Survival outcomes and prognostic factors in luminal A (LumA) early breast cancer (EBC): Real-world evidence from a public cancer center in Peru.

e23406 Background: Luminal A (LumA) early breast cancer (EBC) generally has excellent prognosis (5-year survival ≥ 95%) due to favorable biology and high response to endocrine therapy (ET). However, prognosis is strongly associated with anatomic burden (T/N and stage). In resource-limited settings lacking genomic assays, adjuvant decisions rely on clinicopathological factors and may affect real-world outcomes. We described outcomes and prognostic factors in Peruvian LumA EBC. Methods: Retrospective cohort of LumA (defined as ER/PgR-positive, HER2-negative, and ki-67≤20% by immunohistochemistry) EBC pts treated at a public cancer center in Peru (2018-2022). RFS and OS were estimated by Kaplan-Meier. Cox models evaluated factors associated with recurrence and death (two-sided p < 0.05). Results: A total of 151 pts were included; median age was 55 years (48-61) and 61% were postmenopausal. Most tumors had ductal histology (89%) and grade 2 (39%). T2 disease was most frequent (40.5%). Nodal status: N0 28.5%, N1 69%, N3 2.5%. Stage distribution was predominantly IIA (59%) and I (24%). Breast-conserving surgery was performed in 55%. Adjuvant ET was administered in 87%, (anastrozole, 43%); adjuvant chemotherapy and radiotherapy were given in 52% and 66.5%, respectively; among stages I and IIA, 32% and 62% received adjuvant chemotherapy (gene expression assays were not available). After a median follow-up of 42 months (3-79), 42 recurrences occurred; 1-, 3- and 5-year RFS were 96%, 84% and 69% (median RFS not reached). Node-positive involvement was associated with higher risk vs. node-negative (HR: 2.12, 95% CI, 1.01-4.45; p = 0.044). In stage IIA, chemotherapy followed by ET was associated with higher recurrence risk compared with ET alone (HR 3.45, 95% CI, 1.42-8.37; p = 0.006), likely reflecting confounding by indication. After a median follow-up of 54 months, 19 deaths occurred; 1-, 3- and 5-year OS were 100%, 93% and 89% (median OS not reached). Stages IIB-III were associated with inferior OS vs. stages I-IIA (HR 3.14, 95% CI, 1.27-7.74, p = 0.013). Conclusions: In this real-world Peruvian cohort of LumA EBC, nodal involvement and higher stage were stronger prognostic factors. Lower 5-year RFS/OS than typically reported may reflect higher baseline anatomic risk and tertiary-center case mix. In stage IIA, the association between chemotherapy use and recurrence reflects treatment selection based on clinicopathological risk and chemotherapy prioritization for high-risk lacking genomic stratification. These outcomes highlight the need to optimize risk stratification when genomic assays are unavailable.