Patterns of locoregional failure following radiotherapy for breast cancer.

Abstract

118 Background: To determine the influence of radiotherapy (XRT) technique on local control in breast cancer (BRCA) patients, we evaluated the relationship between PET/CT-identified locoregional recurrences (LRR) and prior definitive XRT treatment plans. Methods: Twenty-two consecutive LRRs of BRCA were identified by PET/CT in 19 patients treated from 2004-2011 with definitive post-operative radiation for their BRCA. PET/CT and original XRT CT simulation images were fused using a B-spline deformable registration algorithm (VelocityAI, Velocity Medical Solutions, Atlanta, GA) to correct for posture and soft tissue changes. Prior XRT dose to the FDG avid LRR region was calculated. LRR was classified as in-field, marginal or out-of-field. In-field LRRs were defined as ≥95% of the LRR volume receiving ≥95% of the prescribed dose (45–50.4 Gy). Marginal misses were LRRs at the field edge and/or not receiving ≥95% of the prescribed dose. Out-of-field LRRs were LRRs intentionally not treated with the original XRT plan. Results: Four LRRs were in-field, 8 were marginal and 10 were out-of-field, with two patients having both in-field and out-of-field LRRs and one having a marginal and out-of-field LRRs. Three in-field, 3 marginal and 6 out-of-field LRRs occurred simultaneously with distant metastasis (DM). In-field LRRs received a median dose of 48.4Gy (range 45-50Gy). Two in-field LRRs were marginal misses of the additional 15 Gy boost dose (median dose received 53.5 Gy); both patients had high grade, ER/PR-, and HER2 + tumors. In the 7 patients with triple negative tumors, there were 4 marginal, 4 out-of-field, and no in-field LRRs. Among marginal LRRs, a median dose of 33Gy (range 2–45.6Gy) was delivered. Out-of-field LRRs occurred in 5 supraclavicular and 5 internal mammary lymph nodes. With a median follow-up period of 3.7 yrs, the 5-yr disease-free survival (DFS) in patients with an isolated LRR is 55.6%, while in those recurring with DM, the DFS is 22.5%. Conclusions: Our study is one of the first to systematically evaluate the influence of XRT technique over local control in BRCA. Due to the number of marginal and out-of-field LRRs, our data suggest that XRT field design and dose are important factors in preventing LRRs in addition to tumor biology.