Treatments for COVID-19: Lessons from 2020 and new therapeutic options

To face the COVID-19 pandemic, prophylactic vaccines have been developed in record time, but vaccine coverage is still limited, accessibility is not equitable worldwide, and the vaccines are not fully effective against emerging variants. Therefore, therapeutic treatments are urgently needed to control the pandemic and treat vulnerable populations, but despite all efforts made, options remain scarce. However, the knowledge gained during 2020 constitutes an invaluable platform from which to build future therapies. In this review, we highlight the main drug repurposing strategies and achievements made over the first 18 months of the pandemic, but also discuss the antivirals, immunomodulators and drug combinations that could be used in the near future to cure COVID-19.


Introduction
Since its first description in December 2019, coronavirus disease 2019 (COVID-19) has spread throughout the world, causing more than 245 million confirmed cases and w5 million deaths as of November 1st, 2021 (https://covid19.who.int/) [1]. COVID-19 is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), a positive-sense single-stranded RNAenveloped virus from the Coronaviridae family [1,2]. Over the last 18 months, the scientific community has deciphered SARS-CoV-2 biology and COVID-19 physiopathology, and developed preventive and curative treatments [3,4]. Since December 2020, several prophylactic vaccines have been approved, others are under development, and a large vaccination campaign has begun [5e7]. However, to date, only a third of the world population is fully vaccinated, countries have unequal access to jabs, and we are far from the vaccine coverage that could effectively protect the population if vaccine effectiveness was 100% [7]. Thus, complementary therapeutic treatments are necessary to protect vulnerable populations from fatality or prolonged COVID-19 sequelae [8], face the emergence and dissemination of SARS-CoV-2 variants associated with potential resistance to treatments [9e11], and sustainably mitigate the impact of the pandemic. This review takes stock of the main therapeutic strategies evaluated in the past 18 months and discusses promising future strategies.

Physiopathology of COVID-19 disease
The natural history of SARS-CoV-2 in the host has been extensively described elsewhere [3,12]. Accumulating evidence shows that, although the respiratory tract is the primary target of the virus, COVID-19 is a systemic disease that in the worst cases can lead to multiorgan failure [13]. Despite most patients being clinically asymptomatic or having mild-to-moderate clinical manifestations, 10e20% will develop severe pathology that can evolve toward critical illness and eventually death, especially those with risk factors or comorbidities [14,15]. Recent studies have also shown that inborn genetic defects of type I interferon (IFNe I) responses [16] or the presence of neutralizing autoantibodies against IFN-I also contribute to disease severity [17,18]. The physiopathology can broadly be divided into three stages that may overlap: (I) the upper respiratory tract (URT) viral stage, (II) the pulmonary stage, and (III) the hyperinflammatory stage ( Figure 1) [14]. Stage I begins at the time of infection and includes the incubation period and early establishment of the disease; patients whose immune response successfully limits viral spread likely have a good recovery. Others may progress toward moderate illness (stage II), where the main symptoms are pneumonia, with fever, cough, and hypoxemia in some but not all cases. Some patients present with severe/ critical forms of the disease (stage III), with symptoms including severe respiratory distress, shock, and organ failure. Although only a small number of patients reach this stage, the mortality rate is considerably increased in this population [14,15,19].
Mechanistically, COVID-19 can be defined as a twophase disease. While the first phase is mostly driven by viral replication in the URT, the second is mainly due to a host-dependent deregulated inflammatory response, leading to cytokine storm, damage to different organs, and eventually systemic failure. This particular evolution of the clinical presentation is of upmost importance for patient management, for which the choice and timing of the appropriate treatment depends on the phase of the pathology [20].

Treatments for COVID-19: lessons from 2020
During 2020, considering the urgency for therapeutic solutions to fight COVID-19, drug repurposing strategies were favored. Among the hundreds of clinical trials completed or still in progress, many antiviral molecules and immunomodulators were administered to patients with severe COVID-19. This section focuses on those treatments evaluated in large and adaptive randomized clinical trials (RCTs) with robust statistical validation, which have proved to bear the highest predictive value ( Table 1).
The SOLIDARITY and RECOVERY trials were the flagship international adaptive RCTs, designed to study the efficacy of several molecules in large cohorts of moderate/severe COVID-19 patients. Candidates included antiviral molecules such as experimental remdesivir (RDV), FDA-approved lopinavir/ritonavir (LPV/r), and hydroxychloroquine (HCQ), antibiotics such as azithromycin, and immunomodulators such as   Time course of COVID-19 progression and therapeutic options. Time course of COVID-19 progression (blue line), shows an increasing disease severity (asymptomatic to critical) through the three stages of the disease: upper respiratory tract viral stage (I), pulmonary stage (II) and hyperinflammatory stage (III). The average (non-adjusted by age, comorbidities, etc) percentage of patients in each stage (orange line) shows that only a small proportion of patients (up to 10-20%) develop a severe to critical disease. The "therapeutic options" section presents the evolution regarding the administration time point of therapeutic molecules in 2020 vs 2021. dexamethasone, convalescent plasma, and interferon beta-1a (IFNb-1a). In October 2020, the WHO SOLI-DARITY trial published interim results showing that RDV, HCQ, LPV/r, and IFNb-1a regimens had little or no effect on overall mortality, initiation of ventilation, or duration of hospital stay in hospitalized patients with COVID

New therapeutic options
This section discusses relevant preclinical studies and clinical trials evaluating promising new monotherapies and drug combinations (Tables 1e3).
Antivirals: target the virus at the early stage of infection Antiviral therapies can target viral and/or host determinants that are crucial for viral replication (Table 1). SARS-CoV-2 entry relies on the angiotensin-converting enzyme-2 (ACE2) as a receptor and also transmembrane serine protease 2 (TMPRSS2) that cleaves and activates the viral spike (S) protein [31]. One strategy to inhibit S/ACE2 interaction is the use of neutralizing antibodies. Treatment of high-risk adult outpatients with mild/moderate COVID-19 with sotrovimab (STV), an anti-S monoclonal antibody (mAb), resulted in an 85% reduction in the risk of hospitalization or death [32]. According to GSK (https://www.gsk. com/en-gb/media/press-releases/gsk-and-virbiotechnology-announce-sotrovimab-vir-7831-receivesemergency-use-authorization-from-the-us-fda/), STV maintains in vitro activity against all known variants of concern, including B.1.617.2, and it recently received an FDA Emergency Use Authorization (EUA). mAbs are at the frontline of currently developed COVID-19 treatments, but their use should be based on an accurate characterization and selection, since they might either inhibit or enhance S-mediated membrane fusion and formation of syncytia [33]. Moreover, mAb monotherapies have already shown their limits, as exemplified by the resistance of B.167.2 to bamlanivimab [11], therefore current directives suggest the use of mAb cocktails to target multiple and complementary epitopes [34,35]. Camostat mesilate, a TMPRSS2 inhibitor approved for treatment of pancreatitis in Japan [36,37], was repurposed as a potent inhibitor of SARS-CoV-2 entry in vitro. However, no improvement was observed in the first clinical trial including 137 hospitalized patients with mild/moderate pathology [38]; other clinical studies are ongoing.
Nucleoside analogs (NAs) have historically been primary options against emerging RNA viruses and this was no exception for SARS-CoV-2. Molnupiravir (MPV), a potent experimental NA and prodrug of the synthetic nucleoside derivative N4-hydroxycytidine, introduces copying errors during viral RNA replication [39,40]. Initially developed by Merck to inhibit influenza viruses [41], MPV has the advantage over RDV of being administered orally. Preclinical data showed that it efficiently inhibited SARS-CoV-2 replication in vitro and in in vivo models [42e44], and its therapeutic potential is currently being tested in patients with mild/moderate COVID-19. In patients with confirmed SARS-CoV-2 infection and symptom onset within 7 days, MPV cleared infectious viruses from nasopharyngeal swabs at day 5 and reduced by 50% the time to elimination of SARS-CoV-2 RNA (14 vs 27 days for placebo) [45]. However, some concerns have been raised regarding MPV due to its toxicity and potential teratogenic effects [40,46]; a phase 2 trial is currently evaluating its safety and efficacy on SARS-CoV-2 shedding in hospitalized patients. Other repurposed drugs potentially targeting multiple steps of the SARS-CoV-2 life cycle are currently being tested, such as ivermectin (IVM), a broad-spectrum antiparasitic agent with previously reported antiviral activity in vitro [47e49]. Although a small study on 57 patients with mild COVID-19 showed a reduction in symptom duration and viral detection in the URT of treated patients [50], no robust evidence to support its clinical use was provided by other studies [51,52] or by a recent Cochrane meta-analysis [53].
Preclinical evaluation of other molecules (Table 2), such as soluble ACE2 peptides that compete with human ACE2 for the interaction with S, or inhibitors of cathepsin L (SLV213), a cysteine protease that plays a key role in SARS-CoV-2 entry, is ongoing [54]. Both soluble ACE2 peptides and SLV213 blocked SARS-CoV-2 infection in vitro with a 50% inhibitory concentration (IC50) in the nanomolar range [55,56]. A phase 2 clinical trial in patients with mild/moderate COVID-19 is planned. Another approach evaluates virus neutralization with a new class of genetically engineered antibody mimetic drugs (DARPinÒ) [57]. Pre-treatment using ensovibep, a DARPin targeting three domains of S, reduced virus load in the URT and LRT in vivo by four and two log, respectively [58]. Ensovibep is currently under evaluation in a phase 2/3 clinical trial in patients with mild COVID-19 and it recently received FDA Fast Track Designation for use in both hospitalized and ambulatory settings. Plitidepsin, an inhibitor of eukaryotic translation elongation factor 1A, reduced  lung SARS-CoV-2 viral load by two-fold in vivo in hACE2-transgenic mice [59], and its safety and efficacy is currently being evaluated in a phase 3 trial in patients with moderate COVID-19. Niclosamide, an anthelminthic and antibacterial drug, was formulated with human lysozyme for its delivery to the URTand LRT. Its intranasal administration improved survival and reduced lung SARS-CoV-2 viral load by three log in hACE2transgenic mice [60]; its safety and efficacy is being currently evaluated in mild/moderate COVID-19 patients. In March 2021, Pfizer initiated a phase 1 study of PF-07321332, a novel antiviral targeting the SARS-CoV-2 3CL protease and concluded that it has a potent in vitro antiviral activity against SARS-CoV-2 and other coronaviruses.
Immunomodulators: restoration of the antiviral immunity and control of cytokine storm The use of immunomodulators has proved to be an efficient strategy to dampen hypercytokinemia and improve patients' clinical outcomes (Table 1). IFN-I and type III interferons (IFN-III) are central players of the early immune response, rapidly induced during viral infection to restrict virus propagation [61]. Even if SARS-CoV-2 has evolved mechanisms to evade IFN-I and IFN-III responses [62], restoring potent IFN responses at early stages of infection could be beneficial. In 2020, subcutaneous administration of type I IFNb-1a in 2063 hospitalized patients with severe COVID-19 was not associated with improved clinical outcomes [21]. However, in 2021, treatment with an inhaled formulation of IFNb-1a showed more than three-fold greater odds of improvement and recovery on day 28 versus the placebo group in 48 hospitalized patients with moderate/severe disease [63]. Another trial evaluated a subcutaneous administration of type III peg-IFNl-1 in 30 outpatients with moderate and mild disease. By day 7, decline in SARS-CoV-2 RNA in nasal swabs was 2.42 log greater with peg-IFNl-1 and viral clearance was two-fold greater for patients with baseline RNA of 10⁶ copies/ml or greater versus the placebo group [64]. Finally, our group has shown that diltiazem, an old antihypertensive drug (calcium channel blocker) has in vitro and in vivo host-directed anti-influenza properties, mainly due to its capacity of inducing the IFN-III antiviral response within reconstituted human airway epithelium (HAE) [65]. Diltiazem also displays antiviral effects against SARS-CoV-2 in in vitro and nonhuman primate (cynomolgus) models ( [66,67] and unpublished data). A double-blind randomized phase 2 clinical study in patients with mild/moderate COVID-19 is planned at the end of 2021.
Given the positive results obtained with dexamethasone, the COVIDICUS trial is currently testing low versus high doses of this drug in severe COVID-19 patients. Other corticosteroids are also being evaluated for their control of immune and inflammatory responses. Inhaled glucocorticoids are commonly used to treat patients with asthma or chronic obstructive pulmonary disease and, interestingly, some studies have shown that these pathologies were underrepresented among hospitalized COVID-19 patients [68,69]. To assess whether the widespread use of inhaled glucocorticoids in these patients might account for this observation, inhaled anti-inflammatory budesonide was administered to 73 outpatients with mild COVID-19 and chronic respiratory disease. Results showed a 91% reduction in the likelihood of needing urgent medical care and a clinical recovery 1 day shorter versus the usual care group [70]. More recently, inhaled budesonide was tested in 2530 mild/moderate COVID-19 non-hospitalized patients and showed an improved time of recovery (11.8 vs 14.7 days) and reduced hospital admissions and deaths (6.8% vs 8.8%) versus the usual care group [71]. In addition to an uncontrolled inflammatory response, COVID-19 is often associated with coagulopathy and thromboembolic events in severe patients [72]. The use of anticoagulants has been considered to treat patients, but there are contradictory meta-analysis data [72,73]. Recently, two studies from combined clinical trial programs showed that therapeutic-dose heparin increased the probability of survival and reduced need for organ support in moderate but not severe hospitalized patients versus usualcare thromboprophylaxis [74,75]. However, more major bleeding events were observed in therapeutic versus thromboprophylaxis treatment, raising caution regarding the use of such molecules, with the WHO advising low doses to be used in hospitalized patients. Alternatively, fluvoxamine, an antidepressant with potential immunomodulatory effect and putative antiviral properties, has been recently evaluated in a phase 3 clinical trial among 1497 high-risk symptomatic patients [76e78]. Preliminary results suggest a potential positive effect of the drug on COVID-19 patients but further investigations are needed.
Interleukin-1 and -6 (IL-1 and IL-6) are two of the main pro-inflammatory cytokines produced during infection and are also drivers of cytokine storms [79]. Tocilizumab, an anti-IL-6 receptor mAb, was evaluated in 2375 hospitalized patients with severe COVID-19 with both hypoxia and systemic inflammation. Compared with placebo, it reduced mortality by 12%, increased chances of successful hospital discharge by 14%, and reduced the chances of requiring invasive mechanical ventilation by 20% [80,81]. Tocilizumab recently received an EUA for the treatment of hospitalized adults and pediatric patients who are receiving systemic corticosteroids and require supplemental oxygen. Anakinra, an IL-1 receptor antagonist, was also evaluated in 59 hospitalized patients with mild-to-moderate COVID-19 pneumonia but showed no benefit and resulted in an increased risk of bacterial sepsis and hepatic cytolysis [82]. Nonetheless, an observational study in 62 patients admitted to hospital for severe COVID-19 with respiratory insufficiency and hyperinflammation showed a 12% reduction in mortality versus the placebo group without adverse events [83].
Internalization of ACE2 receptors, while in complex with S, induces a loss of its enzymatic function, and eventually increases angiotensin II plasma levels, promoting inflammation and leading to ARDS [84]. Treatment with the angiotensin receptor blocker telmisartan resulted in a reduction in the median discharge time by 6 days, in ICU admission by 2.4 fold, and of up to 80% in mortality at day 30 versus placebo in 78 hospitalized COVID-19 patients [85].
Altogether, further large and adaptive RCTs are still needed to assess the therapeutic potential of IFNs, inhaled glucocorticoids, IL-1/IL-6 blockers, and angiotensin receptor blockers in COVID-19.

Drug combinations
Although the complexity of COVID-19 makes unlikely the discovery of a unique drug suitable to treat all patients, combining virus-and host-targeting agents could be an alternative to monotherapy approaches, which, up to now, have mostly failed (Tables 2 and 3).
Cocktails of neutralizing mAbs targeting S are currently used to block viral attachment and entry. Casivimab and imdevimab, two non-competing antibodies, are being evaluated in mild/moderate nonhospitalized COVID-19 patients [34]. Interim results show a reduction of 0.56 log10 viral copies/ml in the serum of primary infected patients, with the largest benefit observed in patients with higher viral loads at baseline (reduction of up to 1.03 log10 copies/ml vs placebo). These promising results prompted an EUA for this cocktail. Copin et al. also showed that this cocktail is effective against all current variants in vitro, in vivo, and in SARS-CoV-2 isolates of 1000 COVID-19 non-hospitalized patients, while an additional third mAb further increased protection against viral escape [35]. Administration of bamlanivimab and etesevimab in 577 nonhospitalized patients with mild/moderate disease resulted in a reduction of 0.57 log10 viral load in treated patients [86]. Early administration of this treatment in nonhospitalized patients with mild and moderate symptoms and at high risk for severe disease reduced both hospitalization rate (2.1% vs 7% in treated vs placebo) and viral load (À1.2 log) [87]. This cocktail received an EUA in February 2021; however, a recent update now advises against its use due to reduced or lack of activity against the B.1.351, P.1, and B.1.617 variants [11]. A preclinical study in macaques evaluating C135-LS and C144-LS, mAbs with extended half-life and targeting complementary epitopes of S, showed a significant reduction in virus replication in the URT and LRT, coupled with reduced lung inflammation and improved clinical scores [88]. Safety, pharmacokinetics and efficacy of this cocktail in COVID-19 patients are currently under investigation in phase 1 and as part of the ACTIV-2 trial, respectively. Preclinical experiments on tixagevimab and cilgavimab, two synergistic long-acting anti-S mAbs, showed that this cocktail is able to protect animal models from infection and demonstrated potent neutralizing activity including against B.  [89,90]. This cocktail is currently being evaluated for postexposure prophylaxis, but also as a treatment for outpatients and inpatients.
Preclinical evaluation of the antifungal itraconazole and the antidepressant fluoxetine, both proposed to have antiviral activity against SARS-CoV-2 most likely through modulation of the endosomal cholesterol pathways, have shown synergistic activity in combination with RDV when compared with monotherapy [91]. Our group has shown that diltiazem displays synergistic effects against SARS-CoV-2 in HAE when combined with RDV [66]. Moreover, a suboptimal low-dose combination of a coronavirus RNA-dependent RNA polymerase enzyme inhibitor (GS441524) and an inhibitor of the M pro protease (GC376) synergistically blocked SARS-CoV-2 infection in the mouse respiratory tract [92]. Of note, GS441524 is the parent nucleoside of RDV. It has a 1000-fold higher concentration in serum than RDV when the latter is administered by IV injection and a better safety profile than the prodrug in vitro and in animal models; it also has antiviral activity comparable with RDV and is well-suited for pneumocytetargeted delivery [93,94].
The dual viral-immunologic nature of COVID-19 sets a strong case for combining antiviral and immunomodulatory molecules. In 2020, a clinical trial tested the triple association of LPV/r, ribavirin, and IFNb-1b versus LPV/ r alone in 127 patients with mild/moderate COVID-19 [95]. The combination shortened the time to negative nasopharyngeal swab (7 vs 12 days), alleviated symptoms in 4 days (8 in control), and reduced hospital stays to 9 days (14.5 in control). Treatment with RDV and baricitinib, a selective inhibitor of Janus kinase 1 and 2 that inhibits production of cytokines [96,97], in 1033 hospitalized patients with moderate/severe disease resulted in faster recovery, particularly in patients under ventilation (10 vs 18 days with placebo) [98]. A preclinical study evaluated the efficacy of RDV associated with the glucocorticoid methylprednisolone against SARS-CoV-2 infection in hamsters. Early treatment with methylprednisolone significantly reduced pulmonary inflammation and alleviated tissue damage, although it increased tissue viral RNA loads and titers and suppressed the development of anti-S antibodies [99]. The combination inhibited viral replication and reduced tissue inflammation. This dichotomy highlights the need of adapting treatment regimens to the disease stage and patient conditions.

Conclusions
Over the last 18 months, knowledge of COVID-19 has dramatically increased, vaccines have become available, but curative treatments are still lacking [4, 14,15]. The holy grail will probably not be found, but taking account of patients' clinical picture, comorbidities, and disease stage will maximize the chances of success. As such, the combination therapy approach, already used to treat other viral diseases [100e102], is promising. From onset of symptoms, a combination of antiviral molecules, such as FDA-approved mAbs combinations [34,35] or new combinations including NAs plus mAbs (e.g., MPV þ casirivimab) and mAbs or NAs plus host-targeting agents with antiviral activity (e.g., MPV þ diltiazem), could be interesting. For patients with comorbidities and for those with moderate/ severe COVID-19, combining antiviral molecules with immunomodulators should be considered. Combining NAs with Janus kinase inhibitors (e.g., MPV þ baricitinib) or mAbs plus corticosteroids (e.g., casirivimab þ imdevimab þ dexamethasone) are possible options. Combination therapies have the potential to prevent disease progression in vulnerable populations, and notably in the context of the potential emergence of virus variants with increased pathogenicity and/or reduced susceptibility to preventive/ therapeutic treatments. However, strong validation in relevant preclinical models is mandatory before clinical evaluation. Eventually, identification of early prognostic biomarkers such as IL-6 and autoantibodies against IFN-I, reported to predict poor evolution of the disease and whose levels seem to correlate with disease severity, could also be helpful [17,18,103,104].

Author contributions
FS, TL, EL, AP, and MR-C conceived and discussed the content of the manuscript. FS, TL, and EL drafted the manuscript. FS and TL made the tables and figures. AP and MR-C revised the manuscript. All authors revised the final version of the manuscript.

Conflict of interest statement
APand 693-704. In this international adaptive randomized clinical trial, the authors report that in patients hospitalized with COVID-19, the use of dexamethasone was beneficial among those who were receiving either invasive mechanical ventilation or oxygen alone at randomization but not among those receiving no respiratory support.