Prevention of infectious diseases through immunization is one of the greatest public health accomplishments of all time. Like all therapies, vaccination also carries the risk of adverse events. However, the risk of serious events was documented to be very low for all approved types of vaccination and neurological adverse events are exceedingly rare (1).

Notable examples include the 1976 H1N1 influenza vaccine which was withdrawn from the market in the USA due to a small increase of Guillain Barré syndrome (GBS) cases (an additional one case of GBS for every 100,000 vaccines (2)), or the intranasal influenza vaccine withdrawn in Switzerland in 2001 due to a small increase in peripheral facial palsy (an extra 13 cases of Bell’s palsy per 10,000 vaccinations (3)).

However, all cases of reported neurological complications following vaccination should be carefully evaluated to confirm if a neurological disorder is indeed present (which is not obvious as a variety of functional neurological disorders were reported after vaccination (4)) as well as to prove causality (1, 5). The latter aspect is difficult to accomplish as there are no distinguishing clinicopathologic findings in vaccine-induced neurological adverse events (1). Nevertheless, as vaccine uptake is directly influenced by public confidence in the safety of the vaccine, it is crucial that those reporting potential neurological adverse events employ rigorous methodological strategies when determining causal relationships (5).

Coronavirus disease 19 (COVID-19) vaccines comprised adenovirus-vectored preparations, including ChAdOx1nCoV-19 (Oxford–AstraZeneca) and Ad26.COV2.S (Janssen), as well as RNA-based immunizations, including BNT162b2 (Pfizer–BioNTech) and mRNA-1273 (Moderna) (5). All of these vaccines have been approved due to their excellent safety and efficacy data from clinical trials (6, 7). Nevertheless, the clinical trials were underpowered to identify rare adverse events crucial for continuous risk-benefit assessments (8). The observed heightened susceptibility to cerebral venous sinus thrombosis following the ChAdOx1nCoV-19 vaccine serves as an example of a rare adverse neurological event with an immune-mediated pathogenesis (9).

The occurrence of immune-mediated neurological complications after vaccination is biologically plausible considering that vaccines can cause a strong expression of proinflammatory cytokines and a T cell response, leading to neuroinflammation after microglia activation (10). This was confirmed for peripheral complications such as GBS in the setting of adenovirus-vectored vaccines as the observed-to-expected ratios for GBS following Ad26.COV2.S vaccination were ≥1.5-fold higher than background rates in one study (11), whereas another study estimated 38 excess cases of GBS per 10 million people receiving ChAdOx1nCoV-19 (8). In both scenarios it was observed that this risk was mitigated by the fact that the rate of neurological events after acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection was up to 617-fold higher than after COVID-19 vaccination (11) and, specifically, 145 excess GBS cases per 10 million people are estimated after a positive SARS-CoV-2 test result (8). The impact of COVID-19 infection as compared to vaccination could be even more prominent, considering that a subtle neuronal damage has been demonstrated in patients with minor neurological involvement, or no symptoms at all (12, 13). In fact, neuroaxonal injury is often subclinical, and surrogate markers such as neurofilament light chain protein (NfL) are key to its detection; though significantly correlating with clinical severity, NfL also appears to be elevated in COVID-19 cases without prominent CNS manifestations (14).

Besides epidemiological evidence, a causal association can be demonstrated also if neurological complications show a specific clinical phenotype divergent from that of the naturally occurring disease. In this regard, pharmacovigilance data demonstrated that GBS occurring after administration of adenovirus-vectored vaccines present with an abnormally high incidence of facial palsy (facial diplegia in most cases), which supports a causal relationship between such exposure and this syndrome (15).

Conversely, little has been done to gather information on central nervous system (CNS) adverse events of COVID-19 vaccination for which only isolated case reports and small case series exist (16–21), despite the fact that detailed assessments of potential neurological adverse events associated with COVID-19 vaccines as well as infection are urgently needed.

Herein, we aimed at characterizing the clinical spectrum and immunological features of CNS immune-related events following SARS-CoV-2 vaccination and assessing the potential for a causal association using validated diagnostic criteria.

Nineteen patients were included in the study (11 patients from the 6 Italian participating centers and 8 patients from the French Reference Center). The median age was 56 years (range: 20-83 years); 10/19 patients were male (53%). The most frequently administered vaccine was Pfizer-BioNTech BNT162b2 (10/19, 53%), followed by AstraZeneca ChAdOx1 nCov-19 (7/19, 37%) and Moderna mRNA-1273 (2/19, 11%). None of the patients received Janssen/Johnson&Johnson Ad26.COV2.S. Overall, 12/19 (63%) patients received an mRNA-based vaccine, while the remaining an adenovirus-vectored vaccine (7/19, 37%).

For comparison, the proportion of people receiving the different types of COVID-19 vaccine in the Italian and French population was as follows: 92% mRNA-based vs. 8% adenovirus vectored-base in Italy (from Rapporto Vaccini Anti-COVID-19, Ministero della Salute, accessed on 15 December 2022) and 94% mRNA-based vs. 6% adenovirus vectored-base in France (from Agence nationale de sécurité du médicament et des produits de santé, ANSM, accessed on 15 December 2022), suggesting a possible over-representation of the adenovirus-vectored vaccines group in the present cohort.

Neurological symptoms presented after the first vaccine dose in in 8/19 (42%), the second in 5/19 (26%), the third in 4/19 (21%), and the fourth in 2/19 (11%). The median time lag between vaccination and symptoms onset was 14 days (range: 2-41 days). A history of systemic autoimmune disease was identified in three patients (3/19, 16%), and it consisted in autoimmune thyroid disorder in all of them.

AE was the most frequent diagnosis (11/19, 58%), followed by myelitis (4/19, 21%), ADEM (3/19, 16%), and brainstem encephalitis (1/19, 5%). The level of diagnostic certainty was “definite” for most patients (10/19, 53%), and “possible” for the remaining ones (9/19, 47%).

The most frequent symptoms were altered mental status (7/19, 37%) and myelopathy (7/19, 37%), followed by subacute working memory deficit (6/19, 32%). Other symptoms included: movement disorder (5/19, 26%), seizures (5/19, 26%), new focal CNS deficit (5/19, 26%), ataxia (4/19, 21%), CN palsies (2/19, 11%), language disturbance (2/19, 11%) and psychiatric symptoms (2/19, 11%). None of the patients developed autonomic dysfunction/central hypoventilation nor ophthalmoplegia. Table 1 summarizes the main clinical features of the patients.

In this study, we reported 19 cases of CNS immune-mediated disorders, all occurring within 6 weeks of SARS-CoV-2 vaccination, which were identified from 7 tertiary referral hospitals across Italy and France (one of them being a national referral center for AE), over almost 1 year and half of vaccination campaign. Even though this study was not designed to capture the impact of COVID-19 vaccination on the epidemiology of CNS inflammatory disorders, these data suggest that such adverse events are very rare, in line with previous reports (8, 11, 19). Conversely, the occurrence of both peripheral (e.g. GBS (27–29)) and central (e.g. AE, myelitis, ADEM (18, 30)) immune-mediated complications of COVID-19 was much higher (11), especially in the first waves of the pandemic, likely reflecting a protective role of vaccines not only in reducing SARS-CoV-2 infections, hospitalizations, and deaths but also neuroimmune disorders triggered by the infection itself (8).

Herein, we observed that three main clinical patterns (AE, myelitis, and ADEM) characterize CNS immune-mediated disorders in patients recently vaccinated from SARS-CoV-2, a type of involvement that parallels the one triggered by COVID-19 (30). Also, similarly to neuroimmune disorders induced by the infection, most of the cases were Ab-negative and cytokine analysis in a small subgroup of cases was in favor of a robust T-cell activation, with elevated levels of CXCL10 (IP-10). CXCL10 is a proinflammatory chemokine which has a key role in the priming phase of the T cell response, as well as in the recruitment of CD8⁺ and Th1-type CD4⁺ effector T cells (31). However, the small sample size prevents us from drawing any robust conclusion, and our findings require confirmation in a larger cohort of patients.

Vaccinations, as it was demonstrated for the ChAdOx1nCoV-19 vaccine (32), can lead to the production of proinflammatory cytokines and T cell responses. This can result in the release of pyrogenic cytokines into the bloodstream, similar to the response seen in natural infections (16). Interestingly, a protective clinical effect from SARS-CoV-2 vaccination is seen within 11 days, when neutralizing Abs are hardly detectable, suggesting that vaccine-induced CD8+ T cells may therefore be the main mediators of protection at this early stage (31, 33, 34). In rare cases, vaccination might lead to neuroinflammation due to microglial activation, influenced by individual genetic and immune memory factors (35). The mechanisms by which vaccinations may potentially induce de novo CNS inflammatory disorders include molecular mimicry, epitope spreading, or demasking/release of epitopes (17, 36). It is possible that similarities between foreign epitopes (provided by vaccine exposure) and self-antigen leads to aberrant activation of autoreactive T or B cell clones (molecular mimicry) (37). Alternatively, an initial antigenic stimulus may induce spreading of the specificity of the immune response, which includes self-epitopes other than the one which initiated inflammation (epitope-spreading) (38). SARS-CoV-2 vaccine may as well have the potential to activate autoreactive clones in an antigen-independent, non-specific manner, mediated by indirect signals that favor an inflammatory milieu (bystander activation) (39). It is difficult to hypothesize which of these mechanisms could be involved, and whether different classes of autoimmune complications may underlie different pathogeneses. Interestingly, a recent paper on acute inflammatory diseases of the CNS after SARS-CoV-2 vaccination has shown that almost all patients included with detectable autoantibodies received ChAdOx1S, suggesting that different types of vaccines may trigger specific forms of CNS autoimmunity (17). In addition, vaccination might also trigger the exacerbation of pre-existing, subclinical, neuroimmune disorders (36), similarly to what it was observed for other adverse events triggered by nonspecific T cell activation, such as those of ICIs (40, 41). Finally, a temporal coincidence of the 2 events (vaccination and CNS immune-mediated disorders) cannot be completely excluded considering the high global vaccination rates (18, 36).

In this cohort, the diagnosis was supported by:

Interestingly, all but 1 patient with an AE phenotype were neuronal Ab-negative, while the remaining patient was positive for Hu Abs (confirmed using 2 distinct techniques at the French Center for PNS). This case showed several distinctive features as compared to the classic anti-Hu phenotype, which belongs to the “high-risk” Ab group, being associated to cancer (usually small-cell lung cancer) in >85% of cases (26). Few non-paraneoplastic cases have been described in children (44), but none after vaccination. In addition, the response to treatment is usually scarce and unsatisfactory (45). Conversely, a prolonged beneficial response to first-line treatments only (corticosteroids and IVIg) was observed in this case, while multiple oncological screenings (including whole-body PET) were negative. Although we need to carefully follow this case over time before making a conclusion on the possible causal link with vaccination, it is interesting to note that an increase in the detection of Hu-Abs (along with GFAP-Abs) was observed at the Barcelona Referral Center by comparing the pre–COVID-19 (2017–2019) and COVID-19 (2020–2021) periods, while the authors did not detect a substantial increase of encephalitis mediated by Abs against neural-surface antigens (46). Supporting this observation, most (>90%) of the cases of AE temporally related to vaccination are Ab-negative in both the present series and previous systematic review of the topic (42). Few cases of anti-LGI1 AE developing after COVID-19 vaccination have been reported, including 4 cases from Italy (20) and 1 from Israel (21), but none of the AE patients in the present series tested positive for LGI1-Abs and none showed the characteristic faciobrachial dystonic seizures pathognomonic of the disorder.

LGI1-Ab AE, being an immunoglobulin G4 (IgG4)-Ab disorder, is strongly associated with particular human leukocyte antigen (HLA) class II haplotypes, in particular DRB1*07:01, which is carried by nearly 90% of anti-LGI1 patients (47). Interestingly, an association between the HLA-DRB1*07:01 allele and greater immunogenicity after the administration of mRNA-1273 vaccine was observed (48), suggesting that future studies aiming at better exploring the immunogenetic factors associated with the development of these rare adverse events are definitely warranted.

Another interesting observation from the present study was the neuroradiological involvement of cerebellar peduncles in 2 cases, all showing contrast enhancement, which was also suggested in previous reports (17, 19, 42, 49).

This study is limited by its retrospective nature and the small sample size. Also, the multicenter design does not allow us to comprehensively study the potential effect of both COVID-19 infection and vaccination on the epidemiology of these conditions.

However, this represents one of the largest series of CNS disorders developing in close temporal association with vaccination, in which a probable link with immunization was established using validated diagnostic criteria.

In conclusion, CNS immune-mediated adverse events of COVID-19 vaccination appear to be very rare and consist mostly of Ab-negative AE, myelitis, and ADEM developing approximately 2 weeks after vaccination, when a mounting robust CD8+ T cell response is suggested by the increased CXCL10 CSF levels. Clinicians should be aware of these rare complications, as most patients improved following immunomodulatory treatment. Conversely, neuroimmune complications after COVID-19 infection are far more frequent, therefore neurologists should continue to promote vaccination as benefits largely outweigh risks at the population level.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

The studies involving humans were approved by Institutional Review Board (IRB-DAME). The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin because the study was retrospective, observational, and non-interventional, focusing solely on the analysis of pre-existing records, without direct participant interaction or intervention.

AV: Conceptualization, Data curation, Funding acquisition, Methodology, Supervision, Validation, Writing – original draft, Writing – review & editing. ST: Data curation, Formal Analysis, Investigation, Writing – review & editing. MD: Data curation, Formal Analysis, Investigation, Writing – review & editing. MF: Investigation, Methodology, Writing – review & editing. FB: Formal Analysis, Software, Writing – review & editing. MB: Investigation, Writing – review & editing. LV: Investigation, Writing – review & editing. DB: Investigation, Writing – review & editing. MG: Investigation, Writing – review & editing. PT: Investigation, Writing – review & editing. MZ: Investigation, Writing – review & editing. LZ: Investigation, Writing – review & editing. APi: Investigation, Writing – review & editing. APa: Investigation, Writing – review & editing. MV-G: Investigation, Writing – review & editing. VD: Investigation, Writing – review & editing. GG: Resources, Writing – review & editing. JH: Investigation, Writing – review & editing. MV: Funding acquisition, Resources, Writing – review & editing.