Inter-laboratory evaluation of a novel DEPArray-HER2 FISH assay.

e12506 Background: Fluorescent in Situ Hybridization (FISH) is a method currently used for detection and assessment of HER2 gene amplification. Although clinical guidelines set forth by CAP/ASCO exist to ensure accuracy, limitations in HER2test results due to sample preparation, assay-conditions and tumor heterogeneity remain unresolved. We have successfully demonstrated analytical confidence in performing HER2 FISH on DEPArray™ sorted and recovered tumor cells. In this study, we aimed to evaluate inter-laboratory concordance of the DEPArray™ HER2-FISH assay. Methods: Three laboratories equipped with the DEPArray™ were designated as testing sites for this study. Positive control SKBr3 cells embedded in paraffin as well as 20 invasive breast carcinoma FFPE samples were blinded and evaluated by each of the three labs. Control and patient samples were processed through the DEPArray™ beginning with dissociation of the FFPE curls followed by single-cell image-based cell sorting to separate and recover pure distinct tumor cell populations prior to HER2 FISH analysis. Data was only obtained when ∼200 intact cytokeratin+/vimentin-/DAPI+ tumor cells from each sample were recovered and used for subsequent FISH using a standard dual-color HER2/CEP17 FISH procedure. Results: Overall, 80% concordance between DEPArray™-HER2 and conventional HER2 (6 HER2 negative and 10 HER2 positive) was observed between lab(s) and the conventional HER2 method. In each of 4 cases, a discordant HER2 result was reported by one of three sites. In three of these discordant cases, the DEPArray™ HER2 ratio was reported as amplified while the conventional result was negative. In the remaining discordant case, the converse was observed by one site; however, this case was initially evaluated 15 years ago. All three sites correctly scored the SKBr3 positive control cells. Conclusions: The results showed a high concordance rate of correct HER2 status classification. This data further supports the understanding that tissue heterogeneity can indeed give rise to discordant results that may consequently affect treatment options for patients. We demonstrate that sample preparation by DEPArray™ may aid in a more precise classification for tumor biomarker status.