This site is reader-supported. When you click through links on our site, we may be compensated.
On Friday, some good news in the fight against SARS-CoV-2 was published in The New England Journal of Medicine. The antiviral drug remdesivir—originally developed as a potential treatment for Ebola—was shown to shorten recovery time for patients infected with the coronavirus. In late April, early results from this phase 3 clinical trial suggested that remdesivir might be of value in treating COVID-19 patients—this new paper confirms that. It's not a cure, but the drug shortened the recovery time from an average of 15 days to 11 days.
The trial involved 1,059 COVID-19 patients across 60 different sites in the United States, Europe, and Asia. Five hundred and thirty-eight patients were treated with a 10-day course of remdesivir; the other 521 patients were given a course of placebo on the same schedule. The patients were assessed daily, both to determine the severity of their symptoms as well as any side effects that could be caused by the drug, which interferes with the the virus' ability to copy its RNA.
What was this trial looking at?
The main thing being measured in this study was how long a patient took to recover, using an eight-point clinical scale that ranged from “not hospitalized,” through increasing levels of care required, and up to “death.” Secondary outcomes for the trial looked at mortality at two and four weeks after treatment began, as well as any serious side effects that occurred during the trial.
There had been some controversy about the trial because when it started in February 2020. Initially, the primary outcome measurement was how well a patient was doing on day 15. However, in late March, the trial's statisticians changed that to a secondary outcome, replacing it with the outcome described above. But these statisticians had no access to the data showing which participants were receiving the drug and which the placebo, and neither did they have any knowledge of the outcome data. The tweak to the study's parameters happened because of a growing awareness by scientists during those few weeks that COVID-19 was a more protracted disease than first thought, and therefore it made sense to study recovery over 28 days, not 15.
In late April, it was time to look at the initial results of the trial. And these results showed enough of a clinical benefit from remdesivir that the researchers had an ethical obligation to share their initial findings with the broader medical community. This also meant that patients receiving the placebo could be given the drug.
11 days < 15 days
Overall, treatment with remdesivir shortened a patient's time to recovery compared to the placebo group, from an average of 15 days to 11 days. Improvements occurred whether or not the patient was receiving supplemental oxygen. What's more, the data lays to rest any worries that remdesivir has to be given very early after the onset of symptoms. In fact, those participants who entered the trial more than 10 days after the onset of symptoms actually showed a better response to remdesivir than those who started being treated during the first 10 days of being symptomatic.
The trial's main secondary outcome—how a participant was doing on day 15—also showed that remdesivir was significantly better than placebo. And the total number of deaths was lower in the remdesivir group (21 versus 28) at this point, although that difference was not statistically significant. (An analysis of mortality at day 28 is still ongoing, given that enrollment in the study only ended in late April.)
The researchers note that remdesivir treatment is unlikely to be sufficient on its own given that it has, at best, a moderate impact on mortality, so studies that combine the drug treatment with other therapies should be explored. But in comparison to another recent study on the effect of hydroxychlorquine on COVID-19—which suggests that drug causes a marked increase in death—this work should definitely be considered a success.
The New England Journal of Medicine, 2020. DOI: 10.1056/NEJMoa2007764 (About DOIs).