Drug repurposing for COVID-19: current evidence from randomized controlled adaptive platform trials and living systematic reviews

Abstract

Introduction

The coronavirus disease 2019 (COVID-19) pandemic resulted in a race to develop effective treatments largely through drug repurposing via adaptive platform trials on a global scale. Drug repurposing trials have focused on potential antiviral therapies aimed at preventing viral replication, anti-inflammatory agents, antithrombotic agents and immune modulators through a number of adaptive platform trials. Living systematic reviews have also enabled evidence synthesis and network meta-analysis as clinical trial data emerge globally.

Sources of data

Recent published literature.

Areas of agreement

Corticosteroids and immunomodulators that antagonize the interleukin-6 (IL-6) receptor have been shown to play a critical role in modulating inflammation and improving clinical outcomes in hospitalized patients. Inhaled budesonide reduces the time to recovery in older patients with mild-to-moderate COVID-19 managed in the community.

Areas of controversy

The clinical benefit of remdesivir remains controversial with conflicting evidence from different trials. Remdesivir led to a reduction in time to clinical recovery in the ACTT-1 trial. However, the World Health Organization SOLIDARITY and DISCOVERY trial did not find a significant benefit on 28-day mortality and clinical recovery.

Growing points

Other treatments currently being investigated include antidiabetic drug empagliflozin, antimalarial drug artesunate, tyrosine kinase inhibitor imatinib, immunomodulatory drug infliximab, antiviral drug favipiravir, antiparasitic drug ivermectin and antidepressant drug fluvoxamine.

Areas timely for developing research

The timing of therapeutic interventions based on postulated mechanisms of action and the selection of clinically meaningful primary end points remain important considerations in the design and implementation of COVID-19 therapeutic trials.

Introduction

The coronavirus disease 2019 (COVID-19) pandemic to date has resulted in >600 million COVID-19 infections with over 6 million fatalities¹ and spurred rapid, collaborative, international, clinical translational research of drug repurposing that is unprecedented. Novel drug discovery is associated with significant resource implications in terms of financial costs and time, with an average duration of 10–15 years to take a drug from bench to bedside.² Although the majority of COVID-19 patients to date have been managed in primary care, a subset develop severe COVID-19, of which acute respiratory distress syndrome, cytokine release syndrome and multiorgan involvement are prominent features.³ To deal with the scale and speed of the COVID-19 pandemic, drug repurposing for COVID-19 has focused on four broad strategies: antiviral therapies aimed at preventing viral multiplication, anti-inflammatory agents, antithrombotic agents and immune modulators.

There are multiple known variants of SARS-CoV-2, with the World Health Organization (WHO) classifying coronavirus variants into ‘variants of interest’, ‘variants of concern’ and ‘variants of high consequence’. Variants have significant consequences for COVID-19 diagnostics and therapeutics alongside disease transmissibility, severity, host immune response and vaccine coverage.^4,5 The generalized approach to therapeutics for COVID-19 is targeted at managing viral load and inflammation. The risk of developing inflammation depends on many factors including viral variant and vaccine status. To date, variants of concern can reduce efficacy of antibody-based interventions.⁵ Antivirals may be less likely to have altered effectiveness because these are generally not subject to the same selective pressures on variation.

The rapid implementation of adaptive, pragmatic, multicenter, collaborative randomized national and international trials has enabled a record number of candidate drugs to be robustly investigated, with tens of thousands of patients enrolled in hundreds of countries. The Multi-Arm Multi Stage (MAMS) trial design was first developed to increase the speed and efficiency of large, randomized trials in the field of oncology and other complex interventions.⁶ In a classical randomized clinical trial design, patients are allocated at random to a study treatment arm from a short list of alternative interventions and clinical trial questions are traditionally tested sequentially. In comparison, the adaptive platform trial design allows multiple intervention arms to be tested systematically and simultaneously. Adaptive trials enable pre-planned interim analyses throughout the lifetime of the trial so that drugs that display poor efficacy are dropped and others added as the trial progresses.^6,7 These trials can be efficient to deliver from a resource point of view, require fewer patients and less recruitment time compared to traditional two arm randomized controlled trials.

Adaptive trial platforms for COVID-19 include RECOVERY,⁸ WHO SOLIDARITY Plus,⁹ REMAP-CAP,¹⁰ ACTT-1,¹¹ ACTT-2,¹² DISCOVERY,¹³ ACTIV¹⁴ and PRINCIPLE.¹⁵ In just over 12 months, adaptive trials delivered conclusive evidence on the efficacy and safety of various drugs repurposed for the treatment of COVID-19.

Alongside these platform trials, the ‘British Medical Journal’ living systematic review¹⁶ and network meta-analysis allows assessments of comparative effectiveness of multiple COVID-19 therapies as new trial data emerge. The WHO has also published living guidelines¹⁷ on therapeutic drugs for COVID-19.

We review current clinical evidence derived from adaptive platform trials in the management of COVID-19, also summarized in Tables 1 and 2.

Table 1.

COVID-19 adaptive platform trial treatments effective in the management of COVID-19

Trial	Intervention/setting	Patient numbers (n)	Proposed mechanism of action	Primary and secondary outcome measures	Results
RECOVERY18	Dexamethasone PO/IV 6 mg OD for up to 10 days vs SoC Clinical setting: hospitalized patients	Dexamethasone—2104 SoC—4321	Anti-inflammatory Immune modulation	1-°to 28-day mortality	• 482 patients (22.9%) in the dexamethasone group vs 1110 patients (25.7%) in the usual care group died within 28 days following randomization (age-adjusted rate ratio, 0.83; 95% confidence interval [CI], 0.75 to 0.93; P < 0.001)
RECOVERY19	Tocilizumab IV (dosage based on weight; 400 mg–800 mg) vs SoC Clinical setting: hospitalized patients	Tocilizumab—2022 SoC—2094	IL-6 receptor antagonist Immune modulation	1- to°28-day mortality	• 621 (31%) out of 2022 patients allocated to tocilizumab and 729 (35%) out of 2094 patients allocated to usual care died within 28 days (rate ratio 0·85; 95% CI 0·76 to 0·94; P = 0·0028)
RECOVERY20	Baricitinib 4 mg PO OD for 10 days vs SoC	Baricitinib—4148 SoC—4008	Immune modulation	1- to°28-day mortality	• 513 (12%) of 4148 patients allocated to baricitinib and 546 (14%) of 4008 patients allocated to usual care died within 28 days, representing a reduction of 13% (age-adjusted rate ratio 0·87; 95% CI 0·77 to 0·98; P = 0·026) • No significant excess in deaths or infection due to non-COVID-19 causes or excess of thrombosis were observed in the baricitinib group • Benefit of baricitinib was consistent regardless other COVID-19 treatments that patients were receiving (corticosteroids, tocilizumab or remdesivir)
REMAP-CAP10	Tocilizumab IV (dosage based on weight; 400 mg–800 mg) vs sarilumab IV 400 mg single dose vs SoC Tocilizumab could be repeated at 12—24 h later at the discretion of the treating clinician if clinical improvement was judged insufficient Clinical setting: hospitalized patients	Tocilizumab – 353 Sarilumab—48 SoC—402	IL-6 receptor antagonist Immune modulation	1—cardiovascular and respiratory organ support–free days	• Tocilizumab—median number of organ support-free days was 10 (interquartile range, –1 to 16). Median adjusted cumulative odds ratio was 1.64 (95% credible interval, 1.25 to 2.14) compared with control • Sarilumab—median number of organ support-free days was 11 (interquartile range, 0 to 16). Median adjusted cumulative odds ratio was 1.76 (95% credible interval, 1.17 to 2.91) compared with control • Tocilizumab and sarilumab were found to be equally effective for patients with severe COVID-19 compared to ‘no immune modulation’
ACTT-111	Remdesivir (200 mg loading dose on day 1, followed by 100 mg daily for up to nine additional days) vs placebo Clinical setting: hospitalized patients	Remdesivir—541 Placebo—521	Antiviral	1—time to recovery (defined by either discharge from the hospital or hospitalization for infection-control measures)	• Remdesivir—median recovery time of 10 days (95% CI, 9 to 11), compared with 15 days (95% CI, 13 to 18) • Placebo arm—median recovery time of 15 days (95% CI, 13 to 18) • Rate ratio recovery 95% CI, 1.12 to 1.49; P < 0.001, by a log-rank test • Kaplan–Meier estimates of mortality at day 29 were 11.4% in the remdesivir arm and 15.2% in the placebo arm (HR, 0.73; 95% CI, 0.52 to 1.03)
ACTT-212	Remdesivir (≤10 days) + baricitinib (≤14 days) vs Remdesivir + placebo (control) Remdesivir dose (200 mg loading dose on day 1, followed by 100 mg daily for up to nine additional days) Baricitinib dose (4 mg OD for duration of hospital stay up to 14 days) Clinical setting: hospitalized patients	Remdesivir + baricitinib—515 Remdesivir + placebo—518	Antiviral Immune modulation	1—time to recovery	• Remdesivir + baricitinib—Median time to recovery in the baricitinib arm was 7 days (95% CI, 6 to 8). • Remdesivir + placebo—median time to recovery was 8 days in the control arm (95% CI, 7 to 9) (rate ratio for recovery, 1.16; 95% CI, 1.01 to 1.32; P = 0.03) • In patients receiving remdesivir + baricitinib there was a 30% higher odds of improvement in clinical status at day 15 (odds ratio, 1.3; 95% CI, 1.0 to 1.6)
PRINCIPLE21	Budesonide inhaler (800 μg twice daily for 14 days) vs SoC Clinical setting: community patients	Budesonide—787 SoC alone—1028	Anti-inflammatory	1—first self-reported recovery, and hospitalization/death related to COVID-19 at 28 days (co-primary endpoints)	• Budesonide group first self-reported recovery—11.8 days [95% BCI 10·0 to 14·1] • SoC group first self-reported recovery—14.7 days [12·3 to 18·0] • HR 1·21 [95% BCI 1·08 to 1·36]), with a probability of superiority greater than 0·999 • Budesonide group estimated rate of hospital admission or death 6·8% (95% BCI 4·1 to 10·2) SoC group estimated rate of hospital admission or death 8·8% (5·5 to 12·7); (estimated absolute difference 2·0% [95% BCI -0·2 to 4·5]; odds ratio 0·75 [95% BCI 0·55 to 1·03]), with a probability of superiority 0·963, below the prespecified superiority threshold of 0·975

Key: OD—once daily; BD—twice daily; PO—per oral; IV—intravenous.

Table 2.

COVID-19 adaptive platform trial treatments not effective in the management of COVID-19

Trial	Intervention/clinical setting	Patient numbers (n)	Proposed mechanism of action	Primary and secondary outcome measures	Results
RECOVERY22	Aspirin 150 mg PO OD + SoC vs SoC until discharge Clinical setting: hospitalized patients	Aspirin + SoC—7351 SoC alone—7541	Antithrombotic Anti-inflammatory	1- °to 28-day mortality	• A total of 1222 (17%) patients allocated to aspirin and 1299 (17%) patients allocated to usual care died within 28 days (rate ratio 0·96; 95% confidence interval [CI] 0·89 to 1·04; P = 0·35) • No significant difference between treatment arms
RECOVERY23	Colchicine BD PO/NG for 10 days + SoC vs SoC Colchicine dose (1 mg after randomization followed by 500 mcg 12 h later and then 500 mcg BD PO/NG for 10 days in total or until discharge) Clinical setting: hospitalized patients	Colchicine + SoC—5610 SoC—5730	Anti-inflammatory Immune modulation	1- to°28-day mortality	• Proportion of patients discharged from hospital alive within 28 days Colchicine (70%) vs SoC (70%); rate ratio 0.98; 95% CI 0.94 to 1.03; P = 0.44) or the duration of hospitalization (median 10 days vs 10 days) • No significant difference between treatment arms
RECOVERY24	Azithromycin 500 mg OD PO/IV + SoC for 10 days or until discharge vs SoC alone Clinical setting: hospitalized patients	Azithromycin + SoC- 2582 SoC alone—5181	Antiviral Immune modulation	1- to°28-day mortality	• 561 (22%) patients allocated to azithromycin and 1162 (22%) patients allocated to usual care died within 28 days (rate ratio 0·97, 95% CI 0·87 to 1·07;  P = 0·50) • No significant difference was seen in the proportion meeting the composite endpoint of death or risk of invasive mechanical ventilation (risk ratio 0·95, 95% CI 0·87 to 1·03; P = 0·24)
RECOVERY25	Lopinavir-ritonavir (400 mg and 100 mg respectively) PO BD vs SoC alone for 10 days or until discharge, whichever occurred sooner Clinical setting: hospitalized patients	Lopinavir-ritonavir −1616 SoC—3424	Antiviral Immune modulation	1- to°28-day mortality	• A total of 374 (23%) patients allocated to lopinavir-ritonavir and 767 (22%) patients allocated to usual care died within 28 days (rate ratio 1·03, 95% CI 0·91-1·17; P = 0·60). • No significant difference was seen in the proportion of patients who met the composite endpoint in patients not on invasive mechanical ventilation at baseline (risk ratio 1·09, 95% CI 0·99 to 1·20; P = 0·092)
RECOVERY26	Hydroxychloroquine PO vs SoC Hydroxychloroquine loading dose 800 mg at baseline and at 6 h, followed by 400 mg at 12 h after the initial dose and then every 12 h for the next 9 days or until discharge, whichever occurred sooner Clinical setting: hospitalized patients	Hydroxychloroquine—1561 SoC—3155	Antiviral Anti-inflammatory	1- to°28-day mortality	• Death within 28 days occurred in 421 patients (27.0%) in the hydroxychloroquine group and in 790 (25.0%) in the usual-care group (rate ratio, 1.09; 95% CI, 0.97 to 1.23; P = 0.15) • There was no significant difference in 28-day mortality between hydroxychloroquine recipients and the usual care group

(Continued)

Table 2.

Continued.

Trial	Intervention/clinical setting	Patient numbers (n)	Proposed mechanism of action	Primary and secondary outcome measures	Results
WHO SOLIDARITY9	Remdesivir 200 mg IV on day 0 and 100 mg on days 1—9 vs Lopinavir (without IFN) 400 mg PO BD for 14 days vs IFNβ-1a vs IFNβ-1a plus Lopinavir 400 mg PO BD for 14 days vs SoC IFNβ-1a dose- 44 μg SC three doses Day 0,3,6/subcutaneous IFNβ-1a; 10 μg IV daily for 6 days Clinical setting: hospitalized patients	Remdesivir—2750 Hydroxychloroquine—954 Lopinavir—1411 IFN—1412 IFN + Lopinavir—651 SoC—4088	Antiviral Anti-inflammatory Immune modulation	1—in-hospital mortality 2°—progression to ventilation if not already ventilated, and time-to-discharge from hospital	• On interim analysis: • Death occurred in 301 of 2743 patients randomized to remdesivir vs 303 of 2708 in the control group (rate ratio, 0.95; 95% CI, 0.81 to 1.11; P = 0.50) • Death occurred in 104 of 947 patients randomized to hydroxychloroquine vs 84 of 906 in the control group (rate ratio, 1.19; 95% CI, 0.89 to 1.59; P = 0.23) • Death occurred in 148 of 1399 patients randomized to lopinavir vs 146 of 1372 in the control group (rate ratio, 1.00; 95% CI, 0.79 to 1.25; P = 0.97) • Death occurred in 243 of 2050 patients randomized to IFN vs 216 of 2050 in the control group (rate ratio, 1.16; 95% CI, 0.96 to 1.39; P = 0.11) • None of the therapies showed a significant reduction in mortality, duration of hospitalization or initiation or duration of ventilation
DISCOVERY27	Lopinavir (400 mg)/ritonavir (100 mg) PO BD for 14 days vs Lopinavir/ritonavir PO BD for 14 days –IFN-β1a vs Hydroxychloroquine vs SoC IFNβ-1a dose- 44 μg SC three doses Day 0, 3, 6/subcutaneous IFNβ-1a; 10 μg IV daily for 6 days Hydroxychloroquine loading dose 800 mg at baseline and at 6 h, followed by 400 mg at 12 h after the initial dose and then every 12 h for the next 9 days or until discharge, whichever occurred sooner Clinical setting: hospitalized patients	Lopinavir/ritonavir −145 Lopinavir/ritonavir–IFN–β-1a—145 Hydroxychloroquine—145 SoC—148	Antiviral Anti-inflammatory Immune modulation	1—clinical status at day 15 on the 7-point ordinal scale of the WHO Master Protocol	• Lopinavir/ritonavir vs control, adjusted odds ratio (aOR) 0.83, (95% CI 0.55 to 1.26, P = 0.39) • Lopinavir/ritonavir–IFN–β-1a vs control, aOR 0.69 (95% CI 0.45 to 1.04, P = 0.08) • Hydroxychloroquine vs control, aOR 0.93 (95% CI 0.62 to 1.41, P = 0.75)
DISCOVERY28	Remdesivir 200 mg IV on day 0 and 100 mg on days 1—9 + SoC vs SoC only Clinical setting: hospitalized patients	Remdesivir—429 SoC—428	Antiviral Immune modulation	• 1°- Clinical status at day 15 on the 7-point ordinal scale of the WHO Master Protocol	• No significant difference in WHO Ordinal status between treatment groups at day 15 (odds ratio 0·98 [95% CI 0·77 to 1·25]; P = 0·85), time to hospital discharge, 28-day all-cause mortality, or SARS-CoV-2 viral kinetics. • Exploratory subgroup analysis—in patients not on mechanical ventilation or extracorporeal membrane oxygenation (ECMO) at randomization, the hazard for the composite endpoint of death, ECMO or new mechanical ventilation lower in the remdesivir group vs SoC alone (HR 0·66 [95% CI 0·47 to 0·91]; P = 0·010)

(Continued)

Table 2.

Continued.

Trial	Intervention/clinical setting	Patient numbers (n)	Proposed mechanism of action	Primary and secondary outcome measures	Results
ACTIV-629	Ivermectin 400 μg/kg for three consecutive days vs Placebo Clinical setting: Community patients with mild-to-moderate COVID-19	Ivermectin—1591 Placebo—774	Antiviral	1—time to sustained recovery, defined as the third of three consecutive symptom free days 2°—hospitalization or death at day 28	• A hazard ratio (HR) of 1.07 was reported for improvement in time to recovery (95% credible interval [CrI], 0.96-1.17; posterior P value [HR >1] = .91). The median time to recovery in the ivermectin group was 12 days (IQR, 11-13) versus 13 days (IQR, 12-14) in the placebo group. Ten hospitalizations/ deaths were reported in the ivermectin group versus 9 events in the placebo group (1.2% vs 1.2%; HR, 1.1 [95% CrI, 0.4-2.6]) • Treatment did not lower incidence of hospitalization or death among outpatients with COVID-1929
ACTIV-630	Inhaled fluticasone furoate 200 μg once daily for 14 days vs Placebo Clinical setting: Community patients with mild-to-moderate COVID-19	Fluticasone furoate—656 Placebo—621	Anti-inflammatory	1—time to sustained recovery, defined as the third of three consecutive symptom free days. 2°—Hospitalization or death at day 28	• No improvement in time to recovery was observed with fluticasone compared with placebo (HR 1.01, 95% credible interval [CrI] 0.91–1.12; posterior probability for benefit [HR > 1] = 0.56). • A total of 24 participants (3.7%) in the fluticasone arm required urgent care or were hospitalized compared with 13 (2.1%) in the placebo arm (HR 1.9, 95% CrI 0.8–3.5; posterior probability for benefit [HR < 1] = 0.03).
PRINCIPLE31	Oral doxycycline (200 mg on day 1, then 100 mg once daily for the following 6 days) plus usual care vs Usual care plus other interventions vs Usual care alone	Doxycycline + usual care—780 Usual care plus other interventions—780 Usual care alone—948	Antiviral	1—time to first self-reported recovery and hospital admission or death related to COVID-19 at 28 days (co-primary endpoints)	• There was limited evidence of a difference in median time to first self-reported recovery between the usual care plus doxycycline group and the usual care only group (9·6 [95% Bayesian credible interval [BCI] 8·3 to 11·0] days vs 10·1 [8·7 to 11·7] days, HR 1·04 [95% BCI 0·93 to 1·17]). The estimated benefit in median time to first self-reported recovery was 0·5 days [95% BCI -0·99 to 2·04]. The probability of a clinically meaningful benefit (defined as ≥1·5 days) was 0·10. Hospitalization or death related to COVID-19 occurred in 41 (crude percentage 5·3%) participants in the usual care plus doxycycline group compared to 43 (4·5%) in the
PRINCIPLE32	Azithromycin 500 mg daily for 3 days vs Usual care plus other interventions vs Usual care alone	Azithromycin + usual care—540 Usual care + other interventions—850 Usual care alone—875	Antiviral	1—Time to first self-reported recovery and hospital admission or death related to COVID-19 at 28 days (co-primary endpoints)	• No clinically meaningful benefit in the azithromycin plus usual care group in time to first reported recovery versus usual care alone was observed (HR 1·08, 95% BCI 0·95 to 1·23), equating to an estimated benefit in median time to first recovery of 0·94 days (95% BCI −0·56 to 2·43). The probability of a clinically meaningful benefit of at least 1·5 days in time to recovery was 0·23. A total of 16 (3%) of 500 participants in the azithromycin plus usual care group and 28 (3%) of 823 participants in the usual care alone group were hospitalized (absolute benefit in percentage 0·3%, 95% BCI −1·7 to 2·2).
PRINCIPLE33	Colchicine vs Usual care vs Usual care + other interventions	Colchicine—156 Usual care alone—1145 Usual care + other interventions—1454	Antiviral Anti-inflammatory	1—Time to first self-reported recovery and hospital admission or death related to COVID-19 at 28 days (co-primary endpoints)	• Time to first self-reported recovery was similar compared with usual care with an estimated HR of 0.919 [95% credible interval 0.72 to 1.16] and an estimated increase of 1.14 days [−1.86 to 5.21] in median time to self-reported recovery for colchicine versus usual care. The probability of meaningful benefit in time to recovery was determined to be low at 1.8%. The rate of COVID-19 related hospitalizations/deaths were similar in the colchicine group versus usual care, with an estimated odds ratio of 0.76 [0.28 to 1.89] and an estimated difference of −0.4% [−2.7% to 2.4].

Key: SC—subcutaneous.

COVID-19 adaptive platform trials

RECOVERY trial

The RECOVERY trial has recruited > 48 000 patients across 200 active sites in the UK to date.⁸ Evaluations have been published for drug repurposing with aspirin, azithromycin, colchicine, dexamethasone, hydroxychloroquine, lopinavir-ritonavir and tocilizumab.

Dexamethasone

Dexamethasone is a glucocorticoid used to treat a variety of inflammatory conditions.²¹ Glucocorticoids were chosen as a drug to repurpose in COVID-19 due to their potential to modulate inflammation-mediated lung injury, thereby reducing risk of progression to respiratory failure and death. However, at the start of the pandemic they were not initially favored as a treatment because glucocorticoids had not previously shown effectiveness in severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS).^34,35

Hospitalized patients with COVID-19 were randomly assigned to receive oral or intravenous dexamethasone (at a dose of 6 mg once daily) for up to 10 days or to standard care alone.¹⁸ The primary outcome was 28-day mortality. Overall, 2104 patients were assigned to treatment with dexamethasone and 4321 to standard of care (SoC). A total of 482 patients (22.9%) in the dexamethasone group and 1110 patients (25.7%) in the usual care group died within 28 days following randomization (age-adjusted rate ratio [RR], 0.83; 95% confidence interval [CI], 0.75 to 0.93; P < 0.001).¹⁸ There was considerable variability in the proportional and absolute between-group differences in mortality according to the level of respiratory support that the patients were receiving at the time of randomization. Among patients receiving mechanical ventilation, the incidence of death was lower in the dexamethasone group compared to the usual care group (29.3% vs 41.4%; RR, 0.64; 95% CI, 0.51 to 0.81). In patients receiving oxygen without invasive mechanical ventilation the risk of death was also lower compared to those in the usual care group (23.3% vs 26.2%; RR, 0.82; 95% CI, 0.72 to 0.94). No difference in risk of death was seen in those who did not need respiratory support at randomization (17.8% vs 14.0%; RR, 1.19; 95% CI, 0.92 to 1.55).

Dexamethasone was the first drug to be recommended by the WHO for COVID-19 treatment in September 2020.

Tocilizumab

Tocilizumab is a monoclonal antibody that antagonizes the membrane bound and soluble forms of the IL-6 receptor (IL-6R/sIL-6R) licensed for the treatment of rheumatoid arthritis.³⁶ IL-6 is a pleiotropic cytokine that regulates the immune response to infections and elevated levels are associated with severe clinical outcomes in COVID-19, including respiratory failure and death. Hospitalized patients with hypoxia (oxygen saturation < 92% on air or requiring oxygen therapy) and evidence of systemic inflammation (C-reactive protein ≥75 mg/l) were randomly assigned in a 1:1 ratio to usual SoC alone versus usual SoC plus intravenous tocilizumab at a dose of 400 mg–800 mg (dosage was based on weight).²² Clinicians could choose to give a second dose 12–24 h later if the patient’s condition had not improved. Twenty-eight day mortality in the intention-to-treat population was the primary outcome measure.

A total of 4116 adults of were included in the assessment of tocilizumab, including 3385 (82%) patients receiving systemic corticosteroids. A total of 621 (31%) out of 2022 patients allocated to tocilizumab and 729 (35%) out of 2094 patients allocated to usual care died within 28 days (RR 0.85; 95% CI 0.76 to 0.94; P = 0.0028).¹⁹ Patients allocated tocilizumab were less likely to reach the composite endpoint of invasive mechanical ventilation or death if not on mechanical ventilation at baseline (35% vs 42%; risk ratio 0.84; 95% CI 0.77 to 0·92; P < 0.0001). There was an increased likelihood of being discharged from hospital within 28 days in patients allocated to the tocilizumab arm compared to standard care (57% vs 50%; RR 1.22; CI 1.12 to 1.33; P < 0.0001). These clinical benefits of tocilizumab were additional to the clinical benefits of systemic corticosteroids and were independent of the level of respiratory support. Tocilizumab was the second drug to be recommended by the WHO for COVID-19 treatment in July 2021. The recommendation is that tocilizumab should be given in combination with corticosteroid therapy.

Aspirin

Aspirin was investigated as a candidate drug for the treatment for COVID-19 due to its antithrombotic properties. It is currently licensed as an analgesic and an antithrombotic in ischemic stroke and ischemic heart disease.²¹ The RECOVERY trial randomized hospitalized patients with COVID-19 to either usual SoC plus aspirin 150 mg once daily versus usual SoC until discharge.²² The primary outcome was 28-day mortality. A total of 7351 patients were randomly allocated to receive aspirin and 7541 patients to receive usual care alone. Overall, there was no significant difference in 28-day mortality between the two groups. A total of 1222 (17%) patients allocated to aspirin and 1299 (17%) patients allocated to usual care died within 28 days (RR 0.96; 95% confidence interval [CI] 0.89 to 1.04; P = 0.35).²² An absolute reduction in thrombotic events of 0.6% (SE 0.4%) and an absolute increase in major bleeding events of 0.6% (SE 0.2%) were associated with the use of aspirin. There was no significant difference in the proportion meeting the composite endpoint of invasive mechanical ventilation or death in patients not on invasive mechanical ventilation at baseline (21% vs 22%; risk ratio 0.96; 95% CI 0.90 to 1.03; P = 0.23).²²

Colchicine

Colchicine is a potent anti-inflammatory agent licensed as a treatment for gout and prophylaxis against recurrent polyserositis in familial Mediterranean fever.³⁶ Hospitalized patients were randomly allocated to either usual standard care or standard care plus colchicine twice daily for 10 days or until discharge with the primary outcome measure being 28-day mortality.²³ A total of 5610 patients were randomly allocated to receive colchicine and 5730 patients to receive usual care alone. No significant difference was seen in the proportion of patients discharged from hospital alive within 28 days (70% vs 70%; RR 0.98; 95% CI 0.94 to 1.03; P = 0.44) or the duration of hospitalization (median 10 days vs 10 days). No significant difference was seen in the proportion meeting the composite endpoint of invasive mechanical ventilation or death (25% vs 25%; risk ratio 1.02; 95% CI 0.96 to 1.09; P = 0.47) in patients not on invasive mechanical ventilation at study enrolment.²³

Azithromycin

Azithromycin is an antimicrobial used to treat a variety of bacterial infections.³⁶ It also displays immunomodulatory actions. Patients were randomized to SoC plus azithromycin 500 mg once per day by mouth or intravenously for 10 days or until discharge (or allocation to one of the other RECOVERY treatment groups) versus SoC alone.²⁴ Twenty-eight-day all-cause mortality was the primary outcome measure, assessed in the intention-to-treat population.

A total of 2582 patients were randomly allocated to receive azithromycin and 5181 patients were randomly allocated to usual care alone. No significant difference was seen in the proportion of patients discharged from hospital alive within 28 days (RR 1.04, 95% CI 0.98 to 1.10; P = 0.19).²⁴ Overall, 561 (22%) patients allocated to azithromycin and 1162 (22%) patients allocated to usual care died within 28 days (RR 0.97, 95% CI 0.87 to 1·07; P = 0.50) and no significant difference was seen in the proportion meeting the composite endpoint of death or risk of invasive mechanical ventilation (risk ratio 0.95, 95% CI 0.87 to 1.03; P = 0.24).²⁴

Lopinavir-ritonavir

Lopinavir-ritonavir is an antiretroviral treatment licensed for the treatment of human immunodeficiency virus (HIV).³⁶ On the basis of preclinical and observational studies, lopinavir-ritonavir was identified as a potential candidate for drug repurposing for COVID-19. Patients were randomly allocated to either usual SoC plus lopinavir-ritonavir (400 mg and 100 mg, respectively) by mouth for 10 days or until discharge (or one of the other RECOVERY treatment groups: hydroxychloroquine, dexamethasone or azithromycin) or usual SoC alone.²⁵ The primary outcome was 28-day all-cause mortality in the intention-to-treat population.

A total of 1616 patients were randomly allocated to receive lopinavir-ritonavir and 3424 patients to receive usual care. There was no significant difference in the proportion of patients discharged from hospital alive within 28 days (RR 0.98, 95% CI 0.91 to 1.05; P = 0.53).²⁵ A total of 374 (23%) patients allocated to lopinavir-ritonavir and 767 (22%) patients allocated to usual care died within 28 days (RR 1.03, 95% CI 0.91 to 1·17; P = 0.60). No significant difference was seen in the proportion of patients who met the composite endpoint in patients not on invasive mechanical ventilation at baseline (risk ratio 1.09, 95% CI 0.99 to 1.20; P = 0.092).²⁵

Hydroxychloroquine

Hydroxychloroquine and chloroquine are licensed for the treatment of malaria, rheumatoid arthritis and systemic lupus erythematosus.³⁶ Based on in vitro activity and data from uncontrolled studies and small, randomized trials they were investigated as drug repurposing candidates for COVID-19 early in the pandemic.

A total of 1561 patients were randomized to receive hydroxychloroquine and 3155 to receive usual care with a primary outcome of 28-day mortality.²⁶ There was no significant difference in 28-day mortality between hydroxychloroquine recipients and the SoC group. Death occurred within 28 days in 421 patients (27.0%) in the hydroxychloroquine group and in 790 (25.0%) in the SoC group (RR, 1.09; 95% CI, 0.97 to 1.23; P = 0.15). Patients who were not receiving mechanical ventilation and were randomized to the hydroxychloroquine group had a higher risk of progressing to invasive mechanical ventilation or death compared to the SoC group (30.7% vs 26.9%; risk ratio, 1.14; 95% CI, 1.03 to 1.27). A small numerical excess of cardiac deaths (0.4 percentage points) but no difference in the incidence of new major cardiac arrhythmia was reported in patients allocated to hydroxychloroquine treatment.²⁶

Baricitinib

Baricitinib, a Janus kinase inhibitor, is an immunomodulatory drug used to treat rheumatoid arthritis and atopic eczema.³⁶ Patients were randomly allocated (1:1) to either usual SoC alone or SoC plus baricitinib 4 mg once daily by mouth for 10 days or until discharge if sooner.²⁰ The primary outcome was 28-day mortality as assessed in the intention-to-treat population. A total of 8156 patients were randomly allocated to receive usual care plus baricitinib versus usual care alone. At randomization, 95% of patients were receiving corticosteroids, 23% receiving tocilizumab and 20% remdesivir. Overall, 513 (12%) of 4148 patients allocated to baricitinib versus 546 (14%) of 4008 patients allocated to usual care died within 28 days, representing a reduction of 13% (age-adjusted RR 0.87; 95% CI 0.77 to 0.98; P = 0.026). No significant excess in death or infection due to non-COVID-19 causes or excess of thrombosis was observed in the baricitinib group. The benefit of baricitinib was consistent regardless of other COVID-19 treatments that patients were receiving (corticosteroids, tocilizumab or remdesivir).²⁰

Remdesivir is a nucleotide prodrug of an adenosine analog that binds to viral RNA-dependent RNA polymerase, prematurely terminating RNA transcription and inhibiting viral replication.³⁶ It was initially developed as a treatment for Ebola virus, although it did not receive regulatory approval.

Other treatments under investigation

Other drugs currently investigated for drug repurposing in the RECOVERY trial⁸ include empagliflozin (a drug used to treat type 2 diabetes)³⁶ and high-dose versus standard dose corticosteroids.

WHO Solidarity PLUS trial

The WHO Solidarity PLUS trial is an international platform trial coordinated by the WHO.⁹ To facilitate participation from hospitals with a high burden of COVID-19, particularly in low and middle income countries, the protocol design was rigorous but pragmatic. The goal of the Global Health Network community is to support COVID-19 research across different clinical contexts and research settings to ensure equity in terms of research engagement and clinical applicability of research findings globally. Solidarity has randomized over 14 000 hospitalized patients in 52 countries to date. Key factors that influence selection of drug repurposing candidates include safety and tolerability, ease of administration, availability, cost and scalability of manufacturing. The WHO Solidarity PLUS trial has investigated remdesivir, hydroxychloroquine, lopinavir/ritonavir and interferon (IFN-β1a).⁹

Hospitalized patients with COVID-19 were randomized equally between one of the locally available trial drug regimens and an open control arm (up to five options, four active and the local SoC).⁹ In the remdesivir arm patients received 200 mg IV on day 0 and 100 mg on days 1–9. Participants were randomly allocated to receive whichever of the four study drugs (lopinavir, hydroxychloroquine, IFN-β1a or remdesivir) locally available at that time or no study drug (control). Allocation was balanced and no placebos were administered.

Patients allocated to a lopinavir arm received 400 mg PO BD for 14 days. Patients allocated to an IFN-β1a arm received a dose of 44 μg SC for three doses subcutaneously on day 0, 3, 6 or IFN-β1a; 10 μg IV daily for 6 days.

The primary analysis was based on intention-to-treat and examined in-hospital mortality in four pairwise comparisons of each trial drug and a control arm (drug available but patient assigned to the same SoC without that drug intervention). Relative risk of death was calculated and stratified according to age and mechanical ventilation status at trial enrolment. The primary endpoint was in-hospital mortality, subdivided according to disease severity. Secondary endpoints were progression to ventilation if not already ventilated and time-to-discharge from hospital.

Lopinavir, hydroxychloroquine and IFN-β1a

At interim analysis involving 11 330 hospitalized patients, 2750 were assigned to receive remdesivir, 954 to hydroxychloroquine, 1411 to lopinavir (without IFN), 2063 to IFN (including 651 to IFN plus lopinavir) and 4088 to no trial drug.⁹ Death occurred in 104 of 947 patients receiving hydroxychloroquine versus 84 of 906 in the control arm (RR, 1.19; 95% CI, 0.89 to 1.59; P = 0.23), in 148 of 1399 patients receiving lopinavir versus 146 of 1372 in the control arm (RR, 1.00; 95% CI, 0.79 to 1.25; P = 0.97), in 243 of 2050 patients receiving IFN versus 216 of 2050 in the control arm (RR, 1.16; 95% CI, 0.96 to 1.39; P = 0.11) and in 301 of 2743 patients receiving remdesivir versus 303 of 2708 in the control arm (RR, 0.95; 95% CI, 0.81 to 1.11; P = 0.50).⁹ No drug definitely reduced mortality, overall or in any subgroup, or reduced initiation of ventilation or hospitalization duration. Lopinavir, hydroxychloroquine and IFN-β1a were subsequently discontinued but the remdesivir arm continued to recruit.

Remdesivir

A further analysis of 8275 patients were randomly allocated (1:1) either to remdesivir (10 daily infusions, unless discharged earlier) or to a control arm (allocated no study drug although remdesivir was locally available).³⁷ Patients were randomly allocated to daily remdesivir infusions (vs open-label control) during a 10-day treatment period in patients allocated to daily remdesivir infusions compared to open-label control. Overall, in the remdesivir arm, 602 (14.5%) of 4146 died versus 643 (15.6%) of 4129 in the control arm (mortality RR 0.91 [95% CI 0.82 to 1.02], P = 0.12). In patients ventilated at time of commencing study drug, 151 (42.1%) of 359 assigned to remdesivir died versus 134 (38.6%) of 347 assigned to control (RR 1.13 [0.89 to 1·42], P = 0.32). In patients receiving oxygen but not ventilation, 14.6% assigned to remdesivir died versus 16.3% assigned to control (RR 0.87 [0.76 to 0.99], P = 0.03). In 1730 patients not on oxygen at time of study commencement, 2.9% assigned to remdesivir died versus 3·8% assigned to control (RR 0.76 [0.46 to 1·28], P = 0.30). In a combined analysis of all those not initially ventilated, 11.9% assigned to remdesivir died versus 13.5% assigned to control (RR 0.86 [0.76 to 0.98], P = 0.02) and 14.1% versus 15.7% progressed to ventilation (RR 0.88 [0.77 to 1.00], P = 0.04).³⁷ The non-prespecified composite outcome of death or progression to ventilation occurred in 19.6% assigned to remdesivir versus 22.5% assigned to control (RR 0.84 [0.75 to 0.93], P = 0.001). Overall, remdesivir did not have a significant effect on patients with severe COVID-19 already on ventilation.³⁷ A small effect against death or progression to ventilation (or both) was seen in hospitalized patients not requiring ventilation.

Other treatments under investigation

Solidarity PLUS is currently investigating the antimalarial drug artesunate, the tyrosine kinase inhibitor imatinib and the immunomodulatory drug infliximab in the next stage of the trial.⁹

REMAP-CAP trial

REMAP-CAP is an international, multifactorial, adaptive platform trial for community acquired pneumonia.¹⁰ Interleukin-6 (IL-6) receptor antagonists tocilizumab and sarilumab were evaluated for efficacy in critically ill patients with COVID-19.¹⁰ Patients were randomly assigned to receive sarilumab (400 mg), tocilizumab (8 mg per kilogram of body weight) or SoC (control), within 24 h after starting organ support in the intensive care unit (ICU). The primary outcome measure was cardiovascular and respiratory organ support-free days, on an ordinal scale combining in-hospital death (assigned a value of –1) and days free of organ support to day 21. A Bayesian statistical model with predefined criteria for efficacy, equivalence, superiority or futility was applied. An odds ratio greater than 1 represented more organ support-free days, improved survival or both. A total of 353 patients were assigned to tocilizumab, 48 to sarilumab and 402 to control. The median number of organ support-free days was 10 (interquartile range, –1 to 16) in the tocilizumab cohort, 11 (interquartile range, 0 to 16) in the sarilumab cohort and 0 (interquartile range, –1 to 15) in the control cohort. In the tocilizumab cohort, the median adjusted cumulative odds ratio was 1.64 (95% credible interval [CrI], 1.25 to 2.14) and 1.76 (95% CrI, 1.17 to 2.91) in the sarilumab cohort compared with control. The posterior probabilities of superiority to control for tocilizumab were >99.9% and for sarilumab was >99.5%.¹⁰ Improved 90-day survival was seen in the pooled IL-6 receptor antagonist groups, yielding a hazard ratio (HR) for the comparison of the intervention with the control group of 1.61 (95% CrI, 1.25 to 2.08) and a posterior probability of superiority of >99.9%.¹⁰

Based on these results, tocilizumab and sarilumab were found to be equally effective for patients with severe COVID-19 compared to ‘no immune modulation’. As a result, the ‘no immune modulation’ arm was closed, and the COVID-19 Immune Modulation Domain arm continued with the aim of comparing other immune modulating drugs (sarilumab, tocilizumab, IFN-β1a and anakinra). A further 900 patients were recruited to the ‘immune modulation domain’. Following the most recent adaptive analysis by Data Monitoring Safety Board (DSMB) communicated to the REMAP-CAP International Trial Steering Committee, tocilizumab and sarilumab were found to have reached the pre-specified statistical trigger for equivalence and further recruitment to this domain halted.¹⁰ A full analysis is due to be published.

The WHO REACT investigators also published a meta-analysis highlighting that IL-6 receptor antagonists have only been effective to date when given to patients with severe COVID-19 treated with concomitant glucocorticoids and thus should be reserved for patients already commenced on glucocorticoid therapy.³⁸

Other treatments under investigation

REMAP-CAP investigators are currently testing infliximab, imatinib (both target tumor necrosis factor alpha [TNF-α]) and namilumab. Infliximab is a chimeric monoclonal antibody that blocks TNF-α and is licensed for the treatment of autoimmune conditions, including Crohn’s disease, ulcerative colitis, ankylosing spondylitis and rheumatoid arthritis.²¹ Namilumab is a human monoclonal antibody that targets granulocyte macrophage-colony stimulating factor 2 and is being researched as a treatment for rheumatoid arthritis.³⁹

Adaptive COVID-19 Treatment trials 1 and 2

The Adaptive Covid-19 Treatment trials 1 and 2 (ACTT-1 and ACTT-2) is an adaptive platform trial that investigated remdesivir¹¹ and baricitinib¹² in COVID-19 in the USA.

In ACTT-1 patients hospitalized with COVID-19 and were randomly assigned to receive either remdesivir (200 mg loading dose on day 1, followed by 100 mg daily for up to nine additional days) or placebo for up to 10 days.¹¹ The primary outcome measure was time to recovery, which was defined by either discharge from the hospital or hospitalization for infection-control measures. A total of 1062 patients underwent randomization (with 541 assigned to receive remdesivir and 521 to receive placebo). Patients in the remdesivir arm had a median recovery time of 10 days (95% CI, 9 to 11), compared with 15 days (95% CI, 13 to 18) in the placebo arm (RR for recovery, 1.29; 95% CI, 1.12 to 1.49; P < 0.001, by a log-rank test). The Kaplan–Meier estimates of mortality at day 29 were 11.4% in the remdesivir arm and 15.2% in the placebo arm (HR, 0.73; 95% CI, 0.52 to 1.03).¹¹

In ACTT-2 patients were randomized to receive remdesivir (≤10 days) and either baricitinib (≤14 days) or placebo (control).¹² The primary outcome measure was time to recovery. A total of 1033 patients underwent randomization (515 patients assigned to combination treatment and 518 patients assigned to control). The median time to recovery in the baricitinib arm was 7 days (95% CI, 6 to 8), as compared with 8 days in the control arm (95% CI, 7 to 9) (RR for recovery, 1.16; 95% CI, 1.01 to 1.32; P = 0.03). In patients receiving remdesivir and baricitinib, there was a 30% higher odds of improvement in clinical status at day 15 (odds ratio, 1.3; 95% CI, 1.0 to 1.6). Time to recovery in patients receiving high-flow oxygen or non-invasive ventilation at enrolment was 10 days with combination treatment and 18 days with control (RR for recovery, 1.51; 95% CI, 1.10 to 2.08).¹² The 28-day mortality was 5.1% in the combination treatment group compared to 7.8% in the placebo group (HR for death, 0.65; 95% CI, 0.39 to 1.09).

The National Institute of Health subsequently recommended baricitinib or tocilizumab in combination with dexamethasone for the treatment of hospitalized patients with severe COVID-19.⁴⁰

DISCOVERY trial

DISCOVERY was established as a phase III, adaptive, controlled, multicenter clinical trial researching therapeutic treatments in patients hospitalized with COVID-19 requiring oxygen therapy in France.¹³ Patients were intended to be randomized between five arms: (1) a control group managed with SoC and four therapeutic arms of SoC with re-purposed antiviral agents: (2) remdesivir + SoC, (3) lopinavir/ritonavir + SoC, (4) lopinavir/ritonavir associated with IFN-β1a + SoC and (5) hydroxychloroquine + SoC. The primary outcome measure was clinical status at day 15 on the 7-point ordinal scale of the WHO Master Protocol with a target sample size of 3100 patients (620 patients per arm). In April 2020, DISCOVERY was incorporated as an add-on trial of the Solidarity consortium of trials conducted by the WHO in Europe and worldwide.

Hospitalized patients were randomly assigned (1:1:1:1:1) to receive SoC alone or in combination with remdesivir, lopinavir–ritonavir, lopinavir–ritonavir and IFN-β1a or hydroxychloroquine. Patients were stratified on severity of disease at inclusion.

A total of 583 participants were randomized to lopinavir/ritonavir (n = 145), lopinavir/ritonavir–IFNβ-1a (n = 145), hydroxychloroquine (n = 145) and control (n = 148).²⁷ No improvement was seen in day 15 clinical status with the investigational treatments: lopinavir/ritonavir versus control, adjusted odds ratio (aOR) 0.83 (95% CI 0.55 to 1.26, P = 0.39), lopinavir/ritonavir–IFN-β1a versus control, aOR 0.69 (95% CI 0.45 to 1.04, P = 0.08), and hydroxychloroquine versus control, aOR 0.93 (95% CI 0.62 to 1.41, P = 0.75). The investigational treatments did not have a significant effect on SARS-CoV-2 clearance.

The remdesivir arm was reported separately.²⁸ Remdesivir was administered as 200 mg intravenous infusion on day 1, followed by 100 mg intravenous once daily for up to 9 days, with a total duration of 10 days.²⁸ The primary outcome was clinical status at day 15 measured by the WHO ordinal scale, in the intention-to-treat population. A total of 857 participants were enrolled and randomly assigned to remdesivir plus SoC (n = 429) or SoC only (n = 428). Fifteen participants were excluded from analysis in the remdesivir group, and ten in the control group. There was no significant difference in WHO Ordinal status between treatment groups at day 15 (odds ratio 0.98 [95% CI 0.77 to 1.25]; P = 0.85), time to hospital discharge, 28-day all-cause mortality or SARS-CoV-2 viral kinetics. Remdesivir was therefore not found to be of clinical benefit in hospitalized patients on oxygen or with symptom duration of >7 days.²⁸ In an exploratory subgroup analysis, the hazard for the composite endpoint of death, extracorporeal membrane oxygenation (ECMO) or new mechanical ventilation was lower in the remdesivir group versus SoC alone (HR 0.66 [95% CI 0.47 to 0.91]; P = 0.010) in patients not on mechanical ventilation or ECMO at randomization.

ACTIV-6 trial

The ACTIV-6 trial is a randomized, placebo-controlled phase III clinical trial aimed at testing repurposed drugs for the management of mild-to-moderate COVID-19 in the community in the USA.¹⁴ To date, two experimental arms have reported on ivermectin 300–400 μg/kg for three consecutive days and fluticasone 200 μg inhaler once daily for 14 days. Primary outcome measures include rate of hospitalization, rate of death and number of patient reported symptoms. Secondary outcome measures include changes in the COVID-19 Clinical Progression Scale, patient reported Quality of Life and a composite score of care visits, emergency room visits and hospitalizations. A third trial arm of fluvoxamine 50 mg twice a day for 10 days has not yet been reported.

Ivermectin

Ivermectin is an antiparasitic drug used for the treatment of onchocerciasis and strongyloidiasis.³⁶ The efficacy of ivermectin 400 μg/kg daily for 3 days compared with placebo for the treatment of early mild-to-moderate COVID-19 was evaluated. Non-hospitalized adults aged ≥ 30 years with confirmed COVID-19 experiencing ≥2 symptoms of acute infection for ≤ 7 days were randomized to receive ivermectin 400 μg/kg daily for 3 days or placebo. Primary outcome measure was time to sustained recovery, defined as at least three consecutive days without symptoms. Secondary outcomes included a composite of hospitalization or death at day 28. A total of 1591 participants were randomized to receive ivermectin 400 μg/kg (n = 817) or placebo (n = 774). Forty-seven percent of participants reported receiving at least two doses of SARS-CoV-2 vaccination. Recruitment covered the delta and omicron variant time periods. A hazard ratio (HR) of 1.07 was reported for improvement in time to recovery (95% credible interval [CrI], 0.96-1.17; posterior P value [HR >1] = .91). The median time to recovery in the ivermectin group was 12 days (IQR, 11-13) versus 13 days (IQR, 12-14) in the placebo group. Ten hospitalizations/deaths were reported in the ivermectin group versus 9 events in the placebo group (1.2% vs 1.2%; HR, 1.1 [95% CrI, 0.4-2.6]). Treatment with ivermectin therefore was not found to significantly improve time to recovery in outpatients with mild to moderate COVID-19, compared to placebo.²⁹

Fluticasone

Fluticasone is an inhaled steroid that has a wide variety of anti-inflammatory effects, used primarily to treat chronic obstructive pulmonary disease and asthma.³⁶ Non-hospitalized adults aged ≥30 years experiencing ≥2 symptoms of acute infection for ≤ 7 days were randomized to inhaled fluticasone furoate 200 μg once daily for 14 days or placebo.³⁰ Primary outcome measure was time to sustained recovery, defined as the third of three consecutive symptom free days. Secondary outcomes included hospitalization or death at day 28.

A total of 656 were randomized to inhaled fluticasone, while 621 were randomized to placebo. No improvement in time to recovery was observed with fluticasone compared with placebo (HR 1.01, 95% CrI 0.91 to 1.12; posterior probability for benefit [HR > 1] = 0.56). A total of 24 participants (3.7%) in the fluticasone arm required urgent care or were hospitalized compared with 13 (2.1%) in the placebo arm (HR 1.9, 95% CrI 0.8 to 3.5; posterior probability for benefit [HR < 1] = 0.03).³⁰

PRINCIPLE trial

Treatments effective in the management of COVID-19

The PRINCIPLE trial is a UK-wide clinical study researching repurposed drugs for the management of mild-to-moderate COVID-19 in the community.¹⁵ The PRINCIPLE trial has reported on the evaluation of four treatments to date—budesonide, colchicine, azithromycin and doxycycline. Patients aged 65 years or older or 50 years or older with co-morbidities, unwell for up to 14 days with suspected COVID-19, without hospitalization, were eligible for trial participation. Time to first self-reported recovery, and hospitalization/death related to COVID-19 at 28 days from randomization were co-primary endpoints and analyzed using Bayesian models.¹⁵

Budesonide

Patients with mild COVID-19 managed in the community were randomized to SoC or SoC plus inhaled budesonide (800 μg twice daily for 14 days).²¹ A total of 1073 participants were randomly assigned to budesonide and 1988 to usual care alone. At an interim analysis with 787 randomized to the budesonide group and 1028 randomized to the usual care group, a benefit in time to first self-reported recovery was seen in the budesonide group versus the usual care group (11.8 days [95% BCI 10.0 to 14.1] vs 14·7 days [12.3 to 18.0]; HR 1.21 [95% BCI 1.08 to 1.36]), with a probability of superiority greater than 0.999. The estimated rate of hospital admission or death was 6.8% (95% BCI 4.1 to 10.2) in the budesonide group versus 8.8% (5.5 to 12.7) in the usual care group (estimated absolute difference 2.0% [95% BCI –0.2 to 4.5]; odds ratio 0.75 [95% BCI 0.55 to 1.03]), with a probability of superiority 0.963, below the prespecified superiority threshold of 0.975.²¹ Serious adverse events (hospital admissions unrelated to COVID-19) were observed in two participants in the budesonide groups and four participants in the usual care group.

Doxycycline

Doxycycline is a broad-spectrum tetracycline-class antibiotic used to treat bacterial and certain parasitic infections.²¹ In this arm of the study, participants were randomly assigned using response adaptive randomization to usual care only, usual care plus oral doxycycline (200 mg on day 1, then 100 mg once daily for the following 6 days), or usual care plus other interventions.³¹ Co-primary endpoints were time to first self-reported recovery, and hospitalization or death related to COVID-19 at 28 days from randomization. A total of 780 (31.1%) were randomized to the usual care plus doxycycline group, 948 in the usual care only group (37.8%) and 780 (31.1%) in the usual care plus other interventions group. In the primary analysis model, there was limited evidence of a difference in median time to first self-reported recovery between the usual care plus doxycycline group and the usual care only group (9.6 [95% BCI 8.3 to 11.0] days vs 10.1 [8.7 to 11.7] days, HR 1.04 [95% BCI 0.93 to 1.17]). The estimated benefit in median time to first self-reported recovery was 0.5 days [95% BCI –0.99 to 2.04]. The probability of a clinically meaningful benefit (defined as ≥1.5 days) was 0.10. Hospitalization or death related to COVID-19 occurred in 41 (crude percentage 5.3%) participants in the usual care plus doxycycline group compared to 43 (4.5%) in the usual care only group (estimated absolute percentage difference − 0.5% [95% BCI –2·6 to 1.4]). Doxycyline was therefore not found to be associated with a clinically meaningful reduction in recovery time, hospital admissions or COVID-19 related death.³¹

Azithromycin

Azithromycin is an antimicrobial with anti-inflammatory and antiviral properties.²¹ The PRINCIPLE study assigned 540 participants to usual care plus azithromycin 500 mg daily for 3 days, 850 participants to usual care plus other interventions and 875 participants to usual care alone.³² Co-primary endpoints included time to first self-reported recovery and hospital admission or death related to COVID-19 at 28 days. A total of 402 (80%) of 500 participants in the azithromycin plus usual care group and 631 (77%) of 823 participants in the usual care alone group reported having recovered within 28 days. No clinically meaningful benefit in the azithromycin plus usual care group in time to first reported recovery versus usual care alone was observed (HR 1.08, 95% BCI 0.95 to 1.23), equating to an estimated benefit in median time to first recovery of 0.94 days (95% BCI –0.56 to 2.43). The probability of a clinically meaningful benefit of at least 1.5 days in time to recovery was 0.23. A total of 16 (3%) of 500 participants in the azithromycin plus usual care group and 28 (3%) of 823 participants in the usual care alone group were hospitalized (absolute benefit in percentage 0.3%, 95% BCI –1.7 to 2.2). On this basis, it was concluded that azithromycin was not an effective treatment for reducing the time to recovery or risk of hospital admission for COVID-19 patients in the community.³²

Colchicine

Colchicine is an anti-inflammatory used in the treatment of gout and other inflammatory conditions including Familial Mediterranean fever and Bechet’s disease.³⁶

Participants were randomized to usual care plus colchicine (500 μg daily for 14 days), usual care or usual care plus other interventions.³³ Co-primary endpoints included time to first self-reported recovery and hospitalization/death related to COVID-19, within 28 days. The hypothesis for the time to recovery endpoint was evaluated first, and if superiority was declared on time to recovery, the hypothesis for the second co-primary endpoint of hospitalization/death was then evaluated. A prespecified clinically meaningful benefit in time to first reported recovery as a HR of 1.2 or larger was used to determine futility (equating to approximately 1.5 days benefit in the colchicine arm, assuming 9 days recovery in the usual care arm). The primary analysis model included 2755 SARS-CoV-2 positive participants, randomized to colchicine (n = 156), usual care (n = 1145) and other treatments (n = 1454). In the colchicine group, time to first self-reported recovery was similar compared with usual care with an estimated HR of 0.919 [95% CrI 0.72 to 1.16] and an estimated increase of 1.14 days [–1.86 to 5.21] in median time to self-reported recovery for colchicine versus usual care. The probability of meaningful benefit in time to recovery was determined to be low at 1.8%. The rate of COVID-19 related hospitalizations/deaths was similar in the colchicine group versus usual care, with an estimated odds ratio of 0.76 [0.28 to 1.89] and an estimated difference of –0.4% [–2.7% to 2.4]. Only one serious adverse event was observed in the colchicine group and one in the usual care group. Colchicine was not found to improve time to recovery in people at higher risk of complications with COVID-19 in the community.³³

Other treatments

Current treatments under investigation in PRINCIPLE include favipiravir (an antiviral drug) and ivermectin (an antiparasitic drug).¹⁵

Living systematic reviews and guidelines

BMJ Living Systematic Review

The BMJ Living Systematic Review searched the multi-lingual WHO COVID-19 database, and six additional Chinese databases up to February 20, 2021.¹⁶ Randomized controlled trials of patients with suspected, probable or confirmed COVID-19 who were randomized to a study drug intervention or SoC or placebo were screened and selected by paired reviewers. A Bayesian network meta-analysis was performed and the risk of study bias was assessed using a modification of the Cochrane risk of bias 2.0 tool. Interventions were evaluated using the GRADE approach (from the most to the least beneficial or harmful).

A total of 463 trials enrolling 166 581 patients were included. Three drugs reduced mortality in patients with mostly severe disease with at least moderate certainty compared with SoC. These included systemic corticosteroids (risk difference 23 fewer per 1000 patients, 95% CrI 40 fewer to 7 fewer, moderate certainty), IL-6 receptor antagonists when given with corticosteroids (23 fewer per 1000, 36 fewer to 7 fewer, moderate certainty) and Janus kinase inhibitors (44 fewer per 1000, 64 fewer to 20 fewer, high certainty).¹⁶

Two drugs probably reduce hospital admission in patients with non-severe disease when compared to SoC. These included molnupiravir (19 fewer per 1000, 29 fewer to 5 fewer, moderate certainty) and nirmatrelvir/ritonavir (36 fewer per 1000, 41 fewer to 26 fewer, moderate certainty). The systematic review found that at present, remdesivir may reduce hospital admission (29 fewer per 1000, 40 fewer to 6 fewer, low certainty).¹⁶

Hydroxychloroquine, lopinavir-ritonavir, azithromycin and IFN-β did not appear to reduce the risk of death or have an effect on any other important clinical outcomes. The effects of ivermectin for all clinical outcomes were rated as very low certainty, including mortality. In patients with non-severe COVID-19, colchicine may reduce mortality (78 fewer per 1000, 110 fewer to 9 fewer, low certainty) and progression to mechanical ventilation (57 fewer per 1000, 90 fewer to 3 more, low certainty).

Interventions that did not appear to have a significant benefit on clinical outcomes included aspirin, azithromycin, angiotensin-converting enzyme inhibitors, colchicine, hydroxychloroquine, inhaled corticosteroids, intranasal corticosteroids, IFN-β, vitamin C, ivermectin, lopinavir-ritonavir and umifenovir. Hydroxychloroquine may increase the risk of adverse effects, mechanical ventilation and duration of hospital stay.¹⁶

WHO living guideline

The WHO published a living guideline in response to emerging evidence from thousands of randomized controlled trials investigating repurposed and novel drug treatments for COVID-19 to inform policy and clinical practice worldwide.¹⁷ An international guideline development panel was formed consisting of content experts, methodologists, clinicians, patients and an ethicist. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach was used to assess strength of evidence and The Magic Evidence Ecosystem Foundation (MAGIC) provided methodological support for the guideline. The WHO living guideline is now in its tenth version, containing 19 recommendations.

In September 2020, the panel issued a strong recommendation for systemic corticosteroids in patients with severe and critical COVID-19. In July 2021, following results from the RECOVERY and REMAP-CAP studies, the WHO guideline provided a strong recommendation for the use of IL-6 receptor blockers (tocilizumab or sarilumab) in patients with severe or critical COVID-19.¹⁷

In January 2022, a strong recommendation was published for the use of baricitinib in combination with corticosteroids in patients with severe COVID-19, as an alternative to IL-6 receptor blockers.¹⁷

In April 2022, a conditional recommendation for the use of remdesivir in patients with non-severe COVID-19 at the highest risk of hospitalization was made. In July 2022, a strong recommendation was made against the use of colchicine in patients with non-severe COVID-19.

Strong recommendations were released against the use of hydroxychloroquine and lopinavir-ritonavir in patients with COVID-19, regardless of disease severity. A recommendation was also issued not to use ivermectin in patients with COVID-19 except in the context of a clinical trial.¹⁷

Conclusion

As the race to develop effective and affordable therapeutics for COVID-19 continues, a number of important lessons for reforming global health research and accelerating therapeutic drug discovery can be learnt from the pandemic so far. The progress achieved by the RECOVERY and WHO Solidarity PLUS trials demonstrates that large global trials are possible, even under pandemic conditions, and offer the promise of quickly and reliably answering critical public health questions concerning therapeutics. Open collaboration, data sharing, harmonized regulatory frameworks and training materials are vital to ensure these trials can be delivered robustly and effectively.

Drug repurposing during the pandemic has also been a reminder that well-designed, robust clinical trials are needed to derive definitive efficacy and safety data when repurposing drugs for new conditions. Driven by the desperation brought on by the pandemic, there was a rush in some quarters to adopt into clinical practice a number of potential drug repurposing candidates before clinical trial results were available. Many of these drugs were later found to be futile. Information about when an intervention is not useful, or may be harmful, is just as valuable as interventions that improve outcomes. Drug repurposing offers the potential of discovering affordable therapeutics with global accessibility compared to the high cost that is often associated with first-in-class novel therapeutics. Vital research must continue to identify effective, safe and accessible treatments for COVID-19 across the disease spectrum, including disease prevention.

Funding

H.M.S. is supported by the Wellcome Trust Institutional Strategic Support Fund (204809/Z/16/Z) awarded to St. George’s University of London.

Conflict of interest

S.K. has participated in the World Health Organization (WHO) Guideline Development Group (GDG) for the treatment of COVID-19. The views expressed here are personal and do not necessarily represent those of the GDG.

Data availability

No new data were generated or analyzed in support of this review.