ESTRO 2024 - Abstract Book

S6019

RTT - Treatment planning, OAR and target definitions

ESTRO 2024

1 The Royal Marsden NHS Foundation Trust, Radiotherapy, Sutton, United Kingdom. 2 The Institute of Cancer Research, Radiotherapy and Imaging, Sutton, United Kingdom. 3 The Royal Marsden Hospital and the Institute of Cancer Research, Joint department of physics, Sutton, United Kingdom

Purpose/Objective:

Online adaptive radiotherapy (oART) creates and delivers a new treatment plan for every fraction, offering opportunity to reduce uncertainty by accounting for anatomical changes. Magnetic resonance imaging (MRI) gives superior soft-tissue visualisation compared to computed tomography, boosting confidence to deliver ultra hypofractionated regimes [1] . For prostate cancer (PCa) the dosimetric benefit of moderately-hypofractionated MRI guided adaptive RT (MRIgART) is not evident for all [2] . This work investigates the dosimetric value of daily MRIgART in patients receiving ultra hypofractionated PCa RT and the impact of organ at risk (OAR) volume on this.

Material/Methods:

Ten patients who received ultra-hypofractionated radiotherapy (36.25Gy in 5-fractions) on the Unity MR-linac (MRL) (Elekta, Stokholm, Sweden) were retrospectively investigated. All were treated with oART [3] . At each fraction a T2 weighted MRI was acquired for replanning (MRI session ). Next the prostate, seminal vesicles and proximal OAR structures were contoured by a clinical oncologist competent in MRI delineation. A second T2w MRI (MRI verification ) was acquired at the latter stages of plan optimisation to confirm prostate position before treatment. Translational deviations, seen on MRI verification , were corrected before treatment if the prostate had moved ≥3mm i.e., outside the smallest PTV margin. Offline, therapeutic radiographers (RTT) with MRL contouring competency refined the rectum, bladder, urethra, penile bulb, femoral heads and external contours on MRI session . This step created anatomically accurate OAR volumes; online adequate OAR volumes were accepted. Structure volumes were copied to the corresponding MRI verification , and the prostate, seminal vesicle and OAR contours edited to accurately represent patient’s anatomy. Target structures including CTVpsv_4000 (prostate plus 1cm of proximal seminal vesicles) and PTVpsv_3625 (CTVpsv_4000 plus 5mm/3mm posteriorly) were generated. Patients’ reference plans were reoptimised on the edited MRI session anatomy in Monaco (V5.40.01, Elekta, Stokholm, Sweden), PACE_C clinical goals were applied [4] . Where necessary target coverage was compromised to meet mandatory OAR constraints. MRI session plans were then recalculated (but not adapted) on the anatomy of the corresponding MRI verification , to give a truer representation of delivered dose. To mimic a non-adapted online IGRT treatment, reference scans were rigidly registered to the prostate on the contoured MRI verification images. The reference plan was re-calculated (but not adapted) on the MRI verification anatomy. Plan groups are named:

• Session (reference plan adapted to MRI session anatomy) • Verification (session plan recalculated on MRI verification anatomy) • Not-adapted (reference plan recalculated on MRI verification anatomy)