FLASH radiotherapy maintains tumor control and enables safe re-irradiation while preserving normal tissue in breast cancer PDX models

Background: Radiotherapy is central to breast cancer treatment but is limited by acute and cumulative skin toxicity, especially in large-field treatments and re-irradiation. Ultra-high-dose-rate FLASH radiotherapy (FLASH-RT, ≥ 40 Gy/s) may widen the therapeutic window by reducing normal-tissue injury without compromising tumor control. However, breast-directed and patient-derived models remain underexplored, including the effects of FLASH under repeated-irradiation conditions. Methods: We integrated an orthotopic triple-negative breast cancer (TNBC) patient-derived xenograft (PDX) model and hemithoracic normal-tissue models to compare FLASH-RT (180 Gy/s) with CONV-RT (0.03 Gy/s). TNBC-bearing NRG mice received single-fraction of 30Gy electron irradiation via a custom stereotactic jig enabling mammary-targeted or hemithoracic fields. A separate cohort of non-tumor-bearing NRG mice underwent left-chest re-irradiation to assess cumulative tolerance. Endpoints included tumor regression, recurrence, survival, and graded skin toxicity; ongoing analyses incorporate histopathology and single-cell/spatial transcriptomics to elucidate FLASH-mediated tissue responses and mechanisms of normal-tissue sparing. Results: FLASH-RT achieved equivalent tumor control to CONV-RT in TNBC-PDX models, with both modalities inducing complete regression by day 16 and maintaining clearance for two weeks before recurrence at day 32 post-RT. In contrast, normal-tissue responses diverged markedly: FLASH-RT significantly reduced acute skin toxicity (median score 0 vs. 5; p<0.0001), eliminated ulceration, and extended survival (120 vs. 90 days post-implantation) in tumor bearing mice. In non-tumor-bearing NRG mice, FLASH-RT also improved tolerance to cumulative thoracic irradiation; mice receiving a second 25Gy left-chest FLASH irradiation dose showed no clinical decline, whereas CONV-RT animals developed progressive toxicity requiring euthanasia within three months of re-irradiation. Multi-omics analyses are underway to define mechanisms of early tissue sparing and improved re-irradiation response. Conclusions: FLASH-RT maintains tumor-control efficacy equivalent to CONV-RT while significantly reducing skin toxicity in TNBC-PDX models and improving normal-tissue tolerance to re-irradiation in non-tumor-bearing NRG mice. These findings support FLASH-RT as a clinically promising strategy that may expand safe re-treatment options and broaden curative radiotherapy opportunities in breast cancer. Mechanistic studies are ongoing to elucidate the biological basis of early tissue sparing and guide translation into breast-conserving and post-mastectomy treatment settings.