Document Type


Publication Date



Automated image analysis systems for breast cancer immunohistochemistry promise efficiency and reliability in the quantification of therapy targets such as the estrogen receptor (ER) or human epidermal growth factor receptor (Her2). Thomas Jefferson University Hospital owns two such systems, the Aperio ScanScope AT (Leica Biosystems) and Ventana iScan Coreo (Roche). A comparison study was performed to determine if choice of system affects target quantification and subsequent clinical tumor classification. Tumor expressions of ER, progesterone receptor (PR), proliferation marker Ki67, and Her2 were quantified with both systems for tissue samples from twenty breast cancer patients. Positive tumor classification was based on percent positivity values of ER >1%, PR >1%, Ki67 >10%, and Her2 score of 3+. Percent agreement for tumor classification was ER 100 (95% CI 83.2-100), PR 95 (75.1-99.9), Ki67 90 (68.3-98.8) and Her2 100 (83.2-100). While agreement was high large variation was observed in percent positivity scores (ER, PR, Ki67).