Gastric HER2 testing study (GaTHER): Evaluation of gastric cancer testing accuracy in Australia.

Abstract

15 Background: Amplification and/or overexpression of human epidermal growth factor receptor 2 (HER2) has been reported in 6%-33% of advanced gastric cancers (GC), and trastuzumab, an anti-HER2 monoclonal antibody, has been shown to provide a survival benefit in HER2-positive disease. Thus accurate testing will be critical to ensure HER2 utility as a predictive marker in GC. HER2 in situ hybridisation (ISH) testing is currently undertaken for all breast cancer in Australia as part of a national program, however pathologists have limited experience with GC tissue. This study evaluated testing methods and determined inter-laboratory reproducibility of HER2 scoring in GC across 9 Australian reference laboratories. Methods: A tissue microarray (TMA) comprising 100 pre-screened gastric or gastro-oesophageal junction (GOJ) adenocarcinoma samples with a range of HER2 positivity was prepared from 1-mm cores. Each laboratory assessed TMA HER2 status by both immunohistochemistry (IHC) (n=9) and ISH [chromogenic (n=3), silver (n=6), fluorescent (n=1)], using its validated testing methods and applying GC scoring criteria. Inter-laboratory agreement on HER2 ISH scoring, using both HER2 copy number and HER2:chr17 ratio, and agreement on IHC scoring (ToGA method), was calculated (kappa statistics). Results: There was only moderate agreement between laboratories on IHC scoring (κ = 0.46). Agreement on CISH/SISH scoring was good/very good when HER2 copy number, stratified into negative, equivocal and positive for amplification, was used (κ = 0.68-0.86), but was reduced when HER2:chr17 ratio was used (κ = 0.59-0.70). Agreement between CISH/SISH and FISH using HER2 copy number was excellent (κ = 0.88-0.91) but was again reduced when HER2:chr17 ratio was used (κ = 0.84-0.86). Good/very good agreement between IHC and FISH (κ = 0.85) or CISH/SISH (κ = 0.72-0.87) was also achieved when HER2 copy number was used. Conclusions: Good concordance across laboratories can be achieved with FISH, or single-probe CISH or SISH, scored using HER2 copy number, with use of chr17 probe in chromogenic assays best restricted to equivocal cases. HER2 IHC alone is not recommended for determining HER2 status in gastric and GOJ cancers. [Table: see text]