OBJECTIVES: This network meta-analysis (NMA) compared thrombopoietin receptor agonists (avatrombopag [AVA] and lusutrombopag [LUS]) in patients with chronic liver disease (CLD) and severe thrombocytopenia undergoing invasive procedures.

METHODS: A systematic literature review identified randomised, controlled trials (RCTs) assessing AVA or LUS in CLD. For the NMA, efficacy outcomes included: proportions of patients requiring ≥1 platelet transfusion, rescue procedures for bleeding events, or rescue transfusions for bleeding events; response rates (% patients not requiring platelet transfusions or any rescue procedure for bleeds ≤7 days post-procedure); and platelet counts ≥50×10⁹/L. Safety outcomes included risks of portal vein thrombosis, serious adverse events (AEs) and any AE. Data were analysed in a Bayesian framework and results reported as odds ratios (ORs) with 95% credible intervals (CrIs).

RESULTS: The NMA included a total of six phase II or III placebo-controlled RCTs (three for each drug). AVA and LUS provided statistically significant differences versus placebo for all efficacy endpoints, except the proportion of patients requiring a rescue transfusion due to bleeding events (AVA and LUS) and the proportion requiring a rescue procedure for bleeding events (LUS only). Rescue transfusions were numerically less frequent in both active groups compared with placebo (OR=0.5 for both), but the differences were not significant due to low frequency of this event. Compared with LUS, significantly more patients receiving AVA achieved a platelet count ≥50×10⁹/L (OR=3.40 [1.41–8.34]); there were no significant between-treatment differences for the other efficacy outcomes. No significant differences were observed for AVA versus placebo or versus LUS for any of the safety outcomes.

CONCLUSIONS: AVA compared with LUS provided a higher likelihood of achieving a target platelet count of ≥50×10⁹/L and was similar for other efficacy outcomes in patients with CLD and severe thrombocytopenia undergoing invasive procedures. The safety profile of AVA was comparable to placebo and LUS.