Digital Image Analysis as an Adjunct in the Analysis of Her2/Neu Staining of Breast Biopsies
Bryan J Platt, Thomas A Summers, Justin M Wells, Ross Barner. Walter Reed National Military Medical Center, Bethesda, MD
Background: Clinically relevant immunohistochemical (IHC) stains are used to assist pathologists in determining susceptibility to targeted therapies in patients diagnosed with cancer, and comprise those stains that markedly impact patient care with prolonged courses of expensive, highly-effective, and targeted medications. However, interpretations of IHC staining parameters can be subjective and are prone to inter- and intra-observer variability. Virtual microscopy digitizes glass slides, creating digital image environments for subsequent computer assisted interpretation of pathologic information.
Design: Automated HER2/neu IHC staining (HerCepTest, Dako, Carpinteria, CA) was performed on newly diagnosed, invasive breast cancers, and scored according to the current ASCO/CAP guidelines. First tier HER2/neu IHC analysis was performed manually by the staff pathologist of record, while a second tier manual review was subsequently performed by the quality control (QC) reviewer for all HER2/neu stains. All HER2/neu stains were then digitized for analysis (IHC HER2 IA Algorithm; Aperio Scanscope XT; Vista, CA). A minimum of 1000 cells were analyzed for each case, ensuring proper representation of all staining patterns present on individual slides and only scoring invasive tumor elements. The digital IA results were then provided to the QC pathologist to aid in rendering the most accurate and informed final interpretation; however, the QC-assisted manual scores were recorded in order to compare to the digital IA scores.
Results: 191 HER2/neu stains were evaluated. Concordance between the ASCO/CAP scores (negative: 0, 1+, equivocal: 2+, and positive: 3+) for manual and digital analysis methods was 84% (160/191). Of the 31 discordant cases, two cases were clinically meaningful, demonstrating negative FISH results after receiving a digital score of 2+ and a manual score of 3+. These two cases were interpreted as equivocal in the final pathologic diagnosis.
Conclusions: The adjunctive use of automated digital IA in the evaluation of clinically pertinent IHC stains may increase the specificity and positive predictive value of interpretations by increasing objectivity and therefore helping to standardize and enhance quantification.
Disclaimer: The views expressed in this presentation are those of the author and do not necessarily reflect the official policy or position of the Department of the Army, Department of the Navy, Department of Defense, or the U.S. Government.
Category: Quality Assurance
Tuesday, March 5, 2013 9:30 AM
Poster Session III # 272, Tuesday Morning