LONDON - EMA Director General Guido Rasi has hit out at the lack of coordination in COVID-19 clinical trials, questioning if the huge number of small studies that are running will support regulatory decision-making.
“This is something very, very important,” Rasi told the European Parliament’s public health committee on May 18. “We have been engaged, with a lot of frustration. Clinical trial authorization is not for EMA, it is member states’ prerogative; we really tried hard to foster [coordination],” he said.
On March 16, the EMA published a statement calling for the EU to set up large, multicountry, randomized clinical trials. That has made little impression, with the only such study, called Discovery, managing to recruit just one patient outside France, from where it is being managed, since getting off the ground on March 22.
The EU clinical trials database, Eudract, shows there are 268 COVID-19 trials planned or running in the EU. Meanwhile, 1,591 COVID-19-related studies are listed on the U.S. website clinicaltrials.gov, and a further 1,114 on the World Health Organization’s (WHO) International Clinical Trials Registry (which does not duplicate any studies listed on clinicaltrials.gov).
Aside from the important resource issue and the ethical issue of enrolling patients in small, individual control groups, the obvious concern is that these trials will not deliver robust data.
As one very infamous case in point, anecdotal evidence abounds on the effectiveness, or otherwise, of hydroxychloroquine. Eudract lists 55 trials investigating use of the antimalarial drug in COVID-19, at various stages of disease pathology and in various combinations with other drugs. Meanwhile, clinical trials.gov lists 197 such studies.
Rasi told members of the European Parliament (MEPs) that “less fragmentation” would be better. With many small trials in a very heterogeneous population, it is very difficult to detect an effect. “It could have been done faster with a much bigger trial, with a single protocol,” he said. “There are still many [hydroxychloroquine] trials ongoing, and I hope they will come in with data to show which patients benefit and when, and for how long.”
The slow progress of the Discovery trial, being run by France’s National Institute of Health and Medical Research, highlights the problems encountered in setting up randomized multicountry European trials in the midst of a pandemic.
The study aims to recruit 3,200 patients, but to date has managed around 700, all but one in France. In comparison, the large-scale U.K. Recovery trial, which like Discovery is testing approved drugs, has recruited 10,420 patients since March 23. The adaptive design has allowed protocol amendments, with two therapies added since the study began.
Why is Discovery “not going too well?” Belgian MEP Frédérique Ries asked Rasi. “What is the problem? What is missing here?” she said.
The emergency situation and lack of an approved therapy for COVID-19 means some fragmentation is to be expected, Rasi said. “Clinicians had to find whatever they could to address and mitigate symptoms.”
In addition, the Discovery trial found itself in competition with WHO’s Solidarity clinical trial, with some member states preferring to sign up for that.
“Another key problem was that this kind of trial needed a sponsor and there wasn’t one,” said Rasi. Instead, funding was coming from a number of different EU clinical research projects.
The European Commission has stepped in to back Discovery through emergency finance put in place to deal with the pandemic. “It will now get EU funds, which I think is the best way to go. I hope many countries will join in, since, unfortunately, there are still many patients,” Rasi said.
Lessons from past pandemics
Rasi was answering MEPs questions just after he and seven senior colleagues at the EMA published a pointed criticism of the lack of European coordination of COVID-19 clinical research, saying such a large collection of small studies is unlikely to support regulatory decision-making.
“For every week that trials don’t deliver, more and more patients are exposed to the wrong treatment, which well-designed and rapidly run clinical trials could have taken off the table, making space to pursue other, and ultimately more meaningful therapeutic options,” Rasi and his co-authors say in an article published in Clinical Pharmacology & Therapeutics.
To make things more complex, it is improbable that a single drug will be able to control the diverse respiratory, inflammatory and cardiovascular symptoms of COVID-19.
Lessons from past pandemics and epidemics show the window for running adequately powered trials “remains open for just so long,” Rasi writes. In the Ebola epidemic in 2014 – 2016 in West Africa, just one randomized controlled trial of a therapeutic got off the ground. It did not complete, and no therapies were approved for treating the infection.
Similarly, despite compressed development plans, no study of the handful of Ebola vaccines tested in West Africa completed and only one generated sufficient data to gain approval. In the case of Zika virus, by the time vaccines were ready to be tested at scale, the virus was disappearing.
A similar situation could arise for COVID-19 vaccine development, as transmission is brought under control just as candidates are ready for field trials.
Of note, a disclaimer on the article says the views expressed in it are the personal views of Rasi and the seven members of the EMA scientific committees who are co-authors. The article may not be quoted as being on behalf of, or reflecting the position of the EMA or other regulatory agencies by which they are employed.
Curious then, that the EMA issued a press release drawing attention to the article, saying the “EMA authors” have set out “concrete actions” anyone involved with COVID-19 clinical trials should take “to generate the type of conclusive evidence needed to enable rapid development and approval of potential treatments and vaccines against COVID-19.”
As Rasi told MEPs, such coordination is “absolutely mandatory if Europe wants to have faster data, more robust data and play a fuller role in research.”