SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Controversies exist regarding the morbidity and mortality of patients undergoing liver transplantation at the extremes of the body mass index (BMI). A review of the United Network for Organ Sharing database from 1987 through 2007 revealed 73,538 adult liver transplants. Patients were stratified into 6 BMI categories established by the World Health Organization: underweight, <18.5 kg/m2; normal weight, 18.5 to <25 kg/m2; overweight, 25 to <30 kg/m2; obese, 30 to <35 kg/m2; severely obese, 35 to <40 kg/m2; and very severely obese, ≥40 kg/m2. Survival rates were compared among these 6 categories via Kaplan-Meier survival curves with the log-rank test. The underweight and very severely obese groups had significantly lower survival. There were 1827 patients in the underweight group, 1447 patients in the very severely obese group, and 68,172 patients in the other groups, which became the control. Groups with extreme BMI (<18.5 and ≥40) were compared to the control to assess significant differences. Underweight patients were more likely to die from hemorrhagic complications (P < 0.002) and cerebrovascular accidents (P < 0.04). When compared with the control, the very severely obese patients had a higher number of infectious complications and cancer events (P = 0.02) leading to death. In 3 different eras of liver transplantation, multivariable analysis showed that underweight and very severe obesity were significant predictors of death. In conclusion, liver transplantation holds increased risk for patients at the extremes of BMI. Identifying these patients and instituting aggressive new policies may improve outcomes. Liver Transpl 15:968–977, 2009. © 2009 AASLD.

Liver transplantation is the definitive management option for patients with end-stage liver disease. However, there has been continued controversy regarding liver transplantation in patients with an elevated body mass index (BMI).1–3 Conversely, little has been reported regarding liver transplantation in patients with an extremely low BMI, which is defined as a BMI < 18.5 kg/m2 according to the World Health Organization.4

Obesity has reached pandemic proportions and is a major contributor to the global burden of chronic disease and disability. The World Health Organization states that obesity rates have risen more than 3-fold in North America since 1980.5, 6 Approximately 20% of the US population is obese, as defined by a BMI > 30 kg/m2.1 As a result, the number of obese, severely obese, and very severely obese patients being referred and evaluated for liver transplantation has also increased.2

Numerous studies have shown conflicting results in obese patients undergoing liver transplantation. A retrospective review of the Scientific Registry of Transplant Recipients database conducted by Pelletier et al.2 demonstrated that there was a survival benefit from transplantation not only for obese patients but also for patients at the extremes of BMI. Other smaller studies have also shown acceptable results after transplantation for obese patients.3, 7 However, other studies have been performed, such as a study by Nair et al.,8 who reviewed the United Network for Organ Sharing (UNOS) database from 1988 through 1996 and demonstrated decreased survival in morbidly obese patients undergoing liver transplantation. Suboptimal results have also been demonstrated in various studies of obese patients undergoing kidney, kidney/pancreas, and heart transplantation.9–12

It has been suggested that patients with an extremely low BMI who undergo renal, liver, or cardiac transplantation may have suboptimal outcomes.9, 13, 14 Similar results have also been shown in such patients who undergo major intra-abdominal cancer surgery and major cardiac surgery.15, 16 An extremely low BMI is associated with a severely malnourished state. Malnutrition is common in critically ill patients, and this state can be aggravated by surgery.

The objective of our study was to investigate the impact of pretransplantation BMI on post–liver transplantation patient survival. Specifically, we hypothesized that individuals at the extremes of BMI (BMI < 18.5 or BMI ≥ 40) were at increased risk of death following liver transplantation. By reviewing associated factors related to their deaths, we aimed to develop insights to improve the medical management of these patients.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

After approval by the University of Washington institutional review board, we conducted a retrospective review of the UNOS Standard Transplant Analysis and Research Files from 1987 through 2007. Data were analyzed for all adult (≥18 years old) patients undergoing orthotopic liver transplantation to assess patient survival according to BMI at the time of transplantation. The BMI is calculated by the division of the person's weight in kilograms by the square of his height in meters (kg/m2) according to World Health Organization guidelines.4 Patients were stratified into 6 groups based on their BMIs. The groups included underweight (BMI < 18.5 kg/m2), normal (18.5 kg/m2 ≤ BMI < 25 kg/m2), overweight (25 kg/m2 ≤ BMI < 30 kg/m2), obese (30 kg/m2 ≤ BMI < 35 kg/m2), severely obese (35 kg/m2 ≤ BMI < 40 kg/m2), and very severely obese (BMI ≥ 40 kg/m2) patients. BMIs were obtained from the UNOS Transplant Recipient Registration Form. If there were missing data or conflicting information (ie, confusion of pounds with kilograms or centimeters with inches), the Transplant Candidate Registration Form was used to confirm or adjust the BMIs. Patients with no BMI information or no follow-up forms were excluded from our study.

Recipient data that were collected included age, sex, ethnicity, and etiology of liver disease, including acute hepatic necrosis, hepatitis B as the only diagnosis, hepatitis C together with any other diagnosis, cryptogenic cirrhosis (as labeled in the UNOS Standard Transplant Analysis and Research Files), alcoholic liver disease as the only diagnosis, cholestatic liver disease (primary biliary cirrhosis or primary sclerosing cholangitis), metabolic liver disease (α-1-antitrypsin deficiency, Wilson's disease, hemochromatosis, and others), and presence of a tumor (any type) as the primary diagnosis. Other recipient factors that were collected included pretransplantation complications from liver disease (ascites and spontaneous bacterial peritonitis) and retransplants.

Recipient factors at the time of transplantation were collected, including the calculated Model for End-Stage Liver Disease (MELD) score (after February 2002) and the individual laboratory values for total bilirubin (mg/dL), serum creatinine (mg/dL), and albumin (g/dL). Other conditions at the time of transplantation were collected, including intensive care unit status, any form of life support (ventilator, cardiac balloon pump, and other), days waiting for a transplant, need for dialysis, and requirement for a liver and kidney transplant.

Donor factors that were collected included age, BMI, type of donor (donation after brain death or donation after cardiac death), living donor, type of liver (whole versus segmental), cold ischemia time in hours (time period from the excision of the donor liver and placement on ice until the removal from ice at transplantation), warm ischemia time in minutes (time period from the removal of the donor liver from ice at transplantation until the reestablishment of blood flow), and national allocation source of the donor liver.

Posttransplantation factors that were collected included length of stay from transplantation to discharge. Posttransplantation causes of death, such as intraoperative, hemorrhagic, or cardiac factors, cerebrovascular disease, graft failure, respiratory difficulties, infectious events, cancer, and renal failure, were also evaluated.

The main outcome measure that was assessed was patient survival. Survival rates during 3 different eras [1987-1992, 1993 to February 2002 (the 10 years prior to implementation of the MELD system for donor liver allocation in the United States), and March 2002 to 2007 (after institution of the MELD system)] were analyzed. For each time period, significant variables for patient survival were determined for both univariable analysis and multivariable analysis by Cox proportional hazards methods.17 The MELD score was substituted for creatinine and total bilirubin for time period 3.

Statistical Analysis

Only data with >80% collection for each variable were noted with no data imputation for statistical analysis. Continuous variables were given as the mean ± standard deviation, and categorical variables were presented as proportions. Nonparametric tests were used for continuous variables, and the chi-square test was used for categorical variables. Kaplan-Meier analysis with the log-rank test was used for comparing the survival among the underweight, very severely obese, and control groups. The Cox proportional hazards model was used to determine significant univariables and multivariables predicting post–liver transplant survival. For all of the analyses, a P value ≤ 0.05 was considered significant. The data were evaluated with JMP 7.0 statistical software (SAS Institute, Cary, NC).

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

From 1987 through 2007, 73,538 adult liver transplants were performed in the United States, as shown by the UNOS Standard Transplant Analysis and Research Files. There were 1309 patients with no BMI information and 783 patients with no Transplant Candidate Registration Forms, for whom it was necessary to rule out conflicting information. Our final study population therefore consisted of 71,446 patients. Not all patients had complete data for all variables, and we subsequently used only data that were >80% complete.

Kaplan-Meier survival curves (Fig. 1) were generated, and they demonstrated that there was statistically significant decreased survival in patients with a BMI < 18.5 kg/m2 or a BMI > 40 kg/m2. As a result, patients with a BMI < 18.5 kg/m2 or a BMI > 40 kg/m2 were assigned to the underweight or very severely obese study group, respectively. Patients with a BMI between 18.5 and 40 kg/m2 were assigned to our control group because they had similar survival rates. The distribution of 1827 patients transplanted in the underweight group, 1447 patients in the very severely obese group, and 68,172 patients in the control group is shown in Fig. 2.

thumbnail image

Figure 1. Patient survival according to the body mass index (BMI) groups.

Download figure to PowerPoint

thumbnail image

Figure 2. Distribution of body mass indices (BMIs) for all 71,446 patients in the final study population.

Download figure to PowerPoint

A comparison of recipient, donor, and follow-up variables between our 2 groups and the controls is demonstrated in Table 1. Compared with the control group, both the underweight and very severely obese groups were younger and female, spent time in the intensive care unit, and required any type of life support. Both the underweight and severely obese groups had a lower incidence of a primary tumor as their indication for liver transplantation. At the time of transplantation, both the underweight and severely obese groups received a higher number of donors via national allocation and experienced longer lengths of hospital stay after admission for transplantation than did the control group.

Table 1. Comparison of Recipient, Donor, and Follow-Up Variables Among Study Subjects
VariableControl (n = 68,172 Patients)BMI < 18.5 (n = 1827 Patients)P*BMI ≥ 40 (n = 1447 Patients)P*
  • Abbreviations: BMI, body mass index; ICU, intensive care unit; LOS, length of stay; MELD, Model for End-Stage Liver Disease.

  • *

    P value compared to the control group.

  • Last laboratory before transplant.

  • For time period 3 only.

Recipient     
 Age (yrs)50.3 ± 10.845 ± 13.6<0.000149.7 ± 9.10.002
 Male0.640.43<0.00010.48<0.0001
 Caucasian0.840.79<0.00010.870.002
 African American0.080.1<0.00010.080.4
 Hispanic0.040.030.50.040.
 Asian0.040.07<0.00010.004<0.0001
 Unknown or other race0.0090.0060.60.0070.4
 Acute hepatic necrosis0.070.070.50.080.06
 Hepatitis B0.040.0410.01<0.0001
 Hepatitis C (any)0.370.24<0.00010.490.006
 Cryptogenic cirrhosis0.040.0410.07<0.0001
 Alcoholic liver disease0.20.16<0.00010.14<0.0001
 Cholestatic liver disease0.140.28<0.00010.05<0.0001
 Metabolic disease0.030.06<0.00010.060.2
 Tumor, primary0.070.05<0.00010.04<0.0001
 Other disease0.040.09<0.00010.040.3
 Ascites0.840.76<0.00010.870.003
 Bacterial peritonitis0.120.120.90.10.01
 Retransplant0.110.21<0.00010.090.03
Recipient condition at transplant     
 Laboratory MELD20.1 ± 8.920.7 ± 90.222.2 ± 90.004
 Bilirubin, mg/dL7.9 ± 10.610.2 ± 12.2<0.00018.1 ± 10.60.4
 Serum creatinine (value?)1.5 ± 1.31.4 ± 1.40.041.6 ± 1.3<0.0001
 Albumin (value?)2.9 ± 0.73 ± 0.80.12.8 ± 0.60.1
 Any form of life support0.10.15<0.00010.130.01
 On dialysis at transplant0.070.10.00020.080.2
 Requiring liver/kidney transplant0.040.050.0040.030.03
 Days waiting for transplant215 ± 353161.4 ± 305<0.0001228 ± 3380.08
 In ICU at time of transplant0.190.25<0.00010.20.02
Donor data at transplant     
 Whole liver0.960.950.060.98<0.0001
 Warm ischemia time, minutes51.4 ± 23.851.8 ± 24.70.752.7 ± 25.70.02
 Cold ischemia time, hours8.7 ± 4.89.3 ± 5<0.00018.8 ± 4.20.06
 Donor from national allocation0.10.14<0.00010.080.04
 Donor age (yrs)37.5 ± 1733 ± 18<0.000139.1 ± 17.20.0003
 Donor BMI (kg/m2)25.4 ± 5.223.4 ± 5.2<0.000126.3 ± 5.3<0.0001
 Donation after cardiac death0.030.0310.030.5
 Living donor0.030.030.20.010.0001
Post-transplant     
 LOS following transplant20.8 ± 24.828.5 ± 33.7<0.000121.9y ± 25.10.001
 Required retransplant0.090.110.010.090.9
 Death during follow-up0.3650.5<0.00010.390.02
 Cause of death     
  Cardiac0.120.10.060.130.8
  Graft failure0.150.130.30.151
  Respiratory0.040.040.90.050.3
  Renal failure0.130.110.10.150.4
  Hemorrhagic0.040.070.0020.060.2
  Infection0.220.240.10.260.02
  Cancer0.120.110.60.080.004
  Intraoperative0.010.020.50.011
  Cerebral vascular disease0.030.050.040.030.9
  Other cause of death0.140.130.60.080.01

Patients in the underweight group were more likely than the control group (Table 1) to be Asian or African American, were more likely to have undergone a previous liver transplant, and were more likely to have cholestatic liver disease or metabolic liver disease (Table 1). Of note, these patients also had lower rates of ascites and spent less time on the waiting list. The underweight patients showed no difference in the albumin level in comparison with the control group, probably because of therapy received in the intensive care unit. At the time of transplantation, underweight patients required a longer cold ischemia time, more dialysis, and a higher rate of combined liver-kidney transplantation.

Compared with the control group, the underweight patients had a higher retransplantation rate due to graft failure and were more likely to die from hemorrhagic complications or cerebrovascular accidents.

Comparing the change in survival of the underweight patients across the 3 eras of transplantation [1987-1991, 1992 to February 2002 (start of the MELD allocation system in the United States), and March 2002 to 2007] reveals no improvement (P = 0.3) in the survival of this group (Fig. 3A). This is in contrast to the increased survival of the control group for the last 2 eras compared to the 1987-1992 era (Fig. 3B).

thumbnail image

Figure 3. (A) Patient survival of the underweight group in 3 different eras (1987-1992, 1993 to February 2002, and March 2002 to 2007) in liver transplantation. The P value compares survival in all 3 eras. (B) Patient survival of the control group in 3 different eras (1987-1992, 1993 to February 2002, and March 2002 to 2007) in liver transplantation. The P value compares survival between the 1987-1992 and 1993-2007 eras. (C) Patient survival of the very severely obese group in 3 different eras (1987-1992, 1993 to February 2002, and March 2002 to 2007) in liver transplantation. The P value compares survival between the 1987-1992 and 1993-2007 eras.

Download figure to PowerPoint

For all 3 time periods, underweight as an independent variable was a significant predictor of risk to survival (Table 2, n = 10,125 patients; Table 3, n = 33,612 patients; Table 4, n = 27,709 patients). In period 1, after adjustments for other factors, underweight increased the chance of poor survival by 14% (P < 0.0001). In period 2, the risk of poor survival was 17% (P < 0.0001), and in period 3, the increased risk was 73% (P < 0.0001).

Table 2. Analysis of Variables as Predictors of Risk to Survival During Period 1 (1987-1991; n = 10,125 Patients)
VariableUnivariate AnalysisMultivariable AnalysisRank- Adjusted Risk
RRPRRP
  1. Abbreviations: BMI, body mass index; ICU, intensive care unit; RR, relative risk.

Recipient     
 Age, years1.015<0.00011.018<0.000112
 Male1.13<0.00011.09<0.00018
 BMI < 18.5 (kg/m2)1.070.021.14<0.00016*
 BMI ≥ 40 (kg/m2)1.170.011.170.015*
 BMI ≥ 18.5 and < 40 (kg/m2)Reference    
 CaucasianReference    
 African American1.13<0.00011.22<0.00014
 Hispanic0.950.03   
 Asian1.2<0.00011.080.039
 Acute hepatic necrosis0.89<0.0001   
 Hepatitis B1.15<0.00011.14<0.00016
 Hepatitis C (any)1.0050.8   
 Cryptogenic cirrhosis1.37<0.00011.38<0.00012
 Alcoholic liver disease1.1<0.00011.12<0.00017
 Cholestatic liver diseaseReference    
 Metabolic disease0.90.004   
 Tumor1.38<0.00011.46<0.00011
 Retransplant1.38<0.00011.33<0.00013
Recipient condition at transplant     
 Log bilirubin1.020.1   
 Log serum creatinine1.37<0.00011.17<0.00015
 Log albumin0.80.00010.850.005Lower risk
 Any form of life support1.23<0.00011.050.0311
 Requiring liver/kidney transplant1.050.2   
 In ICU at time of transplant1.2<0.00011.070.00310
Donor data at transplant     
 Segmental or reduced liver1.30.1   
 Cold ischemia time, hours1.00050.8   
 Donor from national allocation1.050.0004   
 Donor age1.009<0.00011.006<0.000113
 Donor BMI1.0130.0005   
 Living donor2.10.2   
Table 3. Analysis of Variables as Predictors of Risk to Survival During Period 2 (1992 to February 2002; n = 33,612 Patients)
VariableUnivariate AnalysisMultivariable AnalysisRank- Adjusted Risk
RRPRRP
  1. Abbreviations: BMI, body mass index; ICU, intensive care unit; RR, relative risk.

Recipient     
 Age, years1.014<0.00011.012<0.000112
 Male1.03<0.0001   
 BMI < 18.5 (kg/m2)1.19<0.00011.17<0.00017*
 BMI ≥ 40 (kg/m2)1.060.051.070.0511*
 BMI ≥ 18.5 and < 40 (kg/m2)Reference    
 CaucasianReference    
 African American1.12<0.00011.15<0.00018
 Hispanic0.920.020.790.003Lower risk
 Asian0.940.02   
 Acute hepatic necrosis1.020.4   
 Hepatitis B0.87<0.0001   
 Hepatitis C (any)1.1<0.00011.25<0.00014
 Cryptogenic cirrhosis1.56<0.00011.71<0.00011
 Alcoholic liver disease1.030.0031.18<0.00016
 Cholestatic liver diseaseReference    
 Metabolic disease0.9<0.0001   
 Tumor1.2<0.00011.4<0.00012
 Ascites1.12<0.0001   
 Retransplant1.43<0.00011.28<0.00013
Recipient condition at transplant     
 Log bilirubin1.06<0.0001   
 Log serum creatinine1.42<0.00011.2<0.00015
 Log albumin0.7<0.00010.78<0.0001Lower risk
 Any form of life support1.3<0.00011.1<0.00019
 On dialysis at transplant1.33<0.00011.080.00110
 Requiring liver/kidney transplant1.15<0.0001   
 In ICU at time of transplant1.2<0.0001   
Donor data at transplant     
 Segment or reduced liver1.040.1   
 Cold ischemia time, hours1.0050.0031.0080.000314
 Donor from national allocation1.14<0.0001   
 Donor age1.011<0.00011.01<0.000113
 Donor BMI1.0070.0002   
 Donation after cardiac death1.170.0041.150.048
 Living donor1.050.1   
Table 4. Analysis of Variables as Predictors of Risk to Survival During Period 3 (March 2002 to 2007; n = 27,709 Patients)
VariableUnivariate AnalysisMultivariable AnalysisRank- Adjusted Risk
  1. Abbreviations: BMI, body mass index; ICU, intensive care unit; MELD, Model for End-Stage Liver Disease; RR, relative risk.

RecipientRRPRRP 
 Age, years1.01<0.00011.02<0.000113
 Male0.980.5   
 BMI < 18.5 (kg/m2)1.55<0.00011.73<0.00014*
 BMI ≥ 40 (kg/m2)1.44<0.00011.410.00057*
 BMI ≥ 18.5 and < 40 (kg/m2)Reference    
 CaucasianReference    
 African American1.34<0.00011.3<0.000110
 Hispanic1.130.1   
 Asian0.930.2   
 Acute hepatic necrosis1.37<0.00011.330.00059
 Hepatitis B0.68<0.0001   
 Hepatitis C (any)1.18<0.00011.52<0.00015
 Cryptogenic cirrhosis3.15<0.00013.71<0.00011
 Alcoholic liver disease0.980.6   
 Cholestatic liver diseaseReference    
 Metabolic disease0.910.02   
 Tumor1.130.0061.74<0.00013
 Ascites1.150.0004   
 Retransplant2.12<0.00011.76<0.00012
Recipient condition at transplant     
 Laboratory MELD1.02<0.00011.01<0.000114
 Log albumin0.62<0.00010.72<0.0001Lower risk
 Any form of life support2.26<0.00011.44<0.00016
 On dialysis at transplant1.7<0.00011.230.000211
 Requiring liver/kidney transplant1.180.004   
 In ICU at time of transplant2.05<0.00011.3<0.000110
Donor data at transplant     
 Segment or reduced liver0.74<0.0001   
 Cold ischemia time, hours1.02<0.00011.010.000214
 Donor from national allocation1.4<0.00011.18<0.00112
 Donor age1.01<0.00011.01<0.000114
 Donor BMI1.0060.02   
 Donation after cardiac death1.31<0.00011.36<0.00018
 Living donor0.7<0.0001   

A comparison of recipient donor and follow-up variables between patients in the very severely obese group and the controls is demonstrated in Table 1. The very severely obese patients were more likely than the controls to be Caucasian with end-stage liver disease secondary to hepatitis C and to have cryptogenic cirrhosis or other conditions (including the patients with nonalcoholic steatohepatitis and the patients not requiring retransplantation). The very severely obese patients were more likely than the controls to have severe ascites.

At transplantation, the very severely obese patients received a higher number of whole liver transplants than the controls. The very severely obese patients also had higher MELD scores (March 2002 to 2007) than the control group. The survival rate of the very severely obese patients with a MELD score ≤ 22 was equivalent to that of those controls with a MELD score > 22 (Fig. 4). They also had longer warm ischemia times and older donors with higher BMIs in comparison with the controls. After transplantation, the severely obese patients experienced higher rates of death due to infectious complications and cancer.

thumbnail image

Figure 4. Survival of the control group with Model for End-Stage Liver Disease (MELD) scores ≤ 22 or > 22 compared to the survival of the very severely obese group with MELD scores ≤ 22 or > 22. This graph shows data only from period 3: March 2002 to 2007.

Download figure to PowerPoint

The improvement in survival of the very severely obese group mirrors the increase in survival of the control group across the 3 eras of transplantation (Fig. 3C). Severely obese posttransplantation patients experienced a significant increase in survival in the 2 latest eras (1992 to February 2002 and March 2002 to 2007) in comparison with the first era (1987-1991).

For all 3 time periods, very severe obesity as an independent variable was a significant predictor of risk to survival (Tables 2–4). In period 1, after adjustments for other factors, very severe obesity increased the chance of poor survival by 17% (P < 0.01). In period 2, the risk to poor survival decreased for very severely obese patients to 7% (P = 0.05). However, in the third period, with allocation by MELD score, the risk of very severe obesity was 41% (P = 0.0005). Of note, the presence of ascites in the third period was a univariable risk factor (relative risk, 1.15; P = 0.0004) for poor patient survival; however, on multivariable analysis, ascites was not an independent factor. The presence of ascites did not interact statistically with very severe obesity as an independent risk factor. However, the presence of ascites was highly related (P < 0.0001) to elevated serum creatinine and thus to the MELD score in period 3.

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

This is the largest known study of liver transplantation at the extremes of BMI. Our review demonstrates that patients undergoing liver transplantation who are underweight or very severely obese experience significantly higher rates of morbidity and mortality in comparison with patients in the middle BMI categories. Also, across the 3 different eras of liver transplantation, underweight and very severe obesity are independent variables of death in a multivariable analysis. We do not suggest that patients at the extremes of BMI are not transplant candidates or do not enjoy a transplant benefit (higher survival with a transplant than without a transplant), but they do have poorer survival. What can be done for these patients to improve their survival? Does our review offer any insights into providing better care for these patients?

First of all, there is a tremendous cost required to take care of these patients at the extremes of BMI. We found that both the underweight and severely obese groups had longer lengths of stay following transplantation and utilized more intensive care unit services. A transplant program taking care of these patients must be aware of the increased cost.

The higher mortality in obese patients in the first era of liver transplantation was thought to be partly technically related to the difficulty of surgery in the presence of obesity. The survival curve in the 1987-1992 liver transplantation era does have an early mortality component that could be traced to the operative procedure. In the later eras, possibly because of better equipment, the early mortality rate decreased in the very severely obese patient population.

Some transplant programs limit the upper weight range of very severely obese patients to a certain BMI and require the patients to lose weight before being placed on the transplant list. In our study, the severely obese patients had a higher MELD score at transplantation than the non–severely obese patients. To improve the survival of very severely obese patients soon after transplantation, it might be better to perform transplantation with a higher BMI but a lower MELD score. The extra time spent on the effort to lose weight can result in an increase in MELD scores. As we have shown, the very severely obese patients in our study group had significantly better survival rates when transplanted at a lower MELD score. These patients should be transplanted sooner without weight loss and at a lower MELD score, or consideration should be given to assigning extra MELD points to this group so that these patients can undergo transplantation with lower calculated MELD points. Additionally, repeated attempts at weight loss have been seen to result in higher mortality rates in obese patients.18 Thus, weight loss just prior to liver transplantation might actually result in higher short-term mortality. The best strategy might be weight loss following transplantation. A recent case report on gastric banding suggests that a bariatric surgical procedure can be performed at the time of transplantation.19 Some have suggested weight-loss reduction surgery after the patient is stable following transplantation.20 Likewise, enforced weight loss prior to transplantation does little to prevent long-term posttransplantation complications of diabetes, hypertension, coronary artery disease, and stroke.21, 22 In our experience, most posttransplant patients quickly regain their amount of pretransplantation weight loss, with nothing having been done to treat their metabolic syndrome. Weight-loss programs are difficult to sustain in the long term in these patients; thus, surgical weight-loss procedures after transplantation may be the better treatment option. Certainly, more experience and study are needed with surgical methods of weight loss in the obese transplant population.

Our study is consistent with other investigations, which have shown that patients with an elevated BMI have higher rates of infectious complications and cancer after transplantation.8, 9, 15, 23 In fact, in our study, infectious complications were the only statistically significant complications that accounted for the deaths of these patients. We propose that 1 mechanism for this apparent immune deficiency is the presence of diabetes in patients with a BMI > 40 kg/m2. It has been shown in previous studies that diabetic patients are at increased risk of infectious complications after surgical procedures.24 This increased risk of infections is secondary to the blunted immune system, which does not respond appropriately to invading pathogens.24 In addition, supplemental immunosuppressive medication may further exacerbate this process. In addition to an appropriate weight-based immunosuppressive regimen, careful management of severely obese patients' comorbidities (diabetes and hypertension) and aggressive facilitation of weight reduction can optimize the health of these patients and potentially improve patient outcomes.

In our study, obese patients were more likely than controls to have ascites. Previous studies have suggested that patients with an elevated BMI may have a falsely increased BMI secondary to ascites and that if ascites were controlled for, then BMI would not be an independent predictor of poor survival.25, 26 Leonard et al.25 stated that ascites (and not BMI) was associated with an increased risk for postoperative morbidity and mortality. In that study, each liter of ascites removed was associated with a 7% increased relative risk of mortality. In our study, ascites was a univariate predictor of poor survival but fell out as a multivariable predictor. Throughout the 3 periods of transplantation, a BMI > 40 was a multivariable predictor of poor survival.

The early survival of very severely obese patients has improved over the 3 eras of liver transplantation (Fig. 3C), but their long-term survival is decreasing in the latest era. We suggest that improved surgical instrumentation (such as retractors) and surgical techniques have decreased the technical complications in this group; however, the incidence of comorbidities such as diabetes and hypertension has not changed, and this has resulted in the observed decreased long-term survival. Further research needs to be conducted to evaluate whether a more aggressive listing policy for these patients and transplantation at a lower calculated MELD score, in addition to optimization of their comorbidities pre-transplantation and post-transplantation, may improve the overall survival of these patients.

The underweight patients may actually be more difficult to manage. Patients who are underweight and undergo complex surgical procedures have a significantly increased risk of postoperative mortality as a result of poor nutritional status.15 Malnutrition is common in patients with advanced liver disease and is a significant prognostic indicator of poor outcomes after liver transplantation.27, 28 Among our study subjects, more of the underweight patients received retransplantation in comparison with controls and had a longer length of stay in the hospital post-transplantation. We suspect that the underweight patients were more likely to be receiving hyperalimentation, albumin infusion, or increased amounts of nutritional supplements. Likewise, they were transplanted with a MELD score similar to that of controls and after a significantly shorter waiting time than that for the controls. It seems unlikely that shorter waiting times, produced by the assignment of higher MELD scores, or increased hospitalization with more nutritional support would benefit this group.

Underweight patients were more likely than controls to die from cerebrovascular accidents. Patients in the underweight group were mostly Asians and African Americans. Previous studies in Japan and Korea have shown a relationship between a low BMI (<18.5 kg/m2) and increased risk of fatal strokes in their study populations.29–31 These patients should either be screened in the evaluation phase or be given special vigilance in the posttransplantation period to prevent strokes.

Underweight patients have been a difficult group to treat over the different eras of liver transplantation, with no improvement in survival across the eras, as seen in the severely obese and control groups. Radically new developments are needed to improve the survival of critically ill underweight transplant candidates.

In conclusion, liver transplantation is a life-saving measure for people with end-stage liver disease. There is currently a disparity between supply (organs) and demand (patients on the waiting list). Because of the organ shortage, appropriate patient selection for liver transplantation is essential. Donor livers are precious, and our goal should be to maximize both patient and graft survival. It is presently the policy of our institution to not transplant a patient whose BMI is ≤18.5 kg/m2 or ≥40 kg/m2, but we are looking for innovative ways to take care of these high-risk groups. For patients who are severely obese, we have tried to resolve their comorbidities (diabetes, hypertension, and hyperlipidemia) and help them achieve weight loss prior to transplantation. A better approach might be to transplant these patients sooner by not requiring weight loss or working with UNOS for a policy change to assign additional MELD points for severe obesity, as is done for patients with hepatocellular carcinoma. However, because of the increased risk of infection in this patient population, we recommend a posttransplantation immunosuppressive regimen favoring fewer immunosuppressive medications without steroids and low-dose tacrolimus based on the ideal body weight. Aggressive management of the patients' comorbid factors and posttransplantation weight loss is a must. For patients who are underweight, we recommend close follow-up with a nutritionist. If the patients are unable to meet their caloric intake prior to transplantation, they should then be admitted to the hospital for aggressive nutritional supplementation such as tube feedings. This aggressive regimen is continued after transplantation. We also suggest a more aggressive immunosuppressive regimen with higher doses of tacrolimus and mycophenolate mofetil.

Further research in the areas of obesity, bariatric surgery, nutrition, and immunosuppressive management will hopefully bring further insights into the transplant care of these difficult-to-manage patients.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES