Detection of cerebrospinal fluid leakage: initial experience with three-dimensional fast spin-echo magnetic resonance myelography

• Published in: Acta radiologica (1987) Vol. 49; no. 2; p. 197
• Main Authors: Tomoda, Y, Korogi, Y, Aoki, T, Morioka, T, Takahashi, H, Ohno, M, Takeshita, I
• Format: Journal Article
• Language: English
• Published: England 01.03.2008
• Subjects: Adult; Aged; Echo-Planar Imaging - methods; Feasibility Studies; Female; Humans; Hypovolemia - diagnosis; Hypovolemia - etiology; Imaging, Three-Dimensional - methods; Indium Radioisotopes; Intracranial Hypotension - diagnosis; Intracranial Hypotension - etiology; Male; Middle Aged; Observer Variation; Radionuclide Imaging; Reference Values; Retrospective Studies; Spinal Cord - diagnostic imaging; Spinal Cord - pathology; Subdural Effusion - diagnosis; Time Factors
• DOI: 10.1080/02841850701769785

The pathogenesis of cerebrospinal fluid (CSF) hypovolemia is supposed to be caused by CSF leakage through small dural defects. To compare source three-dimensional (3D) fast spin-echo (FSE) images of magnetic resonance (MR) myelography with radionuclide cisternography findings, and to evaluate the feasibility of MR myelography in the detection of CSF leakage. A total of 67 patients who were clinically suspected of CSF hypovolemia underwent indium-111 radionuclide cisternography, and 27 of those who had direct findings of CSF leakage were selected for evaluation. MR myelography with 3D FSE sequences (TR/TE 6000/203 ms) was performed at the lumbar spine for all patients. We evaluated source images and maximum intensity projection (MIP) images of MR myelography, and the findings were correlated with radionuclide cisternography findings. MR myelography of five healthy volunteers was used as a reference. The MR visibility of the CSF leakage was graded as definite (leakage clearly visible), possible (leakage poorly seen), or absent (not shown). CSF leakage was identified with source 3D FSE images in 22 (81.5%) of 27 patients. Of the 22 patients, 16 were graded as definite and six were graded as possible. For the definite cases, 3D FSE images clearly showed the extent of the leaked CSF in the paraspinal structures. In the remaining five patients with absent findings, radionuclide cisternography showed only slight radionuclide activity out of the arachnoid space. Source 3D FSE images of MR myelography seem useful in the detection of CSF leakage. Invasive radionuclide cisternography may be reserved for equivocal cases only.