FISH Testing for HER2 Expression May Yield Discordant Results Depending on Reporting Criteria Used.
Suzanne M Dintzis, Kimberly H Allison, Rodney A Schmidt. University of Washington Medical Center, Seattle
Background: Amplification status of the HER2 gene may be assayed using FISH. Currently, the established CAP/ASCO guidelines allow two different systems for reporting the results of the FISH assay: average HER2 signals per cell or the average ratio of HER2 and chromosome 17 (CEP17) signals. Although both resulting systems classify cases as amplified, not amplified or equivocal, the two systems may yield different interpretations for the same case. We analyzed a large dataset to further examine concordance between the two methodologies.
Design: HER2 and CEP17 counts were collected from 32,116 tumor cells from 1329 consecutive breast cancer cases using data from Excel counting sheets imported to a SQL Server database for analysis and statistics. We specifically computed concordances between results generated using HER2 signals per cell versus HER2:CEP17 ratio. Discordant cases were reviewed to obtain concurrent HER2 immunohistochemical staining, tumor grade and stage.
Results: Amplification status derived from HER2 signals per cell or HER2:CEP17 ratio was concordant in 89.8% of cases. Most discordances involved cases classified as equivocal by either method. Equivocal classification occurred in 7.8% of cases reported by HER2 signals per cell and 4.0% of cases reported by HER2:CEP17 ratio. 1% of cases were amplified by one method and not amplified by the other. Of these, 4 cases were classified as amplified using HER2 signals per cell only (> 6 signals per cell) and not amplified using HER2:CEP17 ratio (< ratio 1.8). On review of HER2 immunohistochemical stains, none of these cases scored 3+ positive; 3 cases were 2+ indeterminate and 1 case was 1+. An additional 9 cases were classified as amplified using the HER2/CEP17 ratio (>2.2) but not amplified by HER2 signals per cell. All "ratio only" amplified cases showed mean CEP17 of less than 1.5. None of the "ratio only" positive cases showed 3+ positive immunohistochemical staining; 3 were IHC negative and 6 were 2+ indeterminate. In addition, the "ratio only" positive cases were generally lower Nottingham grade tumors.
Conclusions: This very large dataset shows that Her2 FISH reporting using the two established CAP/ASCO guidelines is frequently discordant for equivocal cases and rarely more seriously discordant. Because treatment decisions are made based on HER2 status, a two system approach which yields discordant results is not satisfactory. Further research is needed to determine which approach is likely to yield the most clinically relevant results.
Wednesday, March 2, 2011 9:30 AM
Poster Session V # 10, Wednesday Morning