Ablative Bone Metastasis Radiotherapy Using Diagnostic Computed Tomography Scans for Planning

• Published in: International journal of radiation oncology, biology, physics Vol. 120; no. 2; p. e193
• Main Authors: Taylor, E., Javor, J., Tadic, T., Mesci, A., Wong, P., Cummings, B.J., Tsai, C.J.
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01-10-2024
• ISSN: 0360-3016
• DOI: 10.1016/j.ijrobp.2024.07.431

Generating radiotherapy plans on diagnostic computed tomography (CT) images presents substantial benefits for patients and radiotherapy (RT) clinic efficiency, especially in the palliative context. The elimination of simulation CT's presents several challenges to the accurate delivery of radiation, notably in terms of unknown geometric accuracy, CT number-to-density table, and differences in patient setup. Here we describe a method to ensure accurate RT delivery to bone metastases using plans generated on diagnostic CT's.
Volumetric Modulated Arc Therapy (VMAT) RT plans were generated retrospectively on diagnostic CTs for 18 patients with lesions in four non-spine/vertebral bone sites: pelvis (N = 3), hip/upper femur (N = 7), sternum (N = 5), and scapula (N = 3). Using a generic 100kV CT number-to-density table, the diagnostic CT plans were evaluated on simulation CTs acquired for each patient using a translation-only registration based on bone match (“simulated delivered”), simulating our standard institutional cone-beam CT (CBCT) image-guidance (IGRT) process. Differences in dose volume histogram (DVH) metrics between the diagnostic CT plan and simulated delivered plan were quantified and correlated to registration parameters. Muscle, air, and a modified bone contour were proposed as checks for CT number accuracy and geometric integrity, respectively. The workflow was implemented clinically for 4 patients to date, as part of standard-of-care for select patients. CBCT-based dose reconstruction was used to verify DVH metrics.
In the development workflow, setup differences between diagnostic and simulation CT's using bone-match registrations was strongly predictive of target coverage: for a 5mm PTV margin, the difference between planned and simulated delivered PTV D95% was inversely correlated with the root mean square rotation angle (r = -0.90; P < 0.00001). Maximum dose to critical organs-at-risk was not significantly correlated with setup differences (r = -0.13; P = 0.61). Proposed contour-based checks of geometric integrity and CT number accuracy worked in all 18 test cases. Applying our results to clinical treatments using 5° rotation thresholds for bone-match registrations resulted in a high clinical delivery accuracy for 4 treated patients who received ablative doses (>10Gy) in single-fraction VMAT treatments: based on dose reconstruction, CTV D99% was within 1% of planned values for all cases.
Implementing an off-line CT number check and on-line contour based geometric integrity checks while controlling setup differences using thresholds for bone-match registration rotation angles, the developed simulation CT free-based VMAT protocol can deliver dose accurately to palliative patients with metastatic bone disease.