3472 - Risk of Vertebral Compression Fractures after Hypofractionated Spine Stereotactic Body Radiation Therapy

Purpose/Objective(s): Vertebral compression fracture (VCF) is the most common complications after spine SBRT and can cause significant pain and/or neurological deficits. We aimed to evaluate the incidence and risk factors of VCF after spine SBRT with a single-dose segmentation protocol at our institution.

Materials/Methods: The patients treated with spine SBRT from January 2021 to January 2024 were enrolled retrospectively. The prescribed dose is 30-35 Gy/5 fraction. The primary endpoint was the diagnosis of VCF, which was defined as radiologically new or worsening fracture based on MRI and/or CT that was not associated with concurrent tumor progression, The secondary endpoints were pain relief rates at 1, 3, and 6 months. Patient characteristics and therapeutic factors were analyzed. Patient factors included osteolytic or osteogenic, oligometastatic or not, the primary tumor, single or multiple bone metastases, the site of vertebral metastasis, age, gender, SINS, PS, VAS, Bilsky score, and baseline VCF. Therapeutic factors include pre-treatment vertebroplasty or not, the use of bone-modifying drugs or not, systemic treatment, and dosimetric parameters such as the volume of PTV, the D_max and D_mean of PTV, PTV_D_10%-90% (the radiation dose received by 10%-90% of PTV), V_15-35Gy (the volume irradiated by 15Gy-35Gy radiation dose), the D_max?D_0.3cc?D_1.2cc of the spinal cord PRV. Univariate and multivariate modified Poisson regression analyses were performed to determine independent risk factors for VCF.

Results: There were 113 patients treated to 159 segments in the analysis, with a median follow-up of 12 months (range 6-35 months). 13 (11.50%) cases of progressive VCF were diagnosed, including 2 cervical, 7 thoracic, and 4 lumbar, of which 23.08% (3 patients) required surgical intervention to stabilize the disease. The median time of VCF was 5 months (range 1-11 months). The rates of complete pain relief at 1, 3 & 6 months were 29.3 %?45.2% and 46.3%, respectively. Osteogenic and osteolytic lesions accounted for 82.92% and 17.08%, respectively. High-grade spinal cord compression was present in 27 patients (23.89%), of which 96.3% had a decrease in Bilsky score after SBRT. 110 (96.60%) of cases were treated with bone-modifying drugs, and 52 (59.09%) patients with effective systemic therapy. 11 (9.73%) patients underwent vertebroplasty prior to radiotherapy. In univariate analysis, Bilsky score, high-level epidural spinal cord compression, baseline VCF, and V_15Gy?V_20Gy?V_25Gy?V_30Gy?and V_35Gy were predictors of VCF. Significant predictors of VCF on multivariable analyses included Bilsky score, V_25Gy, and baseline VCF. The area under the ROC curve for the logistic regression prediction model is 0.866 (P=0.001, 95%CI 0.729-0.985).

Conclusion: The incidence of VCF after spine SBRT at our institution was 11.50%, and V_25Gy was a dosimetric predictor of VCF.