CT–Guided Bone Biopsies in Metastatic Castration-Resistant Prostate Cancer: Factors Predictive of Maximum Tumor Yield

Abstract

PurposeTo evaluate the success rate of CT–guided bone biopsies in metastatic castration-resistant prostate cancer (mCRPC) and to investigate associated technical, imaging, and clinical parameters affecting diagnostic yields. Materials and MethodsEighty CT–guided bone biopsy specimens were obtained from 72 men (median age, 68 y; range, 49–89 y) enrolled in a multicenter trial to identify mechanisms of resistance in mCRPC. Successful biopsy was determined by histologic confirmation of tumor cells and successful isolation of RNA for molecular analysis. ResultsThe overall success rate of CT–guided bone biopsies was 69% (55/80) based on histology and 64% (35/55) based on isolation of molecular material for RNA sequencing. Biopsies performed in lesions with areas of radiolucency had significantly higher diagnostic yields compared with lesions of predominantly dense sclerosis (95% vs 33%; P = .002) and lesions of predominantly subtle sclerosis (95% vs 65%; P = .04). Success rates increased in lesions with density ≤ 475 HU (79% for ≤ 475 HU vs 33% for > 475 HU; P = .001) and in lesions with ill-defined margins (76% for ill-defined margins vs 36% for well-circumscribed margins; P = .005). Alkaline phosphatase was the only clinical parameter to correlate significantly with diagnostic yield (83% for > 110 U/L vs 50% for ≤ 110 U/L; P = .001). ConclusionsImage-guided bone tumor biopsies can be successfully used to acquire cellular and molecular material for analyses in patients with osteoblastic prostate cancer metastases. Diagnostic yields are significantly increased in lesions with areas of radiolucency, density ≤ 475 HU, ill-defined margins, and interval growth and in patients with alkaline phosphatase > 110 U/L.