Our data shows a case of highly active multiple sclerosis in a young patient following vaccination against SARS-CoV2. As demonstrated through previous reports, this is a rare event. As already stated, among the 28 cases described, none sustained a major disability characterized by an EDSS 6.0 or higher (Table 4). Conversely, follow-up data are lacking in 6 cases. Considering the 22 cases with reported follow-up, follow-up time was not specified in most cases, while in the remaining ones this time was one year at the longest. Conversely, longer follow-up times would be needed to evaluate if relapse risk in those cases is similar to that normally seen in MS. Even in the case here described, where we were confronted with a severe disease course, remission could be reached, and no major sequelae remained after 2 years of onset.
Another the gap in understanding the phenomenon is the tenuous relationship between vaccination and disease development. Concerning MS, in our literature review, evaluated reports included cases where the time for onset after vaccination greatly varied. The periods lasted from 1 to 60 days. Furthermore, the cases occurred after exposition to 5 different vaccines. It could be implied the autoimmune insult is rather relative to SARS-CoV2 exposition, which may also occur as consequence of infection, than to a specific vaccine formulation or technology. Epidemiological evidence supports this position, with an increased risk of developing different autoimmune diseases after COVID-19, a risk which seems reduced for vaccinated patients (Peng et al., 2023). In accordance, there are also reports of both MS and NMOSD development after COVID-19 in unvaccinated patients (Avila et al., 2023; Pignolo et al., 2021; Palao et al., 2020). A systematic review found 18 cases of MOGAD following infection (Mirmosayyeb et al., 2023), while another one revealed 15 cases of NMOSD (Harel et al., 2023). Another review, which encompassed any case of CNS inflammatory demyelinating disease (IDD), found higher numbers for new onset MOGAD, with 27 cases (Lotan et al., 2022). The same study described 10 new diagnoses of MS following infection, and 63 new relapses in previously diagnosed patients. It is of note, though, the relapse percentage in vaccinated MS patients appears to be no different from unvaccinated ones (Kong et al., 2022) and similar to those seen after COVID-19 (Stastna et al., 2022). Tumefactive demyelinating lesions were also observed after COVID-19 (Erdoğan et al., 2023), with another case which further developed into relapsing disease in the form of seronegative NMOSD. In this scenario, we could not find any studies which would present causal relationship between SARS-CoV2 vaccination and MS, or evidence demyelination risk in vaccination is higher than in COVID-19 or other viral expositions. On the contrary, a 2022 evaluation on the World Health Organization (WHO) pharmacovigilance database, which compiled 2,840 cases of CNS demyelination reported as adverse reactions following COVID-19 vaccines, showed no difference in proportion of those events when compared to what is seen in other viral vaccines (Kim et al., 2022).
Both animal and in vitro models showed the potential of the SARS-CoV2 spike protein S1 subunit to induce a pattern of neuroinflammation, with TLR4 activation and increased IL-1β, interleukin-6 (IL-6), TNF, chemokine C-X-C motif ligand 1 (CXCL1) and IL-10 production by microglia (Frank et al., 2022). It has been known for long that viral invasion and autoimmune insult lead to neuroinflammation and immune cell infiltration in CNS through differing mechanisms of brain blood barrier disruption (Fabis et al., 2008). SARS-CoV2 was already demonstrated to be capable of infecting CNS cells in preclinical models, in a process dependent of spike protein binding to angiotensin converting enzyme 2 (ACE2) receptors (Song et al., 2021). Autopsy data show tropism to the brainstem, and that CNS invasion by SARS-CoV2 is independent of disease severity or encephalopathy (Stein et al., 2022). Interestingly, there was low evidence for viral induced neuroinflammation. Another study, however, of 4 patients with COVID-19 and spike protein detected in brainstem, showed both brainstem and the cerebrum to have higher transcription of pro-inflammatory factors (Roczkowsky et al., 2023). In COVID-19, CSF evaluations demonstrate elevated IL-6, interleukin-16 (IL-16) and chemokine C-X-C motif ligand 10 (CXCL10) concentrations in CNS, but in CSF/serum indices which suggest cytokine leakage from periphery through the blood-brain barrier (BBB), rather than CNS intrinsical response or immune cell invasion (Reinhold et al., 2023). Nonetheless, microglia and astrocytes of SARS-CoV2 infected patients presented higher reactivity and impaired anti-inflammatory response (Madden et al., 2023). Hyperalbuminorrachia and increased astrocytic S100b further indicate a BBB dysfunction (Perrin et al., 2021).
Further evidence for vascular impairment in CNS complications of SARS-CoV2 rests in experimental studies. Endothelial cells of post-mortem brain tissue can present inclusion of spike fragments and other SARS-CoV2 membrane proteins. When the S1 subunit was injected in mice, it induced encephalopathy (Nuovo et al., 2021). The injected S1 colocalized within endothelial cells to caspase-3, ACE-2, IL-6 and complement component 5b-9 (C5b-9), indicating microvessel pathology. Integrity of endothelium and BBB disruption was also induced by the spike protein alone (Reynolds et al., 2021) and is mechanistically involved oxidative stress and contraction of pericytes (Khaddaj-Mallat, 2021). Antigen exposition may then itself be a factor for immune trafficking to the CNS. The S1 subunit is also known to cross the BBB (Rhea et al., 2021), where it induces microgliosis with oxidative stress (Clough et al., 2021), activation of resident T cells (Ayasoufi et al., 2023) and IL-1β and TNFα production (Lykhmus et al., 2024), furthering immune invasion.
In a case such as the one here described, where the patient presents genetic risk variants for higher pro-inflammatory cytokine production and autoimmune disease development, including MS, the viral antigen exposition in vaccination probably led to an inflammatory insult which aggravated a subclinical disease, in the same vein of the other cases seen in literature. The most notable difference rests in the atypical presentation of ADEM-like syndrome, which would delay MS diagnosis. ADEM is an expected adverse event after viral infection and vaccination in general and it has not been different for COVID-19 (Anilkumar et al., 2023). ADEM has been reported after SARS-CoV2 vaccination (Permezel et al., 2022; Ahmad et al., 2022), with 85.1% of patients achieving clinical improvement (Nabizadeh et al., 2023). COVID-19 is not only capable triggering a distinct form of ADEM higher vascular commitment, with microhemorrhages and poorer clinical outcomes (Reichard et al., 2020; Manzano et al., 2021), but also of exacerbating pre-existing ADEM (Hussein et al., 2020). This can occur even in absence of advanced infection (Shahmirzaei et al., 2021).
The main controversy in this situation is the differential diagnosis between MS and multiphasic form of ADEM (MDEM). Recurrence of ADEM was already reported in the event of SARS-CoV2 antigen exposition (Poli et al., 2023). MDEM is described in literature as uncommon and generally presents itself as two distinct and subsequent episodes of ADEM. There are no well-established predictors for relapses (Paolilo et al., 2020) and the occurrence of more than one relapse is rare (Kariyawasam et al., 2015). A minority of MDEM subjects have oligoclonal bands in their CSF and encephalopathy symptoms seem less frequent than monophasic ADEM (Zang et al., 2023). However, this definition may overlap with other syndromes with tumefactive demyelination, such as Baló’s concentric sclerosis form of MS, NMOSD or pediatric MOGAD. A retrospective study with patients first presenting with tumefactive lesions demonstrated roughly a quarter of them (33 of 116) later converted to a diagnosis of MS (Li et al., 2022). Patients with this profile can have disease courses varying from full remission to death (Pervin et al., 2024). The distinction is important as MS DMT may be ineffective for MDEM.
Although MRI lesions showed confluence (Fig. 2), a pattern of periventricular and oval lesions, non-continuous myelitis, the existence of black holes in T1-weighted images, the presence of oligoclonal bands in CSF, lack of peripheral nervous system involvement and predominance of motor-sensitive symptoms furthered the probability of MS diagnosis. It is also possible to note the occurrence of at least three distinctive periods of disease exacerbation, without new triggers. Those could be understood as individual relapses (Fig. 1), something less typical of MDEM (Shah et al., 2018). In this context, we considered the patient fulfilled criteria for MS. The short time in achieving a high EDSS value with high lesion count and short intervals between periods of disease activity would then qualify the case as highly active MS (Díaz et al., 2019). There was indication of high efficacy DMT due to cumulative disability risk, which prompted treatment with natalizumab.
Natalizumab acts as an integrin receptor antagonist which has relapse control capacity comparable to those of primarily immunosuppressant drugs (Boz et al., 2023) and the adoption of natalizumab therapy early in disease leads to better disability prognosis when compared to an escalation approach (Iaffaldano et al., 2021). Its main action mechanism is blockage of lymphocyte migration into the CNS by competitive antagonizing of the very-long antigen 4 (VLA-4) region of the α4 integrin present in endothelial cells constituting the BBB, impeding B, T and natural killer (NK) cells infiltration without immunosuppression (Khoy et al., 2020). At the same time, natalizumab has secondary effects stimulating leukocyte activation in the periphery, as can be seen for monocytes (Frisch et al., 2023) and memory cells (Planas et al., 2012). Its high efficacy may be attributed to diminished junctional integrity with higher adhesion molecule expression in MS endothelium (Nishihara et al., 2022). Choice of natalizumab was individualized within the clinical scenario reported using the mapping for genetic risk variants. We considered the possibility of inadequate response to anti-CD20 therapy with increased inflammatory activity; the onus of a potential immunosuppression in a patient with higher COVID-19 morbidity risk during the pandemics (Table 2); and the hypothesis of a higher BBB involvement as seen in other forms of SARS-CoV2 associated neurological complications.
The subject achieved sustained remission. Consistent reduction of plasmatic NfL occurred only after DMT introduction, along that of other CNS lesion biomarkers (Fig. 3). As expected with natalizumab, peripheral immune activity persisted, demonstrated by higher titers of IL-1β, IL-17A and CRP. Similarly, TGFβ, which in the CNS is mainly associated with glial reactivity and may act as a neuroprotective response (Diniz et al., 2019), decreased in plasmatic concentration. Notably, the greatest variation was seen for ROBO-4 (Fig. 4). ROBO-4 is part of a family of axon-guidance neurotrophic factors signaling through a Slit pathway. Indeed, ROBO-4 acts mainly as an angiogenic factor implicated in vasculature homeostasis (Dai et al., 2019). This effect may be amplified in a context of overexpression during inflammatory insult as a compensatory mechanism (London et al., 2011). This dynamic interplays with the activity of other endothelium-relevant molecule, VEGF, which may have synergic or antagonizing effects depending on pathology (Gong et al., 2019). Concurrently to ROBO-4 decrease, the subject presented increasing VEGF after disease control. Those variations may represent the BBB influx normalization after DMT.
As no new signs of disease activity appeared, natalizumab proved useful in this case. The patient is seronegative for JCV, thus the current risk of progressive multifocal leukoencephalopathy (PMLC) is neglectible. Although it is safe to maintain the proposed treatment for now (Butzkueven et al., 2020), it is of future concern to evaluate how long the patient will need DMT and, if she changes her JCV status, what would be an adequate strategy to minimize relapse risk durgin the natalizumab washout period.