- Email: [email protected]
Posttransplantation Diabetes Mellitus: A Long-Term Retrospective Cohort Study
Posttransplantation Diabetes Mellitus: A Long-Term Retrospective Cohort Study M. Allamani, J. Sennesael, and E. Vendemeulenbroucke ABSTRACT Background. Posttransplantation diabetes mellitus (PTDM) is a seious complication of transplantation which is caused by immunosuppressive drugs and adversely affects the survival of the transplant recipient and the long-term survival of the graft. In this study, we assessed the incidence of PTDM and the factors associated with its development during long-term follow-up of renal transplant recipients. We also investigated the influence of PTDM on the cardiovascular risk (CVR) profile. Methods. We retrospectively reviewed the records of the patients who underwent renal transplantation at our center between 1986 and January 2007. Diabetes was diagnosed according to American Diabetes Association criteria. The CVR factors were analyzed at the time of transplantation as well as at 1 and 3 years follow-up. Results. We included 136 nondiabetic transplant recipients. The PTDM incidences at 1, 3, 5, and 10 years were 9%, 12%, 13%, and 16.4%, respectively. The cumulative incidence during follow-up was 17.6%. On both univariate and multivariate analyses body mass index (BMI) was significantly associated with PTDM. Patients with BMI 25 ⫽ 30 kg/m2 had an odds ratio [OR] of 3.53 (95% confidencial interval [CI] 1.26 –9.90; P ⫽ .017) and those with BMI ⬎30 kg/m2 had an OR of 4.58 (95% CI 1.4 –14.01; P ⫽ .012). There were no significant differences in gender distribution, age, pretransplant dialysis period, acute rejection rate, or immunosuppressive regimens between patients with (n ⫽ 24) versus without (n ⫽ 112) PTDM. Conclusion. The risk of PTDM increased continuously with time after transplantation. BMI was an independent predictor of PTDM. Among all of the risk factors for PTDM, obesity is the only modifiable risk factor before transplantation. PTDM was associated with a worse traditional CVR profile; a better control of CVR factors should be performed to prevent long-term morbidity and mortality in this population. iabetes mellitus (DM) is a metabolic disorder in 1%–2% of the normal population. Posttransplantation diabetes mellitus (PTDM) is a common complication, although its incidence varies depending on the criteria used for its definition. PTDM has been deined as a fasting plasma glucose of ⬎140 mg/dL,1 as the need for treatment either with oral hypoglycaemic agents or insulin,2 or as the need for treatment with insulin for more than 30 days,3 following the American Diabetes Association (ADA) criteria,4 or based upon an oral glucose tolerance test (OGTT).5,6 Although most patients develop diabetes mellitus in the first 3 months after transplantation, the incidence of PTDM increases with follow-up.2,4 The incidence of PTDM across various studies varies between 2% and
D
53%.2,4,5,7,8 Patient age, nonwhite ethnicity, high body mass index (BMI), high doses of glucocorticoids, and immunosuppression with high doses of cyclosporine (CsA) or tacrolimus have been identified as risk factors for the development of PTDM.1–3,9 –11 The diabetogenic potential of tacrolimus is greater than that of CsA.4,12,13 PTDM is a form of type 2 DM, thought to develop in response to relative insulin deficiency resulting from inFrom the Departments of Internal Medicine (M.A., E.V.) and Nephrology (J.S.), University Hospital Brussels, Brussels, Belgium. Address reprint requests to Jacques Sennesael, Department of Nephrology, UZ Brussel, Laarbeeklaan 101, 1090 Brussels, Belgium. E-mail: [email protected]
0041-1345/10/$–see front matter doi:10.1016/j.transproceed.2010.07.009
© 2010 Published by Elsevier Inc. 360 Park Avenue South, New York, NY 10010-1710
4378
Transplantation Proceedings, 42, 4378 – 4383 (2010)
POSTTRANSPLANTATION DIABETES
creased resistance to impaired production of insulin, or from a combination of both.8,12,14 Prevention of PTDM is an important goal in renal transplantation. Uncontrolled diabetes is associated with increased cardiovascular morbidity and mortality, which increase the risk of death with a functioning graft.8,15 Compared with the general population, kidney transplant recipients already have a higher risk of death from cardiovascular (CV) disease, because of the high prevalence of cardiovascular risk (CVR) in dialysis patients together with the adverse effects of immunosuppressive agents.15 CsA is associated with higher blood pressure and lipid levels than tacrolimus. The present study sought to determine the incidence of PTDM among kidney transplant recipients by defining diabetes according to the ADA criteria.16 We sought to identify risk factors for the development of PTDM and analyzed the evolution of the CVR factors during the first 3 years after transplantation. METHODS Patient Selection For this retrospective study we analyzed the files of 200 tansplanted patients between 1986 and December 31, 2006. We excuded all patients with pretransplantation diabetes (n ⫽ 38) as well as those with ⬎1 year of follow-up (n ⫽ 21), or those for whom data were not fully available (n ⫽ 4). Also 1 patient who underwent a partial pancreatectomy after transplantation, consequently developing diabetes was excluded for this analysis. Thus, 136 patients were included in the study.
Diagnosis of Diabetes The diagnosis of PTDM was made according to the criteria of the ADA.16 In 1 patient, diabetes was diagnosed using an oral glucose tolerance test (2-hour post– glucose load glycemia ⱖ200 mg/dL). In all of the other patients (n ⫽ 23), the diagnosis was based on increased fasting (ⱖ126 mg/dL) or random glycemia (ⱖ200 mg/ dL). All transplant recipients who developed PTDM were tested for antibodies to exclude type 1 DM.
4379 Acute rejection episodes were treated with intravenous 1-g methylprednisolone pulses for 3 days. Because mainly corticoids, CsA, tacrolimus, and sirolimus are known to be associated with PTDM, we restricted our analysis to these 4 immunosuppressive drugs and did not take into account the type of antiproliferative drug.
Statistics All statistical tests were performed in 2-tailed fashion using SPSS for Windows 16.0 statistical software package (SPSS, Chicago, IL, USA) and considered to be significant when P ⬍ .05. Differences between groups were assessed by Mann-Whitney U tests for continuous variables and by chi-square tests using Fisher exact test for categoric variables. We investigated differences between diabetesfree survivals and the incidence of diabetes by Kaplan-Meier analysis and log-rank tests. Cox regression analysis was used to determine risk factors for diabetes development.
RESULTS General Characteristics of Patients at Transplantation
We included 136 transplant patients in this study their baseline characteristics are presented in Table 1. The mean ⫾ SD age of the patients at transplantation was 51 ⫾ 12 years, and the mean follow-up time after transplantation was 88 ⫾ 59 months. The youngest patient was 15 years old and the oldest 80 years old. Similar numbers of patients received kidney transplantations because of end-stage renal disease due to glomerulonephritis (n ⫽ 34; 25%), or nephrangiosclerosis (n ⫽ 30; 22%), interstitial nephropathy (n ⫽ 32; 24%), or congenital kidney disease (n ⫽ 30; 22%). In 10 patients (7%), the etiology of renal failure was unknown. Mean duration of dialysis before transplantation was 26.3 ⫾ 21 months (median, 20.7 mo). Two patients did not undergo dialysis before transplantation. Description of Immunosuppressive Regimens
At the time of transplantation all of the patients were treated with corticoids. The mean maintenance dose at the
Data Collection and Analysis The following recipient variables were evaluated at the time of transplantation: age, gender, ethnic background, etiology of renal failure, and duration of dialysis. Because no information was available about recipient family histories of DM that variable was not considered in this study. At transplantation as well as after 1, 3, 5, and 10 years, we recorded the following parameters: body weight, height, and acute rejection episodes. Blood pressure and lipid profiles were analyzed at the time of transplantation as well as 1 and 3 years thereafter.
Immunosuppression Induction immunosuppression consisted of corticoids, a calcineurin inhibitor, and imuran or mycophenolate mofetil (MMF). Before 1986, all kidney transplant recipients received CsA as the calcineurin inhibitor in association with imuran and corticoids. From 1996 onwards, the immunosuppression consisted of CsA or tacrolimus or sirolimus in combination with MMF plus corticoids. Tacrolimus doses were adjusted to maintain drug levels of 8 –12 ng/dL, and CsA to 100 –150 ng/mL.
Table 1. General Characteristics of Included Patients at Transplantation Number of patients Age BMI, kg/m2 Male/female ratio Ethnicity, n (%) Caucasian North African Asian Renal pathology, n (%) Nephrangiosclerosis interstitial nephropathy Congenital kidney disease Glomerulonephritis Unknown Duration of dialysis, mo Follow-up, mo Age and BMI are expressed as mean ⫾ SD.
136 51 ⫾ 12 24.9 ⫾ 4.2 74/62 127 (93) 8 (6) 1 (1) 30 (22) 32 (24) 30 (22) 34 (25) 10 (7) 26 ⫾ 21 88 ⫾ 59
4380
ALLAMANI, SENNESAEL AND VANDEMEULENBROUCKE
moment of transplantation was 24.6 ⫾ 4.6 mg; (n ⫽ 136); the mean dose at 1 year after transplantation was 7.3 ⫾ 2.9 mg (n ⫽ 136). At 3, 5, and 10 years the number of patients treated with corticoids were 110/112, 91/92, and 11/51, respectively. The mean dose was tapered to 6.l ⫾ 2.7, 5.2 ⫾ 2.2, and 3.3 ⫾ 2.5 mg, respectively. Ten years after transplantation, 50% of PTDM patients and 82% of patients without PTDM were still prescribed corticoids. At transplantation, 123 (90.4%) patients received CsA with a mean dose of 8.7 ⫾ 2.5 mg/kg, and the average (SD) doses at l, 3, 5, and 10 years were 3.5 ⫾ 1.6, 2 ⫾ 1.6, 2.8 ⫾ 0.9, and 2 ⫾ 1.4 mg/kg, respectively. The numbers of patients receiving CsA were 98 (72%), 79 (70.5%), 65 (70%), and 35 (69%) after 3, 5 and 10 years, respectively. Eleven transplant recipients (8%) received tacrolimus at transplantation at 1, 3, 5, and 10 years, 20 (15%), 15 (13%), 8 (7%), and 8 (16%) patients were on tacrolimus, respectively. At the time of transplantation, 61 recipients (45%) received MMF and 73 (54%) imuran. Incidence of PTDM
Twenty-four (17.6%) recipients developed PTDM. The diagnosis of PTDM was established between 10 days and 11 years after transplantation; the median interquartile range was 12 months (1.4 – 61.1). One-half of the patients (n ⫽ 12) developed PTDM within the first year after transplantation, 16 within 3 years, and 19 within 5 years after transplantation. Figure 1 shows the incidence of PTDM over the follow-up. General characteristics for patients with PTDM are presented in Table 2. Mean age, gender, dialysis duration, and etiology of renal failure were similar between those subjects with PTDM and those without. The mean body mass index (BMI) at transplantation in recipients with PTDM was significantly higher than in those without
Fig 1.
Cumulative incidence of PTDM (Kaplan-Meier plot).
Table 2. Comparison of the General Characteristics at Transplantation for the Patients Who Developed Diabetes after Transplantation (PTDMⴙ) Compared with Those Who Did Not (PTDMⴚ) PTDM⫹
Number of patients Age, y BMI, kg/m2 Male/female ratio Ethnicity, n (%) Caucasian North African Asian Renal pathology, n (%) Nephrangiosclerosis Interstitial nephropathy Congenital kidney disease Glomerulonephritis Unknown
24 51 ⫾ 8 26.7 ⫾ 3.3 14/10
PTDM⫺
112 51 ⫾ 13 24.5 ⫾ 4.3 60/52
21 (88) 3 (12) 0 (0)
106 (95) 5 (4) 1 (1)
8 (33) 1 (4) 5 (21) 8 (33) 2 (8)
22 (20) 31 (28) 25 (22) 26 (23) 8 (7)
P Value
.954 .006 .67 .288
.126
Age and BMI are expressed as mean ⫾ SD.
PTDM [26.7 ⫾ 3.3 vs 24.5 ⫾ 4.3; P ⫽ .006]. There were no differences in gender or ethnicity distribution between patients who developed PTDM and those who did not. Risk Factors for Developing PTDM
Age. The mean age of patients with PTDM at transplantation was 51 ⫾ 12 years (Table 2), which was similar to the mean age of patients without PTDM (51.8 ⫾ 13 years). The median age of all patients was 50 years. Patients were divided into 2 groups: ⱖ50 years versus ⬍50 years old. We did not find a significant difference in the development of PTDM among patients ⱖ50 years old compared with those ⬍50 years old (log-rank: P ⫽ .2). BMI. Mean BMI at the time of transplantation was significantly greater among patients with PTDM compared with without patients (26.7 ⫾ 3.3 vs 24.5 ⫾ 4.3; P ⫽ .006). The study population was arbitarily subdivided according to BMI into 3 groups: ⬍25 kg/m2; 25–30 kg/m2 and ⬎30 kg/m2. Figure 2 shows the Kaplan-Meier plot of the incidence of PTDM and diabetes-free survival among these recipients, showing a significant difference (log-rank: P ⫽ .013). We repeated this analysis using 2 groups of patients, BMI ⬍25 kg/m2 versus ⱖ25 kg/m2, showing again a significant difference (log-rank: P ⫽ .004). Duration of dialysis. Patients were than subdivided into 2 groups according to their median time on dialysis (21 months): ⱖ21 versus ⬍21 months. The mean time on dialysis was similar between patients with and without PTDM (27 ⫾ 21.5 vs 23 ⫾ 19 months; log rank: P ⫽ .4). There was no significant difference in the incidence of PTDM between the groups. Immunosuppression. Corticosteroids. All patients received similar doses of corticoids for induction at transplantation. However in cases of acute rejection episodes (ARE), we administered high doses of corticoids. To determine the contribution of this immunosuppressive drug to the devel-
POSTTRANSPLANTATION DIABETES
4381
parameter that was withheld in the analysis was BMI. Compared with recipients with BMI ⬍25 kg/m2, the odds ratio [OR] for recipients with BMI 25–30 kg/m2 was 3.53 (95% CI 1.26 –9.90; P ⫽ .017). Compared with patients with BMI ⬍25 kg/m2, patients with BMI ⬎30 kg/m2 showed an OR of 4.58 (95% CI 1.4 –14.01; P ⫽ .012). The interaction of age and BMI and the combination of ARE and immunosuppression with tacrolimus did not influence the risk for developing PTDM. Differences in CV Events and CVR Factors between the Two Groups during Follow-Up
Fig 2. Incidence of PTDM according to BMI.
opment of PTDM, we compared the occurrence of ARE as a surrogate marker for total corticosteroid burden. Among the group of patients with PTDM, 21% of recipients developed ARE, as did 26% without diabetes (P ⫽ .6). We also examined the effect of the combination of hig-dose corticoids for ARE and tacrolimus on the incidence of PTDM, showing no significant influence. Calcineurin inhibitors. At the time of transplantation, 92% of the patients received CsA (92% in non-PTDM group vs 91.7% in PTDM group) and 8% tacrolimus (8% vs 8.3%; P ⫽ .96 by Mann- Whitney). The mean daily dose of both calcineurin inhibitors was decreased over the follow-up period: 0.19 to 0.057 mg/kg and 8.7 to 2 mg/kg for tacrolimus and CsA, respectively. Among the patients initially treated with CsA, 22 developed PTDM as did 2 in the tacrolimus group. Multivariate analysis of risk factors for the development of PTDM. Using demographic and immunosuppressive factors as independent variables we performed a multivariate analysis of potential risk factors for the development of PTDM; namely, age, gender, BMI at transplantation, time on dialysis, type of calcineurin inhibitor, and ARE. The only
PTDM patients developed more CV events than patients without PTDM (n ⫽ 7 with vs 2 without PTDM). Among these patients, 5 with diabetes and 1 without diabetes needed surgical treatment (P ⫽ .03). Classic CVR factors, such as blood pressure (BP) and dyslipidemia, were different between groups. Although systolic BP was similar at and 1 year after transplantation, it was significantly higher after 3 years in the PTDM group: 151 mm Hg vs 139 mm Hg in the non-PTDM group (P ⫽ .006). At 3 years after transplantation, diastolic blood pressure was also higher in this group: 91% mm Hg vs 84 mm Hg; P ⫽ .03). There were no differences in total cholesterol levels. In contrast, we observe significantly higher tiglyceride (TG) levels at the time of (214 mg/dL vs 168 mg/dL; P ⫽ .006) and 3 years after transplantation (218 mg/dL in PTDM vs 165 mg/dL in non-PTDM patients; P ⫽ .013). Table 3 shows the CVR factors in the 2 groups. DISCUSSION
PTDM is an important cause of morbidity and mortality among renal transplant recipients, with an incidence ranging from 2% to 53%.2,4,5,7,8 This retrospective study of 136 patients who underwent renal transplantation at our center showed a 17.6% cumulative incidence of PTDM. The main risk factor for the development of PTDM in our population was BMI at transplantation. No relationship was noted between PTDM development and other demographic characteristics or the type of immunosuppressive regimen. The wide range of incidences of PTDM described in the literature can be explained by various factors. First, the citeria to diagnose diabetes varied among the studies.1– 6 In the present study, diabetes was diagnosed according to ADA criteria,16 using fasting or random glycemia, a method
Table 3. CVR Factors in the PTDMⴙ and PTDMⴚ Groups At Transplantation
1 y after Transplantation
3 y after Transplantation
CVR Factor
PTDM⫹
PTDM⫺
P Value
PTDM⫹
PTDM⫺
P Values
PTDM⫹
PTDM⫺
P Values
Systolic BF (mm Hg) Diastolic BP (mm Hg) Total cholesterol (mg/dL) Triglycerides (mg/dL)
14.3 ⫾ 14 83 ⫾ 11 207 ⫾ 47 214 ⫾ 91
145 ⫾ 21 85 ⫾ 14 185 ⫾ 50 168 ⫾ 95
ns ns ns ns
145 ⫾ 14 85 ⫾ 10 204 ⫾ 37 166 ⫾ 14
137 ⫾ 19 83 ⫾ 12 230 ⫾ 48 174 ⫾ 79
ns ns ns .006
151 ⫾ 13 91 ⫾ 13 240 ⫾ 53 218 ⫾ 95
139 ⫾ 19 84 ⫾ 12 230 ⫾ 49 165 ⫾ 66
.006 .003 ns .013
The values are expressed as mean ⫾ SD. Abbreviation: BP, blood pressure.
4382
that was consistent with some earlier studies.1– 4 In a recent work, however, it was concluded that fasting blood glucose might not be the optimal screening test for the diagnosis of PTDM in kidney transplant recipients, that the OGTT was more accurate and reproducible.5 At our institution, an OGTT was implemented in 2006 as part of the yearly follow-up of transplant recipients, but in the present study, only 1 patient was diagnosed by an OGTT. For this reason we concluded that we may have underestimated the incidence of PTDM. Long-term follow-up of patients transplanted since 2006, and thus submitted yearly to an OGTT, showed clarify this issue in the coming years. Furthermore, the follow-up times after transplantation varied considerably among studies2,3,4,7,15,17,18; it was often limited to 2 years.4,15 As shown herein, the risk of PTDM was the highest during the first year, but it can appear up to 11 years later. Different immunosuppressive regimens also may explain variations in the incidences of PTDM.2,7,18 The introduction of calcineurin inhibitors has reduced the incidence of PTDM by lowering the doses of corticoids. This beneicial effect is somewhat counterbalanced by the intrinsic diabetogenic effect of these agents, especially tacrolimus.1,3,4,7,11,13,14 Finally, the diversity in patient demographics among studies may influence the reported incidence of PTDM. Older age and African-American ethnicity have been associated with higher incidence of PTDM.2,7,12 In the present study, we also analyzed risk factors for the development of PTDM. Upon univariate and multivariate analysis, only BMI at transplantation significantly correlated with diabetes development. Patients with an initial BMI ⬎25 kg/m2 showed an OR of 3.53 (P ⫽ .017), and recipients with BMI ⬎30 kg/m2 an OR of 4.58 (P ⫽ .012) compared with lean recipients (BMI ⬍25 kg/m2), a finding consistent with earlier observations in transplant recipients.9,17,18 Weight gain is a common problem after transplantation. In the present study, we observed weight gain in both groups of patients, but those with PTDM gained significantly more weight during the first year after transplantation (4.5 ⫾ 6.7 kg) than those without PTDM (1.8 ⫾ 7 kg; P ⫽ .04). Obese patients may be more insulin resistant, which could, in combination with the immunosuppressive treatment, lead to a relative insulin deficiency and with increased blood glucose to the development of PTDM. However, we did not have data on insulin resistance in our study population, because models used to compute insulin resistance have not been validated in end-stage renal disease patients.19 Elderly recipients, especially those ⬎45 years old have been reported to be at greater risk for PTDM.2,11 Perhaps, because of the small number of patients in the present study, we could not identify this variable as an independent risk factor for PTDM. Neither could we assess ethnicity, probably because of small number of noncaucasians among our study population. In most studies, tacrolimus has been more frequently associated with PTDM than CsA; however, we did not note a significant difference between the cal-
ALLAMANI, SENNESAEL AND VANDEMEULENBROUCKE
cineurin inhibitors, possible due to the small number of patients on tacrolimus. Our study has some limitations. It was retrospective and did not include a large number of cases, a limitation in all single-center experiences with PTDM.6,7,11,18 OGTT was not routinely performed as a screening test for kidney transplant recipients until 2006, but more recently transplanted patients are undergoing this test yearly. Their future follow-up may show whether this test has an important impact on the observed incidence of PTDM. Unfortunately, we did not have data on the family histories of diabetes among transplant recipients, a major risk factor for diabetes development. The strengths of this study are the long-term follow-up after transplantation with few patients lost. Diabetes diagnosis was more robust compared with other studies on PTDM.1– 4,7,18 This single-center study had a uniform immunosuppressive treatment and homogeneous follow-up. Overweight is associated, in the general population with some traditional CVR factors, such as dyslipidemia, hypertension, and insulin resistance. Likewise, in renal transplant patients, obesity has been associated with a worse CVR profile.15,20 Dyslipidemia is another CVR factor highly prevalent among kidney transplant recipients.8,15 We observed higher TG levels among diabetic patients, but no significant difference in total cholesterol levels. Although a greater number of patients with PTDM received lipidlowering medications, it has been suggested that more aggressive therapy would probably reduce the risk of CV disease in this population. Hypertension is another traditional CVR factor frequently observed in transplant patients. The mean blood pressure was significantly higher among the PTDM group in the present study. In general, no antihypertensive agent is contraindicated in renal transplant; this risk factor needs to be controlled rigorously. We concluded that PTDM is a serious complication of renal transplantation, with an incidence of 17.6% over 11 years. Most cases occur during the first year after transplantation. Obesity was the only independent predictor of PTDM. However, it is a modifiable risk factor; obese patients should be counseled to lose weight before transplantation. Furthermore, prevention of weight gain after transplantation may protect patients from the development of PTDM. Although the immunosuppressive regimen with the least diabetogenic potential should be used, this issue may be less important among nonobese patients. PTDM was associated with a worse CVR profile. Better control of modifiable risk factors, such as lipid profile and blood pressure, may improve the prognoses of these patients.
REFERENCES 1. Weir MR, Fink JC: Risk for posttransplant diabetes mellitus with current immunosuppressive medications. Am J Kidney Dis 34:1, 1999
POSTTRANSPLANTATION DIABETES 2. Cosio FG, Pesavento TE, Osei K, et al: Post-transplant diabetes mellitus: increasing incidence in renal allograft recipients transplanted in recent years. Kidney Int 59:732, 2001 3. Margreiter R: Eficacy and safety of tacrolimus compared with cyclosporine microemulsion in renal transplantion: a randomised multicenter study. Lancet 359:741, 2002 4. Marcen R, Morales JM, del Castilo D, et al: Posttransplant diabetes mellitus in renal allograft recipients: a prospective multicenter study at two years. Transplant Proc 38:3530, 2006 5. Armstrong KA, Prins JB, Beller EM, et al: Should an oral glucose tolerance test be performed routinely in all renal transplant recipients? Clin J Am Soc Nephrol 1:100, 2006 6. Hur KY, Kim MS, Kim YS, et al: Risk factors associated with the onset and progression of posttransplantation diabetes in renal allograft recipients: Diabetes Care 30:609, 2007 7. Numakura K, Satoh S, Tschiya N, et al: Incidence and risk factors of clinical characteristics, tacrolimus pharmacocinetics, and related genomic polymorphisms for posttransplant diabetes mellitus in early stage of renal transplant recipients. Transplant Proc 37:1865, 2005 8. Ravidran V, Baboolal K, Moore R. Posttransplant diabetes mellitus after renal transplantation: The emerging clinical challenge. Yonsei Med J 45:1059, 2004 9. Kasiske BL, Synder JJ, Gilbertson D, et al: Diabetes mellitus after kidney transplantation in the United States. Am J Transplant 3:178, 2003 10. First MR, Gerber DA, Hariharan S, et al: Posttransplant diabetes mellitus in kidney allograt recipients: incidence, risk factors, and management. Transplantation 73:379, 2002
4383 11. Kazuyuki N, Shigeru S, Norihiko T, et al: Clinical and genetic risk factors for posttransplant diabetes mellitus in adult renal transplant recipients treated with tacrolimus. Transplantation 80:1419, 2005 12. Salvadori M, Bertoni E, Rosati A, Zanazi M: Post-transplant diabetes mellitus; reviev. J Nephrol 16:626, 2003 13. First MR: Tacrolimus based immunsuppression: J Nephrol 17:25, 2004 14. van Hooff JP, Christiaans MHL, van Duijnhoven EM: Evaluating mechanisms of posttransplant diabetes mellitus. Nephrol Dia Transplant 19:6, 2004 15. Gonzales-Posada JM, Hernandez D, Genis BB, et al: Increased cardiovascular risk profile and mortality in kidney allograft recipients with post-transplant diabetes mellitus in Spain. Clin Transplant 20:650, 2006 16. American Diabetes Association: diagnosis and classification of diabetes mellitus. Diabetes Care 28(suppl 1):537, 2005 17. Parikh CR, Klem P, Wong C, et al: Obesity as an independent predictor of posttransplant diabetes mellitus. Transplant Proc 35:2922, 2003 18. Tetsuhiko S, Akemi I, Kazuharu U, et al: Diabetes mellitus after transplant: relationship to pretransplant glucose metabolism and tacrolimus or cyclosporine A– based therapy. Transplantation 76:1320, 2003 19. Wallace TM, Levy JC, Mathews DR, et al: Use and abuse of HOMA modeling. Diabetes Care 27:1487, 2004 20. Davidson J, Wilkinson Q, Dantal J, et al: New-onset diabetes after transplantation: 2003 international consensus guidelines. Transplantation 75:SS3, 2003
D
53%.2,4,5,7,8 Patient age, nonwhite ethnicity, high body mass index (BMI), high doses of glucocorticoids, and immunosuppression with high doses of cyclosporine (CsA) or tacrolimus have been identified as risk factors for the development of PTDM.1–3,9 –11 The diabetogenic potential of tacrolimus is greater than that of CsA.4,12,13 PTDM is a form of type 2 DM, thought to develop in response to relative insulin deficiency resulting from inFrom the Departments of Internal Medicine (M.A., E.V.) and Nephrology (J.S.), University Hospital Brussels, Brussels, Belgium. Address reprint requests to Jacques Sennesael, Department of Nephrology, UZ Brussel, Laarbeeklaan 101, 1090 Brussels, Belgium. E-mail: [email protected]
0041-1345/10/$–see front matter doi:10.1016/j.transproceed.2010.07.009
© 2010 Published by Elsevier Inc. 360 Park Avenue South, New York, NY 10010-1710
4378
Transplantation Proceedings, 42, 4378 – 4383 (2010)
POSTTRANSPLANTATION DIABETES
creased resistance to impaired production of insulin, or from a combination of both.8,12,14 Prevention of PTDM is an important goal in renal transplantation. Uncontrolled diabetes is associated with increased cardiovascular morbidity and mortality, which increase the risk of death with a functioning graft.8,15 Compared with the general population, kidney transplant recipients already have a higher risk of death from cardiovascular (CV) disease, because of the high prevalence of cardiovascular risk (CVR) in dialysis patients together with the adverse effects of immunosuppressive agents.15 CsA is associated with higher blood pressure and lipid levels than tacrolimus. The present study sought to determine the incidence of PTDM among kidney transplant recipients by defining diabetes according to the ADA criteria.16 We sought to identify risk factors for the development of PTDM and analyzed the evolution of the CVR factors during the first 3 years after transplantation. METHODS Patient Selection For this retrospective study we analyzed the files of 200 tansplanted patients between 1986 and December 31, 2006. We excuded all patients with pretransplantation diabetes (n ⫽ 38) as well as those with ⬎1 year of follow-up (n ⫽ 21), or those for whom data were not fully available (n ⫽ 4). Also 1 patient who underwent a partial pancreatectomy after transplantation, consequently developing diabetes was excluded for this analysis. Thus, 136 patients were included in the study.
Diagnosis of Diabetes The diagnosis of PTDM was made according to the criteria of the ADA.16 In 1 patient, diabetes was diagnosed using an oral glucose tolerance test (2-hour post– glucose load glycemia ⱖ200 mg/dL). In all of the other patients (n ⫽ 23), the diagnosis was based on increased fasting (ⱖ126 mg/dL) or random glycemia (ⱖ200 mg/ dL). All transplant recipients who developed PTDM were tested for antibodies to exclude type 1 DM.
4379 Acute rejection episodes were treated with intravenous 1-g methylprednisolone pulses for 3 days. Because mainly corticoids, CsA, tacrolimus, and sirolimus are known to be associated with PTDM, we restricted our analysis to these 4 immunosuppressive drugs and did not take into account the type of antiproliferative drug.
Statistics All statistical tests were performed in 2-tailed fashion using SPSS for Windows 16.0 statistical software package (SPSS, Chicago, IL, USA) and considered to be significant when P ⬍ .05. Differences between groups were assessed by Mann-Whitney U tests for continuous variables and by chi-square tests using Fisher exact test for categoric variables. We investigated differences between diabetesfree survivals and the incidence of diabetes by Kaplan-Meier analysis and log-rank tests. Cox regression analysis was used to determine risk factors for diabetes development.
RESULTS General Characteristics of Patients at Transplantation
We included 136 transplant patients in this study their baseline characteristics are presented in Table 1. The mean ⫾ SD age of the patients at transplantation was 51 ⫾ 12 years, and the mean follow-up time after transplantation was 88 ⫾ 59 months. The youngest patient was 15 years old and the oldest 80 years old. Similar numbers of patients received kidney transplantations because of end-stage renal disease due to glomerulonephritis (n ⫽ 34; 25%), or nephrangiosclerosis (n ⫽ 30; 22%), interstitial nephropathy (n ⫽ 32; 24%), or congenital kidney disease (n ⫽ 30; 22%). In 10 patients (7%), the etiology of renal failure was unknown. Mean duration of dialysis before transplantation was 26.3 ⫾ 21 months (median, 20.7 mo). Two patients did not undergo dialysis before transplantation. Description of Immunosuppressive Regimens
At the time of transplantation all of the patients were treated with corticoids. The mean maintenance dose at the
Data Collection and Analysis The following recipient variables were evaluated at the time of transplantation: age, gender, ethnic background, etiology of renal failure, and duration of dialysis. Because no information was available about recipient family histories of DM that variable was not considered in this study. At transplantation as well as after 1, 3, 5, and 10 years, we recorded the following parameters: body weight, height, and acute rejection episodes. Blood pressure and lipid profiles were analyzed at the time of transplantation as well as 1 and 3 years thereafter.
Immunosuppression Induction immunosuppression consisted of corticoids, a calcineurin inhibitor, and imuran or mycophenolate mofetil (MMF). Before 1986, all kidney transplant recipients received CsA as the calcineurin inhibitor in association with imuran and corticoids. From 1996 onwards, the immunosuppression consisted of CsA or tacrolimus or sirolimus in combination with MMF plus corticoids. Tacrolimus doses were adjusted to maintain drug levels of 8 –12 ng/dL, and CsA to 100 –150 ng/mL.
Table 1. General Characteristics of Included Patients at Transplantation Number of patients Age BMI, kg/m2 Male/female ratio Ethnicity, n (%) Caucasian North African Asian Renal pathology, n (%) Nephrangiosclerosis interstitial nephropathy Congenital kidney disease Glomerulonephritis Unknown Duration of dialysis, mo Follow-up, mo Age and BMI are expressed as mean ⫾ SD.
136 51 ⫾ 12 24.9 ⫾ 4.2 74/62 127 (93) 8 (6) 1 (1) 30 (22) 32 (24) 30 (22) 34 (25) 10 (7) 26 ⫾ 21 88 ⫾ 59
4380
ALLAMANI, SENNESAEL AND VANDEMEULENBROUCKE
moment of transplantation was 24.6 ⫾ 4.6 mg; (n ⫽ 136); the mean dose at 1 year after transplantation was 7.3 ⫾ 2.9 mg (n ⫽ 136). At 3, 5, and 10 years the number of patients treated with corticoids were 110/112, 91/92, and 11/51, respectively. The mean dose was tapered to 6.l ⫾ 2.7, 5.2 ⫾ 2.2, and 3.3 ⫾ 2.5 mg, respectively. Ten years after transplantation, 50% of PTDM patients and 82% of patients without PTDM were still prescribed corticoids. At transplantation, 123 (90.4%) patients received CsA with a mean dose of 8.7 ⫾ 2.5 mg/kg, and the average (SD) doses at l, 3, 5, and 10 years were 3.5 ⫾ 1.6, 2 ⫾ 1.6, 2.8 ⫾ 0.9, and 2 ⫾ 1.4 mg/kg, respectively. The numbers of patients receiving CsA were 98 (72%), 79 (70.5%), 65 (70%), and 35 (69%) after 3, 5 and 10 years, respectively. Eleven transplant recipients (8%) received tacrolimus at transplantation at 1, 3, 5, and 10 years, 20 (15%), 15 (13%), 8 (7%), and 8 (16%) patients were on tacrolimus, respectively. At the time of transplantation, 61 recipients (45%) received MMF and 73 (54%) imuran. Incidence of PTDM
Twenty-four (17.6%) recipients developed PTDM. The diagnosis of PTDM was established between 10 days and 11 years after transplantation; the median interquartile range was 12 months (1.4 – 61.1). One-half of the patients (n ⫽ 12) developed PTDM within the first year after transplantation, 16 within 3 years, and 19 within 5 years after transplantation. Figure 1 shows the incidence of PTDM over the follow-up. General characteristics for patients with PTDM are presented in Table 2. Mean age, gender, dialysis duration, and etiology of renal failure were similar between those subjects with PTDM and those without. The mean body mass index (BMI) at transplantation in recipients with PTDM was significantly higher than in those without
Fig 1.
Cumulative incidence of PTDM (Kaplan-Meier plot).
Table 2. Comparison of the General Characteristics at Transplantation for the Patients Who Developed Diabetes after Transplantation (PTDMⴙ) Compared with Those Who Did Not (PTDMⴚ) PTDM⫹
Number of patients Age, y BMI, kg/m2 Male/female ratio Ethnicity, n (%) Caucasian North African Asian Renal pathology, n (%) Nephrangiosclerosis Interstitial nephropathy Congenital kidney disease Glomerulonephritis Unknown
24 51 ⫾ 8 26.7 ⫾ 3.3 14/10
PTDM⫺
112 51 ⫾ 13 24.5 ⫾ 4.3 60/52
21 (88) 3 (12) 0 (0)
106 (95) 5 (4) 1 (1)
8 (33) 1 (4) 5 (21) 8 (33) 2 (8)
22 (20) 31 (28) 25 (22) 26 (23) 8 (7)
P Value
.954 .006 .67 .288
.126
Age and BMI are expressed as mean ⫾ SD.
PTDM [26.7 ⫾ 3.3 vs 24.5 ⫾ 4.3; P ⫽ .006]. There were no differences in gender or ethnicity distribution between patients who developed PTDM and those who did not. Risk Factors for Developing PTDM
Age. The mean age of patients with PTDM at transplantation was 51 ⫾ 12 years (Table 2), which was similar to the mean age of patients without PTDM (51.8 ⫾ 13 years). The median age of all patients was 50 years. Patients were divided into 2 groups: ⱖ50 years versus ⬍50 years old. We did not find a significant difference in the development of PTDM among patients ⱖ50 years old compared with those ⬍50 years old (log-rank: P ⫽ .2). BMI. Mean BMI at the time of transplantation was significantly greater among patients with PTDM compared with without patients (26.7 ⫾ 3.3 vs 24.5 ⫾ 4.3; P ⫽ .006). The study population was arbitarily subdivided according to BMI into 3 groups: ⬍25 kg/m2; 25–30 kg/m2 and ⬎30 kg/m2. Figure 2 shows the Kaplan-Meier plot of the incidence of PTDM and diabetes-free survival among these recipients, showing a significant difference (log-rank: P ⫽ .013). We repeated this analysis using 2 groups of patients, BMI ⬍25 kg/m2 versus ⱖ25 kg/m2, showing again a significant difference (log-rank: P ⫽ .004). Duration of dialysis. Patients were than subdivided into 2 groups according to their median time on dialysis (21 months): ⱖ21 versus ⬍21 months. The mean time on dialysis was similar between patients with and without PTDM (27 ⫾ 21.5 vs 23 ⫾ 19 months; log rank: P ⫽ .4). There was no significant difference in the incidence of PTDM between the groups. Immunosuppression. Corticosteroids. All patients received similar doses of corticoids for induction at transplantation. However in cases of acute rejection episodes (ARE), we administered high doses of corticoids. To determine the contribution of this immunosuppressive drug to the devel-
POSTTRANSPLANTATION DIABETES
4381
parameter that was withheld in the analysis was BMI. Compared with recipients with BMI ⬍25 kg/m2, the odds ratio [OR] for recipients with BMI 25–30 kg/m2 was 3.53 (95% CI 1.26 –9.90; P ⫽ .017). Compared with patients with BMI ⬍25 kg/m2, patients with BMI ⬎30 kg/m2 showed an OR of 4.58 (95% CI 1.4 –14.01; P ⫽ .012). The interaction of age and BMI and the combination of ARE and immunosuppression with tacrolimus did not influence the risk for developing PTDM. Differences in CV Events and CVR Factors between the Two Groups during Follow-Up
Fig 2. Incidence of PTDM according to BMI.
opment of PTDM, we compared the occurrence of ARE as a surrogate marker for total corticosteroid burden. Among the group of patients with PTDM, 21% of recipients developed ARE, as did 26% without diabetes (P ⫽ .6). We also examined the effect of the combination of hig-dose corticoids for ARE and tacrolimus on the incidence of PTDM, showing no significant influence. Calcineurin inhibitors. At the time of transplantation, 92% of the patients received CsA (92% in non-PTDM group vs 91.7% in PTDM group) and 8% tacrolimus (8% vs 8.3%; P ⫽ .96 by Mann- Whitney). The mean daily dose of both calcineurin inhibitors was decreased over the follow-up period: 0.19 to 0.057 mg/kg and 8.7 to 2 mg/kg for tacrolimus and CsA, respectively. Among the patients initially treated with CsA, 22 developed PTDM as did 2 in the tacrolimus group. Multivariate analysis of risk factors for the development of PTDM. Using demographic and immunosuppressive factors as independent variables we performed a multivariate analysis of potential risk factors for the development of PTDM; namely, age, gender, BMI at transplantation, time on dialysis, type of calcineurin inhibitor, and ARE. The only
PTDM patients developed more CV events than patients without PTDM (n ⫽ 7 with vs 2 without PTDM). Among these patients, 5 with diabetes and 1 without diabetes needed surgical treatment (P ⫽ .03). Classic CVR factors, such as blood pressure (BP) and dyslipidemia, were different between groups. Although systolic BP was similar at and 1 year after transplantation, it was significantly higher after 3 years in the PTDM group: 151 mm Hg vs 139 mm Hg in the non-PTDM group (P ⫽ .006). At 3 years after transplantation, diastolic blood pressure was also higher in this group: 91% mm Hg vs 84 mm Hg; P ⫽ .03). There were no differences in total cholesterol levels. In contrast, we observe significantly higher tiglyceride (TG) levels at the time of (214 mg/dL vs 168 mg/dL; P ⫽ .006) and 3 years after transplantation (218 mg/dL in PTDM vs 165 mg/dL in non-PTDM patients; P ⫽ .013). Table 3 shows the CVR factors in the 2 groups. DISCUSSION
PTDM is an important cause of morbidity and mortality among renal transplant recipients, with an incidence ranging from 2% to 53%.2,4,5,7,8 This retrospective study of 136 patients who underwent renal transplantation at our center showed a 17.6% cumulative incidence of PTDM. The main risk factor for the development of PTDM in our population was BMI at transplantation. No relationship was noted between PTDM development and other demographic characteristics or the type of immunosuppressive regimen. The wide range of incidences of PTDM described in the literature can be explained by various factors. First, the citeria to diagnose diabetes varied among the studies.1– 6 In the present study, diabetes was diagnosed according to ADA criteria,16 using fasting or random glycemia, a method
Table 3. CVR Factors in the PTDMⴙ and PTDMⴚ Groups At Transplantation
1 y after Transplantation
3 y after Transplantation
CVR Factor
PTDM⫹
PTDM⫺
P Value
PTDM⫹
PTDM⫺
P Values
PTDM⫹
PTDM⫺
P Values
Systolic BF (mm Hg) Diastolic BP (mm Hg) Total cholesterol (mg/dL) Triglycerides (mg/dL)
14.3 ⫾ 14 83 ⫾ 11 207 ⫾ 47 214 ⫾ 91
145 ⫾ 21 85 ⫾ 14 185 ⫾ 50 168 ⫾ 95
ns ns ns ns
145 ⫾ 14 85 ⫾ 10 204 ⫾ 37 166 ⫾ 14
137 ⫾ 19 83 ⫾ 12 230 ⫾ 48 174 ⫾ 79
ns ns ns .006
151 ⫾ 13 91 ⫾ 13 240 ⫾ 53 218 ⫾ 95
139 ⫾ 19 84 ⫾ 12 230 ⫾ 49 165 ⫾ 66
.006 .003 ns .013
The values are expressed as mean ⫾ SD. Abbreviation: BP, blood pressure.
4382
that was consistent with some earlier studies.1– 4 In a recent work, however, it was concluded that fasting blood glucose might not be the optimal screening test for the diagnosis of PTDM in kidney transplant recipients, that the OGTT was more accurate and reproducible.5 At our institution, an OGTT was implemented in 2006 as part of the yearly follow-up of transplant recipients, but in the present study, only 1 patient was diagnosed by an OGTT. For this reason we concluded that we may have underestimated the incidence of PTDM. Long-term follow-up of patients transplanted since 2006, and thus submitted yearly to an OGTT, showed clarify this issue in the coming years. Furthermore, the follow-up times after transplantation varied considerably among studies2,3,4,7,15,17,18; it was often limited to 2 years.4,15 As shown herein, the risk of PTDM was the highest during the first year, but it can appear up to 11 years later. Different immunosuppressive regimens also may explain variations in the incidences of PTDM.2,7,18 The introduction of calcineurin inhibitors has reduced the incidence of PTDM by lowering the doses of corticoids. This beneicial effect is somewhat counterbalanced by the intrinsic diabetogenic effect of these agents, especially tacrolimus.1,3,4,7,11,13,14 Finally, the diversity in patient demographics among studies may influence the reported incidence of PTDM. Older age and African-American ethnicity have been associated with higher incidence of PTDM.2,7,12 In the present study, we also analyzed risk factors for the development of PTDM. Upon univariate and multivariate analysis, only BMI at transplantation significantly correlated with diabetes development. Patients with an initial BMI ⬎25 kg/m2 showed an OR of 3.53 (P ⫽ .017), and recipients with BMI ⬎30 kg/m2 an OR of 4.58 (P ⫽ .012) compared with lean recipients (BMI ⬍25 kg/m2), a finding consistent with earlier observations in transplant recipients.9,17,18 Weight gain is a common problem after transplantation. In the present study, we observed weight gain in both groups of patients, but those with PTDM gained significantly more weight during the first year after transplantation (4.5 ⫾ 6.7 kg) than those without PTDM (1.8 ⫾ 7 kg; P ⫽ .04). Obese patients may be more insulin resistant, which could, in combination with the immunosuppressive treatment, lead to a relative insulin deficiency and with increased blood glucose to the development of PTDM. However, we did not have data on insulin resistance in our study population, because models used to compute insulin resistance have not been validated in end-stage renal disease patients.19 Elderly recipients, especially those ⬎45 years old have been reported to be at greater risk for PTDM.2,11 Perhaps, because of the small number of patients in the present study, we could not identify this variable as an independent risk factor for PTDM. Neither could we assess ethnicity, probably because of small number of noncaucasians among our study population. In most studies, tacrolimus has been more frequently associated with PTDM than CsA; however, we did not note a significant difference between the cal-
ALLAMANI, SENNESAEL AND VANDEMEULENBROUCKE
cineurin inhibitors, possible due to the small number of patients on tacrolimus. Our study has some limitations. It was retrospective and did not include a large number of cases, a limitation in all single-center experiences with PTDM.6,7,11,18 OGTT was not routinely performed as a screening test for kidney transplant recipients until 2006, but more recently transplanted patients are undergoing this test yearly. Their future follow-up may show whether this test has an important impact on the observed incidence of PTDM. Unfortunately, we did not have data on the family histories of diabetes among transplant recipients, a major risk factor for diabetes development. The strengths of this study are the long-term follow-up after transplantation with few patients lost. Diabetes diagnosis was more robust compared with other studies on PTDM.1– 4,7,18 This single-center study had a uniform immunosuppressive treatment and homogeneous follow-up. Overweight is associated, in the general population with some traditional CVR factors, such as dyslipidemia, hypertension, and insulin resistance. Likewise, in renal transplant patients, obesity has been associated with a worse CVR profile.15,20 Dyslipidemia is another CVR factor highly prevalent among kidney transplant recipients.8,15 We observed higher TG levels among diabetic patients, but no significant difference in total cholesterol levels. Although a greater number of patients with PTDM received lipidlowering medications, it has been suggested that more aggressive therapy would probably reduce the risk of CV disease in this population. Hypertension is another traditional CVR factor frequently observed in transplant patients. The mean blood pressure was significantly higher among the PTDM group in the present study. In general, no antihypertensive agent is contraindicated in renal transplant; this risk factor needs to be controlled rigorously. We concluded that PTDM is a serious complication of renal transplantation, with an incidence of 17.6% over 11 years. Most cases occur during the first year after transplantation. Obesity was the only independent predictor of PTDM. However, it is a modifiable risk factor; obese patients should be counseled to lose weight before transplantation. Furthermore, prevention of weight gain after transplantation may protect patients from the development of PTDM. Although the immunosuppressive regimen with the least diabetogenic potential should be used, this issue may be less important among nonobese patients. PTDM was associated with a worse CVR profile. Better control of modifiable risk factors, such as lipid profile and blood pressure, may improve the prognoses of these patients.
REFERENCES 1. Weir MR, Fink JC: Risk for posttransplant diabetes mellitus with current immunosuppressive medications. Am J Kidney Dis 34:1, 1999
POSTTRANSPLANTATION DIABETES 2. Cosio FG, Pesavento TE, Osei K, et al: Post-transplant diabetes mellitus: increasing incidence in renal allograft recipients transplanted in recent years. Kidney Int 59:732, 2001 3. Margreiter R: Eficacy and safety of tacrolimus compared with cyclosporine microemulsion in renal transplantion: a randomised multicenter study. Lancet 359:741, 2002 4. Marcen R, Morales JM, del Castilo D, et al: Posttransplant diabetes mellitus in renal allograft recipients: a prospective multicenter study at two years. Transplant Proc 38:3530, 2006 5. Armstrong KA, Prins JB, Beller EM, et al: Should an oral glucose tolerance test be performed routinely in all renal transplant recipients? Clin J Am Soc Nephrol 1:100, 2006 6. Hur KY, Kim MS, Kim YS, et al: Risk factors associated with the onset and progression of posttransplantation diabetes in renal allograft recipients: Diabetes Care 30:609, 2007 7. Numakura K, Satoh S, Tschiya N, et al: Incidence and risk factors of clinical characteristics, tacrolimus pharmacocinetics, and related genomic polymorphisms for posttransplant diabetes mellitus in early stage of renal transplant recipients. Transplant Proc 37:1865, 2005 8. Ravidran V, Baboolal K, Moore R. Posttransplant diabetes mellitus after renal transplantation: The emerging clinical challenge. Yonsei Med J 45:1059, 2004 9. Kasiske BL, Synder JJ, Gilbertson D, et al: Diabetes mellitus after kidney transplantation in the United States. Am J Transplant 3:178, 2003 10. First MR, Gerber DA, Hariharan S, et al: Posttransplant diabetes mellitus in kidney allograt recipients: incidence, risk factors, and management. Transplantation 73:379, 2002
4383 11. Kazuyuki N, Shigeru S, Norihiko T, et al: Clinical and genetic risk factors for posttransplant diabetes mellitus in adult renal transplant recipients treated with tacrolimus. Transplantation 80:1419, 2005 12. Salvadori M, Bertoni E, Rosati A, Zanazi M: Post-transplant diabetes mellitus; reviev. J Nephrol 16:626, 2003 13. First MR: Tacrolimus based immunsuppression: J Nephrol 17:25, 2004 14. van Hooff JP, Christiaans MHL, van Duijnhoven EM: Evaluating mechanisms of posttransplant diabetes mellitus. Nephrol Dia Transplant 19:6, 2004 15. Gonzales-Posada JM, Hernandez D, Genis BB, et al: Increased cardiovascular risk profile and mortality in kidney allograft recipients with post-transplant diabetes mellitus in Spain. Clin Transplant 20:650, 2006 16. American Diabetes Association: diagnosis and classification of diabetes mellitus. Diabetes Care 28(suppl 1):537, 2005 17. Parikh CR, Klem P, Wong C, et al: Obesity as an independent predictor of posttransplant diabetes mellitus. Transplant Proc 35:2922, 2003 18. Tetsuhiko S, Akemi I, Kazuharu U, et al: Diabetes mellitus after transplant: relationship to pretransplant glucose metabolism and tacrolimus or cyclosporine A– based therapy. Transplantation 76:1320, 2003 19. Wallace TM, Levy JC, Mathews DR, et al: Use and abuse of HOMA modeling. Diabetes Care 27:1487, 2004 20. Davidson J, Wilkinson Q, Dantal J, et al: New-onset diabetes after transplantation: 2003 international consensus guidelines. Transplantation 75:SS3, 2003