Hypereosinophilic syndrome: endomyocardial biopsy versus echocardiography to diagnose cardiac involvement

Objective: To compare echocardiograms and endomyocardial biopsies to diagnose cardiac involvement in hypereosinophilic syndrome. Methods: We examined the agreement between echocardiography and endomyocardial biopsies to detect cardiac involvement in hypereosinophilic syndrome by reviewing cases identified as hypereosinophilia or hypereosinophilic syndrome in Mayo Clinic databases from January 1978 through June 2009. Single-organ cases of eosinophilia such as eosinophilic fasciitis and eosinophilic gastroenteritis were excluded. We recorded echocardiogram and endomyocardial biopsy results including biopsy staining for eosinophil granule major basic protein (if performed). Clinical and laboratory features documented included presenting symptom(s), maximum total eosinophil count, dose of prednisone (if any) and eosinophil count at the time of endomyocardial biopsy, cardiac enzymes, serum tryptase level, electrocardiogram result, the result of testing for the FIP1L1-PDGFRA fusion gene, complications associated with the biopsy procedures and available follow-up information. Results: From a total of 387 patients’ records screened 288 met the criteria for hypereosinophilic syndrome and of these 240 had echocardiograms. Among these patients there were 138 normal echocardiograms, 67 had echocardiograms without findings of hypereosinophilic syndrome but with one or more other abnormalities, and 35 had echocardiograms with findings consistent with hypereosinophilic syndrome. Twenty-five patients from this group of 35 patients had both echocardiogram and endomyocardial biopsy. In 15 patients there was agreement between both endomyocardial biopsy and echocardiography as to the presence (n = seven) or absence (n = eight) for findings of cardiac involvement. In 10 of 25 patients test results diverged: 3 patients with positive echocardiographic changes did not have confirmatory findings by endomyocardial biopsy and seven patients with positive biopsy findings had echocardiograms without findings of hypereosinophilic syndrome. Conclusions: Echocardiograms and endomyocardial biopsies agree for presence or absence of cardiac involvement 60% of the time. Endomyocardial biopsy detected cardiac involvement in 7 patients in whom the echocardiogram was negative for findings of hypereosinophilic syndrome.