Author: Laura Buchanan, Samantha G. Hedrick, Stephen L. Mahan, Isabella Pfeiffer, Chester R. Ramsey 👨🔬
Affiliation: Thompson Proton Center 🌍
Purpose: The CQ Medical kVue Pronumbra Head & Neck (HN) System is a new immobilization device for proton therapy. The mask includes 4 large knobs that attach to the board with built-in shims. These knobs are potentially in the beam path for a HN proton plan, so this study presents the first robust evaluation of the Pronumbra board.
Methods: A head phantom was simulated on the Pronumbra board with a mask. Bilateral and ipsilateral target volumes were deformed to the phantom and normal contours drawn by auto-contouring. Five-field bilateral and 3- and 2-field ipsilateral plans were developed. The 5-field used our standard technique, avoiding posterior beams through the shoulders, using MFO planning. The shoulder-avoid region is below the knobs, so posterior beams treated through them. The 3-field plan used ipsilateral beams and MFO optimization, allowing a single beam to treat through the knobs. The 2-field plan was worst-case scenario, treating a beam through the knob with SFO optimization. All plans were robustly optimized with 0.3cm shifts and 3.5% range uncertainty.
Results: Each plan had the following perturbations in all directions: +/-0.3cm shifts, +/-3 degrees rotations, +/-3.5 range uncertainty, and a beam shifted independently of the others +/-0.3cm. For the 5-field plan, the average change in CTV D95 for 20 perturbations was -1.13%, indicating a reduction in coverage. The minimum was -1.88%, and maximum was 0.13%. For the 3-field, the average, minimum, and maximum change in CTVD95 was -0.94%, -2.31%, and -0.16%, respectively. For the 2-field, the average, minimum, and maximum change in CTVD95 was -1.01%, -2.36%, and -0.15%, respectively. These are within our clinic standard of no more than 3% change in CTV D95.
Conclusion: Treating through the knobs of the Pronumbra HN proton board is safe and robust. They minimally impact the beam, even with a 2-field worst-case scenario.