Abstract PS4-07-02: Determinants of early and late relapse in patients with early triple negative breast cancer from the prospective CANTO Cohort

Résumé fourni par la source

Abstract Background: TNBC exhibits a high risk of early relapse, yet predictors of time to relapse remain poorly defined. Early relapses (while on or within 12 months after ending curative therapies) are associated with chemotherapy-resistant disease and poor prognosis (median overall survival <12 months; Grinda et al., Eur J Cancer 2023). Identifying key factors associated with early and late relapse may help guide therapeutic interventions. This study explores associations between clinical factors and timing of relapse in patients (pts) with TNBC from the prospective CANTO cohort (NCT01993498). Methods: Early-stage TNBC pts enrolled in the CANTO cohort between 2012 and 2023 were included. Descriptive statistics were used to summarize clinical variables. Invasive disease-free survival (IDFS) was estimated with Kaplan-Meier curves from the time of study entry. The proportional hazards assumption was assessed with Schoenfeld residuals and cumulative hazard plots. A time-varying fixed-effect multivariable Cox model was built using stratification at 24 months for distinguishing early and late relapses. To identify the best-fitting multivariable model for IDFS, stepwise selection (stepAIC) was applied starting from a priori covariates, including age, grade, surgery type, and TNM stage , with additional candidates tested based on univariate results and clinical relevance. Results: Among the 999 pts included, median age was 52.9 years (IQR: 45.1-61.9). Most pts presented with stage II disease (56.3%), followed by stage I (34.4%) and stage III (9.3%). Chemotherapy was administered in 93.5% of pts: adjuvant in 51.3% and neoadjuvant in 42.2% of cases. Breast-conserving surgery was performed in 71.7% pts, mastectomy in 28.3%. Around half of the pts underwent axillary dissection (47.6%).Median follow-up was 74.3 months (IQR: 51.2-99.4). During this period, 239 iDFS events occurred, 49% within the first 24 months. The best final time-varying fixed effect multivariate model for predicting iDFS included age, BMI at diagnosis, smoking, type of surgery, grade, stage and chemotherapy exposure. TNM stage showed a strong prognostic role across the entire follow-up, with stage III disease consistently associated with higher relapse risk compared to stage I (HR: 2.44; 95% CI: 1.44-4.14; p < 0.001). For early relapse, underweight status (BMI < 18.5 Kg/m2) was the strongest predictor (HR: 3.35; 95% CI: 1.58-7.11; p = 0.002). Axillary dissection (HR: 1.60; 95% CI: 1.00-2.55; p = 0.05) and receipt of (neo)adjuvant chemotherapy (HR: 1.51; 95% CI: 0.95-2.40; p = 0.078) trended toward a signifcant increased risk of early relapse. Despite adjustment for TNM stage to account for disease severity, these treatment-related variables may act as proxies for disease burden, and some residual counfounding cannot be excluded. Late relapse was associated with older age (HR: 1.03; 95% CI: 1.01-1.04; p = 0.003), while smokers had a borderline increased risk versus non-smokers (HR: 1.61; 95% CI: 0.98-2.63; p = 0.06). Conclusions: This large cohort study shows that early and late relapses in early-stage TNBC are driven by distinct clinical factors. Our findings integrate clinicopathological variables to support the development of models reflecting individual relapse risk and timing. Early relapses, accounting for half of the events, were strongly associated with underweight status, possibly reflecting patient frailty and comorbidities, and high disease burden at diagnosis, as indicated by stage III disease and high treatment intensity, including axillary dissection and chemotherapy. Late relapses were predicted by older age and stage III disease. These data support time-aware prognostic models to guide individual-risk adapted strategies and underscore the need for biological investigations to better understand early relapse in TNBC. Citation Format: C. Benvenuti, L. Perotti, A. Martin, C. Gaudin, C. Jouannaud, M. Fournier, C. Kaderbhaï, A. Kieffer, F. Cherifi, M. Campone, F. Lerebours, P. Cottu, F. André, A. Bertaut, D. Antonio, S. Michiels, B. Pistilli, T. Grinda. Determinants of early and late relapse in patients with early triple negative breast cancer from the prospective CANTO Cohort [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 PS4-07-02.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.