ESTRO 2024 - Abstract Book

S4283

Physics - Intra-fraction motion management and real-time adaptive radiotherapy

ESTRO 2024

Purpose/Objective:

Respiratory motion poses a significant challenge when it comes to accurately targeting esophageal tumors during radiotherapy. The effects of said motion is particularly noticeable in distal and gastroesophageal junction tumors, with median motion of up to 8.2mm in the CC direction [1]. To address this issue, large internal clinical target volume (ICTV) margins are typically used to encompass the tumor position throughout the entire respiratory cycle. Moreover, due to the considerable variability in respiratory motion amplitudes between treatment sessions, substantial PTV margins are necessary to ensure tumor coverage. Unfortunately, these extensive margins increase the risk of radiation-associated toxicity in adjacent healthy tissues. Previous research has explored the use of non-invasive mechanical ventilation to induce regularized breathing (RB) at a frequency of 60 breaths per minute (brpm) with positive end-expiratory pressure (PEEP). This resulted in a 76% reduction of median respiratory motion amplitude of the diaphragm from 37mm during free breathing (FB) to 11mm during RB, as measured on cine-MRIs [2]. The current study aims to quantify the potential dosimetric advantages of employing RB at 60 brpm with PEEP compared to FB in patients with esophageal cancer. Ten patients with carcinoma of the distal esophagus (N=8) and gastroesophageal-junction (N=2) underwent 23x1.8Gy Cone-Beam CT-based image-guided radiotherapy. Before treatment, a 4D-CT planning CT (pCT) was acquired, dividing the respiratory cycle into ten phases, from the initiation of expiration to the completion of inhalation. Throughout the imaging and treatment procedures, patients breathed freely. To simulate the effects of mechanical ventilation at a frequency of 60 brpm with PEEP, characterized by shallow breathing at larger lung volumes, a new RB average CT (RB CT) was created by averaging the two phases around the end inhalation position of the FB 4D-CT scan. Using the newly generated RB-CT scan, new ICTV heart and lung structures were delineated by a trained physician. Prior (unpublished) results found a reduction in mean interfractional respiratory motion in the CC direction from 6.3mm during FB to 2.0mm during 60 brpm with PEEP RB. Based on these results, the PTV margin was reduced by 2mm in the CC directions in this study as compared to conventional clinical standards. This resulted in a final PTV margin of 8mm in the CC direction, whereas the LR and AP margin remained at 7mm. Using the new RB-CT scan and delineations, new treatment plans were generated. Subsequently, differences in target volumes and planned dose volume parameters between FB-CT and RB-CT treatment plans were assessed: PTV (D0.1cm 3 and V95%), ICTV D98%, heart (Dmean and V30Gy), and lung (Dmean, V10Gy, and V20Gy). Statistical significance was determined through the Wilcoxon signed-rank test (p≤0.05). Limiting respiratory motion in the RB-CT significantly reduced CC diaphragm motion from a mean of 13.7 (range: 6 18) mm in the FB-CT scan to 4.3 (range: 1-8) mm in the RB-CT scan. The mean reduction of the ICTV volume was 11.4% (SD: 6.3%) in the RB-CT scan compared to the FB-CT scan. The increased stability of interfraction respiratory motion variation led to an overall 10.0% (SD: 6.1%) reduction of the PTV volume in the RB-CT scan. In figures 1 and 2, a comparison between the RB and FB dose volume metrics is given. The target volume reductions did not yield any statistically significant differences in the target volume dosimetric parameters over those with the FB scan. Specifically, the median[IQR] PTV D0.1cm 3 was 44.4[0.7]Gy in the RB-CT plan, versus 44.4[0.5]Gy in the FB-CT plan (p=0.19), and the median[IQR] V95% of the PTV was 97.9[0.5]% in both the RB and FB treatment plans (p=0.50). Similar results were found for the ICTV with both the RB and FB plan median[IQR] D98% being 40.9[0.1]Gy (p=0.95). The reduction in target volumes did result in statistically significant reductions of heart and lung dose. Namely, the median Material/Methods: Results: