Biomarker-based risk stratification for predicting late recurrence and benefit of extended endocrine therapy: A secondary analysis of the IDEAL trial.

BACKGROUND: Hormone receptor-positive, HER2-negative (HR+/HER2-) early breast cancer (EBC) is characterized by a persistent risk of late recurrences up to 15 years from diagnosis. Phase III randomized IDEAL (Investigation on the Duration of Extended Adjuvant Letrozole) trial was conducted to optimize the duration of extended endocrine therapy (EET). This blinded prospective-retrospective study investigates the potential of CanAssist Breast (CAB) to predict late recurrences and identify patients who may benefit from EET. METHODS: Study includes: i) pooled cohort of 3052 HR+/HER2- EBC patients from previously published CAB retrospective studies; ii)subset of 627 patients from the original IDEAL cohort treated with either 2.5- or 5-years of EET and followed up for 15 years. CAB stratifies patients as low-risk (LR) or high-risk (HR), with ≤9% or 9.1-100% risk of distant recurrence. CAB-HR category was refined to include an ultra-high-risk (UHR) subgroup (>20% recurrence risk) to identify patients at elevated risk of recurrence, and assess potential EET benefit using Kaplan-Meier analysis. RESULTS: In the pooled 3052 and IDEAL cohorts, CAB-LR:HR:UHR proportions were 69:20:11 and 60:28:12. CAB risk score distributions across the cohorts were similar, despite the IDEAL cohort consisting 73% node-positive patients compared to 43% in pooled cohort. CAB-UHR group showed improvement in DRFS in the 5-year versus 2.5-year EET arms (ΔDRFS-10%) compared to CAB-LR (ΔDRFS-1%) and CAB-HR (ΔDRFS-1.5%) groups. CONCLUSION: The refined CAB-based risk stratification identifies a UHR category in the IDEAL cohort, characterized by an increased risk of late recurrence up to 15 years from surgery, who may benefit from EET.