The impact of pretransplant 25-hydroxy vitamin D deficiency on subsequent graft function: An observational study

One hundred forty-four patients who had been hospitalized for KT at Yonsei University Health System in Seoul, Korea (latitude: 37.5°N; annual average sunshine hours: 5.8 hours per day) between April 1, 2002 and June 30, 2004 were enrolled and had their 25-OHD level assessed. All participants were ethnically homogeneous Korean population. At enrollment, patients who were less than 16 years old, had active liver disease or liver cirrhosis, or had received prior KT were excluded. During follow-up period, patients who were forced to take vitamin D supplements after KT for compelling indications, such as pre-existing uremic osteodystrophy or post-transplant osteoporosis, or who did not have an equal number of observations (estimated renal function) made at fixed regular intervals were also excluded from the analysis without imputation of missing data to minimize estimation bias in statistics used in this study (details are below) [13]. Therefore, 106 participants, who maintained functioning grafts and were regularly assessed for graft function without missing data throughout the 36-month study period, were only included in the final analysis (Figure 1).

Baseline demographic and clinical characteristics and laboratory data including 25-OHD level were obtained during hospitalization prior to transplantation. For correcting serum total calcium for the effects of confounders, such as protein level or nutritional status, albumin-corrected calcium was calculated as previously described [14]. Predominant renal replacement therapy (RRT) during the pretransplant end-stage renal disease (ESRD) course was defined as the RRT modality for >50% of the ESRD period.

Of the screened 144 patients, 16 patients (11.1%) experienced acute rejection episodes. Diagnosis of all these rejection episodes was confirmed by biopsy. Of whom, 11 patients were diagnosed with acute cellular rejection and were treated with 1 g boluses of intravenous methyl prednisolone for 3 consecutive days or additional polyclonal/monoclonal antibody therapies for steroid-resistant cases (3 of 11 episodes), which enabled them to maintain graft patency and be included in the final analysis. In contrast, 5 patients were diagnosed with acute mixed humoral/cellular rejection and lost their grafts despite receiving intensive courses of immunosuppressive therapy and returned to dialysis (Figure 1).

These patients were prospectively evaluated for graft function at six-month intervals for 36 months following KT by estimated glomerular filtration rate (eGFR) using the Modification of Diet in Renal Disease (MDRD) equation. When analyzing serum 25-OHD concentrations in this population, 25-OHD deficiency was defined as a serum 25-OHD < 10 ng/mL because this level is associated with an increased risk of rickets in children and osteomalacia in adults [15]. This study was approved by the Institutional Review Board of Severance Hospital, Yonsei University College of Medicine (no. 4-2009-0382) and written informed consent was obtained from all subjects.

All specimens measured for 25-OHD were examined in the endocrinology laboratory at the Medical Research Center of Yonsei University Health System. 25-OHD level was evaluated using a Diasorin 25-OHD ¹²⁵I radioimmunoassay kit [16], sensitivity ≤ 1.5 ng/mL; inter-assay coefficient of variation (CV) = 11.7% at 8.6 ng/mL and 8.6% at 33.0 ng/mL.

During the study period, most patients were treated with a triple regimen of calcineurin inhibitor (cyclosporine or tacrolimus), purine synthesis inhibitor (azathioprine or mycophenolate mofetil), and prednisolone.

Data analysis was performed using SPSS for Windows, version 12.0 (SPSS, Chicago, IL, USA). All results are presented as the mean ± standard error of the mean (SEM). The independent t-test was used to compare means and analysis of variance (ANOVA) with Bonferroni post-hoc test was used for multiple comparisons between groups. The Chi-square test was used to compare proportions between variables. Logistic regression analysis was used to estimate the odds ratios and to identify the independent pretransplant risk factors for 25-OHD deficiency; all variables at the P < 0.1 level in univariate analyses and recipient age at KT were included in the model.

Demographic characteristics of the patients analyzed are listed in Table 1. The mean age was 40.2 ± 1.1 years (range: 16.8–60.6 years), and 31 participants (29.2%) were women. Forty-nine (46.2%) patients had received hemodialysis, and 22 (20.2%) had received peritoneal dialysis as a predominant RRT modality prior to transplantation. The mean duration of dialysis for these 71 patients was 16.8 ± 2.7 months. The remaining 35 subjects (33.0%) received pre-emptive transplants.

The mean serum 25-OHD concentration was 13.1 ± 0.6 ng/mL; 41 participants (38.7%) had vitamin D deficiency defined as less than 10 ng/mL. Considering the latitude of Korea, it can be assumed that there is a seasonal variation in ultraviolet exposure, which is a major determinant of 25-OHD level. Therefore, we analyzed the seasonal variation of pretransplant 25-OHD levels. As shown in Figure 2, there was a trend that serum 25-OHD levels were higher in summer (14.3 ± 0.9 ng/mL)/autumn (14.4 ± 1.4 ng/mL) than in spring (11.4 ± 0.9 ng/mL)/winter (11.4 ± 1.2 ng/mL), but did not reach statistical significance both in ANOVA (P = 0.081) and post-hoc multiple comparison using Bonferroni test.

Characteristics of the patients according to serum 25-OHD level are shown in Table 2. Results showed that the patients divided into two groups according to 25-OHD level were similar with regard to donor age and gender, percentage of living and deceased donors, the number and locus of HLA mismatch, percentage of ABO mismatch, etiology of renal failure, and duration of dialysis. However, the patients with 25-OHD deficiency at KT were more likely to be female (P = 0.002), had received maintenance peritoneal dialysis as a predominant RRT modality prior to KT (P < 0.001), experienced more frequent biopsy-proven acute rejection episodes (P = 0.003), and had lower serum levels of hemoglobin (P = 0.028) and albumin (P < 0.001). Serum calcium level was also lower in the patients with 25-OHD deficiency (P = 0.004), but lost significance after correcting for serum albumin level (P = 0.121).

Among the pretransplant parameters, an adjusted multivariate logistic regression model showed that female gender was independently associated with 25-OHD deficiency (odds ratio [OR] 3.30, 95% CI 1.33–8.21, P = 0.010), and increased serum albumin level was associated with lower odds of 25-OHD deficiency (OR 0.35, 95% CI 0.13–0.98, P = 0.047). In addition, predominant RRT modality before KT was also associated with 25-OHD deficiency (P < 0.001) (Table 3).

A repeated measures ANCOVA was performed to explore whether pretransplant 25-OHD level had an impact on changes in eGFR during the 36-month follow-up period as a main between-subjects factor. We also entered recipient gender and predominant RRT modality before KT as between-subjects factors, and serum albumin level as a covariate into the model because they were potential pretransplant contributors to differences in 25-OHD level (Table 3). Since an acute rejection episode is an established precedent for chronic graft dysfunction [17] and the incidence of posttransplant biopsy-proven acute rejection was significantly different according to 25-OHD level (Table 2), a biopsy-proven acute rejection episode was also included as a between-subjects factor. The within-subjects factor was represented by the eGFR at each six-month interval after KT. Upon multivariate test of within-subjects effect, there was no significant main effect for eGFR at each 6-month interval after KT [Wilks’ lambda = 0.92, F(6, 83) = 1.39, P = 0.228], whereas eGFR at each 6-month interval after KT by group (25-OHD deficiency vs. control) interaction was significant [Wilks’ lambda = 0.84, F(6, 83) = 3.56, P = 0.003]. This indicates that, while overall graft function did not change significantly throughout the 36 months after KT, the patterns of changes in graft function were not consistent and depend on 25-OHD level. Subsequent analysis of between-subjects effects revealed that only the main effect of group according to 25-OHD level (25-OHD deficiency vs. control) was significant, and none of the interactions associated with group according to 25-OHD level (25-OHD deficiency vs. control) were significant [F(1, 88) = 12.07, P = 0.001] (Figure 3) (Table 4). While the main effect of biopsy-proven acute rejection episode was not significant, there were significant interactions of Biopsy-proven acute rejection episode with recipient gender [F(1, 88) = 6.78, P = 0.011] and predominant RRT modality before KT [F(1, 111) = 3.84, P = 0.025]. These results suggest that an acute rejection episode affects post-KT graft function differently according to recipient gender and predominant RRT modality before KT. Further, post hoc repeated measures ANCOVA analyses with the Bonferroni corrections revealed that, after an biopsy-proven acute rejection episode, males experienced a persistently lower level of graft function [F(1, 73) = 7.57, P = 0.007] but females did not [F(1, 29) = 0.65, P = 0.426], and only the graft function of patients receiving pre-emptive KT was affected adversely by a biopsy-proven acute rejection episode [F(1, 33) = 5.86, P = 0.020] but patients receiving HD [F(1, 47) = 1.80, P = 0.185] or PD [F(1, 20) = 0.28, P = 0.605] as predominant RRT modality before KT were not adversely affected.