Retrospective Analysis of Replan Frequency and Causes in Esophageal Cancer Patients Treated with Spot Scanned Proton Therapy

Abstract

<h3>Purpose/Objective(s)</h3> For patients with esophageal cancer treated with intensity-modulated proton therapy (IMPT) at our institution, weekly verification CT scans are obtained to ensure that the plan is still acceptable. If unacceptable deviations from the intended dose are observed, replanning is performed. This study aims to investigate the frequency and causes of replans to determine the appropriate frequency of performing verification scans. <h3>Materials/Methods</h3> Data from a cohort of 205 esophageal cancer patients treated with IMPT between 2015 and 2020 were retrospectively analyzed. Patients were typically positioned head first supine with arms up in a vacuum immobilization device, and a four-dimensional CT (4DCT) simulation scan was obtained. The typical prescription was 50 Gy in 25 fractions over 5 weeks, with contours delineated as per consensus guidelines. For tumor motion ≤ 10 mm, patients were treated free-breathing with isolayer repainting, while for motion > 10 mm, gating was also added. Two posterior oblique beams were typically used, and single field optimization was performed. Robust optimization was utilized with 5-mm translational perturbations and 3% range uncertainty, with dose evaluation, also performed on the extreme respiratory phases. Weekly 4DCT verification scans in treatment position were acquired. Rigid image registration to the original planning CT image set was performed, and the care team reviewed the recalculated dose distributions to determine the need for replan. CTV (V95%) % was selected as a metric to compare the difference in target volume dose between verification scans that did [CTV_replan (V95%) %], versus did not [CTV_no-replan (V95%) %], lead to a replan. <h3>Results</h3> The primary tumor location was upper/middle esophagus (16%) or lower/gastroesophageal junction (83%). 21 patients (10%) required replanning, including 3 patients with 2 replans. Replans occurred during the 1^st (20%), 2^nd (33%), 3^rd (20%), or 4^th (25%) weeks of treatment. Primary reasons for replanning were a superior shift of the diaphragm baseline position affecting the range of the beams entering near this interface (46%), changes in tumor size or position (17%), new or resolving pleural effusion affecting the range of beams (13%), changes in stomach filling affecting the target geometry (13%), and patient setup reproducibility issues (13%). The average value of CTV_no-replan (V95%) % = 98% ± 3% vs. CTV_replan (V95%) % = 96% ± 1%. The modest size of this difference suggests that other factors, including poor fidelity of rigid CTV contours, also contribute to clinical decision-making. <h3>Conclusion</h3> In this large cohort of patients with esophageal cancer treated with robustly planned IMPT, the rate of adaptive replanning was 10% and resulted from 5 discrete changes as described above. Every other week frequency of verification scans should be considered.