2206 - Feasibility of a Multi-Isocenter Technique for Long Target Volume MR-Guided Radiotherapy

E. E. Ahunbay¹, X. Chen¹, A. Angerud², M. J. Awan¹, and E. S. Paulson¹; ¹Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI, ²Raystation, Stockholm, Sweden

Purpose/Objective(s): 1 commercially available MR Linac imposes 22 cm field length restriction along the superior/inferior (S/I) direction. Limitation denies patients with long planning target volumes (PTVs) benefits of receiving MR-guided radiation therapy (MRgRT). We investigate feasibility of multi-isocenter technique using alternative planning system to enable MRgRT in patients with S/I PTV lengths exceeding 22cm. Materials/

Methods: Research version of treatment planning system, with prototype Monte Carlo photon beam model incorporating 1.5T transverse magnetic field, was installed on local machine network of precision radiation medicine company. Dual isocenter plans created for 1 head & neck (HNC) patient and 1 prostate bed and nodes (PBN) patient. Distance between isocenters along S/I direction was 6 and 10 cm for HNC and PBN plans. For each plan, 2 separate identical beam sets simultaneously optimized to generate uniform dose distributions throughout target volume. Online workflow was simulated in phantom: 2 Rad-Rx groups containing longitudinal couch coordinates at 2 isocenters created in electronic patient information management system. Treatment session was initiated at 1st isocenter (Iso1). After simultaneous optimization, Iso1 plan was downloaded to sequencer and delivered to patient-specific quality assurance. 2nd treatment session was then initiated at the second isocenter (Iso2), couch was moved to Iso2, and Iso2 plan was immediately downloaded to sequencer and delivered to patient-specific quality assurance. Patient-specific quality assurance measurements performed by shifting couch by 1mm increments to increase effective “detector resolution” in S/I direction. Couch accuracy and reproducibility evaluated by imaging MV alignment phantom in 0.1mm increments and registering with reference image with ball bearings. Treatment plans tested for robustness by recalculating with lower isocenter shifted relative to upper, by +/- 0.2, 0.5 and 1mm; and DVHs of important structures were compared over whole points.
Results: Over 1-year period, 25 HNC patients and 4 rectal cancer and nodes patients in clinic were denied MRgRT due to PTV lengths exceeding 22cm. Treatment planning system dual isocenter plans calculated, uploaded to electronic patient information management system, and delivered successfully. Repeated image registration with MV alignment phantom revealed couch accuracy and reproducibility within 0.2mm. Patient-specific quality assurance comparison plans passed at >99% using TG218 criteria. Plans with shifted lower isocenter by +/-0.2mm, 0.5mm and 1mm had 99% of all DVH points of important OARs within 1%, 3% and 5% in Dose and Volume axis of each other, except small OARs such as optic nerves and inner ears. Target volume DVHs were all within 1% both dose and volume wise of unshifted plans.
Conclusion: Efficient multi-isocenter treatments on precision radiation medicine company feasible, enabling benefits of MRgRT for patients with long target volumes.