21-Gene Recurrence Score Predictive for Patients Most Likely to Benefit From Postoperative Radiation Following Breast Conserving Surgery for T1N0 Breast Cancer

Abstract

<h3>Purpose/Objective(s)</h3> The 21- gene RT-PCR recurrence score (RS) is performed in patients with hormone receptor positive, HER2 negative, node- negative breast cancer to determine which patients will likely benefit from chemotherapy following surgery. Whether the RS can predict for patients likely to benefit from radiation therapy (RT) following breast conserving surgery (BCS) has not previously been examined. The purpose of this study was to use a large set of patients to determine if RS could predict for patients who would benefit from RT following BCS. <h3>Materials/Methods</h3> The National Cancer Database (NCDB) was queried (2004-2017) for female patients with pT1N0 ER+ PR+ HER2- breast cancer treated with lumpectomy and an available RS. Patients were stratified based on their RS (low risk [LR] = 1-10, intermediate risk [IR] = 11-25, high risk [HR] = 26-99). Propensity score matching (PSM) was conducted to create a 1:1 matched cohort of patients who received radiotherapy and patients who did not. Kaplan-Meier analysis with log-rank testing was used to evaluate overall survival (OS). Univariable (UVA) and multivariable (MVA) analysis were conducted using Cox proportional hazard models to determine which clinical and treatment factors were prognostic for OS in both the overall population and after stratifying patients by RS. <h3>Results</h3> A total of 89,604 patients met the selection criteria: 21,414 in the LR cohort, 59,157 in the IR cohort, and 9,033 in the HR cohort. A total of 92% received RT: 91% in the LR cohort, 93% in the IR cohort, and 92% in the HR cohort. Overall, the 5-year OS was 97% for those who received RT and 91% for those who did not (<i>P</i> < 0.001). After PSM, the 5-year OS in the LR cohort was 94% for those who received RT and 92% for those who did not (<i>P</i> = 0.054). In the IR cohort, the 5-year OS was 95% for those who received RT and 92% for those who did not (<i>P</i> < 0.001). In the HR cohort, the 5-year OS was 93% for those who received RT and 85% for those who did not (<i>P</i> = 0.001). On MVA in the overall cohort, RT (<i>P</i> < 0.001) and endocrine therapy were predictive of improved OS while increasing age, black race, CDCC comorbidity score, tumor size, and increasing RS were predictive for worse OS. On MVA in the LR cohort, RT (<i>P</i> = 0.113) was not predictive of improved OS. However, MVA demonstrated that RT was a positive prognostic factor for OS in both the IR cohort (<i>P</i> < 0.001) and HR cohort (<i>P</i> < 0.001). Chemotherapy was not predictive of improved OS overall or in any cohort on MVA. <h3>Conclusion</h3> This is the first study investigating the predictive value of RS in patients treated with BCS with respect to use of RT. The results suggest that patients with an IR or HR RS may be more likely to benefit from postoperative RT than patients with a LR RS, as an OS benefit was observed with RT use in patients with IR or HR RS, but not in patients with LR RS. Further prospective evaluation is warranted to define the use of RS to guide management with RT.