David Matthews Interview – UK’s poor pandemic response has been shaped by a lack of specific expertise about coronaviruses

In my latest in-depth coronavirus interview, I was joined by Dr David Matthews, a reader in Virology at Bristol University. We discussed the reasons for the government’s failed pandemic response. Matthews argues that what he calls the “flu playbook” prepared us poorly for a coronavirus pandemic. He also makes a powerful case for learning more about viruses in general – this pandemic isn’t the first and won’t be last time a new disease emerges. Next time, we need to be prepared, he tells me.

Alastair Benn: What did SAGE get wrong early on in our pandemic response?

David Matthews: Take the issue of whether or not we should have screened people from flying from Wuhan into the UK early on. The standard answer was: “It’s pointless because if you catch the virus just before you get on the plane, we won’t be able to detect it with our methods for four or five days.” Well, that’s a debatable point. Half of the people who have the virus don’t show any symptoms. So you pick them up. And if it’s so completely pointless to screen people at airports, why did we do it for the Ebola outbreak? At the time, people coming back from West Africa were all screened. Public Health England put a lot of effort into it.

AB: What other mistakes were made in the first stages?

DM: There are large aspects of our response that I did not understand. It only really made sense in the context of a flu pandemic playbook which has been around for years. The impression I got was that they picked it up and started reading from page one and followed the instructions. First of all, this isn’t flu. Secondly, even if you accept that it is like flu, there are still things that I find baffling.

The government kept saying in the beginning that risk to the UK was low or medium. If the reproduction number is 3, then the risk to the UK is certain. It’s not low, or medium or maybe or probable or even very high. It is 100% certain that the virus is coming. In 2009, we had the Swine flu pandemic that went round the globe pretty much as the models would tell you it would go round the globe. I thought it was set in stone that any respiratory virus with an R0 of three would infect somewhere between a quarter and a third of the globe in about a year.

AB: What would a coronavirus playbook look like?

DM: Countries like China and Hong Kong, which experienced SARS full on, learnt that you need to stockpile PPE because you’re going to need a lot of it in order to deal with patients. The flu playbook is informed by cultural underlying assumptions we have about a flu pandemic. There are bunch of things we know about flu that we don’t know about coronaviruses.

With flu, you know there will be a vaccine in four to six months because we have vaccines for flu. We have manufacturing facilities for flu vaccines and we distribute them at large scale every year. We know that the cavalry is on its way as soon as the pandemic starts. In the meantime, we have drugs against flu, drugs that actually work. The list goes on. We know how to test for flu at scale. We have lots of different types of tests for flu. We have tests that show if people have been exposed to flu. We know how flu immunity works. We know how long that immunity will last. We even know what aspects of the genetic makeup of a particular strain of flu might make it more or less dangerous to humans.

We have tonnes and tonnes of information and infrastructure around flu. None of that exists for coronaviruses.

We know very little about the immunology of coronaviruses. We know relatively little about repeat reinfections. We have never had to test at scale for coronaviruses. We have never had to test at scale for coronavirus antibodies. There are no vaccines for coronaviruses and there are no drugs. Dexamethasone is not a drug against the virus – it is a drug against the immune response. All these things are missing.

We needed a playbook for a virus where you assume a standing start, where you have nothing. That may have been the problem. The flu pandemic playbook reflects in subtle ways deep underlying assumptions about how that kind of pandemic works. If you look at who dominates SAGE and NERVTAG in terms of the virus expertise, it is all flu people. A Corona-virologist was not brought in until April.

AB: What about the other coronaviruses? When did they come on the scene?

DM: The other coronaviruses have been around for a very long time. There’s a strain called OC43. A paper came out a few years ago that theorised that OC43 was a bovine coronavirus that jumped into humans around the 1880s and 90s. There was a global outbreak then of what was thought to be flu. The team who produced the paper did an analysis of the genetic makeup of OC43, wound back the evolutionary clock and came upon the explanation that OC43 jumped from cows to humans in the late 1800s. At that time there was an outbreak of bovine pneumonia which wiped out loads of herds of cattle that was closely followed by what was thought to be an outbreak of flu.

There’s no reason to suggest that that couldn’t be true. If you look for coronaviruses in cows you will find some. If you look for coronaviruses in humans you will find some. If you look in pretty much any mammal or in fact pretty much any vertebrate – snakes, frogs even –  you will find some. Every type of virus we know of in humans exists in other vertebrates. I’ve worked on adenovirus for most of my life. There are human adenoviruses. There are snake adenoviruses. There are frog adenoviruses.

AB: What’s the spark that sets the fire ablaze. How do viruses make the jump?

DM: Probably, these sorts of events are happening quite a lot. A virus that has successfully established itself in one species occasionally makes the jumps into others. If the virus can make a living in its new host, then it will do. If, in its business of replicating in its new host, say in jumping from cows to humans, that virus does enough damage and makes enough virus in your nose or in your mouth so that you can cough or sneeze the virus onto someone else, then that will sustain the infection.

The question is – what are the rules governing that? While humans and cows are both mammals, we do have a lot of differences. Because the virus infects individual cells in your body and it interacts with and uses all the machinery in your own cells, depending on how that virus replicates inside the cells, if the levers of power it needs are similar enough in humans, then the virus gets going and will become a success.

Viruses can mutate at quite a quick rate. If it turns out that when the virus replicates its own genetic instructions it will make some mistakes; if it turns out that some of those mistakes lead to progeny viruses that are better able to interact with proteins and levers of power inside your own cells, then that virus will get better at it.

That’s probably what happened when it jumped from cows to people. The cow would have been snotty, would have infected somebody. That virus would have infected somebody, then replicated in that individual, then maybe a small part of that virus would have emerged that would have been better able to infect humans, and that virus would have spread to someone else…

The second person would have been infected by hundreds and thousands of virus particles, some of which would have been genetically slightly different and better able to replicate in humans. They would start to dominate as they can replicate faster.

AB: What’s the best metaphor for the life of a virus?

DM: It’s a parasite, an extreme parasite. It carries very little information of its own, just enough to make copies of its own genetic instructions and form a protective particle around itself. Take foot and mouth disease, which ran through cattle a couple of decades ago in this country. It didn’t infect anybody. The proteins it finds in cows are presumably just too different in humans and so nothing happens. Your genetic makeup simply doesn’t interact with the virus so the virus replication doesn’t get off the ground.

AB: Do we have some protection from the novel coronavirus by virtue of exposure to the other coronaviruses?

DM: There is an argument that there are parts of the other coronaviruses that cause the common cold which look similar enough to this coronavirus that your immune response to the common cold coronaviruses might help you in a fight against this new coronavirus. That’s conceivable but it’s difficult to know how big a role that is playing.

AB: What do you expect will happen to SARS-CoV-2?

DM: Classically, what you would expect when a new virus establishes itself in a new population, that virus should become less dangerous. Less lethal versions of the virus spread further and faster in the population than lethal ones. The downside to that is that one way to become less lethal is to become worse at replicating, making it harder to spread. There’s a trade-off there.

The basis of that thinking comes from when we tried to control rabbits by using Myxomatosis in Australia. When they first introduced it, it was hugely lethal to rabbits. It had a 90 per cent fatality rate among rabbits. Over time that fatality rate fell quite dramatically. Rabbits became more resistant. They were genetically harder for the virus to make a living in. Those two things came together and now myxomatosis is endemic but it doesn’t have spectacular effects.

The difference here is that this virus is not 98% lethal: 98% of people survive it. It’s already not hugely lethal. There’s not really huge pressure on this virus to become less lethal. In people who die, they don’t catch the virus and die very quickly. They clearly are able to spread it around a fair bit before they become seriously ill.

What I expect to happen is several things – from a clinical management point of view, we will get better at handling patients. That’s happened already. That comes from experience. We have drugs that help us to manage the symptoms better. Vaccines will slow the spread and reduce your likelihood of needing hospitalisation. Once you’ve slowed the spread and you’ve reduced the need for hospitalisation, you ensure that those who are get the very best care. You’ve stacked the odds in their favour.

It’s almost impossible to say at this stage whether this is something we will have to live with. When we’ve dished out the vaccine to everybody and the virus is now in a world where most people have antibodies, we don’t know if the virus will change its appearance slightly so that the vaccines are not 100 per cent effective anymore. In order to do that, will the virus become less dangerous? Will we need to make new vaccines? Or will the ones we have be good enough to keep people out of intensive care?

AB: Why do children transmit it more poorly and seem less affected by it?

DM: It’s hard to think of a respiratory virus that, if it kills people, doesn’t kill children. That’s a real surprise. We have got to learn quickly about coronaviruses. We also have to learn quickly about viruses that we don’t have vaccines for and get some vaccines because there are other viruses out there that could make the jump. They could be incredibly dangerous. There’s a virus that we worked on in Australia called Hendra virus. It only infected a dozen people but it has a thirty per cent fatality rate. There’s a whole list of weird and wonderful viruses out there. That suggests that this may happen again and again and again.

AB: Are diseases of this kind the price of an interconnected world?

DM: The classic example for that is Measles. It could only have sustained itself in the human population in the last 10,000 years. Once you’ve had Measles and recovered you can’t catch it again. The immune system remembers it very well. Measles can only attack new borns. You have to have a certain population size to get a certain turnover rate of new borns in order to sustain Measles in a population. That’s been worked out at around 200,000.

It’s a classic anti-vaxxer argument that there were island communities without measles because they were immune, vaccines were introduced by Europeans, then they died. The reason why island communities didn’t have measles is because they didn’t have children fast enough to sustain it. The virus just died out.

Measles kills 1 in 4,000 people or so. If you remember the Ebola virus outbreak in 2014, that killed 11,000 people and caused a global freak-out. In that same year, measles killed 124,000 people despite a massive worldwide vaccination programme. It killed ten times as many people as Ebola in that year. No-one batted an eyelid. Why? Measles has a far low case fatality rate but it is ferociously spreadable. It has an R0 of 15.

AB: Will vaccines fundamentally change the terms of the debate?

DM: We are probably going to have to live with it still. Even though flu kills thousands of people every year and there is a flu vaccine, many people don’t take it. I imagine this virus is going to be added to the number of the reasons that mean you might end up in hospital with a serious, debilitating disease. It’s not a certainty but it’s a likely outcome of this. Herd immunity will come to us whether we come to it with a vaccine or not. At some point, the virus will have infected most people. We will get the vaccine, roll it out and it will reduce pressure on the healthcare system.

The question then is what the virus will do then. Maybe we will have to get a new generation of vaccines that target parts of the virus that don’t mutate easily. All viruses have bits that they can’t change. The question is working out where they are and how easy they are to target. What I would like is a big effort to study viruses in general. However, there was strong signalling early on from the government that they were only interested in vaccine work and work on drugs that might halt the virus rather than fundamental research into the virus and its virology.

AB: How do you read our failure to scale up testing?

DM: During the foot and mouth outbreak, a real issue was that we couldn’t test for the virus quickly enough. Testing and tracing contacts amongst cows was nearly impossible when you couldn’t test fast enough. The government built a facility that could large-scale test for foot and mouth disease. It was able to test 40 to 50,000 samples a week. To put that into context, the entire testing capacity of Public Health England was around 10,000 a week, maybe less. There was a facility near Weybridge that had five times the national testing capacity.

That was offered to the powers that be and nothing was done with it. That was a shocking ball drop. We could have switched on some lights and told people to come into work. That facility, the Animal and Plant Health Agency, were mandated to bring it up to speed once a year to show that it could operate. It was available at the very beginning of the outbreak. We could have massively raised our testing capacity. They had to abandon community testing. They weren’t interested in screening travellers. That facility was designed to receive clinical samples from all over the country and test them and give a rapid yes/no answer. This is exactly what we needed.

There was an arrogance in that. Why were they trying to reinvent the wheel?

AB: The government has now moved to an explicit mass testing strategy – what are your thoughts on that?

DM: That’s a good idea – test, test, test and test. The resistance to testing comes back to the fact that in the flu playbook testing is not considered key except at the very, very beginning when it is thought to be a good idea to test and isolate as many people as possible. Once you have no control over it, then basically you stop testing because you can’t keep up. Testing hard at the beginning buys you time. As soon as the pandemic starts you click go on the timer and watch it run to four to six months when the vaccine arrives. In that scenario, if you test at the beginning, that gives you time to suppress the virus for as long as you can.

Once it’s got out of hand, that’s when you give up. That’s the flu playbook. We have drugs that work for flu. You stockpile antibiotics because secondary bacterial infections normally kill you, not flu. It’s all well worked out for flu. But when you are faced with a virus where there are no vaccines and no drugs. There are no guarantees any of these vaccines could work. Indeed, there are viruses which have no vaccines even after decades and decades during which very bright immunologists have tried to find solutions, HIV, Hepatitis C, for example. Perhaps, faced with the situation they were in, reverting to the flu playbook felt safe and easy and comfortable.

AB: What role has modelling played in our failures?

DM: Our response has been very heavily led by modellers and modelling. It feels opaque. Nothing that the modellers have said so far has been any surprise given that it has followed pretty much what a virus which infects 3 people would be expected to do. In February, I predicted around 100,000 people. The Chief Medical Officer predicted 20,000. That would have been a miracle. There are aspects of the modelling where you could work it out on the back of a fag packet. You don’t need to spend millions of pounds on modelling to work that stuff out.

Dr David Matthews is a reader in Virology and Molecular Medicine Infection and Immunity at the University of Bristol. He has been a virologist for 30 years and has been working on dangerous human coronaviruses for the last 5 years. Dr Matthews is a founder member of the Bristol University COVID-19 Emergency Research (UNCOVER) Group which brings together a range of teams from all over the University including Biomedical Scientists, Clinicians, Epidemiologists and Social Scientists. 

These groups are working at both an international level as well as with regional PHE laboratories and Bristol City Council to overcome the disease and its impact locally and on the world.