2125 - Molecular Status Predicts for Local Control in Patients with Non-Small Cell Lung Cancer Spinal Metastases Following Spine Stereotactic Body Radiotherapy
Sunday, October 1, 2023
4:45 PM – 6:00 PM PT
Location: Hall B2
Screen: 24
D. Shor1, K. L. Zeng2, H. Chen2, A. V. Louie2, I. Menjak3, E. Atenafu4, C. L. Tseng2, J. Detsky2, J. Larouche5, B. Zhang2, H. Soliman2, P. Maralani6, S. D. Myrehaug2, and A. Sahgal2; 1Sunnybrook Odette Cancer Centre, Toronto, ON, Canada, 2Department of Radiation Oncology, Odette Cancer Centre, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada, 3Sunnybrook Health Sciences Centre, Toronto, ON, Canada, 4Princess Margaret Cancer Centre, University Health Network, Toronto, ON, Canada, 5Division of Orthopedic Surgery, Sunnybrook Health Sciences Centre, Toronto, ON, Canada, 6Department of Neuroradiology, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada
Purpose/Objective(s): We report outcomes after spine stereotactic body radiotherapy (SBRT) in patients with metastatic non-small cell lung cancer (NSCLC), to determine the significance of programmed death-ligand 1 (PD-L1) status and epidermal growth factor (EGFR) mutation on local failure (LF) rate. Materials/
Methods: 165 patients and 389 spinal segments were retrospectively reviewed from 2009 to 2021. Baseline patient characteristics, treatment and outcomes were abstracted. Primary endpoint was LF and secondary outcomes included overall survival (OS) and vertebral compression fracture (VCF) rates. OS was estimated using the Kaplan-Meier method. Cumulative LF and VCF rates were calculated using competing risk analysis method. Multivariable analysis (MVA) evaluated factors predictive of LF and VCF. Results: Median follow-up was 13 months (range, 0.5-95 months). Median OS was 18.4 months (95% CI 11.4-24.6). Median age was 67 years (range, 28.2-89.9). 52% were female, 76% had an adenocarcinoma histology and 61% had a smoking history. 49/165 (29%) had an EGFR mutation. PD-L1 status was analysed in 109/165 (66%) patients with 16% PD-L1 = 50%, 20% PD-L1 1-49% and 35% PD-L1 <1%. Of 389 segments, 79% were denovo and 21% were previously radiated. At baseline, 35% had a VCF, 27% had epidural disease, 27% had paraspinal extension, and 49% were Spinal Instability in Neoplasia Score (SINS) stable. 239/389 (61%) were treated with either 24 or 28 Gy in 2 SBRT fractions. Within 1 month of SBRT, 39/165 (24%) had a tyrosine kinase inhibitor, 27/165 (16%) immunotherapy (IO) with or without chemotherapy, and 31/165 (19%) chemotherapy alone. LF cumulative incidence at 1- and 2-years was 16.3% (95% CI 12.8-20.3%) and 25.4% (95% CI 20.9%-30%), respectively. EGFR positivity (p<0.0001), PD-L1=50% (p=0.013) and treatment with IO within 1 month of SBRT (p=0.004) predicted for improved local control on MVA. The 1- and 2-year LF rate in EGFR-positive vs. negative patients were 12.9% vs. 16.6% and 17.7% vs. 28.8%, respectively, and in those PD-L1 =50% vs PD-L1<50% were 7.8% vs. 19.6% and 7.8% vs. 38.1% respectively. Cumulative incidence of VCF at 1- and 2-years were 6.6% (95% CI 4.4-9.4%) and 8.8% (95% CI 6.1-12.0%). MVA identified prior SBRT to the same treated segment (P<0.0001) and a baseline VCF (p<0.0001) as significant predictors. 18/389 (4.6%) had radiation-induced radiculopathy and no radiation myelopathy events detected. Conclusion: We identify the predictive utility of EGFR mutation and PD-L1 =50% status on local control in NSCLC patients with spinal metastases treated with spine SBRT, and a therapeutic benefit with peri-SBRT IO.