Association of Tumor Stage and Molecular Subtype With Three-Year Recurrence Outcomes After Adjuvant Radiotherapy in Stage I-III Breast Cancer: A Single-Centre 2022 Cohort Study

Background Breast cancer remains a major clinical challenge, and recurrence after curative-intent treatment continues to influence prognosis and follow-up strategies. Tumor stage and molecular subtype are central to treatment planning and risk stratification, yet real-world recurrence data from low- and middle-income country (LMIC) settings, where treatment access, timing, and completion may diverge substantially from controlled trial environments, remain limited. We evaluated the association of stage and molecular subtype with three-year recurrence outcomes in a single-center Stage I-III breast cancer cohort treated with adjuvant radiotherapy. Methods This retrospective cohort study included consecutive Stage I-III breast cancer patients diagnosed in 2022 and treated with adjuvant radiotherapy at a tertiary cancer centre in Lahore, Pakistan. The primary endpoint was recurrence-free survival (RFS), defined from the biopsy date to first locoregional recurrence, distant recurrence, or death. Secondary endpoints were distant metastasis-free survival (DMFS) and locoregional recurrence-free survival (LRRFS). Kaplan-Meier curves with log-rank testing compared outcomes by stage group (I-II vs III) and molecular subtype. Multivariable Cox proportional hazards regression was used for adjusted RFS and DMFS analyses, with a further radiotherapy-start-anchored sensitivity analysis for RFS. Results Of 195 patients analyzed, 115 (59.0%) had Stage I-II disease, and 80 (41.0%) had Stage III disease. Molecular subtypes were Luminal A (57.4%), Luminal B (17.9%), HER2-enriched (10.8%), and triple negative (13.8%). RFS events occurred in 64 patients: 54 unique patients experienced recurrence, including seven with both distant and locoregional recurrence, while 10 additional patients died without prior documented recurrence. Stage III disease was independently associated with significantly worse RFS (hazard ratio (HR) 2.182, 95% CI 1.266-3.761; p = 0.005) and DMFS (HR 2.429, 95% CI 1.337-4.411; p = 0.004) in multivariable models. The stage effect persisted in both sensitivity analyses. Molecular subtype was not significantly associated with any recurrence endpoint on univariable or multivariable testing. Conclusion Tumor stage was the most consistent and robust predictor of recurrence outcomes after adjuvant radiotherapy in this single-centre cohort. Although molecular subtype did not independently predict recurrence in these analyses, likely reflecting statistical limitations inherent to smaller, real-world datasets, these findings support stage-based risk stratification in routine clinical practice. Larger multicentre cohorts with complete systemic treatment data are needed to better characterise subtype-specific recurrence patterns in LMIC settings.