Comparative efficacy and safety of targeted disease-modifying therapies in adults with primary IgA nephropathy: a systematic review and network meta-analysis

Background: Targeted disease-modifying therapies for adults with primary IgA nephropathy are increasing in number. However, head-to-head evidence is limited, and trial endpoints and follow-up durations vary. The relative effects of these therapies on short-term proteinuria control, long-term kidney outcomes, and safety remain unclear. Methods: We searched databases and trial registries for randomized, double-blind, placebo-controlled trials with a planned treatment duration of at least 24 weeks alongside stable or optimized supportive care. Frequentist random-effects network meta-analyses evaluated baseline-adjusted geometric mean ratios (GMRs) for 24-hour urinary protein-to-creatinine ratio at 32 to 40 and 24 to 52 weeks, with Holm correction applied. Total estimated glomerular filtration rate (eGFR) slope and composite kidney outcomes were synthesized using common-effect models. Serious adverse events (SAEs) were analyzed using penalized-likelihood logistic regression within risk windows. Sensitivity analyses included restriction to evidence at low risk of bias, exclusion of interim or stage-specific reports, and assessment of precision-recovery assumptions and time-window boundaries. The Mantel to Haenszel method without continuity correction assessed the robustness of sparse SAE results. RoB 2 and CINeMA assessed risk of bias and evidence certainty, respectively. Results: Seventeen trials were included. At 32 to 40 weeks, telitacicept and subcutaneous sibeprenlimab had the lowest GMR point estimates versus placebo: 0.450 (nominal 95% CI 0.323 to 0.628) and 0.488 (0.355 to 0.671), respectively. None of the seven targeted treatment comparisons reached statistical significance after Holm correction. Iptacopan had the largest point estimate for improvement in annualized total eGFR slope: 3.020 mL/min/1.73 m2/year (95% CI 2.025 to 4.015). Composite kidney outcome hazard ratios were 0.570 (95% CI 0.401 to 0.811) for iptacopan and 0.440 (0.271 to 0.714) for targeted-release budesonide. Telitacicept had the lowest SAE odds ratio point estimate at 36 to 39 weeks: 0.314 (95% CI 0.105 to 0.938), although safety estimates were imprecise. In sensitivity analyses, proteinuria effect directions were generally consistent, but the 95% CIs for all retained comparisons included 1 after excluding interim or stage-specific reports; sparse SAE estimates remained imprecise. Evidence certainty was rated as low or very low. Conclusion: Although some regimens yielded the most favorable point estimates, the available evidence remains insufficient to establish treatment priorities or comparative safety advantages. Head-to-head trials are needed.