Objectives: Although both tacrolimus and myco-phenolate have improved outcomes after kidney transplant, studies regarding effects of exposure on outcomes, specifically related to racial disparities, are sparse.
Materials and Methods: In this 8-year longitudinal cohort study of adult kidney transplant recipients, mycophenolate and tacrolimus levels were compared across transplant vintage stratified by non-African Americans versus African Americans. Data were analyzed with standard univariate tests and multivariable regression models.
Results: Our study included 1217 patients (transplanted from 2005-2013) who had tacrolimus and myco-phenolate exposure data, with follow-up through 2015 (53.7% were African Americans). Mean mycophenolate dose was 1672 ± 463 mg/day during the first 3 years posttransplant. Although transplant vintage did not appreciably impact mycophenolate dosing in non-African Americans (0.7 mg/day/y; P = .903), doses significantly decreased in African Americans across transplant vintage (-20.5 mg/day/y; P < .001). Rate of mycophenolate being held or discontinued based on transplant vintage significantly increased in African Americans but did not change in non-African Americans. At the beginning of the study, mean tacrolimus levels were lower in African Americans; however, levels then slightly decreased in non-African Americans (-0.03 ng/mL/y; P = .279) and slightly increased in African Americans (+0.03 ng/mL/y; P = .247), with similar levels by 2013. Higher tacrolimus levels were protective against rejection in African Americans only but were protective against death-censored graft loss in both race/ethnicity groups. Mycophenolate dosing had no appreciable impact on outcomes in African Americans, but higher mycophenolate dosing was a significant risk factor for death-censored graft loss in non-African Americans.
Conclusions: Tacrolimus and mycophenolate exposure levels have significantly changed over time and differed by race/ethnicity. In non-African Americans, those transplanted more recently tended to have lower tacrolimus but similar mycophenolate exposure. Although mycophenolate exposure in African Americans has recently decreased, tacrolimus has increased. Differences in outcomes likely reflect improved understanding of immunosuppressant tolerability by recipient race/ethnicity.
Key words : Disparity, Drug exposure, Ethnicity, Kidney transplant, Outcomes
Kidney transplant is the treatment of choice for end-stage renal disease (ESRD). Four main classes of drugs are available for maintenance immunosup-pression, including corticosteroids, antimetabolites (azathioprine, mycophenolate mofetil [MMF], and mycophenolate sodium), calcineurin inhibitors (cyclosporine and tacrolimus), and mammalian target of rapamycin inhibitors (sirolimus and everolimus). Conventional immunosuppressive regimens consist of various combinations of 2 or 3 agents from different groups. The rationale for combining medications from different classes is to achieve adequate and optimal immunosuppression while limiting adverse effects.1
The National Health and Nutrition Examination Survey III study showed that the relative risk of ESRD was 5-fold higher in an African American (AA) cohort versus a white cohort. Despite adjustments for age, sex, and diabetes, AA race/ethnicity continued to be associated with higher blood pressures and albuminuria.2 African Americans represent about 12% of the US population but account for 32% of those with chronic kidney failure and are 4 times more likely than white individuals to develop ESRD requiring dialysis or kidney transplant.3
After kidney transplant, some data have shown that AAs tend to be high immune responders,4 placing them at increased risk for allograft rejection. They also tend to have higher rates of delayed graft function5 and comorbid conditions such as hyper-tension6 and posttransplant diabetes mellitus,7 which may contribute to inferior graft survival.8 It is worth noting that oral bioavailability of medications is 20% to 50% lower in AA than in non-AA.9 This may be partly due to AAs having higher prevalence of expression of CYP3A5*1 polymorphism, leading to increased tacrolimus clearance and variability.10
Combined MMF and tacrolimus with or without steroids is a well-known regimen as maintenance immunosuppression in renal transplant recipients. The bioavailability of mycophenolic acid (MPA), the active form of MMF, is about 94%, and it reaches peak plasma concentration about 2 hours after oral administration. Mycophenolic acid undergoes hepa-tic glucuronidation to form mycophenolic acid glucuronide (MPAG), which is pharmacologically inactive. This inactive form is secreted into the bile and converted back to MPA by gut bacteria.11 Mycophenolic acid is then reabsorbed and, via hepatic recirculation, produces a second peak at 8 to 12 hours. Mycophenolate mofetil is excreted in the urine as MPAG, accounting for 90% of the adminis-tered MMF dose.12 Mycophenolate mofetil is most commonly associated with teratogenic, gastrointestinal (nausea, vomiting, diarrhea), and hematologic (leukopenia, thrombocytopenia) adverse effects.13
Tacrolimus is metabolized by the hepatic enzyme system via CYP3A4 and therefore has various drug interactions. Common adverse effects of tacrolimus include nephrotoxicity, neurotoxicity, and metabolic sequelae, particularly diabetes mellitus.14
There is a lack of sufficient data assessing trends of dosing and exposure to immunosuppressive agents on outcomes and disparities according to race/ethnicity. Thus, the aim of this study was to determine changes in MMF and tacrolimus exposure by transplant vintage and race/ethnicity and their impact on clinical outcomes in kidney transplant recipients.
Materials and Methods
This Institutional Review Board-approved, single-center, retrospective, longitudinal cohort study compared groups of adult patients who received kidney transplants at the Medical University of South Carolina between 2005 and 2013. Medical records of 1217 kidney transplant recipients with MMF and tacrolimus exposure data through 2015 were retrospectively reviewed. Intrapatient tacro-limus levels and MMF dosing were assessed and compared across transplant vintage and stratified by race/ethnicity (comparing non-AA with AA patients). Pediatric patients, patients who were not on tacrolimus or MMF regimens, and recipients of nonrenal transplants were excluded.
The primary exposure variables of interest were tacrolimus levels and MMF doses. Variables were assessed at the patient level and compared by transplant vintage and race/ethnicity. Tacrolimus levels were captured from electronic health records (EHR) and included over 60 000 concentrations. Mycophenolate mofetil doses were obtained from manual medical record review of clinic visits or hospitalizations at the following posttransplant intervals: days 1, 3, 5, 7, and 14 and months 3, 6, and 12 and then annually thereafter. Any event in which MMF was stopped due to a particular reason was recorded, with MMF dose denoted as 0 for that time period.
Race/ethnicity was the secondary variable of interest. Race was defined as the patient’s self-report of being AA or non-AA. Our transplant center sees few Asian and Hispanic patients; thus most non-AA recipients were white. Additional variables assessed for this analysis included recipient demographics (age, sex, weight, comorbidities, previous transplant, time on dialysis), donor information (age, sex, race, type), transplant characteristics (panel reactive antibodies, HLA mismatches, cold ischemic time), and immunosuppressive regimen (induction and maintenance therapy). Most data were acquired through the center-specific United Network for Organ Sharing STAR file, supplemented by infor-mation from the EHRs.
Acute rejection and graft loss were the clinical outcomes of interest, assessed using time to event methodology. Rejection was defined as biopsy-proven rejection episode, based on Banff criteria (grade ≥ 1A), which required treatment. Date of rejection was set as the date of kidney biopsy. Graft loss was defined as return to chronic dialysis or retransplant, based on our EHR data.
Data were coded and entered using the Statistical Package for the Social Sciences (SPSS: An IBM Company, version 24, IBM Corporation, Armonk, NY, USA). Data were summarized using means, standard deviation, median, and minimum and maximum in quantitative data and using frequency (count) and relative frequency (percentage) for categorical data. Comparisons between quantitative variables were done using the nonparametric Mann-Whitney test.15 For comparisons of categorical data, chi-square tests were performed. Exact test was used instead when the expected frequency was less than 5. P values less than .05 were considered statistically significant.16
Between 2005 and 2013, our records showed 1217 kidney transplant recipients with MMF and tacrolimus exposure data with follow-up through 2015. Table 1 shows the baseline characteristics of the study cohort, stratified by race. There were 653 AAs (53.7%) and 564 non-AAs (46.3%) in the analysis. There were significant differences between the groups with regard to cause of ESRD, preexisting diabetes mellitus, age, ESRD diagnoses, donor characteristics (age, sex, race, deceased donor), HLA mismatches, delayed graft function, and interleukin 2 receptor induction therapy. In general, the AA cohort was at higher risk for deleterious outcomes based on baseline characteristics. The AA cohort had higher prevalence of hypertension and diabetes, a higher proportion requiring dialysis before transplant, longer dialysis vintage, fewer living-donor transplants, and higher prevalence of delayed graft function.
The mean MMF dose was 1672 ± 463 mg/day during the first 3 years posttransplant. Transplant vintage did not appreciably impact MMF dosing in non-AAs (0.7 mg/day/y; P = .903), whereas MMF dosing significantly decreased in AAs across transplant vintage (-20.5 mg/day/y; P < .001; Figure 1). African Americans also had a significant increase in the rate of MMF being held or discontinued based on transplant vintage, which did not change in non-AAs (Figure 2). Mean tacrolimus levels were lower in AAs versus non-AAs in 2005. However, over time, levels were slightly decreased in non-AAs (-0.03 ng/mL/y; P = .279) and slightly increased in AAs (+0.03 ng/mL/y; P = .247), such that mean tacrolimus levels were similar by 2013 (Figure 3).
In terms of outcomes (Table 2), higher tacrolimus levels were protective against rejection in AAs only but were protective against death-censored graft loss in both AAs and non-AAs. We found that MMF dosing had no appreciable impact on outcomes in AAs, but higher MMF dosing was a significant risk factor for death-censored graft loss in non-AAs. In non-AAs, tacrolimus and MMF levels did not correlate with acute rejection and total deaths, but there was a significant correlation in tacrolimus and MMF levels and death-censored graft loss. On the other hand, in AAs, there was a significant correlation between acute rejection and death-censored graft loss and tacrolimus and MMF levels. There was no significant correlation between tacrolimus and MMF levels and total number of deaths.
The results of this study demonstrate that tacrolimus and MMF exposure levels have changed over time and have significantly differed by race. In non-AAs, those transplanted more recently had lower tacrolimus exposure with similar MMF exposure. In recent years, MMF exposure has decreased in AAs but tacrolimus exposure has slightly increased. These data provide novel insights into the evolving use, dosing, and therapeutic drug monitoring of tacro-limus and MMF in AA and non-AA renal transplant recipients and may represent a better understanding of race/ethnicity-specific tolerability of these agents.
Phase 3 clinical trials have shown that AAs require higher doses of MMF to achieve graft outcomes similar to white renal transplant recipients.17 Results have also revealed that enterohepatic recirculation of MPA was twice as prevalent in white patients.18 Other retrospective studies have shown that an increase of approximately 35% to 40% in MPA dose in AA males is needed to maintain MPA concentration similar to white males. This may be explained by the fact that the pharmacokinetics of these drugs is significantly different in non-AA populations.19-21 Non-AAs have more extensive enterohepatic recirculation compared with AAs. This results in slower MPA clearance and greater exposure to MMF in non-AA patients.22 This may have contributed to death-censored graft loss in this population.
Mycophenolic acid is greatly bound to albumin with competitive binding occurring between MPAG and urea.23,24 Although albumin concentrations were normal, patients in our study had varying degrees of renal dysfunction, with greater variation among non-AAs. This interrelationship between renal dys-function and competitive protein binding may support the slower MPA clearance and greater drug exposure seen in non-AAs on MMF.11,19 In addition, considerable interpatient variability may contribute to delayed enterohepatic recirculation and racial/ethnic differences in the total MPA exposure.
Cyclosporine inhibits enterohepatic recirculation of MPAG via the multidrug resistance-associated protein 2 (MRP-2) transporter, preventing decon-jugation to MPA and subsequent elimination of the second peak.25,26 The MRP-2 efflux transporters found in liver, kidney, and intestinal epithelium apical membrane contribute to enterohepatic circulation of MPAG to MPA.25,26 Inhibition of MRP-2 by cyclosporine varies between patients and may be related to the pharmacokinetics of this drug.18,25 Whether tacrolimus has a less potent or no effect on MPAG metabolism compared with cyclosporine has not yet been clearly elucidated.27
The possibility that MRP-2 may vary between AAs and non-AAs and alter MPA and MPAG pharmacokinetics has not been explored. Limited preliminary in vivo data of MRP-2 expression in the gastrointestinal tracts of a white male patient and Chinese male patient suggested a racial difference.28 Larger populations are needed to clarify whether differences in efflux transporters play a racial role in MPA pharmacokinetics. Prospective studies are needed to confirm this recommendation and evaluate the role of routine therapeutic drug monitoring.
It is conceivable that AAs simply have a more robust immune system, are presented with more disparate HLA antigens, or have more frequent sensitizing events. Alternatively, there could be differences in the sensitivity of intracellular molecules to inhibition with MPA.29 Stabilization in MMF pharmacodynamics reaching a plateau level and a stable condition or tolerance development to the renal graft may explain the drop in the dose of MMF in AAs in later years. Frequent discontinuation or holding of MMF over the duration of the study in AAs may be related to the development of adverse effects of the drug, stemming from relatively higher dosing (leukopenia, viral infection).
Taber and associates30 have shown that mean tacrolimus levels were lower in AA versus non-AA patients (P < .05). During the first year after kidney transplant, AA patients were 1.7 times less likely to achieve therapeutic tacrolimus concentrations (8 ng/mL or higher) compared with non-AA patients (35% vs 21%, respectively; P < .001). In addition,AA patients who did not achieve therapeutic concentrations were 2.4 times more likely to have acute cellular rejection than AAs who achieved therapeutic concentrations (20.8% vs 8.5%, respectively; P < .01) and 2.5 times more likely to have antibody-mediated rejection (8.9% vs 3.6%, respectively; P < .01).30
It is well-established that AAs are at substantially higher risk of acute rejection and graft loss after renal transplant.4,31,32 In addition, AAs require significantly higher doses of tacrolimus to achieve therapeutic trough concentrations.31,32 This is likely a reflection of racial/ethnic differences in gene variants as-sociated with drug absorption and metabolism.33 Other than pharmacokinetic issues, AA patients are also at higher immunologic risk, which may be related to differences in the pharmacodynamics of immunosuppressive therapy, including tacrolimus.31,34
In contrast to non-AAs, AAs who achieve the-rapeutic tacrolimus concentrations have substantially lower acute rejection rates but are at risk of develop-ing interstitial fibrosis and tubular atrophy. These findings may reflect modifiable, time-dependent, racial differences in the concentration-effect relation-ship of tacrolimus. Achievement of therapeutic tacrolimus trough concentrations, potentially through genotyping and more aggressive dosing and monitoring, is essential to minimize the risk of acute rejection in AA kidney transplant recipients.30
So far, no published reports exist on such postulated intracellular racial/ethnic differences. However, consideration of patient sex and race/ethnicity on therapeutic drug monitoring of tacrolimus in kidney transplant, using 12-hour trough concentrations to approximate total exposure, is utilized in most transplant centers.35,36 Despite this, evidence to support tacrolimus in improving clinical outcomes, either through reduction in acute rejection rates or toxicities, is conflicting. Data, pooled from 3 randomized controlled trials and conducted in a predominantly low-risk cohort of kidney transplant recipients, have suggested that achieving therapeutic tacrolimus trough concentrations early post-transplant is not associated with reduced rejection rates.37 However, other studies have demonstrated that tacrolimus trough concentrations are associated with improved efficacy in higher risk patients.11,38
There is a paucity of studies that have assessed whether AA race/ethnicity modifies the impact of tacrolimus trough concentrations on clinical outcomes in kidney transplant. This is likely because of an underrepresentation of AA patients in previous studies on this topic.39-42
These data demonstrate that tacrolimus and MMF exposure have significantly changed over time and have differed by race/ethnicity. In non-AA patients, those who had more recent transplant procedures tended to have lower tacrolimus exposure but similar MMF exposure. In AA patients, MMF exposure has decreased in recent years, whereas tacrolimus exposure has slightly increased. The effects of tacrolimus and MMF exposure on outcomes also differ by race, and these results likely reflect the providers’ improved understanding of immuno-suppressant tolerability by recipient race/ethnicity.
Our study has some limitations, including its retrospective design. Compliance on MMF was not verified because information on administered doses was obtained from medical records. Future prospective studies should focus on enterohepatic recirculation of MMF in different races/ethnicities and on racial/ethnic disparities in the phar-macokinetics and pharmacodynamics of tacrolimus and MMF combination.
DOI : 10.6002/ect.2018.0055
From the 1Division of Nephrology and Hypertension, Department of Medicine, and
the 2Division of Transplant Surgery, Department of Surgery, Medical University
of South Carolina, Charleston, South Carolina, USA; and the 3Medical Services,
Ralph H. Johnson VA Medical Center, Charleston, South Carolina, USA; and 4Cairo
University, Division of Nephrology, Department of Medicine, Cairo, Egypt
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare. This study was previously presented as a poster to the American Transplant Congress, Chicago, IL, USA, April 29 to May 3, 2017.
Corresponding author: Karim M Soliman, Division of Nephrology, Department of Medicine, 96 Jonathan Lucas St., Charleston, SC 29425, USA
Phone: +1 843 906 0489
Table 1. Demographic Data
Table 2. Impact of Tacrolimus and Mycophenolate Mofetil on Outcome by Race/Ethnicity
Figure 1. Change in Mycophenolate Mofetil Dose by Race/Ethnicity Over Time
Figure 2. Change in Mycophenolate Mofetil Discontinuation or Hold Rates by Race/Ethnicity Over Time
Figure 3. Change in Tacrolimus Levels by Race/Ethnicity Over Time