Comparison of diffusion-weighted whole body MRI and skeletal scintigraphy for the detection of bone metastases in patients with prostate or breast carcinoma

To prospectively compare the diagnostic accuracy of diffusion-weighted whole body imaging with background whole body signal suppression (DWIBS) with skeletal scintigraphy for the diagnosis and differentiation of skeletal lesions in patients suffering from prostate or breast cancer. A diagnostic cohort of 36 patients was included in skeletal scintigraphy and 1.5 T DWIBS MRI. Based on morphology and signal intensity patterns, two readers each identified and classified independently, under blinded conditions, all lesions into three groups: (1) malignant, (2) unclear if malignant or benign and (3) benign. Finally, for the definition of the gold standard all available imaging techniques and follow-up over a minimum of 6 months were considered. Overall, 45 circumscribed bone metastases and 107 benign lesions were found. DWIBS performed significantly better in detecting malignant skeletal lesions in patients with more than 10 lesions (sensitivity: 0.97/0.91) compared to skeletal scintigraphy (sensitivity: 0.48/0.42). No statistical difference could be found between DWIBS (0.58/0.33) and skeletal scintigraphy (0.67/0.58) in the sensitivity values for malignant skeletal lesions in patients with less than 5 lesions. For benign lesions, scintigraphy scored best with a sensitivity of 0.93/0.87 compared to 0.20/0.13 for DWIBS. Interobserver agreement with Cohen's kappa coefficient was calculated as 0.784 in the case of scintigraphy and 0.663 for DWIBS. With respect to staging, in prostate and breast carcinoma, the DWIBS technique is not superior to skeletal scintigraphy, but ranks equally. However, in the cases with many bone lesions, markedly more metastases could be discovered using the DWIBS technique than skeletal scintigraphy.