The human coronavirus (HCoV) is now considered an emerging pathogen but its links to multiple sclerosis date back some 40 years.
The first coronavirus strain (JHM) was identified in mice in 1949 and caused encephalomyelitis and myelin destruction (Bailey et al. J Exp Med 1949;90:195-212); other murine coronaviruses (e.g. MHV) were later used as viral models of demyelination. Numerous other strains were subsequently identified in animals, including pigs (TGEV, PRCoV), cows (BCoV), chickens (IBV) and cats (FIPV) (Weiss et al. Microbiol Mol Biol Rev 2005;69:635-664). They received the name “coronavirus” in 1968 due to their crown-like appearance; in 1975 they were categorized as three genera (I, II, III) based on serological cross-reactivity and, subsequently, genome sequencing. At that time, two HCoV strains (229E in group I, OC43 in group II) were known as two of the viruses that caused the common cold but were not associated with serious illness in humans.
Throughout the 1980s, there was considerable interest in identifying an infectious cause of MS and HCoV was investigated. An early theory was that an impaired antiviral response contributed to MS pathogenesis (Neighbour & Bloom. Proc Natl Acad Sci USA 1979;76:476-480), which provided a rationale for investigating interferons (alfa, beta and gamma) as antiviral agents in MS. Another was molecular mimicry, and HCoV 229E-reactive T cells were shown to be more cross-reactive with myelin basic protein (MBP)-reactive T cells in MS patients versus healthy controls (29% vs. 1.3%) (Talbot et al. Ann Neurol 1996;39:233-240).
One issue was the neurotropic potential of coronaviruses. HCoV was isolated from brain tissue samples in autopsies of MS patients (Burks et al. Science 1980;209:933-934. Tanaka et al. J Neurol Sci 1976;28:121-126), and HCoV RNA and viral gene expression were reported in the brain and CSF of MS patients (Murray et al. Ann Neurol 1992;31:525-533). HCoV antigen and RNA have been detected in active demyelinating plaques in MS patients (Murray et al. Ann Neurol 1992;31:525-533). Laboratory studies have also reported that HCoV (229E, OC43) can infect human astrocytes, microglia and oligodendrocytes (Bonavia et al. J Virol 1997;71:800-806. Talbot et al. Adv Exp Med Biol 1993;342:339-346).
HCoV is ubiquitous so its isolation in MS patients may be akin to the colonization seen with other viruses, such as Epstein-Barr or herpesvirus. To date, the presence of HCoV in the CNS has not been shown to be associated with a worse clinical course in MS. A Canadian-UK study found that respiratory coronaviruses (e.g. OC43) have the potential to adapt to the CNS environment, although that study found no association between viral RNA and worsening MS plaques (Arbour et al. J Virol 2000;74:8913-8921).
What makes coronaviruses an emerging pathogen is the increasing virulence of some strains. This was first seen in the SARS-CoV (severe acute respiratory syndrome) outbreak in 2003. Since that time, several pathogenic coronavirus strains have been identified in humans: HKU1 in a case of pneumonia (Woo et al. J Virol 2005;79:884-895); NL63 in cases reported in Canada and elsewhere (Bastien et al. J Infect Dis 2005;191:503-506); MERS-CoV (Middle East respiratory syndrome coronavirus), which was first identified in a patient with severe respiratory symptoms transferred from the Gulf states to the UK (Bermingham et al. Euro Surveill 2012;17:20290); and now SARS-CoV-2 (COVID-19).
While little is known about the pathophysiology and treatment of SARS-CoV-2, some lessons may be learned from SARS-CoV and other coronaviruses. During the acute phase of SARS-CoV infection, there was a rapid depletion of lymphocytes, including CD4+ and CD8+ T cells (Lin et al. Emerg Microbes Infect 2020:1-14). CD8+ cells recovered within 2-3 months followed by CD4+ cells. Memory CD4+ cells normalized but other T cell subsets remained below baseline at 1 year. IgG antibodies to SARS appeared at about 2 weeks in the later acute phase. IgG levels were found to be significantly higher in patients with mild disease compared to those with severe disease.
With COVID-19 infection, WBC has been reported to be normal or slightly low with some lymphopenia observed. More significant lymphopenia appears to be associated with more severe disease (Wang et al. JAMA 2020; epublished February 7, 2020). This was seen with SARS, in which a rapid reduction in CD8+ and CD4+ T cells was associated with worse clinical outcomes (Li et al. J Infect Dis 2004;189:648-651). The mechanisms underlying lymphopenia have not been fully determined but may occur by induction of apoptosis by viral proteins rather than by direct viral infection of lymphocytes (Weiss 2005).
During the initial acute phase of COVID-19 infection, a robust T cell response appears to be important in limiting severity. This is supported by the observation that HCoV OC43 establishes infection by interacting with MHC class I molecules (Collins AR. Immunol Invest 1994;23:313-321), suggesting the importance of an intact CD8+ response.
B cells do not appear to play a significant role during the early infection phase. With SARS, SARS-specific IgG antibodies appeared in the late-acute phase and increased thereafter (Li Gang et al. J Trop Med 2003;3:283-285). A persistent antibody response and higher IgG levels were associated with improved outcomes (Cao et al. Virol J 2010;7:299), suggesting that an intact B cell response may be needed to limit viral replication and prevent tissue damage.
B cell depletion will also reduce the utility of vaccination if a COVID-19 vaccine is developed. In the unpublished VELOCE study, there was significant attenuation of the vaccine response in patients treated with ocrelizumab (tetanus toxoid 23.9% vs. 54.5% with controls) (Stokmaier et al. CMSC 2018; DX43).
These observations provide a rationale for the Canadian Network of MS Clinics’ treatment recommendations during the COVID pandemic (see MS treatment recommendations during COVID-19 pandemic). Caution is advised in initiating or maintaining DMTs that produce lymphopenia, notably the cell-depleting therapies cladribine, ocrelizumab and alemtuzumab. All have prolonged effects on immune function so clinicians have the option of postponing the next dose to allow some degree of lymphocyte recovery; this approach will be difficult if the pandemic persists for the next six months. Fingolimod depletes some T cell subsets (CD4+ > CD8+); vaccine studies have reported a diminished but sufficient response to novel antigens (Kappos et al. Neurology 2015;84:872-879). The current recommendation is to maintain fingolimod, with patients instructed to call their neurologist if they develop COVID-19 infection.