HER2 Dual ISH Determination of HER2 Gene Status in Breast Cancer: Interobserver Reproducibility
Abberly A Lott Limbach, Erinn P Downs-Kelly, Bettina G Papouchado, Raymond R Tubbs, Christopher Lanigan, Christine N Booth. Cleveland Clinic, Cleveland, OH
Background: The HER2 status of breast carcinoma has prognostic and treatment implications. Current guidelines from the American Society of Clinical Oncology and College of American Pathologists include immunohistochemistry and fluorescence in situ hybridization (FISH) as HER2 status testing methods. Disadvantages of FISH include the laborious assay, need for fluorescent microscopy and fluorescent signal degradation over time. The INFORM HER2 DNA Probe Cocktail assay (Dual ISH)(Ventana Medical Systems Inc., Tucson, AZ) is fully automated, is performed on formalin-fixed paraffin-embedded tissue and uses HER2 and Chromosome 17 (CHR17) specific probes to qualitatively and quantitatively assess HER2 gene status using brightfield microscopy.
Design: Prior to scoring the study slides, 5 pathologists completed training in Dual ISH interpretation via review of e-learning and reference materials, an e-learning exam, and scoring of a 30-slide test set of various HER2 amplification states (provided by Ventana). Pathologists then independently scored 86 cases of invasive breast carcinoma, blinded to the historical FISH results. The average HER2 and CHR17 copy number were recorded using a Gestalt interpretation and the HER2/CHR17 ratio was recorded (as per the package insert FDA guidelines) as amplified if the HER2/CHR17 ratio ≥2.0 and as non-amplified if the HER2/CHR17 ratio <2.0. If the HER2/CHR17 ratio was 1.8 to 2.2 (inclusive), 20 additional nuclei were counted and a new ratio was recorded based on 40 nuclei. Interobserver agreement and concordance of each observer's Dual ISH score with historical FISH results were calculated using the kappa statistic. With a kappa value > .80, reproducibility has been inferred from published literature to be “excellent”.
Results: Overall agreement between conventional FISH and the average Dual ISH results was 96.3%. The average interscorer Dual ISH Kappa was 0.87 while the average Dual ISH versus FISH Kappa for all scorers was 0.92. The kappa results of the interobserver Dual ISH scores are summarized in Table 1.