Clinical Outcomes of HER2-Low Versus HER2-Zero in HR-Positive Metastatic Breast Cancer Treated With Endocrine Therapy With or Without CDK4/6 Inhibitors: A Multicenter Retrospective Study

Abstract

Purpose: HER2-low status is a predictive factor for novel anti-HER2 therapies in metastatic hormone receptor–positive breast cancer (HR+ MBC). However, its impact on endocrine therapy outcomes remains uncertain. We aimed to explore the effect of HER2-low and HER2-zero status in HR+ MBC patients treated with first-line aromatase inhibitors (AIs) with or without CDK4/6 inhibitors (CDK4/6i). Methods: We retrospectively reviewed postmenopausal women with HR+ MBC treated with first-line AI ± CDK4/6i between January 1, 2017, and December 31, 2022, from six tertiary hospitals in Thailand. HER2-low was defined as HER2 IHC 1+ or IHC 2+ with ISH-negative. Progression-free survival (PFS) and overall survival (OS) were compared in unadjusted and adjusted cohorts using stabilized inverse probability of treatment weighting (sIPTW), adjusting for age, ECOG performance status, de novo metastasis, endocrine sensitivity, visceral metastasis, number of metastatic sites, and treatment year. Interaction analyses were performed to assess effect modification by HER2 status and other clinical subgroups. Results: Among 504 patients, 219 (43.5%) were HER2-low, and 285 (56.5%) were HER2-zero. Median follow-up was 31 months (IQR 19–47). CDK4/6i + AI was administered to 52.5% of HER2-low and 43.9% of HER2-zero patients. After sIPTW adjustment, CDK4/6i + AI prolonged median PFS to 22.1 months compared with 21.5 months for AI alone in the HER2-low cohort (HR = 0.80, 95% CI 0.54–1.18; p = 0.26) and to 20.1 months compared with 13.5 months in the HER2-zero cohort (HR = 0.65, 95% CI 0.45–0.93; p = 0.02). Median OS was 49.3 months with CDK4/6i + AI versus 48.4 months with AI alone in HER2-low (HR = 0.81, 95% CI 0.48–1.36; p = 0.43) and 45.8 versus 42.3 months in HER2-zero (HR = 0.85, 95% CI 0.43–1.05; p = 0.52). Subgroup analyses showed consistent benefit of CDK4/6i + AI across most clinical categories. The interaction test for treatment × HER2 status was not significant (HR = 1.21, 95% CI 0.74–1.98; p = 0.44), indicating no effect modification by HER2-low status. Conclusions: HER2-low status was not associated with prognosis or predictive value for CDK4/6i efficacy, supporting CDK4/6i + AI as the standard first-line therapy irrespective of HER2 expression level.