<h3>Purpose/Objective(s)</h3> Trials suggest benefit from regional nodal irradiation (RNI) in node-positive breast cancer patients (pts), but absolute benefit is not uniform. SWOG S1007 randomized pts with hormone receptor-positive, HER2-negative breast cancer and 21-gene assay recurrence score (RS) ≤25 to endocrine therapy alone (ET) vs chemotherapy then ET. Little is known about modern RNI practice patterns or rates of locoregional recurrence (LRR) in such pts. Here, we report use of radiation therapy (RT) and patterns of LRR on S1007. <h3>Materials/Methods</h3> RT receipt, targets, and dose were prospectively collected. For analysis of LRR, invasive disease-free survival (IDFS) was separated into local/regional events while other IDFS events were competing risks in a Cox model. Since RT occurred after randomization, landmarked analysis starting at 1 year was used for analysis of LRR incidence and IDFS. Analyses adjusted for menopausal status, treatment arm, RS, tumor size, and number of positive nodes. <h3>Results</h3> Of the 4,983 eligible pts, 4,871 (97.8%) had RT forms and 3,947 (81.0%) reported RT receipt. Of pts who had breast conservation (BCS), 95.7% received RT. Of pts who received BCS+RT, 52.7% had RNI (at least a supraclavicular field). Predictors of RNI after BCS were larger tumor (p<0.0001) and more positive nodes (p<0.0001), but not RS (p=0.92). For pts who had mastectomy (mast), 54.2% received RT, of whom 81.1% had RNI. Significant predictors of RT among mast pts were being premenopausal (p=0.022), larger tumor size (p<0.0001), and more positive nodes (p<0.0001), but not RS (p=0.33). The cumulative incidence of LRR between 1-5 years was 0.85% among pts with BCS+RT with RNI, 0.55% after BCS+RT w/o RNI, 0.11% after mast+RT, and 1.7% after mast w/o RT. In premenopausal pts, BCS+RT had reduced LRR compared to mast w/o RT (HR=0.41, 95% CI 0.19-0.91, p=0.028) adjusting for RS, systemic treatment, tumor size, and # of positive nodes. The reduction in postmenopausal pts was not significant (HR=0.48, 95% CI 0.20-1.13, p=0.09) with similar adjustment. IDFS did not differ among BCS+RT, mast w/o RT, mast+RT groups in premenopausal (p=0.14) or postmenopausal (p=0.98) pts after adjustment for other factors. IDFS also did not differ in these 3 groups in the ET arm after adjustment for other factors. <h3>Conclusion</h3> On S1007, practice was evenly divided for use of RNI in the setting of N1 disease & favorable biology. Rates of LRR were low even in pts who did not receive RNI. IDFS did not differ across locoregional treatment groups, even in the ET only arm. Omission of chemotherapy among pts eligible for this trial does not appear to be an indication for use of RNI. These findings reinforce the importance of prospective trials like NCIC MA-39, randomizing pts with RS ≤25 to RNI or not.
Read full abstract