Salvage Mastectomy Versus Second Conservative Treatment for Second Ipsilateral Breast Tumor Event: A Propensity Score-Matched Cohort Analysis of the GEC-ESTRO Breast Cancer Working Group Database

Purpose Second conservative treatment has emerged as an option for patients with a second ipsilateral breast tumor event after conserving surgery and breast irradiation. We aimed to address the lack of evidence regarding second breast event treatment by comparing oncologic outcomes after conservative treatment or mastectomy. Methods and Materials Oncologic outcomes were analyzed using a propensity score-matched cohort analysis study on patients who received a diagnosis of a second breast event between January 1995 and June 2017. Patient data were collected from 15 hospitals/cancer centers in 7 European countries. Patients were offered mastectomy or lumpectomy plus brachytherapy. Propensity scores were calculated with logistic regression and multiple imputations. Matching (1:1) was achieved using the nearest neighbor method, including 10 clinical/pathologic data related to the second breast event. The primary endpoint was 5-year overall survival from the salvage surgery date. Secondary endpoints were 5-year cumulative incidence of third breast event, regional relapse and distant metastasis, and disease-free and specific survival. Complications and 5-year incidence of mastectomy were investigated in the conservative treatment cohort. Results Among the 1327 analyzed patients (mastectomy, 945; conservative treatment, 382), 754 were matched by propensity score (mastectomy, 377; conservative treatment, 377). The median follow-up was 75.4 months (95% confidence interval [CI], 65.4-83.3) and 73.8 months (95% CI, 67.5-80.8) for mastectomy and conservative treatment, respectively (P = .9). In the matched analyses, no differences in 5-year overall survival and cumulative incidence of third breast event were noted between mastectomy and conservative treatment (88% [95% CI, 83.0-90.8] vs 87% [95% CI, 82.1-90.2], P = .6 and 2.3% [95% CI, 0.7-3.9] vs 2.8% [95% CI, 0.8-4.7], P = .4, respectively). Similarly, no differences were observed for all secondary endpoints. Five-year cumulative incidence of mastectomy was 3.1% (95% CI, 1.0-5.1). Conclusions To our knowledge, this is the largest matched analysis of mastectomy and conservative treatment combining lumpectomy with brachytherapy for second breast events. Compared with mastectomy, conservative treatment does not appear to be associated with any differences in terms of oncologic outcome. Consequently, conservative treatment could be considered a viable option for salvage treatment. Second conservative treatment has emerged as an option for patients with a second ipsilateral breast tumor event after conserving surgery and breast irradiation. We aimed to address the lack of evidence regarding second breast event treatment by comparing oncologic outcomes after conservative treatment or mastectomy. Oncologic outcomes were analyzed using a propensity score-matched cohort analysis study on patients who received a diagnosis of a second breast event between January 1995 and June 2017. Patient data were collected from 15 hospitals/cancer centers in 7 European countries. Patients were offered mastectomy or lumpectomy plus brachytherapy. Propensity scores were calculated with logistic regression and multiple imputations. Matching (1:1) was achieved using the nearest neighbor method, including 10 clinical/pathologic data related to the second breast event. The primary endpoint was 5-year overall survival from the salvage surgery date. Secondary endpoints were 5-year cumulative incidence of third breast event, regional relapse and distant metastasis, and disease-free and specific survival. Complications and 5-year incidence of mastectomy were investigated in the conservative treatment cohort. Among the 1327 analyzed patients (mastectomy, 945; conservative treatment, 382), 754 were matched by propensity score (mastectomy, 377; conservative treatment, 377). The median follow-up was 75.4 months (95% confidence interval [CI], 65.4-83.3) and 73.8 months (95% CI, 67.5-80.8) for mastectomy and conservative treatment, respectively (P = .9). In the matched analyses, no differences in 5-year overall survival and cumulative incidence of third breast event were noted between mastectomy and conservative treatment (88% [95% CI, 83.0-90.8] vs 87% [95% CI, 82.1-90.2], P = .6 and 2.3% [95% CI, 0.7-3.9] vs 2.8% [95% CI, 0.8-4.7], P = .4, respectively). Similarly, no differences were observed for all secondary endpoints. Five-year cumulative incidence of mastectomy was 3.1% (95% CI, 1.0-5.1). To our knowledge, this is the largest matched analysis of mastectomy and conservative treatment combining lumpectomy with brachytherapy for second breast events. Compared with mastectomy, conservative treatment does not appear to be associated with any differences in terms of oncologic outcome. Consequently, conservative treatment could be considered a viable option for salvage treatment.