Technical evaluation of non-coplanar lattice radiotherapy: achieving directional VPDR uniformity with a 5-mm leaf width multi-leaf collimator.

Purpose: Spatially fractionated radiation therapy (SFRT) creates alternating high-dose peaks and low-dose valleys for treating bulky tumors, with optimal biological effects associated with sufficient spatial dose modulation across the tumor volume. Coplanar lattice radiotherapy (LRT) demonstrated significant directional imbalance: superior-inferior VPDR was 2.7-fold lower (14.5%) than anterior-posterior and lateral directions (approximately 40%), with transverse directions exceeding the dosimetric range commonly reported in prior LRT studies. This study systematically evaluated non-coplanar LRT to quantify directional VPDR uniformity improvements using the Millennium 120 multi-leaf collimator. Methods: A 3×3×3 lattice structure with 27 vertices was implemented in a cylindrical phantom across 20 geometric configurations: vertex diameters 0.5-2.0 cm and separations 1.0-5.0 cm. Four non-coplanar arcs utilized couch angles of 0°, 315°, 45°, and 90° with consistent 45°collimator angle, optimized using Eclipse HyperArc. VPDR was analyzed in three orthogonal directions, normal tissue dose characteristics were assessed, and treatment delivery efficiency was evaluated through monitor unit distribution patterns and modulation complexity scores. Results: Direct comparison with coplanar delivery demonstrated substantial improvements in directional uniformity. For representative 1.5 cm diameter with 2.0 cm separation, non-coplanar delivery maintained superior-inferior VPDR at 26.4 ± 5.3% while reducing excessive transverse direction values from approximately 40% (coplanar) to 27.6 ± 1.5% and 25.8 ± 3.5% (non-coplanar), bringing all directions within the dosimetric range commonly reported in prior LRT studies. Excessively high transverse VPDR values (38-44% in coplanar) were reduced, achieving directionally balanced VPDR values across all vertex diameters. Normal tissue intersection volumes with high-dose regions achieved zero at separations ≥2.0 cm for 1.0 cm diameter and ≥3.0 cm for 0.5 cm diameter. Monitor unit distribution remained balanced across couch angles. Conclusions: Non-coplanar delivery successfully brought all three directional VPDR values within the dosimetric range commonly reported in prior LRT studies at separations ≥2.0 cm across vertex diameters, addressing the excessive transverse direction VPDR (>40%) observed in coplanar delivery. This advancement enables directionally balanced SFRT dose distribution in all directions, critical for uniform spatial fractionation effectiveness. This systematic evaluation provides quantitative evidence and parameter selection guidance for clinical implementation using conventional medical linear accelerators.