Clinical Implementation of VMAT CSI with Robust Optimization In Raystation: From Simulation to Treatment Start
Abstract
Purpose
Craniospinal irradiation (CSI) is resource-intensive, requiring complex field junctions and long treatment times that challenge existing clinical workflow. Volumetric modulated arc therapy (VMAT) improves planning target volume (PTV) coverage and organ-at-risk (OAR) sparing, clinical implementation is limited by junction sensitivity and setup variability. This study describes the UCSF workflow for implementing VMAT CSI using RayStation robust optimization, experience treating the first 10 patients, and the workflow efficiency gains achieved.
Methods
VMAT CSI plans were generated in RayStation using overlapping coplanar arcs at multiple isocenters with ±5 mm setup uncertainty per isocenter. Daily CBCT setup corrections were retrospectively analyzed to validate this margin. Standardized protocols were applied for simulation, contouring, and planning. Plan quality was evaluated using PTV V95%, junction dose homogeneity (mean HI = 1.05), conformity index (CI = 0.96), and OAR metrics (e.g., mean lung dose, maximum spinal cord dose). Therapist workflow was assessed using a prospective time-motion study. Patient-specific QA was performed using ArcCHECK with 3%/2mm gamma analysis.
Results
Robustly optimized plans achieved mean PTV V95% of 99% ± 1% while meeting all OAR constraints. Simulated 3 mm isocenter shifts in longitudinal, lateral, and vertical directions produced <4% mean junction dose variation. CBCT analysis from 10 patients over 7 fractions showed average shifts <3 mm and maximum shifts of 5 mm in all directions, confirming the 5 mm robustness margin. Lateral shifts were larger than longitudinal and vertical shifts, suggesting increased lateral robustness may be beneficial. Standardized workflow reduced therapist time by 15–20 minutes per fraction. Patient-specific QA showed a gamma pass rate of 95% ± 2%.
Conclusion
VMAT CSI with 5 mm robust optimization in RayStation provides clinically robust junctions validated by daily image guidance while improving workflow efficiency. This approach enhances reproducibility and reduces setup time, supporting efficient CSI delivery in a clinical practice.