A comparative analysis of demographic, clinical and imaging features of myelin oligodendrocyte glycoprotein antibody positive, aquaporin 4 antibody positive, and double seronegative demyelinating disorders – An Indian tertiary care center prospective study

Inclusion criteria

Patients presenting with a clinical syndrome compatible with demyelination such as optic neuritis, transverse myelitis, encephalitis, and brainstem encephalitis that were either positive for anti-MOG antibody or anti aquaporin 4 antibodies or were negative for both antibodies, that is, seronegative were included in the study.

Exclusion criteria

Patients who had alternative causes such as sarcoidosis, neoplastic, paraneoplastic, infectious etiology, and those with a diagnosis of multiple sclerosis were excluded from the study.

Acute disseminated encephalomyelitis (ADEM) was diagnosed by International Paediatric Multiple Sclerosis Study Group criteria: Revisions to the 2013 definitions.^[2] Encephalitis was defined according to international criteria for inflammatory or infectious encephalitis.^[3] ADEM-like lesions were defined as diffuse, poorly demarcated, and large (>1–2 cm) lesions involving predominantly the cerebral white matter and deep grey matter (e.g., thalamus or basal ganglia). Tumefactive demyelinating lesion (TDL) was defined as a T2-weighted magnetic resonance imaging (MRI) “margin to margin” size of 2 cm or more.^[4] We defined chronic relapsing inflammatory optic neuritis (CRION) as per the criteria given by Petzold and Plant.^[5]

Data collection

All patients were evaluated for demographic and clinical data. Disability was evaluated by Expanded Disability Status Scale (EDSS) and visual acuity in case of ON. Cerebrospinal fluid examination was done to look for infectious causes. Neuroimaging data were collected by doing MRI of the brain, optic nerve, and spinal cord using the Phillips Enginia and GE machine. All serum samples were analyzed to detect AQP4 and MOG antibodies. Cell based immunoassay using transfected cell lines for in vitro determination in serum/plasma for quantitative determination of human IgG antibodies to aquaporin 4 receptors and antibody to MOG protein were used.

Eventually all the patients were divided in three groups

1. MOG antibody positive

2. AQP4 antibody positive

3. MOG/AQP4 antibody negative i.e. Seronegative.

Finally, we performed only pairwise comparisons (MOG vs. AQP4 or MOG vs. seronegative).

Statistical analysis

Central tendencies were expressed as mean or median. The demographic, clinical, laboratory, and outcome parameters between the three groups, MOGAD, AQP4-IgG+NMOSD, and AQP4/MOG antibody negative, that is, seronegative were compared using Fisher’s exact test for categorical variables and independent t-test or Mann–Whitney U test (whichever appropriate) for continuous variables. The variable with P value below 0.05 was considered significant. The statistical analysis was done using SPSS version 20 software.

Demographic data

A total of 103 patients were included in our study – 41 anti-MOG antibody positive, 37 AQP4 antibody-positive, and 25 patients were negative for both antibodies. Pediatric age group (<18 years) patients were more common in the MOGAD (36.5%). The predilection for female sex was higher in the AQP4 antibody group as compared to MOGAD (81% vs. 41%; P = 0.0004). Median follow-up duration was similar across three groups. Relapsing course of the disease was most commonly seen in the AQP4-IgG+NMOSD (86.5%) followed by MOG antibody cohort (56%) [Table 1].

Clinical presentation

At presentation, myelitis was seen more frequently in AQP4-IgG+NMOSD as compared to MOGAD (81% vs. 39%; P = 0.0002, Table 1). Bilateral ON (38 events in 18 patients) was the most frequent phenotype in the MOGAD, whereas myelitis was the most frequent phenotype in the AQP4-IgG+NMOSD (54 events in 30 patients) and in the seronegative group (15 events in 13 patients). Encephalitis was seen significantly more in the MOGAD as compared to the AQP4-IgG+NMOSD (11 vs. 0; P = 0.001) [Table 1]. None of the patients had area postrema syndrome in the MOGAD as well as in seronegative group while it was present in 24% patients of AQP4-IgG+NMOSD. Seven patients had recurrent isolated ON in patients with MOGAD. These patients had normal imaging apart from optic nerve involvement. We diagnosed these patients as chronic relapsing inflammatory optic neuritis (CRION). We could report one patient in seronegative group as CRION.

The proportion of patients with seizures was higher in MOGAD (15%) as compared to AQP4-IgG+NMOSD although it was not statistically significant (6 vs. 1; P = 0.1) [Table 1]. The median age of patients who had seizures was 12 years. Paroxysmal tonic spasms were seen only in the AQP4-IgG+NMOSD.

Neuroimaging

ADEM like lesions (7 vs. 1; P = 0.059), cortical (9 vs. 0; P = 0.003) and juxta-cortical (9 vs. 2; P = 0.03) brain lesions were more often seen in the MOGAD [Table 2, Figures 1 and 2]. Periventricular lesions were more common in the AQP4-IgG+NMOSD (11 vs. 20; P = 0.014) [Table 2]. Anterior segment ON was more common in the MOGAD (24 vs. 10; P = 0.005) while posterior segment ON (9 vs. 19; P = 0.01) and chiasmal involvement (4 vs. 13; P = 0.007) was more frequent with AQP4-IgG+NMOSD. Optic sheath enhancement was observed more often in the MOGAD (7 vs. 1; P = 0.059). All myelitis in the AQP4-IgG+NMOSD were longitudinally extensive myelitis (LETM). Cervical and dorsal segment involvement was seen in equal proportion in the MOGAD and AQP4-IgG+NMOSD. Conus involvement was more prevalent in patients with MOGAD myelitis (29%) as compared to AQP4-IgG+NMOSD (3%) (P = 0.02) [Figure 2]. A total of five patients (MOGAD – four and seronegative cohort – one) had TDL. None of the patients in the AQP4-IgG+NMOSD had TDL. All patients with tumefactive demyelination had a relapsing illness. All the six patients with seizures in the MOGAD had FLAIR cortical hyperintensity along with fever at the time of onset with or without altered sensorium. Cortical lesions and FLAIR hyperintense lesions in anti-MOG associated encephalitis with seizures (FLAMES) were seen exclusively in younger patients. Mean age of patients with FLAMES was 11 years.

Close

Figure 1:

Magnetic resonance imaging images showing leukodystrophy like picture in a case of positive anti-myelin oligodendrocyte glycoprotein antibody – Axial (a-c) and coronal (d) FLAIR sections show multifocal confluent white matter hyperintensity in bilateral middle cerebellar peduncle (a), left cerebellar hemisphere (a), left peritrigonal white matter (b), and along periventricular and subcortical white matter of left posterior parietal lobe (c and d). Post-contrast axial (e) and coronal (f) sections show foci of nodular enhancement in the left parietal lobe.

Export to PPT

Close

Figure 2:

MRI Brain of anti- MOG antibody positive patient (a-d) and spine (e-g) images shows multifocal variable sized FLAIR hyperintense white matter lesions (a-c) in supra and infratentorial compartments with no obvious enhancement on post contrast section (d). Long segment central (H-shaped) intramedullary hyperintensity is seen in cervical-dorsal spinal cord (e-g) extending up to the conus (longitudinally extensive transverse myelitis).

Export to PPT

Treatment and outcome

All patients across the three groups received intravenous methylprednisolone at the time of an acute attack. Plasma exchange or intravenous immunoglobulin (IVIg) were used as rescue therapy if the treatment response to steroid was not satisfactory. Both plasma exchange therapy and IVIg were administered in significantly higher proportions in the AQP4-IgG+NMOSD. Azathioprine was the most common initial steroid sparing immunosuppressive agent used in all the three groups. Use of Rituximab as a steroid-sparing agent was highest in the AQP4-IgG+NMOSD. Escalation therapy with rituximab was required most commonly in the AQP4-IgG+NMOSD (12/31, 38.7%) as compared to the MOGAD (4/38, 10.5%). Ten patients in the AQP4-IgG+NMOSD had changed to injection rituximab because of relapse on the previous therapy (nine patients on azathioprine and one patient on mycophenolate mofetil). All four patients in the MOGAD were changed to injection rituximab therapy because of relapse on azathioprine therapy. About 40% patients in the seronegative group were not given any long-term immunosuppression following a short course steroid therapy and rescue therapy if required. There was no difference in the nadir EDSS across the groups. Improvement following discharge (5 vs. 6; P = 0.4) and at 3 month follow-up was also similar (1 vs. 3.75; P = 0.15). Median EDSS at the last follow-up was better in the MOGAD than AQP4-IgG+NMOSD (1 vs. 3.5; P = 0.03) [Table 2]. About 81% percent patients had vision <3/60 at nadir in our study (85% in NMO group and 84% in MOG group) [Table 2]. There was no difference in visual acuity between the groups at discharge, 3-month follow-up, and at last follow-up.

Limitations

Our study has a few limitations. First, it carries the inherent limitations of all the observational studies. Second, some of the data regarding initial presentation was analyzed retrospectively and may have led to recall bias. Third, we could not assess the long-term outcome of all patients because of the ongoing COVID-19 pandemic.