Coronavirus: the infection paradox

05/05/2020

There are many things which are puzzling about this Covid-19 epidemic – and there are few people who would claim that we know all there is to know about it, or that we are even close to understanding the dynamics of the disease.

But, before one can make any sense of it, there must be an understanding of the characteristics of the virus which affect the individual and then then as to how they shape the profile of the epidemic. This is the essential difference between the practice of medicine, per se, and public health. In public health practice, one is more concerned with the latter.

Here, looking at the way this epidemic panning out, there seems to be an essential mismatch between the number of people affected by the disease (and the death rate), as against the projections for what is considered to be a highly infectious disease.

All things being equal – and especially given the incompetence of respective governments in their handling of this pandemic – one would have expected more cases, with a very much higher death rate. Certainly, the high death rate has not materialised on a global scale.

Trying to explain this is not easy but, if we go back to the original SARS pandemic of 2002-3 – which also failed to live up to its promise as a major killer – there was one particular characteristic which seems to be repeating itself: the most seriously affected victims seemed to be family contacts and healthcare workers who had treated infected patients.

Possibly – just possibly – this might give us a clue as to what we are currently dealing with, and give us some important pointers as to how we define our ongoing response.

Before launching a hypothesis, however, it is necessary to clarify by way of definitions, the nature of a pathogen. In generic terms, this is any organism which has the potential to cause disease. But, in order for it to be responsible for an infectious disease, that capability requires the presence of three distinct features.

Firstly, the organism must be communicable – it must be capable of being spread, either by being carried through the air, by water, food or other means. That, combined with the portal of entry (to the human body) and its normal site of action, will determine how easily it is spread.

This is best illustrated by an infectious disease which doesn't tend to spread very well – Legionnaires' disease, caused by the bacilli Legionella pneumophila. This, as the species name implies, causes a form of pneumonia, which has a relatively high mortality.

However, where this organism has gone wrong on the evolutionary scale is that its primary reservoir is water. Yet, to cause illness, it must normally be breathed in by its victim, in order to gain access to the lungs. If, like typhoid – which is also a waterborne disease – it acted on the gastro-enteric tract, and thus give rise to illness when contaminated water was drunk, it would have a much easier life.

Thus, in Legionella pneumophila, we have a potent pathogen but, because it is not particularly communicable, it relatively rarely causes outbreaks. When they occur, they are most often associated with droplet fallout from air conditioning systems – nature's rather devious way of getting people to breathe in water.

The second characteristic for an aspiring pathogen is infectivity – the ability to penetrate the human body, and to gain access to the site where it will exert its effects, so as to cause disease.

This, however, is not to be confused with the third essential property – that of virulence, which relates to the ability to cause the disease when the other conditions are satisfied.

A surprising number of medics don't actually understand the difference: I once spent six weeks of my life in a Scottish Sheriff's court, where a case over the determination of the unfitness of food (in this case Lanark Blue cheese) rested on whether Listeria monocytogenes (Lm) was necessarily a pathogen.

I found myself up against the head of the Listeria Reference Laboratory, who argued that because the organism was known to cause illness and, in this case, was infective, it was a pathogen and therefore harmful to health.

However, there was ample evidence to show that while Lm was in most cases, infective, only a few – relatively rare - strains had a particular gene (lacking in this case) which switched on its virulence, enabling it to cause disease. This explains why Lm is a very common organism but rarely gives rise to outbreaks. (We won our case).

You can thus have a condition where an organism enters the bloodstream, and can be detected, but it doesn't cause disease. With bacteria, this is known as a bacteraemia and, when viruses are involved, it is known as a viraemia.

And this is where I think we might be with the Covid-19 epidemic. Clearly, the SARS-Cov-2 virus is readily communicable, and can easily pass from one human being to another. That it is also highly infective is also without dispute – as measured by the fabled R0 which, in some iterations has been recorded at a level of 5.7.

However, I am beginning to suspect that – as a generality – the organism is not particularly virulent. This is most recently borne out by German work which suggests that ten times more people in the town of Gangelt in Heinsberg municipality had been infected than had previously been thought – with most of those infections being asymptomatic.

This actually puts the case fatality rate at 0.37 percent – considerably lower than some estimates, and roughly on a par with seasonal flu.

Therein lies the paradox, as we see in the UK and elsewhere massive increases over a short period in excess mortality – apparently undeniable evidence as to the lethality of the organism. But it is there that the experience of family contacts and healthcare workers might be helpful.

When we first heard of the death of the Chinese whistleblower doctor in Wuhan – despite his relatively young age and lack of underlying health conditions – I speculated that there might be a dose response involved. The physician had received such a high dose of the virus that it had overwhelmed his innate defences.

In fact, most likely, he had received not one dose but many – having treated multiple patients. This might apply to most healthcare workers who have succumbed to the disease. When it comes to family contacts, it is not unreasonable to hypothesise that a similar dynamic is at play. In an intimate family environment, the susceptible victim may receive not one but many viral doses from the infected family member.

Much the same might be happening in care homes, where the close mixing of residents over a period of time, with a virus excreter in their midst, may ensure that they receive multiple doses – the cumulative viral load being far higher than any one person might experience through a single contact.

This could also explain why the outcome for so many patients referred to hospitals has been so poor. Cases who acquire their infection in the community are then, in their weakened condition, given further doses of infection which add to the challenge, and intensify the severity of the disease.

While this phenomenon may not be responsible for all episodes of the disease, it could certainly explain why an organism which is so apparently infectious, with the potential to do such great harm, has not caused more havoc than it has. It is, in a sense, more of an opportunistic pathogen, requiring specific conditions before it can exert its maximum lethality.

This does not gainsay the other conditions that might also apply - such as underlying health conditions - but when one sees very elderly people surviving the illness, it does seem to suggest that we are dealing with a complex pathogenicity which does not mirror flu or other analogues.

Should this be the case, then we need to be looking at the epidemiology of this disease in a somewhat different fashion. Arguably, transient contacts with viral carriers – and especially in environments where viral shed is rapidly dispersed - may be of little significance. Many of the social distancing rituals being proposed may be irrelevant.

Multiple exposure may also have implications for the treatment of patients – suggesting that they should be housed in a contamination-free environment, with effective barrier nursing to prevent healthcare staff re-infecting their patients, or adding to the dose to which they are exposed.

Either way, the characteristics of this disease suggest that the traditional flu model is less and less appropriate. It also calls into question Mr Hancock's miracle app (Times paywall), where tracking down all contacts indiscriminately may be a complete waste of time and resource.

Certainly, the concept of a multiple exposure infection is not new to epidemiology, and has been noted as a contributory factor in the aetiology of tuberculosis (in its respiratory form) and leprosy. Of more relevance, it has been associated with the spread of influenza in hospitals.

If it is an explanation for the Covid-19 paradox, then we really do have some rethinking to do.