Background
- Hormone receptor-positive (HR+) early breast cancer carries persistent late distant recurrence (DR) risk. 1
- Extended endocrine therapy (EET; extended letrozole in NSABP B-42) reduces recurrence in some patients, but benefit must be weighed against toxicity, adherence and heterogeneous baseline risk. 2
- Multigene assays (Oncotype DX, MammaPrint, Prosigna, EndoPredict, Breast Cancer Index [BCI]) are prognostic; BCI and MammaPrint have shown predictive signals for EET in trial cohorts, with modest absolute benefit or secondary-endpoint effects, and all require molecular testing. 3–10
- We previously developed and externally validated LDR-R, a prognostic model combining hematoxylin & eosin (H&E) images and clinical variables to predict late DR (10-year absolute risk difference 7.95%; hazard ratio [HR] 5.71; 95% confidence interval [CI] 3.50–9.32). 11
- Routine H&E images, combined with clinicopathologic factors, offer a way to identify EET benefit within the context of late DR risk without molecular testing.