Re: detection rate of MR myelography without intrathecal gadolinium in patients with newly diagnosed spontaneous intracranial hypotension

Abstract

Sir—We read with interest the article by Lee et al.1 Lee S.J. Kim D. Suh C.H. et al. Detection rate of MR myelography without intrathecal gadolinium in patients with newly diagnosed spontaneous intracranial hypotension. Clin Radiol. 2022; 77: 848-854 Abstract Full Text Full Text PDF Scopus (0) Google Scholar assessing the sensitivity of non-invasive magnetic resonance myelography (MRM) without intrathecal gadolinium for cerebrospinal fluid (CSF) leakage in patients with spontaneous intracranial hypotension (SIH). The authors conclude MRM has a detection rate of 88% for CSF leakage and that it should be considered for this purpose in patients with SIH; however, the authors have not defined what constitutes “CSF leakage” on MRM. Spinal imaging studies in SIH can be categorised as those that confirm the presence of a spinal CSF leak and those that localise the leak. Although non-invasive MRM may show epidural fluid collections or paraspinal or foraminal fluid, implying a CSF leak is present, it should not be considered an adequate substitute for dynamic or decubitus computed tomography myelography (CTM) or digital subtraction myelography (DSM) in patients in whom leak localisation is necessary for planning targeted treatment. As previous investigators have found, non-invasive MRM is able to detect CSF leaks but unable to accurately localise them. 2 Tay A.S.M.S. Maya M. Moser F.G. et al. Computed tomography vs heavily T2-weighted magnetic resonance myelography for the initial evaluation of patients with spontaneous intracranial hypotension. JAMA Neurol. 2021; 78: 1275-1276 Crossref Scopus (6) Google Scholar This is also evidenced by the high number of patients in this paper reported to show CSF leakage at multiple levels of the spine, which is known to be a rare occurrence when leaks are localised precisely by DSM. 3 Schievink W.I. Maya M.M. Moser F.G. et al. Multiple spinal CSF leaks in spontaneous intracranial hypotension: do they exist?. Neurol Clin Pract. 2021; 11: e691-e697 Crossref Scopus (10) Google Scholar Fig. 1 of the paper, rather than demonstrating an example of leak localisation, actually shows an example of a false localising sign at C1–C2, which is a well-recognised occurrence due to egress of CSF through loose connective tissue at this level and which does not indicate the level of the leak. 4 Schievink W.I. Maya M.M. Tourje J. False localising sign of C1-2 cerebrospinal fluid leak in spontaneous intracranial hypotension. J Neurosurg. 2004; 100: 639-644 Crossref PubMed Scopus (72) Google Scholar Fig. 2 purports to show a leak at the cervicothoracic junction, but this is not clear from the representative image, which appears more consistent with prominent nerve root sleeves. As the authors acknowledge, non-invasive MRM is also unable to identify spinal CSF–venous fistulas, which are increasingly recognised as an important and curable cause of SIH.