Beigel et al, 2020 – Remdesivir: The ACTT-1 Study | INTENSIVE Review

Author: Dr Matthew Durie
Peer reviewer: Dr Vinodh Nanjayya

Beigel JH, Tomashek KM, Dodd LE, et al. Remdesivir for the Treatment of Covid-19 – Preliminary Report [published online ahead of print, 2020 May 22]. N Engl J Med. 2020;10.1056/NEJMoa2007764.


In a much anticipated paper, preliminary results from the US National Institute of Allergy and Infectious Diseases (NIAID) led ACTT-1 trial have been released, demonstrating the use of remdesivir was associated with a shortened time to recovery in patients hospitalised with COVID-19.

What was the design?

This was an international, multicentre, randomised, placebo-controlled trial of 1063 hospitalised adults with laboratory confirmed COVID-19 who received either remdesivir intravenously (200mg load then 100mg daily) or placebo for up to 10 days. Unlike a prior study of remdesivir (Wang Y et al. 2020) there was no limit on the duration of symptoms before study enrolment. Patients were excluded if they had AST/ALT derangement, renal impairment, pregnancy or breastfeeding, anticipated discharge within 72 hours or were involved in another trial.

The primary outcome was time to recovery measured in days after enrolment, where recovery was defined as either discharge from hospital, or hospitalisation without need for ongoing medical care (ie. for infection control purposes only). Failure to recover was censored at day 29 or death. Secondary outcomes included odds of improvement at day 15 and 29, mortality rate, and subgroup analyses depending on severity of illness at enrolment. The study reported preliminary results up to April 28 (see below).

What were the results?

The study found remdesivir was associated with a median time to recovery of 11 days (95% CI 9 to 12) vs 15 days (13 to 19) days (p< 0.0001), with a ‘recovery rate ratio’ (akin to hazard ratio for death) of 1.32 (95% CI 1.12 to 1.55).

In secondary outcomes, benefit from remdesivir was most apparent in those requiring oxygen therapy at enrolment (recovery rate ratio, RR 1.47, 95% CI 1.17 – 1.84) but not in those needing advanced respiratory supports including high flow oxygen (HFO) or non-invasive ventilation (NIV, RR 1.20, 95% CI 0.79 – 1.81) or invasive mechanical ventilation or extra-corporeal membrane oxygenation (RR 0.95, 95% CI 0.64 – 1.42). Mortality at day 15 was also reduced in this patient group (2.4% vs. 10.9%, HR 0.22, 95% CI 0.08 to 0.58), but not overall (HR 0.70, 95% CI 0.47 to 1.04).

When stratified by time from symptom onset of less than or greater than 10 days, the rate of recovery to recovery was similar, although it is conceivable that patients requiring more advanced respiratory supports and in whom remdesivir was not associated with benefit may have had a longer duration of illness (not reported in the study). There was no difference in the rate of adverse events or drug discontinuation.

Why are the results ‘preliminary’?

The authors describe the results as ‘preliminary’ as at the time of analysis, most patients had yet to reach 28 days following enrolment, as required by the study protocol. This was the result of an early interim safety analysis on April 27. The authors state the early analysis was due to an unexpectedly rapid rate of enrolment at which time the required 400 ‘recoveries’ (per the power calculation of 85% power to detect a recovery rate ratio of 1.35) had been met, however it also coincided with the release of data from another trial of remdesivir in China (Wang Y et al. 2020). This interim analysis was released ahead of the full paper by NIAID and widely publicised in the media (see Intensive review) and importantly, from this time on, the treatment allocation could be unblinded, and patients in the placebo group could be given remdesivir. As 31% of patients at that time were still yet to recover or reach the 29-day end point, this has implications for the validity of the planned day 29 ‘final’ analysis.

This was in fact the second change to trial protocol during the study, as the primary outcome had been changed from the difference in clinical status at day 15, to the time to clinical improvement as above. The authors state that this was done due to evolving understanding of COVID-19 in March, and without knowledge of outcome data.

What are the strengths of the study?

This is an international cohort of well-matched placebo and control groups with adequate power to assess the primary outcome and a clinically relevant end point (as opposed to other trials which have reported outcomes like duration of RNA positivity), although one could argue that reducing mortality remains the hope of all COVID-19 therapies being tested.

What are the weaknesses?

From a multicentre study over two months, involving over 73 sites it is unclear how the 1107 patients who were assessed for eligibility were chosen. For comparison, in New York alone 7914 patients with COVID-19 were admitted to 25 hospitals over a two-week period in March (Rosenberg E. et al. 2020). It is likely that these 1107 patients are only a small fraction of all patients who would have met the inclusion criteria. Secondly, the trial reports difficulty with maintaining an adequate placebo at some sites, with a risk of clinician unblinding. Thirdly, it is difficult to be confident that the design was not unduly influenced by Gilead (the producer of remdesivir) as Gilead employees were involved in protocol design and monitoring of the trial although the authors state that all final decisions were made by the NIAID team.

What does this mean for an intensivist?

Remdesivir may be beneficial to patients with moderate to severe COVID-19, who are receiving supplemental oxygen in hospital. This study does not provide any evidence of benefit for patients with more severe disease, requiring intensive care level respiratory supports (including HFNO, NIV or invasive ventilation). The benefit (if any) of remdesivir for intensive care patients therefore remains very uncertain.


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.