149 - Risk of Radiation Myelitis after Hypofractionated Spine Stereotactic Body Radiation Therapy

C. B. Jackson¹, L. A. Boe², L. Zhang³, A. Apte³, L. M. Ruppert¹, J. Haseltine¹, B. A. Mueller⁴, A. Schmitt¹, J. T. Yang⁵, W. C. Newman¹, O. Barzilai¹, M. Bilsky¹, Y. Yamada⁴, A. Jackson³, E. Lis¹, and D. S. Higginson¹; ¹Memorial Sloan Kettering Cancer Center, New York, NY, ²Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, ³Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, ⁴Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, ⁵University of Washington, Seattle, WA

Purpose/Objective(s): Stereotactic body radiation therapy (SBRT) for spinal metastases improves symptomatic outcomes and local control compared to conventional radiotherapy (RT). Treatment failure most often occurs within the epidural space, where dose is constrained by the risk of radiation myelitis (RM). We conducted a retrospective cohort study of patients treated with 3-fraction spine SBRT to create modern risk assessments of RM. Materials/Methods: Patients were treated at a single institution between 2014-2023 with 3-fraction spine SBRT to a level between C1-L2. Patients were excluded if they had received overlapping prior RT, had less than 1 month of magnetic resonance imaging (MRI) follow up, or had a maximal dose to a voxel of spinal cord (Dmax) of 0 Gy. RM was defined as radiographic evidence of cord injury in the treatment field and further classified into grade (G) 1-4 or G3+ by Common Terminology Criteria for Adverse Events v5.0. We assessed multiple dosimetric parameters of the true spinal cord structure for association with risk of RM. Univariable, cause-specific hazards, competing risk regression models were fit for RM using the Cox proportional hazards model with robust standard errors to account for patient-level clustering, with the dosimetric variables as predictors. To generate a dose-response model of RM, predicted probabilities and 95% confidence interval (CI) limits were estimated at 2 years using the regression models for the dosimetric variables of the minimum dose to the 0.1 cm³ of spinal cord receiving the greatest dose (D0.1cc) or Dmax.

Results: There were 1,423 patients treated to 1,904 segments in the analysis. Median follow up was 13 months (interquartile range 7-26 months). We identified 30 cases of RM, 19 of which were G3+. At 2 years, the cumulative incidence of G1-4 RM was 1.8% (95% CI 1.2%-2.5%), and the rate of G3+ RM was 1.1% (95% CI 0.71%-1.7%). For patients who experienced RM, the median time to myelitis was 10 months (range 3.7-29.7 months). Cord D0.1cc was the most important dosimetric predictor of RM on univariable, cause-specific hazards regression (for G3+ RM, hazard ratio (HR) 2.14, 95% CI 1.68-2.72, p < 0.0001). Dmax was also an important predictor of RM (for G3+ RM, HR 1.82, 95% CI 1.48-2.24, p < 0.0001). According to our dose-response model, a true cord D0.1cc of 19.1 Gy and Dmax of 20.8 Gy predict a 1% risk (95% CI 0.3%-1.6% and 0.4%-1.6%, respectively) of G3+ RM 2 years from SBRT.

Conclusion: This is the largest series of RM cases after spine SBRT. A true cord D0.1cc constraint of 19.1 Gy and a Dmax constraint of 20.8 Gy correspond with a 1% risk of G3+ RM at 2 years. Abstract 149 – Table 1

Cord D0.1cc	Modeled Probability of G1-4 Myelitis (95% CI)	Modeled Probability of G3+ Myelitis (95% CI)
14.2 Gy	0.1% (0%, 0.2%)	0% (0%, 0.1%)
16.2 Gy	0.3% (0%, 0.6%)	0.1% (0%, 0.2%)
17 Gy	0.5% (0%, 1%)	0.2% (0%, 0.4%)
18 Gy	1% (0.3%, 1.7%)	0.4% (0%, 0.8%)
18.2 Gy	1.1% (0.3%, 1.9%)	0.5% (0.1%, 0.9%)
19.1 Gy	1.9% (0.9%, 3%)	1% (0.3%, 1.6%)
19.8 Gy	3% (1.7%, 4.4%)	1.6% (0.7%, 2.5%)
20.6 Gy	5% (3.1%, 6.9%)	3% (1.5%, 4.4%)