Objectives: Previous studies have linked obesity to poor outcomes in renal transplant recipients, prompting many transplant centers to encourage weight loss pretransplant in obese patients. Here, we performed a single-center retrospective study to assess the effects of weight loss on graft and patient outcomes.
Materials and Methods: Data from 893 renal transplant recipients at our center from 2007 to 2011 were analyzed. First, renal transplant recipients with a history of obesity before transplant (42%) were compared with nonobese patients. Second, in the obese group, renal transplant recipients with significant weight loss (> 10%) before transplant were compared with other obese renal transplant recipients without significant weight loss.
Results: Renal transplant recipients were predom-inantly white, with 74% having undergone living-donor transplant. Obese patients were older (56.6 vs 46.7 y old) and had more comorbidities and more surgical complications, in particular wound complica-tions and incisional hernias, posttransplant than nonobese patients (14.7 vs 5.5%, respectively). Patient and graft survival rates were similar to those in nonobese patients. In the obese group, patient characteristics and medical or surgical complications after transplant did not differ between those with or without significant weight loss. However, obese patient and graft survival rates were lower in patients with weight loss than in obese patients without weight loss.
Conclusions: In our study, weight loss before transplant surgery in obese patients had no influence on surgical outcomes but was associated with a higher mortality rate. A prospective assessment of the impact of weight loss before surgery is needed to establish its usefulness.
Key words : Body mass index, End-stage renal disease, Obesity, Transplantation
Renal transplant is the treatment of choice for patients with end-stage renal disease (ESRD). After adjustments for comorbidities, renal transplant recipients have shown better long-term survival and quality of life than patients on transplant wait lists irrespective of body mass index (BMI).1-5 The prevalence of obesity has steadily increased in the United States,6 leading to an increased incidence of obese patients with stage 5 chronic kidney disease. In the United States, about 30% of patients are now obese at initiation of dialysis.7,8 Only a small proportion of these patients will be offered renal transplant9; yet, despite higher rates of wound complications and delayed graft function,10-12 obese patients have graft and patient survival results similar to those in nonobese patients.12-15
Deceased-donor transplant and, to an even greater extent, living-donor transplant procedures reduce the incidence of mortality in obese graft recipients compared with obese patients on dialysis on wait lists irrespective of BMI.2,4,5 The benefit of being transplanted is only uncertain in black obese patients with BMI > 40 kg/m2.4 Standard recom-mendations have suggested losing weight before transplant, and many transplant centers will rule out candidates with BMI > 30 or 35 kg/m2 because of higher rates of surgical complications.16,17 Doctors therefore ask these patients to lose weight before registering on transplant wait lists. However, weight loss has been described as an independent risk factor for wound complications, although some retrospective studies did not show improved posttransplant survival or graft loss.9,18,19
Here, we aimed to analyze the rates of post-transplant complications (medical and surgical) and their effects on patient and graft survival in a large series of obese patients according to weight changes before transplant.
Materials and Methods
Approval for data collection and analysis was obtained from the Mayo Clinic Institutional Review Board (16-000519). This study included all patients transplanted from January 2007 to December 2011 in the Division of Transplantation Surgery (Mayo Clinic, Rochester, MN, USA) who gave written consent to use of their data for analysis. The following data were extracted from the local database: age at time of transplant, sex, ethnic origin, underlying cause of renal failure, height, body weight at time of transplant and during pretransplant follow-up, comorbidities, date of transplant, organ source (living or deceased), warm and cold ischemia duration, delayed graft function, complications after trans-plant, and patient and graft survival. Patients with missing data for height or weight at time of transplant or patients with less than 2 weight measurements pretransplant were excluded.
Surgical complications included intra-abdominal hematoma, wound complications, ureteral stenosis/-obstruction, lymphocele, graft thrombosis, incisional hernia, renal graft arterial disease, and resurgery (composite complications 1). Medical complications included infections (septicemia, urinary tract infection and other infections; composite complications 2), and other complications (acute myocardial infraction, coronary artery disease, vascular disease, congestive heart failure, posttransplant diabetes, atrial fibril-lation, cancer, graft failure, acute rejection, and post-transplant lymphoproliferative disorder; composite complications 3).
In the first part of the study, we compared data from obese versus nonobese transplant patients. Obesity was defined as a BMI > 30 kg/m2 according to WHO classification.20 The obese patient group (OB group) included both patients who were obese at the time of transplant and those who had been obese during pretransplant follow-up, at any time between the referral for transplant and the transplant itself. This group was compared with patients who had never been obese (NONOB group).
The second part of the study focused specifically on the OB group and analyzed the outcomes of patients who significantly lost weight before trans-plant (WL group) compared with those who did not (nWL group). Weight loss was considered as significant when it was greater than 10% of the first recorded weight available before transplant (Figure 1).
Statistical analyses were performed with Stata 13 software (StataCorp LP, College Station, TX, USA). The tests were 2-sided, with a type I error set at α = 0.05. Baseline characteristics are shown as mean and standard deviation or median and interquartile range for continuous data (assumption of normality assessed by Shapiro-Wilks test) and as number of patients and associated percentages for categorical parameters. Between-group comparisons were performed with the chi-square or the Fisher exact test for categorical variables, which were followed, when appropriate (P < .05), by the Marascuilo procedure. Analysis of variance or a Kruskal-Wallis test was used for quantitative parameters (homoscedasticity verified by Bartlett test) followed by an appropriate post-hoc test (Tukey-Kramer or Dunn test) for multiple comparisons. Regression models were used in multivariate settings to take into account adjustments according to univariate analysis and for clinically relevant parameters. The Kaplan-Meier method was used to estimate censored data. The log-rank test was used in the univariate analysis to compare independent groups. A Cox proportional-hazards regression was used in multivariate settings according to univariate results and clinical relevance (hypertension, diabetes, age at transplant, and heart complication [composite endpoint]). Proportional-hazard hypotheses were verified by Schoenfeld test and plotted residuals. When the duration of pretransplant follow-up of patients on the wait list differed between groups, it was used as an adjustment parameter in multivariate analysis.
For graft survival, death must be considered as a competing cause of failure, and censoring patients who die could bias estimates. For this reason, risk factor analysis, with death as a competing risk, was performed using the Fine and Gray method. A sensitivity analysis was performed of BMI at the time to transplant (BMI > 35 and BMI > 40 kg/m2) and weight loss as continuous value. Because of missing data, delayed graft function was not analyzed.
Patient characteristics at the time of transplant
During the 4-year study period, 907 patients received a kidney transplant and 896 signed informed consent for use of their data. Complete data were available for 893 patients. Patient characteristics are given in Table 1. The OB group represented 42% of the population (n = 380). At the time of transplant surgery, 298 patients still had a BMI > 30 kg/m2: 173 in obese class I (35 > BMI ≥ 30 kg/m2), 89 in obese class II (40 > BMI ≥ 35 kg/m2), and 36 in obese class III (BMI ≥ 40 kg/m2) (Figure 1).20
Characteristics of patients in the OB and NONOB groups are shown in Tables 1 and 2. Patients in the OB group were older and had more comorbidity (diabetes, hypertension, coronary bypass, atrial fibrillation, vascular disease) than those in the NONOB group (Table 1). Warm ischemia time during surgery tended to be longer in the OB group (Table 2). Most patients received a kidney transplant from a living donor.
Posttransplant outcomes in the OB and NONOB groups
The warm ischemic time during surgery tended to be longer in the OB group, but the rate of delayed graft function did not differ between the groups. Surgical complications, especially wound complications and incisional hernia, were more frequent in the OB than in the NONOB group. Patient and death-censored graft survival rates were similar in the 2 groups (Table 2 and Figure 2).
In the OB group, 295 patients still had a BMI > 30 kg/m2 at the time of transplant. Compared with those in the NONOB group, patients in the OB group who had a BMI > 30 kg/m2 at time of transplant surgery had a longer warm ischemia time (47 ± 21 vs 44±16 min; P < .01) and higher rates of surgical complications (14% vs 9%; P < .01), particularly wound complications and hernia (7% vs 2% and 4.1% vs 0.4%, respectively; P < .01). Comparative analysis of patients in the OB group who had BMI > 30 kg/m2 at time of transplant and those in the OB group with BMI < 30 kg/m2 at time of transplant showed no significant differences in posttransplant com-plications. Likewise, patient and graft survival rates were not significantly different, with patient survival showing P = .82 in univariate analysis (hazard ratio [HR] = 0.83; 95% confidence interval [CI], 0.40-1.71) and P = .61 in multivariate analysis and graft survival showing P = .90 in univariate analysis (HR = 0.74; 95% CI, 0.18-3.09) and P = .68 in multivariate analysis (Table 3 and Table 4). The sensitivity analyses for weight loss as continuous parameter and for different thresholds for BMI at time to transplant confirmed these results.
Effect of weight loss > 10% before transplant in the OB group
Of the 380 patients in the OB group, 78 had a weight loss of > 10% pretransplant (WL group, 20%) (Figure 1). Table 1 shows patient characteristics, which did not differ between WL and nWL groups, except for a greater proportion of African Americans in the WL group (Table 1). Patients in the WL group had a lower BMI than patients in the nWL group at the time of transplant (33.6 ± 4.4 kg/m2 vs 36.3 ± 5.7 kg/m2; P < .001). Rates of surgical and medical complications were similar in the WL and nWL groups (Table 2). In contrast, patient and graft survival outcomes were lower in the WL group (Figure 3). Patients in the WL group had a higher risk of death after transplant (multivariate analysis HR = 4.24; 95% CI, 1.32-13.6; P = .01) and no statistical difference for organ failure (HR = 2.68; 95% CI, 0.95-7.54; P = .06 for univariate analysis and HR = 1.14; 95% CI, 0.29-4.54; P = .86 for multivariate analysis) compared with patients in the nWL group. Sensitivity analyses with different threshold for BMI at time of transplant and weight loss as continuous parameter confirmed these results.
In the nWL group, 53 patients (18%) had a BMI < 30 kg/m2 at the time of transplant surgery, owing to a slight loss of weight (< 10% of initial weight). In the WL group, 32 obese patients (41%) had a BMI < 30 kg/m2 at the time of transplant.
Effect of body mass index in patients in the OB group at time of transplant
Comparisons of patients with BMI < 30 versus ≥ 30 kg/m2 at the time of transplant are summarized in Tables 3 and 4. Patient characteristics and transplant characteristics did not differ, except for warm ischemia time in obese patients in the nWL group with BMI > 30 kg/m2 at time of transplant, which tended to be longer (45 min [range, 36-51 min] vs 40.5 min [range, 31-49.5 min] vs 41 min [range, 36-50 min] vs 41 min [range, 38-51 min] in the OB nWL group with BMI > 30 kg/m2 vs the OB nWL group with BMI < 30 kg/m2 vs the OB WL group with BMI > 30 kg/m2 vs the OB WL group with BMI < 30 kg/m2, respectively; P = .08) (Table 4). Similarly, having a BMI < 30 or > 30 kg/m2 at time of transplant did not influence posttransplant outcomes for patients in the OB group. A sensitivity analysis was performed of BMI at time to transplant threshold (class II BMI > 35 and class III BMI > 40 kg/m2). Although statistical power was lower, results were analogous and the same conclusions were drawn.
Of note, 6.5% of patients in the WL group versus 7.2% of patients in the nWL group and 16.6% of patients in the WL group versus 22% of patients in the nWL group gained more than 10% or 5% weight, respectively (not significant) in the year after transplant.
The novelty of this work is the finding that (1) patients with a history of obesity who had lost weight pretransplant and achieved for a part of them a BMI < 30 kg/m2 still had a greater risk of wound complications than nonobese patients and (2) weight loss pretransplant is probably detrimental to patient survival after adjustment for confounding factors.
It is well known that kidney transplant in obese individuals entails increased risk of postoperative wound complications.11 Our results confirmed that kidney transplant in obese patients entailed increased risk of postoperative morbidity, especially wound complications, even if patients had lost weight before transplant. However, as reported elsewhere, we found that patient and graft survival outcomes were not different in obese versus nonobese patients after adjustment for comor-bidities.10,12-15 In contrast, our study suggests that weight loss before transplant may be detrimental to patient survival after adjustment for confounding factors.
In contrast, 2 meta-analyses reported a higher long-term mortality in undernourished and obese patients than in normal weight patients.11,21 These findings advise caution in the management of obese patients: transplant candidates should be strictly selected according to their comorbidities and age and undergo close cardiovascular monitoring.15,22,23 However, renal transplant can reduce mortality compared with dialysis in obese patients even when their BMI is > 40 kg/m2.2,4,5,13
An important finding of this study is that patients with a history of obesity who had lost weight pretransplant and achieved a BMI < 30 kg/m2 still had a greater risk of wound complications than nonobese patients. It has been reported that weight loss of > 10 kg is an independent risk factor for the development of wound complications.10,18 This could be explained by the persistently large abdominal cutaneous ring or panniculus, which complicates surgical access and leads to more infections or necrosis. The prevention of postoperative com-plications in obese patients could be improved by limiting the spreading force, ensuring meticulous hemostasis, and performing muscular reconstruction.24 Panniculectomy can be performed before transplant,25 and the efficacy of robot-assisted transplant needs to be fully assessed.26,27
In our study, in contrast to other documented reports, the incidence of delayed graft function did not differ between patients in the OB and NONOB groups. Delayed graft function has been observed in obese patients because operation duration and warm ischemia time are often longer. In our study, warm ischemia time during surgery tended to be longer in the OB group and was longer in recipients with BMI > 30 kg/m2 at transplant. Almost all patients received kidney transplants from living donors with a short cold ischemia time, which could explain the low incidence of delayed graft function.
Our study suggested that weight loss before transplant is detrimental to patient survival after adjustment for confounding factors (hypertension, diabetes, age at transplant, and heart complications). Weight loss may lead to the creation of a sub-population of patients who develop a state of protein malnutrition and heightened systemic inflammation.10,28 The nutritional status of obese patients at the time of transplant is a key issue. Unfortunately, nutritional assessment pretransplant was not available for patients in our study. Protein energy wasting and sarcopenic obesity can occur in overweight patients with ESRD and can be aggravated by weight loss strategies.29-31 Graft and patient survival outcomes decrease when muscle mass (assessed by predialysis creatinine) is low.28 Low pretransplant serum albumin levels and high malnutrition inflammation scores have been shown to be associated with higher mortality and graft loss after kidney transplant.30-32 The nutritional and inflammation status of obese patients engaged in a weight loss program should be rigorously monitored. In another retrospective analysis of patients on wait lists and transplant recipients, the decrease in body weight of those on the wait list was not protective for posttransplant mortality or graft loss.18 In our study, the higher mortality rate observed in patients in the WL group may have been due to infectious diseases (composite infectious com-plications of 26.9% vs 23.5% in the WL vs nWL group). The combination of wound complications and poor nutritional status could explain this higher rate.
Substantial weight loss pretransplant has been associated with rapid weight gain posttransplant.18 In our study, no differences in posttransplant weight gain were observed between patients in the WL and nWL groups; therefore, weight gain could not explain the worse outcomes observed in the WL group.
Weight loss in obese ESRD patients is notoriously difficult, and only a few patients achieve a BMI < 30 kg/m2.2,26 Moreover, difficulty in losing weight can significantly increase wait time on dialysis, which is the strongest modifiable risk factor for renal transplant outcomes.9,33 The use of bariatric surgery before or after transplant to achieve weight loss needs to be further evaluated.26,34
Fat mass distribution can influence mortality in ESRD patients more than obesity per se. Prospective studies are needed to assess the significance of waist circumference before and after transplant and to guide transplant candidate selection.35
Our study has a number of limitations. First, the study design was retrospective. In particular, infor-mation on patient nutritional status was lacking, and we do not know whether weight loss was intentional or not. Similarly, we did not know whether weight loss was or was not accompanied by protein energy wasting at the time of transplant. Another limitation concerns the cohort, which was overwhelmingly composed of individuals with white ethnicity and predominantly composed of living-donor transplant recipients, which makes it difficult to generalize any conclusions to a larger population. The wait time between referral and transplant was probably much shorter for this cohort than it would have been for most patients waiting for a transplant from a deceased donor. Better outcomes with grafts from living donors than from deceased donors could have masked some of the effects of weight loss. Finally, some important pretransplant factors were missing, including dialysis modality, dialysis vintage, and preemptive transplant. These factors have been shown to influence transplant outcomes.
This study, conducted in a predominantly white population of living-donor transplant recipients, confirms that adjusted patient and graft survival outcomes are similar in obese and nonobese patients despite a higher rate of surgical complications in the former. Losing weight did not protect against postoperative complications and may even be detrimental to survival.
Prospective studies are needed to assess the effects of weight loss on transplant outcomes and should include nutritional parameters and abdominal girth and body composition follow-up to better define the determinants of posttransplant events. Another challenge is to look for possible ways to obtain a loss of body weight without compromising postoperative outcomes. In this respect, the value of exercise training, which could maintain lean body mass and body functioning, and early bariatric surgery should be investigated.26,34,36
DOI : 10.6002/ect.2018.0124
From the 1Centre Hospitalo-Universitaire Clermont-Ferrand, Nephrology
Department, Clermont-Ferrand, France; the 2Clermont University, University
d'Auvergne, Clermont-Ferrand, France; the 3UMR 1019, INRA, UMR 1019, UNH, CRNH
Auvergne, Clermont-Ferrand Cedex 01, France; the 4Centre Hospitalo-Universitaire
Clermont-Ferrand, Biostatistics Unit, Délégation Recherche Clinique et
Innovation, Clermont-Ferrand, France; and the 5The William J. von Liebig
Transplant Center, Division of Transplantation Surgery, Mayo Clinic, Rochester,
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare .The authors thank Nicolas Michaud (Novartis Laboratory) and Jeff Watts for their help.
Corresponding author: Anne-Elisabeth Heng, Service de Néphrologie, Hôpital Gabriel Montpied, 58 rue Montalembert, F-63003 Clermont-Ferrand, France
Phone: +33 610844495
Figure 1. Patient Selection Tree
Figure 2. Patient (A) and Graft (B) Survival in Obese and Nonobese Groups
Figure 3. Patient (A) and Graft (B) Survival in Obese Group: Comparison of No Weight Loss Versus Weight Loss Patients
Table 1. Characteristics of Kidney Transplant Recipients With and Without History of Obesity
Table 2. Outcomes in Kidney Transplant Recipients With and Without History of Obesity
Table 3. Characteristics of Obese Kidney Transplant Recipients With or Without Significant Weight Loss Before Transplant According to Body Mass Index < 30 or > 30 kg/m2 at Time of Transplant
Table 4. Outcome in Obese Kidney Transplant Recipients With or Without Significant Weight Loss Before Transplant According to Body Mass Index < or > 30 kg/m2 at Time of Transplant