Comparative Pathologic Analysis of Breast Cancers Classified as HER2/neu-Amplified by FISH Using a Standard HER2/CEP17 Dual Probe and an Alternative Chromosome 17 Control Probe

At our institution, breast cancer cases that generate an equivocal HER2/neu (HER2) result by fluorescence in situ hybridization (FISH) using the dual HER2/chromosome enumeration probe (CEP17) are reflexed to an assay that utilizes an alternative control probe (lissencephaly gene1 [LIS1] [17p13.3]/retinoic acid receptor [RARA] [17q21.2]). This study examines whether cancers that are classified as HER2-amplified with an alternate probe are clinicopathologically similar to those that are classified as such using the HER2/CEP17 probe. Reports for 1201 breast cancers were reviewed, and clinicopathologic findings were compared between HER2/CEP17-equivocal cases that became HER2-amplified using the alternate probe (group A: n=48), HER2-amplified cases using the HER2/CEP17 probe (group B: n=169), and HER2-nonamplified cases using the HER2/CEP17 probe (group C: n=910). Of 1201 cases tested using the HER2/CEP17 probe, 169 (14%) were HER2-amplified, 122 (10%) were equivocal, and 910 (76%) were nonamplified. Additional testing with the alternative probe on the 122 equivocal cases reclassified 48 (39%) of them to HER2-amplified, and such cases comprised 22% of all HER2-amplified tumors. A higher proportion of tumors with HER2 copy number between 5.0 and 5.9 became positive upon additional testing when compared with those with a priori HER2 copy numbers between 4.0 and 4.9 (P=0.0362). Group A cases, compared with group B cases, were more frequently positive for estrogen receptor (97.91% vs. 72.18%, P<0.0001) and progesterone receptor (85.41% vs. 59.17%, P=0.0009). Most group A cases (71%) were HER2 equivocal (score 2+) by immunohistochemistry, whereas most group B cases (60%) were positive (score 3+). Groups A and B showed no significant differences regarding patient age, lymph node status, tumor grade, histotype, and stage distribution. In summary, among our HER2-amplified cohort of breast cancers, alternative probe-detected cases were more frequently estrogen receptor and progesterone receptor positive than HER2/CEP17-detected cases, and were more frequently discordant with HER2 immunohistochemistry results. These findings raise the possibility of underlying biologic differences between these 2 groups, which warrants further study. However, the tumors were largely comparable regarding all other clinicopathologic variables. As it is unknown whether HER2-targeted therapy is truly beneficial in this subgroup of patients, future clinical trials should specifically evaluate this subset.