Effect of rituximab dose on induction therapy in ABO-incompatible living kidney transplantation: A network meta-analysis.

Background: Rituximab is an induction immunosuppressant essential for ABO-incompatible kidney transplantation (ABOi KT). However, studies on its dosing, which differs among countries and transplant centers, are lacking. Therefore, we retrospectively investigated the effectiveness of the induction dose of rituximab against patient mortality, graft failure, and adverse events.

Methods: We included the studies referring to at least 2 of eligible induction doses (200 mg, 200-500 mg, or 500 mg) of rituximab during ABOi KT and relevant outcomes such as patient survival, graft failure, and bacterial and viral infections. We performed direct and indirect network meta-analyses using Bayesian models and ranked different rituximab doses using generation mixed treatment comparison. Publications were retrieved using CENTRAL, MEDLINE, EMBASE, and Science Citation Index Expanded databases from 1970 to February 2020 and analyzed. The GRADE of network meta-analysis approach specified 4 levels of certainty for a given result: high, moderate, low, and very low.

Results: Among the 4256 patients from 21 trials, glomerular filtration rate, graft loss, antibody-mediated rejection, T-cell mediated rejection, fungal infection, bacterial infection, and CMV infection did not differ among ABOi groups treated with different rituximab doses. The effect on mortality was significantly higher in rituximab 200 to 500 mg, and rituximab 500 mg groups (odds ratios [OR] 3.5, 95% CrI: 1.3-9.8, and OR 3.0, 95% CrI 1.1-9.8), but not in rituximab 20 mg group (OR 0.45, 95% CrI 0.036-2.5). The incidence of BK virus was significantly lower in the rituximab 200-mg group than in the other groups.

Discussion: In ABO-incompatible kidney transplantation, low-dose rituximab is more efficacious than higher doses and reduces serious infection risks. Additional randomized controlled trials might be needed to confirm these findings due to small sample size.