Identifying Carditis in Acute Rheumatic Fever With Myocardial T1 Mapping

Abstract

Background: The pathogenesis of acute rheumatic fever (ARF) is incompletely understood. Valvulitis is used as part of the diagnostic criteria, while pancarditis has only been demonstrated at autopsy. Purpose: To identify carditis in patients with ARF using T1 mapping on cardiac magnetic resonance (CMR) imaging and echocardiographic global longitudinal strain (GLS). We hypothesised that prolonged native T1 time, consistent with myocardial inflammation, would be present in ARF patients compared to controls. Methods: We prospectively recruited 36 patients; 15 fulfilling the Australian modified Jones Criteria for the diagnosis of ARF, 12 controls with an inflammatory condition without major criteria for ARF and 9 healthy controls. All patients underwent CMR with assessment of non-contrast myocardial T1 mapping as well as echocardiography for GLS. The primary outcome was differences in T1 time between the groups. Results: Patients with ARF had evidence of carditis demonstrated by markedly elevated T1 times [976(±167 ms)] compared to those with non-cardiac inflammatory conditions [841(±50 ms)] and healthy controls [811(±54 ms)], P = 0.002 (Figure 1). There was no difference in global longitudinal strain between the groups (−19.4% vs −19.5% vs −20.4%, P = 0.66). Conclusions: Patients with ARF have markedly elevated myocardial T1 times on CMR, consistent with active inflammation from carditis. In contrast, GLS was normal in ARF patients. Incorporating CMR T1 mapping into the diagnostic algorithm for ARF may improve diagnostic certainty and lead to more effective delivery of secondary penicillin prophylaxis.