Neuroinflammation triggered by SARS-CoV-2 infection: syndromes and therapies

Older adults and people with comorbidities have been the groups most severely affected by SARS-CoV-2 infection. By contrast, children and adolescents are typically asymptomatic or mildly affected.¹ Neurological complications of COVID-19, particularly stroke, have already been well described in adults.² Initial reports of neurological complications of COVID-19 were in children with paediatric inflammatory multisystem syndrome temporally associated with SARS-CoV-2 (PIMS-TS), a hyperinflammatory syndrome also known as multisystem inflammatory syndrome in children (MIS-C). However, a broader spectrum of neurological complications in children have been less well described.3, 4

In The Lancet Child & Adolescent Health, Stephen Ray and colleagues⁴ report a prospective study of neurological manifestations of SARS-CoV-2 infection using a UK nationwide notification portal for neuroscience bodies (CoroNerve), which was integrated with hospital admission data from National Health Service England. In England, neurological complications were identified in 51 of 1334 hospitalised children and adolescents during the 10 months from April 2, 2020, to Feb 1, 2021 (3·8 cases [95% CI 2·9–5·0] per 100 paediatric patients). The equivalent adult study identified neurological complications in 267 of 30 197 hospitalised adults (0·9 cases per 100 patients) during the same time period.² The 20-fold higher frequency of hospitalisations in adults is striking; however, neurological complications were relatively more common in children and adolescents, especially in those with pre-existing neurological conditions (19% of patients) and those of Black or Asian ethnicity (69%).3, 4

To stratify the neurological complications, Ray and colleagues⁴ described two groups. One group (52% of patients), referred to as the COVID-19 neurology group, had recognised infection-associated neurological complications (such as status epilepticus), or infection-triggered neuroimmune syndromes (eg, acute disseminated encephalomyelitis [ADEM] or Guillain-Barré syndrome). The other group of patients (48%), referred to as the PIMS-TS neurology group, had neurological involvement as part of PIMS-TS, with encephalopathy and neuropathy being the most common neurological features. This group presented later after SARS-CoV-2 infection and were more likely to be admitted to the intensive care unit than those in the COVID-19 neurology group (appendix). Unlike in the adult CoroNerve study where stroke constituted about 50% of neurological complications, stroke was only described in two (4%) of 52 children and adolescents with neurological complications.

Neuroimaging was important to make specific syndromic diagnoses. In the COVID-19 neurology group radiological ADEM was frequent, whereas some patients in the PIMS-TS neurology group had encephalopathy with reversible splenium involvement,⁴ a rare clinicoradiological syndrome, typically triggered by common infections or Kawasaki disease.5, 6

The detailed immunopathogenesis of these neurological complications is unclear. Investigations showed a pronounced inflammatory response, particularly in the PIMS-TS neurology group.4, 5 Most studies to date have not provided evidence of direct CNS viral invasion, although the findings are mixed.⁷ In the study by Ray and colleagues, all children and adolescents who had cerebrospinal fluid (CSF) tested for SARS-CoV-2 had negative test results, supporting the concept of infection-triggered neuroinflammation, rather than direct neurovirulence.⁴ Brain histopathology in patients with fatal COVID-19 has shown profound hyperinflammation and microangiopathy, but typically no SARS-CoV-2 antigen.⁸

Some of the neurological complications, such as status epilepticus in children with pre-existing brain disease, presumably represent infection-provoked exacerbation of underlying vulnerability. By contrast, other complications, such as the acute demyelinating syndromes, are due to adaptive autoreactive lymphocytes or autoantibodies. Interestingly, five (63%) of eight patients with acute demyelinating syndromes had positive myelin oligodendrocyte glycoprotein (MOG) IgG. By contrast, the proposed immunopathogenesis of PIMS-TS is dysregulated cellular and cytokine-driven hyperinflammation and endothelial disease, with immunological similarities to Kawasaki disease and macrophage-activation-like syndrome. Therefore, the immunopathogenesis is heterogeneous, with broad involvement of both the innate and adaptive immune systems.⁵

A strength of the study by Ray and colleagues was the prospective, nationwide data collection overlaid on COVID-19 hospitalisation data, allowing prevalence calculations of neurological complications in hospitalised children. Much harder to determine is a total population prevalence, given the paucity of testing in children and adolescents who are not admitted to hospital.

The limitations of the study include difficulties inherent to making definite associations between non-CNS infections and neurological complications. False-positive associations might be particularly common when using single-positive IgG as causal evidence, given the SARS-CoV-2 seroprevalence of approximately 6% in the UK, with higher prevalence among Black or Asian people.⁹ Performing both respiratory and CSF PCR plus paired acute or convalescent serology could improve causal association between SARS-CoV-2 infection and neurological complications. The data captured by Ray and colleagues included a very broad spectrum of neurological and psychiatric symptoms, which reduced potential preconceived bias, although some recognised complications such as acute necrotising encephalopathy and cerebrovenous sinus thrombosis were not described, and milder syndromes such as headache, parosmia, or behavioural syndromes were likely to be under-reported. A further limitation was the inevitably short follow-up, with concerning early outcome data: 17 (33%) of 52 patients had some disability, and one patient (2%) died.⁴ Given recovery after brain injury continues for more than a year, long-term follow-up is needed to describe the true outcomes. It is also unclear if children and adolescents are vulnerable to post-acute sequelae of SARS-CoV-2 infection (so-called long COVID) as is described in adults, with fatigue, neurocognitive difficulties, and depression that persist after acute disease.⁷

Describing neurological syndromes helps clinicians to make diagnostic and therapeutic decisions. Important analogies to COVID-19 neurology syndromes include descriptions of influenza-induced neurological complications, which affect 7·6% of children admitted to hospital with influenza,¹⁰ and herpes simplex virus (HSV) encephalitis, which can induce a latent autoimmune neurological relapse that responds to immune therapy. Recognising that infections such as influenza, HSV, and SARS-CoV-2 can induce secondary autoinflammation and autoimmunity provides a therapeutic rationale that infection-provoked neuroinflammation requires anti-inflammatory treatments to optimise outcomes.

© 2021 JohnnyGreig/iStock

We declare no competing interests.