Well conceived drug trials have saved hundreds of thousands of lives

IN FEBRUARY 2020 doctors from the World Health Organisation (WHO), visiting China to see the ward- and morgue-filling reality of covid-19, found more than 200 clinical trials under way. Desperate to help dying patients, doctors in Wuhan and elsewhere had been scouring their dispensaries for plausible treatments. They were trying out antiviral drugs developed to treat other viruses, anti-inflammatory and other drugs that modulate immune responses, antibody-rich blood plasma from people who had recovered, traditional Chinese medicines of various sorts and more. In each case there was at least some reason to hope that the treatment might do some good. But the trials were all too small to see if those hopes were being borne out.

Drugs are a difficult, and expensive, technology to work with because—unlike, say, a computer chip—they can be useful even if they only work in some cases, and telling those that work some of the time from those that don’t work at all requires big clinical trials. Seeing the pandemic heading their way, doctors and health systems in other countries scrambled to set some up. Drug companies, for their part, started to work on the only targeted drugs they know how to develop from scratch in months: antibodies.

The need for Solidarity and Resolve

In March 2020 the WHO set up Solidarity, a trial which eventually enrolled 12,000 patients in 30 countries. It looked at the effects of various existing antivirals used to treat HIV and Ebola, as well as an antimalarial drug, hydroxychloroquine, which seemed in laboratory studies to have had an effect on SARS-CoV, the virus behind the SARS outbreak of 2002-03. Six months on, none of the treatments was found to reduce mortality or the need for mechanical ventilation; nor did they shorten hospital stays.

Though disappointing, this was not surprising. Antivirals have never had the stopping power or the broad range of actions that the best antibiotics used to fight bacteria can boast. This is because killing bacteria, as antibiotics do, is a fundamentally simpler problem than stopping viruses. Bacteria are cells in their own right which depend on lots of distinctive molecular mechanisms to stay alive. An antibiotic that knocks out one of those mechanisms can be effective against many sorts of bacteria while leaving human cells, which lack the relevant mechanism, unscathed.

Antivirals, by contrast, must either tweak the inner workings of cells to make them inimical to viruses or target one of the handful of viral proteins required for infection or replication. The first option has historically led to a great many side effects. New technologies, though, might allow a more targeted approach that aims at some factor in the host which the virus absolutely requires and the host can do without. The CRISPR gene-editing system makes it possible to create tens of thousands of cell lines each lacking just one gene; finding out which of these cell lines a virus cannot get on with shows which host proteins it most depends on. Applied to SARS-cov-2 this screening technology has turned up various host factors worth looking at, some of which are already drug targets, but there is no human trial data to go on as yet.

The alternative to going after host factors is targeting a system which the virus provides for itself, such as the one it uses to copy its genome when reproducing. Remdesivir, a drug developed by Gilead, an American pharmaceutical company, does this by mimicking one of the building blocks of RNA. When the virus’s genome-copying mechanism uses this molecule instead of the proper building block, it crashes.

Remdesivir did not pan out as a treatment for hepatitis C, the virus for which it was originally designed, and though it has some effect against Ebola other treatments work better. When Solidarity tried remdesivir against SARS-CoV-2 it found no benefit. But three smaller trials, some with patients in earlier stages of infection when antivirals work best, found it reduced the length of hospital stays (though did not save lives). Fifty countries, including America, have approved it for use, but the WHO remains unconvinced of its merits.

Though there are still some other antivirals in trials, “I think we have pushed about as far as one can with repurposing,” says Francis Collins, the director of America’s National Institutes of Health. “You really need to have something more targeted.” The Rapidly Emerging Antiviral Drug Development Initiative, or READDI, based at the University of North Carolina, aims to make sure that when the next pandemic virus comes along there are better targeted antivirals to meet it. It is seeking to develop broad-spectrum drugs against the viral families with the most pandemic potential: coronaviruses, flaviviruses (such as Zika and West Nile virus) and alphaviruses (such as chikungunya, a mosquito-borne disease). The goal is to have, by 2025, stockpiles of five promising drugs that have passed safety trials in humans and are ready for application. It is focusing on drugs given orally, which can be used early and widely. Remdesivir’s modest results may stem from its need for intravenous administration. It is only used in hospital patients for whom its time has passed.

If antivirals have underwhelmed, a cheap anti-inflammatory drug has been a life saver. The means of demonstrating this was RECOVERY, a clinical trial which covers all of Britain’s hospitals. Martin Landray of Oxford University rushed to get RECOVERY integrated into the clinical-trials system within the National Health Service before wards began to overflow. The effort paid off, allowing recruitment into RECOVERY to become a possibility for most covid-19 patients referred to hospital. It enrolled more than 11,000 patients in 175 hospitals in its first three months, and has published results on six treatments.

Dexamethasone, the benefits of which RECOVERY had demonstrated by June 2020, is a corticosteroid which acts as an anti-inflammatory. Inflammation is part of the body’s natural defence against infections and other challenges. Immune cells sensing something amiss in their neighbourhood send out signals which cause nearby blood vessels to dilate and leak. That causes redness, swelling and pain while allowing more immune cells to gain access to the site. With covid-19, as with a variety of other infections, this response can go too far, compromising infected organs.

Steroids can stop that. The problem is that if they dampen down the immune response before it has the virus at bay they may do more harm than good, says Ralph Baric of the University of North Carolina. RECOVERY showed that the best time to start dexamethasone is when patients first needed to be given oxygen.

Quickly adopted as the standard of care all around the world dexamethasone has been credited with saving more than 600,000 lives. Survival rates have also been improved by experience. Early on, patients with low oxygen levels were put on mechanical ventilators too readily. After blood clots were found to be common in severe covid-19 cases, blood-thinners became a routine hospital treatment. In Britain the death rate among hospitalised patients has fallen from 31% in March-May 2020 to about 12%.

This February the RECOVERY trial showed that tocilizumab, another drug which dampens down immune responses, also reduces the risk of death from covid-19, confirming results from an earlier European trial. Tocilizumab, like all drugs with names ending in -mab, is an antibody—a class of drugs for which the market has quadrupled over the past decade.

Queen Mab gallops, night by night

The immune system uses antibodies to fight pathogens. When they stick to virus particles they can stop the virus from getting into cells. When they stick to cells infected by a virus they tag them as targets for other bits of the immune system to attack.

The pharmaceutical industry tends to use the antibodies it makes in cell cultures to different ends: not to target alien molecules, but to target molecules the body is making to its own detriment, most notably in cancers and autoimmune diseases. All the big-selling antibody drugs cater to these markets. Tocilizumab blocks the action of interleukin-6, a messenger molecule which plays a role in inflammation; it was developed to lessen the symptoms of rheumatoid arthritis. Other antibodies which modulate the immune system are still in trials as covid-19 treatments.

But some antibodies have been developed as treatments for infections. Last year America’s FDA approved two antibody drugs for Ebola; there are now a large number of antibodies for SARS-CoV-2 in trials. It has been possible to get them into the pipeline quickly because of the way they are developed. Finding an immune-system cell that produces the antibody you want, tweaking the relevant genes and putting them into cells devoted to antibody production is a lot faster than sifting through vast libraries of chemicals to find those with the most promise, learning how to make them into medicines, proving them not to be toxic and so on.

A year into the pandemic antibody treatments made by two American drug companies, Regeneron and Eli Lilly, have been approved by regulators. Each is a cocktail of two different antibodies, and each has to be infused through an intravenous drip. But both are approved for use only in patients with mild to moderate illness, or those who are at risk of becoming ill. They are contraindicated for those already sick enough to be admitted to hospital. Even if their potential beneficiaries feel it worth having infusions when not that sick, the practicalities of providing such a service to people who should be isolating at home are daunting. Nearly 1m doses of the two drugs that have been bought and distributed by America’s federal government have gone unused.

What is more, the spike protein at which the antibodies are aimed is evolving. Researchers need antibodies that recognise these new variants, too. George Scangos is the chief executive of Vir Biotechnology, which makes an antibody recently shown in trials to reduce hospitalisation or death by 85%. He describes how his team selected it because it binds to a bit of the SARS-CoV-2 spike protein that looks the same as its equivalent in SARS-CoV, the virus which causes SARS. The thinking was that, if evolution had kept it the same, it was probably because it could not be changed without weakening the virus. But Dr Scangos says it is possible that viral evolution will, in time, catch up with their cleverness, and that the team will have to find further antibodies to be used in combination with this one.

The ideal would be a system that can make lots of slightly different antibodies at once, makes them exactly where the patient wants them, and can be programmed in advance to churn them out as required. Happily, almost all human bodies come equipped with just such a production line, in the cells of the immune system. And in the most impressive of all medicine’s responses to the covid-19 pandemic, new vaccine technology is now putting those production lines to use in millions of people every day. ■