Patterns of Radiation Therapy Boost Administration in 356,160 Breast Cancer Patients Treated With Hypofractionated Versus Conventional Whole-Breast Irradiation

Abstract

Purpose/Objective(s): Randomized trials have demonstrated the local control benefit of radiation therapy boost (RB) to the lumpectomy cavity following whole breast irradiation (WBI) in breast cancer patients treated with conventional fractionation (C-WBI). However, evidence supporting the use of RB in breast cancer patients treated with hypofractionated whole breast irradiation (H-WBI) is lacking. The purpose of this study was to compare the frequency of RB administration in National Cancer Data Base (NCDB) breast cancer patients treated with H-WBI and C-WBI and to identify factors associated with RB treatment. Materials/Methods: Using the NCDB, 356,160 women diagnosed with pathologic Tis-3, N0-1, M0 breast cancer between 2003 and 2013 were identified. All patients received breast conserving surgery followed by WBI. The frequency of RB administration in H-WBI and C-WBI patients was compared. Logistic multivariable analysis was performed to examine demographic, patient, and tumor characteristics associated with RB treatment in the H-WBI and C-WBI patients. Results: Among 356,160 patients, 87.5% received conventional WBI while 12.5% received H-WBI. RB was administered more frequently in women treated with C-WBI versus (vs) H-WBI (88.9% v. 52.2%; P < 0.001). Among H-WBI patients, RB was administered more frequently in women treated in 2004 (74%) and 2005 (60%) than in patients treated between 2006 and 2012 (<55%, P < 0.001). Median RB dose was 10.0 Gy (90% CI: 8.0-16.0 Gy) in H-WBI patients vs. 12.0 Gy (90% CI: 10.0-16.0 Gy) in C-WBI patients. On multivariable analysis, factors associated with RB treatment in H-WBI patients were younger age (age<50 vs >50 years), treatment at community cancer program or comprehensive community cancer program versus academic center, higher T stage, poorly differentiated tumor, node positive disease, positive surgical margins, and treatment with chemotherapy (all P < 0.001). Interestingly, younger age, treatment at a community cancer program or comprehensive community program versus academic center, higher T stage, and treatment with chemotherapy (all P < 0.02), were also associated with RB administration in C-WBI patients but tumor grade, positive nodes, and positive margins were not. Conclusion: Between 2003 and 2013, the majority of C-WBI breast cancer patients received a RB irrespective of patients and tumor characteristics. Although H-WBI patients were less likely to receive a RB, those treated with RB had patient or tumor characteristics consistent with National Comprehensive Cancer Network Guidelines for RB administration including young age, poorly differentiated tumors, and positive surgical margins. Future studies are needed to clarify appropriate RB practice in the setting of H-WBI.