Induction therapy in heart transplantation: a systematic review and network meta-analysis

The utility of induction therapy (IT) in heart transplantation remains contested. Commissioned by a clinical-practice guidelines panel to evaluate the effectiveness and safety of IT in adult heart transplantation patients, we conducted this systematic review and network meta-analysis (NMA) to assess the short and long-term outcomes. We searched bibliographic databases for studies published from January 2000 to October 2022 that enrolled adult (at least 95% of the population ≥18 years) de novo, single-organ heart transplantation recipients, and reporting on the use of any IT agent and the impact on any post-transplantation outcome. Based on patient-important outcomes informed by our patient partners and clinicians, we performed frequentist NMAs separately for RCTs, and observational studies with adjusted analyses. We determined absolute effect estimates using MAGICapp, by determining baseline risks reported in the studies and used the GRADE framework to assess the certainty of evidence and summarize the findings. From 5,156 publications identified, we included 7 RCTs, and 12 observational studies evaluating outcomes using multivariable analysis (Cox proportional hazards models or logistic regression models), and report on the use of two contemporarily-used IT agents – basiliximab and rATG. The RCTs provide only very low certainty evidence, and thus proved uninformative of the effect of the two agents versus no IT or one another. With low certainty in the evidence from observational studies, basiliximab may increase 30-day (OR 1.13; 95%CI 1.06-1.20) and 1-year (OR 1.11; 95%CI 1.02-1.22) mortality compared to no IT. With low certainty from observational studies, rATG may decrease 5-year cardiac allograft vasculopathy (OR 0.82; 95%CI 0.74-0.90) compared to no IT, as well as 30-day (OR 0.85; 95%CI 0.80-0.92), 1-year (OR 0.87; 95%CI 0.79-0.96), and overall (HR 0.84; 95%CI 0.76-0.93) mortality compared to basiliximab. Tables 1a, 1b, and 1c summarize the network estimates, absolute effect estimates, and certainty of evidence. With low and very low certainty in the synthetized evidence, these NMAs provide no compelling evidence for or against the routine use of IT, and highlight the need for future studies. Low certainty evidence suggests possible superiority of rATG versus basiliximab.