Abstract PS5-01-10: Estimating the effect of initiating early maintenance endocrine therapy on brain metastases-free survival and other clinical outcomes in patients with HER2+/HR+ mBC without brain metastases: A target trial emulation

Abstract Background: Maintenance endocrine therapy (mET) is increasingly used as part of the post-chemotherapy treatment for HER2-positive, hormone receptor-positive metastatic breast cancer (HER2+, HR+ mBC). However, the impact of mET on delaying brain metastases (BM) and other clinical outcomes is unclear. As the time from taxane completion to mET initiation can vary considerably, standard cohort methods with fixed index dates can lead to immortal time and selection bias. This study estimates the effect of different mET treatment strategies (TS) on BM-free survival (BMFS), overall survival (OS), and real-world progression-free survival (rwPFS) in patients (pts) with HER2+, HR+ mBC without BM. Methods: The study used the Flatiron Health database (January 2011-September 2024). mET included aromatase inhibitors, selective estrogen receptor degraders or modulators, and ovarian function suppression. Our target trial compared four TS in pts with HER2+, HR+ mBC who had completed ≥5 cycles of 1L HER2-targeted therapy + taxane (baseline was last taxane administration): Arm 1: mET initiation within 3 months (mo) of baseline; Arm 2: no mET initiation within 3 mo of baseline (initiation was possible afterwards); Arm 3: no mET initiation within 60 mo of baseline; and Arm 4: mET initiation between 4-6 mo of baseline. In all TS, CDK4/6 inhibitor use was not permitted and pts who initiated 2L treatment/had a BMFS/OS/rwPFS event were excluded from the TS. In the target trial emulation (TTE), pts were assigned to all TS with which their baseline data were compatible, creating four clones (TS were not distinguishable at baseline), and were censored upon deviation from the assigned TS. BMFS, OS, rwPFS were measured up to 60 mo from last taxane administration (LTA); risk differences and risk ratios were estimated via pooled logistic regression adjusted for baseline confounders (de novo vs. recurrent disease, disease-free interval, treatment at an academic center, Roche Prognostic Score, last HR test being negative despite prior positive tests). Estimates were bounded with percentile-based 95% confidence intervals (CI) estimated via bootstrapping. Sensitivity analyses included changes in grace period, time discretization, functional form of time, censoring of early death events, and simulation of uniform initiation of TS. Results: The TTE (N = 784) evaluated 391 pts who received ET within 3 mo and 393 pts who did not receive ET within 3 mo. At the time of taxane completion, baseline characteristics were generally balanced between pt groups, except ET pts were more likely to be treated in an academic center (25.8% vs. 17.3%) and less likely to have a negative latest HR test (3.6% vs. 12.7%). The difference in the risk of death (95% CI) comparing Arm 1 vs. Arm 3 was -3.2% (-5.6, -0.8) at 12 mo and -6.5% (-10.9, -1.7) at 24 mo (negative estimates favor Arm 1). The corresponding risk differences for rwPFS were -6.8% (-16.0, 0.9) and -10.1% (-20.3, -1.9), and for BMFS, -2.0% (-8.9, 4.7) and -4.7% (-13.0, 2.8). Treatment effect estimates were less precise beyond 36 mo due to loss to follow-up. Similar estimates were found against other TS, although the limited number of pts with data compatible with Arm 4 yielded imprecise estimates. Conclusions: Early mET within 3 mo of LTA was associated with improved clinical outcomes in pts with HER2+, HR+ mBC (although the limited number of BMFS events and modest estimates prevent making any firm conclusions). The TTE framework for comparative effectiveness studies with flexible treatment initiation helps align real-world data with randomized clinical trials, supporting better design of real-world data studies including comparisons of TS that can inform clinical practice. Citation Format: T. Sanglier, I. Gravestock, J. Leone, M. Secrest, C. Tchakoute, E. Restuccia, F. Montemurro, M. Shivhare, A. Knott, P. Lambert, D. Martinez, P. Luhn, X. Garcia de Albeniz Martinez, S. Tolaney, N. Lin, S. Sammons. Estimating the effect of initiating early maintenance endocrine therapy on brain metastases-free survival and other clinical outcomes in patients with HER2+/HR+ mBC without brain metastases: A target trial emulation [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS5-01-10.