Do HER2 low tumors have a distinct clinicopathologic phenotype?

Abstract

570 Background: Clinicopathologic features of breast cancer subtypes defined by hormone receptor (HR) and HER2 status differ. These analyses classified HER2 IHC <3+ with negative FISH as HER2-. With the recognition of the clinical relevance of HER2 low, there is debate as to whether this is a distinct subtype. We sought to determine if features of HER2 low tumors differ from HER2- and HER2+ after controlling for HR status. Methods: Patients undergoing upfront surgery from 1998 to 2010 were identified from a prospectively maintained institutional database. HER2 status was classified by IHC/FISH analysis as HER2-, HER2 low (IHC 1+ or 2+ with negative FISH), and HER2+ (IHC 3+ or FISH +) and stratified by HR status. Univariable (UVA) and multivariable multinomial logistic regression analysis (MVA) were performed to determine associations among variables and subtypes. Results: 11,323 tumors from 11,072 patients were included. 5,104 (45%) were HER2 low, 41.2% were HER2- and 13.6% were HER2+. On MVA stratified by HER2 status only, compared to HER2- tumors, HER2 low was associated with LVI (OR 1.2 [95% CI 1.06-1.36]; p=.003), multifocality (OR 1.26 [95% CI, 1.12-1.42]; p<.001), nodal micrometastasis (OR 1.15 [95% CI, 1.02-1.31]; p=.024), and lower rates of <3 positive nodes (OR 0.77 [95% CI, 0.66-0.90], p=.001). HER2+ was associated with younger age, high grade, multifocality, extensive intraductal component (EIC), and pN2/3 nodal stage. Clinicopathologic features stratified by HR and HER2 status are shown in the table. On UVA, in both HR+ and HR– tumors, age and multifocality were associated with HER2 low, and LVI, multifocality, and EIC with HER2+ compared to HER2- tumors. On MVA, no variables were independently associated with both HR+ and HR-/HER2 low tumors compared to HER2-. In contrast, HER2+ tumors regardless of HR status were significantly associated with multifocality and EIC. Conclusions: HER2 low breast cancer features seem to be driven by HR status and HER2 overexpression. We do not have enough evidence to support the interpretation of HER2 low as a distinct subtype. [Table: see text]