Image-based collection to allow isolation and analysis of pure intact tumor cell populations from FFPE: Implications for more precise HER2 FISH analysis.

Abstract

e12094 Background: Guidelines worldwide focus on the importance of precise and accurate Fluorescent In Situ Hybridization (FISH) methods for testing companion diagnostic markers, including Human Epidermal Growth Factor Receptor 2 (HER2) gene amplification in breast cancer. Despite these guidelines, variations in test results are observed due to pre-analytical sampling. In this study, we demonstrate a unique approach to isolating pure and intact tumor cells from breast cancer Formalin-Fixed, Paraffin-Embedded (FFPE) samples for precise FISH analysis. Methods: Fifty-micron thick FFPE curls from both HER2 non-amplified breast cancer tumors (n = 4; each with a reported HER2/CEP17 ratio 1.8) and positive control SKBr3 breast cancer cells were tested. Isolation of ∼250 pure and intact cytokeratin-positive/vimentin-negative/DAPI positive cells from each sample were captured using the DEPArray platform, an automated system enabling image-based cell sorting with single-cell resolution for pure cell population isolation and collection. Standard dual-color HER2/CEP17 FISH (Vysis PathVysion) analysis was performed on recovered cells. Results: Positive HER2 amplification levels for the FFPE derived control SKBr3 cells were observed HER2/CEP17 ratio > 4.4. Among the patient samples, ≥ 75% of the DEPArray isolated tumor cells were recovered onto slides. Through routine FISH scoring, an expected non-amplified result was observed for each patient sample, with HER2/CEP17 ratios ranging from 1.1 to 1.4. Conclusions: We demonstrate feasibility in performing FISH on pure and intact tumor cells isolated from breast cancer derived FFPE using the DEPArray platform. Using a 50-micron section permitted recovery of whole, intact, tumor cells based on immunostaining for cytokeratin, vimentin, and DAPI. Efficient recovery of the DEPArray sorted cells onto slides further permitted routine FISH analysis of only tumor cells. These preliminary results imply the possibility of more precise FISH analysis when standard FISH results are inconclusive or when insufficient tumor content prohibits downstream analysis. Evaluation of larger numbers of patient samples is underway