Abstract 3185: Use of whole slide digital imaging for determination of caPCNA and Ki-67 labeling index in normal breast reductions and DCIS: Comparison of the two biomarkers using the Aperio image-based positive pixel algorithm

Abstract

Abstract While mitotic rates as well as proliferation indices have been reported as significantly different among normal breast and different grades of breast cancer, a reproducible cut-off value has not emerged. The goal of this study was to determine whether more reproducible algorithms using whole slide digital images (WSDI) can define better cut-off values for the normal breast vs. DCIS. Design: Whole tissue sections from 10 normal breast reductions and 10 cases of DCIS were evaluated by immunohistochemistry for caPCNA and Ki67 (Mib-1). Having identified an isoform of PCNA that was uniquely expressed by cancer cells (caPCNA) and distinct from the isoform expressed by non-malignant and normal cells (nmPCNA), antibody selectively binding to caPCNA was developed and used for these studies. Stained slides were scanned using the Aperio (Vista, California) automated whole slide scanning system (Scanscope CS) and were viewed using ImageScopeTM. Single fields of view from each WSDI measuring ∼420,000 µ2 and representing the highest density of caPCNA and Ki-67 positive nuclei were selected for analysis. Different areas were analyzed in a subset of slides to determine the variability in each case, and the total nuclear labeling index was generated using the Aperio positive pixel algorithm. Results: We analyzed the labeling indices in 10 normal breast reductions and 10 DCIS cases. Both caPCNA and Ki-67 were nuclear specific with minimal cytoplasmic staining. The mean labeling indices for the Aperio automated positive pixel counts were caPCNA (0.15) and Ki-67 (0.12) for normal breast and caPCNA (0.52) and Ki-67 (0.29) for DCIS. The correlation coefficients between automated counts were 0.653 for normal breast and 0.616 for DCIS, with an overall correlation coefficient of 0.708 for the entire series. Conclusions: Our analysis of both biomarkers yielded a similarly increasing labeling index for DCIS cases when compared to their expression in normal breast expression. The positive pixel algorithm method has a sampling bias due to subjective selection of the area being counted. In summary, the present results show similar findings for caPCNA and Ki-67 in normal breast reductions with no evidence of breast disease and in DCIS clinical cases. Comparing caPCNA with Ki-67, and unlike traditional PCNA, caPCNA can potentially better differentiate between normal breast and DCIS, and may be a useful biomarker for early stage cancer diagnosis. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3185. doi:10.1158/1538-7445.AM2011-3185