KEY POINTS
- This preclinical study investigated whether alterations in lipid metabolism contribute to the normal-tissue-sparing effects of proton FLASH radiotherapy in head and neck cancer. Female C57BL/6 mice bearing orthotopic MOC2 oral carcinoma received 10 Gy in one fraction using either conventional proton irradiation or FLASH; tongue tissue and tumor samples were collected 14 days later for lipidomic, proteomic and spatial analyses.
- Irradiation used a clinical Varian ProBeam pencil beam scanning gantry, delivering conventional protons at approximately 1 Gy/s and FLASH protons at 100 Gy/s in transmission mode. The experimental groups included nonirradiated controls, with five animals per group in the tumor-bearing cohort and three biological samples per group used for the principal molecular analyses. Across all tissues, 1,547 lipid species representing 37 lipid classes were quantified after biochemical filtering.
- Normal tongue tissue demonstrated substantially greater lipid-metabolic changes than tumor tissue. Investigators identified 537 differentially regulated lipid species in tongue versus 133 in tumor, with conventional irradiation producing particularly pronounced depletion of triglycerides and other neutral lipids. The tumor lipidome showed relatively limited differences between conventional and FLASH irradiation, without a comparably organized modality-specific response.
- Triglyceride depletion was markedly greater after conventional RT than after FLASH. Among triglyceride species, 61.5% were downregulated after conventional irradiation versus 28.0% after FLASH, while corresponding proportions for oxidized triglycerides were 77.3% versus 45.4%. These percentages describe the fractions of lipid species classified as downregulated relative to controls, not a direct percentage reduction in total tissue triglyceride concentration.
- FLASH also produced a distinct pattern of membrane-associated lipid remodeling. Compared with conventional irradiation, a greater proportion of phosphatidylethanolamine species was upregulated (26.3% vs 10.3%), with a smaller difference for phosphatidylcholine species (18.2% vs 14.2%). Proteomic analysis supported differences in lipid processing, including reduced Mgll and increased Apoe and Lypla2 after FLASH, whereas conventional irradiation showed changes associated with triglyceride catabolism and oxidative stress, including Ces1d suppression and Acox1 induction.
- Spatial lipidomics localized a substantial proportion of the triglyceride signal to minor salivary gland acini within the tongue, with more pronounced loss after conventional irradiation and partial preservation after FLASH. However, only two evaluable biological sections per condition contained the relevant glandular structures, making these results illustrative rather than definitive evidence of anatomically localized radioprotection.
- The authors proposed a Lipid Metabolic Remodelling Hypothesis, in which FLASH-associated preservation and redistribution of lipid reserves may contribute to tissue recovery beyond the initial oxidative injury phase. This remains unproven: the study did not directly measure salivary function, mucositis severity, tumor control or lipid metabolic flux, and candidate molecular changes were identified using small sample numbers and largely exploratory statistical analyses.
CLINICAL TAKEAWAY
FLASH radiotherapy may influence normal-tissue recovery through sustained changes in lipid storage, transport and membrane-associated metabolism, particularly in radiosensitive glandular structures. These findings offer potential molecular targets for future radioprotection research, but functional validation, repeated-fraction experiments and larger biological cohorts are necessary before linking this signature causally to the FLASH effect.