Incidence and predictors of exceptional response (ExRes) in patients (pts) initiating first-line (1L) treatment for HER2-positive (HER2+) metastatic breast cancer (MBC).

Abstract

e13021 Background: HER2-targeted therapies have revolutionized the treatment of HER2+ MBC; however, pt responses can be heterogeneous. ExRes in MBC, defined as progression-free survival (PFS) of ≥3 years following 1L HER2-targeted therapy initiation, has not been widely studied in large datasets derived from routine clinical practice. Methods: This study used the nationwide Flatiron Health electronic health record-derived de-identified database to assemble a cohort of pts with HER2+ MBC who initiated trastuzumab as 1L therapy from January 1, 2011, to September 30, 2021 (median follow-up 37 months); data were collected up to September 30, 2024, allowing for three years of potential follow-up. Pts were categorized as having ExRes, conventional response (ConRes; PFS < 3 years or death), or unknown response (UnRes; censored before 3 years). Descriptive statistics were used to summarize demographics, clinical characteristics, and treatment variables, all measured up to the index date, across response groups. Associations between clinical characteristics and ExRes status were measured using conditional logistic regression models, and univariate Cox regression analyses were conducted to identify potential predictors of ExRes. Results: Among the 2,454 pts selected, 459 (19%) were classified with ExRes, 1,791 (73%) with ConRes, and 204 (8%) with UnRes. Factors associated with ExRes included Eastern Cooperative Oncology Group performance status (1 v 0 odds ratio [OR] 0.52; 95% confidence interval 0.40–0.69; 2+ v 0 OR 0.24; 0.14–0.39), and de novo v recurrent disease (recurrent 3–24 months v de novo OR 0.45; 0.30–0.69; recurrent 25–60 months v de novo OR 0.55; 0.41–0.73; recurrent 60+ months v de novo OR 0.88; 0.67–1.15). HER2 immunohistochemistry (IHC) 3+ status was associated with ExRes (not IHC 3+ v IHC 3+ OR 0.51; 0.37–0.70), whereas hormone receptor (HR) status was not (HR-positive v HR-negative OR 0.83; 0.66–1.04). Metastatic sites were also associated with ExRes. Brain metastases (yes v no OR 0.27; 0.15–0.48), lung/liver metastases (yes v no OR 0.58; 0.48–0.72), and increasing number of organ sites (per one site increase OR 0.60; 0.54–0.68) were negatively associated with ExRes. Pertuzumab use was also associated with ExRes (yes v no OR 1.39; 1.10–1.77). Conclusions: These data from a real-world setting demonstrate that a substantial minority of pts achieve excellent outcomes with trastuzumab-based 1L therapy, confirm the benefit of pertuzumab in optimizing long-term outcomes, and are aligned with data from clinical trials. Factors associated with ExRes include metastatic sites, de novo disease, HER2 IHC 3+ status, and pertuzumab use. Despite these promising findings, further development of predictive models and validation in broader cohorts are needed to improve early identification accuracy and optimize treatment strategies.