ESTRO 2024 - Abstract Book

S3629

Physics - Dose prediction, optimisation and applications of photon and electron planning

ESTRO 2024

2079

Mini-Oral

Four-dimensional dose calculations and treatment planning strategies for dynamic tumour tracking

Emilie E Carpentier 1,2 , Ronan McDermott 3 , Marie-Laure Camborde 1 , Tania Karan 1 , Alanah M Bergman 1 , Tony Mestrovic 1 1 BC Cancer - Vancouver, Medical Physics, Vancouver, Canada. 2 University of British Columbia, Physics and Astronomy, Vancouver, Canada. 3 BC Cancer - Vancouver, Radiation Oncology, Vancouver, Canada

Purpose/Objective:

Dynamic tumour tracking (DTT) is a respiratory motion management technique where the radiation beam follows and treats a moving tumour in real time 1 . However, no commercially available treatment planning systems (TPSs) model panning/tilting beam geometry. Additionally, DTT plans are created and evaluated on a single breathing phase computed tomography (CT) image despite treatment occurring over the entire breathing cycle. These factors leave organs at risk (OARs) vulnerable to exceeding their dose constraints during DTT treatments 2,3 . This work investigates a novel procedure for calculating four-dimensional (4D) dose distributions for a gimballed linear accelerator (linac), the Vero4DRT (Brainlab AG), that models its panning/tilting motion 4,5 . Two 4D treatment planning strategies that incorporate 4D motion into plan optimization are also presented; the “Boolean OAR” method and the “Aperture Sorting” method. These methods provide more representative dose calculations for DTT plans and improve OAR sparing. Step-and-shoot intensity modulated radiotherapy (IMRT) plans were optimized in the RayStation TPS (RaySearch Laboratories) on the exhale-phase CT (50% phase from the 4DCT) for ten previously treated liver cancer patients, ensuring all OARs met institutional dose constraints. The inhale- and exhale-phase CTs (0% and 50% phases from the 4DCT) were rigidly registered by aligning implanted fiducial markers (fids) near the target, since the Vero4DRT tracks fids during DTT. An in-house Python script executed in the TPS transferred the original exhale-phase CT plan to the inhale-phase CT while altering the beams’ angles to model the gimballed linac’s panning/tilting motion (figure 1, step 1). The dose was re-calculated on the inhale-phase CT, then deformed back to the exhale-phase CT for dose accumulation (figure 1, step 2). These panning/tilting 4D dose calculations showed that each CT dataset had an OAR exceeding its dose limit. Two OAR-sparing 4D planning strategies were then applied. The “Boolean OAR” method created a union of an OAR’s exhale- and inhale-phase contours, and this boolean-OAR was protected during plan optimization (figure 1, step 3a). The “Aperture Sorting” method assigned certain IMRT-apertures to the breathing phase where they contributed the least to an OAR’s maximum dose (figure 1, step 3b). Both planning strategies were applied to each patient’s dataset, and a 4D dose calculation was performed as described above. Material/Methods: