PQA 04 - PQA 04 Palliative Care and Central Nervous System Poster Q&A
2599 - Local and Marginal Failures in Spinal and Non-Spinal Osseous Metastases Receiving SBRT and Immunotherapy: A Secondary Analysis of Phase I Trials
L. M. Serra1, C. Lynch1, M. C. Korpics1, M. Gutman1, C. M. Bestvina2, S. Pitroda1, A. Juloori1, S. J. Chmura1, and R. R. Katipally1; 1Department of Radiation and Cellular Oncology, University of Chicago, Chicago, IL, 2Department of Hematology Oncology, University of Chicago Medical Center, Chicago, IL
Purpose/Objective(s): Contouring guidelines for stereotactic body radiotherapy (SBRT) to bone metastases, especially to non-spine sites, are an evolving area. Inadequate CTV margins may lead to marginal failures or failures within the boney compartment. Here, we report the local control outcomes for spine and non-spine bone metastases treated with SBRT and immunotherapy across three prospective phase I trials. Materials/
Methods: Between 2016 and 2020, patients with widely metastatic solid malignancies were treated on three phase I trials (n = 213) combining multi-site SBRT with immunotherapy (regimens included pembrolizumab, ipilimumab with nivolumab, urelumab with nivolumab, or cabiralizumab with nivolumab; administered concurrently or sequentially). In this secondary analysis, patients that received SBRT to a bone metastasis were included. CTV expansions were 0mm for non-spine sites and anatomically defined per NRG-BR001 for spine sites. Local failure (LF) was defined as progression (per SPINO response assessment recommendations and imaging review by Radiation Oncology and Radiology) within the 95% isodose line. Marginal failure was considered progression within the 20%-95% isodose line. Outcomes were analyzed utilizing cumulative incidence with death as a competing risk. Univariable and multivariable Fine-Gray models were utilized to correlate LF rates with tumor, patient, and treatment-related variables, such as age, anatomical site, histology, GTV volume (cc), and PTV V95% (%). Results: 29 patients (median age 58 years; median OS 8.1 months) received SBRT to 37 bone metastases (25 [68%] non-spine; 12 [32%] spine; 95% received 30 Gy in 3 fractions). Non-spine sites included the chest (n = 12), pelvis (n =11), and femur (n = 1). The two most common histologies were lung (13 lesions [35%]) and breast (6 lesions [16%]). Median GTV volume was 7cc and PTV V95% was 98%. There were 4 (10.8%) local failures and zero marginal failures (or bone compartment failures) among treated lesions (median follow-up 8.1 months). In the multivariable Fine-Gray model, both lesion location (P < 0.001) as well as histology (P < 0.001) were associated with LF, driven by lesions in the chest and spine. Cumulative incidence of LF is presented in Table 1. Table 1 Conclusion: In widely metastatic patients receiving immunotherapy, the rate of local failure was low and marginal failures were not observed after SBRT to bone metastases, even with 0mm CTV expansions for non-spine lesions. Both histology and lesion location were associated with local failure. These findings inform adequate target delineation for future trials utilizing SBRT to non-spine and spine bone metastases.
