Skip to content

Science Museum Group

The first encouraging news in the global hunt for treatments to curb the pandemic was reported last week. Roger Highfield, Science Director, describes the race to cure COVID-19.

how is the hunt for treatments going?

We glimpsed light at the end of the tunnel with the announcement of encouraging results from infusions by a drip of remdesivir, a drug originally developed by the Californian company Gilead Sciences to combat Ebola.

The drug infusions enabled hospitalised patients with advanced COVID-19 to recover 31% faster than those who received a placebo, or dummy drug in the international trial in 75 hospitals, co-led by the UK’s Medical Research Council and UCL.

Even though another, smaller, trial of the drug did not show any statistically significant benefit, the US Food and Drug Administration has issued an emergency authorisation to use the drug.

Mahesh Parmar, Director of the MRC Clinical Trials Unit at UCL, said:

‘The data and results need … assessment by the relevant health authorities in various countries. While this is happening, we will obtain more and longer-term data from this trial, and other ones, on whether the drug also prevents deaths from COVID-19.’

Though it could help save lives if the findings of the international trial are confirmed, remdesivir will never be a magic bullet.

Many questions remain, from whether it produces better results if given earlier in the progression of the disease to how to ramp up production and ensure equitable distribution.

Resistance to any single drug might also develop, so more than one drug may be needed in the long term (though, fortunately, coronaviruses evolve more slowly than other viruses).

In general, the dream of repurposing old drugs has fallen short of reality. Until we get a drug customised for this particular virus, existing antiviral drugs are likely to have ‘a moderate effect’ at best, says Chris Whitty, Chief Medical Officer for England.

why is it harder to develop drugs to fight viruses?

Viruses are not easy to kill once they are inside the body, even though they’re tiny, only a scrap of genetic code wrapped in an overcoat of protein and lipid (fat). They measure just one-thousandth of the diameter of a human hair (15 – 300 nanometres across, much smaller than bacteria).

It is even harder to make antiviral medicines than making antibiotics because to multiply in the human body, viruses rely more on our molecular machinery than they do their own.

Viruses hijack the chemical machinery – notably ribosomes in human cells that turn genes into flesh and blood – to create more viruses by turning human cells into virus factories.

Wendy Barclay, Action Medical Research Chair, Virology, at Imperial College London commented:

‘The disease is a result of both viral damage and collateral damage and so the different types of drugs being tested target either one, and the drugs that attempt to inhibit the virus could inhibit either viral genes or human host genes.’

Because they borrow our biochemistry, viruses are hard to combat without also damaging the cells they have infected, causing side effects.

That is why, in the long run, vaccines are preferable to end the pandemic: they prevent infection with the virus in the first place.

what are the possible treatments?

Because there are no proven treatments for COVID-19, three basic approaches are being tried:


Typically these are small molecules that interfere with virus biochemistry, its interactions with proteins in human cells, or the patient’s immune system.

The coronavirus contains 29 proteins –building blocks of living things – and the hunt is on for drugs to interfere with this machinery, so different drugs are required for different protein targets in the virus.

When it comes to remdesivir, the drug looks to the virus like a ‘letter’ of viral genetic code but does not behave like it. As a result, this drug interferes with a protein – called polymerase – that copies viral genetic material.

Another kind of drug is a protease inhibitor. When the virus infects a cell it first makes long ‘polyproteins’ that are chopped into functional protein pieces by an enzyme called a protease.

The main protease in the virus’s life cycle can be disrupted with a protease inhibitor, though we don’t yet have a drug to do this.

Another, quite different, approach is to use drugs to stop the virus from entering human cells, which depends on human proteins, notably ACE2, furin and TMPRSS2.

As one example, a drug called camostat mesylate, already approved for use in Japan as a treatment for pancreatitis, blocks TMPRSS2.

One team has looked at how 26 of the 29 SARS-CoV-2 proteins interact with human proteins, creating a ‘protein interaction map’ to guide drug repurposing.

They found the virus interacts with 332 human proteins and, of these, 66 were targeted by 69 known drugs. Laboratory tests in Paris and New York revealed two groups of drugs with antiviral activity (although no tests were performed in individuals with SARS-CoV-2).

One group of these antiviral agents, including cancer drugs, blocked protein translation (a key process for viral replication) while the second group, including a hormone and anti-allergy drugs, targeted other proteins (Sigma1 and Sigma2, which are so-called receptors that are part of the cellular communication network).

Yet another approach is to use drugs to curb an excessive immune reaction triggered by the virus (a ‘cytokine storm’), which is thought to be linked to lung inflammation and life-threatening disease, ‘acute respiratory distress syndrome.’

Convalescent plasma

An old-fashioned treatment that can be effective, in which a seriously-ill COVID-19 patient is given blood plasma collected from a recovered patient. Their plasma, the yellowish liquid which remains when red and white blood cells have been removed, contains antibodies that neutralise the virus.

There is already some evidence that debilitated patients can rally after a dose of survivors’ blood and the United States has launched a national effort to roll these blood-based therapies out as soon as possible.

In the UK, a trial started last week, using plasma collected by NHS Blood and Transplant more than a month after patients have recovered from infection, when they have developed antibodies to COVID-19.


Identifying the specific antibodies within convalescent plasma that can neutralise the virus, and synthesising them to scale, is the 21st century equivalent of old blood-based convalescent plasma treatments.

Within a week of receiving a blood sample from one of the first American patients who recovered from COVID-19, the company AbCellera screened over 5 million immune cells looking for the ones that made antibodies that helped the patient neutralize the virus and recover.

The Chinese Academy of Sciences and Vanderbilt University Medical Center are providing UK-based AstraZeneca with genetic sequences for antibodies they have discovered against SARS-CoV-2.

Meanwhile, Regeneron has isolated hundreds of virus-neutralizing antibodies from mice which have been genetically-modified to have a human immune system.

do we really have to hang about for trials?

Yes. Because it often takes a long time and lots of studies to show that a drug is really effective.

Many countries stockpiled the drug Tamiflu for pandemic influenza before a study eventually concluded the drug was no more effective than paracetamol.

However, as Prof Wendy Barclay of Imperial College London pointed out:

‘Tamiflu is a very good drug but you have to give it early while virus replication, the target for the drug, is still exponentially increasing – and the problem is that most patients do not present to their GP until several days after they develop symptoms. This is an issue that also hinders COVID-19 therapies.’

Confidence in the reliability of clinical research has been under increasing scrutiny since 2005, when John Ioannidis of Stanford University wrote an influential article: Why Most Published Research Findings are False.

He even coined a phrase, the Proteus phenomenon, to describe a common pattern seen in molecular genetic research, blighted by early contradictory results.

However, as my first blog post pointed out, that is how science works – theory and ideas are endlessly challenged by data, new ideas, and the results of experiments in what is called the scientific method.

Matthew Freeman of the Dunn School, University of Oxford, commented:

‘Like all of life, and even viruses, scientific ideas evolve. The pioneering biologist and Nobel Laureate, Max Perutz, had a favourite saying that “In science, truth always wins”.

Although there will be wrong turns along the way, the scientific method eventually leads to understanding how the world really works. In some ways, this is the main achievement of all experimental science.’

what kind of trials are the best?

The ‘gold standard’ is to conduct what is called a randomised controlled trial, when a number of similar people are randomly assigned to two or more groups to test a specific treatment.

One group (the experimental group) has the treatment, the other (the control group) has a dummy (placebo) or no treatment.

Outcomes are measured at specific times and any difference in response between the groups is assessed statistically.

Much of the confusion about whether a drug really works comes from early reports of small studies which are neither randomised, nor controlled.

Spurred on by the current reward and recognition systems of academia, and borne on a tide of concern about COVID-19, there is a tendency to quickly publish one-off findings which look transformative, rather than invest additional money, energy and time to ensure that these one-off findings are reproducible

‘We need evidence from large, high quality trials,’ said Anthony Gordon, an NIHR research professor at Imperial College London.

‘If you give a drug to a small number of patients and they do well, you will never know for sure if they were always going to do well, even without the drug. That is why you need a control group and to randomise treatment.’

Moreover, all drugs have side effects and most don’t work for everyone (see, for example, cancer drugs).

what are the biggest trials?

The World Health Organization has announced a global trial, SOLIDARITY, on four promising therapies:

  1. Chloroquine and sister compound hydroxychloroquine, a tried and tested antimalarial which is thought to have an antiviral effect by targeting endosomes.
  2. Cellular compartments that cells use to ingest material that coronaviruses also exploit to invade human cells.
  3. Remdesivir; a combination of two HIV drugs, lopinavir and ritonavir, which block viral replication.
  4. The same combination as above plus interferon-beta, an immune system messenger that can trigger cell pathways that stop viruses from replicating.

‘More than 100 countries have joined the Solidarity Trial, and more than 1200 patients have been randomized from the first five countries, to evaluate the safety and efficacy of four drugs and drug combinations,’ said a WHO spokesperson.

is the uk doing drug trials?

Some of the biggest and most systematic COVID-19 drug trials on the planet are taking place in the UK.

There are 31 in all, according to the National Institute for Health Research, the research arm of the NHS, which spends around a £1 billion annually.

The three main national trials are:


Platform Randomised trial of INterventions against COVID-19 In older peoPLE, which will test treatments on 3000 higher risk patients in hundreds of GP practices.

Researchers from the University of Oxford are looking for people aged 50 to 64 with pre-existing major illnesses, and patients aged 65 and above without any other known illness.

Unlike many other clinical trials, which focus on patients with serious symptoms who are in hospital, PRINCIPLE hopes to identify treatments that can slow or halt disease progression and prevent hospitalisation according to Chris Butler, Professor of Primary Care in the Nuffield Department of Primary Care Health Sciences, a part-time GP.

‘What’s unique about the PRINCIPLE trial platform is that it’s so flexible. By setting up a nationwide primary care research network across the NHS, we’re able to rapidly evaluate potential new treatments for COVID-19.’

Once again, they will test hydroxychloroquine. The antibiotic azithromycin will soon be added to this so called ‘platform trial’, which offers a framework to test many drugs.

‘We are focusing on treatments that could be rapidly scaled up for widespread use in the community,’ said Prof Butler. ‘We will not be including remdesivir, which is a treatment that is given intravenously, and is suited for hospital use.’

More patients will be recruited if additional treatments are introduced and the trial design may also be adjusted in the light of results that emerge. ‘It will be a little while before we will be getting the first results from the PRINCIPLE trial,’ said Prof Butler.

‘At the moment, we are only able to include patients who are registered with 420 general practices that are participating in the trial. By the end of next week, we hope to have the trial open to people regardless of their practice.’


Randomised Evaluation of COVID-19 Therapy, led by Peter Horby at the University of Oxford, is the fastest-growing trial in medical history, with 8000 patients recruited in its first month.

This trial will test lopinavir-ritonavir (Kaletra, commonly used to treat HIV); low-dose dexamethasone (a type of steroid, which is used to reduce inflammation); hydroxychloroquine; azithromycin (a commonly used antibiotic); and tocilizumab (an anti-inflammatory treatment given by injection).

Data from the trial will be regularly reviewed so that any effective treatment can be made available to all patients. The RECOVERY team will also review information on new drugs and include promising ones in the trial when it comes to lowering death rates, speeding hospital discharge, and reducing the need for ventilation.


This is an international randomised control trial that has been under development for some years for patients admitted to hospital intensive care unit with Community-Acquired Pneumonia, CAP, which is caused by bacteria, sometimes viruses, such as seasonal flu.

‘We had planned for a pandemic. We knew that when one comes, it is difficult to start from scratch. REMAP-CAP gives us the infrastructure to cope.’ according to the UK lead researcher, Anthony Gordon, an NIHR research professor at Imperial College London.

The ‘Randomised, Embedded, Multi-factorial, Adaptive Platform Trial for Community-Acquired Pneumonia’ aims to enroll patients at more than 100 hospitals within the UK, as well as in 15 other countries. ‘We will need around 600-1000 patients before we might see important results – the number we need varies according to the size of the effect: a dramatic effect will be apparent in smaller numbers, sooner,’ said Gordon.

Once again they will be looking at the anti-viral drugs, lopinavir-ritonavir; hydroxychloroquine, on its own or with the antivirals; hydrocortisone (a type of steroid, which is used to reduce inflammation linked with COVID-19); and immune modulation drugs (interferon beta 1a, interleukin 1 receptor antagonist [Anakinra], and interleukin 6 inhibition [tocilizumab and sarilumab]) to block excessive inflammation.

They have also started to test two more approaches:

Convalescent plasma therapies, which have been used successfully to treat other viral infections. Trials are necessary because there are issues about how long immunity to COVID-19 lasts and how effective that immunity is.

Moreover, there is also a risk, albeit small, of side effects: ‘If you don’t use mature enough antibodies, you can increase the activity of the virus,’ said Gordon. In other words, they need to make sure they use the right quantity and quality of antibodies.

Anticoagulant (blood-thinner) drugs, starting with heparin, because of the suspicion that, in addition to inflammation, COVID-19 causes small blood vessels of the lung to be blocked by clots.

What is puzzling is the phenomenon of ‘happy hypoxics’, where some patients seem unaffected by their extraordinarily low blood-oxygen levels, or hypoxia, and don’t feel short of breath.

what’s innovative about these trials?

RECOVERY has five active arms and is currently the largest trial in the world.

PRINCIPLE is the first trial in the world to work through primary care, trying to prevent people going to hospital in the first place. ‘The beauty of our trial design is that we can report a result as soon as we are confident that we have one, rather than having to wait until the target sample size has been achieved, as one would have to do in a traditional trial design,’ said Prof Butler.

REMAP-CAP is a smart trial in several senses. As well as being modular, so different treatments can be added if other drugs show promise, more than one drug can be tested at a time, said Prof Gordon. ‘These patients are so sick they are likely to need more than one treatment’. In other words, we need to treat both the virus itself and the consequences of infection.

The trial is also ‘adaptive’ and use Bayesian statistics, named after the 18th-century Presbyterian minister Thomas Bayes. Bayes devised a systematic way of calculating, from an assumption about the way the world works, how the likelihood of something happening changes as new facts come to light.

Bayes’ approach uses probabilities to weigh up whether one drug is doing better than another. In this way, after a routine assessment, REMAP-CAP will allocate even more patients to have treatments that seem to be the most beneficial.

‘That way patients can benefit from the insights from the trial before we announce the results to the world, which will take many months,’ said Gordon. ‘This self-learning system means more patients get the best treatment we can offer.’

how else can we use drugs?

Drugs can also be used for prevention – prophylaxis – to shield people who are at risk, just as we protect people at risk of heart disease with drugs called statins, said Chris Whitty, Chief Medical Officer for England. However, these drugs need to have low side effects and to be long-lasting.

did we see this coronavirus pandemic coming?

We saw it coming a mile off, according to Jeremy Farrar, Director of the Wellcome. The threat is on the UK National Risk Register.

We even knew a pandemic would result from this kind of virus (an RNA virus), according to a 2008 study by Mark Woolhouse and Eleanor Gaunt at the University of Edinburgh.

An RNA virus has its genetic recipe encoded through the order of four ‘letters’ in a 30,000 letter stretch of the chain like nucleic acid, RNA.

RNA viruses, notably those originating in wildlife (in the case of COVID-19, bats) are the most common class of pathogens behind novel human diseases.

Other examples of RNA viruses include HIV, Lassa fever, SARS, MERS, Chikungunya, Zika, Ebola and influenza.

Woolhouse added: ‘In 2017 a World Health Organisation committee (of which I was a member) added “Disease X” to a list of ten or so known viruses (Ebola, Lassa fever and so on). Disease X was an explicit recognition that the next pandemic might be caused by something we don’t yet know about. In 2018 the same committee (without me this time) went a bit further in characterising Disease X. They highlighted a novel coronavirus other than SARS or MERS‘.

Prefilled syringes of Influenza vaccine, for 2000-2001 strains. Part of the Science Museum Group Collection.

why are rna viruses, such as sars-cov-2, such a threat?

RNA viruses find it easier to jump from one species into another because they mutate more than other kinds of viruses, evolving to adapt to new species.

Even though only a small percentage of these mutations are helpful, that also means that they can evolve to evade their host’s immune system and defense mechanisms along with the antiviral drugs we use to fight them.

why can this kind of virus evolve so quickly?

The rapid evolutionary rates of RNA viruses result from frequent error-prone replication. However, in the case of SARS-CoV-2 it does not mutate as much as some of its peers since one of its proteins, NSP14, does ‘proof read’ genetic code.

is resistance going to be a problem?

Most RNA viruses do evolve antiviral drug resistance, said Prof Wendy Barclay, but she added that this matters most in acute infections if there is no fitness cost: in other words, the mutation required for the virus to be resistant to a drug does not harm the ability of the virus to replicate in human cells, and moreover, coronaviruses evolve more slowly than other RNA viruses.

how can we combat resistance?

By applying the revolutionary, evolutionary ideas of Charles Darwin.

These viruses mutate so that, when treatment starts, non-resistant viruses are wiped out to leave the few resistant mutants to quickly grow in number.

The experience of treating AIDS, however, shows that all is not lost. By attacking the virus on several fronts, using combinations of drugs with different underlying actions, it is much harder for the virus to evolve resistance.

Understanding the mechanisms by which a virus adapts to each drug and how these may enhance or diminish its ability to adapt to others, will be crucial. It could be that as few as two or three drugs, targeting different pathways, could be enough to control the virus in a given patient.

Charles Darwin portrait image
A Carte de Visite photograph of Charles Darwin from the Science Museum Group Collection.

where do rna viruses ultimately come from?

There is much debate about the origin of viruses and it is interesting to note that many believe the very first life to emerge on Earth four billion years ago may well have relied on RNA.

RNA is a more delicate-yet-flexible kind of genetic material than DNA. RNA not only stores information but, unlike DNA, it can catalyse chemical reactions too.

However, that does not mean all RNA viruses date back to the dawn of life. Andrew Rambaut of the University of Edinburgh said: ‘There is no shortage of RNA in cells. My guess is that RNA viruses represent multiple independent origins.’

drugs typically take more than a decade and billions to develop – how can we speed this up?

We can screen vast numbers of chemicals in a computer to see if they provide leads for drugs that can interfere with virus proteins, or the proteins in human cells that they target.

The biggest supercomputers on the planet have been enlisted for the job by  Peter Coveney of the Computational Biomedicine Centre of Excellence at University College London together with the Director of the Leibniz Rechenzentrum and Gauss Centre for Supercomputing, Prof Dieter Kranzlmüller, and Rick Stevens of Argonne National Laboratory, near Chicago.

The team is using the world’s most powerful supercomputer, Summit at Oak Ridge National Laboratory, Tennessee, the EU’s fastest supercomputer – SuperMUC-NG at the Leibniz Supercomputing Centre in Garching, near Munich, and Scafell Pike, one of the supercomputers at the Hartree Centre in Cheshire.

Cray 1a Supercomputer, serial number 11 on display in the Science Museum.
An early supercomputer was the Cray 1a supercomputer, on display in the Science Museum.

Their primary targets are existing drugs that are currently in manufacturing pipelines and can be repurposed quickly. In one project, they are screening a library of 8000 known drug compounds to find ones that are capable of binding to the protein spikes on the virus to thwart its ability to invade human cells.

They are also ploughing through a library of 100 million small molecules and libraries of billions of compounds that could be manufactured quickly for testing.

Experimental work complements these huge simulations and includes detailed studies of virus proteins. The use of ‘wet labs’ to check that candidate drugs really do interfere with target proteins and work by medicinal chemists, notably at the University of Chicago, to customise synthetic molecules to target viral proteins.

Typical of many scientific responses to COVID-19, this is a vast, cooperative, international exercise.

In this case, it involves more than a hundred researchers from five US national laboratories (Argonne, Brookhaven, Los Alamos, Oak Ridge National Laboratory, Lawrence Livermore National Laboratory), nine universities (University of Chicago, University of Illinois, University of Virginia, Rutgers University, Stony Brook University, George Mason University, University of Texas, University of California San Diego, UCL), a private research centre (the JC Venter Institute), and a public academy (Leibniz Rechenzentrum at the Bavarian Academy of Sciences and Humanities).

what are the major unknowns about covid19?

According to the Chief Medical Office for England, we don’t yet know the proportion of people who are infected and show no symptoms; how long immunity lasts; if blood tests correlate with immunity; if children can transmit the virus; if it is seasonal, and will surge in the coming winter; why serious disease tends to develop in some people after a week, and why men are more likely to die than women.

Fortunately, there is an unprecedented global effort underway by scientists, mathematicians, technicians and engineers to answer these questions.

what is the state of the pandemic?

You can get the latest news on how far the pandemic has spread worldwide from the Johns Hopkins Coronavirus Resource Center or from the Robert Koch-Institute, Berlin, view the UK hotspots identified by an app, check the number of UK COVID-19 lab-confirmed cases and deaths, and the overall number of deaths from the Office of National Statistics.

There is more information in my earlier blog posts, from the UKRI, on this COVID-19 portal and Our World in Data.

The Science Museum Group is collecting objects and ephemera to document this public health emergency for future generations.