After many months of studying severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) – the pathogen responsible for the ongoing coronavirus disease 2019 (COVID-19) pandemic – much is still unknown. However, evidence points to the occurrence of a cytokine storm in severe and critical cases of COVID-19. This is not unique to COVID-19 and has occurred in the majority of severe coronavirus and influenza epidemics.
However, there are some striking differences in the cytokine release syndrome (CRS) in these conditions and that which occurs in COVID-19. A new study published on the preprint server medRxiv* in November 2020 reports some of the unique elements of COVID-19-induced CRS and how understanding these distinctive characteristics of the disease could help in formulating specific interventions and therapies.
A cytokine storm, or CRS, refers to a hyper-inflammatory response involving the dysregulated activation of a large number of immune and inflammatory cells; these cells pour out a flood of pro-inflammatory cytokines. This sets up a positive feedback loop or a vicious cycle of inflammation. Cytokine storms are associated with higher rates of mortality in severe and critical COVID-19 cases.
Earlier studies have described the process as beginning with cytokine-induced widespread tissue damage, followed by multi-organ dysfunction and death in the absence of rapid intervention.
CRS resolution is a typical sequel, and is characterized by high levels of anti-inflammatory cytokines, such as the type I interferon (IFN) signaling cascades mediated by IFN-α and IFN-β. These, in turn, activate IL-12 and IFN-γ, which is a type II IFN. Type I IFN pathways are inhibited by IL-10, and certain viral proteins, which can delay or modulate this response, as observed in severe COVID-19.
Looking for differences in CRS in five viral infections
The fact remains that the immune response in different viral infections, including SARS-CoV and MERS-CoV – two betacoronaviruses in the same family as SARS-CoV-2 – as well as and influenza A, is not specifically understood, especially in their early stages. Such knowledge could help prevent a disease’s progression to its severe stage and the development of CRS.
The current study focuses on the differences in the cytokines released by these viruses. The researchers used available prepublished and published literature for information on 98 cytokines, including interferons, interleukins, tumor growth factors and chemokines associated with CRS.
They examined especially changes in these molecules in patients infected by five important CRS-causing viruses: the two influenza A virus subtypes H5N1 and 91 H7N9, SARS-CoV, MERS-CoV and SARS-CoV-2. They ended up with 38 significant cytokines that had been measured in actual patients.
While 16 cytokines were raised in a specific viral infection in at least one study, only five were raised in all five viral infections. This does not consider the magnitude of change, however.
Disruption of immune response in betacoronaviruses (beta-CoVs)
They observed that all measured cytokines were elevated in influenza A infections, but with the betacoronaviruses (beta-CoVs), the pattern was more nuanced, with some cytokines at baseline levels. The cytokine response in COVID-19 is midway between that of the other beta-CoVs and the influenza A viruses. They concluded that this indicated the ability of beta-CoVs to evade the immune response.
There were eight cytokines clusters based on the direction of change in their secretion following infection by each of these viruses. The first cluster, I, contains TNF-α with IL-2 and IL-10. The first is pro-inflammatory and the others anti-inflammatory. A rise was seen only in influenza.
Similarly, clusters III and VI, containing pro-inflammatory cytokines like IL-6, type I and II IFNs, and several chemokines, generally recorded a rise. Cluster IV includes cytokines like IL-4 and IL-5, mostly unchanged in CoV infections but raised in influenza. IL-4 plays a role in Th2 differentiation, and the latter cells can secrete IL-5 to modulate the recruitment of eosinophils.
With cluster VII and VIII, IL-15 and CCL5 are unchanged in COVID-19, but the former is involved in NK cell differentiation as part of the innate immune antiviral response. CCL5 is responsible for eosinophil infiltration, a key process in post-COVID-19 recovery. Cluster II includes cytokines that were measured only in H7N9, and cluster V only in COVID-19.
CRS in beta-CoV infections
COVID-19 is associated with a beta-CoV-like response in terms of pro-inflammatory cytokines, but preserves some, like IL-2 and IL-10, that would be expected to be raised in a viral infection. Thus, SARS-CoV-2 induces a markedly weaker type I IFN response than the other beta-CoVs or the influenza A viruses.
With SARS, both type I IFN and the downstream IL-12 are activated due to the involvement of mature dendritic cells, with IL-12 indirectly activating IFN-γ, but not IL-10. This could perpetuate the positive feedback loop CRS causes.
In MERS, type I IFN is induced in only some individuals, but not IL-12, IL-10 or IFN-γ. Strangely, IL-10, which was formerly thought to be anti-inflammatory, is now being shown to have pro-inflammatory activity.
Both these beta-CoV infections induce impaired innate immune responses allowing alveolar macrophages or neutrophils to be infected, causing increasingly severe lung injury. The persistent pro-inflammatory cytokine state causes acute respiratory distress syndrome (ARDS) and severe lung damage, leading to higher mortality rates. In fact, SARS was associated with death in around 10% of patients, and MERS in around 34%, compared with 2.3% in COVID-19.
Again, in severe SARS, IL-10 levels are very low but high in MERS. Despite this, in MERS, low IL-4 and IL-2 levels release IFN-γ from inhibition, and high IFN-γ titers induce type II IFN levels.
CRS in influenza A
In influenza A, however, the antiviral response is rapid, with an intact negative feedback loop inhibiting excessive inflammation. IFN-I is modulated by viral proteins, but remains intact and can be hyperactivated to cause mortality in severe influenza. Similarly, failure to trigger TGF-β can increase the severity of illness.
A curious finding is that while the cytokines that regulate inflammatory cascades are enabled in influenza A, CRS can still occur. This may be the result of TGF-β deficit, with, additionally, a characteristic impairment of CD4 and CD8 T cells, despite their abundant number, in CRS. Again, the typical anti-inflammatory phenotype of monocytes that is expected to occur at high antigen presentation, as in progressive infection, fails to manifest. The monocytes thus remain in a chronic pro-inflammatory activation state, which prevents the host response from subsiding.
Further study would help understand how CRS occurs in severe influenza, by outlining the population expansion and activation profile of both innate and adaptive immune cells.
What are the implications?
The primary difference between the beta-CoVs and influenza viruses seems to be that while resolution of CRS is poor in the former, the immune system in the latter infection remains capable of resolving the cytokine storm.
Cytokine storm cell signaling vector horizontal background. Image Credit: Whitedragon / Shutterstock
The unique feature of SARS-CoV-2 is the disruption of downstream signaling after type I IFN activation. This upsets the equilibrium of inflammatory and anti-inflammatory cytokines. In about 80% of cases, however, the inflammation does resolve. Here again, SARS-CoV-2 stands out from the other beta-CoVs in the high degree of CRS resolution.
In short, the researchers say, “SARS-CoV-2-mediated infections are characterized by a clear dysregulation of type-I IFN response, and consequently the downstream cytokine signature such as IL-4, IL-12, IL-2, IL-10 and the downstream type II IFN response.”
By helping to map these responses, it may be possible to develop therapeutics that reduce the severity of CRS these infections, or even prevent its development.
medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.