Obesity is perhaps the most significant public health problem

Transcription

1 Obesity and Its Effect on Survival in Patients Undergoing Orthotopic Liver Transplantation in the United States Satheesh Nair, 1 Sumita Verma, 2 and Paul J. Thuluvath 2 Studies assessing morbidity and mortality in obese patients undergoing orthotopic liver transplantation (OLT) have produced conflicting results, mainly because of the small sample size. The objective of our study was to determine graft and patient survival in obese adults receiving OLT in the U.S. between 1988 through 1996 using the United Network for Organ Sharing (UNOS) database. Among the 23,675 transplantations performed during the 9-year study period, 18,172 (75%) patients fulfilled the inclusion criteria. Of these, 8,382 (46%) were nonobese (body mass index [BMI] < 25 kg/m 2 ), 5,913 (33%) were overweight (BMI, kg/m 2 ), 2,611 (14%) were obese (BMI, kg/m 2 ), 911 (5%) were severely obese (BMI, kg/m 2 ), and 355 (2%) were morbidly obese (BMI, kg/m 2 ). The outcome measures assessed were immediate (30-day), 1-, 2-, and 5-year patient survival. Obese groups had a higher proportion of women, a greater prevalence of cryptogenic cirrhosis (P <.05) and diabetes (P <.05), and a higher serum creatinine. Primary graft nonfunction, and immediate, 1-year, and 2-year mortality were significantly higher in the morbidly obese group (P <.05). Five-year mortality was significantly higher both in the severely and morbidly obese subjects (P <.05), mostly as a result of adverse cardiovascular events. Kaplan-Meier survival was significantly lower in morbidly obese patients, and morbid obesity was an independent predictor of mortality. Obesity is associated with a significant increase in long-term mortality, mostly as a result of cardiovascular events. Weight loss should be recommended for all patients awaiting a liver transplantation, especially if their BMI is more than 35 kg/m 2. (HEPATOLOGY 2002;35: ) Obesity is perhaps the most significant public health problem facing the United States today. 1 The National Health and Nutrition Examination Surveys (NHANES) show that there has been a significant increase in the prevalence of obesity between 1976 through 1989 (NHANES II) and 1988 through 1994 (NHANES III). 2 In NHANES III, 22.5% of all adults aged between 20 to 74 years had a body mass index (BMI) of more than 30 kg/m 2. This increase was seen in men and women of all ages and all racial and ethnic groups. Perioperative morbidity and mortality are increased in obese patients after major surgical procedures because of concomitant problems such as coronary artery disease, hyperlipidemia, and restrictive pulmonary function. 1,3 There are only limited data on the effect of obesity on the outcome after transplantation. It has been shown that obese renal-transplant recipients have comparatively poorer graft and patient survival and a higher rate of wound infection compared with nonobese recipients. 5-7 In one study, actuarial Abbreviations: NHANES, National Health and Nutrition Examination Surveys; BMI, body mass index; OLT, orthotopic liver transplantation; UNOS, United Network for Organ Sharing; OR, odds ratio. From the 1 Division of Gastroenterology and Hepatology, Ochsner Clinic, New Orleans, LA; and 2 The Johns Hopkins University School of Medicine, Baltimore, MD. Received July 2, 2001; accepted September 25, Address reprint requests to: Paul J. Thuluvath, M.D., FRCP, Room 429, 1830 Bldg East Monument Street, The Johns Hopkins School of Medicine, Baltimore, MD pjthuluv@jhmi.edu; fax: Copyright 2002 by the American Association for the Study of Liver Diseases /02/ $35.00/0 doi: /jhep patient and graft survival were not different in obese and nonobese patients undergoing renal transplantation, though the former had an increased incidence of wound infection and weight gain during the first year following transplantation. 4 There are few studies that had analyzed the impact of obesity on orthotopic liver transplantation (OLT) In one study, Braunfeld et al. reported similar incidence of multiorgan dysfunction, intra- and postoperative pulmonary and cardiac complications, infection, and length of hospital stay in 40 morbidly obese (BMI 40kg/m 2 ) patients compared with controls. 9 However, another study found a higher immediate postoperative morbidity after OLT, but comparable long-term survival in severely obese patients (BMI 35 kg/m 2 ). 10 In a recent study involving 121 patients, of whom 57 were obese (i.e.,bmi 27.3 kg/m 2 in women and 27.8 kg/m 2 in men), we reported that intraoperative complications and 3-year survival were similar in all the groups studied. 11 However, postoperative complications (respiratory and vascular), length of hospital stays, and cost of OLT were significantly higher in the obese subjects. One of the major limitations of the previous studies, including ours, that had assessed the impact of obesity on morbidity and mortality after OLT was the small sample size of the obese patients. This made it almost impossible to show a difference in the longterm outcome, because there are many confounding factors, in addition to obesity, that affected survival after OLT. In addition, the definition of obesity and outcome measures were different in previous studies, contributing to the conflicting results obtained so far. The United Network for Organ Sharing (UNOS) database has many limitations, but it provides a large sample size from different 105

2 106 NAIR, VERMA, AND THULUVATH HEPATOLOGY, January 2002 Table 1. Demographic and Clinical Characteristics of Patients at Entry Nonobese (<25 kg/m 2 ) (N 8382) Overweight ( kg/m 2 ) (N 5913) Obese ( kg/m 2 ) (N 1611) Severely Obese ( kg/m 2 ) (N 911) Morbidly Obese (>40 kg/m 2 ) (N 355) Age (yr) Sex (female) 48% 37% 41% 49% 56%* BMI (kg/m 2 ) DM (%) 10% 15% 18%* 22%* 22%* Creatinine (mg/dl) * Etiology Hepatitis C 16%* 21% 25% 26% 25% Alcohol 15% 19% 17% 15% 13% Hepatitis B 5% 4% 4% 2% 2% PBC 14%* 7% 6% 4% 4% Cryptogenic 8% 10% 15% 18%* 18%* UNOS status Status 1 (%) 14% 13% 13% 16% 16% Status 11 (%) 21% 20% 18% 19% 21% Status 111 (%) 37% 43% 46% 46% 45% Donor age (yr) Cold ischemia time (h) NOTE. All values expressed as mean with standard deviation. All percentage figures are rounded off to the nearest whole number. Abbreviations: DM, diabetes mellitus; PBC, primary biliary cirrhosis. *P.05; patients with more than one etiology not listed. UNOS status based on previous system. parts of the country. Moreover, it is reliable for more hard facts such as graft and patient survival. Height and weight are also routinely recorded in the UNOS database for matching donors and recipients. The objective of our study was to determine the shortand long-term patient and graft outcome in obese patients undergoing OLT. Patients and Methods A detailed study of the UNOS database was performed between 1988 through We included only adults (older than 18 years) undergoing OLT during this period. The following information was collected: age (both of recipient and donor), sex, height, weight, prevalence of diabetes, serum creatinine, etiology of liver disease, UNOS status, and cold ischemia time. Diabetes was defined as fasting blood glucose more than 7 mmol/l. 12 BMI was used to define the degree of obesity, and BMI was calculated by dividing the weight in kilograms (at the time of listing) by the square of the height in meters (kg/m 2 ). 13 On the basis of BMI, patients were classified into the following 5 groups: nonobese ( 25 kg/m 2 ), overweight ( kg/m 2 ), obese ( kg/ m 2 ), severely obese ( kg/m 2 ), and morbidly obese ( kg/m 2 ). 14,15 We excluded patients whose height or weight was not recorded. We also excluded those patients who had a BMI 50 kg/m 2, because of concerns regarding erroneous recording of height or weight. The main outcome measures assessed were (1) primary graft nonfunction; (2) relisting for transplantation within 30 days; (3) graft survival at 1 and 2 years; and (4) immediate (30-day) and mortality at 1, 2, and 5 years. The cause of death, when available, was further classified into operative, cardiovascular (cardiac arrhythmia, heart failure, or myocardial infarction), infection, and cerebrovascular. Statistical Analysis. The differences between the various BMI groups were compared using the 2 test and ANOVA for categorical and continuous variables, respectively. Mortality statistics were calculated at 1, 2, and 5 years (using December 31, 1999, as the last follow-up date) as the percentage of patients who died within 1, 2, or 5 years after OLT. For the 1-, 2-, and 5-year mortality calculations, those patients who did not complete the term (either 1, 2, or 5 years) and those who did not have complete follow-up data were excluded. Kaplan-Meier analysis with the log rank test was used for comparing the survival between the groups. Cox regression analysis was used to determine whether obesity was an independent predictor of mortality. For all analyses, a 2-tailed P value of 0.05 or less was considered significant. The data were analyzed using the statistical software SPSS 10.0 (Chicago, IL). Results Of the 23,675 liver transplantations performed during the 9-year period from 1988 through 1996, there were 20,281 adult recipients; 1,948 patients had no recording of either height or weight. One hundred ninety-one patients had a BMI 50 kg/m 2 ; they were excluded. The remaining 18,172 (77%) fulfilled the inclusion criteria. Of these, 8,382 (46%) were nonobese, 5,913 (33%) were overweight, 2,611 (14%) were obese, 911 (5%) were severely obese, and 355 (2%) were morbidly obese. Overall, 3,811 patients (21%) had a BMI over 30 kg/m 2 ; this is similar to the prevalence of obesity that was reported in the general population in the NHANES III study. 2 The demographics and the baseline clinical characteristics of patients are shown in Table 1. The groups were similar with regard to age (both donor and recipient), cold ischemia time, and UNOS status. There was a higher proportion of women in the obese groups, but this was statistically significant only in the morbidly obese group. The etiology of the liver disease also varied between the different groups. There was a higher prevalence of primary biliary cirrhosis in nonobese patients. Cryptogenic cirrhosis was more common in severely and morbidly obese patients. Serum

3 HEPATOLOGY, Vol. 35, No. 1, 2002 NAIR, VERMA, AND THULUVATH 107 Table 2. Outcome Measures in Obese and Nonobese Patients Nonobese (N 8,312) Overweight (N 5,913) Obese (N 1,611) Severely Obese (N 911) Morbidly Obese (N 355) Primary graft nonfunction 6% 7% 7% 9% 10%* Relisted for transplantation in 30 days 5% 6% 5% 6% 7% Graft survival 1-year 75% 75% 76% 74% 72% 2-year 70% 70% 70% 68% 64% Mortality 30-day 6% 7% 8% 8% 12%* 1-year 16% 14% 14% 18% 22%* 2-year 25% 24% 25% 26% 33%* 5-year 44% 46% 47% 51%* 57%* Cause of death Cardiovascular 16% 20% 22%* 28%* 27%* Infection 39% 39% 42% 32% 44% CVA 9% 7% 6% 4% 6% Operative mortality 6% 7% 6% 10% 7% NOTE. All percentage figures are rounded off to the nearest whole number. Abbreviation: CVA, cerebrovascular accident. *P than.05. Complete information available for all patients. Calculated mortality may be higher than the actual mortality at 5 years, because patients with incomplete data were considered lost for follow-up, and there may be bias in reporting deaths. creatinine was higher in the morbidly obese group. As expected, the prevalence of diabetes was significantly higher in the obese patients. The main outcome measures are described in Table 2. Morbidly obese patients had a significantly higher prevalence of primary graft nonfunction. There were no significant differences in 1- and 2-year graft survival between the different groups. Thirty-day and 1- and 2-year mortality was significantly higher in the morbidly obese group (P.01). The 5-year mortality was significantly higher in the severely obese (51%) and morbidly obese (58%) groups compared with the nonobese group (P.02). Five-year survival data were reliably documented only in 10,038 patients. Those patients with incomplete data were considered lost for follow-up for 5-year survival and censored for Kaplan-Meier survival analysis. It is therefore probable that we may have overestimated 5-year mortality in all groups, but the proportion of missing data was similar in all 5 groups, which maintained validity for comparisons between groups. Intraoperative mortality and death directly related to infections were comparable between the obese and nonobese subjects. The major cause of the increased mortality in the obese groups was related to adverse cardiovascular events (Table 2). Kaplan-Meier survival curves for the different groups are shown in Fig. 1. The morbidly obese patients had a significantly lower survival compared with the other 4 groups (P.001 by log rank test). The survival curve for the morbidly obese started separating within the first year, as shown in Fig. 1. The Cox proportionalhazard model for survival at 2 years showed that morbid obesity was an independent predictor of mortality (odds ratio [OR], 1.52; 95% CI, ; P.02). Severe obesity was not an independent predictor of mortality. Recipient s age (OR, 1.01; 95% CI, ; P.001), serum creatinine (OR, 1.09; 95% CI, ; P.001), diabetes (OR, 1.32; 95% CI, ; P.001), UNOS status I (OR, 3.9; 95% CI, ; P.02), and UNOS status II (OR, 2.45; 95% CI, ; P.03) were other independent predictors of mortality. Discussion The increasing prevalence of obesity appears to be a global phenomenon and is a major cause of concern in the United States. Successive cross-sectional, nationally representative surveys indicate that the crude prevalence of overweight and obesity (BMI 25 kg/m 2 ) was 59.4% for men and 50.7% for women. 2 Our study is consistent with these findings and showed that 54% of patients undergoing OLT were either overweight or obese. However, in our study, we found a significantly higher proportion of females in the morbidly obese (BMI 40 kg/m 2 ) group. It is probable that we may have overestimated obesity, because many patients with endstage cirrhosis had fluid overload; however, our comparison be- Fig. 1. Kaplan-Meier survival in patients based on their BMI. (1) Nonobese, (2) overweight, (3) obese, (4) severely obese, (5) morbidly obese.

4 108 NAIR, VERMA, AND THULUVATH HEPATOLOGY, January 2002 tween the groups is still valid, because fluid overload would have affected all groups. The survival was similar in the nonobese, overweight, and obese group, but 5-year mortality was higher in the severely obese group. Thirty-day and 1-, 2-, and 5-year mortality was significantly higher in the morbidly obese group. As we have previously suggested, the actual mortality may be lower than we have reported, because we excluded those patients who had incomplete 5-year data, thereby reducing the actual patients at risk at 5 years. Kaplan-Meier survival showed a significantly lower survival in morbidly obese patients, and Cox regression suggested morbid obesity was an independent predictor of mortality. In the morbidly obese group, 10% of patients had primary graft nonfunction, yet only 7% were listed for retransplantation. It was difficult to determine the minor differences in relisting between the groups from the database. However, we can only speculate that patients with severe or morbid obesity may have developed other complications in the immediate postoperative period, such as an increased requirement for ventilatory support, wound dehiscence, and wound infection. It is possible that these additional complications may have discouraged the patients physicians from relisting those patients. The findings of our study suggest that liver-transplant recipients should be strongly advised to maintain their BMI ideally below 35 kg/m 2, and, more importantly, that morbid obesity (BMI 40 kg/m 2 should be considered as a relative contraindication for OLT. Although all 3 obesity groups (obese, severely obese, and morbidly obese) showed a significantly higher cardiovascular mortality, this translated to a poorer 5-year survival only in severely and morbidly obese patients. Our findings of increased cardiovascular mortality in obese patients were not unexpected. In this study, it was difficult to determine other risk factors for cardiovascular disease, such as diabetes, hypertension, and hyperlipidemia, in a reliable fashion, but it is reasonable to assume that the morbidly obese patients had many other risk factors for cardiovascular diseases in addition to obesity. These risk factors may have contributed to the morbidity and mortality. Previous studies have not been able to detect a survival difference in severely and morbidly obese groups, mainly because of small sample size. The large sample size also allowed us to stratify patients into 5 groups, and we were also able to determine the influence of other confounding factors on survival. Overall infection was the commonest cause of death in all groups, accounting for 32% to 44% of all deaths, with no significant differences among the different groups. This is consistent with the findings of our earlier study in which infection accounted for 57% to 66% of all deaths. 11 Four small studies had previously analyzed the mortality of obese patients undergoing OLT In one study, Sawyer et al. reported a comparable survival in their 40 obese recipients (BMI 30 kg/m 2 ) compared with nonobese recipients. 8 In our study, we showed that immediate and 5-year mortality are similar in the obese group (BMI, kg/ kg/m 2 ) compared with the nonobese group. The main weakness of previous studies was the small sample size and relatively short follow-up. Despite the inherent deficiencies of the UNOS database (missing data on pre-olt risk factors, inability to confirm cause of death, incomplete 5-year survival data), we believe that we have adequate information to suggest that morbid obesity should be considered a relative contraindication for OLT in the era of severe organ shortage. In a recent study, Modlin et al. 7 reported that obese renal-transplant recipients had a significantly higher pretransplantation history of angina and myocardial infarction, 7 and in their study, the 5-year survival in obese subjects was significantly lower than in nonobese patients (67% vs. 89%; P.0002). Cardiac disease was the leading cause of death in the obese groups. Based on their findings, the authors recommended weight reduction to 30 kg/m 2 in all patients awaiting a renal transplantation, and also suggested that obese patients with a history of cardiac disease should not be transplanted until weight reduction has been achieved. We would concur with this advice, but based on our current study, we would modify this to include only those who weigh more than 35 kg/m 2. Moreover, unlike patients with renal failure, patients with liver failure have no option other than OLT. Morbidly obese (BMI 40 kg/m 2 ) subjects had a significantly higher prevalence of primary graft nonfunction. We do not have an adequate explanation for this. We can only speculate that technical difficulties and hepatic artery thrombosis may explain primary graft nonfunction in this group. The prevalence of cryptogenic cirrhosis was significantly higher in patients with a BMI 35 kg/m 2. This is consistent with other studies that suggested that a significant proportion of patients with cryptogenic cirrhosis have underlying nonalcoholic steatohepatitis (NASH). 16,17 In conclusion, 7% of patients undergoing OLT in the U.S. are severely or morbidly obese (BMI 35 kg/m 2 ), and this is associated with an increased prevalence of adverse cardiovascular events leading to a significantly lower 5-year survival. In addition, these patients are likely to gain additional weight after OLT; therefore, patients with a BMI 35 kg/m 2 awaiting liver transplantation should be actively encouraged to lose weight to a target weight of 30 kg/m 2. We believe that those weighing over 40 kg/ m 2 (morbid obesity) should be listed on a highly individualized basis. The findings of our study may be used as a guideline by transplantation centers when markedly obese subjects are considered for liver transplantation. Acknowledgment: the database. References The authors thank the UNOS for providing 1. Flancbaum L, Choban PS. Surgical implications of obesity. Annu Rev Med 1998;49: Flegal KM, Carroll MD, Kuczmarski RJ, Johnson CL. Overweight and obesity in the United States: prevalence and trends, Int J Obes Relat Metab Disord 1998;22: Choban PS, Flancbaum L. The impact of obesity on surgical outcome: a review. J Am Coll Surg 1997;185: Merion RM, Twork AM, Rosenberg L, Ham JM, Burtch GD, Turcotte JG, Rocher LL, et al. Obesity and renal transplantation. Surg Gynecol Obstet 1991;172: Gill IS, Hodge EE, Novick AC, Steinmuller DR, Garred D. The impact of obesity on renal transplantation. Transplant Proc 1993;25: Johnson CP, Kuhn EM, Hariharan S, Hartz AJ, Roza AM, Adams MB. Pre-transplant identification of risk factors that adversely affect length of stay and charges for renal transplantation. Clin Transplant 1999;13:

5 HEPATOLOGY, Vol. 35, No. 1, 2002 NAIR, VERMA, AND THULUVATH Modlin CS, Flechner SM, Goormastic M, Goldfarb DA, Papajcik D, Mastroianni B, Novick AC. Should obese patients lose weight before receiving a kidney transplant? Transplantation 1997;64: Sawyer RG, Pelletier SJ, Pruett TL. Increased early morbidity and mortality with acceptable long-term function in severely obese patients undergoing liver transplantation. Clin Transplant 1999;13: Braunfeld MY, Chan S, Pregler J, Neelakanta G, Sopher MJ, Busuttil RW, Csete M. Liver transplantation in obese patients. J Clin Anesth 1996;8: Keeffe EB, Gettys C, Esquivel CO. Liver transplantation in patients with severe obesity. Transplantation 1994;57: Nair S, Cohen DB, Cohen MP, Tan H, Maley W, Thuluvath PJ. Postoperative morbidity, mortality, cost and long-term survival in severely obese patients undergoing orthotopic liver transplantation. Am J Gastroenterol 2001;96: Ledru F, Ducimetiere P, Battaglia S, Courbon D, Beverelli F, Guize L, Guermonprez JL, et al. New diagnostic criteria for diabetes and coronary artery disease: insights from an angiographic study. J Am Coll Cardiol 2001;37: Ravussin E, Swinburn BA. Pathophysiology of obesity. Lancet 1992; 340: World Health Organization. Report of a WHO consultation on obesity. Obesity: preventing and managing the global epidemic. Geneva: World Health Organization, NHLBI Obesity Education Initiative Expert Panel on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults the evidence report. Obes Res 1998;6:51S-209S. 16. Caldwell SH, Oelsner DH, Iezzoni JC, Hespenheide EE, Battle EH, Driscoll CJ. Cryptogenic cirrhosis: clinical characterization and risk factors for underlying disease. HEPATOLOGY 1999;29: Poonawala A, Nair SP, Thuluvath PJ. Prevalence of obesity and diabetus in patients with cryptogenic cirrhosis: a case controlled study. HEPATOLOGY 2000;32: