Introduction

In December 2019, the world witnessed an emergence of a novel coronavirus, severe acute respiratory syndrome coronavirus 2 (SARS-CoV 2). As per the latest World Health Organization report, there were 611,421,786 confirmed cases on September 24, 2022, along with 6,512,438 deaths [1]. Apart from mortality, this pandemic had a devastating impact on the economy and social life globally.

The world’s scientific community responded unprecedently and rapidly to this pandemic, by developing novel COVID-19 vaccines. Multiple randomized controlled trials observed that vaccines were the most effective tool to reduce mortality and morbidity during this pandemic [2,3,4]. There are four types of vaccines against COVID-19: nucleic acid (mRNA, DNA), protein-based (virus-like particle) vaccines, viral vector (replicating, non-replicating) and whole-virus (live attenuated, inactivated) vaccines [5]. Indian data as on September 24, 2022, reveal that a total of 1,02,61,06,833 of the Indian population have received at least the first dose, while 94,74,14,286 have completed the vaccination schedule of 2 doses and 20,08,20,471 have taken booster dose [6]. COVID-19 vaccines approved by the government of India are ChAdOx1-S (Recombinant, non-replicating adenovirus vector encoding the SARS-CoV-2 Spike glycoprotein) (Covishield/Vaxzervria, Astra-Zeneca), BBV152 (whole-virus inactivated) (Covaxin), Gam-COVID-Vac (recombinant replication-deficient adenovirus-based vaccine) (Sputnik V), RBD protein subunit COVID-19 vaccine (Corbevax) and SARS-CoV-2 rS (Recombinant spike protein nanoparticle vaccine) (Covovax) [6].

Neurological adverse events after immunization (AEFI) like cerebral venous sinus thrombosis, Bell’s palsy and Guillain–Barré syndrome (GBS) have been reported in clinical trials and real-world settings post-COVID-19 vaccination [2, 7, 8]. Central nervous system inflammatory demyelinating diseases (CNS-IDD) were also reported in clinical trials. Two cases of transverse myelitis in the vaccine group were reported in the phase III trial ChAdOx1 vaccine [2]. However, prospective studies of CNS-IDD after COVID-19 vaccines are still lacking. We collected data from patients with CNS-IDD, suggestive of temporal associations with COVID-19 vaccination. We report 16 patients with different neurological manifestations of CNS-IDD within 6 weeks of COVID-19 vaccination.

Methodology

All patients were recruited prospectively at a tertiary care hospital in North India from June 2021 to May 2022. The following inclusion criteria were used: (1) evidence of CNS demyelination proved by clinical examination and radiological findings on magnetic resonance imaging (MRI); (2) recipient of COVID-19 vaccine, either first or second dose, within the past 6 weeks from the onset of symptoms; (3) no evidence of COVID-19 infection within the past 12 weeks from the onset of symptoms; and (4) no evidence of CNS demyelination within 12 months of the first attack. The aim of this study is to describe the characteristics of individuals who experience their first demyelinating episode post-COVID vaccination. We excluded patients diagnosed with CNS demyelination before receiving their first dose of vaccination. Patients were diagnosed as having MS according to the 2017 McDonald criteria, NMOSD was diagnosed according to the 2015 International Panel for NMO Diagnosis (IPND) criteria, and myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) was diagnosed according to the recent international MOGAD panel criteria [9,10,11]. ADEM was diagnosed by the International Paediatric Multiple Sclerosis Study Group criteria for paediatric multiple sclerosis and immune-mediated central nervous system demyelinating disorders: revisions to the 2013 definitions [12]. Patients were diagnosed to have tumefactive demyelination if they had a lesion on the FLAIR sequence measuring > 2 cm [13].

Patients fulfilling inclusion criteria were studied for demographic profile, clinical features, type of COVID-19 vaccine exposure, duration between COVID-19 vaccination and symptom onset, investigations including cerebrospinal fluid (CSF) analysis, serum Aquaporin 4 IgG (AQP 4) antibodies, myelin oligodendrocyte glycoprotein (MOG) IgG antibodies, MRI of the brain and/or spine, and ancillary investigations to exclude alternative aetiologies: serum antinuclear antibodies (ANA) profile, C-reactive protein (CRP) antineutrophil cytoplasmic antibodies (ANCA), paraneoplastic antibodies and autoimmune encephalitis antibodies.

The institutional ethics committee approved this study, and written consent was taken from all participants.

Statistical analysis

Data were expressed as descriptive statistics, such as mean ± SD or median with the range for continuous variables and frequency, and percentage for categorical variables.

Results

We analysed a cohort of 16 patients (8 female and 8 male) who received COVID-19 vaccination and developed central demyelination. The majority of the patients were young with a median age was 40 years (range 16–66) (Table 1). All sixteen patients had new onset demyelination (Table 1). Clinically, the most common presentation was myelitis (7/16) or optic neuritis (6/16) (Table 2, Fig. 1). Three patients presented with hemiparesis and three patients presented with altered sensorium. One patient presented with signs of raised intracranial tension (Table 2).

Table 1 Demographic and Laboratory details
Full size table
Fig. 1
figure 1

MRI spine (a–c) shows long segment intramedullary central (grey matter) T2 hyperintensity is noted in spinal cord from D1–D10 vertebra level with mild cord expansion s/o longitudinally extensive transverse myelitis

Full size image
Table 2 Examination and radiological findings of all patients
Full size table

In the present cohort, patients received ChAdOx1 nCoV-19 (14 patients) or BBV152 (2 patients). After the first dose of vaccination, 13/16 patients developed demyelinating event and 5/16 patients developed demyelinating events after the second dose of vaccine (Table 2). Two patients developed demyelinating events after both doses of vaccines. Median time duration of presenting symptoms after vaccination was 18 days (range 3–40 days) (Table 1).

MOG-IgG antibody was the most common antibody (8/16) detected in our cohort. One patient tested positive for ANA antibody (Table 2). None of the patients tested positive for AQP4 antibody, ANCA antibody, paraneoplastic or autoimmune encephalitis antibodies. We performed CSF studies on twelve patients. Median CSF white cell counts were 13 cells/µl (range 0–200 cells/µl). Oligoclonal bands (OCB) in CSF were detected in two patients. Ten patients (62.5%) had elevated CSF protein (> 45 mg/dl). Median CSF protein levels were 66 mg/dl (28–468 mg/dl) (Table 1).

Four patients had LETM on MRI (Fig. 1). Two patients had radiological findings consistent with multiple sclerosis. Tumefactive demyelinating lesions (TDL) were present in four patients (25%) (Fig. 2).

Fig. 2
figure 2

MRI (a–e) images shows a well-defined FLAIR hyperintense mass lesion (a) involving subcortical, deep and periventricular white matter of left parietal and occipital lobe extending into splenium of corpus callosum on left side with peripheral diffusion restriction (b and c) and incomplete rim enhancement on post-contrast scan (d). MRS (e) shows large lipid–lactate peak from the peripheral enhancing edge of the lesion. Features are suggestive of tumefactive demyelination

Full size image

In the present cohort, two patients were diagnosed with RRMS (Fig. 3), eight patients were diagnosed with MOG antibody-associated disease (MOGAD), three patients were diagnosed as seronegative TDL, one patient had seronegative cortical encephalitis and one patient had seronegative multifocal central demyelination (Table 2).

Fig. 3
figure 3

MRI images (a–c) shows multiple FLAIR hyperintense lesions oriented perpendicular to the periventricular origin (a) and involving calloso-septal interface (b). FLAIR hyperintensity is also seen along left optic nerve (c). These imaging features are in favour of demyelination disease (multiple sclerosis)

Full size image

All patients received injectable methylprednisolone and oral steroids during the acute event. Five patients received plasma exchange therapy due to inadequate response to steroids. Eight patients were started on azathioprine, three patients received rituximab, one patient received an injection of cyclophosphamide and one patient was started on mycophenolate mofetil (MMF) to prevent further disability. Two patients were not kept on any long-term immunomodulators. As none of the patient in our study relapsed till last follow-up, no patient was checked for positivity of MOG antibody (Table 3).

Table 3 Treatment details and clinical outcome of all patients
Full size table

Discussion

Since the eighteenth century, vaccines have proved to be an effective solution in the prevention of infectious diseases. Rarely, however, vaccines may induce an unexpected inflammatory reaction by breaking the mechanism of self-tolerance, or the production of autoantibodies, causing evident autoimmune diseases [14].

A previous systematic literature review reported 71 cases of post-vaccination central demyelination. Influenza and human papillomavirus vaccines were the most common cause of central demyelination [14]. Neurological adverse events after COVID-19 vaccination are monitored by various regulatory authorities. Two cases of transverse myelitis were reported in the vaccine group in the phase III trial of the ChAdOx1 vaccine [2]. Vaccine adverse event reporting system (VAERS) has reported transverse myelitis and ADEM post-COVID-19 vaccine [15].

Myelitis was the most common presentation in our study. In the systematic review of 32 patients of CNS-IDD, transverse myelitis was most commonly reported following COVID-19 vaccinations [16]. In a study from southern India, transverse myelitis and optic neuritis were the most common presentations associated with ChAdOx1 vaccine [17]. In a recent systematic review, 17/32 cases were reported after mRNA vaccination and 10/32 were reported after viral vector vaccines [16]. mRNA-based vaccines were the most common culprit in South Korea [18]. Female preponderance is usually seen in CNS-IDD, but in this cohort, we found equal cases in both gender.

The most common antibody detected in our cohort of post-vaccinal CNS-IDD was the MOG IgG antibody. Around 20% of cases of MOGAD are associated with infectious agents or vaccines [19]. Post-vaccine MOGAD has been explained with measles–rubella, Japanese encephalitis, diphtheria, pertussis and tetanus (DPT), influenza and polio vaccines [19, 20]. In a systematic review of post-COVID-19 vaccine MOGAD, 18/20 (90%) patients received the ChAdOx1 vaccine [21]. In our series, 7/8 MOGAD patients received the ChAdOx1 vaccine. None of the patients were evaluated for MOG IgG antibody in the follow-up. Previous studies suggest that MOGAD after COVID-19 vaccination is primarily a monophasic illness with a lack of persistent positive MOG antibody [17].

We reported two RRMS patients in our cohort. Apart from yellow fever vaccines, previous studies did not support the association between vaccination and multiple sclerosis relapses. However, relapses and new onset RRMS diagnosis have been observed previously [22]. In both case series, all patients had received mRNA COVID-19 vaccination [23]. Both RRMS patients received the ChAdOx1 vaccine.

The mechanism of CNS-IDD after COVID-19 vaccination is not clearly understood. Susceptibility of patients for autoimmune diseases, vaccine antigens and immune adjuvants used in the vaccine to enhance immune responses are the few factors speculated to have a role in the post-vaccination CNS-IDD [16]. Similarities between viral protein used in vaccine and self-antigens (i.e. myelin, oligodendrocyte specific protein) can trigger immune response (molecular mimicry) [24]. This inflammatory response caused by molecular mimicry can further increased by mechanisms like bystander activation and epitope spreading [25]. Bystander activation involves the activation of costimulatory pathways and cytokine release. In epitope spreading, antigen-specific immune response leads to tissue destruction. This releases endogenous epitopes, causing cytokine release and tissue damage [16]. Vaccine-related factors like type, route of administration and dose could possibly trigger of demyelination [24]. The occurrence of asymptomatic lesions and subclinical involvement has been reported in MS and MOGAD [26, 27]. It may be possible some of the patients in our cohort already had an underlying subclinical/asymptomatic demyelinating illness that was unmasked following the administration of COVID-19 vaccine.

ChAdOx1 vaccine viral vector vaccine using modified chimpanzee DNA adenovirus. This DNA later replicates in host DNA and produces viral spike protein which can elicit an immune response [5]. BBV152 is a whole-virus inactivated vaccine produced from the SARS-CoV-2 strain, in which the genetic material of the vaccine is destroyed but the antigen remains intact [28]. Vaccines deliver the SARS-CoV-2 antigen into dendritic cells (DC) using lipid adenoviral vectors in the case of the ChAdOx1 vaccine and the case of BBV152 by using the inactivated SARS-CoV-2 virus [5, 28].

ChAdOx1 vaccine triggers innate sensors like double-strand DNA sensor TLR9, and BBV152 vaccine activates TLR7/8 (IMDG) and the NLRP3 inflammasome. This results in the production of IL-12 drives naïve T-cells towards the pro-inflammatory Th1-biased response that generates strongly neutralizing antibodies and virus-specific CD8 effector T-cells to clear the infection. Antigen-specific memory B- and CD8 T-cells afford long-term protection [28, 29]. Researchers studied the cross-reactions between host and microbial proteins using the Protein Data Bank, showing that SARS-CoV-2 proteins possibly interact with proteins involved in axonal transport, neuronal transmission, synaptic vesicle trafficking, endocytosis, axonal transport and blood–brain barrier [29, 30].

Finally, the literature has reported more post-COVID-19 infection CNS-IDD cases than post-COVID-19 vaccines. This is in line with the available literature on various vaccines which suggest risk of demyelination to be substantially lower when compared to infections [16]. Population-based studies from Europe showed no increased risk of neuroimmunological events post-COVID-19 vaccination like transverse myelitis, meningitis and encephalomyelitis. However, people with COVID-19 infections have a higher risk of transverse myelitis or encephalomyelitis [31, 32]. A systematic review by Ostovan et al. showed that apart from a lower incidence of transverse myelitis associated with COVID-19 vaccines, the intensity of post-vaccine myelitis was lesser than myelitis post-COVID-19 infection [33]. Thus, although the risk of CNS-IDD post-vaccines is there, the benefits of vaccination overshadow the possible risk of CNS inflammation.

The risk of vaccine-related CNS-IDD is low, even in patients with known autoimmune CNS diseases [34]. Retrospective studies have shown that vaccines BNT162b2-Pfizer-BioNTech and mRNA-1273-Moderna benefit from vaccination and outweigh the risk of vaccine-related CNS-IDD reactivations [34,35,36]. In the study of 26 AQP4-IgG + NMOSD and 30 MOGAD patients, one patient had a relapse within 4 weeks of vaccination [34]. Seven out of 324 (2.2%) patients with multiple sclerosis reported relapse within two months post-BNT162b2-Pfizer-BioNTech vaccination [36].

This study has limitations. Establishing causality between COVID-19 vaccines and CNS-IDD may be uncertain as this is a case series. We did not report the rate of CNS-IDD post-COVID-19 infection in the community. We did not compare probable CNS-IDD cases with controls, i.e. diagnosed with one of the CNS-IDD and were not vaccinated within 6 weeks of the demyelination.

Conclusion

We describe the temporal association between COVID-19 vaccines and CNS-IDD. Differences between episodes of demyelination after a particular type of vaccine, temporally associated demyelination events and the presence of MOG-IgG antibody in a significant proportion of patients could increase the likelihood of immunogenic phenomenon post-vaccination in the particular patient. Long-term post-marketing surveillance and large prospective controlled studies are required to establish a possible causal relationship between the COVID-19 vaccine, CNS-IDD and safety of individual vaccines.