In 1991, a storm struck North America’s eastern seaboard. Only this was a storm like no other. There was a strange set of circumstances that built this storm, with conditions rarely ever seen independently, let alone together. There was warm air that was moving in from a general low pressure area over the Atlantic Ocean, cold air coming in from a high pressure area closer to the shore and an unusual level of moisture in the air. This unique set of weather conditions led to what Sebastian Junger called, “the perfect storm” in his book of the same title.

A satellite image of the 1991 “Perfect Storm”
This occult combination of circumstances led to an Air National Guard helicopter crashing into the ocean, and a sword fishing boat going down, with all hands lost.
Similarly, there are certain conditions within some COVID-19 patients with can rustle up a different kind of storm; a cytokine storm.
What is a cytokine storm?
Cytokines are soluble factors that are produced to direct the aggressiveness and nature of the body’s immune response against a pathogen. In essence, a cytokine storm occurs when the immune system produces excessive, uncontrolled quantities of cytokines in response to a pathogen, which can lead to severe conditions, such as Acute Respiratory Distress Syndrome (ARDS) and multi-organs system failure.
According to WHO reports on the first outbreak in China, 13.8% of COVID-19 patients had severe responses to the disease, with difficulty breathing and blood oxygen saturation below 93% (blood oxygen saturation is ≥ 95% in healthy individuals). 6.1% of patients were classed as critical cases, with respiratory failure, septic shock and/or multi organ system failure.
A probable cause of these severe and critical clinical manifestations of the disease is cytokine storm; it can lead to respiratory failure, septic shock and multi organ system failure1. In fact, some of the severe cases of COVID-19 might be experiencing early signs of cytokine storm, and perhaps determining if this is the case might be helpful.
How does SARS-CoV-2 cause cytokine storms?
As I explained in my previous post, viral particles released by infected cells can be detected by alveolar macrophages (a white blood cell which is a part of the innate immune system). If it has been primed by interferon gamma, it can become hyperactivated, and will produce TNF-A, IL-1 and IL-6. These are all cytokines. The hyperactivated macrophages will also secrete chemokines, a type of cytokine that serves as a signal to attract specific cell types to the source. In this case, the CCL2, 3 and 5 chemokines act on circulating neutrophils and monocytes and pull them towards the site of infection, from the blood in neighbouring capillaries.
Elevated cytokine and chemokine levels in the lungs of SARS-CoV-1 patients2 is linked to increased neutrophil and monocyte levels in the lungs. This encourages a positive feedback cycle, because when these neutrophils and monocytes enter the lungs and encounter the viral particles, they too will secrete pro-inflammatory cytokines, mainly comprising of the same trinity we encountered earlier; TNF-A, IL-1 and IL-6. This leads to cytokine levels building up, potentially to very high levels.
To circle back chronologically, let us consider those primed macrophages, which were ready to be activated by the detection of viral particles. What primed them? In this case, it seems likely that natural killer cells (NKs) might have done the job. When NK cells encounter viral particles, they produce interferon gamma which hyperactivates the macrophages. They also produce interferon alpha and beta!
These two interferons are quite important at the start of an infection. This is because their binding to interferon receptors on infected cells triggers the “suicide” of the infected cell. If it encounters an uninfected cell, it “primes” the cell, such that when it does become infected, it will commit suicide. Therefore, early on in an infection, interferons help to clear off a fair number of infected cells, which at the time, would constitute a larger portion of the viral load. Quite an elegant system (when it works).
A problematic situation might develop if interferon alpha and beta are only released in the later stages of an infection, when the monocytes and neutrophils have already entered the lungs. These interferons act on the monocytes and neutrophils to trigger the production of more chemokines (like CCL2) that bring even more monocytes and neutrophils in. This exacerbates the issue of overproduction of cytokines. Therefore, it would appear that the timing of interferon release is quite important; early on, it has a greater chance of slaying the virus, whereas in later stages, this might come at the cost of rampant cytokine production.
In addition, it was found that COVID-19 patients requiring ICU admission had higher levels of CCL-2 and TNF-A compared to COVID-19 patients with less severe disease.
This paints the picture of a physiological landscape filling up with excessive levels of cytokines, in an uncontrolled fashion.
Why is this so dangerous?
In the words of Jim Morrison, with such high levels of cytokines, “there’s a killer on the road.” The riders on this chemical storm enter the pulmonary circulation, which collects newly oxygenated blood from the lungs, and delivers it to the heart. From here, the blood containing the cytokines is pumped throughout the body. These cytokines are now placed to affect any organs that are perfused by the systemic circulation, and to exert their key effects.
IL-1 and TNF-A trigger vasodilatation of blood vessels, widening them, and thus decreasing the resistance against the flow of blood. This therefore drops the total peripheral resistance of the body’s circulation. These 2 cytokines also increase capillary permeability, allowing more fluid from the capillaries to enter tissues, decreasing the volume of the fluid in the circulation.
These 2 changes have quite a significant impact on arterial blood pressure. The drop in total peripheral resistance directly drives a decrease in arterial blood pressure, because arterial blood pressure is equal to total peripheral resistance multiplied by cardiac output!
The drop in circulating blood volume decreases the stroke volume of the heart (the volume pumped out of the heart with each contraction), which in turn decreases the cardiac output, driving a further decrease in arterial blood pressure.
The drop in arterial blood pressure, if severe, is very dangerous. It essentially reduces the perfusion of the organs in the body, and this compromises their abilities to carry out their various tasks. They become starved of oxygen and other key nutrients, leading to multi organs system failure.
Chemokines and Neutrophils
As I mentioned earlier, chemokines act as chemoattractants for inflammatory cells. Neutrophils that have infiltrated the lungs, upon detection of the presence of viral particles, release CXCL-10, which helps to recruit even more inflammatory cells in , through a positive feedback cycle. This particular chemokine is closely linked to the manifestation of ARDS, and high serum levels of CXCL-10 are found in patients with severe cases of COVID-19.
What is interesting here is that there may be a causative link between the damage that comes with ARDS and the release of CXCL-10. A study in mice found that if one were to induce ARDS (via the injection of an endotoxin), there would be higher serum levels of CXCL-10, and the administration of an anti CXCL-10 antibody treatment actually reduced the damage that the ARDS caused to the lungs.
It’s an interesting point in favour of using anti CXCL-10 antibodies to treat ARDS in COVID-19 patients, but obviously, what may be true in a mice might not hold true in humans, so further research is needed.
IL-6
IL-6 actually has quite a fascinating activation pattern; although it is produced by macrophages in tandem with IL-1 and TNF-A, both of these cytokines serve to increase production of IL-6 by white blood cells, driving a further increase in IL-6 levels.
IL-6 serves to trigger the formation of Th17 helper T cells from naïve helper T cells, and these Th17 cells produce IL-17, which recruits more neutrophils to the site of infection (which is most likely to be the lungs at the start of a SARS-CoV-2 infection). As we’ve established, this isn’t necessarily a great thing, because the neutrophils both directly and indirectly drive further cytokine production.
IL-6 also plays a role in triggering a litany of pro-inflammatory events, contributing to the terrible outcomes associated with cytokine storm. In fact, serum levels of IL-6 in COVID-19 were found to be correlated with the severity of the disease, potentially reflecting the magnitude of its role in cytokine storm. IL-6 also has some interesting effects on the liver, such as ramping up production of C-reactive protein (CRP). This can act as a useful biomarker of cytokine storm in a clinical setting, but further tests would still be necessary to confirm the diagnosis.
Tocilizumab
In light of IL-6’s involvement in COVID-19 induced cytokine storms, a drug previously approved to treat cytokine storm coming about from CAR-T therapy, the “wonder” immunotherapy for cancer (which isn’t quite as wonderful as it would seem), is being investigated for treating COVID-19 patients. The drug, known as tocilizumab, is a humanised anti IL-6 receptor antibody, which binds to IL-6 receptors to inhibit downstream signal transduction. This has the effect of inhibiting the pro-inflammatory effects of IL-6.
An analysis published in the Lancet on 24 June 2020 looked at the results of a retrospective study that compared tocilizumab treatment with the standard of care (which included hydroxychloroquine) across 3 tertiary care centres in Italy. The study was specifically concerned with 544 COVID-19 patients with severe pneumonia, so this was looking at patients that fit within the WHO’s definitions of “severe” and “critical” cases. Interestingly, there was a statistically significant difference in deaths between the tocilizumab group (7%) and the standard care group (20%), with a p value < 0.0001. In fact, the group that received tocilizumab intravenously had a hazard ratio of 0.55 compared to the standard care group when invasive ventilation and death were considered as a single composite outcome (p value = 0.044).
Admittedly, I was quite excited by this paper when it was first announced, because the authors initially planned to use the Sequential Organ Failure Assessment (SOFA) as one of their outcome measures, which would help to establish a stronger link to cytokine storm suppression. Unfortunately, they replaced this metric with a more generic comorbidity index in the final results. Therefore, these results don’t strictly show if tocilizumab was working to improve outcomes by disrupting cytokine storms.
There are also some caveats to tocilizumab use that we have to consider. First, by actively suppressing the immune system, there is going to be an increased risk of secondary infection, as this study proved, showing that 13% of the tocilizumab treatment arm were diagnosed with new infections, versus 4% for the standard treatment arm. Secondly, tocilizumab has some prominent side effects, such as hypertension and headaches.
Admittedly, the theory supporting the use of tocilizumab, and indeed this paper, makes the idea of administering it to treat possible COVID-19 induced cytokine storm quite seductive. However, we have to be aware that there still is not a large enough body of evidence that supports or discourages its use to treat COVID-19 induced cytokine storm. As with many facets of COVID-19, more data is needed.
Corticosteroids
Another class of drugs that have been put forward as a possible treatment line for the overactive immune response to COVID-19 are corticosteroids. These drugs have the capacity to decrease interleukin production, so in theory, this could be helpful for treating cytokine storm in COVID-19 patients.
However, a systematic review in the European Respiratory Journal looking at 771 publications found that there was no significant evidence that the use of inhaled corticosteroids directly led to beneficial outcomes with regards to the treatment of ARDS (which can be a manifestation of cytokine storm) in COVID-19 patients.
On the hand, a cohort study completed just earlier this month compared ICU admissions and in-hospital deaths between a group of COVID-19 patients (whose disease had progressed to ARDS) that was given methylprednisolone (MP), and a group that was not given methylprednisolone. 18.1% of the group treated with methylprednisolone were admitted to the ICU over the course of the study (versus 30% in the non-MP group). 7.2% of the MP treated arm of the study died in hospital, whereas 23.3% of the non-MP treated group died in hospital. This suggests that intravenous MP might bring about positive improvement in outcomes for COVID-19 induced ARDS, which occurs in part due to overactive cytokine production.
Conversely, we need to consider the fact that this study only included 173 patients, and a full review of the data has not been completed yet; I’ve been eyeballing the raw data thus far. Further statistical analysis is definitely needed, preferably as part of a larger meta-analysis.
I’d be remiss not to discuss dexamethasone, another corticosteroid that has been widely reported to be effective in the treatment of COVID-19 patients with severe infections. It’s important to note that the RECOVERY trial which tested its efficacy only showed this benefit for critically ill patients, with severe complications including cytokine storm. There’s no need to stock up on dexamethasone for your house!
To sum up
It’s strange to think that the SARS-CoV-2 virus itself does not directly cause some of the most severe symptoms associated COVID-19. A lot of the damage comes from the immune system activating excessive pro-inflammatory responses through uncontrolled cytokine production, leading to ARDS and, possibly, multi-organs system failure.
I suppose a question that might be worth looking into is why some individuals are more vulnerable to COVID-19 induced cytokine storms relative to others. Perhaps this might have to do with higher TLR (toll-like-receptors) or PRR (pattern recognition receptors) expression due to a preceding infection, allowing for an overly robust response to SARS-CoV-2 viral particles.
However, even then, a major challenge remains with pre-empting a cytokine storm. Timing immunosuppression would have to be key; too early and the virus gains a foothold early, too late, and the cytokine storm begins its deadly rampage.
Identifying the right moment to act will be key.
Footnotes
1 There are other mechanisms by which COVID-19 can cause multi-organ system failure, but I’m restricting my discussion to cytokine storm induced multi-organ system failure
2 There’s not a lot of research that has been done on this specific relationship for SARS-CoV-2 (that I could find anyway!)
Further Reading
- Tocilizumab in treating COVID-19 – https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7118634/
- Systematic Review on Inhaled Corticosteroid Use – https://erj.ersjournals.com/content/early/2020/04/20/13993003.01009-2020
- Methylprednisolone study – https://clinicaltrials.gov/ct2/show/results/NCT04323592
- WHO report on the first outbreak in China – https://www.who.int/docs/default-source/coronaviruse/who-china-joint-mission-on-covid-19-final-report.pdf
- A discussion on COVID-19, cytokine storms and how we can treat cytokine storm – https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7194613/pdf/main.pdf