Whole Breast Irradiation in Comparison to Endocrine Therapy in Early Stage Breast Cancer-A Direct and Network Meta-Analysis of Published Randomized Trials.

Abstract

Multiple randomized trials have established adjuvant endocrine therapy (ET) and whole breast irradiation (WBI) as the standard approach after breast-conserving surgery (BCS) in early-stage breast cancer. The omission of WBI has been studied in multiple trials and resulted in reduced local control with maintained survival rates and has therefore been adapted as a treatment option in selected patients in several guidelines. Omitting ET instead of WBI might also be a valuable option as both treatments have distinctly different side effect profiles. However, the clinical outcomes of BCS + ET vs. BCS + WBI have not been formally analyzed. We performed a systematic literature review searching for randomized trials comparing BCS + ET vs. BCS + WBI in low-risk breast cancer patients with publication dates after 2000. We excluded trials using any form of chemotherapy, regional nodal radiation and mastectomy. The meta-analysis was performed using a two-step process. First, we extracted all available published event rates and the effect sizes for overall and breast-cancer-specific survival (OS, BCSS), local (LR) and regional recurrence, disease-free survival, distant metastases-free interval, contralateral breast cancer, second cancer other than breast cancer and mastectomy-free interval as investigated endpoints and compared them in a network meta-analysis. Second, the published individual patient data from the Early Breast Cancer Trialists' Collaborative Group (EBCTCG) publications were used to allow a comparison of OS and BCSS. We identified three studies, including a direct comparison of BCS + ET vs. BCS + WBI (n = 1059) and nine studies randomizing overall 7207 patients additionally to BCS only and BCS + WBI + ET resulting in a four-arm comparison. In the network analysis, LR was significantly lower in the BCS + WBI group in comparison with the BCS + ET group (HR = 0.62; CI-95%: 0.42-0.92; p = 0.019). We did not find any differences in OS (HR = 0.93; CI-95%: 0.53-1.62; p = 0.785) and BCSS (OR = 1.04; CI-95%: 0.45-2.41; p = 0.928). Further, we found a lower distant metastasis-free interval, a higher rate of contralateral breast cancer and a reduced mastectomy-free interval in the BCS + WBI-arm. Using the EBCTCG data, OS and BCSS were not significantly different between BCS + ET and BCS + WBI after 10 years (OS: OR = 0.85; CI-95%: 0.59-1.22; p = 0.369) (BCSS: OR = 0.72; CI-95%: 0.38-1.36; p = 0.305). Evidence from direct and indirect comparison suggests that BCS + WBI might be an equivalent de-escalation strategy to BCS + ET in low-risk breast cancer. Adverse events and quality of life measures have to be further compared between these approaches.