KEY POINTS
- The study combined 2D and 3D breast cancer experiments with an in-silico treatment-planning analysis. MDA-MB-231 triple-negative and MCF7 ER-positive cells were irradiated with photons, protons, helium-4 and carbon-12 ions; oxygen-16 was additionally tested in the triple-negative model.
- Particle beams were delivered using clinically realistic 3-cm spread-out Bragg peaks, with LET values spanning approximately 3–165 keV/μm. Across clonogenic and spheroid experiments, all tested ions showed greater biological effectiveness than photons, while helium-4 generally produced a response intermediate between protons and heavier carbon ions.
- A particularly notable signal involved tumor behavior rather than cell killing. Photon irradiation increased migration and invasion, whereas particle irradiation suppressed both. Even in an exploratory comparison of 2 Gy particles versus 4 Gy photons, migration remained significantly reduced with particles through 27 hours (p<0.001) and invasion remained lower at 24 hours (p<0.01).
- The effect persisted in a three-dimensional triple-negative breast cancer spheroid invasion assay. 4 Gy photons increased invasion by approximately 30% at 48 hours, whereas all particle beams suppressed it; helium-4 produced relative invasion of 0.60±0.10 at 48 hours and remained reduced at 0.74±0.05 after four days.
- Tumor-conditioned media after helium-4 and carbon-12 irradiation also significantly reduced endothelial network formation (p<0.001), while proton-treated media reduced network number (p<0.01). VEGF-A concentrations fell by approximately 50% after particle irradiation, whereas photon exposure increased proangiogenic signaling.
- Biological results were then incorporated into plans for 10 left-sided chest-wall breast cancer patients prescribed 40.05 Gy in 15 fractions. Protons used fixed RBE 1.1, while helium-4 and carbon-12 plans used biological optimization based on the experimentally derived triple-negative breast cancer α/β of approximately 9.4 Gy.
- Particle plans consistently predicted less cardiopulmonary toxicity than VMAT. Major coronary-event risk was 2.24% with VMAT versus <2% with all particle modalities, and predicted radiation-pneumonitis risk was approximately 10% with VMAT versus substantially lower with particle plans. Protons and helium generally provided better normal-tissue sparing than carbon ions, while carbon target coverage was limited by biological-model effects at tissue interfaces.
CLINICAL TAKEAWAY
The most interesting observation is that particle radiation may influence migration, invasion and angiogenesis differently from photons even after accounting approximately for greater cell killing. Helium-4 appears particularly interesting as a potential middle ground between proton dosimetry and carbon-ion biology, but this remains preclinical and in-silico evidence without proof of better breast cancer outcomes in patients.
SOURCE
International Journal of Radiation Oncology, Biology, Physics