Vertebral and femoral bone mineral density and bone strength in prostate cancer patients assessed in phantomless PET/CT examinations

Bone fracture risk assessed ancillary to positron emission tomography with computed tomography co-registration (PET/CT) could provide substantial clinical value to oncology patients with elevated fracture risk without introducing additional radiation dose. The purpose of our study was to investigate the feasibility of obtaining valid measurements of bone mineral density (BMD) and finite element analysis-derived bone strength of the hip and spine using PET/CT examinations of prostate cancer patients by comparing against values obtained using routine multidetector-row computed tomography (MDCT) scans—as validated in previous studies—as a reference standard. Men with prostate cancer (n = 82, 71.6 ± 8.3 years) underwent Fluorine-18 NaF PET/CT and routine MDCT within three months. Femoral neck and total hip areal BMD, vertebral trabecular BMD and femur and vertebral strength based on finite element analysis were assessed in 63 paired PET/CT and MDCT examinations using phantomless calibration and Biomechanical-CT analysis. Men with osteoporosis or fragile bone strength identified at either the hip or spine (vertebral trabecular BMD ≤ 80 mg/cm3, femoral neck or total hip T-score ≤− 2.5, vertebral strength ≤ 6500 N and femoral strength ≤ 3500 N, respectively) were considered to be at high risk of fracture. PET/CT- versus MDCT-based BMD and strength measurements were compared using paired t-tests, linear regression and by generating Bland-Altman plots. Agreement in fracture-risk classification was assessed in a contingency table. All measurements from PET/CT versus MDCT were strongly correlated (R2 = 0.93–0.97; P < 0.0001 for all). Mean differences for total hip areal BMD (0.001 g/cm2, 1.1%), femoral strength (− 60 N, 1.3%), vertebral trabecular BMD (2 mg/cm3, 2.6%) and vertebral strength (150 N; 1.7%) measurements were not statistically significant (P > 0.05 for all), whereas the mean difference in femoral neck areal BMD measurements was small but significant (− 0.018 g/cm2; − 2.5%; P = 0.007). The agreement between PET/CT and MDCT for fracture-risk classification was 97% (0.89 kappa for repeatability). Ancillary analyses of BMD, bone strength, and fracture risk agreed well between PET/CT and MDCT, suggesting that PET/CT can be used opportunistically to comprehensively assess bone integrity. In subjects with high fracture risk such as cancer patients this may serve as an additional clinical tool to guide therapy planning and prevention of fractures.