Resolving Equivocal HER2 Status in Breast Cancer by Automated and Quantitative RNA Chromogenic In Situ Hybridization (CISH)
Zhen Wang, Son Bui, Hongwei Wang, Nan Su, Xiao-Jun Ma, Yuling Luo, Raymond R Tubbs. Cleveland Clinic, Cleveland, OH; Advanced Cell Diagnostics, Hayward, CA
Background: Fluorescence in situ hybridization (FISH) for HER2 gene amplification and immunohistochemistry (IHC) for HER2 protein overexpression both can generate “equivocal” HER2 results (FISH HER2/CEP17 ratio 1.8-2.2 or IHC score 2+), and some cases generate equivocal results by both methods. This study explores the potential of a novel automated and quantitative HER2 mRNA CISH assay based on the recently developed RNAscope technology for resolving equivocal HER2 status.
Design: Formalin-fixed, paraffin-embedded (FFPE) breast cancer specimens from a non- consecutive series of 73 cases were analyzed for HER2 mRNA using a fully automated RNAscope CISH assay. There were 30 negative and 22 positive cases based on the combined FISH and IHC results, and 21 cases were equivocal by both methods (“dual equivocals”). These cases were pre-allocated into a training set (n=38) and a validation set (n=35). Automated image analysis of HER2 mRNA staining was used to count the number of punctate “dots” per cell, each dot corresponding to a single HER2 RNA transcript. and correlated to HER2 FISH, IHC and HER2 mRNA RT-PCR. A probabilistic linear discriminant analysis model for HER2 status based on the training set was built and applied to the validation set.
Results: Evaluable HER2 mRNA CISH results were obtained for 67 cases (92%). HER2 mRNA dots per cell correlated strongly to FISH (Spearman r=0.81) and RT-PCR (r=0.86) and definitively separated HER2 positive and negative cases in the training set. A predictive model based on HER2 mRNA dots/cell in the training set correctly identified 11/12 negative and 5/5 positive cases (concordance=94%) in the validation set. The only discrepant case was reviewed and found to have both amplified ductal carcinoma in situ (DCIS) and Non-amplified invasive components; the HER2 RNA CISH results were from the DCIS. When analyzed on the invasive component only, this case was correctly classified as HER2 negative. The model classified all 16 double equivocal cases in the validation set into positives (n=2) and negatives (n=14).
Conclusions: This quantitative HER2 mRNA CISH assay was highly accurate in assessing HER2 status and may provide an effective means to resolve FISH/IHC dual equivocal cases. The walk-away automation and image analysis-based quantification should minimize both analytical and post-analytical variability in HER2 testing. Quantification of single RNA transcripts in situ in routine clinical specimens demonstrates great potential in predictive biomarker analysis.
Wednesday, March 21, 2012 9:30 AM
Poster Session V # 50, Wednesday Morning