ESTRO 2021 Abstract Book

S13

ESTRO 2021

This combined treatment was previously demonstrated to be safe and feasible for conventional EBRT by the multicentre randomised controlled phase III FLAME trial. The current phase II hypo-FLAME trial showed that a focal SBRT boost to the macroscopic tumour(s) is associated with acceptable acute GU and GI toxicity in addition to whole gland prostate SBRT. Considering the demonstrated potential benefit in tumour control by focal boosting and extreme hypofractionation this schedule could be the new standard of care in all risk categories. References Brand DH, Tree AC, Ostler P et al PACE Trial Investigators. Intensity-modulated fractionated radiotherapy versus stereotactic body radiotherapy for prostate cancer (PACE-B): acute toxicity findings from an international, randomised, open-label, phase 3, non-inferiority trial. Lancet Oncol. 2019 Nov;20(11):1531- 1543. Fransson P, Nilsson P, Gunnlaugsson A et al. Ultra-hypofractionated versus conventionally fractionated radiotherapy for prostate cancer (HYPO-RT-PC): patient-reported quality-of-life outcomes of a randomised, controlled, non-inferiority, phase 3 trial. Lancet Oncol. 2021 Feb;22(2):235-245. Draulans C, van der Heide UA, Haustermans K, et al. Primary endpoint analysis of the multicentre phase II hypo-FLAME trial for intermediate and high risk prostate cancer. Radiother Oncol. 2020 Jun;147:92-98.

SP-0039 Brachy P. Hoskin United Kingdom

Abstract not available

Poster highlights: Poster Highlights 1: Treatment planning 1

PH-0041 Introducing a probabilistic target presence in expected-value optimization G. Buti 1 , K. Souris 1 , A.M. Barragan-Montero 1 , J.A. Lee 1 , E. Sterpin 1 1 UCLouvain, Institute of Experimental and Clinical Research, Brussels, Belgium

Purpose or Objective In recent literature, a probabilistic target has been proposed as a promising alternative to the CTV. However, the question of how a probabilistic target should be incorporated in the treatment planning process still remains open. This study proposes a method that combines the probabilistic target with other geometric uncertainties in a probabilistic optimization framework. Materials and Methods A probabilistic target is defined in order to model the decreasing probability of tumor presence with distance from the GTV edge. Tumor infiltration errors ( i ) are assumed to be distributed according to a histological model rho(i) from literature (Fig 1, left). A generic tumor probability curve (probability of tumor presence P(d) vs. distance d ) is derived as the probability that a tumor infiltration will take a value greater than or equal to d : P( d ) = rho ( i >= d ) (Fig 1, right). A probabilistic target is then obtained by (1) computing a 3D Euclidean distance map where each voxel represents the distance from the GTV, and (2) using Fig 1 (right) to convert the distance map into a 3D probability map, followed by a correcting for anatomical barriers. The uncertainties are considered in the optimization process by computing the expected value of the objective function. Specifically, for the tumor infiltration errors, the value of each target voxel is weighted with the probabilities defined by the probabilistic target. The optimization strategy is implemented in an open-source treatment planning system MIROpt. Three lung tumor patients with motion are used as test cases, treated with IMPT. Uncertainties include setup errors, range errors, breathing motion and tumor infiltration errors. The results are compared to CTV-based worst- case robust optimization. Treatment plan robustness is evaluated with the Monte Carlo dose engine MCsquare: first, 250 error scenarios are randomly sampled and the dose is recomputed to reach a statistical noise below 2%. To evaluate tumor infiltration errors, a target realization is obtained by isotropically dilating the GTV with the sampled infiltration error and correcting for anatomical barriers. Second, the 10% worst scenarios (based on target D95) are discarded in order to generate a 90% confidence interval in dosimetric space.

Results Table 1 compares dose-volume histogram (DVH) metrics of the target and OARs for the obtained treatment