3022 - Platform-Independent Valley Dose Control by Lattice Spacing across Distinct Linac Geometries in Spatially Fractionated Radiotherapy
Presenter(s)
K. Hirotaki1, H. Tachibana2, Y. Ebinuma1, and M. Ito1; 1National Cancer Center Hospital East, Kashiwa, Japan, 2Section of Radiation Safety and Quality Assurance, National Cancer Center Hospital East, Kashiwa, Japan
Purpose/Objective(s): Spatially fractionated radiotherapy (SFRT) produces heterogeneous dose distributions consisting of high-dose peaks and low-dose valleys. While development has focused on peak escalation, valley dose may influence normal tissue response and therapeutic ratio. However, the relative contributions of peak escalation and valley suppression to peak-to-valley dose ratio (PVDR), and whether their effects are determined by delivery geometry or lattice spacing, remain unclear. Lattice spacing is a modifiable planning parameter across delivery platforms and may serve as a practical lever for controlling low-dose exposure. We performed a controlled dosimetric comparison across two linac geometries and two lattice spacings to determine dominant PVDR drivers and evaluate whether spacing functions as a platform-independent control parameter in SFRT planning.
Materials/Methods: Eleven patients were retrospectively replanned using lattice configurations with fixed 10-mm spherical targets and center-to-center spacings of 20 mm and 30 mm. For each spacing, plans were generated on a biaxial rotational ring-gantry system and a conventional C-arm linac using consistent optimization priorities. Primary endpoints included valley mean dose, PVDR, and normal tissue exposure (BODY–GTV V3). Paired comparisons were performed within spacing and machine. Two-way repeated-measures analysis of variance evaluated Machine × Spacing interaction effects. Correlation analysis assessed relationships between valley dose and PVDR.
Results: Increasing lattice spacing from 20 mm to 30 mm reduced valley dose and normal tissue exposure while producing modest changes in peak dose parameters. Valley mean dose decreased by 42.8%, from 9.35 Gy to 5.35 Gy (p < 0.001), with a reduction in BODY–GTV V3 from 8.26% to 3.52% (–57.4%, p = 0.023). PVDR increased from 3.46 to 7.58 (+119%, p = 0.005) across both delivery systems. Machine-related differences were limited and mainly observed in selected peak metrics, with interaction effects for vertex mean dose but not valley mean dose. PVDR showed an inverse association with valley mean dose (r = –0.88, p < 0.001), whereas correlations with peak metrics were weaker. These findings were consistent across delivery systems and spacing conditions.
Conclusion: Lattice spacing—not delivery geometry—was the principal determinant of valley dose and PVDR. While peak metrics showed geometry-dependent variation, clinically relevant PVDR behavior was primarily governed by valley dose suppression. These findings identify lattice spacing as a platform-independent planning parameter capable of controlling valley dose and PVDR across linac geometries. Spacing selection may represent a generalizable lever for controlling low-dose exposure and influencing normal tissue risk in SFRT. This supports prioritizing spacing optimization over platform selection and may facilitate more consistent SFRT planning across institutions.