Portal de Periódicos da CAPES

Cyclosporine Lowering With Everolimus or Mycophenolate to Preserve Renal Function in Heart Recipients: A Randomized Study

2010; Wolters Kluwer; Volume: 89; Issue: 2 Linguagem: Inglês

10.1097/tp.0b013e3181c42b95

1534-6080

Luciano Potena, Isidoro Giorgio Bianchi, Gaia Magnani, M. Masetti, Fabio Coccolo, Francesco Fallani, Antonio Russo, Francesco Grigioni, Angelo Branzi, Claudio Ponticelli,

Organ Transplantation Techniques and Outcomes

Calcineurin inhibitors (CNIs) represent the cornerstone of immunosuppressive therapy in heart transplant (HT) recipients. CNI nephrotoxicity, however, may cause irreversible renal failure, which is a major mortality risk factor (1–4). To reduce these risks, successful attempts to minimize or avoid CNI by introducing newer immunosuppressants, such as mycophenolate mofetil (MMF) and mammalian target of rapamycin (mTOR) inhibitors, have been reported (5–15). However, randomized studies comparing the safety and efficacy profile of CNI reduction under the umbrella of MMF or an mTOR inhibitor are lacking. METHODS This single-center randomized controlled study in stable HT recipients (registered at clinicaltrials.org, NCT00420537) was designed, conducted, and analyzed independently of the study drug manufacturers. The study objective was to demonstrate the possibility of limiting renal disease progression by reducing cyclosporine doses (Sandimunn Neoral, Novartis Pharma, Basel, CH) while adding MMF (Cellcept, Roche, Basel, CH) or the mTOR inhibitor everolimus (Certican, Novartis Pharma). Inclusion criteria comprised time from transplant between 1 and 4 years and calculated creatinine clearance (CrCl) between 20 and 60 mL/min (16). Study protocol was approved by the local ethical committee. After a baseline clinical and laboratory evaluation, patients were randomly assigned to receive either everolimus adjusted to 3 to 8 ng/mL blood trough levels (C0) and cyclosporine adjusted to 40 to 90 ng/mL C0 or MMF 1.5 g twice daily and cyclosporine adjusted to 100 to 150 ng/mL C0. Everolimus was started at 0.25 mg twice daily with concomitant 25% to 30% cyclosporine dose reduction and MMF or azathioprine withdrawal. After 5 to 7 days, white cell count and drugs C0 were checked, with further doses adjustment if needed. Allocation was balanced by the minimization method (17) for presence of baseline CrCl more than or equal to 40 mL/min, diabetes, and pretransplant coronary artery disease. The primary endpoint was the difference in CrCl between the two study groups. Assuming a 22% coefficient of variability of mean CrCl, a sample of 70 patients was needed to provide 80% power to detect a difference of 15% in CrCl between the study arms 12 months after randomization. Secondary endpoints were change of CrCl within and across the study arms, clinical and laboratory predictors of CrCl change, incidence of biopsy-proven acute rejection, and incidence of infection requiring specific antimicrobial treatment. RESULTS Thirty-four patients, transplanted 2.5±1.3 years before with baseline CrCl of 43.5±9.1 mL/min, were allocated to either everolimus (n=17) or MMF (n=17) arm. Six to 8 months before baseline, CrCl was 47.1±11.5 mL/min, indicating a progressive loss of renal function before study interventions (P=0.02 vs. baseline). Pulmonary infections caused one death in the MMF arm and one drug discontinuation in everolimus arm. Study enrollment was halted when only half of the planned population entered the study because of safety issues raised after 6 months of follow-up by 5 (29%) patients developing bacterial infections in the everolimus arm versus only one (6%) on MMF (P=0.04). However, at the end of study period, the estimated incidence of infection was similar in both arms (29±11 vs. 24±10%; P=0.5). Mean cyclosporine levels lowered by 65% in the everolimus arm (from 175±75 ng/mL to 60±32 ng/mL) and by 31% in the MMF arm (from 198±55 ng/mL to 136±34 ng/mL). After cyclosporine reduction, overall CrCl improved during the first 2 months (47.2±13.6 mL/min; P=0.03) and then stabilized toward the end of the study period (47.1±13.8 mL/min; P=0.1; all P values calculated for paired analysis). Both study interventions halted prestudy decline in CrCl, with no significant difference in CrCl at month 12 (49.8±13.4 vs. 44.5±13.8 mL/min; P=0.5; Fig. 1), suggesting overall efficacy of study interventions in reducing the progression of renal dysfunction.FIGURE 1.: (A) Time course of mean creatinine clearance (CrCl) in the two study groups. Asterisks indicate significant difference from baseline by Student's t test for paired data. Error bars indicate standard deviation. Horizontal dashed line indicates the level of baseline CrCl. Although CrCl initially improved only in mycophenolate mofetil (MMF) patients, CrCl decline was arrested in both arms, and at 12 months follow-up, CrCl was not significantly different between everolimus and MMF patients. Prestudy CrCl was obtained post hoc (gray lines). (B) Baseline to month 12 mean CrCl in the two study groups according to baseline proteinuria. Lack of proteinuria seem to identify patients who may gain significant benefit from everolimus introduction. P values were calculated by Student's t test for paired data. Error bars indicate standard deviation.A baseline proteinuria less than 150 mg/24 hr was the only independent predictor of CrCl improvement during the study period (odds ratio [95% confidence interval]=5.5 [1.05– 29.4], P=0.04). To explore the interaction of proteinuria with study interventions, we stratified the paired comparisons between baseline and month 12 CrCl within study arms. Proteinuria less than 150 mg/24 hr was significantly associated with CrCl improvement in the everolimus group (from 45.8±7.7 to 50.5±8.6; P=0.02) and with borderline significant CrCl improvement in the MMF arm (from 42.7± 6.9 mL/min to 47.9±8.1 mL/min; P=0.06). Conversely, proteinuria more than or equal to 150 mg/24 hr identified everolimus patients with numerically worsening renal function (Fig. 1). At study entry, 31 (91%) patients were on statin therapy. Although we found no significant difference in final low-density lipoproteins levels between study arms (104±38 mg/dL vs. 91±31 mg/dL; P=0.3), five patients on everolimus required an increase in statin dose. Moreover, everolimus intake was associated with a significant increase in triglycerides concentration (142±77 mg/dL vs. 204±95 mg/dL; P=0.01 for paired analysis). DISCUSSION This randomized study shows the possibility of halting the progression of renal dysfunction in midterm HT recipients through the introduction of everolimus or MMF and concomitant cyclosporine reduction. Despite a greater cyclosporine reduction in everolimus patients, at the end of the study, CrCl was similar between study groups. In addition, exploratory analyses suggest that an even minor proteinuria can identify HT recipients unlikely to respond to mTOR conversion, as already reported in kidney and lung transplant recipients (18–20). Study enrollment was interrupted because an increased number of bacterial infections during the initial period after everolimus start. However, infectious episodes were later balanced in the MMF arm with no final difference between the groups. On the basis of this experience, we believe that the use of low-dose antibacterial prophylaxis for urinary and lung infection and a strict clinical monitoring during the first months after everolimus introduction may be appropriate. The major limitation of this study is the reduced sample size that did not make it possible to reach the planned power to detect differences in renal function in the study groups. Thus, most of the findings should be considered only hypothesis-generating because deriving from secondary endpoint analysis. However, to our knowledge, this is the first randomized study testing cyclosporine reduction with everolimus or MMF for progression of CNI-dependent renal dysfunction in HT recipients, and we believe that it may provide some clinical hints in the choice for the optimal kidney sparing strategy. Indeed, within analogous overall efficacy of both treatments in reducing CNI nephrotoxicity, MMF showed a smoother safety profile than everolimus. Conversely, lack of proteinuria may help to select optimal candidates to everolimus-based CNI minimization, who may also take advantage of the potential benefits of everolimus antiproliferative properties (21–23). ACKNOWLEDGMENT The authors thank the nursing team of the Bologna Heart Transplant Center for the help in sample collections and study visits organization. Luciano Potena Isidoro G. Bianchi Gaia Magnani Marco Masetti Fabio Coccolo Francesco Fallani Antonio Russo Francesco Grigioni Angelo Branzi Cardiovascular Department University of Bologna Bologna, Italy Claudio Ponticelli Division of Nephrology IRCCS Humanitas Milan, Italy