Objectives: Guidelines suggest that secondary reflex testing may be useful for resolving HER2 status in breast cancers with equivocal results by both immunohistochemistry (IHC) and in situ hybridization (ISH). We compared two reflex ISH assays and a polymerase chain reaction (PCR) assay for this application.
Methods: Twenty-nine breast cancers with equivocal IHC and ISH results were retested two ways: (1) ISH using differentially labeled probes targeting ERBB2 (HER2, 17q12) and either RAI1 (17p11.2) or ORC4 (2q22.3-2q23.1) in two separate assays and (2) real-time quantitative PCR amplification of ERBB2 and a control locus (EIF5B, 2q11.2).
Results: Results of the HER2/RAI1 and HER2/ORC4 ISH assays were concordant for 21 (72%) cases, and results of all three secondary reflex assays were concordant for only 18 (62%) cases. Result discrepancies between the two ISH readers were observed for cases close to the cutoff threshold.
Conclusions: Use of different control loci for ISH is associated with discordant results, and PCR is more likely to classify cases as nonamplified, possibly due to contamination with nontumor cells. While resolution of HER2-equivocal results is desirable from a clinical perspective, different secondary reflex assays yield different results, and the correlation of these results with clinical outcomes is unknown.