KEY POINTS
- The retrospective dosimetric study included 12 patients with thoracic or lumbar spine metastases. Target volumes ranged from 13.0 to 133.3 cm³, with a median of 57.3 cm³.
- Photon VMAT and intensity-modulated proton plans were generated for 20 Gy in one fraction. Proton plans used two posterior-oblique fields and were robustly evaluated with ±3-mm setup and ±3.5% range uncertainties.
- Target coverage was comparable: median CTV D95 was 20.2 Gy with photons versus 20.1 Gy with protons (p=0.505), and D90 was 20.6 versus 20.5 Gy (p=0.778). Worst-case proton CTV D95 averaged 19.28 Gy.
- Proton planning reduced spinal-cord D0.5cm³ from 11.9 to 7.1 Gy (p=0.008) and D0.35cm³ from 12.8 to 7.8 Gy (p=0.006). Worst-case proton D0.35cm³ averaged 14.18 Gy.
- Oesophageal D5cm³ decreased from 8.0 to 1.9 Gy (p<0.001), while mean lung dose fell from 0.7 to 0.2 Gy (p=0.040).
- Bowel D0.5cm³ decreased from 6.1 to 3.2 Gy (p=0.016), and bowel D5cm³ decreased from 5.3 to 2.2 Gy (p=0.005). Heart, kidney, cauda-equina, and sacral-plexus differences were not statistically significant.
- All plans met AAPM TG-101 organ-at-risk limits. The cohort was small and selected, required at least a 3-mm separation between epidural disease and the spinal cord, and provided no evidence that the dosimetric differences reduce myelopathy, gastrointestinal toxicity, or vertebral fracture.
CLINICAL TAKEAWAY
Protons may be useful when single-fraction spine SBRT is limited by the spinal cord or nearby thoracoabdominal organs while equivalent target coverage remains achievable. The absolute benefit will be anatomy-dependent, and clinical superiority cannot be inferred from this planning study.