Neuromyelitis optica spectrum disorder, and related disorders : epidemiology, imaging biomarkers, and electrophysiology

<p dir="ltr">Background: Acute myelitis is often associated with central nervous system inflammatory diseases as multiple sclerosis (MS), or with the antibody-driven diseases neuromyelitis optica spectrum disorder (NMOSD, i.e., neuromyelitis optica (NMO)), and myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD). Studies on the incidence and long-term outcome of acute myelitis are scarce, and much is still unclear the epidemiology and pathophysiology of NMOSD and MOGAD. Both NMOSD and MOGAD are commonly treated with the immunosuppressive agent rituximab, but only a few reports provide real-world data on efficacy and safety.</p><p dir="ltr">Enlarged perivascular spaces (EPVS) are fluid-filled lesions on brain MRI, possibly involved in autoimmune CNS processes. Studies on EPVS in NMOSD and MOGAD are few, and comparative studies between neuroinflammatory diseases are lacking.</p><p dir="ltr">An important and challenging differential diagnosis for acute myelitis is spinal cord infarction (SCI). Motor and sensory evoked potentials (MEP/SEP) test the integrity of spinal cord pathways and have been used in intraoperative monitoring to detect spinal cord ischemia.</p><p dir="ltr">Objectives: The aim of this thesis was: to estimate the incidence and prevalence of NMO/NMOSD in Sweden (Study I); to explore how patients with NMOSD and MOGAD fare on rituximab in terms of relapse incidence and risk of severe infections (Study II); to describe the incidence and clinical presentation of acute myelitis, as well as functional outcome and relapse risk (Study III); to examen the distribution of EPVS in patients with various neuroinflammatory diseases (Study IV); and to test if evoked potentials can differentiate SCI from acute myelitis (Study V).</p><p dir="ltr">Methods: Study I, included a cohort of individuals with a diagnostic code in the national patient registry (NPR) for NMO/NMOSD, or with a positive aquaporin-4 antibody (AQP4-IgG) analysis, in Sweden from 1987 to 2013 (n = 294). The 2006 NMO diagnostic criteria was used to identify NMO patients (n = 51), and individuals with AQP4-IgG but a more limited clinical presentation were categorized as NMOSD (n = 41).</p><p dir="ltr">Study III, a cohort study, included individuals with diagnostic codes in the NPR likely to include acute myelitis in Stockholm County from 2008 to 2018 (n = 2321). A modified version of the 2002 Transverse Myelitis Consortium Working Group criteria, which also allows for partial myelitis, was used to identify patients with acute myelitis (n = 461). Follow-up data was collected for patients not receiving an initial diagnosis of MS or NMOSD. Uni- and multivariate analyses were used to explore factors of poor functional outcome, and factors related to risk of relapse after an idiopathic myelitis.</p><p dir="ltr">In Studies II and IV, patients with NMOSD and MOGAD at Karolinska University Hospital were included. Study III retrospectively studied a cohort of NMOSD, both AQP4-IgG positive (AQP4+NMOSD) (n = 24) and seronegative (n = 10), as well as MOGAD (n = 8) patients treated with rituximab between 2012 and 2021. Study IV prospectively included all available AQP4+NMOSD (n = 22), seronegative NMOSD (n=8), and MOGAD (n = 10) patients during 2021-2022. Additionally, Study IV included three MS cohorts (n = 146) previously collected as part of other studies and healthy controls (n = 70).</p><p dir="ltr">Incidence rates for relapse and severe infectious events (SIE) during rituximab treatment, in Study II were compared between groups in a multivariate analysis.</p><p dir="ltr">In Study IV, all brain MRI examinations (n=256) were performed with the same protocol on the same 3 Tesla MRI machine. Using 3D T1-weighted images, brain volumetry was collected with FreeSurfer, and EPVS count, mean size, and volume was quantified with 3D computational framework with a morphological-based Frangi filter. Multivariable analysis was used to compare EPVS metrics between groups, and Spearmann for testing for correlations across disability level within groups.</p><p dir="ltr">For Study V, patients with acute myelitis or spinal infarction at Karolinska University Hospital were prospectively included within 30 days of onset and underwent neurophysiological examination of lower limb MEP and SEP. Fisher's exact test was used to compare MEP and SEP findings between groups.</p><p dir="ltr">Results: In Study I, the incidence of NMO/NMOSD in Sweden, per million person-years, increased from 0.3 between 1987 and 2006 to 0.8 between 2007 and 2013. The prevalence at the end of 2013 was 10.4 per million individuals. Most patients (78%) had relapsing disease with a median time to first relapse of 1.4 years.</p><p dir="ltr">Study II found that half of all patients relapsed on rituximab. Patients with MOGAD had 3 times higher incidence risk ratio (IRR) for relapses compared to AQP4+NMOSD patients. Conversely, the IRR for SIE was around 5 times higher in AQP4+NMOSD than MOGAD.</p><p dir="ltr">Study III found acute myelitis to have a stable incidence of 24.9 (95% CI 16.7-33.9) per million person-years in Stockholm between 2008 and 2018. Around 11% had poor function outcome and an increase in risk with age over 50 years, transverse spinal cord lesions, elevated count of polymorphonuclear cells, and elevated CSF/serum albumin ratio. Patients with idiopathic myelitis relapsed in 27% of cases, with increased risk related to findings of oligoclonal bands, transverse spinal cord lesions, and multifocal spinal cord lesions. MS was diagnosed at some point in 48% of acute myelitis patients.</p><p dir="ltr">Study IV found the highest EPVS burden in MOGAD patients, compared to healthy controls, followed by seronegative NMOSD and MS. The EPVS burden was driven by higher counts in MOGAD and larger EPVS mean size in MS. Disability level was not found to correlate with EPVS counts.</p><p dir="ltr">Study V found that MEP abnormalities, particularly non-responsive MEP, was more frequent in SCI compared to acute myelitis, and could discriminate SCI from acute myelitis with moderate sensitivity and high specificity. No difference in SEP abnormalities was found between the groups.</p><p dir="ltr">Conclusions: NMO/NMOSD incidence rates were similar to other predominantly Caucasian countries. The increase in incidence rates during the study period was probably related to increased access to AQP4-IgG analysis and increased awareness of the disease.</p><p dir="ltr">MOGAD patients are at substantial risk of relapses during rituximab treatment and more effective options are needed. Clinicians should be watchful for infections in AQP4+NMOSD patients.</p><p dir="ltr">Acute myelitis is a relatively common neurological condition with high risk of having an underlying chronic neuroinflammatory disease (particularly MS). A thorough clinical workup and follow-up are recommended to quickly identify those at risk.</p><p dir="ltr">EPVS metrics varied over CNS inflammatory diseases suggesting distinct characteristics and demonstrate the utility of automatic EPVS quantification.</p><p dir="ltr">MEP show promise as a supplementary tool when differentiating SCI from acute myelitis but larger studies are needed to validate the findings and further develop the method.</p><h3>List of scientific papers</h3><p dir="ltr">I. <b>Jonsson DI</b>*, Sveinsson O*, Hakim R, Brundin L. Epidemiology of NMOSD in Sweden from 1987 to 2013: A nationwide population- based study. Neurology. 2019 Jul 9;93(2):e181-e189. <a href="https://doi.org/10.1212/WNL.0000000000007746" rel="noreferrer" target="_blank">https://doi.org/10.1212/WNL.0000000000007746</a></p><p dir="ltr">II. Carlsson O, <b>Jonsson DI,</b> Brundin L, Iacobaeus E. Relapses and Serious Infections in Patients with Neuromyelitis Optica Spectrum Disorder Treated with Rituximab: A Swedish Single-Center Study. J Clin Med. 2024 Jan 8;13(2):355. <a href="https://doi.org/10.3390/jcm13020355" rel="noreferrer" target="_blank">https://doi.org/10.3390/jcm13020355</a></p><p dir="ltr">III. <b>Jonsson DI</b>, Sveinsson O, Moeini N, Pivac E, Wirdefeldt K, Brundin L, Iacobaeus E. Incidence, Etiology, and Long-Term Outcome of Acute Myelitis in Stockholm County, Sweden: A Population-Based Study. Neurol Neuroimmunol Neuroinflamm. 2025 Nov;12(6):e200472. <a href="https://doi.org/10.1212/NXI.0000000000200472" rel="noreferrer" target="_blank">https://doi.org/10.1212/NXI.0000000000200472</a></p><p dir="ltr">IV. <b>Jonsson DI</b>, Ouellette R, Hagbohm C, Park D, Coello RD, Ganeshan AR, Chamyani N, Ineichen BV, Evertsson B, Valdes Hernandez M, Wardlaw J, Jagodic M, Kockum I, Brundin L, Fink K, Piehl F, Iacobaeus E*, Granberg T*. Enlarged perivascular spaces differ across MOGAD, NMOSD, and MS. *Shared last authorship. [Manuscript]</p><p dir="ltr">V. <b>Jonsson DI</b>, Gripenland J, Iacobaeus E. The Use of Combined Sensory-Evoked Potentials and Motor-Evoked Potentials for Differential Diagnostics of Spinal Infarction and Acute Myelitis. - A Single Center Pilot Study. [Manuscript]</p>