MOG antibody–associated encephalomyelitis mimicking bacterial meningomyelitis following ChAdOx1 nCoV-19 vaccination: a case report

Abstract

We report a case of anti-myelin oligodendrocyte glycoprotein (MOG) antibody-associated encephalomyelitis following vector-based vaccination against SARS-CoV-2 that mimicked bacterial meningomyelitis upon initial presentation. A 43-year-old woman who had received a first dose of ChAdOx1 nCoV-19 (Vaxzevria; Astra Zeneca, UK Limited) 9 days earlier presented with subacute sensorimotor paraparesis, urinary retention, headache, meningism, and fever. Clinical findings and cerebrospinal fluid (CSF) features were highly suggestive of bacterial infection; however, despite receiving broad anti-infective treatment alongside with high-dose glucocorticoids, symptoms deteriorated. Imaging findings and the detection of immunoglobulin G against MOG substantiated diagnosis of an anti-MOG associated disorder. Treatment with high-dose intravenous (IV) methylprednisolone and plasma exchange resulted in substantial clinical improvement, which sustained under monthly regimen of IV Tocilizumab at 3-month follow-up. Awareness of this post-vaccinal presentation of a rare autoimmune disorder is important to not miss potential treatment options.

Introduction

Immunoglobin G (IgG) antibodies against myelin oligodendrocyte glycoprotein (MOG) are associated with autoimmune inflammatory conditions of the central nervous system (CNS). In neuromyelitis optica spectrum disorder (NMOSD), up to 42% of aquaporin 4 (AQP4) antibody–negative patients harbor antibodies against MOG. ¹ However, the clinical spectrum of MOG antibody–associated disorders (MOGAD) is broad and goes beyond the phenotype of classical NMOSD with its predominant affection of the optic nerves and spinal cord. Encephalitic presentations involving the supra- and/or the infratentorial brain have been described in adults as well as in pediatric populations, where up to 58% of children with acute disseminated encephalomyelitis (ADEM) are seropositive for MOG IgG.^2–4

As in AQP4-positive NMOSD, onset of MOGAD is preceded by acute infection or vaccination in some cases. ⁵ With regard to the ongoing global pandemic, several cases of MOGAD following SARS-CoV-2 infection have been reported.^6–8 There also have been sporadic cases of inflammatory CNS disorders after vaccination against SARS-CoV-2, including a seronegative NMOSD-like presentation in a patient with longstanding stable multiple sclerosis (MS) after a vector-based vaccine, as well as a case of AQP4-positive NMOSD following administration of an inactivated virus vaccine.^9,10 However, cases of MOGAD following COVID-19 vaccination have not been described in the literature so far.

Here, we present a case of MOG antibody–associated encephalomyelitis following vaccination with ChAdOx1 nCoV-19 with unusual cerebrospinal fluid (CSF) features that mimicked bacterial meningomyelitis upon initial presentation.

Case description

A 43-year-old woman presented to our hospital with sensorimotor paraparesis, urinary retention, and headache for 24 hours. Nine days earlier, she had received a first dose of ChAdOx1 nCoV-19 (Vaxzevria; Astra Zeneca), a vector-based vaccine against SARS-CoV-2. Medical history was unremarkable except for migraine. On admission, the patient had mild paraparesis, hyperreflexia, bilateral positive Babinski sign, a thoracic sensory level (T10), meningism, and fever of 38°C, scoring 5.0 on the Expanded Disability Status Scale (EDSS). Spinal magnetic resonance imaging (MRI) revealed T2 hyperintense lesions involving C6 to T1 as well as T3 and T4, consistent with transverse myelitis (Figure 1(a1)). Despite sporadic T2 hyperintense foci that were interpreted as unspecific, MRI of the brain did not show any abnormalities (Figure 1(a2) and (a3)). Initial laboratory tests showed a white blood cell count of 11.52/nl, and serum C-reactive protein and procalcitonin were not elevated. CSF analysis revealed extensive predominant granulocytic pleocytosis of 545 cells/μl (Figure 2), elevated lactate and CSF protein (4.4 mmol/l and 135 mg/dl, respectively), as well as a reduced CSF to serum glucose ratio (Table 1). Oligoclonal bands were negative and no other immunoglobin abnormalities were detected. An autoimmune disorder was suspected; bacterial meningomyelitis, however, was considered the main differential diagnosis. Broad empiric treatment with 1000 mg of intravenous (IV) methylprednisolone (IVMP) daily alongside with administration of ceftriaxone and ampicillin was initiated. Furthermore, plasma exchange (PLEX) was planned. During the next 5 days, this regimen of antibiotics, cumulative administration of 5 g of IVMP and one session of PLEX resulted in slight clinical improvement. However, on the fifth day after initiation of treatment, the patient developed a stuporous to comatose state with fever over 40°C as well as sensorimotor tetraparesis (EDSS 9.0) requiring monitoring on our intensive care unit (ICU). On a second CSF examination, further increase in granulocytic pleocytosis (720 cells/μl) was detected. The anti-infective medication was escalated to meropenem, and extensive infectious evaluation was performed, including next generation sequencing (NGS) for over 1500 pathogens. All results were negative except for an increased CSF-concentration of Cutibacterium acnes DNA upon initial evaluation (in 912 of 1824 reads), which was not confirmed in a follow-up CSF examination and interpreted as contamination. Early follow-up MRI of the brain showed new T2 hyperintense lesions involving frontal cortex, periventricular space, pulvinar thalamic nuclei, brain stem, and cerebellar peduncles (Figure 1(b2) and (b3)). Spinal lesions were progressive with additional involvement of C3 to C5 (Figure 1(b1)). Accordingly, the anti-infective therapy was discontinued. The diagnostic workup resulted in the discovery of antibodies against MOG in CSF and serum with titers of 1:32 and 1:1000, respectively. Furthermore, serologic testing revealed a slightly increased antinuclear antibody titer of 1:320; however, anti-extractable nuclear antigens tested negative. IgG antibodies against AQP4, glial fibrillary acidic protein (GFAP), N-methyl-d-aspartate (NMDA) receptor, or γ-aminobutyric acid receptor B (GABA B) were neither detectable in CSF nor serum.

Figure 1.

MRI upon admission (a), early follow-up 5 days after admission (b), and follow-up at 3 months (c). Initial sagittal T2-weighted spinal images with hyperintense lesions extending from C6 to T1 as well as T3 and T4 (a1) and no abnormalities on axial fluid-attenuated inversion recovery images of the brain (a2, a3). MRI at 5-day follow-up showing progressive spinal lesions with additional involvement of c3 to c5 (b1) and new hyperintense lesions of the subcortical white matter (b2) and bilateral pulvinar (b3). Partial resolution of former findings in cervical spine (c1), subcortical white matter (c2), and bilateral pulvinar (c3) at 3 months.

Figure 2.

A high-grade granulocyte-dominated granulo-lymphomonocytic pleocytosis is observed. The cell number increase can be well assessed in (a). In (b), neutrophilic granulocytes (exemplary marking with arrows), lymphocytes (exemplary marking with arrowheads), and monocytes (exemplary marking with asterisks) can be well differentiated in the higher magnification. Scale bars: (a) = 100 μm, (b) = 5 0μm.

Table 1.

CSF characteristics and serum MOG antibody titers at initial presentation, early follow-up 5 days after admission, and follow-up at 3 months.

	Initial presentation	Follow-up at 5 days	Follow-up at 3 months	Reference value
CSF WCC (cells/µl)	545	720	27	<5
CSF lactate (mmol/l)	4.4	3.6	1.8	1.2–2.1
CSF total protein (mg/dl)	135	61	50	15–45
CSF/serum glucose	0.5	0.6	0.6	>0.5
CSF MOG-IgG-Ab titer	1:32	1:10	1:3.2	N/A
Serum MOG-IgG-Ab titer	1:1000	1:320	1:320	<1:10

Ab, antibody; CSF, cerebrospinal fluid; IgG, immunoglobin G; MOG, myelin oligodendrocyte glycoprotein; WCC, white cell count.

A diagnosis of MOGAD was made and in addition to further administration of IVMP, treatment with PLEX was continued. Fever remitted soon afterward and neurological symptoms improved. In total, the patient received a cumulative dose of 11 g of IVMP and 7 sessions of PLEX in the hyperacute phase. Nonetheless, symptoms occurred again during tapering of glucocorticoids. Due to the unfavorable course of the disease and to keep the option for further COVID-19 vaccination, a monthly regimen of 400 mg IV Tocilizumab was started in addition to treatment with oral prednisone, and the patient was discharged to a rehabilitation facility. At 3-month follow-up, the patient showed sustained clinical improvement; however, light cerebellar syndrome with dominating intention tremor of the right hand persisted (EDSS 2.0). MRI of the brain revealed partial resolution of findings (Figure 1(c1)–(c3)), and antibodies against MOG still remained positive in CSF and serum (titers 1:3.2 and 1:320, respectively). Notably, the subject exhibited a rather low anti-SARS-CoV-2 serum titer of 75.7 BAU/ml 2 weeks after vaccination. After clinical recovery and under temporary intensification of glucocorticoid treatment (100 mg of prednisone daily for five days), the patient received a second COVID-19 vaccination, this time with an mRNA-based vaccine. As a result, the antibody titer increased up to >2080 BAU/ml, without any relapse of symptoms.

Discussion

To our knowledge, this is the first report on MOG antibody–associated encephalomyelitis following a vector-based vaccine against SARS-CoV-2, and there are several interesting aspects to the presented case: Clinical findings and CSF characteristics were highly suggestive of bacterial meningomyelitis, including meningism, high fever, granulocytic CSF pleocytosis of up to 720 cells/μl, and increased CSF lactate of 4.4 mmol/l. Although Jarius et al. ¹¹ found that granulocytic pleocytosis and increased lactate levels are common in acute MOG antibody–associated encephalomyelitis, the authors emphasized that elevation of these parameters usually occurs at much lower levels as compared with bacterial meningitis. In their large study on CSF findings in MOG antibody–associated encephalomyelitis, less than 2% of cases exceeded a CSF cell count > 300 cells/μl or a lactate level > 4 mmol/l, respectively. Nonetheless, rare cases of NMOSD mimicking bacterial infection on CSF evaluation have been reported.^12,13 In addition, fever has also been described as a clinical feature of some MOGAD phenotypes. ¹⁴

Imaging findings included an ADEM-like pattern of poorly demarcated, fluffy T2-hyperintensities and longitudinal extensive transverse myelitis (LETM), both common radiological features of MOG antibody–associated encephalomyelitis.^5,15 However, these findings developed with delay as compared with the clinical course. Of further note was the presence of a pulvinar sign, a finding that is mostly recognized as a feature of variant Creutzfeldt-Jakob-Disease (vCJD) ¹⁶ but has been observed in cases of MOG antibody–associated encephalomyelitis as well.^3,17

In the past, some authors have characterized MOGAD as a monophasic, steroid-sensitive condition bearing a rather favorable prognosis.^18,19 Some of these findings have been revised by larger case series with longer follow-up that revealed frequent relapses and severe disease courses with reduced steroid responsiveness.^5,20 Our patient did partially respond to high-dose IVMP and some symptoms occurred again when glucocorticoids were tapered, which is a common issue in MOGAD.^15,21 Nonetheless, there are no established guidelines for long-term immunosuppressive treatment and follow-up data are scarce. Tocilizumab could be a promising long-term treatment option in patients with severe, relapsing MOGAD, and a growing number of favorable treatment responses is being reported.^22–25 In particular during the COVID-19 pandemic, an immunomodulation that does not hamper antibody response to vaccination is of advantage. ²⁶ Despite being used almost exclusively as a second- or third-line treatment in MOGAD, the presented case highlights the potential use of Tocilizumab as a safe, effective, and well-tolerated first-line, long-term immunosuppressive treatment option in patients with severe relapsing disease, which has been discussed by other authors as well. ²⁷ Large-scale prospective data would be desirable to provide better guidance on treatment decisions.

Regarding the general safety profile of COVID-19 vaccines in patients with neuroimmunological disorders, observational studies did not find an increased risk for relapse.^28–30 Interestingly, the majority of patients in these studies received disease-modifying treatments by the time they were vaccinated, as was the case in our patient when receiving the second mRNA vaccine. These findings suggest a potential relationship between immunotherapy and vaccine tolerability in these patients; however, the underlying mechanisms remain unclear. Nonetheless, considering the potential harm of infection with SARS-CoV-2, patients with neuroimmunological disorders should be encouraged to undergo vaccination.

Conclusion

The present case expands the spectrum of MOG antibody–associated encephalomyelitis to coronavirus vaccinations, and clinicians should be aware of atypical presentations that mimic bacterial infection. Since the individual disease course can be disabling, early diagnosis is of great importance. Tocilizumab might be an option in patients with MOGAD unresponsive to steroids. The use of mRNA-based SARS-CoV-2 vaccines may be considered in patients with vector vaccine–associated MOG antibody–associated encephalomyelitis.