Comparison of Image Analysis System for Quantification of HER2 and ER in Breast Cancers

Indianapolis, IN

Breast Cancer is one of the most common cancer among women today. During their lifetime, about 1 in 8 U.S. women will develop invasive breast cancer. However, the rates have been decreasing the U.S. since 2000. New biomarkers are needed to predict patient outcomes and what drugs should be given to the patients in this era of precision medicine. In this study, two tissue microarrays (TMAs) were evaluated, each containing approximately 80-90 cores (2 per patient) on each TMA. These TMAs were evaluated using two different computer image analysis systems, the Aperio and Qupath Imaging systems. The biomarkers examined were Estrogen Receptor (ER), which is expressed in the nucleus, and HER2, which is expressed on the cell membrane of breast cancer cells. The immunostaining was preformed using the DAKO-flex staining platform system. The TMA slides were imaged using the Aperio Digital Imaging System. Biomarker expression was quantified by a positive pixel algorithm and a calculated histoscore. Both systems were comparable in analysis efficiency and data accuracy by comparing positive pixel results and H score results. An R-squared line showed the similarity between the two systems within 10% on both positive pixel and H score analysis. Both can quantitate biomarkers accurately.