Trial analysis  
We focused our analysis on the most representative ones including HCQ, remdesivir, lopinavir/ritonavir, favipiravir and tocilizumab with the intent that this will clarify inconclusive results from inadequate designs and guide potential future strategies.
1) hydroxychloroquine (HCQ) trials
A search (until April 7th, 2020) was performed on clinical trial registries of privately and publicly funded clinical trials (https://clinicaltrials.gov/) focusing on HCQ and COVID-19. A total of 44 registered clinical trials of HCQ were retrieved. Thirty (30) of these were therapeutic trials for active infection and 21 were prophylactic trials, with several studies having multi-purpose designs. The prophylactic trials were separated into 15 pre-exposure (PrEP) and 6 post-exposure prophylaxis (PEP) trials. These trials have exploded in number over the past month (see Figure. 3), especially in preventative trials.
CQ was first discovered to have potential activity against COVID-19 in China46 and Chinese experts recommended CQ to patients diagnosed as mild, moderate and severe cases of COVID-19 pneumonia47. A dose of 500 mg bid for 10 days is recommended for chloroquine phosphate. HCQ has emerged as a drug of interest due to its relatively good clinical safety. HCQ has had some promising preliminary results. Compared to standard care, HCQ treatment has been associated with increased symptomatic pneumonia improvement (80.6% vs. 54.8%)4 and higher levels of viral clearance (70.0% vs 12.5%)48 when compared to the standard of care, though only in relatively small studies. On the other hand, HCQ has not shown an effect in all studies, with Chen J et al. reporting no difference in HCQ-treated and control group patients’ viral clearance49. These findings are summarized in Table 3.
HCQ is also being studied in prophylactic trials, both pre- and post- viral exposure with a much larger sample size in the former (see Table 4). Most of these use a regimen of 800 mg of HCQ sulfate for day 1 and 400 mg daily for the rest of the trial. Median trial duration is 7 days. These trials are still on-going, but it should be noted that in pre-exposure studies, the population is mostly healthcare workers who likely have a disproportionately high amount of viral exposure, versus a more general lay population in post-exposure trials. Thus, some of these trials may not be generalizable to a larger population.
Due to the high transmission of SARS-Cov-2, close contacts in household setting (~15%) and other high-risk close contacts (~10%) have a high secondary infection rate.51 A preventive strategy using antiviral drugs is being adopted aimed at reducing transmission and hospitalization. As an existing drug, HCQ has the advantages of sufficient drug stockpile, good drug safety and economic cost, which provide the rationale that HCQ is being widely used.52
In treatment or PEP purpose studies, the most common dosage regimen in these trials is 800 mg for the first day + 400 mg daily for maintenance dosing (HCQ sulfate), where the dose interval is mostly twice a day. In the treatment purpose study, the median treatment period was 7 days (range: 5-14 days); which is longer than the median treatment period in the PEP study (median: 5 days, range: 3-11 days). For PrEP study, designs mainly focused on healthcare workers who are at potential risk to exposure to the COVID-19 from patients and HCQ is given over a longer treatment duration by daily regimen or regimens with longer than 24 hour dosing interval.
Although these large scale prevention trials should only be conducted after clinical confirmation of anti-viral efficacy by HCQ, their data, if released, will confirm the efficacy and safety of HCQ against the COVID-19.
Remdesivir trials
Remdesivir was originally developed by Gilead Science for Ebola virus infections 39 and is still under investigation. After a case report of a severe case of COVID-19 in the US being cured by remdesivir53, a number of clinical trials on remdesivir started worldwide.
These trials include the Chinese studies (NCT04252664, NCT04257656) initiated by Dr. Cao Bin, the global multicenter study initiated by NIAID in the US (NCT04280705), the solidarity study initiated by WHO (remdesivir as a group, NCT04330690, NCT04321616) and 2 studies initiated by Gilead Science (NCT04292730, NCT04292899). More details of these trials are summarized in Table 2. Since there is no documented clinical data so far, it is still unknown whether remdesivir is effective for COVID-19.
Remdesivir is an adenosine analogue that can incorporate into the nascent viral RNA chains and cause the termination of viral replication.54 In a mouse model of SARS-CoV, remdesivir significantly reduced lung viral load, improved pulmonary function, and reduced clinical disease.37 According to previous reports, remdesivir is a potent inhibitor of SARS-CoV-2 replication in infected cells suggesting potential therapeutic activity against COVID-19.36
The treatment protocol for these trials is 200 mg IV remdesivir on Day 1, followed by 100 mg IV remdesivir as a daily maintenance dose for the duration of hospitalization, up to 5 or 10 days maximum, depending on the trial. This is the dose of remdesivir used to treat Ebola virus. However, this may not be an optimized dose for treatment of COVID-19. Early studies on remdesivir’s action on COVID-19 report a much higher EC50 and EC90 (0.77 and 1.76 μM) than that for remdesivir’s action against Ebola (EC50 ≈0.1 μM). It should also be noted that the Ebola remdesivir trials from which this dosing regimen was borrowed showed negative results.39
These trials also use different dose timings, using both symptomatic disease and RT-PCR to define their therapeutic windows. For example, the Chinese trials give doses within 8 days of symptom onset for mild and moderate cases, or 12 days for severe cases while the NIAID trial uses a window of 3 days after PCR confirmation. The Gilead trials use a window of 4 days after PCR confirmation. It is unknown what therapeutic timing window will be most effective.
These trials also study slightly different populations. Only NCT04252664 assesses remdesivir in patients with mild COVID-19. All other studies are limited to moderate and severe cases.
A recent case series on 53 severe COVID-19 patients who received remdesivir in the context of compassionate care shows some encouraging results.55 36 patients (68%) had an improvement in oxygen requirement or ventilatory support, 25 patients (47%) were stabilized and discharged, and 7 patients (13%) died. Although this trial was not conclusive and did not have a control group, historical comparisons with general care or other clinical trial, such as 22% mortality overall in 201 In Wuhan, China56 and 22% mortality in a recent randomized, controlled trial of lopinavir–ritonavir3. The compassion-use data suggest that remdesivir may have clinical benefit in patients with severe COVID-19. This study is limited by the small size of the cohort, the relatively short duration of 28 days follow-up, potential missing data, and the lack of a randomized control group. However, these results are encouraging and open the door for future randomized, placebo-controlled trials of remdesivir therapy.
3) Lopinavir-ritonavir
Lopinavir-ritonavir is originally a medication used for HIV but has been used for other RNA viruses including SARS-CoV. An open-label study published in 2004 suggested, by comparison with a historical control group that received only ribavirin, that the addition of lopinavir–ritonavir (400 mg and 100 mg, respectively) to ribavirin reduced viral load, incidence of acute respiratory distress syndrome, and mortality among patients with SARS.57 These results suggest potential activity for SARS-CoV-2 as well.
Compared to HCQ and remdesivir, fewer trials exist for lopinavir-ritonavir treatment of COVID-19. In the largest trial of lopinavir-ritonavir in 199 adult patients hospitalized for severe COVID-19, there was no difference in time to clinical improvement between patients given 400 mg lopinavir and 100 mg ritonavir twice daily for 14 days compared to those given standard care.3Given these weak results, there has been less interest in lopinavir-ritonavir as a treatment for COVID-19.
However, these discouraging results may partially be explained by SARS-CoV-2’s viral dynamics. The peak of SARS-CoV-2 viral load is around the time of symptom onset58, but the average time between symptom onset and randomization in this study was 13 days. At this time, the application of antiviral drugs may be too late to have significant effect. Indeed, subgroup analysis showed a minor acceleration of clinical recovery (16.0 days vs. 17.0 days) and a reduction in mortality (19.0% vs. 27.1%) in those treated within 12 days of symptom onset. In addition, this trial did not assess effect at different doses and there is no preliminary data to suggest that the trial dose is effective for COVID-19. Finally, the trial was not blinded and the authors acknowledge the potential for bias in the subjective criteria used3. Ultimately despite the negative results on lopinavir-ritonavir for COVID-19, it is still uncertain whether or not this drug combination is useful for COVID-19.
4) Favipiravir
Favipiravir is a purine nucleic acid analogue that is ribosylated and phosphorylated intracellularly into its active metabolite, which may interfere with viral replication by inhibiting RNA-dependent RNA polymerase. The EC50 of fapiravir against SARS-COV-2 in vitro is 62 μM (9.7 mg/L)36, which is much higher than the EC50 for influenza A and B viruses (0.014~0.55 mg/L). The steady-state peak concentration of fapiravir in human plasma is 62 mg/L (600 mg, BID).59 Its serum protein binding in the concentration range of 0.3~30 mg/L is 54 % and radio-tracer studies in monkeys showed that labeled fapiravir in the lung was approximately 51% of that in plasma at 0.5 h after dosing. Assuming that the fapiravir EC50 concentration in vitro represents unbound drug and that the free concentration of fapiravir in the lung is equal to that in plasma (i.e. about 46%) then the free concentration of fapiravir in lung is about 14.5 mg/L or about 50% greater than the EC50 value.
An open-label, nonrandomized, controlled trial was conducted at the Third People’s Hospital of Shenzhen to evaluate favipiravir in patients with COVID-19 pneumonia.60 Patients with mild or moderate pneumonia within 7 days of onset were selected for this study, and they were more appropriate subjects for antiviral drug trials compared to the severe patients. A total of 80 patients with mild or moderate pneumonia treated within 7 days of symptom onset were assigned to the test group (35) to receive favipiravir (1600 mg twice daily on Day 1 or 600 mg twice daily on Days 2−14 orally) or to the active control group (45), to receive lopinavir-ritonavir (400 mg/100 mg twice daily orally), respectively. In addition, all subjects received aerosolized interferon-α1b (60μg twice daily). Treatment continued until viral clearance was confirmed or for 14 days. The results showed that the favipiravir arm had shorter viral clearance time (median (interquartile range, IQR), 4 (2.5 – 9) d versus 11 (8 – 13) d, P < 0.001) and higher radiographic improvement rate (91.43% versus 62.22%, P = 0.004) compared to the control arm. However, this trial also has some limitations, such as the small number of enrolments and like most trials, it is unblinded, which leads to inevitable selection bias in patient recruitment.
5) Tocilizumab
Cytokine release syndrome (CRS) occurs in many patients with severe COVID-19, which is also an important cause of death. Interleukin-6 (IL-6) plays an important role in CRS.61 As an IL-6 receptor blocker, tocilizumab has been included as an immunotherapeutic agent for severe and critical patients in the ”Novel Coronavirus Pneumonia Diagnosis and Treatment Protocol (Trial Seventh Edition)”. At present, there are several registered clinical trials of tocilizumab in the treatment of patients with COVID-19. Most trials have targeted severe patients, and some trials include an increase of inflammatory factors such as IL-6, CRP or ferritin as one of their inclusion criteria.