Corpus Callosum Involvement in MOG Antibody-Associated Disease and Comparison to AQP4-IgG-Seropositive Neuromyelitis Optica Spectrum Disorder and Multiple Sclerosis

Abstract

Background

Data on corpus callosum involvement in myelin-oligodendrocyte-glycoprotein-antibody-associated disease (MOGAD) are limited.

Objective

To compare callosal lesions in MOGAD, multiple sclerosis (MS) and aquaporin-4-IgG-positive-neuromyelitis-optica-spectrum-disorder (AQP4+NMOSD).

Results

Callosal lesion frequency was similar in MOGAD (38/171[22%]), MS (24/72[33%]) and AQP4+NMOSD (18/63[29%]). Clinical phenotypes included: encephalopathy (47%), focal supratentorial (21%) or infratentorial deficits (45%). None had callosal-disconnection syndromes. Maximal callosal-T2-lesion diameter (median[range]) in mm was similar in MOGAD (21[4-77]) to AQP4+NMOSD (22[5-49];p=0.93) but greater than MS (10.5[2-64];p=0.04). Extracallosal extension (21/38[55%]) and T2-lesion resolution (19/34[56%]) favored MOGAD.

Conclusions

Despite similar frequency and imaging overlap, larger lesions, sagittal midline involvement and lesion resolution favored MOGAD.

Background

Myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) is a central nervous system (CNS) demyelinating disease distinct from multiple sclerosis (MS) and aquaporin-4-IgG-positive neuromyelitis optica spectrum disorder (AQP4+NMOSD).¹ Corpus callosum lesions are frequent in CNS demyelinating diseases, but detailed descriptions of callosal involvement in MOGAD and comparison to MS and AQP4+NMOSD are not available.

We aimed to analyze the frequency and imaging characteristics of callosal lesions in MOGAD and compare to MS and AQP4+NMOSD.

Methods

MOGAD patients with a compatible syndrome, and serum MOG-IgG1 positivity (testing methodology as described previously²) were identified from a Mayo Clinic database from 01-Jan-1996 to 31-Dec-2020. Medical records were reviewed for MRI brain scans within 6 weeks of new/worsening symptoms. Demographics, presentation details at the time of attack, and serum MOG-IgG1 titers were collected.

The comparative cohorts included consecutive patients from a Mayo Clinic AQP4+NMOSD database and from two of the author’s MS Clinics (N.H.C., E.P.F.). Included patients were required to meet the same radiological criteria as the MOGAD cohort. AQP4+NMOSD patients fulfilled 2015 NMOSD diagnostic criteria, while the MS patients met 2017 MS diagnostic criteria and additionally tested negative for AQP4-IgG and MOG-IgG1. There were a total of 38 patients (MOGAD 12, AQP4+NMOSD 12, MS 14) from this study included in a prior publication that focused only on lesion evolution and did not evaluate corpus callosum involvement.²

Scans across all three cohorts were performed on a range of different MRI scanners both at Mayo Clinic and from other institutions. MRI callosal lesions were confirmed by consensus of two authors (N.H.C., E.P.F.). Recorded characteristics included: section (genu, body, splenium), longitudinally-extensive (>2.5 cm) or midline involvement, maximal callosal T2-lesion diameter, callosal lesion number, and gadolinium enhancement. Lesions were subependymal (arising from ependymal surface) or had extracallosal extension (>50% lesion external to corpus callosum). Callosal lesion resolution at follow-up was assessed.

Continuous and categorical variables were reported as median (range) and number (%), respectively. Comparisons were performed with Wilcoxon rank-sum test for continuous and Fisher exact test for categorical variables. A p-value <0.05 was considered statistically significant (BlueSky statistics 7.4).

Standard protocol approvals, registrations, and patient consents

Mayo Clinic’s Institutional Review Board approved the study. All participants consented to use of their medical records for research.

Data availability

Anonymized data from this study will be made available on request.

Results

MOGAD callosal involvement

There were 171 MOGAD scans from 119 unique patients. Callosal T2-hyperintense lesions were identified in 38/171 (22.2%) MOGAD scans. In those with callosal involvement, median age was 17.5 years (range 2-56) and 52.6% were female. The median serum MOG-IgG1 titer was 1:100 (range 1:20-1:1000).

Most patients with callosal lesions had cerebral presentations including one or more of: focal deficits referable to other lesions in supratentorial (8/38 [21.1%]) or infratentorial (17/38 [44.7%]) brain, or encephalopathy (18 [47.4%]). Callosal disconnection syndromes were not identified. Isolated myelitis and optic neuritis were rare (Table 1). Callosal lesions were highly variable: details in Table 1 and MRI examples in Figure 1.

Table 1 –

Comparison of clinical and radiological features in patients with callosal lesions

	MOGAD N=38	AQP4+NMOSD N=18	MS N=24	p MOGAD vs AQP4+NMOSD	p MOGAD vs MS	p AQP4+NMOSD vs MS
Callosal lesion scans/all scans	38/171 (22.2%)	18/63 (28.6%)	24/72 (33.3%)	0.31	0.08	0.58
Median age (range)	17.5 (2-56)	48 (14-89)	31.5 (18-52)	<0.001	<0.001	0.003
Female	20 (52.6%)	16 (88.9%)	19 (79.2%)	0.009	0.058	−0.68
Clinical presentation	N=38	N=18	N=24
Cerebral	33 (86.8%)	10 (55.6%)	15 (62.5%)	0.04	0.03	1
Optic neuritis	13 (34.2%)	3 (16.7%)	5 (20.8%)	0.15	0.39	1
Isolated optic neuritis	3 (7.9%)	3 (16.7%)	4 (16.7%)	0.37	0.41	1
Myelitis	11 (28.9%)	12 (66.7%)	6 (25.0%)	0.01	0.78	0.01
Isolated myelitis	1 (2.6%)	4 (22.2%)	5 (20.8%)	0.03	0.03	1
Headache	17 (44.7%)	4 (22.2%)	1 (4.2%)	0.14	<0.001	0.15
Anatomical involvement
Genu	12 (31.6%)	12 (66.7%)	10 (41.7%)	0.02	0.43	0.13
Body	27 (71.1%)	14 (77.8%)	15 (62.5%)	0.75	0.58	0.33
Splenium	19 (50.0%)	9 (50.0%)	17 (70.8%)	1	0.12	0.21
Sagittal midline involvement	18 (47.4%)	13 (72.2%)	2 (8.3%)	0.09	0.002	<0.001
Longitudinally extensive	4 (10.5%)	6 (33.3%)	1 (4.2%)	0.06	0.64	0.03
Subependymal	13 (34.2%)	18 (100%)	19 (79.2%)	<0.001	<0.001	0.06
Extracallosal extension	21 (55.3%)	2 (11.1%)	7 (29.2%)	0.003	0.07	0.26
Lesion characteristics
Median diameter of largest T2-lesion in mm (range)	21 (4-77)	22 (5-49)	10.5 (2-64)	0.93	0.04	0.07
Median number of lesions (range)	1 (1-9)	2 (1-12)	1.5 (1-8)	0.13	0.73	0.31
Gadolinium enhancing	12/36 (33.3%)	8/17 (47.1%)	10/23 (43.5%)	0.38	0.58	1
Lesion resolution	N=34	N=13	N=23
Complete	19 (55.9%)	2 (15.4%)	3 (13.0%)	0.02	0.002	1
Median time to complete resolution scan; months (range)	4 (0-94)	19 (17-21)	11 (9-23)	0.05	0.04	0.8
Median interval to last follow-up scan	55 (2-146)	28 (7-141)	57 (6-271)	0.34	0.3	0.16

AQP4+NMOSD = aquaporin-4-IgG-positive neuromyelitis optica spectrum disorder; MOGAD = myelin oligodendrocyte glycoprotein antibody-associated disorder; MS = multiple sclerosis

Figure 1. Examples of MOGAD corpus callosum involvement and comparison of lesion evolution acutely and at follow-up.

Callosal T2-hyperintense lesions demonstrated by white arrows on FLAIR and T2-weighted brain MRI images; unfilled arrows point from acute to follow-up MRI. A: Diffuse splenium T2-hyperintensity. B: Genu T2-hyperintense lesion involving the septum pellucidum (enhancing; post-gadolinium image not shown). C: Callosal body T2-hyperintense lesion involving the midline and extending to involve the parasagittal white matter; right brainstem T2-hyperintense lesion also seen (arrowhead). D: Longitudinally-extensive T2-hyperintense lesion involving nearly the entire corpus callosum. E. Multiple large callosal T2-hyperintense lesions (extending to parasagittal white matter; not shown) and other subcortical/juxtacortical T2-hyperintense lesions (arrowheads). F: Large genu T2-hyperintense lesion; also pictured multiple other T2-hyperintense lesions including the right basal ganglia, left internal capsule, left insula and left occipital white matter (arrowheads). G: Extracallosal extension of T2-hyperintensity to bilateral parasagittal frontal white matter (arrowheads). H: Longitudinally-extensive T2-hyperintense lesion extending to the parasagittal frontoparietal white matter (arrowhead). Complete resolution of MOGAD T2-hyperintense lesions: midline genu at only 6 weeks (I) and heterogeneous T2-hyperintense lesion extending to the bilateral frontal cortex after 5 months (J). AQP4+NMOSD T2-hyperintense lesions with reduced size but persistence in the body at 54 months (K) and splenium at 39 months (L). MS T2-hyperintense lesions, persistent with minimal size reduction at follow-up in left splenium at 8 months (M) and callosal body at 14 months (N).

Comparison of callosal lesions in MOGAD, MS and AQP4+NMOSD

Callosal lesion frequency in MOGAD (38/171 [22.2%]), was similar in MS (24/72 [33.3%]; p=0.08) and AQP4+NMOSD (18/63 [28.6%]); p=0.31). MOGAD patients were younger; demographic, clinical and MRI features are compared in Table 1.

Maximal callosal T2-lesion diameter was larger in MOGAD than MS, while lesion number and enhancement rates were similar across groups (Table 1). Midline involvement was infrequent in MS while longitudinally-extensive involvement was most common in AQP4+NMOSD (Table 1). Subependymal lesions were more common in MS and universal in AQP4+NMOSD; extracallosal extension occurred less in AQP4+NMOSD (Table 1). A characteristic pattern of extracallosal extension to bilateral parasagittal frontal/parietal cortices was uniquely observed in 6/38 (15.8%) MOGAD cases (Figure 1). Complete callosal lesion resolution was more common in MOGAD (Table 1) with comparative examples in Figure 1.

Discussion

Callosal involvement occurred in 22% of MOGAD scans, similar to prior reports of 18-33%.^3–5 Most had cerebral symptoms and few had isolated optic neuritis or myelitis. Unlike in MS and AQP4+NMOSD,^{6, 7} MOGAD callosal lesion characterization and comparison has not been reported. We found MOGAD callosal lesions to be variable, but often large and involving the sagittal midline. Extracallosal extension was frequent; bilateral spread to parasagittal frontoparietal cortices may be a distinctive MOGAD pattern, with similar cases reported previously.⁸

Callosal lesion frequency was similar in all disorders, although others reported higher frequencies in MS.⁹ MOGAD callosal lesions were larger and more commonly involved the midline versus MS, while extracallosal extension was more common than in AQP4+NMOSD. MOGAD callosal lesion resolution was more frequent compared to MS and AQP4+NMOSD, as demonstrated previously.²

Limitations of this study include its retrospective nature and variable MRI timing and scanners. Moreover, while callosal clinical syndromes were not documented this may reflect real-life challenges of assessing young children or encephalopathic patients. Documentation of headache may be inconsistent and is not a typical symptom of CNS demyelinating disease, but headache was frequent in MOGAD potentially reflecting the high frequency of leptomeningeal involvement reported with some cerebral attacks.¹⁰ The younger median age than other MOGAD cohorts reflects a high proportion of pediatric patients predisposed to cerebral presentations.

In summary, we highlight clinical and radiologic features of corpus callosum involvement in MOGAD and identify discriminators from MS and AQP4+NMOSD that may be useful in clinical practice.