Interobserver variability in the pathologic interpretation of endomyocardial biopsy results

Abstract

Controversy exists over the role of endomyocardial biopsy in evaluating patients with dilated cardiomyopathy, particularly in detecting myocarditis and in assessing prognosis. Interobserver variability, if high, could explain conflicting reports. To assess this possibility, we submitted biopsy specimens from 16 patients with dilated cardiomyopathy to seven cardiac pathologists. The same slides were independently reviewed by each and assessed for fibrosis, hypertrophy, nuclear changes on a 0 to 3+ scale, mean lymphocyte count per high-power field, and myocarditis. The prevalance of significant fibrosis ranged from 25% to 69%, hypertrophy from 19% to 88%, nuclear changes from 31% to 94%, and abnormal lymphocyte count from 0 to 38%. One or more pathologists diagnosed definite or possible myocarditis in 11 of the 16 patients. Of these 11 patients, three pathologists agreed about three and two pathologists agreed about five. Myocarditis was diagnosed by a single pathologist in three cases. We conclude that interobserver variability is high in interpreting biopsy specimens from patients with dilated cardiomyopathy and that quantitative and standardized methods are needed to increase diagnostic consistency.