Evaluation Of Positioning Accuracy For Stereotactic Reirradiation Of Recurrent Head And Neck Cancer Background: Reirradiation For Unresectable Head And Neck Locoregional Recurrences (Lrr) Is Clinically And Technically Challenging And Portends

Reirradiation for unresectable head and neck locoregional recurrences (LRR) is clinically and technically challenging and portends a risk of severe treatment toxicity. We instituted the use of stereotactic ablative radiation therapy to the head and neck (HN-SABR) to deliver highly conformal ablative radiation doses for treatment of LRR. We evaluated reproducibility and performed an uncertainty analysis of our initial clinical setup to guide PTV margin determination. Seven patients were simulated supine using a moldable head and neck cushion, a shoulder-length thermoplastic mask with bite block, with 6 stereotactic x-ray infrared (IR) balls placed on the mask. Two 1-mm slice-thickness CT scans were obtained and fused with diagnostic MRI or PET images. Patients were treated with a medical linear accelerator system with high-definition MLC using 2 arc VMAT plans. Patients were initially positioned using stereotactic x-ray IR tracking. Translational and rotational set-up errors were determined by comparing stereotactic x-ray positioning to verification x-ray; subsequently, soft tissue alignment was confirmed using cone-beam CT (CBCT) and auto-registration software. Differences in setup by stereotactic x-ray versus CBCT were determined. Using pre-beam x-rays, the intrafractional positioning errors were also investigated. The systematic and random errors were analyzed. Initial translational setup errors were: vertical (AP), -1.2±1.6mm; longitudinal (SI) -0.1±1.4mm; lateral (LR), 0.2±1.8mm (n=52 alignments). Initial rotational setup errors were: yaw 0.1±0.6°; roll, -0.2±0.6°; pitch, 0.1±0.9°. Individual patient intrafraction uncertainty by stereotactic x-ray alignment was: AP, 0.0±0.7mm; SI, -0.3±0.7mm; LR, -0.1±0.5mm; yaw 0.1±0.5°; roll, -0.1±0.4°; pitch, -0.1±0.4°. Mean agreement between CBCT and stereotactic x-ray was: AP, 0.4±0.5mm; SI, -0.2±0.7mm; LR, 0.2±0.4mm; yaw 0.0±0.7°; roll, -0.1±0.7°; pitch, 0.2±0.1°. The calculated CTV to PTV margins in AP, SI, and LR directions were 1.0mm, 1.3mm, and 1.4mm, respectively, when image guidance was used for online setup correction. The minimum PTV margin to achieve a 95% confidence level to cover 90% of the prescribed dose was 1.0mm AP, 1.5mm SI, and 1.4mm LR. Stereotactic x-ray online image guidance can correct initial translational setup and rotational errors to within 1mm and 1°, respectively. These early results indicate that our patient positioning system provides a precise and robust platform for incorporating HN-SABR in treating patients with recurrent head and neck cancers.