3570 - Safety and Within-Session Tongue Position Stability with a Patient-Specific Negative-Pressure Intraoral Spacer Assessed on Fluoroscopy and MRI: Proof-of-Concept Studies in Healthy Volunteers and Tongue Cancer Patients

Purpose/Objective(s):

Reproducible tongue immobilization is critical for high-precision image-guided radiotherapy workflows. We evaluated the safety and within-session tongue position stability of a patient-specific, 3D-printed negative-pressure intraoral spacer, quantified by radiopaque tongue-marker displacement on fluoroscopy and MRI.

Materials/Methods:

Two prospective single-center proof-of-concept (PoC) studies placed radiopaque markers at three midline tongue sites (tip, midpoint, base).

PoC1 (healthy volunteers): Posteroanterior (PA) and lateral fluoroscopy were acquired immediately after spacer insertion and at 20 minutes. A 10-mm copper wire embedded along the device midline provided in-field pixel-to-mm calibration. Displacement was defined as the 2D change in marker position between time points within each projection. One radiologist performed triplicate readouts per site/view and averaged measurements.

PoC2 (tongue cancer): Pre- and post-contrast T1-weighted MRI were acquired ~20 minutes apart (in-plane pixel size 0.5 mm). A spacer-embedded high-signal fiducial served as the stationary reference; rigid registration used this fiducial to define a spacer-fixed frame. Displacement was defined as the 3D Euclidean distance between marker centroids across time. Two radiologists measured independently; inter-reader agreement was assessed using the intraclass correlation coefficient (ICC) on patient-level mean displacement (averaged across three sites). Safety endpoints were device- and imaging-related adverse events (AEs).

Results:

PoC1: Four participants were analyzable. Participant-level mean 2D displacement (median [IQR]) was 0.45 (0.35–0.73) mm in PA and 0.55 (0.41–0.68) mm in lateral views. Across all sites/views, 91.7% of measurements were =1 mm and 100% were =2 mm (maximum 1.44 mm PA; 1.30 mm lateral). No serious AEs occurred; two participants reported four transient Grade 1 symptoms (e.g., nausea, local discomfort) without mucosal injury.

PoC2: Five patients were analyzed. Patient-level mean 3D displacement (median [IQR]) was 0.87 (0.71–1.17) mm (reader 1) and 0.71 (0.50–1.65) mm (reader 2). Across all sites, the proportions of measurements =2 mm were 93.3% and 86.7%, respectively. Inter-reader agreement for patient-level mean displacement was excellent (ICC=0.93). No device- or imaging-related AEs were reported.

Conclusion:

A patient-specific negative-pressure intraoral spacer demonstrated favorable safety and small within-session tongue displacement on fluoroscopy and MRI. Fluoroscopy showed predominantly sub-millimeter 2D displacement per projection, and spacer-referenced MRI demonstrated predominantly =2 mm 3D displacement with excellent inter-reader agreement, supporting feasibility for high-precision tongue radiotherapy workflows. Ongoing work will assess inter-fraction reproducibility and quantify comparative dosimetric impact versus standard immobilization approaches.