Utilizing Image Analysis in the Assessment of Decalcification Effects on Prognostic Breast Markers
Shawn Maclary, Sonia Kalaria, Arkadiusz Gertych, Fai Chung, Beatrice Knudsen, Shikha Bose, Bonnie Balzer. Cedars-Sinai Medical Center, Los Angeles, CA
Background: Treatment options for breast cancer bone metastases are often determined by the immunohistochemical expression of predictive biomarkers- Estrogen (ER), Progesterone (PR), Ki-67, HER-2/neu and p53. Bone biopsies require decalcification prior to histologic evaluation, however the effects of decalcification on expression of biomarkers are unknown. Using an experimental model and time course, this study is designed to quantitatively assess whether or not decalcification affects the results of the biomarker measurements.
Design: Residual tumor samples were prospectively collected from 9 breast cancer specimens: 1 grade 2 and 8 grade 3 invasive ductal carcinomas. Samples were formalin fixed according to CAP guidelines, and placed in an acidic decalcification solution (Decal State, Decal Chemical Corp., Tallman, NY) at ambient temperature for 0, 1, 6, and 24 hours(hrs). After routine processing, sections were stained by immunohistochemistry utilizing the Ventana System (Oro Valley, AZ). Stained slides were digitized at x20 magnification on a Leica SCN 400 (Buffalo Grove, IL) scanner. Tumor areas were randomly selected and image analysis was executed using the Tissue IA 2.0 (Leica) software. After delineating cell membranes and nuclei, signals were quantified in subcellular compartments. Numerical scores were converted to categorical scores of weak, moderate and strong to recreate the conventional reporting of biomarker expression by pathologists.
Results: Upon image analysis, the signal intensities and percentages of positive cells gradually decreased with the length of decalcification and a significant decrease was observed after 6 hrs for all markers. Progressive loss was further confirmed microscopically and noticeable at decalcification times ≥ 1hour. It correlated with a significant reduction in the percentage of positive cells. No increase in immunoreactivity was observed for any of the markers after decalcification.
Conclusions: Decalcification reduces the immunoreactivity of predictive breast cancer markers. The time of decalcification may endanger optimal treatment decisions by causing artificially low protein expression signals. This project demonstrates that image analysis can provide an objective and precise assessment of the adverse effects of decalcification on biomarker measurements and can lead to the development of a strategy for correcting the loss of immunoreactivity in decalcified tissues.
Wednesday, March 6, 2013 1:00 PM
Poster Session VI # 276, Wednesday Afternoon