In the present study, we investigated the potential for routine staging computed tomography (CT) to assess fracture risk in breast cancer patients.Patients and Methods:
A total of 184 randomized CT data sets of women with breast cancer were studied. Intellispace Density software (Philips, Amsterdam, Netherlands) was used to determine the mean volumetric bone mineral density (vBMD) of 3 nonfractured vertebral bodies (T12-L4) and predictors according to the quantitative and qualitative tumor morphology. Two radiologists detected prevalent vertebral fractures and classified them (osteoporotic vs. metastatic). For the determination of an association between bone mineral density, metastatic vertebral bone involvement, and the prevalence of vertebral fractures, a statistical analysis was conducted using multivariable logistic regression and receiver operating characteristic analyses.Results:
Of 184 women, 50 (27%) were diagnosed with prevalent vertebral metastases, of whom, 42 had vertebral fractures (23%). Of these fractures, 20 were found to be osteoporotic (11%), and 22 were associated with metastatic bone disease (12%). The vBMD T score showed the strongest association with osteoporotic fractures (odds ratio, 2.9; 95% confidence interval, 1.4–6.0; area under the curve [AUC], 0.89; P < .01). Metastatic vertebral fractures showed a strong association with the simple prevalence of metastases, regardless of tumor size (odds ratio, 9.8; 95% confidence interval, 4.3–22.3; AUC, 0.90; P < .01). Compared with the prevalence of metastases in the multivariate receiver operating characteristic statistic, the further detection of tumor size T0 and cortical bone infiltration showed a nonsignificant greater association with prevalent vertebral fractures (AUC, 0.92, P = .3; and AUC, 0.93, P = .07).Conclusion:
vBMD measurements and the detection of metastases and their morphology from routine staging CT allow the discrimination of breast cancer with and without vertebral fractures. The simple prevalence of vertebral metastases is strongly associated with prevalent metastatic fractures.Micro-Abstract:
The present study is the first to investigate routine computed tomography as a method to assess vertebral fracture risk in patients with breast cancer. The prevalence of mixed and osteolytic bone lesions and the detection of reduced bone density on computed tomography can discriminate breast cancer patients with and without vertebral fractures. Lesion size did not improve the fracture models significantly.