Consultation in breast surgical pathology: Interobserver diagnostic variability of atypical intraductal proliferative lesions

PURPOSE: To evaluate the agreement about the histopathological diagnosis of intraductal proliferative breast lesions between general pathologists and a specialist in breast pathology. METHODS: This was an observational, cross-sectional study of 209 lesions received in consultation at the Breast Pathology Laboratory of the School of Medicine, Federal University of Minas Gerais, from 2007 to 2011, comparing the original diagnosis and the review. We included only cases with a formal request for review and cases in which the original diagnosis or reviewer's diagnosis showed proliferative lesions, pure ductal carcinoma in situ, ductal carcinoma in situ associated with microinvasion or associated with invasive carcinoma. The kappa index and percent concordance were used in the statistical analyses. RESULTS: A moderate agreement was observed between the original histopathological diagnosis and the second opinion (kappa=0.5; percentual concordance=83%). After the review, the diagnosis of malignancy was confirmed in 140/163 cases (86%) and the diagnosis of benign lesions was confirmed in 34/46 cases (74%). Regarding specific diagnosis, we observed moderate agreement between the original diagnosis and the reviewer's diagnosis (136/209 cases; kappa=0.5; percent concordance=65%). The highest disagreement was observed in cases of ductal carcinoma in situ with microinvasion (6/6 cases; 100%). Important discordance was observed in cases of atypical ductal hyperplasia (16/30 cases; 53%) and ductal carcinoma in situ (25/75 cases; 33%). Regarding the histological grade of ductal carcinoma in situ, we observed good agreement between the original diagnosis and the review (29/39 cases; kappa=0.6, percent agreement=74%). CONCLUSION: Our data confirm that intraductal proliferative breast lesions, especially atypical ductal hyperplasia, ductal carcinoma in situ and ductal carcinoma in situ with microinvasion show relevant discrepancies in the histopathological diagnoses, which may induce errors in therapeutic decisions.