ESTRO 2026 JAMA Oncol. 1 août 2023;9(8):1056 - 64. Keywords: Radiotherapy, Head and Neck Neoplasms Mini-Oral 4980 Evaluation of automated proton planning for Head and Neck Cancer: Robustness, dosimetric performance and NTCP impact. Heidi Gulbrandsen, Camilla Grindeland Boer, Grete May Engeseth Cancer Clinic, Haukeland University Hospital, Bergen, Norway
S2803
RTT - RTT contouring, target definition, and treatment planning
the quality of irradiation can be compromised by morphological changes in patients during treatment, particularly in the absence of adaptation. Adaptive radiotherapy (ART), while resource-intensive, could offer an improved balance between efficacy and toxicity, even in resource-constrained settings.The aim of the study was to analyze the impact of anatomical variations during treatment on dose distribution to target volumes and organs of interest (OARs). Material/Methods: Retrospective study including patients with HNC treated with volumetric modulated arc therapy (VMAT). All patients underwent daily cone-beam computed tomography (CBCT) for image-guided radiation therapy (IGRT). A mid-treatment CBCT was used for re-contouring of target volumes and specific organ of interest (spinal cord, brainstem, parotid glands). Conversion of Hounsfield units to electron density allowed for recalculation of the initial plan's dosimetry on the anatomy of the day. Results: We included 30 patients. Ten patients were replanned between the 3rd and 4th week of treatment due to significant weight loss or tumor progression. In non- replanned patients, tumor coverage was significantly reduced on mid-treatment CBCTs, with a mean decrease in D95% of -1.45% (p = 0.0399) for the high- risk planning target volume (PTV) and -2.46% (p = 0.0019) for the low-risk PTV.Conversely, an increase in doses to OARs was observed. In the spinal cord, a mean increase in maximum dose (Dmax) of +0.56 Gy was noted, with a maximum observed increase of +2.85 Gy. For the brainstem, a mean increase in Dmax of +0.98 Gy was observed, peaking at +3.3 Gy. In the parotid glands, a mean increase in mean dose of +1 Gy and in the dose to 50% of the volume (D50) of +1.28 Gy was detected.For the replanned patient group (n=10), there was no significant difference in tumor coverage (D95% for both PTVs). A notable reduction in doses to OARs was achieved: -1.81 Gy in spinal cord Dmax, -6.25 Gy in brainstem Dmax, and -7.64 Gy in parotid gland mean dose.Regarding constraint violations, dose constraints for the spinal cord and brainstem were not exceeded in any patient in either group. For the parotid glands, 85% of non-replanned patients exceeded the mean dose constraint of 26 Gy, and 27% exceeded the D50 constraint of 30 Gy. Conclusion: Although ART is time-consuming, a judicious selection of patients eligible for replanning could maximize the cost-effectiveness of ART in resource-limited countries. References: 1) Castelli J, Thariat J, Benezery K, Hasbini A, Gery B, Berger A, et al. Weekly Adaptive Radiotherapy vs Standard Intensity-Modulated Radiotherapy for Improving Salivary Function in Patients With Head and Neck Cancer: A Phase 3 Randomized Clinical Trial.
Purpose/Objective: This study investigated the development and
evaluation of an automated planning script for head and neck cancer (HNC) patients treated with intensity- modulated proton therapy (IMPT). The study assessed the feasibility, performance, and potential clinical relevance of automated proton plan-generation. Material/Methods: Fifteen andeighteen patients were included for training and validation of the planning script, respectively. The automatic planning framework employed fixed field arrangements, incorporated Dose Fall Off objectives, and integrated region-avoidance constraints to ensure clinically robust plan generation. Each patient had three plans generated for comparison: a manually IMPT plan, an automatically generated IMPT plan, and a clinically accepted photon- based photon plan (VMAT). All plans had three dose levels: 50Gy, 60Gy, and 68Gy, respectively. Robust evaluation had a dose coverage criterion of D99% >= 95% in 45 uncertainty scenarios (0.4 cm setup- and 3.5% density uncertainty) for all CTV’s. Manual and auto-generated IMPT plans were evaluated for robustness, dosimetric performance, and estimated NTCP for dysphagia and xerostomia (1). The difference in estimated NTCP ( Δ NTCP) between VMAT and IMPT, and VMAT and auto-generated IMPT plans was compared to assess the potential impact on model- based patient selection. Results: All planning approaches were robust. The mean dose (Dmean) to organs at risk (OARs) was slightly higher in the auto-generated plans, except for the oral cavity, with dose differences ranging from –0.1 Gy(RBE) to 2.3 Gy(RBE) (Figure 1). The median NTCP for dysphagia was 13.0%, 4.7%, and 5.2% for VMAT, manual IMPT, and auto-generated IMPT plans, respectively, whereas for xerostomia it was 21.9%, 19.2%, and 20.6%, respectively. (Figure 2). The median Δ NTCP for dysphagia was 6.5% for the manual and 5.5% for the auto-generated plans, while the corresponding values for Δ NTCP for xerostomia were 2.5% and 1.9%,
Made with FlippingBook - Share PDF online