Controlling treatment allocation bias in a registry analysis when comparing calcineurin inhibitors☆

Abstract 

Previous analyses of outcomes between immunosuppressive regimens included data from the early years of tacrolimus use and frequently included all Tacrolimus- or cyclosporine-based regimens. We now evaluate clinical outcomes associated with only the two most commonly prescribed regimens—Tacrolimus (Tac)-mycophenylate mofetil (MMF) and cyclosporine (CsA)-MMF—using recent data reported to the UNOS Scientific Renal Transplant Registry. Data from living donor kidney transplants was chosen to minimize selection bias between treatment groups. Outcomes are reported only for recent years (1998 to 1999 with 3-year follow-up) because acute rejection rates were markedly higher in 1995 to 1997 compared to 1998 to 2000 (38% to 68% vs 21% to 32%) and clinical practice has evolved since 1995, which would have biased results in favor of more recent immunosuppressive regimens. Three-year graft survival for patients transplanted from 1998 to 1999 was significantly higher in living donor kidney transplant patients receiving CsA-MMF (91.1%, n = 4686) versus Tac-MMF (88.1%, n = 2393) (P = .0006). After adjustment for potential confounding variables, risk of graft failure at 3 years was significantly higher in patients receiving Tac-MMF versus CsA-MMF for both all-cause graft failure (hazard ratio 1.28, 95%CI 1.09 to 1.49, P = .002) and death-censored graft failure (hazard ratio 1.25, 95%CI 1.05 to 1.49, P = .013). In view of the reduced rejection rate that has been reported using Tac-MMF versus CsA-MMF in clinical trials, it is possible that the nonimmunologic effects of calcineurin inhibitors may now play an increasingly important role in determining graft survival rates. In conclusion, this large-scale registry analysis demonstrates that graft survival in living donor kidney transplant patients is significantly improved using CsA-MMF compared to Tac-MMF.