Remdesivir in COVID-19: A tale of two trials

The COVID-19 pandemic continues to cause perturbation worldwide, with healthcare systems struggling to cope with second waves and increasing incidence of newly infected patients. India may be witnessing a plateau effect, but the worst may be far from over. Today, we have a clearer understanding of effective therapeutic interventions compared to the early phase of the pandemic when little was known about a novel virus. The overall care and patient outcomes have distinctly improved over the past several months. The antiviral drug, remdesivir, was one of the therapeutic options that suggested benefit in a preliminary report, leading to its widespread use.¹ The full report of this randomized controlled trial was published earlier this month,² followed by the much larger SOLIDARITY trial, conducted by the World Health Organization (WHO).³ In light of these pivotal studies, let us consider the role of remdesivir in combating SARS-CoV-2 infection. 

The ACTT-1 trial

Study population

The Adaptive COVID-19 Treatment Trial (ACTT-1) was conducted between February 21, 2020, and April 19, 2020, involving 60 centers across 10 countries. The study population included hospitalized adult patients with COVID-19 who had evidence of lower respiratory tract infection. SARS-CoV-2 infection was confirmed by PCR or other types of assay. Patients had infiltrates on the chest radiograph or CT scan, or SPO_{2 ≤} 94% on room air, or required supplemental oxygen, or were mechanically ventilated. Patients with elevated AST or ALT levels more than five times normal, and evidence of renal dysfunction with an eGFR of less than 30 ml/min were excluded. Remdesivir was administered to 541patients, while 521 patients received a placebo.

Intervention arm 

In the intervention arm, intravenous remdesivir, 200 mg was administered on day 1, followed by 100 mg daily from day 2 to 10 or until discharge from hospital or death.

Control arm 

In the control arm, patients received matching placebo. Patients in both arms received standard care; other specific treatments for COVID-19 were followed according to local written policies or guidelines. If there was no written policy or guideline, other experimental therapies were prohibited during the study period. 

Outcomes

Primary outcome

The clinical status was evaluated based on an 8-category ordinal scale. 

1. Not hospitalized, no limitation of physical activity
2. Not hospitalized, with limitation of physical activity or requirement for home oxygen 
3. Hospitalization only as part of an infection control strategy 
4. Hospitalized, requiring medical care 
5. Hospitalized, requiring supplemental oxygen 
6. Hospitalized, requiring high-flow oxygen or non-invasive ventilation 
7. Hospitalized requiring invasive mechanical ventilation or extracorporeal membrane oxygenation 
8. Death 

The primary outcome was the time to recovery, defined as the time to reach categories 1,2, or 3 on the ordinal scale. Patients who received remdesivir recovered more rapidly compared to the placebo group. The median time to recovery was 10 (IQR: 9–11) days in the remdesivir arm compared with 15 (IQR: 13–18) days in the control arm. The rate ratio for recovery was 1.29 (1.12–1.49) in favor of remdesivir. The rate ratio for recovery was most pronounced among patients who required supplemental oxygen alone (category 5). Furthermore, the rate ratio for recovery was significant in the subgroup of patients who received remdesivir within 10 days of onset of illness 1.37 (95% CI, 1.14 to 1.64) compared to later 1.20 (95% CI, 0.94 to 1.52), suggesting that later treatment may not be effective. 

Secondary outcomes 

1. Clinical improvement at day 15: improvement in clinical status according to the ordinal scale score was higher in the remdesivir group, compared to the placebo group (OR: 1.5; 95% CI, 1.2 to 1.9). 
2. Mortality at day 15: was lower in the remdesivir group by Kaplan Meir estimate 6.7% (4.8–9.2) vs. 11.9% (9.4–15.0). Hazard ratio: 0.55 (0.36–0.83).
3. Mortality at day 29: was not significantly different between groups by Kaplan Meir estimate 11.4% (9.0–14.5) vs. 15.2% (12.3–18.6). Hazard ratio: 0.73 (0.52–1.03). A favorable effect of remdesivir on mortality was seen among patients in category 5 (requiring supplemental oxygen alone).
4. The requirement for non-invasive ventilation or high-flow oxygen among patients who did not require these interventions initially, was lower in patients who received remdesivir.

The SOLIDARITY trial 

Design 

The World Health Organization (WHO)-led SOLIDARITY trial involved the evaluation of four antiviral therapies involving drugs that were repurposed for the treatment of COVID-19. These included remdesivir, hydroxychloroquine, the lopinavir-ritonavir combination, and interferon-β1a. The trial employed an adaptive design; drugs that appeared ineffective could be discontinued during the course of the trial. The study involved 11,330 patients admitted to 405 hospitals across 30 countries. Patients were followed up until death or discharge from hospital. 

Study population

Hospitalized adult patients aged ≥18 years were eligible; patients were randomized to receive any of the study drugs available locally and matched with controls in a 1:1 ratio. Remdesivir was administered to 2743 patients compared with 2708 patients who acted as controls.

Intervention arm

In the intervention arm, remdesivir, 200 mg was administered intravenously on the first day, followed by 100 mg per day for 9 more days (same regime as the ACTT-1 trial). 

Control arm 

Local standard of care was followed; no placebo was administered. 

Results 

Primary outcome

The primary outcome, the in-hospital mortality, was not different between the remdesivir and the control groups: remdesivir vs. control: 301/2743 (11.0%) vs. 303/2708 (11.2%). Death rate ratio for remdesivir: 0.95 (0.81-1.11); p = 0.50. The mortality rate was not significantly different regardless of age, gender, the requirement for supplemental oxygen or mechanical ventilation, and presence of co-morbidities including diabetes, heart disease, chronic lung disease, and chronic liver disease. The mortality among patients who were on mechanical ventilation at randomization (invasive and non-invasive) was 98/254 (43%) in the remdesivir arm compared with 71/233 (37.8%) in the control arm. Mortality among patients who required supplemental oxygen at randomization was 12.2% vs. 13.8% and among those who did not require oxygen at randomization was 2.0% vs 2.1%. 

Figure 1. Kaplan-Meir curves for in-hospital mortality- remdesivir vs. control groups

Secondary outcomes

Requirement ventilator support 

Among remdesivir-treated patients, 2489/2743 patients did not require ventilator support at randomization; 295 of these patients (11.9%) received mechanical ventilation subsequently. In the control group, 2475/2708 patients did not require ventilator support at randomization; 284 of these patients (11.5%) required mechanical ventilation subsequently. The difference between patients progressed to require mechanical ventilation was not different between groups. 

Table 1. Number and outcomes of patients who required mechanical ventilation during the course of the trial  

	Remdesivir	Control
Patients who were not ventilated at randomization	2489	2475
Ventilated later, total	295 (11.9%)	284 (11.5%)
Ventilated later, died	117 (39.7%)	108 (38.0%)
Ventilated later, discharged	139 (47.1%)	146 (51.4%)
Ventilated later, still in hospital	39 (13.2%)	30 (10.6%)

Duration of hospitalization

The percentage of discharged patients who were still in hospital, remdesivir vs. control,  at day 7 (69% vs. 59%), day 14 (22% vs. 19%), and day 21 (9% vs. 8%) were similar between groups. 

Table 2. Percentage of patients who remained in hospital at different time points 

	Remdesivir: % of patients still in hospital	Control: % of patients still in hospital
Day 7	69%	59%
Day 14	22%	19%
Day 21	9%	8%

Comparison between the two trials 

The ACTT-1 trial was sponsored by the manufacturer with a more elaborate design and comprehensive data collection. This 1:1 randomized controlled trial demonstrated that remdesivir administration leads to a reduction in the duration of hospitalization, especially if administered within 10 days of onset of illness. However, no mortality difference was noted between groups at the 29-day follow-up. The beneficial effect on the duration of hospitalization seemed to be confined to patients who required supplemental oxygen alone, without the requirement for mechanical ventilation. It remains unknown if remdesivir may have a significant clinical benefit among the more severely ill patients. 

The SOLIDARITY trial was much larger, involving 30 countries, many from the relatively less developed world, at the peak of the pandemic wave. The study adopted a simple design, with a pragmatic limitation of data collection in constrained healthcare settings. Four repurposed antiviral therapies were evaluated against their respective controls, without the use of a placebo. Considering the extenuating circumstances under which the SOLIDARITY trial was conducted, there were several limitations.

The SOLIDARITY trial does not provide detailed information on the severity of illness or the duration of symptoms at randomization. The actual duration of hospital stay is not compared between groups. Details of treatment modalities employed in the control group is not provided and may have been highly variable across healthcare settings in different countries. The characteristics and outcomes of patients who were on invasive vs. non-invasive mechanical ventilation were not addressed separately. 

The available evidence from ACTT-1 suggests that remdesivir may be efficacious only if administered early in the course of the disease; the SOLIDARITY trial does not address outcomes related to the duration of symptoms. Besides, SOLIDARITY was conducted over a period of more than 6 months; the overall approach to the care of COVID-19 patients may have changed significantly over time, particularly regarding the modalities of respiratory support. Many guidelines and recommendations had initially suggested avoiding high-flow oxygen devices and non-invasive ventilation; however, these modalities have become the standard of care in early disease. It is also important to note that SOLIDARITY has been published online on a preprint platform and is yet to be peer-reviewed. 

Summary 

• The ACTT-1 trial suggests that remdesivir administration may reduce the hospital length of stay if administered within the first 10 days of onset of symptoms among less severely ill patients
• No mortality benefit was demonstrated with remdesivir in the ACTT-1 trial; a trend towards improved survival by day 29 was observed in patients who were only on supplemental oxygen, without the requirement for mechanical ventilation 
• The SOLIDARITY trial also did not reveal improved survival among patients who received remdesivir; furthermore, in contrast to the ACTT-1 trial, the number of patients who continued to receive hospitalized care did not differ with remdesivir compared to the control group at days 7, 14, and 21
• Detailed information regarding the severity of illness at baseline, the time point of remdesivir administration during the course of illness, and information regarding other treatment modalities employed are not provided in the SOLIDARITY trial
• It is likely that early remdesivir administration among patients with less severe illness may reduce the time to recovery; however, future controlled trials among subgroups of patients with less severe illness are required to confirm this hypothesis.

References 

1.         Remdesivir for the Treatment of Covid-19 — Preliminary Report. N Engl J Med. 2020;383(10):992-994. doi:10.1056/NEJMc2022236

2.         Beigel JH, Tomashek KM, Dodd LE, et al. Remdesivir for the Treatment of Covid-19 — Final Report. N Engl J Med. Published online May 22, 2020:NEJMoa2007764. doi:10.1056/NEJMoa2007764

3.         WHO Solidarity trial consortium, Pan H, Peto R, et al. Repurposed Antiviral Drugs for COVID-19 –Interim WHO SOLIDARITY Trial Results. Infectious Diseases (except HIV/AIDS); 2020. doi:10.1101/2020.10.15.20209817