237 - Five vs. Three Fractions of Prone Accelerated Partial Breast Irradiation: Results of a Randomized Controlled Phase 3 Non-Inferiority Trial

N. K. Gerber¹, J. Goldberg², O. G. Maisonet¹, J. Xiao¹, T. Hochman³, D. R. Paez Davila⁴, D. Axelrod⁵, F. R. Schnabel⁶, A. Guth⁵, J. Purswani⁷, C. Hardy Abeloos⁴, O. Cahlon¹, C. A. Perez⁸, and S. C. Formenti⁹; ¹New York University Grossman School of Medicine, Department of Radiation Oncology, New York, NY, ²NYU School of Medicine, New York, NY, ³NYU Langone Medical Center, New York, NY, ⁴NYU Langone Health, New York, NY, ⁵Department of Surgery, NYU Langone Health and Perlmutter Cancer Center, New York, NY, ⁶New York University School of Medicine, New York, NY, ⁷Albert Einstein College of Medicine, Montefiore Medical Center, Bronx, NY, ⁸St. Jude Childrens Research Hospital, Memphis, TN, ⁹Weill Cornell Medicine, New York, NY

Purpose/Objective(s): Accelerated partial breast irradiation (APBI) has been shown to be equivalent to whole breast irradiation in multiple randomized clinical trials though fractionation, technique, and patient position varies widely between trials. Building on our institutional experience of prone APBI delivered in 5 fractions, we designed a randomized trial hypothesizing that 3-fraction APBI would be non-inferior to 5-fraction APBI with respect to fibrosis. Materials/

Methods: Post-menopausal women who had undergone breast-conserving surgery for invasive disease = 2 cm (any grade) or for screen-detected ductal carcinoma in-situ = 2.5 cm (grade 1-2), negative margins, and clinically or pathologically node negative were enrolled on this randomized, controlled, non-inferiority trial. Patients were randomly assigned (1:1) to receive APBI delivered prone 30Gy in 5 fractions (5F) daily (control) or 24Gy in 3 fractions (3F) every other day. Standard tangential beams were used to deliver radiation using field-in-field or intensity modulated radiation therapy. The primary endpoint was that the rate of post-treatment radiation fibrosis (grades 2+) on the 3F arm is not more than 10% worse than the expected rate on the 5F arm by 2 years with one sided alpha of 0.025 and power of 80%. Secondary endpoints included local recurrence (LR), regional recurrence (RR) distant metastasis (DM), contralateral breast cancer, and physician and patient rated cosmesis. This trial is registered with ClinicalTrials.gov, NCT02276885.
Results: Between October 2014 and May 2021, we obtained consent from 320 patients of whom 308 were randomized (12 screen-failures), 155 to 5F arm and 153 to 3F arm. After excluding 2 additional screen failures and 22 patients who withdrew consent, 284 patients were treated, 144 patients in the 5F arm and 140 in the 3F arm. Within 2 years, grade 2+ fibrosis occurred in 18.3% (25/137) of patients in the 5F arm and 12.3% (17/138) in the 3F arm meeting criteria for non-inferiority (lower edge of 95% CI -0.025). At a median follow up of 54 months (range: 1-108) the rate of RFS including LR, RR or DM was 95% (3 deaths, 1 LR and 2 DM) (95% CI 88%-98%) in the 5F arm and 93% (5 LR, 1 RR, 4DM, and 5 deaths including 2 of the DM) (95% CI: 86%-96%) in the 3F arm. Within 2 years, 93.4% (128/137) patients were recorded as having good/excellent physician-rated cosmesis in the 5F arm and 97.8% (135/138) in the 3F arm. Within 2 years, 83.6% (112/134) of patients rated their own cosmesis as good/excellent in the 5F arm and 78.1% (107/137) in the 3F arm.
Conclusion: APBI delivered prone with 24Gy in 3 fractions is non-inferior to 30Gy in 5 fractions with respect to fibrosis. Both groups demonstrate low rates of LR. This trial supports the use of APBI in this low-risk patient population and provides novelty with regard to an alternative fractionation scheme and the use of mini-tangents in the prone position.