Accuracy of digital image analysis (Dia) of borderline human epidermal growth factor receptor (her2) immunohistochemistry in invasive ductal carcinoma

Abstract

Background: Digital image analysis (DIA), used to extract information from pathology slides, provides better precision and no limitation regarding different interpretations by observers. Objectives: The present study aimed at evaluating the accuracy of DIA in the interpretation of borderline (2+) human epidermal growth factor receptor 2 (HER2) immunohistochemistry (IHC) slides of invasive ductal carcinoma of the breast. Methods: Sixty pathology samples with invasive ductal carcinoma of the breast were extracted based on HER2 (2+) and their fluorescence in situ hybridization (FISH), and chromogenic in situ hybridization (CISH) responses (as reference standard). The slides were digitized and, then, two pathologists examined the slides and documented diagnosis. DIA was performed by a free web application. Results: Totally, 307 digital images with 298 megabytes volume were extracted. The accuracy, sensitivity, and specificity values of DIA were 86 %, 46.1 %, and 97.8 %, respectively, with 8 false-negative cases. There was moderate agreement between the pathologist 1 (kappa = 0.42) and pathologist 2 (kappa = 0.41) with DIA. Conclusions: DIA had good accuracy and could be used for the interpretation of borderline HER2 IHC method in invasive ductal carcinoma.