Survival after breast-conserving surgery and radiotherapy versus mastectomy: propensity score analyses within a...

Large randomized studies show equivalent survival after breast-conserving surgery (BCS) + adjuvant radiotherapy (RT) and mastectomy. In contrast, more recent observational studies suggest BCS + RT to be superior, but it is questionable whether patients in these treatment arms are comparable. Here, overall survival (OS) and breast cancer-specific survival (BCSS) after BCS + RT and mastectomy are compared within a randomized trial comparing intravenous and inhalation anaesthesia during breast cancer surgery.

The patient cohort was recruited from the randomized CAN-study. Patients with tumours >30 mm, patients with tumours <10 mm, and patients who underwent BCS without RT were excluded. OS and BCSS were estimated using multivariable Cox regression analyses and three different propensity score models.

The final study cohort included 830 women, of whom 601 underwent BCS + RT (median age 64 years) and 229 underwent mastectomy (median age 68 years). Women who underwent mastectomy had more co-morbidities and more unfavourable tumour characteristics. Mastectomy was associated with significantly less favourable OS in unadjusted, adjusted, and two of the three propensity score analyses. BCSS was inferior in the mastectomy group in the unadjusted analysis, with an HR of 2.27 (95% c.i. 1.20 to 4.30). In the adjusted and three propensity score analyses, BCSS was equal in the treatment groups, with an adjusted HR of 1.02 (95% c.i. 0.43 to 2.42).

In this study, which was designed to approximate a randomized trial as closely as possible, no significant difference in BCSS was observed between BCS + RT and mastectomy. Differences in OS likely reflect occult selection bias of patients with higher co-morbidity burden to mastectomy.