Correlation of HER2 low status in I-SPY2 with molecular subtype, response, and survival.

Abstract

514 Background: HER2low, defined as immunohistochemical (IHC) 1+ or 2+ without HER2 gene amplification, predicted improved progression free and overall survival with trastuzumab deruxtecan (TDXd) compared to chemotherapy in patients (pts) with metastatic breast cancer in Destiny Breast04. Controversy exists regarding the correlation of HER2low with molecular subtypes and outcome. We evaluated these associations in the neoadjuvant ISPY2 trial. Methods: We investigated HER2 IHC class in pts with clinically HER2-negative breast cancer (BC) enrolled in the first 10 arms of ISPY2. To explore the biology of HER2low, we used Fisher’s exact test to assess association of IHC class (0 vs 1+ or 2+) with Response Predictive Subtypes-5 (RPS-5; based on immune, DNA repair damage (DRD), HER2, and luminal markers; PMID: 35623341), SET index [endocrine responsiveness], and Mammaprint high1 (MP1) vs high2 (MP2). Association with pathologic complete response (pCR) was assessed using Fisher’s exact test, and association with distant recurrence free survival (DRFS) was assessed using Cox Proportional Hazards modeling. Results: Of 742 HER2negative pts enrolled in the first 10 arms of ISPY2, local HER2 IHC is available for 585; 299 hormone receptor + (HR+) and 286 triple negative (TN). 63% of pts are HER2low, with HER2low status more frequent in HR+ (71%) compared to TN BC (55%; OR = 1.97; p = 0.00011). There was no clear relationship between HER2low IHC and RPS-5 subtypes or endocrine sensitivity measured by the SET index. There was no significant association of HER2low with pCR overall or within treatment arms (p > 0.05); this was true for TN and HR+ subsets. 562 pts with HER2negative disease have distant recurrence free survival (DRFS) data with a median follow-up of 4.23 yrs. Pts with pCR had excellent outcome (5yr DRFS > 94%), with no impact from HER2low vs 0 status. For pts with nonpCR, after adjusting for HR+ status, the DRFS hazard ratio for HER2low vs 0 status is not significant [0.68(0.46-1.02)]. Within HR+, HER2low is found in > 70% for both SET low and high. HR+/MP1 pts (with generally low pCR rates) are significantly more likely to be HER2low (75%; (172/219)) compared to HR+/MP2 (61%; (49/80)) (OR: 1.79; Fisher p = 0.044). Importantly, 70% of HER2-Immune-/DRD- (78% HR+ and 56% TN), a subtype with low response to all ISPY2-tested agents, are HER2Low. Conclusions: In ISPY2 which enrolls patients with MP high risk BC, HER2low by IHC was frequent and higher in HR+ than TN disease. There was no clear association with molecular markers, pCR or DRFS. The frequency of HER2low in the immune-/DRD- subtype raises the potential for exploring T-DXd in this high risk, low-response setting. Clinical trial information: NCT01042379 .