Risk models for ICI-associated myocarditis occurrence and severity in patients receiving immune checkpoint inhibitors

This retrospective unmatched case-control study examined risk factors and predictive models for ICI-associated myocarditis in a cohort of 196 patients receiving immune checkpoint inhibitors. The population consisted of 98 confirmed myocarditis cases and 98 controls. The study aimed to identify baseline characteristics associated with the occurrence and severity of this condition.

The primary analysis assessed a baseline model incorporating an elevated eosinophil ratio, reduced lymphocyte ratio, and elevated myoglobin. This model demonstrated an area under the curve (AUC) of 0.699 (95% CI, 0.626-0.772) for predicting the occurrence of ICI-associated myocarditis. For predicting severe myocarditis, a combined ECG and enzymatic signature at onset was evaluated, yielding an AUC of 0.769 (95% CI, 0.655-0.882).

Safety, tolerability, discontinuations, and adverse events were not reported in this study. The study phase and setting were not specified. Key limitations include the retrospective unmatched design, which may introduce selection bias, and the absence of absolute numbers for specific risk stratification thresholds. The study was not funded by industry sources, though specific conflicts were not detailed.

While a two-tiered risk stratification was established, early identification of high-risk patients remains challenging. Clinicians should interpret these AUC values as indicative of moderate predictive performance rather than definitive diagnostic tools. The lack of reported safety data limits the ability to assess the impact of these risk factors on patient management or treatment decisions.