Late-onset immune checkpoint inhibitor-associated myocarditis: A systematic review and patient-level synthesis

BACKGROUND: Immune checkpoint inhibitor (ICI)-associated myocarditis is usually considered an early toxicity, but late presentations remain poorly characterised. We aimed to describe the timing, clinical features, management and outcomes of late-onset ICI-associated myocarditis. METHODS: We conducted a PRISMA 2020 systematic review and patient-level synthesis of PubMed and Embase from inception to 25 August 2026. Late-onset myocarditis was defined as presentation ≥ 90 days after the first ICI dose. Borderline or diagnostically uncertain cases were examined in sensitivity analysis. RESULTS: Thirty-two studies were included; 31 provided extractable data for 39 potentially eligible cases. Thirty patients met primary-analysis criteria and nine were sensitivity-only. Median onset was 9.0 months (IQR 6.2-12.8); 6/30 (20.0%) presented at 3 to < 6 months, 13/30 (43.3%) at 6 to < 12 months, 8/30 (26.7%) at 12 to < 24 months and 3/30 (10.0%) at ≥ 24 months. Among patients with known treatment status, 8/21 (38.1%) presented after ICI discontinuation. Major arrhythmic or conduction complications occurred in 10/19 (52.6%), left ventricular ejection fraction was < 50% in 18/26 (69.2%), and mechanical circulatory support was required in 5/29 (17.2%). Nine of 29 patients died (31.0%), including three myocarditis/cardiac deaths. Among 19 patients with ascertainable recovery, 11 (57.9%) achieved complete recovery, four (21.1%) partial/incomplete recovery and four (21.1%) no recovery. CONCLUSIONS: Late-onset ICI-associated myocarditis can occur months to years after exposure, including after discontinuation, with heterogeneous severity.