S2035
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
Digital Poster 5196 Column Constrained Optimization for Spatially Fractionated Radiotherapy Evan M Porter 1 , Theodore Geoghegan 2 , Krystal Straessler 1 1 Radiation Oncology, University of California San Francisco, San Francisco, USA. 2 Radiation Oncology, University of California Davis, Irvine, USA Purpose/Objective: Spatially fractionated radiotherapy (SFRT) for treatment of unresectable sarcomas has shown promise in limited clinical studies [1,2,3]. Planning practices vary across centers, but common approaches use brass applicators, static-field IMRT, conformal arc, or VMAT. We present a proof-of- concept for a novel planning approach: column- constrained VMAT (ccVMAT) optimization. Material/Methods: Patients previously treated for un-resectable soft tissue sarcoma (N=3) were planned de novo using three techniques: static field IMRT, VMAT and ccVMAT. Using the RayStation scripting API, the treatment lattice was generated as a rectangular grid of 1.5cm diameter spheres, separated by 3.0cm center-to- center. Spheres which intersected the 1.0cm margin from the planning target volume (PTV) or any organs at risk (OARs) were excluded.All plans were created for a Varian TrueBeam with 6MV flattening-filter-free beams, chosen to minimize delivery time. Each plan delivered 18Gy in one fraction, prescribed to the median (50%) of the lattice volume. Optimization constraints prioritized minimizing OAR dose over coverage, limited maximum dose to 21Gy, and minimized dose between the lattice vertices.Static field IMRT employed beams separated by 45° gantry rotation. VMAT plans used three co-planar arcs with collimators angles of 5°, 355° and 85°. For the ccVMAT plans, each superior-inferior column of the treatment lattice was treated with a dedicated arc. MLCs were confined to each column using RayStation's treat margin and were constrained to within 2.5mm of the column surface. Results: Compared to the IMRT and VMAT plans, the ccVMAT plans demonstrated steeper dose gradients and lower doses between the lattice vertices. However, ccVMAT plans required longer delivery times than VMAT (582 vs 478 seconds), due to the smaller field sizes. Planning statistics for each case are provided in Table 1.PatientPTV Volume (cc)Delivery MethodDelivery Time (sec)PTV-Grid Mean Dose (cGy)Mean Dose (1-10 mm) (cGy)Gradient Index1559IMRT373685107433.32VMAT143681108131. 67ccVMAT38343773914.602840IMRT50358392317.90V MAT32351389918.51ccVMAT47944078215.113696IMR
and underdose in the uncertainty shift plans calculated with preset monitor units (figure 2). The average difference between Eclipse and the point dose measurements was -0.8±2.4%. Film gamma analysis passed the set criteria, although the pass rate of plans with auto-feathering off was significantly worse, indicating that auto-feathering improves plan robustness.All plans were clinically reviewed, with target and OAR doses comparable to Tomotherapy. A 90-115% variation of the uncertainty plans with a set- up error of 3-5mm was considered acceptable.
Conclusion: A multi-isocentre VMAT technique planned with the auto feathering tool produced plans that are robust, easier to deliver compared to moving junction techniques, and dosimetrically comparable to Tomotherapy. This technique has been used clinically to treat both craniospinal and long-limb volumes. References: [1] [1] Matsumoto K, Monzen H, Kubo K, et al. (March 21, 2023) Volumetric Modulated Arc Therapy Planning for Craniospinal Irradiation with a New O-ring Linac. Cureus 15(3): e36493. DOI 10.7759/cureus.36493[2] Misa J, McCarthy S, St. Clair W, et al. (September 20, 2023) Development and Quality Assurance of Multileaf Collimator (MLC) Auto- Feathering Junctions for Multi-Isocenter Supine Volumetric Modulated Arc Therapy (VMAT) Craniospinal Axis Irradiation on Halcyon. Cureus 15(9): e45640. DOI 10.7759/cureus.45640 Keywords: Multi-isocenter, long treatment volumes
Made with FlippingBook - Share PDF online