Top 10: May 5

Home | April | ENG | ESP | ITA | DEU | FRA | TUR | VNM

By Christian Hoffmann &
Bernd S. Kamps

5 May

Top 10 Special on


Get it done!

On remdesivir, rumors of recovery, rolling reviews, and random noise

“That’s very exciting. Get it done, Daniel.” If you want to get an idea about the incredible pressure on and expectations from researchers, then please read the protocol from the White House (Trump 2020). NIAID’s Anthony Fauci and Gilead’s CEO Daniel O’Day make heroic attempts to explain the situation to decision makers. This pressure has its consequences. What we’ve seen during the last few days has probably never happened during the last 500 years on this planet: a drug the authorities (read: FDA) give “Emergency Use Authorization”, two days after the first randomized trial practically shows ineffectiveness. Crazy times. What is whispered at press conferences counts more than peer-reviewed scientific evidence. Because the window for learning is so short, the need to balance scientific rigor against speed seems inevitable. But should that really be the case? Last Friday, a smart comment in SCIENCE argued “against pandemic research exceptionalism”. Even in such a crisis, the rules of good science should not be thrown overboard. Releasing the full data is essential to allow scientists to understand studies (London 2020).

Let’s take a closer look on the scarce data we have on remdesivir:

  1. Compassionate Use Program: this was a fragmentary case series (Grein 2020) on some patients (only 53/61 patients were analyzed) with varying disease severity. Some improved, some didn’t: random noise. We believe, for a number of reasons, this case series published in the New England Journal of Medicine is a cautionary tale for “science in a hurry”, arousing false expectations. It might have been preferable to postpone the publication. However, Daniel O’Day, Gilead’s CEO, wrote the same day that “the majority” of patients “demonstrated clinical improvement”.
  2. NCT04257656: This multicentre trial, funded by the Chinese Ministry of Science and Technology, was conducted between Feb 6 and March 12 at ten hospitals in Hubei (Wang 2020). A total of 237 patients with pneumonia confirmed by chest imaging, oxygen saturation of 94% or lower on room air and within 12 days of symptom onset were randomized to receive 10 days of single infusions or placebo. Clinical improvement was defined as the days to the point of a decline of two levels on a six-point clinical scale (from 1=discharged to 6=death) or discharged alive from hospital, whichever came first. Patients were 65 years old (IQR 56–71), 56% male, many were co-treated with lopinavir (28%) and corticosteroids. The trial did not attain the predetermined full sample size because the outbreak was brought under control in China. However, from the analyzable data, remdesivir was not associated with a difference in time to clinical improvement (hazard ratio 1.23, 95% CI 0.87–1.75). Clinical improvement rates were 27% versus 23% at day 14 and 65% versus 58% at day 28. Day 28 mortality was 14% versus 13%. Of note, the viral load decreased similarly in both groups. Some patients with remdesivir had dosing prematurely stopped due to adverse events (12% versus 5%, mainly gastrointestinal symptoms and increases of liver enzymes). But let’s think positive: Time to recovery was “numerically” shorter in the remdesivir group than the control group, particularly in those treated within 10 days of symptom onset.
  3. SIMPLE 1: This Phase III trial evaluated 5-day and 10-day dosing durations in 397 hospitalized patients with severe COVID-19. On April 29, Gilead announced that no difference was seen in clinical improvement (odds ratio: 0.75, 95% CI 0.51 – 1.12] on day 14. The most common adverse events were nausea (9.5 %) and acute respiratory failure (8.3%). Grade 3 or higher liver enzyme elevations occurred in 7.3%, with 3.0 % discontinuing remdesivir. Gilead plans to submit the full data for publication in a peer-reviewed journal “in the coming weeks”. An expansion phase will enrol an additional 5,600 (!) patients around the world.
  4. ACTT (Adaptive COVID-19 Treatment Trial): Sponsored by NIH, this was the first phase III study launched in the United States. ACTT began on February 21 (the first participant was repatriated after being quarantined on the Diamond Princess) and enrolled 1,063 hospitalized patients with advanced COVID-19 and lung involvement. The design was “adaptive” to incorporate additional investigative treatments (and, surprisingly, to adapt new response criteria, see below). A total of 68 sites participated, among them 47 in the US and 21 in Europe and Asia. On April 16, it was decided to modify the primary endpoint (from mortality to time of recovery), facing “evolving clinical data”. Only 11 days later, an independent DSMB noted that remdesivir was better than placebo with regard to the new primary endpoint, time to recovery (defined as being well enough for hospital discharge or returning to normal activity level). With the drug, recovery was 31% faster (11 versus 15 days, p<0.001). For the initial primary endpoint, mortality, results suggested an only marginal benefit (8.0% versus 11.6%, p=0.059). “Whenever you have clear-cut evidence a drug works, you have an ethical obligation to immediately let the people in the placebo group know so they can have access to it”, said Anthony Fauci. Fine. But now we have waited another week. So where is the data? It will be probably not that easy to explain why such a major trial switched the key outcome measure only a few days before the interim analysis was done. At that time, they were still blinded for the results, correct? There is no doubt that this paper will have a complex and interesting discussion.

In the meantime, both the optimists and the pessimists can hold onto their opinions.


Table 1. Remdesivir, optimistic and pessimistic view
Study Optimistic view Pessimistic view
Compassionate Use Encouraging, the majority demon-strated clinical improvement No control group, fragmentary data without any message
SIMPLE Similar efficacy with 5- and 10-day dosing, no new safety signals No control group. Placebo for 5 vs 10 days would’ve produced the same results
NCT04257656 Time to recovery numerically shorter No effect on mortality, no effect on viral load, some side effects
ACTT Faster time to recovery, strong trend towards lower mortality (8 versus 12%) No significant effect on mortality in >1,000 patients, and is 4 days of “faster recovery” relevant?

What comes next?

Several additional trials are ongoing. Some have been suspended such as NCT04252664, a trial in adults with mild and moderate COVID-19, because during the last few weeks no eligible patients could be recruited. The second SIMPLE trial, NCT04292730 (GS-US-540-5774) is probably the most interesting study, evaluating the efficacy of two remdesivir regimens compared to standard of care in 600 patients with moderate COVID-19, with respect to clinical status assessed by a 7-point ordinal scale on day 11. Estimated study completion date is May 2020. INSERM in France has initiated a study evaluating remdesivir and other potential treatments, using a master protocol (SOLIDARITY) developed by WHO. This study (NCT04315948) is a multi-centre, adaptive, randomized, open clinical trial of the safety and efficacy of treatments of COVID-19 in hospitalized adults. Adults hospitalized for severe COVID-19 will be randomized to one of 4 treatment arms, including standard of care, remdesivir, lopinavir/r plus interferon ß-1a and hydroxychloroquine.

In the meantime, EMA’s human medicines committee (CHMP) has started a ‘rolling review’ of data. This speeds up the assessment of a promising investigational medicine during a public health emergency but does not imply that its benefits outweigh its risks. We’ll see what happens. By the way, the EUA allows for the distribution and emergency use of remdesivir only for the treatment of COVID-19; remdesivir remains an investigational drug and has not been approved anywhere. The fact sheet for health care providers is found here: FDA 2020.

Yes, very exciting.


FDA. Fact sheet for health care providers. Emergency use authorization (EUA) of Remdesivir. (GS-5734™).

Gilead Sciences. Gilead Announces Results From Phase 3 Trial of Investigational Antiviral Remdesivir in Patients With Severe COVID-19. Press release 30/04/2020. Full-text:

Grein J, Ohmagari N, Shin D, et al. Compassionate Use of Remdesivir for Patients with Severe Covid-19. N Engl J Med. 2020 Apr 10. PubMed: Full-text:

London AJ, Kimmelman J. Against pandemic research exceptionalism. Science. 2020 May 1;368(6490):476-477. PubMed: Full-text:

NIH. NIH clinical trial shows Remdesivir accelerates recovery from advanced COVID-19. Press release. Full-text.

Trump D et al. Remarks by President Trump and Members of the Coronavirus Task Force in Meeting with Pharmaceutical Companies. March 2, 2020. Full-text:

Wang Y, Zhang D, Du G, et al. Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial. April 29, 2020. Fulltext PDF: Full-text web page: