Table 1.

Proposed mechanism of action and clinical data for drugs still under investigation or of unproven efficacy for COVID-19

Atovaquone	Unknown mechanism of action but hypothesized to exert antiviral activity through binding of SARS-CoV-2 spike protein [90] and potential modulation of cellular nucleotide pools, thus disrupting viral replication [91]. Identified as a potential candidate with activity against SARS-CoV-2 via computational molecular docking simulations due to its capability for spike protein binding [90]. A double-blind, RCT has been completed evaluating (atovaquone plus SOC) vs. (placebo plus SOC) with results pending (NCT04456153)
Chloroquine (CQ) or hydroxychloroquine (HCQ) ± azithromycin	Chloroquine, hydroxychloroquine (hydroxyl analog of chloroquine) and azithromycin have shown in vitro activity against SARS-CoV-2 in infected Vero E6 cell [92, 93]. Also believed to have immunomodulatory activity which could theoretically benefit patients with COVID-19. Multiple large retrospective studies and RCTs have shown no convincing benefit of CQ or HCQ in treatment [94–96] or prophylaxis [97, 98] of COVID-19. Similarly, benefit was not demonstrated in outpatients with COVID-19 [99, 100]. The combination of CQ/HCQ with azithromycin is also associated with increased risk for cardiovascular toxicity (e.g., QTc prolongation)
Famotidine	Famotidine is a H2-receptor antagonist hypothesized to inhibit replication of SARS-CoV-2 by binding to its papain-like protease [101]. However, no in vitro antiviral activities were shown in SARS-CoV-2 infected Vero E6 cells [102]. A meta-analysis of five retrospective studies showed no significant protective effect of famotidine in decreasing the rates of severe illness, intubation, or death in patients with COVID-19 [103]
Favipiravir	Favipiravir is an RNA-dependent RNA polymerase inhibitor. Not available in the USA. In an open-label, non-randomized clinical trial in China, favipiravir was associated with reduced duration of viral shedding and time to improvement of chest imaging compared to lopinavir/ritonavir [104]. In other RCTs, favipiravir showed no significant protective effect in viral clearance [105] or time to clinical recovery [106]
Fluvoxamine	Selective serotonin re-uptake inhibitor (SSRI) with several hypothesized mechanisms of action including anti-inflammatory cytokine regulation via inactivation of sigma-1 receptor activity, interference of lysosomal trafficking of SARS-CoV-2, and decreased platelet aggregation [107]. One RCT (N = 152) found a lower likelihood of clinical deterioration in adult outpatients with symptomatic COVID-19 treated with fluvoxamine 100 mg three times daily for 15 days compared to placebo [108]. Observational data have shown that other SSRIs (e.g., escitalopram and fluoxetine) were associated with decreased risk of intubation or death [109]. Additional studies are needed to determine fluvoxamine’s place in therapy as an antiviral or adjunct treatment for minimizing inflammatory or procoagulant effects of COVID-19
Interferons (IFNs)	Interferons upregulate the body’s natural immune system which theoretically can benefit patients with viral infections. IFN alpha and beta have shown in vitro activity against SARS-CoV-2, and IFN beta seems to produce a stronger effect on SARS-CoV-2 than on SAR-CoV-1 [110, 111]. Clinical trial data that specifically evaluates interferon therapy in COVID-19 infection are lacking. Available trials evaluated IFNs as a component of a combination therapy. An open-label RCT showed that the combination therapy of IFN beta-1b, ribavirin, plus LPV/RTV was more effective than LPV/RTV monotherapy in shortening the time to negative SARS-CoV-2 polymerase chain reaction (PCR) results and symptom resolution in patients who had < 7 days of symptoms [112]. Another open-label RCT in Iran among patients with severe COVID-19 illness showed significantly reduced 28-day mortality with subcutaneous interferon 1a compared to the control group [113]
Ivermectin	Conflicting in vitro evidence of activity against SARS-CoV-2. Several retrospective cohort studies and small RCTs have shown mixed outcomes with ivermectin. An RCT Iraq (N = 118) with mild to severe COVID-19 patients showed that ivermectin plus doxycycline was associated with reduced time to recovery [114], and another study conducted in Bangladesh (N = 72) among patients with mild COVID-19 showed a faster viral clearance with ivermectin monotherapy without similar effect on clinical symptoms [115]. However, a larger RCT in Colombia among hospitalized patients with mild COVID-19, 5-day ivermectin therapy did not improve median time to resolution of symptoms compared with placebo [116]. Furthermore, the high risk of bias, lack of blinding, wide heterogeneity, and some concerns over data veracity among available RCTs with ivermectin means definitive interpretation of ivermectin’s role in COVID-19 awaits larger, RCTs [117]
Lopinavir with ritonavir (LPV/RTV)	Lopinavir is proposed to block the main protease of SARS-CoV-1 and inhibit viral replication. An open-label RCT from China early in the pandemic suggested that LPV/RTV treatment given within 12 days from symptom onset was associated with faster recovery and lower mortality than those in the standard-of-care (SOC) group [118]. However, subsequent RCTs, including two large open-label RCTs among hospitalized patients with COVID-19, LPV/RTV was not effective in reducing the time to mortality, hospital discharge, and risk for progression to mechanical ventilation [19, 119]
Molnupiravir	Prodrug of ribonucleoside analog beta-d-N4-hydroxycytidine which competes as a substrate for viral ribonucleic acid (RNA) polymerases, causing viral error catastrophe due to increased viral mutation beyond a biologically tolerable threshold, impairment of viral fitness [120]. In vivo data demonstrated significantly decreased SARS-CoV-2 transmission in ferret models and potent antiviral effect in human epithelial cells (half maximal inhibitory concentration [IC50] 0.15 uM) [121, 122]. Phase 1 and 2 data showed potential efficacy and safety in humans [120]. A phase 3 study evaluating molnupiravir for the outpatient management of COVID-19 was stopped early at interim analysis due to favorable reduction in risk of hospitalization or death, but the full results have not been peer-reviewed or published (NCT04575597) [123]. Manufacturer terminated the phase 3 portion of its study evaluating molnupiravir for hospitalized patients with COVID-19 due to lack of perceived clinical benefit [124]
Nitazoxanide	Antiprotozoal drug with in vitro activity against a range of viruses (e.g., respiratory syncytial virus [RSV], hepatitis B virus [HBV], hepatitis C virus [HCV]). It is hypothesized that nitazoxanide affects SARS-CoV-2 entry and fusion into host cells, disrupts viral genome synthesis and translation, packaging and release of virions [125]. In vitro evidence suggests similar efficacy against SARS-CoV-2 replication in cell culture assays (EC50 2.12 uM) compared to remdesivir (EC50 0.77uM) and chloroquine (EC50 1.13uM) [92]. One RCT (N = 465) found nitazoxanide 500 mg three times daily for 5 days for mild COVID-19 significantly reduced the viral load but showed no difference in symptom resolution compared to placebo [126]. Majority of clinical data are observational or retrospective with high risk of bias. A placebo-controlled RCT in preprint also found a significant decline in viral load in mild or moderate COVID-19 with nitazoxanide but no clinical outcomes were measured [127]
Nitric oxide (NO)	Although studied in non-COVD-19 patients with acute respiratory distress syndrome (ARDS), inhaled NO has limited evidence in COVID-19. It is recommended as a rescue therapy for severe ARDS in COVID-19 patients where other options fail to improve oxygenation (weak recommendation, CIII) [3•, 67]. A Cochrane review of 13 trials in non-COVID-19 ARDS patients did not show a mortality benefit (38.2% vs. 37.5%, RR = 1.04; 95% CI, 0.9 to 1.19) and only a transient improvement in oxygenation [128]
Ribavirin	Ribavirin was not found to be effective against SARS-CoV-1 in vitro [129], and the combination of ribavirin with interferon-alpha against MERS-CoV did not reduce mortality in a multicenter observational study [130]
Ruxolitinib	Janus-associated kinase (JAK) 1 and 2 inhibitor capable of suppressing cytokine signaling responsible for hyperinflammatory response and organ damage in moderate to severe COVID-19 [131]. One RCT (N = 43) comparing ruxolitinib to placebo found no statistical difference any clinical outcome but numerically favored ruxolitinib for faster median time to clinical improvement (12 vs. 15 days), improvement on imaging at 14 days (90% vs. 61.9%), and 28-day mortality (0% vs. 14.3%) [132]. Similar mechanism to baricitinib so the lack of statistical difference in the RCT may be due to small sample size
Vitamin D	Hypothesized to play a role in ARDS [133]; Vitamin D deficiency (defined as serum 25-hydroxyvitamin D concentration < 20 ng/mL) may be associated with higher risk of developing several infections [134]; In an open-label RCT among hospitalized patients with COVID-19 pneumonia (N = 76), vitamin D supplementation was associated with reduced rate of ICU admissions (vitamin D vs. SOC; 50% vs. 2%; p < 0.001) [135]. In another double-blind, placebo-controlled RCT among hospitalized adults with moderate to severe COVID-19 (N = 240), vitamin D supplementation increased serum 25-hydroxyvitamin D concentrations but failed to improve clinical outcomes compared to placebo [136]