Breast-Conserving Therapy Including Whole-Breast Irradiation is Associated with Better Overall Survival Compared to Mastectomy Alone in Patients with Early-Stage, Lymph Node-Negative Breast Cancer with High OncotypeDX Recurrence Score: an NCDB Analysis

Document Type


Publication Date


Publication Title

International Journal of Radiation Oncology, Biology, Physics


Purpose/Objective(s) Previous single-institutional retrospective studies suggest improved overall survival (OS) in women with early stage, lymph node negative breast cancer treated with breast conserving therapy including whole breast irradiation (BCT), compared to those treated with mastectomy alone. In this study, we compared OS between these two cohorts using the NCDB. Additionally, we examined OS after stratifying by OncotypeDX recurrence score. Materials/Methods We performed a retrospective cohort analysis of NCDB patients with non-metastatic, pT1-2, pN0 breast cancer treated between 2006 and 2014. Patients were classified into two categories: those undergoing BCT (breast conserving surgery followed by whole breast irradiation), and those undergoing mastectomy without post-mastectomy irradiation. Patients were matched using propensity scores with inverse probability of treatment weighting (IPTW) with stabilized weights. Patients were matched for race, age, grade, pT stage, LVI, receptor status, hormone therapy, chemotherapy, OncotypeDX recurrence score, comorbidity score, year of diagnosis, treatment facility type, insurance status, median income, US region, laterality, and tumor quadrant. The primary objective was to compare OS between BCT and mastectomy patients. Results After IPTW-matching, the analysis included 149,273 and 87,073 patients undergoing BCT and mastectomy, respectively. There were no differences between the matched groups in any of the included variables after matching. Median follow up for the matched cohorts were 48.5 (BCT) and 47.3 months (mastectomy). OS was better in the BCT group at 5 years (94% vs 92%, P<0.0001) and 7 years (81.1% vs 73.3%, P<0.0001). In univariate Kaplan-Meier (UV) and multivariate cox proportional hazard analyses (MVA), treatment with BCT was a positive predictor of OS with a mastectomy HR of 1.4 in MVA (95% CI 1.3-1.4, P=0.0001), when setting BCT as reference. We then stratified patients by OncotypeDX RS into low RS (≤25), 36, 571 patients with BCT and 17,785 with mastectomy, or high RS (>25), 7,018 patients with BCT and 3,811 patients with mastectomy. Median follow up for patients with available RS were 42.3 months (BCT) and 41.6 months (mastectomy). OS was not different for patients undergoing BCT or mastectomy in the low RS cohort at 5 or 7 years (P=0.339 and 0.289). For those with high RS, OS was better in the BCT group at 5 years (95% vs 93%, P=0.001) and 7 years (93% vs 88%, P=0.0001). In MVA, treatment with BCT was a positive predictor of OS with a mastectomy HR of 1.5 (95% CI 1.2-1.9, P<0.0001), when setting BCT as reference. Conclusion Our analysis suggests that BCT is associated with better OS compared to mastectomy alone in patients with early stage, lymph node negative breast cancer with a high OncotypeDX recurrence score. We hypothesize that incidental irradiation to regional lymphatics, may underly the improvement in OS with BCT in this high-risk cohort. Further investigation into patterns of failure in these patient cohorts is needed.





First Page


Last Page