Diagnosis of vertebral metastasis, epidural metastasis, and malignant spinal cord compression: are T(1)-weighted sagittal images sufficient?

The objective of this study was to determine whether T(1)-weighted sagittal images alone are adequate in the diagnosis of vertebral metastasis, epidural metastasis, and malignant spinal cord compression. Ninety-four complete magnetic resonance (MR) studies of the spinal column (a complete study consisting of T(1)-weighted sagittal images, T(2)-weighted sagittal images, and T(1)- and/or T(2)-weighted axial images) and 94 T(1)-weighted sagittal images alone (a subset of the complete studies) from 57 consecutive cancer patients over the last 2 years with clinically suspected cord compression were blindly and independently evaluated by four radiologists. The complete MR studies were used as the standard. Overall, the sensitivity of T(1)-weighted sagittal images alone to vertebral metastasis (87%) was statistically greater than cord compression (70%) (p = 0.05), and statistically greater than epidural metastasis (46%) (p </= 0.02). The specificity for cord compression (97%) was greater than the specificity for epidural metastasis (89%) (p = 0.03), and greater than the specificity for vertebral metastasis (83%) (p </= 0.02). There was a strong trend for better detection of cord compression overall and better detection of vertebral metastasis in the cervical spine by the most experienced radiologist. Complete studies of the spine are necessary in the diagnosis of vertebral metastasis, epidural metastasis, and cord compression, particularly with less experienced radiologists.