Imaging bone metastases in breast cancer: Evidence on comparative test accuracy

Background: Numerous imaging modalities may be used to detect bone metastases (BM) in women with breast cancer. Methods: Systematic evidence review, including quality appraisal, of studies reporting on comparative imaging accuracy for detection of BM from breast cancer. Results: Eligible studies (N = 16) included breast cancer subjects who had imaging evaluation for suspected BM or for staging/restaging in suspected local or distant relapse. Median prevalence of BM was 34.0% (range 10.0%-66.7%). There was substantial heterogeneity in the quality of reference standards and in the prevalence of BM, which could account for some of the differences in reported comparative accuracy. Most frequently, bone scan (BS) was compared with newer imaging modalities in subjects selected to both tests; therefore, results could be affected by selection bias. There was some evidence that positron emission tomography (PET), and limited evidence that PET/computed tomography (CT), CT, and magnetic resonance imaging (MRI), may provide small increments in accuracy relative to BS as add-on tests; there was little evidence regarding single photon emission computed tomography or whole-body MRI. Conclusions: There is some evidence of enhanced incremental accuracy for some of the above-mentioned tests where used as add-on in subjects selected to more than one imaging modality, with little evidence to support their application as a replacement to BS in first-line imaging of BM. PET/CT appears to have high accuracy and is recommended for further evaluation.