The FDA authorized emergency use of remdesivir in hospitalized patients with severe COVID-19 in May 2020. The authorization occurred after publication of the ACTT-1 trial, which was reviewed in a previous EBM Focus earlier this year. The ACTT-1 study population was quite sick: about 25 percent were intubated or on extracorporeal membrane oxygenation (ECMO), and 90 percent required oxygen. Data from those 1,059 patients indicated a shorter time to recovery with remdesivir (11 vs. 15 days) but no significant difference in mortality; however, the possibility of benefit could not be excluded. A subsequent NIH guideline on treatment of COVID-19 recommended 10 days of remdesivir in patients requiring high-flow oxygen, BiPAP, mechanical ventilation, or ECMO, with five days for all other patients on supplemental oxygen.

In September, an open-label trial sponsored by Gilead was published in JAMA evaluating remdesivir for hospitalized patients less ill than those in the ACTT-1 trial. This trial was followed by an expanded FDA authorization of remdesivir for any patient hospitalized with COVID-19. In this trial, 596 patients at least 12 years of age (median age 57 years, 61 percent male) hospitalized with laboratory-confirmed COVID-19 and moderate pneumonia (pulmonary infiltrates and room air oxygen saturation over 94 percent) were randomized to one of three groups: standard care alone, or standard care plus either five or 10 days of remdesivir (200 mg IV on day one followed by 100 mg IV daily for either four or nine days).

…the lack of benefit for any specific outcome such as mortality or length of hospital stay supports the impression that while remdesivir may have a benefit, it is a modest one at best.

Outcomes were evaluated on a seven-point ordinal scale of morbidity where death was one and discharge was seven, with improvement defined as an increase of at least two points on this seven-point scale. Median length of treatment was five days in the five-day remdesivir group and six days in the 10-day remdesivir group. At day 11, at least two points of improvement were observed in 65 percent of the 10-day remdesivir group, 70 percent of the five-day remdesivir group, and 61 percent of the standard-care-only group. The only significant difference was between five days of remdesivir and standard care alone. There were no significant differences in mortality, time to recovery, duration of oxygen therapy, or length of hospitalization amongst the three groups.

This trial suggests remdesivir may help patients improve a little faster, consistent with the ACTT-1 trial. There are, however, numerous concerns with the data. First, seeing benefit in the five-day but not the 10-day remdesevir group seems peculiar, and raises the possibility that this is a chance finding. It could also suggest that there is some harm associated with prolonged treatment, but that seems less likely given that the median duration of treatment in the 10-day group was only six days. Second, since the scale used is ordinal and not cardinal, improving by two points means different things depending on where on the scale you start. For example, getting off a ventilator is more meaningful than decreasing the level of oxygen support, so an outcome capturing the percent of patients who are improving by two points may miss subtle differences in the group. Third, the dichotomous outcome reporting treats patients who worsened and those who do not get better as one group. Finally, the lack of benefit for any specific outcome such as mortality or length of hospital stay supports the impression that while remdesivir may have a benefit, it is a modest one at best. This skepticism is reflected in the IDSA’s recommendation for five days of remdesivir only in hospitalized patients sick enough to require supplemental oxygen, with longer use reserved for those on mechanical ventilation or ECMO.

For more information, see the free topic COVID-19 (Novel Coronavirus) in DynaMed.

EBM Focus articles provide concise summaries of clinical trials most likely to inform clinical practice curated by theDynaMed editorial team.