Effects of the change in cutoff values for human epidermal growth factor receptor 2 status by immunohistochemistry and fluorescence in situ hybridization: a study comparing conventional brightfield microscopy, image analysis-assisted microscopy, and interobserver variation.

CONTEXT New guidelines for HER2 testing have been introduced. OBJECTIVES To evaluate the difference in HER2 assessment after introduction of new cutoff levels for both immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) and to compare interobserver agreement and time to score between image analysis and conventional microscopy. DESIGN Samples from 150 patients with breast cancer were scored by 7 pathologists using conventional microscopy, with a cutoff of both 10% and 30% IHC-stained cells, and using automated microscopy with image analysis. The IHC results were compared individually and to HER2 status as determined by FISH, using both the approved cutoff of 2.0 and the recently introduced cutoff of 2.2. RESULTS High concordance was found in IHC scoring among the 7 pathologists. The 30% cutoff led to slightly fewer positive IHC observations. Introduction of a FISH equivocal zone affected 4% of the FISH scores. If cutoff for FISH is kept at 2.0, no difference in patient selection is found between the 10% and the 30% IHC cutoff. Among the 150 breast cancer samples, the new 30% IHC and 2.2 FISH cutoff levels resulted in one case without a firm diagnosis because both IHC and FISH were equivocal. Automated microscopy and image analysis-assisted IHC led to significantly better interobserver agreement among the 7 pathologists, with an increase in mean scoring time of only about 30 seconds per slide. CONCLUSIONS The change in cutoff levels led to a higher concordance between IHC and FISH, but fewer samples were classified as HER2 positive.