Age-dependent relative and absolute benefit of radiotherapy after breast-conserving surgery: a population-based...

Radiotherapy (RT) after breast-conserving surgery (BCS) improves breast cancer outcomes. However, its clinical relevance may vary with age, particularly in the presence of competing risks and treatment burden. How the association between RT and breast cancer-specific survival changes across the age spectrum has not been well characterized.

Patients aged ≥ 50 years who underwent BCS for breast cancer were identified from the SEER database between 2000 and 2022. Patients with unknown or refused RT status, unknown cause of death, or follow-up < 3 months were excluded. Inverse probability of treatment weighting (IPTW)-adjusted Cox proportional hazards models with age-by-treatment interaction terms were used to estimate age-specific associations between RT and breast cancer-specific survival (BCSS). Absolute effects were assessed using cumulative incidence functions accounting for competing risks, and continuous age effects were evaluated using restricted cubic spline models.

Among 532,118 eligible patients, 409,652 (77.0%) received RT and 122,466 (23.0%) did not. RT was associated with improved BCSS across all age groups. The relative association attenuated with increasing age, with hazard ratios increasing from 0.54 (95% CI, 0.51-0.58) among patients aged 50-54 years to 0.81 (95% CI, 0.75-0.88) among those aged ≥ 85 years (p for interaction < 0.001). Absolute risk reduction also decreased with age, from approximately 5.2-5.6% in younger patients to 2.9-3.7% in patients aged 70-84 years. Restricted cubic spline analysis showed a gradual attenuation of the RT-associated survival advantage with age, without evidence of an abrupt threshold.

RT after BCS was associated with improved BCSS across the age spectrum, although both relative and absolute benefits became smaller with increasing age. These findings support individualized treatment decisions that consider absolute benefit, competing risks, and patient-specific factors rather than relying only on fixed age thresholds.