OBSERVER VARIATION IN AGNOR COUNTS IN NEOPLASTIC BREAST-LESIONS

OBJECTIVE: To determine interobserver and intraobserver variability of AgNOR quantitation in neoplastic lesions of the breast. STUDY DESIGN: Forty-Jive cases, 20 benign and 25 malignant lesions, were included i n the study. Counts were performed on one slide from each case within a pre-marked area of about 1 cm(2) in a standardized manner by two obs ervers blind to the histopathologic diagnosis and independent of each other and repeated after two weeks. Interobserver and intraobserver ag reement was assessed using the Bland-Altman method. RESULTS: Our resul ts showed small mean interobserver and intraobserver differences but w ide limits of agreement. CONCLUSION: Observer variation in AgNOR count s is too high for the method to be of any diagnostic or prognostic rel evance.