Last updated: November 22, 2021
On this page:
- Drugs for Which There Is Evidence of Harm
- Other Drugs
- Table of Drugs Under Early Investigation for COVID-19
A variety of therapeutics have been studied for treatment of COVID-19 for which there are varying degrees of evidence. For most, there are currently insufficient clinical data to recommend either for or against their use. For some, there is clear evidence of harm and no evidence of benefit. This overview is not a comprehensive summary, but includes therapeutics with strong biological plausibility that are available in the United States and are or will be studied by clinical trial.
The following section addresses agents for which there is evidence of harm and no clear evidence of benefit for the treatment of COVID-19.
Hydroxychloroquine (HCQ) and chloroquine (CQ) are antimalarial agents that are also used to treat certain autoimmune disorders. HCQ has a lower incidence than CQ of adverse events with chronic use (Ben-Zvi, January 2011). Both drugs have immunomodulatory effects on various cytokines, including IL-1 and IL-6 (Ben-Zvi, January 2011), and prior to the COVID-19 pandemic were known to have in vitro effects against various viruses, including SARS (Keyaerts, October 2004; Vincent, August 2005; Savarino, September 2011). Early in the course of the pandemic both agents were also found to have in vitro activity against SARS-CoV-2 (Wang, February 2020; Liu, March 2020; Yao, March 2020). Given this potential biological plausibility, numerous clinical studies were initiated to examine efficacy for COVID-19 treatment.
In March 2020, the FDA issued an emergency use authorization (EUA) for the use of HCQ/CQ in hospitalized patients with COVID-19. In April 2020, the FDA released a statement cautioning against use of HCQ or CQ for COVID-19 outside of the hospital setting or a clinical trial due to risk of heart rhythm problems. Based on subsequent clinical trial data showing a lack of efficacy and concern for safety signals, the FDA revoked the EUA in June 2020 (see Safety section for data on mortality).
Since the beginning of the pandemic there has been extensive research related to HCQ, including more than five randomized controlled trials and several large observational cohort studies. Given the availability of higher-level data, in the following review we have focused on these studies. In non-hospitalized patients with asymptomatic or mild COVID-19, hydroxychloroquine has not been shown to reduce SARS-CoV-2 RNA viral load or disease progression in several small randomized trials (Skipper, October 2020; Mitja, July 2020). Similarly, a randomized double-blind, placebo-controlled trial of 821 participants in the United States found no impact on development of confirmed or probable COVID-19 when hydroxychloroquine was used as post-exposure prophylaxis within four days of a high-risk exposure (Boulware, August 2020).
- IDSA guidelines recommend against the use of hydroxychloroquinefor hospitalized patients with COVID-19.
- IDSA guidelines recommend against the use of hydroxychloroquineplus azithromycin for hospitalized patients with COVID-19.
- NIH guidelines recommend against the use of hydroxychloroquine and/or azithromycin for the treatment of COVID-19 in hospitalized and non-hospitalized patients.
In summary: While observational studies of the use of hydroxychloroquine (HCQ) and chloroquine (CQ) in patients with COVID-19 have had mixed results, several randomized controlled trials conducted in hospitalized patients with COVID-19 have not shown clinical benefit, nor have randomized controlled trials focused on post-exposure prophylaxis. In addition, some data suggest the use of HCQ may be associated with significant cardiac adverse events in patients with COVID-19. The use of HCQ or CQ is not recommended in patients with COVID-19.
Hydroxychloroquine in Hospitalized Patients with COVID-19 (The RECOVERY Collaborative Group, October 2020).
Overall, in this randomized, open-label trial, among patients hospitalized with COVID-19, those who received hydroxychloroquine did not have a lower incidence of death at 28 days compared with those who received standard of care.
- Randomized, controlled, open-label platform trial comparing a range of possible treatments with usual care in patients hospitalized with COVID-19 in the U.K.
- 1,561 patients randomly assigned to receive hydroxychloroquine and 3,155 to receive standard of care.
- The mean age was 65.4 (SD ±15.3) years.
- A history of diabetes was present in 27% of patients, heart disease in 26% and chronic lung disease in 22%, with 57% having at least one major coexisting illness.
- At randomization, 17% were receiving invasive mechanical ventilation including extracorporeal membrane oxygenation, 60% were receiving oxygen only (with or without noninvasive ventilation) and 24% were receiving neither.
- The median duration of treatment was 6 days (interquartile range, 3-10 days).
- 28-day mortality.
- Death at 28 days occurred in 421 of 1,561 patients (27%) in the hydroxychloroquine group and in 790 of 3,155 patients (25%) in the usual-care group (rate ratio, 1.09; 95% confidence interval [CI], 0.97 to 1.23; p=0.15).
- Patients in the hydroxychloroquine group had a longer duration of hospitalization than those in the usual-care group (median, 16 days vs. 13 days) and a lower probability of discharge alive within 28 days (59.6% vs. 62.9%; rate ratio, 0.90; 95% CI, 0.83 to 0.98).
- Patients in the hydroxychloroquine group had a greater risk of death from cardiac causes (mean [±SE] excess, 0.4±0.2 percentage points) and from non–SARS-CoV-2 infection (mean excess, 0.4±0.2 percentage points).
- Single-country study; patient population and results may not be generalizable to other countries.
- Physiologic, electrocardiographic, laboratory or virologic measurements were not collected.
Retrospective Cohort Study of HCQ +/-Azithromycin in Patients Hospitalized with COVID-19 (Arshad, July 2020).
Overall, in this retrospective cohort study, treatment with HCQ or HCQ + azithromycin was associated with reduction in COVID-19-associated mortality; however, the lack of balance between the groups and the lack of clarity around how treatment decisions were made limits the generalizability of these results.
- 2,541 hospitalized COVID-19 patients in Michigan.
- Patients received HCQ + azithromycin (783 patients), HCQ alone (1202 patients), azithromycin alone (147 patients) or neither drug (409 patients).
- HCQ + azithromycin were given only to select patients with severe COVID-19 and minimal cardiac risk factors.
- Corticosteroid use varied by group: HCQ, 78.9%; HCQ + azithromycin, 74.3%; azithromycin, 38.8%; neither medication, 35.7%.
- Mechanical ventilation was required in 13% of HCQ alone arm, 29% among HCQ + azithromycin, and in 8% that did not receive either medication.
- In-hospital mortality.
- Overall in-hospital mortality was 18.1% (45% in ICU patients).
- Mortality by treatment category was as follows: HCQ + azithromycin, 157/783 (20.1% [95% CI: 17.3%–23.0%]), HCQ alone, 162/1202 (13.5% [95% CI: 11.6%–15.5%]), azithromycin alone, 33/147 (22.4% [95% CI: 16.0%–30.1%]) and neither drug, 108/409 (26.4% [95% CI: 22.2%–31.0%]).
- Compared to neither treatment, HCQ alone was associated with a 66% HR reduction (p < 0.001), while HCQ + azithromycin were associated with an HR reduction of 71% (p < 0.001).
- A matched propensity scoring of HCQ versus no HCQ (190 patients in each group) showed that HCQ compared to no HCQ was associated with decreased mortality by a mortality HR decrease of 51% (p<0.009).
- The retrospective nature of this study could have allowed for confounding.
- Determination of treatment decisions is not well described in the manuscript.
- Information on duration of symptoms prior to hospitalization is not available.
- There were substantial differences between the treatment arms with respect to corticosteroid use and their need for mechanical ventilation.
A Randomized Controlled Trial of HCQ in Outpatients With Early COVID-19 (Skipper, July 2020).
Overall, in this randomized controlled trial of outpatients with early COVID-19 disease, the use of HCQ was not associated with a change in symptom severity; adverse effects were common in the HCQ group compared to the placebo group.
- 491 symptomatic, non-hospitalized adults with four days or less of symptoms and laboratory-confirmed COVID-19, probable COVID-19 or high-risk exposure within the preceding 14 days.
- 423 patients contributed to the primary endpoint: 241 (57%) health care workers, 106 (25%) household contacts and 76 (18%) with other exposures.
- Patients were randomized to receive HCQ or placebo.
- Change in overall symptom severity over 14 days.
- Change in symptom severity over 14 days did not differ between the HCQ and placebo groups (difference in symptom severity: absolute, −0.27 points; P = 0.117).
- At 14 days, 24% (49/201) of participants receiving HCQ had ongoing symptoms compared with 30% (59/ 194) receiving placebo (P = 0.21).
- Medication adverse effects occurred in 43% (92/ 212) of participants receiving HCQ versus 22% (46/211) receiving placebo (P < 0.001).
- There was no difference in hospitalization (10 in the placebo group versus 4 in the HCQ group; P = 0.29).
- Participants were enrolled through internet-based surveys, which may have skewed the study population.
- The study population was heterogeneous.
- Only 58% of the sample were tested for SARS-CoV-2; it is possible not all the patients in the study had COVID-19.
Randomized Controlled Trial of HCQ in Patients with Mild to Moderate COVID-19 (Tang, May 2020).
Overall, in this randomized controlled trial that was halted early due to low enrollment, HCQ had no effect on the duration of viral detection, despite the use of high doses and a prolonged 2-3 duration.
- 150 hospitalized patients in China with laboratory-confirmed COVID-19.
- Illness severity at admission was primarily mild (15%) or moderate (84%); 1% of cases were severe.
- 75 patients were randomized to receive HCQ + standard of care, and 75 were randomized to receive standard of care. Study arm assignment was not blinded.
- The dose of HCQ used was 1,200 mg daily for 3 days, followed by 800 mg daily for 2-3 weeks.
- In comparison, the most used regimen in the United States is 800 mg on the first day followed by 400 mg daily for four days.
- The mean duration from symptom onset to randomization was 16.6 days.
- Viral clearance by 28 days.
- Secondary endpoint: rate of improvement in symptoms over 1 month (defined as defervescence, improved oxygen saturation and disappearance of respiratory symptoms).
- The probability of negative conversion by 28 days in the standard of care + HCQ group was 85.4% (95% CI: 73.8% to 93.8%), similar to that in the standard-of-care group (81.3%, CI: 71.2% to 89.6%).
- Adverse events were reported in 21/70 (30%) HCQ recipients versus 7/80 (9%) individuals not receiving HCQ.
- Blinding was not utilized during assignment to the study arms; selection bias is possible.
- The study was terminated early due to low enrollment and the impression of benefit (which may have led to underpowering).
- The time between symptom onset and randomization was 16.6 days; patients may have been too far along in the course of their illness to derive a beneficial anti-viral effect.
A Multicenter, Open-Label, Randomized, Placebo-Controlled Trial Evaluated Patients With Mild-to-moderate COVID-19 (Cavalcanti, July 2020).
Overall, in this open-label randomized clinical trial, in patients hospitalized with mild-to-moderate COVID-19, HCQ +/- azithromycin did not improve clinical status at 15 days compared to standard of care. Higher adverse events occurred in the patients who received HCQ +/- azithromycin.
- 504 hospitalized patients with confirmed mild-to-moderate COVID-19 were randomized to receive either standard of care, HCQ + standard of care or HCQ + azithromycin + standard of care.
- There were 173 patients in the control group, 159 patients in the HCQ group and 172 patients in the HCQ + azithromycin group.
- The median time from symptom onset to randomization was 7 days.
- Clinical status at 15 days using a seven-level ordinal scale.
- Compared to standard of care, the proportional odds of having a worse score on the ordinal scale at 15 days was not affected by either HCQ (OR 1.21; P=1.00) or HCQ + azithromycin (OR 0.99; P=1.00).
- Compared to standard of care, more adverse events were reported in patients that received HCQ + azithromycin (39.3%) or HCQ alone (33.7%).
- QTc interval prolongation and liver enzyme elevations were most common.
- The median time of symptom onset to randomization was 7 days, which limits the generalizability of these results.
- Blinding was not employed during randomization, which could have introduced several levels of bias.
- A good number of patients had already received HCQ and/or azithromycin prior to study entry, which could have affected the results.
RECOVERY, a Controlled, Open-Label Trial Comparing a Range of Possible Treatments in Patients Hospitalized with COVID-19 (Holby, July 2020).
Overall, in this randomized controlled open-label trial of patients hospitalized with COVID-19, HCQ was not associated with 28-day mortality reduction, but was associated with an increased length of hospital stay and increased risk of progressing to invasive mechanical ventilation or death.
- 1561 patients were randomized to receive HCQ; 3155 patients were randomized to standard of care.
- 28-day mortality.
- 28-day mortality did not differ in those patients that received HCQ vs. usual care (26.8% vs. 25.0% rate ratio 1.09; 95% CI 0.96 to 1.23).
- Patients allocated to HCQ were less likely to be discharged from hospital alive within 28 days (60.3% vs. 62.8%; rate ratio 0.92; 95% CI 0.85-0.99).
- Patients allocated to HCQ and not on invasive mechanical ventilation at baseline were more likely to reach the composite endpoint of invasive mechanical ventilation or death (29.8% vs. 26.5%; risk ratio 1.12; 95% CI 1.01-1.25).
- The mortality rate of patients with severe COVID-19 disease found in this study is higher than what has been generally found in this group in the United States.
- Open-label design; researchers and patients in the study knew who was receiving which treatment. This could have introduced bias into the results.
- Not all patients had proven SARS-CoV-2 infection via RT-PCR, but in post-hoc analysis of patients with a positive PCR result (90% of the sample), the results were similar.
Prospective Observational Study of HCQ in Hospitalized Patients With COVID-19 (Geleris, May 2020).
Overall, in this prospective observational study of patients hospitalized with COVID-19, HCQ was not associated with an increase or reduction in the composite endpoint of death or intubation.
- 1376 non-intubated patients hospitalized with COVID-19 at a single institution in New York City.
- 811 (58.9%) received HCQ; 45.8% were treated within 24 hours and 85.9% within 48 hours.
- Composite of intubation or death in a time-to-event analysis.
- There was no significant association between HCQ use and intubation or death (HR 1.04, 95% CI 0.82-1.32).
- This is an observational study; therefore, confounding is possible.
- The decision to administer HCQ was made by the treating physician; selection bias may have occurred.
- The study occurred at a single institution; therefore, results may not be generalizable.
Randomized Controlled Trial of HCQ in Patients With COVID-19 (Chen, April 2020).
Overall, in this randomized controlled trial (RCT) of patients with mild COVID-19, HCQ was associated with a shorter time to clinical improvement (defined by resolution of fever and cough) and radiographic improvement. However, at study entry many patients had already met the time to clinical recovery study endpoints; it is not clear how the analysis accounted for these patients.
- 62 hospitalized patients with non-severe COVID-19.
- 31 patients were randomized to receive 5 days of HCQ, and 31 to receive standard of care.
- Standard of care included antiviral agents, antibacterial agents and immunoglobulin with or without corticosteroids.
- Time to clinical recovery (defined as becoming afebrile and cough relief for 72 hours or more), clinical characteristics and changes in chest CT.
- Compared with the control group, body temperature recovery time was significantly shortened in the HCQ treatment group (2.2 ± 0.4 days vs 3.2 ± 1.3 days).
- Cough duration was shorter in the HCQ group compared to the control group (2.0 ± 0.2 days vs 3.1 ± 1.5 days).
- More patients had radiographic improvement with HCQ [25/31 (81%) vs. 17/31 (55%), p=0.05].
- No baseline characteristics of patients were provided.
- The endpoint focused on resolution of fever and cough; the clinical relevance of this endpoint is not clear.
- At entry into the study, in the HCQ arm nine patients had no fever and nine had no cough. In the control arm 14 patients had no fever, and 16 had no cough. It is not clear how these patients were accounted for in the analysis.
- No information about viral load was included.
Across the body of evidence from several RCTs, treatment with HCQ may increase the risk of experiencing adverse events (Tang, May 2020, Cavalcanti, July 2020). One RCT suggests increased risk of QT prolongation among patients treated with HCQ+AZ compared to those not receiving HCQ (RR: 8.50; 95% CI: 1.16-62.31) (Cavalcanti, July 2020). Early reports raised concern about the association of hydroxychloroquine with in-hospital mortality. A systematic review and meta-analysis (Fiolet, August 2020) of the effect of hydroxychloroquine with or without azithromycin on the mortality of patients with COVID-19 found that hydroxychloroquine alone was not associated with any mortality impact (RR = 1.09 [95% CI 0.97-1.24, n = 3 studies] for randomized controlled trials), but that the combination of hydroxychloroquine and azithromycin was associated with a significant increase in mortality (RR = 1.27; 95% CI 1.04-1.54, n = 7 studies). Because of this finding, many clinical studies, including the European DisCoVeRy clinical trial or the WHO international Solidarity clinical trial, discontinued all treatment arms with hydroxychloroquine.
The known adverse events of ivermectin include gastrointestinal and neurological symptoms, and overdoses can be associated with decreased consciousness, hallucination, coma, and death. As with many other highly-protein bound drugs, ivermectin has an effective concentration for killing SARS-COV2 (EC50) that cannot be achieved in humans using safe doses; this has led to people taking large doses of various formulations of ivermectin in attempt to gain clinical effect, with sometimes disastrous consequences (Schmith, October 2020). In April 2020, FDA issued a warning that formulations of ivermectin intended for veterinary use should not be repurposed and used in humans. On August 26, 2021, CDC issued a Health Advisory warning against the use of ivermectin to prevent or treat COVID-19. The advisory noted that prescriptions of ivermectin have increased twenty-four-fold in the United States from March 2019 to August 2021, and that there has been a parallel increase in the number of calls to Poison Control Centers relating to adverse effects from ivermectin. IDSA suggests against the use of ivermectin for the prevention or treatment of COVID-19 in hospitalized and non-hospitalized adults. NIH’s COVID-19 Treatment Guidance Panel has determined that there is insufficient evidence to recommend for or against the use of ivermectin for COVID-19 treatment. Recent legal efforts to force hospitals to employ ivermectin for the treatment of COVID-19 have resulted in a joint statement from IDSA, HIVMA, and SHEA that such actions expose patients to serious harm.
Molnupiravir (name derived from Mjölnir, the hammer of the Norse thunder god, Thor; formerly called EIDD-2801 and MK-4482) is a readily bioavailable prodrug of a ribonucleoside analogue that interferes with multiple SARS-CoV-2 viral processes, including replication. It also acts potently against several other RNA viruses, including Ebola, influenza, MERS-CoV and Venezuelan equine encephalitis virus.
In human airway epithelial cells, molnupiravir has potent effect against SARS-CoV-2, potently reducing viral production with an in vitro IC50 of 0.024 µM, without observed cytotoxicity (Sheahan, April 2020). In mice and ferret models of SARS-CoV-2, molnupiravir has shown efficacy in prevention and treatment of infection, laying a strong foundation and rationale for clinical studies (Cox, January 2021; Wahl, March 2021).
In a Phase 1 randomized controlled single and multiple ascending dose pharmacokinetic study, molnupiravir had rapid appearance in plasma of ß-d-N4-hydroxycytidine (NHC, also called EIDD-1931)with a time of peak concentrations 1 to 1.75 hours after the dose and plasma concentrations which exceeded those expected to be efficacious based on preclinical data (Holman, August 2021; Painter, March 2021).
Phase 2 studies to examine safety, tolerability and antiviral effect followed, in both outpatients with COVID-19 (Study MK-4482-006, also known as EIDD-2801-2003) and inpatients (the ongoing END-COVID study):
Study MK-4482-006 assessed safety, tolerability and antiviral effect of molnupiravir given within 7 days of symptom onset in nonhospitalized adults with confirmed COVID-19. A total of 202 participants were randomized to receive molnupiravir (200 mg, 400 mg or 800 mg twice daily for 5 days) versus placebo (N=23, 62, 55 and 62 in the respective groups). Endpoints included the time to undetectable levels of nasopharyngeal viral RNA (by RT-PCR) and time to elimination of replication-competent (i.e., infectious) virus from nasopharyngeal The drug was well tolerated; rates of any adverse event were 48% in the molnupiravir 200 mg recipients, 32.3% in the 400 mg recipients, 20% in the 800 mg recipients and 29% in placebo recipients, with similar very low rates of drug discontinuation across the groups. At day 3 after treatment initiation, there was lower recovery of virus from recipients of molnupiravir 800 mg (1.9%) versus placebo (16.7%) (p=0.02). This effect was also seen at day 5 (0% virus recovery in 400 mg or 800 mg recipients; 11.1% in placebo recipients; p=0.03). The time it took to clear viral RNA was also shorter in participants who got 800 mg of molnupiravir as compared to placebo (p=0.01). The END-COVID study in hospitalized patients with COVID-19 is still enrolling participants, with results anticipated soon (Fischer, June 2021 – preprint, not peer-reviewed).
Several Phase 3 trials of molnupiravir have been initiated for COVID-19:
MOVe-AHEAD, an ongoing Phase 3 trial of molnupiravir for prevention of COVID-19, MOVe-OUT (also called MK-4482-002), a Phase 2/3 randomized, placebo-controlled trial of efficacy, safety and PK of molnupiravir among outpatients with PCR-confirmed COVID-19.
MOVe-IN, a phase 2/3 randomized, placebo-controlled study of safety, efficacy and PK of molnupiravir among hospitalized An interim analysis of the MOVe-IN study data concluded that there was no meaningful benefit of molnupiravir in hospitalized patients, and at that time, the study was discontinued.
The manufacturers of molnupiravir issued an October 2021 press release announcing a positive result in the interim analysis of 770 participants in the randomized, placebo-controlled, multisite, Phase 3 MOVe-OUT trial of oral molnupiravir for nonhospitalized adults with COVID-19 who were at risk of progression to severe disease.
In MOVe-OUT, participants were eligible to be randomized if they were within 5 days of original symptom onset. The findings, not yet peer reviewed or published, were that 7.3% (28/385) of participants in the molnupiravir group were hospitalized at day 29 post treatment, compared with 14.1% (53/377) of the participants in the placebo group. This represented a 50% reduction in the risk of hospitalization for outpatients with mild or moderate COVID-19, and the hospitalization rate was found to be significantly different in the two groups, with a p-value of 0.0012. There was also a difference between the two groups in the number of participants who died by day 29 post treatment (0/385 deaths in the molnupiravir group compared with 8/377 deaths in the placebo group.) Based on these results, the Data Safety and Monitoring Board for the study recommended stopping the trial early (around 90% of the originally planned sample size had been enrolled at that time).
Shortly after the MOVe-OUT trial results, on Oct. 11, 2021, Merck and Ridgeback announced submission of an EUA application to FDA for oral molnupiravir, based on the MOVe-OUT trial data.
Efforts are underway to supply molnupiravir (once approved for emergency use) to governments of high-income countries as well as low- and middle-income countries. The manufacturer (in North America, Merck & Co., Inc. holds the rights to the trademark "Merck," but globally the company trades under the name MSD) has committed to nonexclusive voluntary licensing agreements with eight generic manufacturers, with a goal of distributing molnupiravir to 100 low- and middle-income countries. Notably, the same week that the MOVe-OUT trial results were announced, two Indian generic manufacturers (Aurobindo Pharma Ltd. and MSN Laboratories) announced that they would end trials of a generic version of molnupiravir in patients with moderate COVID-19 (defined as COVID-positive with oxygen saturation of >90%), due to a lack of efficacy. The MOVe-OUT trial, however, had specified that patients with oxygen saturation lower than 93% should be excluded, so there were probably slightly sicker participants in the Indian trials, and it is not known at this time whether some were hospitalized. At this time, the two generic manufacturers are continuing Phase 3 trials of generic molnupiravir in patients with mild disease.
There is accumulating human safety data for molnupiravir, though as yet no long-term safety data, which may be more relevant for genetic or reproductive toxicities. In Phase 1 single-dose and multiple ascending dose studies, more participants in the placebo arm experienced adverse events than in the molnupiravir arm (43.8% versus 35.4%, respectively, for the single-dose study, and 50% versus 42.9%, respectively, for the multiple-dose study) (Painter, May 2021). These AEs were mostly mild; the only grade 2 AEs in the molnupiravir group were headache and oropharyngeal pain. The only AE deemed potentially treatment related by the investigator was a pruritic rash, which resolved after discontinuation of the study drug. No concerning trends in clinical laboratory abnormalities were observed. In the Phase 2 and 3 studies, the participants who received molnupiravir had similar rates of AEs to placebo recipients (Fischer, June 2021 – preprint, not peer-reviewed).
Theoretical safety concerns have arisen from the drug’s mechanism. The prodrug molnupiravir is quickly converted into NHC. NHC in turn gets converted by intracellular kinases/phosphatases into the triphosphorylated active drug, NHC-TP (or EIDD-2061). NHC-TP is recognized by the RNA-dependent RNA polymerase of the coronavirus, which confuses it with cytidine, in a manner that is not detected and corrected by the proofreading exonuclease enzyme of coronavirus (an advantage in preventing the acquisition of drug resistance). NHC-monophosphate gets incorporated into the coronavirus genome and causes lethal mutagenesis (an accumulation of detrimental mutations resulting in viral error catastrophe) in the nascent strand of viral RNA, interfering further with viral functioning.
As the presumed mechanism of action of molnupiravir is to induce lethal mutagenesis in the nascent SARS-CoV-2 viral strand, questions have previously arisen about the potential for an NHC metabolite to be incorporated into human DNA and induce mutagenesis in human cells (Stuyver, January 2003). The basis for this concern is that a common ribonucleoside diphosphate form of NHC can be transformed into ribonucleoside triphosphates (causing damage to viral RNA) or be recognized by the host ribonucleotide reductase enzyme and transformed into 2′-deoxyribonucleoside triphosphates (which are potentially incorporated into and cause damage to host cell DNA in actively dividing cells). One study examined this potential using a hypoxanthine phosphoribosyltransferase gene mutation assay in cells derived from hamster ovaries and exposed the cells to high doses of molnupiravir for 32 days. The study found that 3 µM concentrations of NHC were associated with mutagenesis in cell culture (Zhou, August 2021). It is worth noting that the maximum course being studied in clinical trials is 5 days of twice daily molnupiravir, and that the positive control, 1 minute of UV light, caused a higher mutation rate than any of the drug concentrations.
As in vitro Ames testing for NHC did show positive results with mutagenicity for two of six bacterial strains, likely on the basis of the incorporation of the molecule into bacterial DNA, the manufacturer of molnupiravir has followed that with extensive in vivo preclinical toxicology studies to assess whether molnupiravir has toxicities in small mammals. This testing included the Big Blue assay, which uses transgenic rodents carrying a mutation (lacI) that allows for detection of small mutations and deletions in tissues, to test whether high doses of molnupiravir induced mutagenesis within the animals. It also included the PIG-a assay, which uses flow cytometry to detect mutations in a reporter gene to see whether a drug causes mutations in vivo. The totality of the mutational toxicology assays performed for molnupiravir were reportedly reassuring that the drug was found to be neither mutagenic nor genotoxic in mammals (Painter, October 2021).
FDA’s Antimicrobial Drugs Advisory Committee will hold a hearing on molnupiravir on Nov. 30, 2021, to make further regulatory decisions regarding the drug. Other approved antivirals such as ribavirin and favipiravir have similar (less potently) mutagenic mechanisms (Crotty, June 2001), and favipiravir has restrictions placed on its use by Japan’s regulatory body for that reason (Nagata, February 2015); it is unclear at this time whether there will ultimately be particular populations (e.g., pregnant individuals) in whom the use of molnupiravir will be restricted.
Fluvoxamine is a selective serotonin reuptake inhibitor approved by the FDA since 1994 and used in the treatment of depression (now available at low cost as a generic). Unlike other drugs of its class, fluvoxamine stimulates sigma-1 receptors on the surface of the endoplasmic reticulum, which processes and traffics proteins within cells. This results in dampened inflammatory response to sepsis in laboratory animals, and has also been shown to block SARS-CoV-2 replication (Hashimoto, March 2021). In this way it can be thought of as having both immunomodulatory and antiviral properties. Other hypothesized mechanisms of action of fluvoxamine against SARS-CoV-2 include its inhibition of platelet activation and its targeting of lysosomes (Homolak, August 2020; Schlienger, May 2003). Of note, fluvoxamine is an inhibitor of certain cytochrome P450 metabolizing enzymes, including CYP3A4, CYP1A2 and CYP2C9, so may increase levels of any co-administered drug that is metabolized by those enzymatic pathways.
Because of these theoretical, in vitro, and preclinical benefits, fluvoxamine (50 mg twice daily for 14 days after a loading dose of 50-100 mg) was administered to a cohort of horse track employees who developed COVID-19 in a California outbreak related to shared congregate living conditions. This prospective cohort study found that all 65 recipients of fluvoxamine had resolved their disease symptoms at the end of 14 days, whereas 29 of the 48 people who declined treatment still had residual symptoms at 14 days (and six were hospitalized, two intubated, and one died) (Seftel, February 2021). Subsequently, fluvoxamine (100 mg 3x daily for 14 days) was studied against placebo in a preliminary small-scale study in St. Louis, MO of 152 non-hospitalized individuals with PCR-confirmed COVID-19, and a significantly lower percentage of the fluvoxamine group had clinical deterioration at 15 days than the placebo group (0% as compared to 8.3%, respectively; 95% CI 1.8% to 16.4%; log-rank p=0.009) (Lenze, December 2020). Overall, these studies generated the hypothesis that fluvoxamine may reduce hospitalizations for adult outpatients with symptomatic COVID-19.
A much larger recent randomized platform trial in Brazil, called the TOGETHER trial, was released as a preprint in late August 2021, showing similar preliminary findings. The trial randomized symptomatic adults with confirmed COVID-19 to receive either fluvoxamine (100 mg PO BID x 10 days) or placebo; the primary endpoint was extended observation in the Emergency Department for COVID-19 or hospitalization for COVID-19. A total of 739 participants received fluvoxamine and 733 received placebo. There was an observed lower rate of hospitalization in the fluvoxamine (10.4%) than in the placebo group (14.7%) (ITT RR 0.71; 95% Bayesian Credible Interval 0.54-0.93), with a 99.4% posterior probability that fluvoxamine was superior. The trial was stopped by its data and safety monitoring board for superiority on August 6, 2021. Final data that encompasses all days of follow-up for all participants will be forthcoming.
In late 2021, the World Health Organization plans to launch a new phase of its Solidarity PLUS trial in 52 countries to test three candidate drugs: artesunate (an FDA-approved antimalarial used in intravenous form for the treatment of severe malaria), infliximab (an FDA-approved monoclonal antibody which blocks TNF alpha) and imatinib (an FDA-approved signal transduction inhibitor/tyrosine kinase inhibitor used for the treatment of Philadelphia chromosome positive leukemias).
Artesunate, like imatinib, may have both antiviral and immunomodulatory effects in COVID-19. These are theorized to be due to its inhibition of endocytosis of virus, NF-kappa B mediated dampening of cytokines such as TNF-alpha, IL-6, and IL-1, and its inhibition of MMP-2 and MMP-9 (Magenta, December 2014; Xu, February 2007). It has been shown in vitro in green monkey kidney Vero E6 cells to have anti-SARS-CoV-2 activity with clinically achievable EC50 values of 12.98 ± 5.30μM (Cao, July 2020). A clinical study of 43 patients with COVID-19 reported statistically improved clinical outcomes in the artesunate group as compared to the control group (Lin, April 2020). Some have hypothesized that artesunate might have beneficial effects in mitigating the neurological complications of COVID-19.
Regarding route of dosing, it is important to note that, while the Solidarity PLUS trial will use intravenous artesunate, the oral and intravenous forms of artesunate have been shown to have equivalent efficacy and tolerability in other settings, such as the treatment of Plasmodium falciparum malaria (Alin, December 1995). Infliximab will be also given intravenously, and imatinib orally.
The background for the choice of these three drugs is complex. It has been observed in several settings that patients receiving various types of immune inhibitors have attenuated cases of COVID-19. The SECURE-BID registry found that patients with IBD on anti-TNF antibodies had lower risk of death or hospital admission than those not on these agents (aOR 0.60 [95% CI 0.38-0.96], p=0.03) (Brenner, August 2020). The COVID-19 Global Rheumatology Alliance registry likewise found that patients with rheumatologic disease on anti-TNF agents +/- other immunomodulators had lower rates of hospital admissions (aOR 0.40 [95% CI 0.19-0.80], p=0.01) (Gianfrancesco, July 2020). These observations, combined with the moderate success of other cytokine-blocking agents (e.g. tocilizumab), have led to the study of infliximab in a broader patient population hospitalized with COVID-19. A small (n= 17) open-label phase 2 study (NCT04425538) of infliximab in patients with COVID-19 has completed enrollment and results are forthcoming. The CATALYST phase 2 randomized adaptive trial of N=146 patients with COVID-19 pneumonia in the UK compared infiliximab to either namilumab (a granulocyte-macrophage colony-stimulating factor inhibitor) or standard of care, for reducing inflammation as measured by C-reactive protein (CRP). The study found no benefit from infliximab and a high rate of adverse events in the infliximab group (narilumab, however, did appear to significantly reduce inflammation in hospitalized patients) (Fisher, June 2021 – preprint, not peer-reviewed).
Regarding imatinib, some mechanistic studies have noted that the SARS-CoV and possibly SARS-CoV-2 viruses rely on ABL2 kinases to infect host cells, and therefore that blocking this kinase specifically might have antiviral activity (Sisk, May 2018; Coleman, September 2016). Efforts to identify potentially useful antiviral drugs from large-scale high-throughput screening identified imatinib as a leading anti-SARS-CoV-2 contender, along with mycophenolic acid and quinacrine dihydrochloride (Han, January 2021). It has also been suggested that imatinib may exert its effects partially as an immunomodulator (i.e., not as a direct antiviral) via unclear mechanisms. Indeed, imatinib has been shown to have very low potency against SARS-CoV in Vero E6 cell cultures, with a clinically unachievable EC50, and in ACE2+ human Caco-2 cells, had no inhibitory effect (Zhao, November 2020). Some early clinical data suggested that imatinib may stop pulmonary capillary leak and confer clinical benefit to patients hospitalized with COVID-19 (Aman, June 2021). Case reports of rapid improvement of severe COVID-19 exist (Ortega, September 2020). A randomized placebo-controlled study of imatinib among 385 participants hospitalized with COVID-19 found a lower 28-day mortality in the imatinib group than in the placebo group (HR 0.51 [0.27-0.95]), but when factors associated with mortality were controlled for, aHR was non-significant at 0.52 (95% CI 0.26-1.05). There was however a significantly shorter duration of mechanical ventilation requirement in the imatinib group than the placebo group (median duration 7 days [3–13] vs. 12 days [IQR 6–20] respectively; p=0·0080), and shorter length of intensive care unit stay in the imatinib group (median duration 8 days [5–13] vs. 15 days [7–21] in the placebo group; p=0·025). Notably, 276 of the 385 study participants (72%) also received dexamethasone.
Using high-throughput methods of identifying potential compounds of promise for the treatment of SARS-CoV-2, several drugs have been identified, many of which have entered clinical trials. Compounds that are hypothesized to block the binding of ACE2 and the receptor-binding domain include nobiletin, glycyrrhizin, neohesperidin and SSAA09E2. Antiprotozoal agents such as nitazoxanide and ivermectin have been identified with AI and simulation-based methods as having some interaction with protein targets on SARS-CoV-2, and are theorized to have anti-inflammatory effects. Several “naturally occurring” compounds have been shown to inhibit SARS-CoV-2 in vitro (resveratrol, ginkgolic acid, baicalein) and also to synergistically augment antiviral effect when used in combination with existing antivirals (linoleic acid + remdesivir; cepharanthine plus nelfinavir) (Yang, June 2021).
Other direct antivirals (other than remdesivir) include nucleoside/nucleotide analogues, such as favipiravir (an RdRp blocker that has approval in Japan for the treatment of influenza and is being studied for COVID-19). A recent systematic review and meta-analysis of clinical trials of favipiravir for COVID-19 (with a total of 1,019 participants) found that there was a significantly higher likelihood of clinical improvement in favipiravir recipients than in control recipients in the 14 days after hospitalization ((OR = 3.03, 95% CI = 1.17–7.80), but no significant improvement in mortality. Other RdRp inhibitors include ribavirin, galidesivir and molnupiravir (note: updated information to come following Merck announcement of interim phase III results via press release). Inhibitors of viral protease enzyme, including disulfiram, lopinavir-ritonavir, darunavir, N3, 11a/11b and carmofur, are also of some interest as antivirals against SARS-CoV-2.
TMPRSS2 inhibitors such as camostat mesylate block SARS-CoV-2 cell entry in vitro and have been theorized to block SARS-CoV-2 entry into lung tissue in COVID-19 pneumonia, but a double-blind randomized, placebo-controlled trial in N=205 hospitalized patients with COVID-19 pneumonia failed to show any impact of camostat on mortality (HR 0.82 [95% CI 0.24-2.79; p=0.75]) or on time to clinical improvement (Gunst, May 2021). A related and more potent TMPRSS2 inhibitor, nafamostat, is being investigated in COVID-19 patients in a subsequent randomized placebo-controlled clinical trial (the RACONA study) which opened in June 2021.
As a general rule, in the human immune response to viruses, interferon-mediated responses precede pro-inflammatory responses (Galani, January 2021); this rule does not apply to SARS-CoV-2, where patients with severe COVID-19 had delayed and muted IFN production but robust pro-inflammatory cytokines. However, it has been observed that higher IFN-ɣ responses correlate with lower bronchial SARS-CoV-2 viral load and improved outcomes, and that lower IFN-ɣ is associated with lung fibrosis in COVID-19 patients (Hu, September 2020). In keeping with this observation, interferons have also been considered in the treatment of COVID-19, with the COVIFERON randomized controlled trial of IFNβ1a and IFNβ1b in patients with severe COVID-19 showing a significant difference in time to clinical improvement in the IFNβ1a arm as compared to the control group (HR 2.36[95% CI 1.10-5.17], p=0.031) (Darazam, April 2021). There was an extremely high mortality in all three arms (20%, 30% and 45% in the IFNβ1a, IFNβ1b and control arms, respectively).
Various other host-targeted therapeutics have been clinically tested for COVID-19, with the hope that they will prevent the dysregulated inflammatory response and their efficacy will be unchanged regardless of variant. Opaganib is a new oral drug which selectively inhibits sphingosine kinase 2 (SK2), which is an essential component of the body’s signaling response to Tumor Necrosis Factor, leading to the inflammatory cascade. It also has in vitro antiviral activity against a variety of viruses, and has been employed in five patients as treatment of severe COVID-19 under a compassionate use program in Israel, with reportedly salutary effects when compared to matched controls, including lower CRP, faster resolution of lymphocyte count, and comparatively more disease resolution (Kurd, November 2020). It was then studied in a phase 2 randomized, placebo-controlled trial in 40 adults with COVID-19 pneumonia in the U.S., and there was a greater proportion of patients on opaganib who were weaned to room air on day 14 than on placebo (53% as compared to 22%; no statistics shown in the poster presented at the 2021 World Microbe Forum). A larger phase 2/3 study is underway, with results pending.
There are many therapeutics under early investigation for treatment of COVID-19 for which there is currently insufficient clinical data to recommend either for or against. This overview is not a comprehensive summary, but a list of therapeutics with strong biological plausibility that are available in the United States and are or will be studied by clinical trial.