2625 - Extended Field Radiosurgery of Spinal Metastases with MRI-Guided Adaptive Planning: A Dosimetric Study
Presenter(s)
D. J. Lee1, B. Guevara2, C. Montoya1, K. Brown3, P. Kelly4, E. A. Mellon5, and B. Spieler2; 1Sylvester Comprehensive Cancer Center, University of Miami Miller School of Medicine, Miami, FL, 2Department of Radiation Oncology, University of Miami/Sylvester Comprehensive Cancer Center, Miami, FL, 3Department of Radiation Oncology, University of Miami/Sylvester Comprehensive Cancer Center, Miami, FL, Miami, FL, 4Department of Radiation Oncology, Orlando Health Cancer Institute, Orlando, FL, 5Department of Radiation Oncology, Sylvester Comprehensive Cancer Center, Miller School of Medicine, University of Miami, Miami, FL
Purpose/Objective(s): Up to 50% of spinal metastases synchronously involve multiple vertebral levels. Stereotactic body radiation therapy (SBRT) provides quicker and more robust pain relief and superior local control compared to conventionally fractionated radiation. However, subtle movements of the vertebral column between simulation and treatment can result in radiation overdose to the spinal cord and increased risk of myelopathy. Consensus guidelines currently restrict SBRT to =3 contiguous vertebral segments. Adaptive linacs offer live target/organ-at-risk (OAR) tracking and adaptive re-planning. This dosimetric study evaluates the ability of an MRI-based adaptive linac platform to safely deliver SBRT to spinal metastases involving 3-7 contiguous vertebral segments.
Materials/Methods: Ten (10) patients previously treated on a 0.35-T MRI-linac and registered on an IRB-approved database were randomly selected. All selected patients had undergone a “simulation” MRI and “treatment” MRI, with the two MRIs performed =15 days apart and each capturing the T8-L2 vertebrae ± 2 cm. Adaptive and non-adaptive 18 Gy single-fraction plans for the treatment of 3, 4, 5, 6, and 7 contiguous vertebral segments were created in software for each patient. Percent PTV coverage and satisfaction of spinal cord (SC) dose constraints were assessed for each plan. Plans achieving =90% PTV coverage, SC Dmax = 14 Gy and SC V10 Gy = 0.35 cc were considered as meeting accepted planning goals for treatment. The ability of adaptive and non-adaptive plans to meet these constraints were compared, controlling for the number of treated vertebral segments, using paired t-tests and McNemar’s tests in Stata.
Results: For each of 3, 4, 5, 6, and 7 contiguous vertebral segments, adaptive SBRT generated safer plans, meeting spinal cord constraints in all instances, with significantly reduced maximum doses (mean SC Dmax 10.7 Gy adaptive vs. 15.7 Gy non-adaptive; each p<0.01), improved target coverage (mean PTV coverage 92.6% adaptive vs. 88.4% non-adaptive; each p<0.01), and reduced variations in target coverage (mean standard deviation of PTV coverage 0.44% adaptive vs. 1.83% non-adaptive, p=0.003).
Conclusion: In this dosimetric study, adaptive SBRT significantly improved target coverage and SC dose constraint satisfaction when treating extended spine fields >3 vertebral levels compared to non-adaptive SBRT. The positive results of this study warrant safety and efficacy testing in a prospective clinical trial using adaptive SBRT to treat extended spine fields.