Can recumbent magnetic resonance imaging replace myelography or computed tomography myelography for detecting lumbar spinal stenosis?

Abstract

To compare the diagnostic efficacy of recumbent magnetic resonance imaging (MRI), computed tomography myelography (CTM), and myelography, with regard to indications for surgery for lumbar stenosis. In patients with lumbar spinal stenosis-like disorders, small compressions are sometimes observed in magnetic resonance images acquired in the recumbent position, leading to potential misdiagnosis. Few prospective studies have compared the diagnostic accuracy of MRI, myelography, and CTM. Therefore, it is not clear whether myelography is necessary or not. Fifty-four patients fulfilled the criteria. All patients underwent MRI, myelography, and CTM. MRI was performed with the patient in a normal recumbent position, and CTM was performed with the patients in both a recumbent and extended positions. All patients underwent surgery for lumbar spinal stenosis. Findings from visual examinations (sagittal images of MR, axial images of MR, axial reconstruction images of CTM and myelograms) were defined as compression + or -. We analyzed the sensitivity of the different examinations for diagnosis and the relationship among the types of images. Sensitivity was as follows: CTM 94.4 %, myelography 87.0 %, and MRI 75.9 %. In myelography, the images of 37 patients were worsened by dynamic synthesis (Dyn+). Among patients without compression on MRI, 11 showed compression on myelography. Of these 11, 8 of these patients were Dyn+, and 2 patients showed compression on myelography, but not on CTM and were Dyn+. Thus, some compression can be revealed only with myelography. CTM was more sensitive than axial MRI and showed compression in 12 patients that was not detected by axial MRI. Myelography revealed stenosis that was not detected by MRI. CTM with extension is more sensitive for detecting stenosis than MRI. Recumbent MRI cannot replace myelography or CTM in terms of dynamic findings and sensitivity.