Abstract
<title>Abstract</title> <p>Purpose Several mechanistically distinct therapies now have randomized evidence in primary IgA nephropathy, but no trial has compared two active agents directly. We estimated short-term antiproteinuric effects by mechanism class and defined how far those estimates can be interpreted comparatively. Methods We searched MEDLINE/PubMed, Europe PMC, ClinicalTrials.gov and other public registries, trial supplements, protocols and statistical analysis plans through 24 June 2026. Parallel-group randomized trials in adults with biopsy-confirmed primary IgA nephropathy were eligible when a treatment-to-control urine protein-to-creatinine ratio (UPCR) geometric mean ratio (GMR) was extractable at 26–39 weeks. Effects were analysed on the log-GMR scale in a Bayesian random-effects model with mechanism class as the node. Risk of bias was assessed with RoB 2 and confidence with CINeMA. Results Ten randomized contrasts (2,036 participants) informed four mechanism classes and nine agent nodes. APRIL/BAFF-axis therapy showed the largest modelled UPCR reduction (GMR 0.50, 95% credible interval 0.44–0.56), followed by endothelin-pathway therapy (0.62, 0.54–0.71), complement inhibition (0.65, 0.55–0.77) and targeted-release corticosteroid (0.73, 0.61–0.87). Seven trials were at low risk of bias. CINeMA confidence was moderate for class-versus-control estimates and low to very low for every active-active contrast, all of which are indirect. Conclusion APRIL/BAFF-axis therapy produced the largest short-term antiproteinuric signal available to June 2026. This is a class-level indirect estimate, not evidence of head-to-head superiority or of within-class interchangeability. Mature kidney-outcome data remain concentrated in nefecon, sparsentan, atrasentan and iptacopan, and treatment selection should weigh that asymmetry.</p>