In this study, researchers randomly administered either remdesivir or placebo to 237 patients. They found that remdesivir was safe to take but not associated with treatment benefits.
In this rigorously conducted randomized controlled trial, 237 patients with SARS-CoV-2 (COVID-19) pneumonia who were admitted to the hospital in Hubei, China were randomized to receive either remdesivir or placebo. Remdesivir did not change the time to recovery or death rate in patients with COVID-19.
Why is this important?
There has been significant public interest in repurposing drugs that already exist to treat patients with COVID-19. One such drug, remdesivir, has already been studied rigorously in a Chinese trial sponsored by Gilead Sciences, Inc. Within the United States, the National Institute of Allergy and Infectious Diseases (NIAID) issued a press release on April 29, 2020 touting the success of remdesivir in an American clinical trial. On May 22, 2020, the data from that trial was published in the New England Journal of Medicine. That study will be covered in a future Resaerch Explained summary.
What did the study do?
Enrolled patients with COVID-19 from 10 hospitals
Realistic study population
The study population was very applicable to the real world. It consisted of patients who were admitted to the hospital with SARS-CoV-2 pneumonia (confirmed by PCR testing and chest imaging) and who required supplemental oxygen or other respiratory support. The study design required enrollment, and hence, hospital admission, to occur within 12 days of symptom onset. This is probably applicable for most, if not all, COVID-19 patients who come to the hospital in the real world.
Patients used other medications
Patients enrolled in this study were allowed to receive other medications. Although this is not totally avoidable for ethical reasons, these medications included the antiviral medication lopinavir-ritonavir, the antiviral and immune modulating medication interferon alfa-2b, and anti-inflammatory and immunosuppressive steroids. 29% of patients in the remdesivir group received interferon alfa-2b compared to 38% in the placebo group, a potentially noteworthy difference. On the other hand, similar numbers of patients received lopinavir-ritonavir (28% and 29%) and steroids (65% and 68%) in the two groups. Unfortunately, the investigators did not report the doses of these medications so that we could better compare whether the patients were treated similarly in the two groups.
Randomized patients to receive either remdesivir or placebo
Randomized double-blinded placebo-controlled study design
This study was a prospective randomized double-blinded placebo-controlled trial. This is widely considered to provide the highest quality of clinical evidence by preventing confounding (hidden differences that impact outcomes) between the treatment and control groups. The authors assigned patients to remdesivir or placebo in a 2:1 ratio, which is an accepted strategy to control the number of patients in each treatment group. The authors stratified the randomization according to whether or not patients required respiratory support, which partially helped to ensure that the patients in one group of the trial were not far sicker than the patients in the other group at the beginning of the trial.
Compared improvement in remdesivir group with the placebo group
Unusual primary objectve
The study investigators defined a 6-point clinical severity scale, with 6 representing death, 5 representing ECMO or mechanical ventilation, 4 representing non-invasive positive pressure ventilation or high-flow oxygen therapy, 3 representing oxygen therapy, 2 representing hospital admission and 1 representing feeling well enough to go home from the hospital. They looked at how quickly each patient achieved a 2-point reduction on this scale. This clinical scale is reasonable, but it had not been used prior to this study. It is more widely accepted to analyze endpoints such as mortality that can be easily compared across studies.
Analyzed mortality as a secondary objective
The study investigators also compared the mortality between the two arms of the trial at multiple time points. This was useful because it presented the study results in terms of a widely accepted metric.
Use of standardized statistical methods
Although the 6-point clinical scale devised by the investigators was unusual, their use of a statistical technique, called the Cox proportional hazards model, to compare the two study groups is widely accepted in clinical literature.
The study investigators performed intention-to-treat analysis, a form of data analysis that analyzes the patients in their assigned group even if they have to drop out of the study for some reason. This is a widely accepted method to avoid obtaining misleading results.
The patients in the study were followed for 28 days, which was long enough to determine what happened to most, but not all, of the patients. At the end of the study, the clinical outcome had still not been determined for 16% of remdesivir patients and 20% of placebo patients.
No significant difference between remdesivir and placebo groups
Remdesivir Placebo Mortality 14% 13% Days to Recover 21 23 Adverse Events 66% 64%
After 28 days, 14% of remdesivir patients and 13% of placebo patients had died. This difference was not statistically significant. Of the remdesivir patients who survived, it took them an average of 21 days to clinically improve on the 6-point scale devised by the investigators, compared to 23 days for the placebo patients. This difference was not statistically significant. The rate of adverse events was similar between the remdesivir (66%) and placebo (64%) groups, suggesting that the medication was relatively safe.
Meaningful results even though trial was halted early
In the Methods section of the paper, the investigators described how they initially intended to enroll 453 patients (rather than the 236 who were actually enrolled), but the trial had to be halted early because the initial outbreak was controlled in Wuhan, Hubei, China. However, the number of patients enrolled in the trial was probably sufficient to tell us that even if remdesivir helps patients with SARS-CoV-2 pneumonia, it is unlikely to be the magic bullet many had hoped.
The study investigators arrived at a suitable conclusion based upon the data.