KEY POINTS
- This prospective single-centre pilot enrolled 24 patients undergoing definitive radiotherapy: 12 with unresectable stage III non-small cell lung cancer and 12 with limited-stage small cell lung cancer. All received volumetric modulated arc therapy to 60 Gy in 30 fractions.
- Each patient had paired plans generated using internal target volume-based robust optimization with ±5 mm setup uncertainties and conventional planning target volume optimization using a 5 mm isotropic margin.
- Nominal target coverage was comparable, including median internal target volume D95 of 58.5 Gy with both approaches. Robust optimization was selected for clinical treatment in 21 of 24 patients (87.5%).
- Robust optimization reduced median lung V20 from 25.5% to 23.1% (p=0.001), mean heart dose from 11.4 to 9.9 Gy (p=0.006), mean esophageal dose from 22.3 to 19.8 Gy (p<0.0001), and spinal cord maximum dose from 35.98 to 31.81 Gy (p=0.0001).
- In perturbed scenarios, spinal cord constraints were met in 99.4% of robust optimization scenarios versus 91% with planning target volume optimization. Among patients with non-small cell lung cancer, modeled pneumonitis risk decreased from 8.2% to 7.2%, modeled grade 2 or higher esophageal toxicity from 32.3% to 25.6%, and estimated immune-cell dose from 4.7 to 4.2 Gy.
CLINICAL TAKEAWAY
Internal target volume-based robust optimization appears clinically feasible for long-course photon radiotherapy and may reduce thoracic organ exposure without materially compromising target coverage. However, this was a small, non-randomized planning study, and the toxicity findings were model-based; prospective clinical outcomes are still needed.