ESTRO 2024 - Abstract Book

S389

Beachytherapy - Physics

ESTRO 2024

The INTRAGO clinical trial assesses survival in glioblastoma patients treated with intraoperative radiotherapy (IORT). Patients are randomized into the control arm, receiving external beam radiotherapy (EBRT), and patients who receive EBRT in addition to IORT. The IORT of INTRAGO uses the INTRABEAM system with spherical applicators. Treatment planning for INTRABEAM relies on vendor-provided in-water depth dose curves obtained according to the TARGIT dosimetry protocol. However, recent studies have shown discrepancies between the estimated TARGIT dose from the delivered dose. 1 In addition, this approach does not consider polar anisotropy of the source and fails to provide 3D dose distributions required for subsequent EBRT treatment plans. As an alternative, the INTRABEAM was dosimetrically characterized according to the AAPM TG-43 formalism. 2,3 The aim of this work was to investigate the accuracy of the TG-43 dosimetry method when used in treatment planning of INTRAGO patients, and the effect on the dose to OARs. A new treatment planning module for IORT volumetric dose calculations was developed in an in-house TPS called Brems . The system allows imposing calculated dose distributions on patients’ CT and the generation of dose volume histograms on contoured OARs. In this study, the IORT dose was retrospectively calculated in 18 INTRAGO patients. The control arm (9 patients) was included in this study by generating hypothetical IORT dose distributions with the average applicator diameter and dose prescription of the IORT arm. The IORT dose was obtained using (1) the TARGIT method; (2) the TG-43 characterization; (3) MC dose to water (MC w ); and (4) MC dose to tissue (MC het ). The combined OAR doses from IORT and EBRT were determined by converting IORT dose to EQD2. The TG-43 dose distribution was calculated in Brems using published TG-43 parameters according to the recommendations of the AAPM TG-43 protocol, and its evaluation with MC followed recommendations of the TG 186 protocol. The MC simulations were performed using egs_brachy as the simulation engine and calculated from phase space files previously scored at the surface of each applicator and validated with measurements in water. To assess the impact of tissue heterogeneity, published material density and composition for some tissues, namely brain, cranium, pituitary, and eye lens, were assigned to the contoured structures in the MC het simulations. The plans were evaluated based on the maximal point dose (dose to a volume > 0.03 cm3 (D 0.03cc )) to planning risk volumes (organ at risk (OAR) + 3 mm margin) and pre-established dose constraints. Table 1 presents a comparison of the OAR doses obtained with the IORT calculated with the TG-43 and the MC formalisms. According to this evaluation, the OAR dose obtained with the TG-43 distribution was statistically similar to that obtained with MCw, validating the methodology used to retrieve TG-43 dose distributions in the TPS. A Wilcoxon signed rank test showed, however, statistical differences with the MC het distributions. This deviation may be explained, in part, as a contribution of the bone voxels enclosed by the OAR + margins to the absorbed dose in the MC het simulation. Figure 1 shows the D 0.03cc obtained for the EBRT-only and for the combined EBRT + IORT plans. The figure highlights the increased OAR dose exceeding the constraints in some cases when adding the EBRT and IORT plans. These deviations were not identified in the current practice in the INTRAGO study. The EBRT + IORT(TG-43) method resulted in 18 cases of OARs exceeding the dose constraints. This result was identical for the EBRT combined with the TARGIT and the MC w methods. The more accurate MC het calculation led to the highest dosimetric differences, with 5, 6, 3, 3, 5, and 2 cases (out of 18) exceeding the dose constraint to the brainstem, optic chiasm, optic nerves, and lenses, respectively. Moreover, the mean cumulative dose to brainstem exceeded its constraint of 66 Gy with the MC het , which was not evident with the current INTRAGO clinical practice. Material/Methods: Results: