Isabelle Thibault1, Cari M Whyne2, Stephanie Zhou3, Mikki Campbell3, Eshetu G Atenafu4, Sten Myrehaug3, Hany Soliman3, Young K Lee3, Hamid Ebrahimi2, Albert J M Yee5, Arjun Sahgal6. 1. Department of Radiation Oncology, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada; Department of Radiation Oncology, Centre Hospitalier de L'Universite de Québec-Université Laval, Quebec, Quebec, Canada. 2. Orthopaedic Biomechanics Laboratory, Sunnybrook Research Institute, Department of Surgery, University of Toronto, Toronto, Ontario, Canada. 3. Department of Radiation Oncology, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada. 4. Department of Biostatistics, University Health Network, University of Toronto, Toronto, Ontario, Canada. 5. Division of Orthopaedic Surgery, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada. 6. Department of Radiation Oncology, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada. Electronic address: arjun.sahgal@sunnybrook.ca.
Abstract
PURPOSE: To determine a threshold of vertebral body (VB) osteolytic or osteoblastic tumor involvement that would predict vertebral compression fracture (VCF) risk after stereotactic body radiation therapy (SBRT), using volumetric image-segmentation software. METHODS AND MATERIALS: A computational semiautomated skeletal metastasis segmentation process refined in our laboratory was applied to the pretreatment planning CT scan of 100 vertebral segments in 55 patients treated with spine SBRT. Each VB was segmented and the percentage of lytic and/or blastic disease by volume determined. RESULTS: The cumulative incidence of VCF at 3 and 12 months was 14.1% and 17.3%, respectively. The median follow-up was 7.3 months (range, 0.6-67.6 months). In all, 56% of segments were determined lytic, 23% blastic, and 21% mixed, according to clinical radiologic determination. Within these 3 clinical cohorts, the segmentation-determined mean percentages of lytic and blastic tumor were 8.9% and 6.0%, 0.2% and 26.9%, and 3.4% and 15.8% by volume, respectively. On the basis of the entire cohort (n=100), a significant association was observed for the osteolytic percentage measures and the occurrence of VCF (P<.001) but not for the osteoblastic measures. The most significant lytic disease threshold was observed at ≥11.6% (odds ratio 37.4, 95% confidence interval 9.4-148.9). On multivariable analysis, ≥11.6% lytic disease (P<.001), baseline VCF (P<.001), and SBRT with ≥20 Gy per fraction (P=.014) were predictive. CONCLUSIONS: Pretreatment lytic VB disease volumetric measures, independent of the blastic component, predict for SBRT-induced VCF. Larger-scale trials evaluating our software are planned to validate the results.
PURPOSE: To determine a threshold of vertebral body (VB) osteolytic or osteoblastic tumor involvement that would predict vertebral compression fracture (VCF) risk after stereotactic body radiation therapy (SBRT), using volumetric image-segmentation software. METHODS AND MATERIALS: A computational semiautomated skeletal metastasis segmentation process refined in our laboratory was applied to the pretreatment planning CT scan of 100 vertebral segments in 55 patients treated with spine SBRT. Each VB was segmented and the percentage of lytic and/or blastic disease by volume determined. RESULTS: The cumulative incidence of VCF at 3 and 12 months was 14.1% and 17.3%, respectively. The median follow-up was 7.3 months (range, 0.6-67.6 months). In all, 56% of segments were determined lytic, 23% blastic, and 21% mixed, according to clinical radiologic determination. Within these 3 clinical cohorts, the segmentation-determined mean percentages of lytic and blastic tumor were 8.9% and 6.0%, 0.2% and 26.9%, and 3.4% and 15.8% by volume, respectively. On the basis of the entire cohort (n=100), a significant association was observed for the osteolytic percentage measures and the occurrence of VCF (P<.001) but not for the osteoblastic measures. The most significant lytic disease threshold was observed at ≥11.6% (odds ratio 37.4, 95% confidence interval 9.4-148.9). On multivariable analysis, ≥11.6% lytic disease (P<.001), baseline VCF (P<.001), and SBRT with ≥20 Gy per fraction (P=.014) were predictive. CONCLUSIONS: Pretreatment lytic VB disease volumetric measures, independent of the blastic component, predict for SBRT-induced VCF. Larger-scale trials evaluating our software are planned to validate the results.
Authors: Yilin Cao; Gregory C Stachelek; Wei Fu; Daniel Y Song; Russell K Hales; K Ranh Voong; Jeffrey J Meyer; Harry Quon; Chen Hu; Kristin J Redmond Journal: J Radiosurg SBRT Date: 2021