select search filters
briefings
roundups & rapid reactions
Fiona fox's blog
As reported by Stat News and the FT, new data on a trial of remdesivir in COVID-19 patients became available online for a brief period.
Dr Stephen Griffin, Associate Professor at Leeds Institute of Medical Research (LIMR), University of Leeds, said:
“It is difficult to interpret the Chinese study of remdesivir that was mistakenly posted on the WHO site yesterday without full access to the completed, peer reviewed manuscript.
“Whilst the results appeared disappointing it appears that the trial was halted prematurely, which can hamper the correct statistical analysis. Nevertheless, poor performance in a trial this size may have implications for the larger studies currently ongoing, but this remains to be seen.
“Remdesivir, based upon preclinical studies, represents one of the best options for a SARS-COV2 antiviral therapy, making the outcome of this and other trials critical for the longer term management of COVID19.”
Prof Stephen Evans, Professor of Pharmacoepidemiology, London School of Hygiene & Tropical Medicine, said:
“Assuming that this, accidental, release of data from a randomised trial in China are genuine, and there is no reason to believe they are not genuine, then this would give the first reliable evidence on the benefits and an outline of possible harms of remdesivir. Previously released data did not have a proper comparison group, while this trial has a group allocated at random.
“It would seem that the data come from a trial, registered appropriately at the US National Institutes of Health supported registry clinicaltrials.gov, and assuming this is correct we can say more about the trial.
“The description is at https://clinicaltrials.gov/ct2/show/NCT04257656?term=remdesivir&cond=COVID-19&draw=2&rank=4
“The entry there has been noting that the trial was “terminated (The epidemic of COVID-19 has been controlled well in China, no eligible patients can be enrolled at present)” for a week or two.
“It notes that the trial (ClinicalTrials.gov Identifier: NCT04257656) had 237 participants (which is the number reported in the leaked WHO entry). It was intended to recruit 452 patients. It was carried out by Capital Medical University in Beijing. It was in patients with confirmed severe Covid-19 disease who had been admitted to hospital.
“The leaked data suggest that those on remdesivir and on placebo had similar outcomes. The mortality rate on remdesivir seems to be 22/158 (13.9%) while on placebo was 10/78 (12.8%) {One patient in the placebo group did not get treatment so the numbers analysed are 78 instead of 79 allocated}.
“This is a relative risk of 1.09 with a confidence interval from 0.54 to 2.18. This indicates the numbers are too small to exclude a halving or a doubling of the death rate. There is a 1% difference in the death rate (13.9-12.8%), which is compatible with a reduction of 8% and an increase of 10%. It was also noted that a larger number stopped their treatment because of adverse events while on remdesivir. The numbers seem to be 18/158 vs 4/78. The risk ratio is 2.2, compatible with a range from 0.78 to 6.34. The numbers in the trial are too small to draw strong conclusions, but certainly are not any indication of benefit for remdesivir and are slightly more compatible with possible harm.
“Assuming these data are correct, the trial was too small in numbers recruited to detect what could be a clinically important benefit or an important increase in risk. It does not show that remdesivir has a really dramatic improvement in mortality, but could nevertheless be compatible with a small improvement in spite of the results apparently showing remdesivir does slightly worse than placebo.
“The trial protocol required patients to be entered into the trial within 12 days of the onset of symptoms. It has been suggested that these drugs need to be started early in the course of disease but it would seems that there are a number of days after infection before symptoms appear. If the drug only works well when given very early after infection, it may be much less useful in practice. We need appropriate randomised trials to test this hypothesis.
“What is incontrovertible is that randomised controlled trials with a proper comparison group (the best standard of care, ideally with a placebo) are the only ones from which reliable conclusions can be drawn. There are larger trials of that type being carried out now and results will be available later in the year. There is another trial from Wuhan in China which is apparently “Suspended” with the same message “(The epidemic of COVID-19 has been controlled well at present, no eligible patients can be recruited [sic])” and may have recruited 308 patients. Results from it can presumably be expected soon.”
https://www.ft.com/content/0a4872d1-4cac-4040-846f-ce32daa09d99
All our previous output on this subject can be seen at this weblink:
www.sciencemediacentre.org/tag/covid-19
Declared interests
Prof Stephen Evans: “No conflicts of interest.”