key: cord-0906740-gd9rghqa
authors: Nicol, Melanie R.; Joshi, Abhay; Rizk, Matthew L.; Sabato, Philip E.; Savic, Radojka M.; Wesche, David; Zheng, Jenny H.; Cook, Jack
title: Pharmacokinetics and Pharmacological Properties of Chloroquine and Hydroxychloroquine in the Context of COVID‐19 Infection
date: 2020-07-20
journal: Clin Pharmacol Ther
DOI: 10.1002/cpt.1993
sha: c5338f6783207ef6013b82670ef7567988ff8759
doc_id: 906740
cord_uid: gd9rghqa

Chloroquine and hydroxychloroquine are quinoline derivatives used to treat malaria. To date, these medications are not approved for the treatment of viral infections and there are no well‐controlled, prospective, randomized clinical studies or evidence to support their use in patients with Coronavirus Infectious Disease‐2019 (COVID‐19). Nevertheless, chloroquine and hydroxychloroquine are being studied alone or in combination with other agents to assess their effectiveness in the treatment or prophylaxis for COVID‐19. The effective use of any medication involves an understanding of its pharmacokinetics (PK), safety and mechanism of action. This work provides basic clinical pharmacology information relevant for planning and initiating COVID‐19 clinical studies with chloroquine or hydroxychloroquine, summarizes safety data from healthy volunteer studies, and summarizes safety data from Phase 2 and Phase 2/3 clinical studies in patients with uncomplicated malaria, including a Phase 2/3 study in pediatric patients following administration of azithromycin and chloroquine in combination. In addition, this work presents data describing the proposed mechanisms of action against the severe acute respiratory distress syndrome (ARDS) coronavirus–2 (SARS‐CoV‐2) and summarizes clinical efficacy to date.

Since coronavirus disease 2019 , caused by the severe acute respiratory distress syndrome (ARDS) coronavirus-2 (SARS-CoV-2), first emerged in late 2019, there has been a widespread search for effective preventive and treatment options to ameliorate the devastating pandemic. As vaccine availability is likely 12 or more months away, there is an immediate need to identify therapeutic options. Early interest, experimental usage, clinical studies, and emergency use authorizations (EUAs) have focused on several potential therapies including antivirals such as remdesivir or antimalarial agents such as chloroquine (1) and hydroxychloroquine (2) with or without the antibacterial agent azithromycin (3), which has been previously used in the treatment of chloroquine-resistant malaria given reported in vitro synergy (4).

With the recent results from the RECOVERY trial (NCT04381936) as well as the announced suspension of hydroxychloroquine arms in several other trials, there is decreasing support for its use in hospitalized or severely ill patients. However, many trials, particularly in prevention or early treatment settings, or in specific subpopulations, are ongoing. This publication seeks to summarize the current state of knowledge for both chloroquine and hydroxychloroquine, including pharmacokinetic (PK) properties, known pharmacology and evidence related to use in treatment of COVID-19. Prior to the current pandemic, more studies had been reported with use of chloroquine than hydroxychloroquine, thus there is notably more information available for chloroquine regarding metabolism, safety, and overall use.

Both chloroquine and hydroxychloroquine are orally administered 4-aminoquinoline antimalarial compounds. The chemical structures of chloroquine and hydroxychloroquine are shown in Figure 1 .

Chloroquine is administered as chloroquine phosphate and has a molecular weight of 515.9 g/mol (Base: 319.9 g/mol). Hydroxychloroquine is administered as hydroxychloroquine sulfate and has a molecular weight of 433.95 g/mol (Base: 335.9 g/mol).

This article is protected by copyright. All rights reserved Table S1 summarizes limited information on the clinical PK of chloroquine and hydroxychloroquine that is available from ARALEN® (CHLOROQUINE PHOSPHATE, USP) (5) and PLAQUENIL® (HYDROXYCHLOROQUINE SULFATE TABLETS, USP) (6) US product labels.

Information on the PK of pharmacologically active metabolites of chloroquine and hydroxychloroquine is discussed in multiple publications (7-21). A summary of the compiled information is provided below.

The majority of the available PK data in literature are from healthy volunteers receiving a single oral dose containing 300 mg or 600 mg chloroquine base (Table 1) . Cross study comparisons of exposure parameter estimates (Maximum concentration (Cmax), area under the concentration-time curve (AUC), and time to reach Cmax (Tmax)) were not pursued, as different formulations were used across the different studies. Gustafsson et. al. reported the mean (± SD) bioavailability of chloroquine from oral tablet as 89 ± 16% (22). Tmax ranged from 1 to 6 hours across studies and this variability could be due to in vivo PK variability, formulation differences, and/or administration conditions. In addition, despite being in use for years, the available PK data on the effect of food on chloroquine absorption are limited.

Most of the available PK data in literature are from healthy volunteers receiving a single oral dose of 155 mg hydroxychloroquine base (Table 2) . Cross study comparisons of exposure parameter estimates (Cmax, AUC, and Tmax) were not pursued, as the exact information on different formulations that were used across the different studies was not available. Tett et. al. reported the mean (± SD) fraction of the oral dose absorbed as 0.74 (± 0.13) based on the blood and urine PK data 

Estimates reported for apparent volume of distribution (V/F) for chloroquine are relatively large and with notable variation (Table 1) 

This article is protected by copyright. All rights reserved the range of 25-400 ng/mL and (S)-chloroquine binds more to plasma (~67%) than (R)-chloroquine (~49%). Chloroquine is also reported to transfer via placenta and into milk (33, 34).

The reported mean (± SD) blood to plasma hydroxychloroquine concentration ratio is 7.2 (±4.2) (15).

Estimated protein binding is between 30% and 50% and hydroxychloroquine is reported to bind to both albumin and alpha, glycoprotein (35, 36). The findings reported by McLachlan et. al. (35) suggest hydroxychloroquine demonstrates stereoselective protein binding, with (S)hydroxychloroquine binding more to plasma (~64%) than (R)-hydroxychloroquine (~37%), resulting in approximately 52% average protein binding for racemic mixture. Hydroxychloroquine is reported to transfer via placenta and into milk (19, 37-39). (41). The same study noted that at low chloroquine concentrations, CYP2D6 may also play a significant role. Based on literature review, it is proposed that the co-administration of drugs that modulate CYP2C8, CYP3A4, and CYP2D6 could have potential effects on the PK of chloroquine (42). Based on in vitro studies, chloroquine is a weak inhibitor of CYP2D6 but it is unlikely to have a significant effect on the PK of other CYP2D6 substrates in human (42). Additional information on the potential of drug-drug interaction with chloroquine as object and precipitant is detailed by Kiang et. al. (43) (44) (45) . Table 3 presents results of clinical drug interaction studies involving chloroquine as the precipitant and as the object.

This article is protected by copyright. All rights reserved Hydroxychloroquine Similar to chloroquine, information on the metabolism and excretion of hydroxychloroquine is scarce in the literature, including information on specific metabolism pathways. Hydroxychloroquine demonstrates similar PK properties to chloroquine. Hydroxychloroquine's major metabolite is desethylhydroxychloroquine due to its metabolism by Cytochrome P450 enzymes CYP2D6, 2C8, 3A4 and 3A5 (53). Hydroxychloroquine is primarily eliminated through the kidneys. Rainsford, et.

al. discusses the information on hydroxychloroquine metabolism pathway and potential drug interactions from the literature (19). In this review, the most significant drug interactions of relevance noted are with methotrexate. Based on the findings from the two studies (28, 50), Rainsford, et. al.

concluded co-administration of hydroxychloroquine with methotrexate causes reduced Cmax while delaying Tmax and lack of any significant drug interactions related to effects on the PK of hydroxychloroquine by methotrexate. In addition, hydroxychloroquine is considered a weak inhibitor of CYP2D6 and 400 mg daily hydroxychloroquine dose increased metoprolol by 50.7% (54, 55). Table 4 presents results of clinical drug interaction studies involving hydroxychloroquine as the precipitant and the object, respectively.

The estimates for elimination half-life for chloroquine vary significantly (Table 1) . Following multiple doses, chloroquine elimination half-life is reported to range from 30 to 60 days (57, 58).

Krishna et. al. concluded that due to the large V/F (>100 l/kg), distribution rather than elimination processes determine the blood concentration profile of chloroquine in patients with acute malaria (57).

Other potential reasons for high variability in elimination half-life for chloroquine could be differences in PK follow-up duration, improvement in sensitivity of analytical methods over time, and/or different methodologies of calculations. Regarding chloroquine apparent clearance (CL/F), the reported mean estimates ranged from 16 to 40 l/hr (Table 1) . Estimated urinary recovery of chloroquine is reported to be 46% and 55% (20, 22). Augustijns et. al. evaluated whole blood PK of chloroquine enantiomers in humans after a single oral dose of the separate enantiomers and reported Accepted Article that the total body clearance was lower for the (R)-enantiomer (8±2 l/h) than for the (S)-enantiomer (14±4 l/h) (10). The study also reported that terminal half-life and mean residence time (MRT) were longer for (R)-chloroquine (12 days and 16 days, respectively) than for (S)-chloroquine (10 days and 11 days, respectively).

Similar to chloroquine, information on the metabolism and excretion of hydroxychloroquine is scarce in the literature including information on specific metabolism pathways. Fan et. al. reported the mean estimated elimination half-life for hydroxychloroquine to be 11 to 12 days when estimated with hydroxychloroquine plasma concentration over 62 days and 1 to 2 days when estimated with hydroxychloroquine plasma concentration over 72 hours (29). These findings suggest the impact of PK follow-up duration on the reported high variability in elimination half-life for hydroxychloroquine across the studies.

The reported mean amount of hydroxychloroquine excreted unchanged in the urine ranges between 23-27% (15, 16, 69). McLachlan et. al. evaluated renal clearance of the hydroxychloroquine enantiomers and reported that (S)-hydroxychloroquine had a mean (± SD) renal clearance from blood of 41 ± 11 ml min, approximately twice that of (R)-hydroxychloroquine (30). Ducharme et. al. notes similar findings with the reported total urinary excretion of (S)-hydroxychloroquine to be higher than that of (R)-hydroxychloroquine (17). In the same study, the estimated elimination half-life of (S)hydroxychloroquine (19 ± 5 days) was significantly shorter than that of (R)-hydroxychloroquine (22 ± 6 days), partly due to its faster urinary excretion and hepatic metabolism. Its renal clearance was twice that of (R)-hydroxychloroquine (4.61 ± 4.01 vs 1.79 ± 1.30 l/h).

With respect to the use of population PK (POP-PK) analysis approach, three studies for chloroquine (60-62) and four studies for hydroxychloroquine were identified during the literature search (18, 63-65). These studies are summarized in Table 5 and key information is summarized below. It is noteworthy that the intensity of the PK sampling, sensitivity of bio-analytical method used and/or sampling duration may affect the PK model selection process. In addition, the lack of variability in

This article is protected by copyright. All rights reserved intrinsic and extrinsic factors within the patient/subject population, whose PK data are being utilized for POP-PK analysis, may affect the ability of identifying relationship between covariates and PK. Obua et. al. reports that two-compartmental PK model best described the chloroquine pharmacokinetics based on a POP-PK analysis that utilized sparse PK data obtained from finger prick sampling in pediatric malaria patients (59). The same study reports that no correlation was identified between body weight or age with the PK model parameters. Höglund et al. reports that for adult malaria patients, the disposition of chloroquine was also adequately described by the two-compartment model without the use of any covariates on the model parameters (61). POP-PK approach was also used to assess differences in the PK of chloroquine in pregnant and non-pregnant Balevic et. al. reports the use of POP-PK analysis to assess the effect of pregnancy on PK of hydroxychloroquine in patients with Rheumatic Diseases and notes lack of significant changes hydroxychloroquine exposure due to pregnancy in small cohort of patients (64). The study reports that a one-compartment PK model best described hydroxychloroquine PK and body weight was found to be a significant covariate on V/F. Within the POP-PK analyses literature on hydroxychloroquine

This article is protected by copyright. All rights reserved discussed above, the reported BSV (CV%) in CL/F estimates ranged from 16-44% (RSE: 4-41%).

For V/F estimates for central compartment,BSV (CV%) ranged from 16-23% (RSE: ~58%). 

This article is protected by copyright. All rights reserved

The PBPK models and simulations for hydroxychloroquine and chloroquine reiterated prior knowledge that blood concentrations for both exceed that of plasma. More importantly, the full PBPK models indicated that hydroxychloroquine and chloroquine both have a predilection for lung. The model-predicted lung concentrations exceed plasma concentrations at steady state by more than 400fold. In addition, if anti-inflammatory effects and prevention of cytokine storm are instrumental in treating disease, then drug concentrations in the lung may provide beneficial effect.

In addition, PBPK modeling and simulation have been utilized to inform dosing decisions for COVID-19 clinical trials. Ultimately, as efficacy and safety data from multiple COVID-19 prevention and treatment trials accrue, a more precise estimation across studies will help inform on effective doses and exposures over time.

In both chloroquine and hydroxychloroquine use, gastrointestinal discomfort is the most commonly described side effect manifesting primarily as nausea, vomiting, diarrhea and stomach pain. These toxicities are more common at higher doses (69). Hypoglycemia may occur and can be enhanced in patients receiving concurrent hypoglycemic drugs. Rash and itching are well described and due to the long half-lives, they may last for prolonged periods after discontinuation. Hemolytic anemia is a serious adverse effect in individuals with glucose-6-phosphate dehydrogenase (G6PD) deficiency who receive chloroquine or hydroxychloroquine (5, 6). Retinal toxicity is a well-described adverse effect for hydroxychloroquine, however it is primarily a concern with chronic and cumulative doses of >1000 grams (70).

The effects of chloroquine alone as well as chloroquine and azithromycin on QTc prolongation were investigated in a previously unpublished, single center open-label, placebo-controlled, randomized, multiple-dose healthy volunteer study (N=119 total; 24 per group except for the chloroquine+1500mg Az group where 23 subjects were enrolled. Three subjects withdrew from the chloroquine+1500 mg Az group during the treatment phase of the study due to adverse events (1 due to diarrhea, 1 due to loss of appetite, and 1 due to nausea, diarrhea, and vomiting) and thus were not included in the

This article is protected by copyright. All rights reserved pharmacokinetic or QTc analyses). A parallel group design was utilized given the long half-lives of both chloroquine and azithromycin. Subjects were assigned to one of the following five treatment groups:  Placebo QD x 3 days  1000 mg chloroquine phosphate (600 mg base) QD x 3 days  1000 mg chloroquine phosphate plus 500 mg azithromycin QD x 3 days  1000 mg chloroquine phosphate plus 1000 mg azithromycin QD x 3 days  1000 mg chloroquine phosphate plus 1500 mg azithromycin QD x 3 days Triplicate 12-lead ECGs were serially obtained on Day -1 (baseline), Day 1, Day 2 and Day 3 including at 0, 1, 2, 3, 4, 5, 6, 8, 10 and 12 hours on Day 3. Serial plasma samples were obtained for chloroquine and azithromycin quantitation on Days 1-3 including at times immediately following ECG assessments on Day 3. QT interval data were corrected for heart rate using Fridericia's method (QTcF). In all analyses, the means of the ECG parameters within a triplicate were used as the It should be noted that the high chloroquine dose (600 mg twice daily for 10 days or total dose 12g) of a recent phase IIb trial of patients with SARS in Manaus, Brazilian Amazon was stopped because of

This article is protected by copyright. All rights reserved safety. There was a trend for higher lethality compared with the lower dose and one quarter of the high dose patients developed QTc>500 msec (72). All patients in this study also received 5 days of azithromycin, also a QTc prolonging drug.

The effect of hydroxychloroquine on QTc is less well studied than the effect of chloroquine. Case reports have noted prolonged QTc values with hydroxychloroquine toxicity (73-78). In a study of 85 patients with connective tissue disorders who were receiving hydroxychloroquine, the investigators concluded that QT intervals did not appear to be different from what would be considered normal in these patients (79).

Recently there are emerging QT data from 84 patients with SARS-CoV-2 infection treated with a combination of hydroxychloroquine and azithromycin. QTc values were determined for subjects at baseline (before administering hydroxychloroquine and azithromycin) and while on drug, but no reference was given regarding doses administered, the timing of QTc assessment with respect to dose nor the formula used to correct QT for heart rate. QTc values were prolonged maximally (28 msec) from baseline between Days 3 and 4. In 30% of patients, QTc increased by greater than 40 ms. In 11% of patients, QTc increased to >500 ms, representing high risk group for arrhythmia. Development of acute renal failure but not baseline QTc was a strong predictor of greater QTc prolongation (80).

No dose adjustment is suggested by the manufacturer for patients with renal impairment. No dose adjustment is suggested by the manufacturer for those with hepatic impairment; however, caution should be used in patients with hepatic disease and/or alcoholism, as the drug concentrates largely in the liver. Table S1 summarizes the language provided in both the chloroquine and hydroxychloroquine labels regarding both renal and hepatic impairment.

This article is protected by copyright. All rights reserved

No dose adjustment is suggested by the manufacturer for patients with renal impairment. Jallouli, et.

hydroxychloroquine blood concentrations pre-and post-dialysis. Patients received either 200 mg/day (N=2) or 400 mg/day(N=1) of hydroxychloroquine, and blood concentrations did not change significantly pre-or post-dialysis (81). Hydroxychloroquine was not detected in the dialysis bath of all three patients (<50 ng/mL). This suggests hydroxychloroquine is not dialyzable. No dose adjustment is suggested by the manufacturer for those with hepatic impairment; however, caution should be used in patients with hepatic disease and/or alcoholism, as the drug concentrates largely in the liver. Table S1 summarizes the language provided in both the chloroquine and hydroxychloroquine labels regarding both renal and hepatic impairment.

Geriatric Population: Typical adult dosing is recommended for patients greater than 65 years of age.

Hydroxychloroquine and chloroquine clinical trials did not include a sufficient number of geriatric subjects to determine whether they respond differently than younger subjects. However, given both drugs are significantly renally cleared and elderly patients are more likely to have decreased renal function, consideration should be given for dose reduction in this population.

Pediatric Population: Both hydroxychloroquine and chloroquine are preferably dosed based on actual body weight in the pediatric population. The American Academy of Ophthalmology recommends keeping daily doses ≤6.5 mg/kg hydroxychloroquine-sulfate (≤5.0 mg/kg base) and ≤3.8 mg/kg chloroquine-phosphate (≤2.3 mg/kg base) for all patients to reduce the risk of retinopathy and permanent vision loss, though this risk is linked to duration of use over years (82). Approved pediatric doses are above this threshold for treatment of acute uncomplicated malaria.

In infants, children and adolescents, the highest approved dose of chloroquine is for treatment of acute malarial attack. The highest recommended chloroquine dose for the treatment of acute, uncomplicated malarial attack is 10 mg base/kg PO followed by 5 mg base/kg at 6, 24, and 48 hours after initial dose for a total of 4 doses (83). Doses for chemoprophylaxis are suggested to be lower

This article is protected by copyright. All rights reserved (84). Several fatalities have been reported due to accidental ingestion at even lower doses (0.75-1 g chloroquine-phosphate in one 3-year-old child) (5). The injection label notes that in no instance should a single intramuscular or subcutaneous dose exceed 6.25 mg (5 mg base) per kg of body weight, since children are especially sensitive to the effects of the 4-aminoquinolines. Severe reactions and sudden death have been reported following parenteral administration in children (5, 85).

In infants, children and adolescents, 10 mg base/kg followed by 5 mg base/kg hydroxychloroquine at 6, 24 and 48 hours after initial dose is the highest recommended dose for the treatment of acute uncomplicated malarial attack (83). Doses for chemoprophylaxis, juvenile rheumatoid arthritis and systemic lupus erythematosus are suggested to be lower (84).

Based on WHO Guidelines for the treatment of malaria, chloroquine is considered safe in the first trimester of pregnancy and has been studied in combination with azithromycin for intermittent preventive treatment in pregnancy (IPTp) as a fixed dose combination chloroquine and metabolites 

This article is protected by copyright. All rights reserved

American Academy of Pediatrics (AAP) considers use of hydroxychloroquine usually compatible with breastfeeding and hydroxychloroquine is considered appropriate treatment for SLE during pregnancy by the American College of Rheumatology (91, 92) . Unless pregnancy is planned, fetal exposure cannot be avoided by discontinuing hydroxychloroquine at the time pregnancy is discovered, as hydroxychloroquine is stored in the liver with an extended half-life.

Hydroxychloroquine can be detected in cord-blood in concentrations similar to those in post-partum maternal serum (93) and crosses the placenta. One study saw no differences when comparing patients with systemic lupus erythematosus with no hydroxychloroquine exposure (N=163), continuous hydroxychloroquine use (N=56) and those discontinuing hydroxychloroquine during the first trimester when assessing (N=38) for miscarriages, stillbirths, pregnancy losses and congenital abnormalities (94) . Another study by Costedoat-Chalumeau et. al. demonstrated no difference in growth rate and no evidence of visual, hearing or developmental abnormalities in the mean 26-month follow-up of children between hydroxychloroquine and control group (95) . Hydroxychloroquine has been shown to be transferred to human breast milk. Breastfed infants are exposed to ~2% of maternal dose on a body weight basis (37).

Considering the adverse event profile and elimination pathway of both hydroxychloroquine and chloroquine, strong consideration to alternative therapies should be made in patients exhibiting these comorbidities. Arentz et al. recently published a case series report discussing characteristics and outcomes of 21 critically ill patients with COVID-19 in Washington State. Of the 21 patients with PCR confirmed SARS-CoV-2 infection by nasopharyngeal sample, 18 were identified to have comorbidities. Chronic kidney disease (N=10) and congestive heart failure (N=9) were the most common comorbidities (96) .

The presumed mechanism of antiviral activity of chloroquine and hydroxychloroquine is increasing the pH of the endosome that is required for viral/cellular fusion. Using flow cytometry and sorting of

This article is protected by copyright. All rights reserved SARS--CoV spike glycoprotein and ACE2 on Vero E6 cells, CDC investigators also found that chloroquine impairs the glycosylation of ACE2, thereby impairing the binding of the virus to its cellular receptor (97) . To date, there are no data on stereospecific antiviral effects of chloroquine or hydroxychloroquine.

In addition to direct antiviral activity, chloroquine and hydroxychloroquine have known immunomodulatory and anti-inflammatory effects. The anti-inflammatory effects of hydroxychloroquine were first discovered through the serendipitous observation that WWII soldiers receiving chloroquine for malaria prophylaxis noted improvements in skin rashes and arthritis, ultimately leading to the development of this drug as a treatment for systemic lupus erythematosus and other immunologic conditions such as rheumatoid arthritis (98, 99) . Immunomodulatory effects of chloroquine as well as hydroxychloroquine include inhibition of antigen presentation to dendritic cells, reduced cytokine production in macrophages and reduced signaling of both B and T cells.

Therefore, in addition to direct antiviral effects, chloroquine and hydroxychloroquine may play a role in reducing the cytokine storm associated with COVID-19 progression to ARDS. Although these drugs are generally considered more immunomodulatory than immunosuppressive, it is worth pointing out that in early infection, the role of the innate immune system is a critical factor in preventing COVID-19 progression to serious disease. As our understanding of COVID-19 pathogenesis evolves, whether these immunomodulatory effects are beneficial or detrimental will need to be elucidated. 

This article is protected by copyright. All rights reserved chloroquine concentrations as little as 0.1 µM (32 ng/mL) reduced viral spread by 50% while chloroquine concentrations of 100 µM (32,000 ng/mL) reduced viral spread by up to 94% (97) . The half-maximal inhibitory effect was estimated to occur at 4.4 µM +/1.0 µM (1408 +/320 ng/mL).

Pretreatment with chloroquine was found to have an even more potent viral inhibition with 0.1, 1 and 10 µM (32,320, and 3200 ng/mL), reducing infectivity by 28, 53 and 100% respectively, which may suggest lower drug exposures required for prophylaxis than treatment.

Following the COVID-19 outbreak, Wang et al tested the in vitro inhibitory activity and cytotoxicity of a number of potential therapeutics against SARS-CoV-2, including chloroquine in Vero E6 cells (100) . Of the seven drugs tested, chloroquine (along with remdesivir) had one of the highest selectivity indexes with an EC 50 of 1.13 µM (361 ng/mL) and CC 50 >100 µM (>32,000 ng/mL). (Table 7) Hydroxychloroquine Similar to chloroquine, hydroxychloroquine has also been tested in a number of in vitro experiments for activity against SARS-CoV-2. A wide range of EC 50 values have been reported under various conditions (Table 7) , with some comparative studies finding greater potency with hydroxychloroquine

(2) and some less potent (101). Below is a summary of the various experiments, including experimental characteristics and reported EC 50 values is shown in Table 7 .

For clinical use in COVID-19, the CDC provides anecdotal dosing suggestions, but explicitly states that optimal dosing and duration of hydroxychloroquine for COVID-19 are unknown (103).

Translational pharmacokinetic/pharmacodynamic modeling is one approach to propose optimized hydroxychloroquine dosing regimens which ensure the highest likelihood of success as COVID-19 treatment. With lack of known exposure correlates for efficacy to date, several groups have attempted to integrate available pharmacological data and mechanistic knowledge related to COVID-19, including in vitro data for SARS-CoV-2, historical data on population pharmacokinetics, safety data of hydroxychloroquine from large patient cohorts, and newly emerging clinical PK/PD data from patients with COVID-19 (2, 104-106). Early characterization of the clinical exposure-response

This article is protected by copyright. All rights reserved relationship between hydroxychloroquine and SARS-CoV-2 viral load, suggests that the hydroxychloroquine doses needed to cause more rapid viral clearance compared to standard of care could be much higher than those currently being studied for COVID-19 patients in any setting and these doses may lead to substantial risk of cardiac toxicity (104-106). As these investigations have pointed out, when making in vitro-in vivo extrapolations to optimize dosing, there are many considerations to be made including 1) the compartment to be targeted (e.g. plasma, intracellular vs extracellular lung tissue, lung fluid, other organs), 2) free vs total concentrations, 3) plasma vs. serum vs whole blood, 4) EC 50 vs EC 90 or magnitude above these targets.

Initial support for the use of chloroquine and hydroxychloroquine came from a published report which states that results from 100 patients across multiple institutions found that for the treatment of COVID-19, chloroquine was superior to standard of care in preventing exacerbation of pneumonia, reducing days to conversion rate and shortening time to clinical recovery (1). To date, this study has not been published and data regarding magnitude of benefit, specific dosing or timing of treatment initiation, observed toxicities, and treatments received by the control group are not available. Other data with chloroquine, as mentioned above, is from one 81-person trial in Brazile compared 600 mg chloroquine base twice daily for 10 days (12 g total) to 450 mg x2 on day 1, then 450 mg daily for 4 additional days (2.7 g total). This study did not have a placebo arm and was halted prior to reaching enrollment goals because higher mortality was observed in the high dose chloroquine arm (72). One additional small study (n=22) showed no difference between chloroquine and lopinavir/ritonavir for viral clearance (107).

Clinical studies of hydroxychloroquine are heterogeneous in terms of hydroxychloroquine regimen, outcomes measurements (clinical, SARS-CoV-2 viral load), and severity of illness (108, 109). Only one study non-randomized study has reported a significant reduction in time to viral clearance among patients with mild disease who received hydroxychloroquine or hydroxychloroquine with azithromycin, and this small study had significant limitations including a control arm with significantly higher viral loads compared to those who received treatment (110). The largest well characterized observational cohorts of hospitalized patients, hydroxychloroquine with or without azithromycin were not associated with a mortality benefit (111-112).

This article is protected by copyright. All rights reserved Only 2 randomized clinical trials have been published in hospitalized COVID-19 patients, one of which found that high-dose hydroxychloroquine (1200 mg/day for three days followed by 800 mg/day for 2-3 weeks) given to mild or moderately ill patients did not improve time to negative PCR test nor time to alleviation of symptoms (113). A small randomized study in China reported no apparent clinical benefit of 400 mg daily for 5 days compared to placebo in COVID-19 patients, the majority of whom had mild disease (114). Recently, the RECOVERY trial (NCT04381936) stopped enrollment into the hydroxychloroquine arm of their trial in hospitalized COVID-19 patients, reporting no difference in mortality, hospital stay, or other outcomes, between 1542 subjects on hydroxychloroquine and 3132 subjects on standard care. Subsequently, the WHO-sponsored SOLIDARITY trial and an NIH-sponsored trial have suspended enrollment of their hydroxychloroquine arms, reporting no benefit observed although the data have not yet been released.

Hydroxychloroquine has also been proposed for use in prophylactic settings. The first randomized clinical trial published in an outpatient setting found that high dose hydroxychloroquine (800 mg loading dose followed by 600 mg/day for 4 days) given within 4 days of a high-risk exposure did not reduce likelihood of acquiring COVID-19 (115). Studies in high risk health care workers evaluating a role for pre-exposure prophylaxis are currently underway.

In summary, to date, there is a lack of evidence from controlled, randomized clinical trials powered for efficacy endpoints (the gold standard for evidence-based medicine) for use of chloroquine or hydroxychloroquine to treat COVID-19. Increasingly, evidence points to a lack of benefit in hospitalized and severely ill patients; whether a role exists in early treatment or prevention remains to 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved Table S1 Accepted Article

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved Precipitant term refers to the drug that causes an effect on the substrate drug by inhibiting or inducing enzymes

Object term refers to the drug whose exposure may or may not be changed by a precipitant drug Accepted Article 

This article is protected by copyright. All rights reserved 

This article is protected by copyright. All rights reserved 

Transfer of chloroquine and desethylchloroquine across the placenta and into milk in Melanesian mothers

Chloroquine and its metabolites in human cord blood, neonatal blood, and urine after maternal medication

Health Organization Guidelines for the treatment of malaria

Excretion of chloroquine and desethylchloroquine in human milk

Transfer of drugs and other chemicals into human milk

American College of Rheumatology Guideline for the Management of Reproductive Health in Rheumatic and Musculoskeletal Diseases

Evidence of transplacental passage of hydroxychloroquine in humans

Hydroxychloroquine in lupus pregnancy

Safety of hydroxychloroquine in pregnant patients with connective tissue diseases: a study of one hundred thirty-three cases compared with a control group

Characteristics and Outcomes of 21 Critically Ill Patients With COVID-19 in Washington State

Chloroquine is a potent inhibitor of SARS coronavirus infection and spread

Chloroquine analogues in drug discovery: new directions of uses, mechanisms of actions and toxic manifestations from malaria to multifarious diseases

Hydroxychloroquine: from malaria to autoimmunity

Remdesivir and chloroquine effectively inhibit the recently emerged novel coronavirus (2019-nCoV) in vitro

Accepted Article