Abstract
1936
Objectives Spontaneous spinal cerebrospinal fluid (CSF) leak is a cause of orthostatic headache due to intracranial hypotension. In-111 DTPA cisternography has been used in patients with orthostatic headache to diagnose and localize CSF leak (Schievink WI, JAMA, 2006). Sensitivity and specificity of In-111 DTPA cisternography, however, have not been elucidated in large number of patients. We conducted multicenter prospective study to reveal the accuracy of In-111 DTPA cisternography and other MR-based imaging modalities.
Methods One hundred patients with orthostatic headache were registered from 11 medical institutions in Japan between May 2008 and March 2011 under the control of the Cerebrospinal Fluid Hypovolemis Research Group. These patients were examined with brain MRI (n=89), In-111 DTPA cisternography (n=89), MR myelography (n=86), axial T2-weighted MRI of the spine (n = 70), and/or CT myelography (n = 2). The spinal CSF leak was diagnosed by experienced neuroradiologist (TH) and nuclear medicine physician (JH) who were blinded to patient identity. The diagnosis of spinal CSF leak was based on all the available imaging findings.
Results Sixteen patients were diagnosed in consensus as having spinal CSF leak. 76 patients showed no imaging evidence of CSF leak. The location of CSF leak was cranial to cervical portion in 1 patient, cervical in 4, cervical to thoracic in 6, thoracic in 3, and lumbar in 2. Sensitivity and specificity of the In-111 cisternography for spinal CSF leak was 94% and 83%, respectively. Sensitivity and specificity of the MR myelography was 81% and 94%, respectively.
Conclusions In-111 DTPA cisternography remains useful especially when spinal MR/MR myelography can not localize CSF leak