Efficacy and safety of subsequent pharmacological therapies for previously treated adults with primary immune thrombocytopenia: a network meta-analysis and certainty-of-evidence assessment

This is the authors' abstract. We don't add key points for this paper.

Background Adult primary immune thrombocytopenia (ITP) often relapses after first-line corticosteroid therapy, but comparative evidence for subsequent pharmacological therapies in previously treated adults remains fragmented. Objective To compare the relative efficacy and safety of currently available subsequent pharmacological therapies for previously treated adults with primary immune thrombocytopenia and to assess the certainty of evidence for clinically relevant treatment comparisons using a network meta-analysis framework. Methods We performed a systematic review and network meta-analysis of randomized controlled trials in adults with persistent or chronic primary ITP who had previously received at least one ITP-directed treatment and were receiving subsequent pharmacological therapy. PubMed, Embase, CENTRAL, and Web of Science were searched from inception to July 1, 2026. Eligible interventions included eltrombopag, romiplostim, avatrombopag, hetrombopag, fostamatinib, rituximab, and recombinant human thrombopoietin (rhTPO). The primary outcome was platelet response (platelet count ≥50 × 10 ^ 9/L). Secondary outcomes were bleeding, rescue therapy, any adverse event, and serious adverse events. Random-effects network meta-analysis estimated risk ratios (RRs) with 95% confidence intervals (CIs), and certainty of evidence was assessed using CINeMA. Results Twenty-two randomized trials involving 2,753 participants and eight treatment nodes were included. Compared with placebo, rhTPO (RR 5.34, 95% CI 2.39 to 11.92) and avatrombopag (RR 4.67, 95% CI 2.10 to 10.39) showed the largest improvements in platelet response, while romiplostim, hetrombopag, and eltrombopag also favored treatment. Fostamatinib was associated with a lower incidence of any reported bleeding event (RR 0.24, 95% CI 0.07 to 0.79; high-certainty evidence). rhTPO reduced rescue therapy use (RR 0.18, 95% CI 0.07 to 0.46; high-certainty evidence), followed by hetrombopag and eltrombopag. For any adverse event, no treatment showed a clearly reduced risk vs. placebo, although rhTPO ranked most favorably. For serious adverse events, eltrombopag (RR 0.39; 95% CI, 0.26–0.58), hetrombopag (RR 0.46; 95% CI, 0.27–0.79), and romiplostim (RR 0.65; 95% CI, 0.47–0.88) were associated with lower risks than placebo. Estimates for rhTPO and several other treatments, however, were imprecise. Conclusions Different subsequent pharmacological therapies appear to offer outcome-specific advantages in previously treated adults with primary ITP, but no single therapy dominated all endpoints. Treatment selection should integrate efficacy, safety, and certainty of evidence rather than ranking alone. Systematic Review Registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420261455143 , identifier PROSPERO CRD420261455143.