PGI2 - Bayesian Hierarchical modeling of randomized and non-randomized studies comparing ciclosporin with immediate- and prolonged-release tacrolimus in Liver Transplant Recipients

Abstract

Several meta-analyses comparing ciclosporin with immediate-release (IR) tacrolimus have been conducted. Only recently have sufficient data become available for evidence synthesis including prolonged-release (PR) tacrolimus. In the present study, a network meta-analysis (NMA) was conducted to compare the efficacy of ciclosporin, IR and PR tacrolimus in liver transplant recipients. Systematic literature review of PubMed, EMBASE and the Cochrane Library identified randomized controlled trials and large-scale observational studies (≥500 patients) published since January 2000 comparing IR tacrolimus with PR tacrolimus or ciclosporin following primary liver transplant. A Bayesian NMA was conducted to evaluate likelihood of death, graft loss, and acute rejection (AR) at 12 months. Outcomes were adjusted for recipient gender and age, hepatitis C (HCV) status, hepatocellular carcinoma, mycophenolate mofetil, azathioprine, and steroid use. Head-to-head study results were available for ciclosporin versus IR tacrolimus (n=14) and PR tacrolimus versus IR tacrolimus (n=2). Relative to ciclosporin, IR and PR tacrolimus were associated with reduced likelihood of death within 12 months of transplant (median odds ratios [OR] of 0.78 and 0.60, respectively). Mortality outcomes were superior with PR versus IR tacrolimus (median OR of 0.76). AR was less common with IR tacrolimus compared with ciclosporin (median OR of 0.69), whereas limited data for PR tacrolimus was evidenced by large credible intervals. There were no significant differences with respect to graft loss. Sensitivity analyses indicated that results were robust to assumptions and changes in confounders. Confounder analysis found that patient age was associated with death, whereas HCV status was associated with AR and graft loss. Patient and treatment factors beyond immunosuppressive treatment influence outcomes. Results of this confounder-controlled analysis provide further evidence that tacrolimus has substantial benefits over ciclosporin in liver transplant recipients. Compared with IR tacrolimus, data indicate that PR tacrolimus is superior with respect to patient survival.