Comparison of FISH and SISH Methods for HER2 Testing in Breast Carcinoma: A Validation Study Emphasizing Automated Methods
Martin C Chang, Marianne Rogers, Gordana Kuruzar, Mona Reid, Mendes Mendes, Philip Plotnick, Azar Azad. Mount Sinai Hospital, Toronto, ON, Canada; Univ of Toronto, Toronto, ON, Canada
Background: Amplification for the HER2 gene is of prognostic and predictive significance in breast carcinoma. The standard methodology for assessing HER2 amplification is FISH. Silver-based methods (SISH) represent newer technology that allow assessment of HER2 status, but using brightfield microscopy. Our goal is to evaluate and validate more automated methods of HER2 testing.
Design: Cases of invasive breast carcinoma tested for HER2 status by FISH were identified and selected to provide a sample representative of the patient population. A total of 100 cases were selected: 78 HER2-negative, 21 HER2-positive, and 1 HER2-equivocal. FISH processing used Vysis Pathvision probes (HER2 and CEP17, Abbott Laboratories). The original interpretation was by manual counting. Additional 4 μm sections were cut and SISH performed (Ventana Benchmark Ultra), with HER2 and Chromosome 17 tests on 2 separate slides. Manual counting of both slides was performed to calculate the ratio of HER2 to Chromosome 17. SISH slides were interpreted by automated imaging (Ventana VIAS), and the results compared to the manual counts (both SISH and FISH). FISH slides were interpreted using the Visiopharm Integrator System (Visiopharm, Denmark). The discordances between methods were analyzed, with the original FISH result as the "gold standard".
Results: Using both manual counting and automated scoring, SISH was concordant with manual FISH in 95% of cases. These discordances were either false negatives (up to 2.6%), or cases in which there was disagreement between positive and equivocal interpretations. The manual and automated scoring methods of SISH were 98% concordant with each other. The automated method of interpreting FISH was 93% concordant with the manual method. Compared to FISH, the SISH method demonstrated the following advantages: automated slide processing, faster interpretation by brightfield microscopy, and room-temperature slide storage. The main source of discordances between FISH and SISH was difficulty in the latter of interpreting clustered HER2 and Chr17 signals.
Conclusions: The FISH and SISH methods are highly concordant for the determination of HER2 status. FISH has the advantage of enabling the spectral filtering of signals for counting, and for evaluating the ratio within each cell. SISH, although requiring separate scoring of HER2 and Chr17 signals, has the advantages of automated processing and enabling scoring under brightfield microscopy. Both methods are amenable to automated interpretation.
Tuesday, March 20, 2012 9:30 AM
Poster Session III # 37, Tuesday Morning