ESTRO 2024 - Abstract Book

S3460

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

ESTRO 2024

177

Digital Poster

Dose homogeneity of a novel autoplanning technique for total body irradiation at extended distance

Ruud van Leeuwen 1 , Manwel Barsegyan 2 , Drean Verwegen 1 , Erik van der Bijl 1 , René Monshouwer 1

1 Radboudumc, Radiotherapy, Nijmegen, Netherlands. 2 Saxion University of Applied Sciences, Department of applied physics, Enschede, Netherlands

Purpose/Objective:

Total Body Irradiation (TBI) is a critical component of conditioning regimens for hematopoietic stem cell transplantation. We present a TBI autoplanning forward IMRT technique for non-myeloablative TBI, designed to improve dose homogeneity. Next, we compare homogeneity of the autoplanning dose distribution with that of conventional treatment planning techniques used in clinical practice, using only open beams. Since we expected robustness to be compromised when using IMRT segments, we tested the effect of small shifts on the dose distribution.

Material/Methods:

Our planning study utilized CT scans of 31 patients who underwent non-myeloablative TBI (1x200cGy) between January 2021 and September 2023 at our institution. Planning computed tomography (CT) scans were acquired in supine position on a vacuum matress, with knees bent in order to fit our maximum (160 cm) treatment length. Treatment plans were created using 10 MV photon beams from left and right, at extended distance (source midline 350cm). Calculation included a 1 cm PMMA spoiler screen to increase surface dose. The autoplanning technique employed in this study was based on an in-house developed software tool. The key steps of the autoplanning process included: 1. Calculation of an initial dose distribution in treatment planning system Pinnacle 16.0.1 using lateral beams, and exporting the dose to the software tool via DICOM. 2. Calculation in the software tool of a 2D sagittal dose map using the 1D average along the lateral axis and subsequent discretization of the dose map in four dose levels (Fig. 1). Representative dose levels were determined using the histogram of the 2D dose map. 3. Creation of MLC beam segments with leaf edges aligned with the steps in the dose map. For each direction, three MLC segments at zero degree collimator angle were created, along with an open beam at 45 degrees. Monitor units (MUs) were assigned to all eight beams to compensate for attenuation in the corresponding dose level, thereby optimizing homogeneity. 4. Beam data (angles, shapes, MUs) were sent to Pinnacle digitally for an ultimate dose calculation. We conducted a planning study in which autoplanning dose distributions (Auto) were compared to dose distributions obtained using lateral (LRRL) and anterior-posterior (APPA) open beams. Auto, LRRL and APPA plans were created on the same CT scan for all 31 patient data sets. For the APPA plans, gantry angles were set around 0 and 180 degrees, representing irradiation of the patient in lateral recumbent position. Two normalization methods were investigated: to the whole-body volume mean dose, and to a prescription point as indicated on CT, to represent workflows not using the full dose distribution for MU calculation. To assess robustness of our technique, additional calculations were performed with the patient shifted 1cm in the caudal, cranial, dorsal and ventral direction. Dose distributions were compared using the whole-body and brain mean dose, D1% and D99%