Suitability criteria have broadened over time resulting in more end-stage renal disease (ESRD) patients with complex medical conditions receiving kidney transplants. This change in practice has occurred due to the increasing number of patients waiting for transplants (1) and the improvements in both duration (2) and quality of life (3) after kidney transplantation. These benefits have been reported in groups once deemed to be inappropriate for transplantation (4–7).
An area of ongoing controversy in kidney transplantation is the acceptance of obese patients as kidney transplant candidates. The most widely used measure of obesity is the body mass index (BMI) defined as the weight (in kilograms) divided by the height (in meters) squared. Although this measure does not take into account lean body mass or the distribution of adiposity, it is easily measured and has been endorsed by the World Health Organization. In many centers, increased BMI above a certain threshold (e.g., >35 kg/m2) is an exclusion criterion for kidney transplantation (8) due to its implications for surgical complications and posttransplantation cardiovascular morbidity/mortality (9–13).
More patients on dialysis now have larger BMI than in the past (14), resulting in many patients being overweight or obese at the time of kidney transplantation (15). This has also led to greater numbers of patients with increased BMI presenting for transplant assessment. Studies examining the impact of BMI on kidney transplant outcomes have been conflicting (10, 16, 17). The uncertainty about the results of kidney transplantation in these patients led us to investigate this issue in our own center. Specifically, this report examines the role of BMI at the time of kidney transplantation in predicting adverse posttransplantation outcomes, including biopsy-proven acute rejection (BPAR).
After applying the inclusion and exclusion criteria, 1151 patients comprised the final study cohort (see SDC 1, http://links.lww.com/TP/A869). A total of 236 delayed graft function (DGF) events, 202 first BPAR episodes, and 165 all- cause graft failures (89 graft losses and 76 deaths) accrued over 4405 patient-years of follow-up. The distribution of BMI in the study population was generally right skewed with a mean (SD) of 26.4 (5.3) and median (interquartile range [IQR]) of 25.6 (6.8). A quarter of all patients had BMI<22.5, whereas a quarter of patients had BMI>29.4. The range of BMI in the study cohort spanned from 15.6 to 54.9. Since 2006, patients in the upper BMI categories (BMI≥30) represented 25% to 30% of all patients in our program (see SDC 2, http://links.lww.com/TP/A869).
Table 1 shows the baseline characteristics of the study patients by BMI categories. Recipients in the highest BMI group were more likely to be older, female, Caucasian, first transplants, and had ESRD due to diabetes. The lowest BMI group had a higher proportion of female recipients versus the middle three BMI groups. Expanded-criteria donor kidneys were less prevalent in the highest (and lowest) BMI groups. There were no systematic trends observed in transplant factors. Median calcineurin inhibitor blood levels, mean mycophenolate mofetil dose, and mean prednisone dose were generally somewhat higher in the upper BMI categories (see SDC 3–6, http://links.lww.com/TP/A869). Of note, 25 of 74 patients in the highest category had BMI≥40 kg/m2 (i.e., morbid obesity).
Table 2 displays results from the logistic regression models examining the risk of DGF across each BMI category (reference, 20–24.9 kg/m2). The proportion of patients with DGF in the <20, 20 to 24.9, 25 to 29.9, 30 to 34.9, and ≥35 kg/m2 BMI groups were 15.3%, 17.5%, 19.3%, 25.5%, and 37.2%, respectively (P<0.001). We constructed a series of models, each incorporating a more extensive set of covariates. The unadjusted model (Model 1) showed that recipient BMI of 30 to 34.9 and ≥35 kg/m2 were associated with a significantly increased risk of DGF (odds ratio [OR; 95%confidence interval [CI], 1.62 [1.07–2.45] and 2.79 [1.66–4.71], respectively). The fully adjusted model (Model 4) showed a consistently increased risk of DGF in the two highest BMI groups (OR [95% CI], 1.96 [1.19–3.24] and 4.31 [2.19–8.49], respectively).
Figure 1 depicts the cumulative probability of BPAR by BMI category. The highest BMI group showed the greatest risk of BPAR in the early posttransplantation period and this finding persisted throughout follow-up (log-rank P<0.001). The cumulative probability of BPAR at 10 years in the <20, 20 to 24.9, 25 to 29.9, 30 to 34.9, and ≥35 kg/m2 BMI groups were 24.8%, 20.6%, 19.0%, 20.1%, and 43.9%, respectively. The lowest BMI group showed an intermediate risk of BPAR, but this risk was not statistically different from the three middle BMI groups. There were no systematic trends in the severity or type of acute rejection across BMI groups (see SDC 7, http://links.lww.com/TP/A869).
Table 3 shows multivariable Cox proportional hazards models for the risk of BPAR according to BMI category. Theunadjusted Cox model (Model 1) revealed the hazard ratio (HR; 95% CI) for BPAR among patients with BMI≥35 kg/m2 (vs. 20–24.9 kg/m2) to be 2.19 (1.38–3.47). The fully adjusted Cox model (Model 4) showed a similar HR (95%CI) of 2.18 (1.34–3.56). When T-cell– and antibody-mediated BPAR were evaluated separately, there was a moremarked (and statistically significant) increase in therisk of T-cell–mediated BPAR (HR [95% CI], 2.52 [1.44–4.39]) versus antibody-mediated BPAR (HR [95% CI], 1.79 [0.83–3.84]).
Table 4 depicts multivariable Cox proportional hazards models for the outcomes of all-cause graft failure, death-censored graft failure, and death with graft function. The results are shown without and with the inclusion of BPAR as a time-varying covariate. Fully adjusted models indicated that the risk of both all-cause graft failure and death-censored graft failure was significantly increased in patients with BMI≥35 kg/m2 versus the reference group, but the HRs were significantly attenuated after inclusion of BPAR as a time-varying covariate. Although the HR for death with graft function was elevated among patients with BMI≥35 kg/m2 versus the reference group, it was less marked and not statistically significant.
Sensitivity analyses on subcohorts of patients with zero peak panel-reactive antibody (PRA), immediate graft function, and deceased-donor kidney transplant recipients revealed similar findings to the primary analysis across all outcomes. For example, among patients with no PRAs at kidney transplantation, BMI≥35 kg/m2 continued to be associated with a significantly increased risk of BPAR (HR [95% CI], 2.79 [1.35–5.77]). Of note, the overall results were also robust to the inclusion of patients with primary non-function. Finally, reclassifying the lowest category of BMI to less than 18.5 kg/m2 did not appreciably alter the results (see SDC 8–10, http://links.lww.com/TP/A869).
Our analysis reveals that increased recipient BMI atthe time of kidney transplantation is a risk factor for both short-term and long-term adverse outcomes after transplantation. Recipient BMI≥30 kg/m2 was associated with an increased risk of DGF, whereas BMI≥35 kg/m2 was associated with increased risks of BPAR, all-cause graft failure, and death-censored graft failure. These associations persisted after adjustment for recipient, donor, and transplant factors and were robust to various sensitivity analyses.
To date, studies examining the association of increased BMI and posttransplantation outcomes have been conflicting. A number of reports have found no relationship between increased BMI and the occurrence of DGF (10, 17–19). These studies tended to have smaller sample sizes than the studies showing an increased risk of DGF in patients with larger BMI (16, 20–22). We observed a strong association between BMI and the risk of DGF, particularly in patients with BMI≥30 kg/m2. Some of the technical challenges of transplanting obese patients may contribute to this risk (18). Moreover, obesity may be associated with increased thrombin formation leading to a prothrombotic state and the subsequent development of graft microthrombi, which may increase the risk of DGF (21).
The association of recipient obesity with BPAR is more controversial, with no observed increased risk in several reports (10, 16–18). However, more recent studies have identified an increased risk associated with obesity (20, 22, 23). Our findings corroborate this latter set of studies. Notably, there appeared to be a threshold effect for patients with BMI≥35 kg/m2. Although there may have been inadequate statistical power to establish a dose–response effect, the relative hazard for BPAR in the lower BMI categories was close to 1, whereas the highest category showed a clear increment in risk.
One of the most commonly purported reasons for the increased risk of BPAR in obese patients is the challenge in achieving adequate exposure to maintenance immunosuppression. Cyclosporine is lipophilic and distributes preferentially to adipose-rich tissues (24). Altered pharmacokinetics with obesity have also been associated with difficulties in dose adjustments because the relation between oral dose and area under the concentration time curve is likely nonlinear (25). Other possible mechanisms increasing the risk of acute rejection episodes in obese patients include ischemia-reperfusion injury due to increased warm ischemia time during the transplant procedure and reduced adiponectin levels (26). The latter is associated with obesity and has been shown to increase the risk of acute rejection in a mouse model of cardiac transplantation (27).
Some data exist supporting the use of rabbit antithymocyte globulin over interleukin-2 receptor blockers to improve outcomes in obese patients (28). Dosing of rabbit antithymocyte globulin may be insufficient in obese patients because there may be discomfort by clinicians to use very large doses based on actual body weight. Although the actual doses given to patients in our cohort were higher in larger patients, the weight-based doses of rabbit antithymocyte globulin tended to be slightly less in the upper BMI categories (see SDC 11, http://links.lww.com/TP/A869). The best balance between dosing of induction therapy to maximize antirejection efficacy versus reducing the risk of future infectious and cancer-related complications in obese patients requires further study.
We found that the risk of all-cause graft failure and death-censored graft failure were significantly increased in recipients with BMI≥35 kg/m2. A similar but nonsignificant trend was witnessed for the risk of death with graft function. A number of studies in the literature support our findings (9, 16, 18, 19, 22, 23), although some have found no association between BMI and graft and/or patient outcomes (10, 17, 20). Of note, these studies have varied with respect to patient characteristics, length of follow-up, sample sizes, and analytical approaches. Interestingly, the risk of graft failure as a function of BMI was attenuated after adjustment for BPAR as a time-varying covariate. Although the loss of statistical significance may partly reflect a reduction in power, it is also possible that acute rejection mediates at least some of the detrimental effect of increased BMI on long-term outcomes (as evidenced by the change in point estimates). The latter is further supported by the presence of a threshold effect at the highest BMI category for both BPAR and graft outcomes (before BPAR adjustment).
The relation between increased BMI and graft failure may also relate to altered intrarenal hemodynamics, which may lead to chronic allograft dysfunction and graft loss (29, 30). Endothelial dysfunction has also been suggested to result from obesity-related increases in the nitric oxide synthase inhibitor asymmetric dimethylarginine, leading to vascular dysfunction and reductions in adiponectin (26, 31, 32). Adiponectin has antiatherosclerotic/anti-inflammatory properties and is reduced in the blood of obese kidney transplant recipients with the metabolic syndrome (33). Reduced pretransplantation adiponectin levels have also been associated with allograft failure over 3 years of follow-up (34).
Our study confirms the association of recipient BMI atkidney transplantation and adverse posttransplantation outcomes. In particular, it extends previous observations that higher BMI may increase the risk of BPAR. We showed these results in a large cohort, with more than 4000 person-years of follow-up, under a common immunosuppression protocol and using multivariable modeling strategies to reduce the potential effects of confounding bias.
Despite its strengths, some limitations of our study deserve note. First, we did not follow the trajectory of BMI to evaluate whether weight gain after kidney transplantation may be a predictor of outcomes. This may have greater relevance for long-term endpoints such as death with graft function and deserves a separate analysis. Second, the single-center nature of this study may limit the generalizability of the results to other settings. However, we examined a large cohort that is likely representative of many Canadian and U.S. transplant centers; thus, we expect our findings to reasonably translate to those settings. Third, despite the large size of the overall cohort, the number of patients in the highest BMI category was relatively small (n=74). This may limit the precision of some of our estimates. Finally, the observational nature of this study may make it susceptible to selection, information, and confounding biases. To mitigate this possibility, we outlined explicit a priori selection criteria for study inclusion/exclusion, used our Comprehensive Renal Transplant Research Information System (CoReTRIS) database to assemble the cohort, applied multivariable modeling strategies to reduce potential confounding, and performed sensitivity analyses to test the robustness of our results.
In summary, increased BMI is an important independent risk factor for adverse posttransplantation outcomes in our study population. The excess risk for DGF appears to start at recipient BMI≥30 kg/m2, whereas the risks of BPAR, all-cause graft failure, or death-censored graft failure were significantly increased at BMI≥35 kg/m2. Further research is needed to elucidate the mechanisms that underlie these associations and to determine if pretransplantation weight reduction may improve outcomes.
MATERIALS AND METHODS
Study Design and Participants
This is an observational cohort study of all eligible adult (≥18 years old) kidney transplant recipients at the Toronto General Hospital transplanted from January 1, 2000 to December 31, 2010 and followed until December 31, 2011. Exclusion criteria included recipients of kidney transplants from an outside institution, receipt of a simultaneous or prior nonkidney transplant, primary nonfunction, or the absence of height or weight at the time of transplantation. During the study period, there were no absolute BMI thresholds used for the selection of potential kidney transplant candidates.
All data for this study were retrieved from our in-center research database, the Comprehensive renal Transplant Research Information System (CoReTRIS). CoReTRIS contains an extensive set of recipient, donor, transplant, laboratory, pathology, treatment, and follow-up data on all patients receiving kidney transplants at the Toronto General Hospital since January 1, 2000. These data have been abstracted from patient charts (electronic and paper), entered into the database, and audited for completeness and accuracy.
All recipients universally received either depleting or nondepleting induction therapy. Maintenance immunosuppression included a calcineurin inhibitor, mycophenolate mofetil, and prednisone. Before 2007, the first-line calcineurin inhibitor was cyclosporine microemulsion with C2 level monitoring. Subsequently, tacrolimus with trough level monitoring became the first-line calcineurin inhibitor. Acute rejections were treated with intravenous corticosteroids, rabbit antithymocyte globulin, intravenous immunoglobulin, plasmapheresis, and/or rituximab. Biopsies were performed for indication, reviewed by a renal pathologist, and classified using the Banff criteria (35).
Exposure and Outcome Classification and Assessment
Recipient BMI (in kg/m2) was calculated from weight and height measured at the time of hospital admission for kidney transplantation. BMI was categorized into five prespecified groups: <20, 20 to 24.9 (reference), 25 to 29.9, 30 to 34.9, and ≥35 kg/m2. The following endpoints were chosen a priori to study the impact of BMI on patient outcomes: (a) DGF (i.e., the need for at least one session of dialysis within the first week after transplantation), (b) BPAR, (c) all-cause graft failure (including death), (d) death-censored graft failure, and (e) death with graft function.
The following potential confounders were examined in multivariable models: (a) recipient factors (age, sex, race, cause of ESRD, peak PRA level, re-graft status, history of diabetes mellitus, history of coronary artery disease, and time on dialysis); (b) donor factors (age, sex, donor type, BMI, expanded-criteria donor status, and donation after cardiac death status); and (c) transplant factors (cold ischemia time, calcineurin inhibitor type at transplantation, number of human leukocyte factor mismatches, and transplant era). BPAR was also entered as a time-varying covariate in graft failure and/or death Cox proportional hazards models. We used the method of multiple imputation to impute missing covariate data (36).
To evaluate the robustness of the primary results, the analyses were repeated in the following three subcohorts: (a) patients whose peak PRA level was zero at the time of transplantation (to reduce the influence of preformed antibodies on the risk of BPAR), (b) patients without DGF (to exclude the influence of DGF on the risk of BPAR, graft loss, or death), and (c) deceased-donor kidney transplants. The impact of including patients with primary nonfunction was examined by attributing them 0.5 days of graft function in the survival analysis. Finally, the lower end of BMI was recategorized as less than 18.5 kg/m2 (instead of <20 kg/m2) to reflect the World Health Organization’s classification scheme.
The distributions of baseline characteristics across BMI categories were evaluated using parametric and nonparametric tests as appropriate. The incidence of DGF was assessed graphically in each BMI group. The relation between BMI and DGF was examined in sequentially nested multivariable logistic regression models. Each successive model adjusted for a larger set of covariates representing recipient, donor, and transplant characteristics.
The cumulative probabilities of time-to-event outcomes were graphically assessed using the Kaplan–Meier method and differences across survival distributions were evaluated using the log-rank test. The risks for BPAR, all-cause graft failure, death-censored graft failure, and death with graft function were evaluated in Cox proportional hazards models, adjusting for potential confounders. Plots of the Schoenfeld residuals and the log (cumulative hazard) functions were constructed to assess the proportional hazards assumption. No important departures were detected.
The Research Ethics Board at the Toronto General Hospital approved this study. All statistical analyses were performed using Stata/MP 12 (StataCorp, College Station, TX). A two-sided P<0.05 was considered statistically significant.
The authors thank Elizabeth Murakami for her excellent administrative support and the students of the Multi-Organ Transplant Student Research Training Program for their dedication and diligence in collecting, entering, and auditing data for the CoReTRIS at the Toronto General Hospital, University Health Network.
1. Collins AJ, Foley RN, Chavers B, et al. United States Renal Data System 2011 annual data report: atlas of chronic kidney disease & end-stage renal disease in the United States. Am J Kidney Dis
2012; 59: A7.
2. Wolfe RA, Ashby VB, Milford EL, et al. Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med
1999; 341: 1725.
3. Jofre R, Lopez-Gomez JM, Moreno F, et al. Changes in quality of life after renal transplantation. Am J Kidney Dis
1998; 32: 93.
4. Glanton CW, Kao TC, Cruess D, et al. Impact of renal transplantation on survival in end-stage renal disease patients with elevated body mass index
. Kidney Int
2003; 63: 647.
5. Johnson DW, Herzig K, Purdie D, et al. A comparison of the effects of dialysis and renal transplantation on the survival of older uremic patients. Transplantation
2000; 69: 794.
6. Khauli RB, Steinmuller DR, Novick AC, et al. A critical look at survival of diabetics with end-stage renal disease. Transplantation versus dialysis therapy. Transplantation
1986; 41: 598.
7. Pelletier SJ, Maraschio MA, Schaubel DE, et al. Survival benefit of kidney and liver transplantation for obese patients on the waiting list. Clin Transpl
8. Pondrom S. The AJT report: news and issues that affect organ and tissue transplantation. Am J Transplant
2012; 12: 1663.
9. Holley JL, Shapiro R, Lopatin WB, et al. Obesity as a risk factor following cadaveric renal transplantation. Transplantation
1990; 49: 387.
10. Johnson DW, Isbel NM, Brown AM, et al. The effect of obesity on renal transplant outcomes
2002; 74: 675.
11. Kuo JH, Wong MS, Perez RV, et al. Renal transplant wound complications in the modern era of obesity. J Surg Res
2011; 173: 216.
12. Lynch RJ, Ranney DN, Shijie C, et al. Obesity, surgical site infection, and outcome following renal transplantation. Ann Surg
2009; 250: 1014.
13. Olarte IG, Hawasli A. Kidney transplant complications and obesity. Am J Surg
2009; 197: 424.
14. Kramer HJ, Saranathan A, Luke A, et al. Increasing body mass index
and obesity in the incident ESRD population. J Am Soc Nephrol
2006; 17: 1453.
15. Friedman AN, Miskulin DC, Rosenberg IH, et al. Demographics and trends in overweight and obesity in patients at time of kidney transplantation
. Am J Kidney Dis
2003; 41: 480.
16. Meier-Kriesche HU, Arndorfer JA, Kaplan B. The impact of body mass index
on renal transplant outcomes
: a significant independent risk factor for graft failure and patient death. Transplantation
2002; 73: 70.
17. Howard RJ, Thai VB, Patton PR, et al. Obesity does not portend a bad outcome for kidney transplant recipients. Transplantation
2002; 73: 53.
18. Modlin CS, Flechner SM, Goormastic M, et al. Should obese patients lose weight before receiving a kidney transplant? Transplantation
1997; 64: 599.
19. Chow KM, Szeto CC, Leung CB, et al. Body mass index
as a predictive factor for long-term renal transplant outcomes
in Asians. Clin Transplant
2006; 20: 582.
20. Chang SH, Coates PT, McDonald SP. Effects of body mass index
at transplant on outcomes
of kidney transplantation
2007; 84: 981.
21. Molnar MZ, Kovesdy CP, Mucsi I, et al. Higher recipient body mass index
is associated with post-transplant delayed kidney graft function. Kidney Int
2011; 80: 218.
22. Gore JL, Pham PT, Danovitch GM, et al. Obesity and outcome following renal transplantation. Am J Transplant
2006; 6: 357.
23. Papalia T, Greco R, Lofaro D, et al. Impact of body mass index
on graft loss in normal and overweight patients: retrospective analysis of 206 renal transplants. Clin Transplant
2010; 24: E241.
24. Akhlaghi F, Trull AK. Distribution of cyclosporin in organ transplant recipients. Clin Pharmacokinet
2002; 41: 615.
25. Shibata N, Hoshino N, Minouchi T, et al. Relationship between area under the concentration versus time curve of cyclosporin A, creatinine clearance, hematocrit value, and other clinical factors in Japanese renal transplant patients. Int J Clin Pharmacol Ther
1998; 36: 202.
26. Zoccali C, Mallamaci F, Tripepi G, et al. Adiponectin, metabolic risk factors, and cardiovascular events among patients with end-stage renal disease. J Am Soc Nephrol
2002; 13: 134.
27. Okamoto Y, Christen T, Shimizu K, et al. Adiponectin inhibits allograft rejection in murine cardiac transplantation. Transplantation
2009; 88: 879.
28. Patel S, Pankewycz O, Kohli R, et al. Obesity in renal transplantation: the role of induction therapy on long-term outcomes
. Transplant Proc
2011; 43: 469.
29. Chagnac A, Weinstein T, Korzets A, et al. Glomerular hemodynamics in severe obesity. Am J Physiol Renal Physiol
2000; 278: F817.
30. Tsuboi N, Utsunomiya Y, Kanzaki G, et al. Low glomerular density with glomerulomegaly in obesity-related glomerulopathy. Clin J Am Soc Nephrol
2012; 7: 735.
31. Teplan V, Schuck O, Racek J, et al. Asymmetric dimethylarginine and adiponectin after renal transplantation: role of obesity. J Ren Nutr
2008; 18: 154.
32. Kielstein JT, Boger RH, Bode-Boger SM, et al. Marked increase of asymmetric dimethylarginine in patients with incipient primary chronic renal disease. J Am Soc Nephrol
2002; 13: 170.
33. Shu KH, Tsai IC, Ho HC, et al. Serum adiponectin levels in renal transplant recipients with and without metabolic syndrome. Transplant Proc
2012; 44: 676.
34. Roos M, Baumann M, Liu D, et al. Low pre-transplant adiponectin multimers are associated with adverse allograft outcomes
in kidney transplant recipients: a 3-year prospective study. Regul Pept
2012; 178: 11.
35. Sis B, Mengel M, Haas M, et al. Banff ‘09 meeting report: antibody mediated graft deterioration and implementation of Banff working groups. Am J Transplant
2010; 10: 464.
36. Sterne JA, White IR, Carlin JB, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ
2009; 338: b2393.