Comparative efficacy of complement inhibitors in complement Inhibitor–Naïve PNH: A network meta-analysis of randomized trials

Abstract Introduction: Paroxysmal nocturnal hemoglobinuria (PNH) is an uncommon, life-threatening disease, caused by intravascular hemolysis by the complement system. The management plan focuses on complement inhibition but, no randomized trial had directly compared all the available agents. In this review, we aim to conduct an indirect non-head-to-head network meta-analysis of the existing randomized trials to compare the efficacy of the available agents across patient-centered outcomes in complement inhibitor-naive patients. Methods: Following PRISMA guidelines,A comprehensive literature search was conducted on PubMed, and CENTRAL for studies published up to April 2025 using keywords and MeSH terms related to 'paroxysmal nocturnal hemoglobinuria’, 'complement inhibitors’, and 'outcomes’. Eligible studies included double-arm randomized trials comparing complement inhibitor agents with supportive care or other inhibitor agents in patients with complement inhibitor naive status. Data on baseline demographics, transfusion avoidance, and FACIT fatigue score were extracted. The risk of bias was assessed using the ROB2 tool. A frequentist model network meta-analyses were conducted in RStudio (v5.4.1) using a common-effects model. Results: A total of 4 randomized controlled trials evaluating 4 complement inhibitor agents (Ravulizumab, Crovalimab, Eculizumab & Pegcetacoplan) were included in this meta-analysis, involving 589 complement inhibitor–naïve adults with PNH. The age of the participants ranged from 17 to 78. The proportion of male patients was 52.6% (n = 310/589). Among 502 patients across three trials, 60.8% were Asian (n = 305/502), 2.3% were Black (n = 12/502), and 18.7% were White (n = 94/502). A total of 353 out of 502 patients (70.3%) received ≥1 unit of packed red blood cells in the 12 months prior to screening. Across the included studies, 107 of 343 patients (31.2%) had a history of aplastic anemia (3 studies), 90 of 536 (16.8%) had a history of major vascular events (3 studies), and 14 of 290 (4.8%) had myelodysplastic syndrome (2 studies). Every trial reported outcomes at a uniform 26-week timepoint. There have been minor variations in definitions of the existing outcomes, core endpoints were largely comparable across studies, supporting robust indirect comparisons. All active treatments demonstrated superiority over placebo across the evaluated outcomes, displaying both clinical benefits and outcome-specific strengths widely. In the network meta-analysis of transfusion avoidance comprising four randomized controlled trials, Pegcetacoplan demonstrated the greatest odds compared to placebo [OR = 181.33, 95% CI: 17.50–1879.39], followed by Ravulizumab [OR = 133.08, 95% CI: 7.31–2422.11], Eculizumab [OR = 93.14, 95% CI: 5.39–1609.09], and Crovalimab [OR = 49.49, 95% CI: 2.65–925.04]. There were no statistically significant differences between the active treatments as all pairwise comparisons yielded overlapping confidence intervals, except Ravulizumab which showed a statistically significant advantage over Crovalimab [OR = 2.69, 95% CI: 1.13–6.41; p = 0.0256]. For the change in FACIT fatigue score, all active treatments demonstrated improvement over placebo. Crovalimab showed the greatest benefit [MD = +13.00, 95% CI: 8.46–17.54], followed by Ravulizumab [MD = +11.07, 95% CI: 6.56–15.58], Eculizumab [MD = +10.40, 95% CI: 6.30–14.50], and Pegcetacoplan [MD = +4.50, 95% CI: –0.20 to 9.20]. Among these, Crovalimab, Ravulizumab, and Eculizumab were statistically superior to placebo, while Pegcetacoplan did not achieve statistical significance. Upon comparison of active treatments with each other, Crovalimab demonstrated significantly greater improvement compared to Pegcetacoplan [MD = +8.50, 95% CI: 1.97–15.03] and Eculizumab [MD = +2.60, 95% CI: 0.65–4.55]. However, there was no statistically significant difference found when compared to Ravulizumab[MD = +1.93, 95% CI: –0.78 to 4.64]. Conclusion: We found no single agent being consistently superior to others across all clinically relevant outcomes. Notably, the treatment of choice should be individualized based on the goals of care and priorities of the patients such as transfusion independence or quality of life. Post-market real-world analysis and comparison of these agents may guide optimal sequencing or cost-effective strategies in the management of PNH.