APBI (Accelerated Partial-Breast Irradiation) Versus WBI (Whole-Breast Irradiation) Delivered with the Same Dose Fractionation: A Propensity Score Matched Analysis of Two Prospective Studied Cohorts

Abstract

The recent updates of the RAPID and RTOG 0413 trials have led to renewed interest in the question of APBI versus WBI outcomes. On the other hand, IMPORT-LOW, which was the only APBI randomized study strictly testing treatment volume without the confounding effect of dose-time factors, reported non-inferiority of APBI with fewer late toxicities. In this study, we report our institutional experience comparing two matched prospective cohorts of breast cancer patients treated with the same dose-fractionation using APBI or WBI. Two phase II trials using the same dose fractionation of 4995 cGy in 15 fractions enrolled early stage breast cancer patients at our institution between 2008 and 2017; one utilizing APBI and the other utilizing WBI. A total of 184 patients were enrolled in both trials (34 on the APBI and 150 on the WBI). Patients who received regional nodal irradiation or who were treated prone on the WBI trial were excluded. We then performed a propensity score 1:1 match between the two cohorts using age, histology, breast volume, and seroma volume. Outcomes of interest included loco-regional control, acute and late toxicity, and cosmesis. Cosmesis was scored using the Harvard Cosmesis Scale; toxicities were recorded using CTCAE version 4.03. A total of 68 patients were matched. All patients had T1-T2, N0 disease with negative margins. Median follow up was 62.5 months [6-96] for the WBI patients versus 92.5 months [23-119] for the APBI patients. Median time to final cosmetic evaluation was 24.5 months [6-67] in the WBI cohort versus 90 months [9-119] in the APBI cohort. In accordance with the propensity score match, breast size mean (1400 cc vs 1600 cc; p =0.15), seroma size mean (37 cc vs 49 cc, p=0.27) and mean age (58 vs 60, p=0.17) were similar in the two groups. Ipsilateral breast recurrence rates were similar between the two groups (2/34, 5.9%) in the APBI versus (3/34, 8.8%) in the WBI, p= 1.0. In the WBI group, 1 patient had distant metastasis 6 years after treatment and 1 patient had a contralateral breast recurrence (CBR). In the APBI group, 4 patients had CBR. At final cosmetic evaluation, 91.2% in each cohort had excellent/good cosmesis, 5.9% in each cohort had fair cosmesis and 2.9% (1 patient in each cohort) had poor cosmesis, p = 0.92. In the WBI cohort, 11/34 patients had grade 2 or 3 acute toxicities (32.4%) versus 2/34 in the APBI group (5.9%), p= 0.01. In the WBI cohort, 12/34 patients had grade 2 or 3 late toxicities (35.3%) versus 7 in the APBI group (20.6%), p= 0.28. Clinical outcomes, toxicity profiles and cosmetic results were similar in patients treated with APBI or WBI, with significantly fewer grade 2–3 acute side effects after APBI. Our results are in agreement with the IMPORT-LOW outcomes and suggest that with similar fractionation schemes APBI and whole breast are associated with similar outcomes.