What are the main conclusions of the Recovery study on dexamethasone?
In patients who have been admitted to hospital with COVID-19, for those who require a ventilator treatment, dexamethasone reduces the risk of dying by one third. Among the patients who do not yet need a ventilator but might be treated with oxygen, it reduces the risk of death by one fifth. In patients who are admitted to hospital but whose lungs are working well, i.e. they do not need oxygen or a ventilator, then dexamethasone has no effect. This drug is for people who have the more severe forms of COVID-19, those affecting the lungs’ function.
How did you come up with the idea of developing the trial when until a few weeks ago the use of corticosteroids had been contraindicated or not recommended during COVID? When I saw the study, I thought that once again, medicine is a combination of art and science.
Yes. Guidelines are based on the best information that one has at the time, and there were no trials of dexamethasone in COVID-19, there was no information. Many of those guidelines, as you say, said ‘not recommended’. But many of those guidelines also say, ‘except in the context of clinical trials’, because we need to know what the answer might be.
What we see with COVID-19 is that it is almost a two-phase disease. There is an early phase, where there is a lot of virus and the body’s immune system is combating the virus. In that phase, suppressing the immune system would not be great. Then follows an inflammatory phase, where the immune system is still fighting the virus. It is in that second phase that using dexamethasone turns out to be effective.
Recovery patients were all hospitalised. Do you see a potential use of corticosteroids in primary care or at home?
Well, we did not study those people. However, it is certainly the case that the people we did study, those whose lungs were working well and did not need oxygen, did not benefit. I think it is a reasonable extension to say that we have no good evidence that corticosteroids work in primary care or out of the hospital, and probably no good reason to think they might work.
I would not want to use dexamethasone outside of the hospital setting. Inside the hospital setting, I would only want to use it when the lungs are not working well, and then it actually works really well. The sicker the patient – particularly the worse their lungs – the better the drug works.
Why did you choose dexamethasone and not another corticosteroid? One might assume that all corticosteroids might, if used in equipotential doses, be equally effective in diminishing the inflammatory response.
We had to choose something and we chose dexamethasone. I think the dosage and the time are more important than which steroid you use. We used a low dose and we have used it in patients who are late into their disease and have bad lung disease. It is a combination of low dose and timing that is significant.
Recovery is defined as a large, pragmatic, randomised controlled adaptive platform trial. This approach seems essential for studies during epidemics, but perhaps also outside epidemics.
In the context of the epidemic, there are a few things that count. It is a new disease with no known treatments. It is a life-threatening disease with one in four patients admitted to hospital who will die. Therefore, we needed trials so that we could study a number of treatments, which is what we have been doing.
You are right; outside the context of an infectious disease epidemic, this same approach could and arguably should be used for other conditions. To give some examples, there is what you might call the slow chronic epidemic of diseases like obesity, diabetes, heart disease or cancer, but also the sort of burning epidemic of antimicrobial-resistant infections, in which this approach might be helpful. So in those circumstances, this sort of large, platform approach where you focus in on the things that matter, you keep it relatively simple, so that every patient can take part: that sort of approach has been really important. It is a bit of a change up for how trials are done.
Corticosteroids have been a tool for any respiratory failure for many years. They have been widely used in syndromes closely related to COVID-19, including SARS, MERS, severe influenza, and community acquired pneumonia. Where is the news with the Recovery results?
We have to be careful, as each of those diseases is slightly different. It will depend heavily on the balance between the virus being the problem and the inflammation being the problem. So timing is one thing, but the amount of inflammation you get, and the amount that causes lung damage, might vary between different infections. So I think we should be careful not to take these results and use them much more widely.
You pointed to the guidelines earlier on, saying that steroids were not recommended: that was because there was not enough information really to make a recommendation. The guidelines are there to guide and are written on the basis of the best information we have at the time, but what is important is that we need the big trials to tell the difference between things we hope are true and things we know are true.
In several European countries there are now new waves of epidemics.
In the UK as elsewhere in the world, we are planning ahead in case there is another resurgence of the epidemic later in the year. From a trial point of view, we are, for example, continuing important treatments like convalescent plasma. The Recovery result was a great start but it is not enough and we need to know more.
We do not have vaccines and we are unlikely to get vaccines in time for this winter. We might get some, but not at millions of doses, so all the social distancing and testing are really important. From a UK perspective, now is the time to do it when things are sufficiently quiet, because when things get busy, it is too late. You can see all around the world that if you do not prepare, by the time the virus really hits it is just too late.
New vaccines and treatments are the results of solid international research. At the risk of being trivial, it is worth repeating that research is crucial.
Someone considered research a luxury for the economy. In this particular case, it has never been more true that if we do not help science and research, you have no economy. If you look around the world, over the last three or four months economies have stopped because of the impact of this virus. So, if you do not have health, you do not have science, you do not have an economy. This is a complete reversal of the traditional way of thinking about it and I hope we will wake up to that.
A final thing to say is that the treatments that come out should have large-scale, affordable and equitable access. One of the things we have to be really guarded about is that anyone of any particular country or geography demands or limits the availability of the treatments to others.