Association of Extension of Cervical Cord Lesion and Area Postrema Syndrome With Neuromyelitis Optica Spectrum Disorder

Neuromyelitis optica spectrum disorders (NMOSDs) are autoimmune disorders of astrocytic water channels associated with the highly specific serum biomarker, aquaporin-4 (AQP-4) IgG.¹ Aquaporin-4 has high expression in the circumventricular organs and spinal cord. The area postrema (AP), located in the dorsal medulla, is the chemosensitive vomiting center and has high AQP-4 expression.² The AP syndrome with unexplained hiccups, nausea, and vomiting is one of the core clinical characteristics in the 2015 diagnostic criteria for NMOSDs and may occur concurrently with a myelitis episode.¹ Prior studies have suggested spinal cord lesions extending to the AP are specific for NMOSDs.³ However, in our clinical practice, we have encountered numerous non-NMOSD long myelopathies extending to the region of the AP. In this study, we evaluated whether longitudinally extensive cervical cord lesions extending to or accompanied by AP/dorsal medulla lesions on magnetic resonance imaging (MRI) are specific for NMOSDs.

We reviewed the records and MRIs of Mayo Clinic patients with cervical spinal cord lesions from our AQP-4 IgG–seropositive NMOSD myelitis (1996-2014) and long myelopathy (1996-2016) databases. The Mayo Clinic institutional review board approved the study, and patients consented to the use of their medical records for research purposes. Demographic and clinical data were collected.

For those with dorsal medulla involvement on MRI, the medical records were evaluated for intractable nausea, vomiting, or hiccups within 6 weeks of MRI. Categorical variables and continuous variables were analyzed using χ² and Mann-Whitney U tests, respectively.

Two hundred thirty six cases of long myelopathy involving the cervical spinal cord were analyzed. Among these, 180 patients had NMOSDs and 56 had other etiologies including sarcoidosis (n = 29), spondylosis (n = 16), lymphoma (n = 6), spinal cord infarction (n = 3), dural arteriovenous fistula (n = 1), and paraneoplastic myelitis (n = 1). Forty seven cases (19.9%) had extension of the cord lesion to or coexisting involvement of the dorsal medulla (Table). There was no difference in the proportion of cases with extension of the lesion to dorsal medulla between NMOSDs and other etiologies (19% vs 21%; P = .44). All NMOSD cases were AQP-4 IgG positive by cell-based assay. Among the cases with extension to or involvement of the dorsal medulla, a higher proportion of patients with NMOSDs had intractable nausea and vomiting compared with other etiologies (51% vs 0%; P < .05; odds ratio, 23.6; 95% CI, 1.3-430.4) (Figure). A higher proportion of NMOSD cases had hiccups (26% vs 0%; P = .05), although the difference did not reach statistical significance.

Our study refuted the misconception that longitudinally extensive spinal cord lesions extending to AP are highly specific for NMOSDs and showed that such lesions accompany diverse etiologies of myelopathy. However, when extension of a spinal cord lesion to the AP is accompanied by intractable nausea and vomiting, it is highly specific for AQP-4 IgG–seropositive NMOSD. This confirms the utility of the updated 2015 NMOSD diagnostic criteria,¹ which emphasize clinical rather than MRI features of AP syndromes and are in conflict with a prior study with much smaller numbers suggesting cervical cord lesion extension to the AP is specific for NMOSDs.³ The AP is highly vascularized and lacks blood-brain-barrier, potentially increasing its exposure to blood-borne AQP-4 IgG and may explain its predilection to cause AP syndromes vs alternative etiologies with MRI abnormalities in this region.² The reason cervical cord lesions extend to the medulla in other etiologies is unclear but spinal cord injury studies have hypothesized this could be from raised intramedullary pressure resulting from narrowed spinal canal, cord swelling, and impaired circulation.⁶ Other radiological factors, such as gadolinium enhancement pattern, may be more helpful than extension of T2 signal to the dorsal brainstem in distinguishing between long myelopathy etiologies (Figure).^4,5

Corresponding Author: Eoin P. Flanagan, MD, Department of Neurology, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (flanagan.eoin@mayo.edu).

Published Online: January 9, 2017. doi:10.1001/jamaneurol.2016.5441

Author Contributions: Dr Flanagan had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Dubey, Pittock, Flanagan.

Acquisition, analysis, or interpretation of data: Dubey, Krecke, Flanagan.

Drafting of the manuscript: Dubey.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Dubey.

Obtained funding: Pittock.

Study supervision: Flanagan.

Conflict of Interest Disclosures: Dr Pittock and the Mayo Clinic have a financial interest in patents that relate to functional aquaporin-4/neuromyelitis optica IgG assays and neuromyelitis optica IgG as a cancer marker. Dr Pittock has provided consultation to Alexion Pharmaceuticals, MedImmune, and Chugai Pharma USA but has received no personal fees or personal compensation for these consulting activities; all compensation for consulting activities is paid directly to Mayo Clinic. Dr Pittock has received a research grant from Alexion Pharmaceuticals for an investigator-initiated study and grant RO1 NS065829-01 from the National Institutes of Health and support from the Guthy-Jackson Charitable Foundation for neuromyelitis optica research. No other disclosures were reported.