2503 - Ecological Markers of Radioresistance in Breast Cancer: A Spatial Multi-Omics Approach

Purpose/Objective(s): Adjuvant radiation therapy (RT) reduces locoregional recurrence in early-stage breast cancer, yet no established method exists to tailor RT to individual radiosensitivity. We hypothesized that pre-existing resistance to oxidative damage promoted by hypoxic tumor microenvironment (TME) habitats defines a cell's evolutionary trajectory and radiosensitivity. Using a spatial multi-omics approach, we aimed to identify ecological markers of radioresistance to stratify at-risk patients and inform individualized radiation dosing.

Materials/Methods: 3D spheroid cultures of MCF7, T47D, MDA-MB231, and MCF10A cell lines were grown as small (<240 µm) or large (>240 µm) to model normoxic and hypoxic habitats, respectively. Spheroids were irradiated (2 to 50 Gy), and growth curves and viability (3D CellTiterGlo) were assessed. Spheroids were collected at 6 hours, 24 hours, and 1 week post-irradiation, sectioned, and mounted for MALDI-based spatial metabolomics/lipidomics and multiplexed immunofluorescence (mIF) for spatial proteomics (markers: viability, proliferation, hypoxia, acidosis, EMT). Sequential MALDI and mIF sections were spatially registered for multi-omics spatiotemporal analysis.

Results: Large spheroids demonstrated greater radioresistance than small spheroids across all cell lines, most strikingly in MCF7. MALDI clustering revealed a Darwinian selection pattern: dominant post-irradiation niches in large spheroids were enriched with monounsaturated fatty acid-containing phospholipids (MUFA-PLs; e.g., PE(18:1/16:0)), while small spheroid niches were enriched with PUFA-PLs (e.g., PE(38:4)). Cross-modality correlation of MALDI and mIF data revealed that MUFA-PL-rich metabolic niches co-localize with distinct acidic niches within the TME. mIF showed large spheroids became highly glycolytic with mesenchymal-like phenotypes over time. Ionizing radiation (IR) promoted proliferation within hypoxic habitats at 1 week post-irradiation, evidenced by Ki67-positive cells emerging in regions that are quiescent in non-irradiated controls, in addition to the expected proliferative activity within normoxic habitats. Priming large spheroids with ferroptosis inducers prior to IR re-sensitized them to radiation.

Conclusion: Hypoxic habitats drive radioresistance through eco-evolutionary selection of MUFA-PL-enriched metabolic niches, which co-occur with distinct acidic niches, suggesting a convergent microenvironmental program that promotes resistance to ferroptosis, a reactive oxygen species-mediated cell death pathway involving PE-PUFA peroxidation. Re-sensitizing cells to ferroptosis collaterally restores radiosensitivity. These findings support hypoxia- and acidosis-associated metabolic ecology as novel, actionable biomarkers for patient stratification and AI-guided individualized radiation dosing.