Component Network Meta-Analysis of Therapies for Kawasaki Disease

Background: Kawasaki disease (KD) treatment combines intravenous immunoglobulin (IVIG), corticosteroids, infliximab (IFX), and aspirin (ASA), yet the independent contribution of each component to coronary artery aneurysm (CAA) prevention remains unclear. Traditional pairwise meta-analyses cannot separate the effects of individual treatment components within multidrug regimens. We performed a component network meta-analysis (cNMA) to quantify the independent effects of IVIG, steroid regimens, IFX, and ASA dose on CAA risk and rank treatment strategies. Methods: We searched Google Scholar, Scopus, Embase, and PubMed from inception to March 31, 2026. Aggregate data from 10 studies (21 treatment arms, 7,083 patients, 1,020 CAA events) were analyzed using frequentist cNMA with additive component modeling to estimate relative risks (RRs) versus standard IVIG plus high-dose ASA. SUCRA values ranked regimens. Global inconsistency, steroid dose-response, subgroup analyses by Kobayashi risk score (≥4 vs <4) and ethnicity, and comparisons of IFX versus second-dose IVIG in IVIG-resistant patients were performed. Results: Compared with standard IVIG plus high-dose ASA, prolonged oral prednisolone reduced CAA risk by 72% (RR 0.28, 95% CI 0.21-0.37), pulse intravenous methylprednisolone (IVMP) by 50% (RR 0.50, 95% CI 0.34-0.74), and IFX by 67% (RR 0.33, 95% CI 0.20-0.54). Low-dose ASA showed no significant difference versus high-dose ASA (RR 1.08, 95% CI 0.86-1.36), while moderate-dose ASA trended toward benefit (RR 0.82, 95% CI 0.72-1.25). IVIG plus prolonged steroid plus high-dose ASA ranked highest (SUCRA 0.91), followed by IVIG plus prolonged steroid plus IFX plus high-dose ASA (0.89), IVIG plus IFX plus high-dose ASA (0.78), IVIG plus pulse IVMP plus high-dose ASA (0.72), and standard IVIG plus high-dose ASA (0.50). IVIG alone ranked poorly (0.22), and ASA monotherapy ranked lowest (0.08). In high-risk patients (Kobayashi score ≥4), prolonged steroids reduced CAA risk by 80% (RR 0.20) versus 35% (RR 0.65) in low-risk patients. Japanese cohorts demonstrated larger steroid effects (RR 0.22) than non-Japanese cohorts (RR 0.48). Among IVIG-resistant patients, IFX reduced CAA risk by 33% versus second-dose IVIG (RR 0.67). Conclusions: Component network meta-analysis reveals that prolonged oral corticosteroids provide the largest independent CAA risk reduction in KD, with greatest benefit in high-risk Japanese patients. Infliximab demonstrates substantial efficacy as both primary adjunctive and rescue therapy for IVIG resistance. Aspirin dose does not significantly affect CAA outcomes, challenging traditional high-dose ASA protocols. These findings support risk-stratified, component-optimized strategies: IVIG plus prolonged prednisolone for high-risk patients, IVIG plus IFX for IVIG-resistant cases, and low-dose ASA for antiplatelet coverage.