Clinical characteristics, immunomodulatory treatments and factors associated with fatal outcomes in immune checkpoint inhibitor–related myositis–myocarditis–myasthenia overlap syndrome (IM3OS): a patient-level systematic review with an illustrative case

Objective To characterize the clinical characteristics, immunomodulatory treatment patterns, and factors associated with fatal outcomes in immune checkpoint inhibitor–related myositis–myocarditis–myasthenia overlap syndrome (IM3OS). Methods This PRISMA 2020-guided systematic review searched PubMed, Web of Science, and CNKI through July 20, 2026 (PROSPERO: CRD420261378477). Patient-level data from case reports and series were analyzed using univariate analyses and exploratory binary logistic regression, with peak creatine kinase included in sensitivity analyses. Immunomodulatory therapies beyond corticosteroids were also summarized. Results We included 107 reports comprising 129 patients. Median onset was 21 days after initial immune checkpoint inhibitor administration, and 89.9% developed IM3OS within the first two treatment cycles. IM3OS-related fatal outcomes were reported in 47 patients (36.4%). Among 114 patients with information on additional immunomodulatory treatment, intravenous immunoglobulin and plasma exchange were reported in 91 and 44 patients, respectively; conventional immunosuppressants and mechanism-targeted therapies were also used. In model 1 (n=86), older age (OR per year, 1.119; 95% CI, 1.049–1.194) and higher peak cardiac troponin I (OR per tenfold increase, 2.427; 95% CI, 1.182–4.984) were associated with fatal outcomes. Associations persisted in model 2 (n=55), whereas the association with methylprednisolone pulse-dose category remained uncertain (≥1,000 versus 500 mg/day: OR, 2.979; 95% CI, 0.517–17.158). Peak creatine kinase was not independently associated with outcome. An illustrative pembrolizumab-associated case in recurrent endometrial cancer demonstrated that biochemical improvement preceded functional recovery. Conclusion Older age and higher peak cardiac troponin I were associated with fatal outcomes in exploratory analyses. Optimal methylprednisolone pulse dosing and the relative efficacy of different immunomodulatory therapies remain undetermined. Early multisystem assessment and individualized immunomodulatory management are warranted. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420261378477 , identifier CRD420261378477.