Intrafraction motion stability of open vs. closed facemasks in head and neck radiotherapy: Insights from the OPEN phase III trial.

Abstract

This OPEN (Optimising Patient Experience in Head and Neck Radiotherapy) phase III trial sub-study, aimed to evaluate intrafraction motion in head and neck (H&N) cancer patients using three different facemask designs. Specifically, we compared intrafraction motion among patients immobilized with a closed facemask or one of two open-face designs, utilizing pre-/post-cone-beam computed tomography (CBCT) and surface-guided radiation therapy (SGRT). A pre-planned interim analysis on the first 56 patients enrolled in the OPEN trial was conducted as a safety checkpoint. In the OPEN trial, patients are randomised into three arms: closed facemask, 3-point open facemask, or 5-point open facemask. Intrafraction motion was assessed using both CBCT and SGRT. CBCT provided deviations in translational and rotational dimensions based on bony alignment, while SGRT offered continuous monitoring of surface motion. Intrafraction motion metrics, (i.e. mean, standard deviation, maximum absolute deviation, and the 95th percentile of surface motion) were recorded for each open mask patient using SGRT data to fully quantify motion variation during treatment. The 95th percentile of SGRT deviations was used for direct comparison with CBCT motion data. Bayesian analysis was conducted to determine the equivalence of motion across mask types and measurement techniques. Margins to account for intrafraction motion were calculated across mask types using Van Herk's formulism. Mean CBCT deviations were less than 0.4 mm and 0.2 degrees, while SGRT recorded 95th percentile deviations of 0.4 mm and 0.8 degrees over all patients. SGRT detected transient maximum deviations not captured by CBCT, particularly in the yaw axis. However, these differences were transient. Bayesian analysis showed no clinically significant differences in intrafraction motion between mask types or measurement methods. No correlation was found between SGRT and CBCT measured motion within the small range of motion recorded. No difference in intrafraction margin requirements were found between arms. Based on CBCT-measured intrafraction motion, margins of 1.8 mm, 1.7 mm, and 1.3 mm were calculated for the vertical (VRT), lateral (LAT), and longitudinal (LNG) directions, respectively, to account for intrafraction motion for all mask types, with SGRT confirming that patient motion during treatment remained within these margins. Intrafraction motion, as measured by both CBCT and SGRT, remains within clinically acceptable limits and yields similar PTV margins across both open and closed mask types. Intrafraction PTV margins were found to be comparable across all mask types. The use of SGRT allowed for the detection of transient deviations and rotational differences that were not detected using CBCT alone. Overall, these findings confirm that both 3-point and 5-point open-face masks achieve intrafraction stability comparable to traditional closed facemasks in head and neck radiotherapy.