Clozapine-induced myocarditis: characterisation using case-control design

Abstract

Purpose: Myocarditis is a hypersensitivity reaction, typically occurring in the third week after commencing clozapine, the most effective treatment available for schizophrenia. The signs and symptoms of myocarditis are notoriously difficult to distinguish from those associated with initiation of clozapine. A case-control design is commonly used to investigate risk factors, we show that it can be used to identify features which assist in the diagnosis of myocarditis. Methods: Cases and controls were documented from patients' medical records. Controls were matched by unit at which clozapine was commenced and approximate start date. Results: 105 cases and 296 controls met entry criteria. Time to onset for cases was 10-33 days, with 82% developing 14-21 days after commencing clozapine. Almost 90% of cases and controls had tachycardia. Eosinophilia developed in 64% of cases and 30% of controls, but onset among cases was delayed 0-8 days after the peak in troponin. However, 87% of cases had C-reactive protein (CRP) > 50mg/L and CRP could be raised up to 5 days before the rise in troponin. Multivariate regression analysis indicated that the risk of myocarditis increased with increasing age (31% per decade; 95% CI 7-60%), increasing rate of clozapine dose titration (26% per 250mg during days 1-9; 95% CI 2-55%) and concomitant sodium valproate (odds ratio 2.59; 95% CI 1.51-4.42). Conclusion: Comparison of cases and controls permitted identification of the features of myocarditis, and avoided confounding by features associated with introduction of clozapine. Monitoring for myocarditis should use troponin and CRP but not eosinophil counts. Clozapine should be introduced by slow dose titration and sodium valproate is best avoided, if clinically feasible.