SUMMARY To define the range of cardiac involvement in the Churg-Strauss syndrome, M mode, continuous wave Doppler, and cross sectional echocardiograms were recorded in twelve patients with the disorder. The M mode recordings were digitised and the cross sectional images were recorded with standardised gain settings to determine regional myocardial echo amplitude. Left ventricular end diastolic and end systolic dimensions were increased above the normal 95% confidence interval in four patients, three of whom showed a depressed shortening fraction. Mitral regurgitation was present in six patients; this was severe enough to need valve replacement in two. Mean echo amplitude in both the septum and the posterior wall was significantly increased above normal by a mean (SD) of4-87 (2-57) dB, suggesting the presence ofmyocardial fibrosis. There was no evidence of subendocardial involvement as there is in other hypereosinophilic syndromes.Mitral regurgitation is common in the Churg-Strauss syndrome. This cannot be ascribed to involvement ofthe cusps or chordae and it occurs even when ventricular function is well preserved. It is suggested that mitral regurgitation is caused by diffuse myocardial fibrosis.The Churg-Strauss syndrome is recognised by the combination of blood eosinophilia, eosinophilic tissue infiltration, extravascular fibrinoid necrotising (allergic) epithelioid and eosinophilic granuloma formation, and disseminated necrotising vasculitis occurring in asthmatic patients.' 2 Lanham et al noted that "the histological components of the syndrome often do not coexist temporally or spatially" and suggested that the clinical pattern of a prodromal phase of allergic disease, followed by a second phase characterised by eosinophilic tissue infiltration and peripheral blood eosinophilia culminating in a third vasculitic phase, should supercede strict adherence to histopathological criteria for the purposes of diagnosis.' Myocardial fibrosis, particularly in the subendocardial region, is a characteristic feature of the idiopathic hypereosinophilic syndrome4 and it has been attributed directly to the eosinophilia itself. Possible cardiac involvement in patients with ChurgStrauss syndrome and comparable blood eosinophilia, however, is less well defined. We therefore studied a series of patients with Churg-Strauss syndrome by M mode, continuous wave Doppler, and cross sectional echocardiography to assess whether they were also at risk for heart disease.Requests for reprints to Dr J M Morgan, National Heart Hospital, London WIM 8BA.Accepted for publication 6 April 1989 Patients and methodsWe studied 12 patients in whom the diagnosis of Churg-Strauss syndrome was made on the basis of histological criteria" and clinical features.' The blood eosinophil count was raised in all (eosinophil counts ranged from 2-3 to 11 x 109 1 at presentation). The results of echocardiography were compared with those in 20 age and sex matched controls with no clinical evidence of heart disease in whom stress test had been negative at high wo...