Diagnosis of spinal metastasis: are MR images without contrast medium application sufficient?

To determine the usefulness of adding contrast-enhanced (CE) magnetic resonance imaging (MRI) to conventional MRI for evaluation of spinal metastases.

One-hundred-and-two whole spine MR examinations, obtained for metastasis work-up within a 2-month period, from 65 men and 37 women (mean age, 64 years) with extra-spinal tumor, who also underwent CE-MRI, were retrospectively evaluated by three radiologists. The number of spine segments with bone marrow involvement was interpreted using a 3-point confidence scale (probable metastasis, equivocal, probably benign) during session 1 (conventional imaging) and session 2 (addition of CE-MRI to conventional imaging). The patients were assigned to 14 categories based on the changes in confidence rating between sessions 1 and 2; these were aggregated to four groups indicating the degree of usefulness of CE-MRI: definitely useful, equivocal, not useful, and presumed non-metastatic groups. Clinical information, metastatic bone type, the number of probably metastatic segments, and anatomical level and position were compared among the former three groups.

The readers assigned 39-53% of cases to the definitely useful group. The number of probably metastatic segments differed significantly among the three groups for all readers (p ≤ 0.046). Age, sex, primary cancer, metastatic bone type, and anatomical level and position were similar.

Adding CE-MRI to conventional MRI was useful for objectively detecting and characterizing spinal segments with metastases in 39-53% of cases. However, there were no clinical or radiological factors that could predict the usefulness of CE-MRI in evaluating spinal metastases, except for the number of metastatic segments.