key: cord-0802403-yc0tmsl5
authors: Ismail, Ismail Ibrahim; Salama, Sara
title: A systematic review of cases of CNS demyelination following COVID-19 vaccination
date: 2021-11-09
journal: J Neuroimmunol
DOI: 10.1016/j.jneuroim.2021.577765
sha: bd2e503f38e8511074a77af27522dadb3c204c40
doc_id: 802403
cord_uid: yc0tmsl5

BACKGROUND: Since the emergency use approval of different types of COVID-19 vaccines, several safety concerns have been raised regarding its early and delayed impact on the nervous system. OBJECTIVE: This study aims to systematically review the reported cases of CNS demyelination in association with COVID-19 vaccination, which has not been performed, to our knowledge. METHODS: A systematic review was performed by screening published articles and preprint of cases of CNS demyelination in association with COVID-19 vaccines in PubMed, SCOPUS, EMBASE, Google Scholar, Ovid and medRxiv databases, until September 30, 2021. This study followed PRISMA guidelines. Descriptive findings of reported cases were reviewed and stratified by demographic and clinical findings, diagnostic work-up, management, and overall outcome. RESULTS: A total of 32 cases were identified, with female predominance (68.8%) and median age of 44 years. Eleven cases were reported after Pfizer vaccine, 8 following AstraZeneca vaccine, 6 following Moderna, 5 following Sinovac/ Sinopharm vaccines, and one following each of Sputnik and J&J vaccines. The majority of cases (71.8%) occurred after the first dose of the vaccine, with neurological symptoms manifesting after a median of 9 days. The most common reported presentations were transverse myelitis (12/32) and MS-like pictures (first diagnosis or a relapse) in another 12/32 cases, followed by ADEM- like (5/32), and NMOSD- like (3/32) presentations. History of a previous immune-mediated disease was reported in 17/32 (53.1%) cases. The mRNA-based vaccines resulted in the greatest number of demyelinating syndromes (17/32), followed by viral vector vaccines (10/32), and inactivated vaccines (5/32). Most MS-like episodes (9/12) were triggered by mRNA-based vaccines, while TM occurred following both viral vector and mRNA-based vaccines. Management included high dose methylprednisolone, PLEX, IVIg, or a combination of those, with a favorable outcome in the majority of case; marked/complete improvement (25/32) or stabilized/ partial recovery in the remaining cases. CONCLUSION: This systematic review identified few cases of CNS demyelination following all types of approved COVID-19 vaccines so far. Clinical presentation was heterogenous, mainly following the first dose, however, half of the reported cases had history of immune-mediated disease. Favorable outcome was observed in most cases. We suggest long-term post-marketing surveillance for these cases, to assess for causality, and ensure the safety of COVID-19 vaccines.

A summary of the clinical characteristics is presented in Table 1 .

When looking at the clinical pictures, the most common presentations were transverse myelitis (12/32), and MS-like (first diagnosis or relapse) pictures (12/32), followed by ADEM-like (5/32), and NMOSD-like (3/18) presentations.

Transverse myelitis presentation was reported in 12 patients, 6 males and 6 females, with a median age of 44.5 years (36-78 years). There was a median interval of 6.5 days On spinal MRIs, simultaneous thoracic and cervical spinal cord involvement was reported in 6/12 cases, followed by isolated thoracic cord affection in 5/12 patients. On the other hand, isolated J o u r n a l P r e -p r o o f cervical cord was noted in 2 patients, while conus medullaris involvement was reported in only one patient. Longitudinally extensive transverse myelitis (LETM) was the main radiological feature in 7/12 patients, while short segment involvement was reported in 5/12. In patient 8, conus pathology was part of a lesion involving the thoracic cord, however, serological testing for AQP4 and MOG antibodies was negative. Brain MRI was abnormal in only two of the myelitis patients with nonspecific hyperintensities showing neither restriction nor contrast enhancement. Generalized tonic-clonic seizures (GTCS) were reported in 3/5 patients, while confusion and headache were seen in 2/5 patients. Serological testing for AQP4 and MOG antibodies was negative in 4 patients and was not tested in the fifth. CSF analysis showed pleocytosis in 3/5 patients (Patient J o u r n a l P r e -p r o o f 2, 4 and 5), while elevated protein was reported in only one patient (0.75g/L). Moreover, OCBs were tested in 4/5 patients and were negative.

All reported MRI lesions were supratentorial. Some lesions showed contrast enhancement (Patient 5).

Additionally, spinal cord was involved in the form of multifocal lesions in one case (Patient 5).

NMOSD-like presentation was reported in 3 cases. One of them (Patient 30) was previously diagnosed with MS, and was on natalizumab for 21 years, when she suffered simultaneous optic neuritis and LETM. Even though she tested negative for AQP4 antibody in serum and CSF, her imaging features were more in favor of NMOSD rather than MS. Moreover, her CSF showed severe pleocytosis (542 cells/μl) with neutrophils predominance and markedly elevated protein (2.2 gm/L).

Unfortunately, testing for MOG antibody and OCBs in CSF was not reported. The other patient (Patient 31) presented by the characteristic area postrema syndrome that was supported by MRI data and positive serology for AQP4 antibodies. CSF analysis had elevated white blood cells count of (31 cells/μl), with mononuclear cells as the predominant type. The third patient (Patient 32) was a 64year-old male with a history of Sjogren's disease. His presentation suggested spinal cord involvement which was supported radiologically by extensive cord involvement from cervical region to conus.

Serological testing was positive for AQP4 antibodies. Brain MRI showed demyelinating patches as well.

When addressing the management plans, treatment comprised high dose methylprednisolone (with or without oral tapering), PLEX, IVIG, or a combination of those. Antiviral/ antibiotics were also given in few cases. Fortunately, interventions achieved either marked/complete improvement in 25/32 (78.1%) cases, stabilized or resulted in partial recovery in the rest.

J o u r n a l P r e -p r o o f Finally, to further characterize the cases, we classified them based on the received vaccine type (Table 2) . CNS demyelination was reported after all types of vaccines in the literature. The mRNAbased vaccines resulted in the greatest number of demyelinating syndromes (17/32), followed by viral vector vaccines (10/32) as opposed to 5 cases following inactivated vaccines. Similarly, most MS-like episodes (9/12) were triggered by mRNA-based vaccines.

However, in TM cases, viral vector vaccines were associated with demyelination in 6/12 cases, followed by mRNA-based vaccines (5/12), and only one case following inactivated vaccines.

Furthermore, the median interval between receiving inactivated COVID-19 vaccines and symptom development was longer; 14 days, as opposed to 7.5 and 6 days for viral vector and mRNA vaccines, respectively.

Data regarding the second dose of vaccination, for those who developed symptoms after the first dose, was available for only 9/32 cases; 4 cases did not adhere to the vaccination schedule, 3 cases switched from viral vector to an mRNA vaccine, and 2 cases received the second dose without any new symptoms.

In order to be defined as vaccine-induced, the WHO had suggested certain criteria to be met 11 , including: (1) temporal relationship (vaccination must precede the occurrence of the event), (2) consistency of evidence (similar or same results generated by studies using different methods in different settings), (3) strength of association (statistical significance to demonstrate that it was not a chance occurrence), (4) specificity (vaccine is the only cause of the event), and (5) biological plausibility and coherence (there must be a biologically plausible mechanism between cause and effect). Such strict criteria were rarely met in the majority of cases in the literature, similar to our review, making inference of causation a challenge. In literature, few post-vaccination autoimmune diseases were firmly and reliably considered as vaccine-associated, such as GBS cases following 1976 swine influenza vaccine. However, other suspected associations, such as the hepatitis B vaccine and MS, and HPV vaccine and ADEM, have not been strongly confirmed 12 The concern of a potential association of CNS autoimmune inflammation and COVID-19 vaccines has been recently raised, as cases began to unfold. According to the interim analysis of four randomized controlled trials of ChAdOx1 nCoV-19 vaccine (AZD1222), 3 cases of ATM were reported, from 11, 636 participants included. One case was considered to be an idiopathic demyelination that is possibly related to the vaccine, while the other two were J o u r n a l P r e -p r o o f most likely a pre-existing or previously unrecognized MS 14 . However, in a recent review of 11 COVID-19 vaccine candidates 6 , the preliminary official data from the vaccine manufactures and the drug authorities suggested that neurologic adverse events were rare, and cases of CNS demyelination were reported in association with viral vector vaccine only.

In our review, CNS demyelination was reported following all types of approved COVID-19 vaccines (no protein-based vaccine was approved at the time of writing). Neurological symptoms appeared within the first 1-2 weeks in most cases. Females comprised the majority of cases, which agrees with data in literature, where around 85% of immune-mediated diseases affects women 15 . This has been attributed to greater immune responses against foreign and self-antigens in women compared to men.

Furthermore, more than half of the cases had history of probable or definite autoimmune diseases, which could make them liable to increased risk of developing other immune-mediated diseases 16 .

The mRNA-based vaccines resulted in the greatest number of demyelinating syndromes (53.1%), followed by viral vector vaccines (31.2%), as opposed to 5 cases (15.6 %) following inactivated vaccines. Similarly, 75% of MS-like episodes were triggered by mRNA-based vaccines. However, it should be noted that more patients with immune-mediated diseases received mRNA-based vaccines compared to the other types combined (10 vs 7 cases, respectively). Furthermore, TM cases were associated with both viral vector and mRNA-based vaccines (50% vs 41.6%, respectively), contrary to earlier published data, which limited TM to viral vector vaccines only 6 .

As of September 2021, VAERS database had 328 reports of suspected cases of TM worldwide, following all types of vaccines 18 . However, a recent analysis of VAERS data revealed no increased risk of neuroautoimmune adverse events from COVID-19 vaccines compared to other vaccines 19 .

At the time of writing, 6.4 billion doses of COVID-19 vaccines have been given, and 2.7 billion were fully vaccinated. Considering the annual incidence of TM of 1.34 to 4.60/ million in general J o u r n a l P r e -p r o o f population, 24.6/ million in cases of acquired demyelination 20 , and 0.5/million in COVID-19 patients 21 , the current incidence of post-vaccination TM would be considered low.

Moreover, few cases of ADEM were reported, given that ADEM is the prototype and one of the most common white matter diseases associated with vaccines. ADEM is usually monophasic with a widely variable clinical presentation and favorable outcome, which was seen in our cases 9 .

Six MS patients had clinical relapses, while another 6 were newly diagnosed with MS following vaccination. Vaccines had long been incriminated in the development of MS, or in triggering MS relapses. However, pooled analysis from multiple studies found no sufficient evidence to support a causal relationship between the onset of MS and various common vaccinations 22 . Moreover, a 2017 systematic review of more than 50 articles 23 found no increased risk in developing MS and in relapses after vaccination.

As regards to COVID-19, Achiron and colleagues 24 found no increased risk of relapse activity in MS patients who received BNT162b2 vaccine as well. The relapse rate was higher following the first dose than the second dose (2.1% and 1.6%, respectively), which was similar to the rate in non-vaccinated patients during the corresponding period. However, in our review, MS relapse rate occurred following first and second doses equally. Moreover, they found that relapse rate was slightly higher in younger patients and in those treated with immunomodulatory drugs. However, their review was limited to cases who received BNT162b2 COVID-19 vaccine only, with a relatively short follow-up period.

Interestingly, NMOSD-like presentation was reported in 3 cases, raising the possibility of crossreactivity between the used viral antigens and aquaporin-4. This predisposition to the spinal cord and the optic nerves has also been reported following other vaccines (e.g. HPV) vaccine 25 .

A favorable outcome was noted in the majority of the reported cases, which was similar to the findings of a recent nationwide study 26 , where most patients with neurological complications experienced complete recovery within days to weeks without long-term sequalae.

The exact mechanism of demyelination after COVID-19 vaccines remains poorly understood, however, it is postulated that a combination of vaccine-related factors, in addition to susceptibility of the patients, could be involved. Molecular mimicry represents one of the main immunopathogenic factors, where similarity between the proteins of the viruses used for the vaccination and self-antigens (e.g. myelin) triggers an undesired immune-response 25 .

In ChAdOx1 vaccine (AZD1222); a SARS-CoV-2 structural surface vector glycoprotein antigen (spike protein; nCoV-19) gene is included in a replication-deficient chimpanzee adenovirus, which could be a possible trigger of demyelination. 17 Another important factor is the pathogenic role of immunologic adjuvants (substances that are used to enhance the antigen-specific immune responses), which can mimic evolutionarily conserved molecules activating both the innate and adaptive immune systems 27 . The mRNA vaccine exhibits a property of self-adjuvantation, where the mRNA acts as both antigen and adjuvant. A theoretical risk of inducing an autoimmune reaction could be related to activation of toll-like receptors TLR7 and TLR8, resulting in type I interferon production, and eliciting robust T and B cell responses, thus activating bystander autoreactive lymphocytes 28 . This bystander activation, along with macrophages J o u r n a l P r e -p r o o f secreting cytokines, can results in local inflammation and the recruitment of additional T-helper cells. 29 Other etiologies include vaccine-related factors such as the type, dose and the route of administration 28 , in addition to a possible immunological and genetic susceptibility of the patients. 

This systematic review has some limitations. Since all available literature were reported as single case reports and one case series so far, there could be some sort of reporting and/or publication bias. In addition, we presented the reports as a set of cases for practical reasons, however, one should be wary of interpreting the data as coming from a uniform cohort, and inferring direct causality from the anecdotal data provided. Moreover, the small number of reported cases, the heterogeneity of clinical data, or the incomplete work-up in some cases, hindered the ability to perform a meta-analysis.

Despite these shortcomings, the current review represents the first preliminary data on the association of COVID-19 vaccination and CNS demyelination, which can help future research. 

In this review, CNS demyelination occurred following all types of approved COVID-19 vaccines.

Clinical presentation was heterogenous, including TM, MS, ADEM and NMOSD. Symptoms occurred within 1-2 weeks, mainly following the first dose of vaccine. Interestingly, more than half of the reported cases had history of immune-mediated diseases. A favorable outcome was observed in the majority of cases after treatment. Currently, the world is facing the largest mass vaccination campaign in history, and cases of demyelination will inevitably occur, either directly following vaccination, or by chance. However, the incidence appears to be low, in comparison to demyelination following COVID-19 infection. Although association does not always imply causation, long-term post-marketing surveillance for cases of demyelination is warranted, to assess for causality, and ensure the safety of COVID-19 vaccines. 

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. J o u r n a l P r e -p r o o f J o u r n a l P r e -p r o o f

World Health Organisation: WHO coronavirus disease (COVID-19) dashboard

An overview of current COVID-19 vaccine platforms

COVID-19 vaccines. Current Opinion in Immunology

Safety of COVID-19 vaccines administered in the EU: Should we be concerned?

The potential neurological effect of the COVID-19 vaccines: A review

Covid-19: Oxford researchers halt vaccine trial while adverse reaction is investigated

Safety and effectiveness of SARS-CoV-2 vaccines: A systematic review and meta-analysis

The spectrum of post-vaccination inflammatory CNS demyelinating syndromes

Neurological side effects of COVID-19 vaccines are rare

Vaccination and autoimmune disease: what is the evidence? The Lancet

Could Autoimmunity Be Induced by Vaccination? International Reviews of Immunology

Transverse myelitis and vaccines: a multi-analysis

Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK. The Lancet

The Prevalence of Autoimmune Disorders in Women: A Narrative Review

Are Individuals With an Autoimmune Disease at Higher Risk of a Second Autoimmune Disorder?

Atlanata, GA: US Department of Health and Human Services, CDC; 2021

Acute Transverse Myelitis (ATM):Clinical Review of 43 Patients With COVID-19-Associated ATM and 3 Post-Vaccination ATM Serious Adverse Events With the ChAdOx1 nCoV-19 Vaccine (AZD1222)

Immunizations and risk of multiple sclerosis: systematic review and metaanalysis

Vaccines and multiple sclerosis: a systematic review

COVID-19 vaccination in patients with multiple sclerosis: What we have learnt by

The spectrum of post-vaccination inflammatory CNS demyelinating syndromes

Systemic and Neurologic Adverse Events Among 704,003 First-Dose Recipients of the Pfizer-Biontech (BNT162b2) mRNA COVID-19 Vaccine in Mexico

Oxford-AstraZeneca COVID-19 vaccine efficacy. The Lancet

Autoimmune/inflammatory syndrome induced by adjuvants (Shoenfeld's syndrome): clinical and immunological spectrum

SARS-CoV-2 vaccines and autoimmune diseases amidst the COVID-19 crisis

The good, the bad and the ugly of vaccination

Association of CNS demyelination and COVID-19 infection: an updated systematic review

FLAIR: fluid attenuated inversion recovery, DW: diffusion weighted, mRNA: messenger ribonucleic acid, IVMP: intravenous methyl prednisolone