ESTRO 2024 - Abstract Book

S4172

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

ESTRO 2024

Material/Methods:

Oligometastatic patients with abdominal lymphadenopathies were enrolled in the STEAL trial, and treated with CyberKnife with an integrated CT scanner on rails [1]. Patients were all treated with a total dose of 45 Gy in 5 daily consecutive fractions of 9 Gy. For each patient, a library of 3 plans was created pretreatment: plan A, based on the planning CT; plan B, with OARs contours based on their location on a diagnostic CT scan both with 45 Gy prescribed to the 90% isodose-line; and plan C, based on the planning CT with 45 Gy prescribed to the 80% isodose-line. Before each fraction a CT scan was made with the in-room CT scanner. Next, the planning CT and fraction CT were rigidly matched based on the spine. Based on this match, the dose of the three plans and the target volumes (GTV and PTV) were rigidly transferred to the fraction CT. Organs at Risk (OAR) as contoured in the planning CT were propagated to the fraction CT scan by deformable image registration (DIR). For each plan, relevant DVH parameters such as PTV coverage and dose to bowel (V35) were computed and presented in a tabular format indicating whether they passed or failed. Following a decision tree, the radiotherapy technologist (RTT) selected the best plan of the day, i.e. the plan with the highest target coverage without exceeding OAR constraints. Plan selection took place without interference of a radiation oncologist. The RTT’s underwent training to ensure safe execution of the workflow. We included 52 patients and performed 55 treatments. The primary tumor was prostate adenocarcinoma in 19 patients (36 %), colorectal in 17 (33%) and 16 patients (31%) had a primary tumor with a different origin. For only 12 patients (22%) plan A was delivered in all fractions and thus received the standard treatment. An adaptive treatment plan was chosen in 58% of the fractions. Plan C was most frequently selected from the library in 45% of the fractions, followed by plan A in 43%. Plan B was chosen in 12% of the fractions. Anatomical variations observed in the fraction CT scan led to replanning for 4 patients, with in 3 cases due to a constraint violation and in 1 patient due to movement of the GTV outside the PTV. A local control of 96% was achieved at 1 year, as shown by the Kaplan Meijer curve displayed in Figure 1. The median overall survival was 49 months (95% CI 39-58). One patient developed a late grade 3 ureter stenosis 2 years after treatment of a left iliacal lymph node. Grade ≤ 2 acute toxicity was reported in 34 patients (62%). Grade ≤2 late toxicity was reported in 4 patients (8%). Results:

Conclusion: