MD, Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (watt.kymberly@mayo. edu).
REFERENCES 1. Adams TD, Gress RE, Smith SC, et al. Long-term mortality after gastric bypass surgery. N Engl J Med. 2007;357(8): 753-761. 2. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis [published correction appears in JAMA 2005;293(14):1728]. JAMA. 2004;292(14): 1724-1737. 3. Chang SH, Stoll CR, Song J, Varela JE, Eagon CJ, Colditz GA. The effectiveness and risks of bariatric surgery: an updated systematic review and meta-analysis, 2003-2012. JAMA Surg. 2014; 149(3):275-287. 4. Teh SH, Nagorney DM, Stevens SR, et al. Risk factors for mortality after surgery in patients with cirrhosis. Gastroenterology. 2007;132(4):1261-1269. 5. Mosko JD, Nguyen GC. Increased perioperative mortality following bariatric surgery among patients with cirrhosis. Clin Gastroenterol Hepatol. 2011;9(10):897-901.
February 2015;90(2):209-215
n
http://dx.doi.org/10.1016/j.mayocp.2014.11.012 www.mayoclinicproceedings.org
BARIATRIC SURGERY IN PATIENTS WITH CIRRHOSIS
6. Dallal RM, Mattar SG, Lord JL, et al. Results of laparoscopic gastric bypass in patients with cirrhosis. Obes Surg. 2004; 14(1):47-53. 7. Kral JG, Thung SN, Biron S, et al. Effects of surgical treatment of the metabolic syndrome on liver fibrosis and cirrhosis. Surgery. 2004;135(1):48-58. 8. Mathurin P, Hollebecque A, Arnalsteen L, et al. Prospective study of the long-term effects of bariatric surgery on liver injury
Mayo Clin Proc. n February 2015;90(2):209-215 www.mayoclinicproceedings.org
n
in patients without advanced disease. Gastroenterology. 2009; 137(2):532-540. 9. Shimizu H, Phuong V, Maia M, et al. Bariatric surgery in patients with liver cirrhosis. Surg Obes Relat Dis. 2013;9(1):1-6. 10. Mummadi RR, Kasturi KS, Chennareddygari S, Sood GK. Effect of bariatric surgery on nonalcoholic fatty liver disease: systematic review and meta-analysis. Clin Gastroenterol Hepatol. 2008;6(12):1396-1402.
http://dx.doi.org/10.1016/j.mayocp.2014.11.012
215
Bariatric Surgery in Patients With Cirrhosis With and Without Portal Hypertension: A Single-Center Experience Laura Pestana, DO; James Swain, MD; Ross Dierkhising, MS; Michael L. Kendrick, MD; Patrick S. Kamath, MD; and Kymberly D. Watt, MD Abstract Objective: To assess safety and outcomes (metabolic and liver) of bariatric surgery in patients with cirrhosis with or without portal hypertension. Patients and Methods: This study is a retrospective review of 14 patients with Child’s A cirrhosis with or without portal hypertension who were prospectively enrolled from February 23, 2009, through November 9, 2011, with 6- to 24-month follow-up after bariatric surgery (11 patients underwent sleeve gastrectomy [78.6%] and 3 gastric bypass [21.4%]). Four patients had portal hypertension detected by esophagogastroduodenoscopy. Results: The mean patient age was 55.5 years, and 10 of 14 patients were women. The mean weight decreased from 12518 to 9417 at 1 year (P<.001) and 9317 kg at 2 years (P<.001) postsurgery. The prevalence of diabetes decreased from 10 of 14 patients to 4 of 12 (P¼.01) and 1 of 6 (P¼.02) at 1 and 2 years postsurgery. The frequency of dyslipidemia and hypertension decreased but was not statistically significant; however, the number of medications required to control them decreased. Hepatic steatosis was detected by perioperative liver biopsy in 13 of 14 patients (5%-30% steatosis in 6 patients, 31%-60% in 6, and >60% in 1). At 1 year postsurgery, only 1 of 8 patients who underwent follow-up ultrasound imaging showed evidence of steatosis. The bilirubin level was above 2 mg/dL in 1 patient at 1 year postsurgery. One patient had encephalopathy at 2 years postsurgery. None of the patients developed peri- or postoperative bleeding or surgical complications. Conclusion: Bariatric surgery in patients with compensated cirrhosis even with mild portal hypertension is well tolerated and safe with minimal risk of postoperative complications if performed in a large referral center. This population can experience the beneficial effects of weight loss and improved metabolic syndrome, as well as reduced hepatic steatosis. ª 2015 Mayo Foundation for Medical Education and Research
O
besity is common in patients with cirrhosis and is associated with an increased risk of hepatic decompensation. Weight loss is a goal in all cases, but it is extremely challenging to achieve. Bariatric surgery is an approved surgical intervention for the treatment of morbid obesity (body mass index [calculated as the weight in kilograms divided by the height in meters squared] 40 kg/m2) or moderate obesity (body mass index 35 kg/ m2) with considerable obesity-related comorbidities.1,2 This operation can provide substantial weight loss with improvement or complete resolution of obesity-related conditions, including diabetes, hypertension, hyperlipidemia, and obstructive sleep apnea.1,2 For patients with advanced fibrosis and cirrhosis, historically,
n
Mayo Clin Proc. 2015;90(2):209-215
bariatric surgery was not advised or offered. Complications of bariatric surgery, including bleeding, gastrointestinal symptoms, nutritional or electrolyte abnormalities, and stomal stenosis, can be seen in 10% to 17% of patients without cirrhosis.3 Patients with cirrhosis who undergo surgery of any kind are at increased risk of mortality from liver failure, renal failure, or even postoperative bleeding due to impaired coagulation. This risk depends on the degree of liver dysfunction or model for end-stage liver disease score.4 According to a National Inpatient Sample study,5 the mortality risk of bariatric surgery in patients without cirrhosis is 2 to 3 times that in patients with well-compensated cirrhosis and 32 times that in patients with decompensated cirrhosis. However, mortality associated with bariatric
Mayo Clin Proc. n February 2015;90(2):209-215 n http://dx.doi.org/10.1016/j.mayocp.2014.11.012 www.mayoclinicproceedings.org n ª 2015 Mayo Foundation for Medical Education and Research
From the Department of Internal Medicine (L.P.), Department of Surgery (J.S., M.L.K.), Division of Biomedical Statistics and Informatics (R.D.), and Division of Gastroenterology and Hepatology (P.S.K., K.D.W.), Mayo Clinic, Rochester, MN.
209
MAYO CLINIC PROCEEDINGS
surgery is considered to be generally low (perioperative mortality ranging from 0.08% to 0.22% and postoperative mortality ranging from 0.31% to 0.35%)1,3; thus, a more relevant issue may be the risk/benefit ratio. The benefits of bariatric surgery in this patient population are not well described in the literature, thus making risk/benefit assessment challenging. Most data on outcomes of bariatric surgery in patients with cirrhosis come from small numbers of patients who are incidentally found to have cirrhosis at the time of surgery,6,7 which limits outcome measures. A large National Inpatient Sample database study5 of patients with cirrhosis undergoing bariatric surgery was limited to survival and length of stay outcomes. It is known that nonalcoholic steatohepatitis (NASH), a common etiology of cirrhosis, is strongly associated with obesity and metabolic syndrome and may improve after bariatric surgery, benefiting the patient beyond weight loss and improved metabolic syndrome.8 Whether any benefit can be derived in a patient with NASH-related cirrhosis is unknown. The aim of this study was to assess the safety and benefits of bariatric surgery in patients with known cirrhosis with and without portal hypertension. This study evaluated surgical and medical outcomes, including effects on metabolic syndrome and liver function, over 6- to 24-month follow-up. PATIENTS AND METHODS This study is a retrospective review of 14 patients with known Child’s A cirrhosis with or without mild portal hypertension who were prospectively enrolled from February 23, 2009, through November 9, 2011, from hepatology or bariatic clinics and referred for bariatric surgery at Mayo Clinic, Rochester, Minnesota. The patients underwent either laparoscopic sleeve gastrectomy (the majority) or laparoscopic Roux-en-Y gastric bypass as determined by the surgeon involved (J.S. or M.L.K.). All patients with portal hypertension underwent laparoscopic sleeve gastrectomy. Preoperative evidence of metabolic syndrome was assessed, including the presence or absence of diabetes, hypertension, dyslipidemia, sleep apnea, and cardiovascular disease. Diabetes, hypertension, and hyperlipidemia were defined by clinical documentation of the endocrine team or on the basis of the 210
Mayo Clin Proc.
n
following standard definitions: the use of insulin or oral hypoglycemic agents, hemoglobin A1c level of greater than 6.5%, or fasting blood glucose level of greater than 126 mg/dL for diabetes. Hypertension was defined as a sustained blood pressure of greater than 140/90 or greater than 130/80 mm Hg for patients with diabetes or the use of antihypertensive medications. Dyslipidemia was defined as the use of lipidlowering therapy or on the basis of laboratory findings of low-density lipoprotein level of 130 mg/dL or more (3.3 mmol/L), triglyceride level of 150 mg/mL or more (1.7 mmol/L), or highdensity lipoprotein level of less than 40 mg/dL (1.0 mmol/L) in men and less than 50 mg/dL (1.3 mmol/L) in women. Improvement in these parameters was defined as discontinuation of medications required to maintain values below the above-mentioned values (while maintaining the required values without using medications). Values for liver enzymes and function tests before bariatric surgery were obtained. The presence or absence of features suggestive of portal hypertension (gastric and esophageal varices, portal hypertensive gastropathy, ascites, and encephalopathy) was assessed before bariatric surgery. The results of liver biopsy at the time of surgery were recorded. Postoperative measures over 2 years included liver function and evidence of decompensation, in addition to surgical and metabolic outcomes. Follow-up included data from 6 months to 2 years after bariatric surgery. This study was approved by the Mayo Clinic Institutional Review Board. STATISTICAL ANALYSES Variables were expressed as either mean SD or prevalence (%), as appropriate. For continuous variables, repeated-measures linear regression models were fit using time (baseline, 6 months, 1 year, and 2 years) as a factor to determine whether mean values changed over time. For binary variables, logistic regression models using generalized estimating equations were fit using time (baseline, 6 months, 1 year, and 2 years) as a factor to determine whether the odds of the outcome changed over time. In both modeling situations, the correlation between all pairs of times was assumed to be equal. If the logistic regression model using generalized estimating equations could not be fit because of the small
February 2015;90(2):209-215
n
http://dx.doi.org/10.1016/j.mayocp.2014.11.012 www.mayoclinicproceedings.org
BARIATRIC SURGERY IN PATIENTS WITH CIRRHOSIS
sample size, then the Wald test of marginal homogeneity was used to compare pairs of time points. RESULTS Demographic characteristics of the patient population are summarized in Table 1. The mean patient age was 55.514.5 years, and 10 of 14 patients were women. Model for end-stage liver disease scores ranged from 6 to 9 for all patients before surgery. Eleven patients underwent laparoscopic sleeve gastrectomy, and 3 patients underwent laparoscopic Roux-en-Y gastric bypass. Eleven patients had NASH, 2 had hepatitis C
TABLE 1. Demographic Characteristics of the Patient Population (N¼14)a,b Age (y) Sex: female Procedure Roux-en-Y gastric bypass Sleeve gastrectomy Underlying disease NASH HCV and NASH Cryptogenic Portal hypertension detected by EGD Portal hypertensive gastropathy Small esophageal varices Presurgery Mean weight (kg) Diabetes Insulin use Dyslipidemia Hypertension Hypertension medication 1 drug 2 drugs 3 drugs Steatosisd None Mild (5%-30%) Moderate (30%-60%) Severe (>60%) Inflammatory graded None Mild (grade 1) Moderate (grade 2)
55.514.5 10 (71.4) 3 (21.4) 11 (78.6) 11 2 1 4c 4 1
(78.6) (14.3) (7.1) (28.6) (28.6) (7.1)
12518 10 (71.4) 9 (90.0) 5 (35.7) 9 (64.3) 2 (22.2) 4 (44.4) 3 (33.3) 1 6 6 1
(7.1) (42.9) (42.9) (7.1)
1 (7.1) 10 (71.4) 3 (33.3)
EGD ¼ esophagogastroduodenoscopy; HCV ¼ hepatitis C virus; NASH ¼ nonalcoholic steatohepatitis. b Values are presented as mean SD or as No. (percentage). c One patient with both portal gastropathy and small varices. d Detected by perioperative liver biopsy: grade 1 inflammation ¼ scattered lobular, minimal portal infiltrate; grade 2 inflammation ¼ mild to moderate lobular, portal and periportal infiltrate. a
Mayo Clin Proc. n February 2015;90(2):209-215 www.mayoclinicproceedings.org
n
and NASH, and 1 had cryptogenic disease. Four patients had portal hypertension detected by esophagogastroduodenoscopy, 3 had only portal hypertensive gastropathy, and 1 had portal gastropathy and small varices. None of the patients had ascites or encephalopathy. Most patients had mild to moderate steatosis and mild to moderate inflammatory infiltrate as detected by perioperative liver biopsy despite the presence of cirrhosis.
Metabolic Outcomes Metabolic outcomes are listed in Table 2. The mean weight decreased from 12518 kg presurgery to 10216 kg at 6 months postsurgery (P<.001 vs presurgery), 9417 kg at 1 year postsurgery (P<.001 vs presurgery), and 9317 kg at 2 years postsurgery (P<.001 vs presurgery). Before surgery, diabetes was present in 10 (71.4%) patients, and this value decreased to 6 (50%) (P¼.09 vs presurgery), 4 (33.3%) (P¼.01 vs presurgery), and 1 (16.7%) (P¼.02 vs presurgery) at 6 months, 1 year, and 2 years postsurgery, respectively. Nine patients (90%) with diabetes were using insulin before surgery. This value decreased to 3 (50%) at 6 months postsurgery (P¼.07) and 1 (25%) at 1 year postsurgery (P¼.03). At 2 years postsurgery, diabetes was present in only 1 of 6 patients, and this patient was using insulin (P¼.14). Dyslipidemia was present in 5 patients (35.7%) presurgery, and these patients were using statins; dyslipidemia remained in these 5 patients at 6 months postsurgery (P¼1.0 vs presurgery), and 3 patients (21%) continued to have dyslipidemia at 1 year postsurgery (P¼.65 vs presurgery), but none had dyslipidemia at 2 years postsurgery (P¼.32 vs presurgery). Although the prevalence of hypertension increased (Table 2) from 9 patients (64.3%) presurgery to 2 patients (33.3%) at 2 years postsurgery, it was not statistically significant (P¼.08); however, the number of medications required to control them decreased. Before surgery, 6 patients (50%) were taking 2 or more antihypertensive medications. Only 4 (33.3%) and 3 (25%) (P¼.39 and P¼.11, respectively) patients required 2 or more medications at 6 months and 1 year postsurgery, respectively; 4 (66.7%) patients required no blood
http://dx.doi.org/10.1016/j.mayocp.2014.11.012
211
MAYO CLINIC PROCEEDINGS
TABLE 2. Metabolic Outcomesa,b Postsurgery Variable
Presurgery (n¼14)
6 mo (n¼12)
1 y (n¼12)
2 y (n¼6)
Weight (kg) P value vs presurgery % total body weight change Diabetes P value vs presurgery Insulin use P value vs presurgery Dyslipidemia P value vs presurgery Hypertension P value vs presurgery Hypertension medication 0 drugs (no HTN) 1 drug 2 drugs P value vs presurgery
12518
10216 <.001 18.4 6 (50.0) .09 3 (50.0) .07 5 (41.7) 1.0 8 (66.7) .90
9417 <.001 24.8 4 (33.3) .01 1 (25.0) .03 3/11c (27.3) .65 6 (50.0) .15
9317 <.001 25.6 1 (16.7) .02 1 (100) .14 0 .32 2 (33.3) .08
4 (33.3) 4 (33.3) 4 (33.3) .39
6 (50.0) 3 (25.0) 3 (25.0) .11
4 (66.7) 2 (33.3) 0 .39
10 (71.4) 9 (90.0) 5 (35.7) 9 (64.3)
5 (35.7) 2 (14.3) 7 (50.0)
HTN ¼ hypertension. Values are presented as mean SD or as No. (percentage). c Only 11 of 12 patients had lipid studies available. a
b
pressure medication at 2 years, but statistical significance was not found (P¼.39). Liver-related Outcomes Hepatic steatosis was detected by perioperative liver biopsy in 92.9% of patients (5%-30% steatosis in 6 patients, 30%-60% in 6, and >60% in 1). At 1 year postsurgery, only 12.5% (1 of 8) of patients who underwent follow-up ultrasound imaging showed evidence of steatosis (4 patients did not undergo ultrasound imaging for various reasons). The bilirubin level was above 2 mg/dL in 1 patient at 1 year postsurgery (bilirubin level 2.3 mg/ dL from 1.0 mg/dL). The bilirubin level of this patient remained stable (2.2 mg/dL) at 2 years postsurgery. The mean alanine aminotransferase (ALT) level decreased from 5131 U/L (n¼13) presurgery to 3521 U/ L at 6 months postsurgery (n¼9; P¼.26 vs presurgery), 3613 U/L at 1 year postsurgery (n¼9; P¼.03 vs presurgery), and 2811 U/L at 2 years postsurgery (n¼5; P¼.02 vs presurgery). The Figure depicts the change in mean weight (kg) and ALT level over time after bariatric surgery. The ALT level was elevated (>45 U/L) in 3 patients postsurgery: in 1 patient, the ALT level was 88 U/L at 6 months postsurgery and increased to 32 U/L at 1 212
Mayo Clin Proc.
n
year postsurgery. The ALT level slightly increased in the second patient at 2 years postsurgery (40 U/L at 1 year postsurgery and 47 U/L at 2 years postsurgery), and the ALT level of the third patient was 68 U/L at 1 year postsurgery, with no value recorded at 2 years postsurgery. None of these 3 patients had elevated bilirubin levels postsurgery. In the entire sample, 1 patient had encephalopathy at 2 years postsurgery, with comorbid mania at that time. This patient had a preexisting transjugular intrahepatic portosystemic shunt and was using sedative medications; thus, her encephalopathy was not considered to be related to bariatric surgery performed 2 years ago. None of the 4 patients with mild endoscopic portal hypertension developed perioperative or postoperative bleeding or surgical complications. This cohort had no surgical complications. DISCUSSION This study helps to better define the risk/ benefit ratio of bariatric surgery in the population with cirrhosis. This weight loss surgery can be safely performed in patients with well-compensated cirrhosis with or without mild portal hypertension in a large bariatric referral center with careful postoperative
February 2015;90(2):209-215
n
http://dx.doi.org/10.1016/j.mayocp.2014.11.012 www.mayoclinicproceedings.org
BARIATRIC SURGERY IN PATIENTS WITH CIRRHOSIS
Mayo Clin Proc. n February 2015;90(2):209-215 www.mayoclinicproceedings.org
n
150 Weight (kg)
ALT (U/L)
125
100 Mean ± SD
follow-up. Importantly, this study also found that bariatric surgery (predominantly sleeve gastrectomy) can result in significant weight loss and improvement in metabolic syndrome in patients with cirrhosis. The long-term benefits of these improvements are not defined in this 2-year follow-up study, but extrapolation from the medical literature would suggest long-term benefits of metabolic syndrome control with possible benefits in cardiovascular disease risk reduction. Liver function in patients with cirrhosis appears to remain stable up to 2 years postsurgery, with a potential improvement in hepatic steatosis, which suggests that possible stabilization or even delayed progression of decompensated cirrhosis could be anticipated. Clearly, further studies are warranted to further assess this. A large National Inpatient Sample database study5 of bariatric surgical outcomes in patients with and without cirrhosis found that patients with compensated and decompensated cirrhosis had higher mortality rates (0.9% [3 times] and 16.3% [32 times], respectively) than did patients without cirrhosis (0.3%). Unfortunately, this database could not provide the cause of death or specific liver-related or metabolic outcomes. It was suggested that centers performing large numbers of bariatric surgery procedures had a lower risk of death in the population with cirrhosis. Our study followed patients up to 2 years postsurgery and observed 100% survival, with only 1 patient showing evidence of hepatic decompensation 2 years after bariatric surgery. This patient had a preexisting transjugular intrahepatic portosystemic shunt, was using sedative medications, and had received a diagnosis of mania confounding her diagnosis of hepatic encephalopathy; thus, her encephalopathy was not considered to be related to bariatric surgery performed 2 years ago. Notably, 4 of our patients had mild portal hypertension detected by esophagogastroduodenoscopy before surgery. None of these patients developed postoperative bleeding complications despite portal hypertensive gastropathy changes in the operative site. This is the first study that purposely assessed bariatric surgery in patients with known portal hypertensive changes. Although more research remains to be done on the safety of bariatric surgery in patients with
75
50
25
0 Baseline
6 months
1 year
2 years
Time
FIGURE. Change in patient’s weight (kg) and ALT level (U/mL) over time after bariatric surgery. ALT ¼ alanine aminotransferase.
advanced liver disease, the limited findings suggest that laparoscopic sleeve gastrectomy may at least be well tolerated in patients with well-compensated cirrhosis with mild portal hypertensive changes. Although studies have consistently found a reduction in metabolic syndrome after bariatric surgery, little data exist on such metabolic outcomes in patients with advanced liver disease. A recent study by Shimizu et al9 considered 23 patients with cirrhosis (12 known and 11 unknown at the time of surgery) who underwent multiple bariatric surgery procedures. They noted improved glucose control in 87% of patients and improvement in hypertension and dyslipidemia in 69% and 67% of patients, respectively. Our study supports these findings, with a decrease in dyslipidemia from 35.7% presurgery to 0% at 2 years postsurgery (P¼.32) and a decrease in diabetes from 71.4% presurgery to 16.7% at 2 years postsurgery (P¼.02). Although there was no statistically significant reduction in the incidence of hypertension in this small sample, there may be clinical relevance because fewer patients with hypertension and fewer drugs to control it (if present) were noted. In our study, as in the study of Shimuzi et al,9 patients with cirrhosis achieve sustained weight loss and
http://dx.doi.org/10.1016/j.mayocp.2014.11.012
213
MAYO CLINIC PROCEEDINGS
had improvement in obesity-related comorbidities after bariatric surgery. Liver function appears to remain stable in our patient cohort after bariatric surgery. We found laboratory evidence of potentially worse liver function (defined as a bilirubin level of >2 mg/dL) in only 1 patient postsurgery (bilirubin level of 2.3 mg/dL at 1 year postsurgery and the level remained stable at 2 years postsurgery). In addition, the mean ALT level decreased to 3521 U/L (within the normal range, 7-45 U/L) at 6 months postsurgery (P¼.26) and decreased further to 2811 U/L at 2 years postsurgery (P¼.02). This improvement in liver transaminases likely reflects improved hepatic steatosis because improvement in NASH after significant weight loss with bariatric surgery has been suggested in patients without cirrhosis.8,10 Improvement in liver steatosis and/or inflammation may benefit these patients by delayed progression of liver disease, but further studies are needed to confirm this conclusion. The type of bariatric surgery procedure performed warrants comment. In patients with cirrhosis, we only selectively perform the Roux-en-Y gastric bypass procedure. In this procedure, more than 95% of the stomach is bypassed. Should the patient have variceal bleeding in the future, the esophagus and the gastric pouch can be visualized by endoscopy, but not the bypassed stomach. In rare cases, bleeding gastric varices may not be amenable to endoscopic treatment. Moreover, endoscopic retrograde cholangiopancreatography is challenging to perform in this group of patients because of the altered anatomy. In contrast, in the gastric sleeve procedure, more than 80% of the stomach is excised, but it does not result in discontinuity or bypass of any portion of the gastrointestinal tract. The entire esophagus and remnant stomach can be visualized by endoscopy should there be any bleeding in the gastrointestinal tract, and ability to perform endoscopic retrograde cholangiopancreatography is not affected. We thus prefer the gastric sleeve procedure as the bariatric surgery procedure in patients with cirrhosis. Our study had several limitations. First, the small sample size of patients with cirrhosis undergoing bariatric surgery limits the statistical power. Second, a 2-year follow-up period limits the extrapolation of long-term benefits 214
Mayo Clin Proc.
n
in this population with respect to delayed disease progression. The lack of postoperative histological data limits the interpretation of overall liver-related outcomes. The purpose of this study, however, was to assess peri- and postoperative safety, which has been established in this small sample. It should be noted that these procedures were performed (and results obtained) in a tertiary referral center with both a large bariatric surgical and a liver transplant program and potentially cannot be extrapolated to other settings. CONCLUSION Bariatric surgery in patients with compensated cirrhosis with or without mild portal hypertension is well tolerated and safe with minimal risk of postoperative complications if performed in a large referral center. This population can experience the beneficial effects of weight loss and improved metabolic syndrome, as well as reduced hepatic steatosis. Careful follow-up of these patients is required because any surgical procedure in individuals with cirrhosis is associated with an increased risk of death or decompensation. The longterm benefits of bariatric surgery in this population remain to be determined. Larger studies are needed with long-term follow-up to determine the true risk/benefit ratio. Abbreviations and Acronyms: ALT = alanine aminotransferase; NASH = nonalcoholic steatohepatitis Correspondence: Address to Kymberly D. Watt, MD, Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (watt.kymberly@mayo. edu).
REFERENCES 1. Adams TD, Gress RE, Smith SC, et al. Long-term mortality after gastric bypass surgery. N Engl J Med. 2007;357(8): 753-761. 2. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis [published correction appears in JAMA 2005;293(14):1728]. JAMA. 2004;292(14): 1724-1737. 3. Chang SH, Stoll CR, Song J, Varela JE, Eagon CJ, Colditz GA. The effectiveness and risks of bariatric surgery: an updated systematic review and meta-analysis, 2003-2012. JAMA Surg. 2014; 149(3):275-287. 4. Teh SH, Nagorney DM, Stevens SR, et al. Risk factors for mortality after surgery in patients with cirrhosis. Gastroenterology. 2007;132(4):1261-1269. 5. Mosko JD, Nguyen GC. Increased perioperative mortality following bariatric surgery among patients with cirrhosis. Clin Gastroenterol Hepatol. 2011;9(10):897-901.
February 2015;90(2):209-215
n
http://dx.doi.org/10.1016/j.mayocp.2014.11.012 www.mayoclinicproceedings.org
BARIATRIC SURGERY IN PATIENTS WITH CIRRHOSIS
6. Dallal RM, Mattar SG, Lord JL, et al. Results of laparoscopic gastric bypass in patients with cirrhosis. Obes Surg. 2004; 14(1):47-53. 7. Kral JG, Thung SN, Biron S, et al. Effects of surgical treatment of the metabolic syndrome on liver fibrosis and cirrhosis. Surgery. 2004;135(1):48-58. 8. Mathurin P, Hollebecque A, Arnalsteen L, et al. Prospective study of the long-term effects of bariatric surgery on liver injury
Mayo Clin Proc. n February 2015;90(2):209-215 www.mayoclinicproceedings.org
n
in patients without advanced disease. Gastroenterology. 2009; 137(2):532-540. 9. Shimizu H, Phuong V, Maia M, et al. Bariatric surgery in patients with liver cirrhosis. Surg Obes Relat Dis. 2013;9(1):1-6. 10. Mummadi RR, Kasturi KS, Chennareddygari S, Sood GK. Effect of bariatric surgery on nonalcoholic fatty liver disease: systematic review and meta-analysis. Clin Gastroenterol Hepatol. 2008;6(12):1396-1402.
http://dx.doi.org/10.1016/j.mayocp.2014.11.012
215