Repurposing Drugs for COVID-19: Pharmacokinetics and Pharmacogenomics of Chloroquine and Hydroxychloroquine

Introduction

SARS-CoV-2 is a new coronavirus type that has not been previously identified in humans. Little is known about the highly infectious virus or how to combat it. The current strategy considers two broad categories of therapies: antivirals, which may target the coronavirus directly, and host modifiers and immune-based medications, which may influence the immune response to the virus. Currently, all the therapeutic agents are repurposed medications.

For approximately 6000 identified medical conditions, only 500 have approved therapies; a critical need currently exists for the availability of drug therapies.^1–3 Drug repurposing helps to minimize the deficiency and delivers a candidate at a shorter development time and a lesser cost.⁴ A key advantage of repurposed drugs is that safety has been established and only efficacy of the new indication needs to be assessed.⁵ Many of the well-known repurposed drugs, including sildenafil, minoxidil and aspirin emerged by chance from “unorganized” drug discovery processes.⁶ Several diverse disease states have common transcriptional inflammatory and metabolic pathways, suggesting that drugs designed for treatment of one disease can potentially be used to treat other diseases.⁷ Drug repurposing is being applied to finding a therapeutic approach for the COVID-19 pandemic. Thirty-one potential broad-spectrum antiviral agents (BSAAs) were recently identified as having potential for treating SARS-CoV-2/COVID-19.⁸ Several existing BSAAs have been initiated into clinical trials (Table 1).

Table 1 Broad-Spectrum Antiviral Agent Candidates in Clinical Trials

Two of the drugs listed in Table 1, hydroxychloroquine and chloroquine, have been proposed as treatments for COVID-19.

Chloroquine (Aralen^®) and hydroxychloroquine (Plaquenil^®) are 4-aminoquinoline medications used to treat several disease states. Chloroquine (CQ) was first developed for the treatment of malaria.⁹ Hydroxychloroquine (HCQ) is β-hydroxylated analogue of CQ.¹⁰ Both medications have been successfully used to treat extraintestinal amebiasis, several infectious (HIV, Q fever, Zika virus, fungal infections) and rheumatological (systemic lupus erythematosus, antiphospholipid syndrome, rheumatoid arthritis, Sjögren’s syndrome) diseases.^11–13

In the therapy of malaria, the agents inhibit the action of heme polymerase, which causes the buildup of toxic heme in Plasmodium species.^14,15 The antiviral activity is not fully understood. The drugs accumulate in human organelles, raise the endosomal pH and prevent viral activity.^16–18 The elevated pH inhibits nucleic acid replication, glycosylation of viral proteins, viral fusion and entry into the cell, viral assembly and release.¹⁹

Adverse Reactions of Chloroquine and Hydroxychloroquine

Postmarketing cases of life-threatening and even fatal events have been reported for chloroquine and hydroxychloroquine. An overdose of CQ can cause acute poisoning and death.²⁰ HCQ was demonstrated to be 40% less toxic than chloroquine, although prolonged and overdose administration can still cause poisoning.^21,22 Patients may present with atrioventricular block, pulmonary hypertension, sick sinus syndrome or with cardiac complications. The most life-threatening adverse reaction is QTc prolongation with subsequent risk of ventricular arrhythmias.²³ Concomitant QTc-prolonging medications may increase the severity of the complication even more.²⁴ The mechanism of QTc prolongation by chloroquine and hydroxychloroquine is unknown, largely depending on the cardio-vascular health of the patients.^25–27 Another complication of the two medications is retinopathy (chloroquine retinopathy), which can result in irreversible impairment of the retina.^28,29 High concentrations of the drugs in the retina, due to binding to retinal melanin, result in the damage of the tissue.³⁰ Hydroxychloroquine may also produce a severe cutaneous adverse effect such as a generalized pustular figurate erythema (GPFE).¹²⁶ High concentrations of the medications in the skin and very slow cutaneous elimination (longer than 6 months) may result in this severe cutaneous reaction.^59,87 The most common adverse effects of the drugs are nausea, vomiting, diarrhea.³¹ Other complications include hypoglycemia in diabetics, hemolytic anemia in G6PD deficiency patients, tinnitus and headache.^32,33

Chloroquine and Hydroxychloroquine for COVID-19 Therapy

Chloroquine and hydroxychloroquine have shown the ability to inhibit replication of multiple coronaviruses in vitro,^34–36 including SARS-CoV-2 in concentration-dependent manner.^{17,19,22,37,38} The anti-SARS-CoV-2 activity of HCQ seems to be less potent compared to CQ. The EC₅₀ for CQ (2.71 μM) was significantly lower than that of HCQ (4.51μM).²² In contrast, hydroxychloroquine was found more potent than chloroquine against SARS-CoV2 when given post-infection and prophylactically.¹⁹

Clinical evidence of the effectiveness of HCQ or CQ for the treatment of COVID-19 is limited. Some small clinical trials have shown therapeutic benefits of the drugs, while others have shown the opposite. In recent clinical trials, over 100 people with COVID-19 have been treated with chloroquine. These patients had less severe disease and a shorter illness duration compared to those who did not receive chloroquine.³⁹ Another open-label non-randomized clinical trial with 36 COVID-19 patients demonstrated that hydroxychloroquine treatment resulted in viral load reduction/disappearance in the patients. The effect was reinforced by azithromycin.⁴⁰ Contrasting results were reported in a small study with 11 hospitalized patients; no difference in clinical outcomes was observed between patients treated with HCQ and azithromycin and patients on standard care.⁴¹ In a randomized trial with 62 hospitalized patients, patients on HCQ had a more substantial proportion of clinical improvement of pneumonia (80% vs 55%) than patients with standard care.⁴² In another clinical trial with 368 COVID-patients, an increased overall mortality was observed in the patients treated with hydroxychloroquine.⁴³ More clinical trials are going on.

The FDA has issued an emergency use authorization for CQ and HCQ to treat COVID-2019 infection, allowing the unapproved use of these medications in light of a public health emergency.^32,33 On April 24, the FDA issued warning against HCQ or CQ unless the therapy is closely supervised by a healthcare professional.⁴⁴ The caution was initiated after the agency received reports of serious adverse effects in COVID-19 patients. These findings do not apply to the use or evaluation of hydroxychloroquine in pre- or post-exposure prophylaxis in patients exposed to COVID-19.

These results bring forward the need for large controlled clinical trials to provide guidance on safe and effective dosing of CQ/HCQ for COVID-19 therapy. Furthermore, response to drugs is subject to inter-individual variability. Patients treated with the same dose of the same drug, may exhibit lack of efficacy, or adverse reactions. The variability, at least in part, is attributed to genetic polymorphisms. Knowledge of pharmacogenomic (PGx) of the drugs is necessary to estimate effective and safe dosing and to avoid/minimize adverse reactions. No PGx studies have been conducted to investigate the inter-patient variability of CQ and HCQ in COVID-19 patients.

The purpose of the study was to highlight the importance of large, randomized, controlled clinical trials and pharmacogenomic studies to assess the optimal dosing of CQ and HCQ in diverse populations as a treatment for COVID-19.

Methods

A review of the literature was performed using Medline/PubMed/Embase database resources for English language papers from 1947 up to July 2020 to identify appropriate articles that addressed the objectives of this review. A variety of keywords were employed in keyword/title/abstract searches that included: chloroquine, CQ, hydroxychloroquine, HCQ, pharmacokinetics, pharmacogenomics, COVID-19, SARS-CoV-2. We obtained 126 appropriate results after screening all sources; relevant and non-relevant. The publications were reviewed independently by two investigators. The investigators extracted the data and inspected each reference identified by the search. In cases where the same studies were reported in more than one publication, the study’s results were accounted for only once. Limits to the search strategy were English language articles and human studies. The electronic search was followed by extensive hand searching using reference lists from the identified articles. The search method was used to strengthen existing concepts and to identify study designs for upcoming research studies.

Results

Pharmacokinetics of Chloroquine and Hydroxychloroquine

This paper presents the current knowledge on chloroquine (CQ) and hydroxychloroquine (HCQ) pharmacokinetics (PK) with a focus on stereoselectivity of their disposition. Both drugs are racemic mixtures, consisting of equal amounts of R(-) and S(+)-enantiomers.⁴⁵ The pharmacokinetics of these two 4-aminoquinolines are similar and regulate dosing of the drugs.

Dosing Considerations

CQ is available as chloroquine phosphate, Aralen^®. Each 500-mg tablet of chloroquine phosphate contains 300 mg of chloroquine. HCQ is available as hydroxychloroquine sulfate, Plaquenil^®. Each 200-mg tablet of hydroxychloroquine sulfate contains 155 mg of hydroxychloroquine. Doses of both agents are based on ideal body weight (IBW). Chloroquine doses are 3.5–4.0 mg/kg/day and produce plasma levels of 6 to 9 × 10⁻⁷ M/L. Doses of hydroxychloroquine are 6.0–6.5 mg/kg/day and produce plasma concentrations of 1.4 to 1.5 × 10⁻⁶ M/L.⁴⁶ An initial adult dose of chloroquine phosphate for malaria therapy is 1 g followed by 500 mg given at 6–8 hours, 24, and 48 hours. Lupus erythematosus and rheumatoid arthritis chloroquine phosphate dosage is 250 mg daily with dose reduction after remission.⁹

Initial adult dose of hydroxychloroquine sulfate is 800 mg followed by 400 mg given at 6–8 hours, 24, and 48 hours. Lupus erythematosus and rheumatoid arthritis initial treatment is 400–600 mg of hydroxychloroquine sulfate daily for several weeks or months.¹⁰

Pediatric dosage of chloroquine and hydroxychloroquine is based on body weight. An initial pediatric dose of chloroquine phosphate for treatment of malaria is 16.7 mg/kg followed by 8.3 mg/kg given at 6, 24, and 48 hours after initial dose. Maximum total dose is 2.5 g.⁹ Initial pediatric dose of hydroxychloroquine sulfate for treatment of malaria is 12.9 mg/kg followed by 6.4 mg/kg given at 6, 24, and 48 hours after the first dose. Maximum total dose is 2 g.¹⁰

CQ and HCQ have been used for prophylaxis and treatment of malaria in pregnant women without evidence of adverse effects on the fetus. Dosing for treatment and prophylaxis of uncomplicated malaria is the same in pregnant and nonpregnant adults. Due to pregnancy-induced physiologic changes, some pharmacokinetic properties of the drugs may be altered, suggesting dose adjustments may be needed. But data are not sufficient to determine an appropriate dosing during pregnancy.^9,10

Small amounts of chloroquine and hydroxychloroquine excrete into breast milk. The amounts of the drugs are not sufficient to harm the infant nor to protect the child from malaria. Weekly CQ/HCQ of 500/400 mg may be given until breastfeeding is completed.^9,10

No information is available on the effect of chloroquine and hydroxychloroquine in geriatric patients. But because CQ and HCQ are mostly excreted in the urine, elderly patients with age-related kidney problems may require caution and a dose adjustment for the patients. The dose adjustment should be based on the kidney function.

The optimal dosing of HCQ and CQ for treatment of COVID-19 is unknown. Most of the published clinical studies had HCQ dosage of 400 mg/5 days or 800 mg on the first day and 400 mg for the next 4 days. The latest regimen was supported by pharmacokinetic modelling, where an oral HCQ sulfate loading dose of 400 mg twice daily, followed by a maintenance dose of 200 mg twice daily for 4 days was able to achieve treatment efficacy and a good safety profile.¹⁹ This regimen reached three times the potency of CQ phosphate given 500 mg twice daily for 5 days.¹⁹ However, more reliable information is required before it can be widely used to treat COVID-19.

Absorption

Oral absorption of chloroquine and hydroxychloroquine in humans is efficient. Both drugs have oral bioavailability of 0.7–0.8.^47,48 Although 2-3-fold difference in the absorbed fraction of oral doses was reported.^49,50 Maximum blood concentrations (Cmax) for the oral doses showed significant differences between subjects (range 135–422 ng/mL), but not within subjects.^51,52 Oral bioavailability of chloroquine is 52–114%. CQ oral tablets have slightly greater bioavailability than oral solutions, 67–114% vs 52–102%, respectively.¹⁶ Hydroxychloroquine has oral bioavailability of 67–74%.^47,53 Antacids may decrease the bioavailability of both drugs.⁵⁴ Oral chloroquine reaches Cmax faster than hydroxychloroquine. Time to reach the maximum level (Tmax) for CQ was estimated at 30 minutes,^16,55–57 while Tmax for HCQ was estimated 3.74 hours.^47,58 Absorption of the R and S enantiomers was not significantly different.⁴⁷

Distribution

CQ and HCQ have multicompartment disposition in humans with wide distribution to the body tissues.⁴⁸ Highest concentrations were found in the melanin-containing cells, the retina and the skin. High levels were observed in the liver, spleen, kidney, and lung.⁵⁹ In the blood, concentrations in erythrocytes were up to 5 times higher than in plasma.⁶⁰ Reported volumes of distribution were 44,000 L and 65,000 L for HCQ and CQ, respectively.^{16,57,61–63}

Plasma protein binding of the drugs ranges between 50% and 60%.^47,64 CQ and HCQ are mostly bound to two plasma proteins, albumins and α-1-acid glycoproteins. Binding of both compounds to plasma proteins is stereoselective. Chloroquine is approximately 60% bound to plasma proteins.^16,65 Extend of S(+)-chloroquine plasma protein binding is greater than binding of R(-)-chloroquine (67% vs 35%).⁶⁶ Binding of hydroxychloroquine to plasma proteins is around 50%, which is less than chloroquine binding. The S-hydroxychloroquine is 64% bound to plasma proteins, while the R-hydroxychloroquine is only 37% protein bound.⁴⁷ Following separate administration of the individual enantiomers of both drugs, R(-)-isomers reach higher and more sustained plasma and ocular concentrations than S(+)-forms.^16,67,68

Metabolism

Chloroquine and hydroxychloroquine have long half-lives and low blood clearance. CQ is rapidly N-desethylated into two major metabolites: desethylchloroquine (40%) and bisdesethylchloroquine (10%).^69,70 Desethylchloroquine is the pharmacologically active metabolite and further metabolizes to bidesethylchloroquine. CQ is metabolized primarily by CYP2C8 and CYP3A4 mediating 80% of the total metabolism of the drug. Other enzymes, CYP3A5 and CYP2D6 break down chloroquine to a lesser extent.^69,71–73 Metabolism of CQ is stereoselective. After administration of the individual enantiomers, the concentration of (R)-chloroquine was 1.3-fold higher compared to concentrations of (S)-chloroquine in plasma and 1.8-fold higher in the blood in patients with rheumatoid arthritis.⁷⁴ Blood concentrations of the active metabolite S(+)-desethylchloroquine exceeded those of the R(-)-forms.^67,75,76

HCQ has similar to chloroquine biotransformation but breaks down into more metabolites. HCQ is N-dealkylated by CYP3A4 to the two active metabolites desethylhydroxychloroquine, desethylchloroquine and an inactive metabolite bidesethylchloroquine. Other cytochrome P450 enzymes (CYP2C8, 2D6, and 3A5) are involved in the metabolism to a lesser extent.^64,77 Biotransformation of HCQ is also stereoselective. Several studies have reported faster hepatic metabolism of S(+)-enantiomers compared to metabolism of R(-)-enantiomers.^{67,76,78–80} The blood and plasma concentrations of R-hydroxychloroquine exceeded those of the S-hydroxychloroquine with the mean R/S ratio of 2.2 in the blood and 1.6 in the plasma.⁸¹ The mean blood concentration ratio R/S for desethylhydroxychloroquine was 0.45 and for desethylchloroquine was 0.56, indicating stereoselective metabolism of the compound.⁸¹

Similar doses of the two drugs produced 11-fold variations in the blood concentrations in patients with rheumatoid arthritis^47,63,82,83 and in healthy volunteers,^52,64 suggesting different extend of metabolism among individuals. Moreover, chloroquine and hydroxychloroquine are involved in several metabolic drug–drug interactions (DDIs). A CYP3A4 inhibitor, cimetidine increased serum concentrations of CQ by 48%.⁸⁴ Another CYP3A4 inhibitor, ketoconazole reduced the formation of active metabolite desethylchloroquine.^71,85

Excretion

Urinary excretion is the main route of elimination for chloroquine and hydroxychloroquine.

The 50% of a chloroquine dose is recovered in the urine as unchanged drug, with 10% of the dose recovered in the urine as its active metabolite desethylchloroquine.¹⁶ The 19% of a CQ dose is recovered in feces.⁸⁶ Small amounts (5%) of the drug eliminate through the skin and up to 45% stored in lean tissues.⁴⁶ Elimination from the skin is very slow. CQ remains in the skin longer than 6 months, a time when the drug is no longer detectable in the plasma.⁸⁷ Chloroquine and active metabolite desethylchloroquine have elimination half-lives of 20 to 60 days and may be detected in urine months after a single dose.¹⁶ Chloroquine has a total clearance of 0.35–1L/h/kg.¹⁶ Renal clearance accounts for half of the total systemic clearance and increases by acidification of the urine.⁸⁷

The renal excretion accounts for 40–50% of HCQ elimination, where only 16–21% is excreted as unchanged drug.⁴⁷ The 24–25% of absorbed dose is excreted in the feces, which is greater than CQ feces excretion.⁴⁶ The elimination through the skin and long-term storage in lean tissues is identical to those of chloroquine, 5% and 45%, respectively.^21,46,88 The total clearance of hydroxychloroquine is 96 mL/min.⁴⁷ IV hydroxychloroquine has a half-life of 40 days (22.4 days in blood, and 123.5 days in plasma).⁸⁹ The elimination half-life of both drugs is significantly longer in patients with chronic renal disease.^16,57 In anuretic patients, the plasma levels were 70% and 25–30% higher for CQ and HCQ, respectively, compared to concentrations in subjects with normal kidney function.⁵³ Enantioselective renal elimination of the medications has been demonstrated in patients.^66,74 (S)-hydroxychloroquine had a mean renal clearance approximately twice that of (R)-hydroxychloroquine. In addition, the mean renal clearance of active (S)-metabolites was also higher than that of (R)-metabolites.⁸¹

The main mechanism of renal elimination of the medications is tubular secretion as renal excretion 7-fold exceeds the glomerular filtration rate.^57,90 The tubular secretion is an active process mediated by membrane proteins. It was reported that chloroquine is a substrate and potent competitive inhibitor of multidrug and toxin extrusion protein 1 (MATE1).^91,92 As substrates and/or inhibitors of active transport and metabolism, CQ and HCQ may be involved in several drug–drug interactions.^93–98

Pharmacogenomics of Chloroquine and Hydroxychloroquine

Response to drugs is subject to inter-individual variability. 40–70% of individuals that receive a drug, exhibit lack of efficacy, or adverse drug reactions. Up to 30% of the variability is attributed to genetic polymorphisms.⁹⁹ Cytochrome P450 enzymes are major determinants of drug response. They are responsible for approximately 80% of Phase I drug metabolism, and 70% of drug clearance.¹⁰⁰ The human CYP supergene family includes 57 genes, 12 of which are responsible for more than 75% of all drug oxidation reactions.¹⁰⁰ The CYP genes are highly polymorphic composed of large numbers of single-nucleotide polymorphisms (SNPs) and copy number variations. The most studied are genes of CYP2D6, 2C9, 2C8, 3A4, and 3A5 enzymes.^101–110 However, drug pharmacokinetics also depend on the renal excretion of the medications. MATE1, encoded by SLC47A1 gene, has been identified as a major efflux transporter involved in the renal excretion of many drugs including chloroquine.^91,111–115

Pharmacogenomics Informing Chloroquine Malaria Pharmacotherapy

Individual variation in drug response is a critical challenge in effective drug pharmacotherapy. Both the nature of the drug, as well as the dose of the drug, are subjected to vary on an individual basis. Genetic polymorphisms in metabolizing enzymes influence the pharmacokinetics and drug response.¹¹⁶

Plasmodium vivax is the major cause of malaria disease outside Africa. The World Health Organization (WHO) recommends chloroquine as a component of the treatment protocol for uncomplicated P. vivax malaria.¹¹⁷ Chloroquine is metabolized by the CYP450 isozymes 2C8, 3A4, 3A5 and 2D6. The CYP2C genes are located in a cluster on chromosome 10q24, organized as Cent-CYP2C18-CYP2C19-CYP2C9-CYP2C8-Tel.¹¹⁸ The CYP2C8 gene is approximately 30 kb in size and includes nine exons.¹¹⁹ CYP2C8 is the most divergent with respect to its protein sequence. Interindividual variability in chloroquine efficacy was previously reported in Africa and Asia and attributed to: P. vivax resistance to chloroquine, noncompliance, suboptimum dose and drug–drug interactions.¹²⁰ In a study reported in 2016, assessment of genetic polymorphisms in chloroquine metabolizing enzymes was identified as a need.¹²¹ To that end, the investigators focused on a cohort consisting of 164 P. vivax malaria patients followed during malaria treatment from 2007 to 2009. The study reported for the first time the influence of the CYP2C8 gene on gametocyte clearance rate with patients undergoing chloroquine/primaquine malaria treatment. From baseline until the first day of treatment, wild-type CYP2C8 homozygous individuals achieved greater reduction in gametocytes as compared to individuals without this genotype. The results suggested that CYP2C8, CYP2C9 and CYP23A5 genetic variants influenced chloroquine malaria treatment.

Pharmacogenomics Informing Hydroxychloroquine Lupus Pharmacotherapy

Discoid lupus erythematosus is the most common form of cutaneous lupus.¹²² Patients diagnosed with systemic lupus erythematosus and rheumatoid arthritis show a positive correlation between whole blood hydroxychloroquine levels and clinical response.⁶² HCQ is metabolized to N-desethylhydroxychloroquine in the liver.¹²³ The reaction is mediated by CYP3A4, CYP2C8, CYP2D6 and CYP3A5 isoforms.^71,123 In a study reported by Lee et al, 194 systemic lupus erythematosus patients were genotyped for 4 SNPs in CYP3A4*18B, CYP2D6*10, CYP3A5*3.¹²⁴ The association of the respective genotypes with blood hydroxychloroquine and N-desethyl hydroxychloroquine was the focus of the investigation. The CYP2D6*10 allelic variants were found to be significantly associated with the N-desethylhydroxychloroquine/hydroxychloroquine ratio. The study demonstrated that this ratio is related to CYP2D6 polymorphisms in systemic lupus erythematosus patients treated with hydroxychloroquine.

A multicenter observational and pharmacogenetic study with 200 discoid lupus erythematosus patients treated with HCQ was reported by Wahie et al.¹²⁵ Thirty-nine percent of the patients failed to respond to hydroxychloroquine, or developed toxicity. The study showed a trend for CYP2C8 variants to be associated with better response.

Discussion

The PK and PD characteristics of chloroquine and hydroxychloroquine need to be evaluated in order to provide safe and effective COVID-19 therapy. The pharmacokinetics of chloroquine and hydroxychloroquine are similar. Oral absorption of the drugs is comparable with bioavailability values of 0.7–0.8.^47,48 But the two medications have significant variations in bioavailability between individuals.^49,50 Genetic polymorphism of CYP enzymes involved in the presystemic metabolism can explain at least in part the individual differences in the oral absorption of the drugs and may reflect the variations in blood and tissue concentrations of the drugs in COVID-19 patients.

Both CQ and HCQ have wide distribution to the body tissues. Plasma protein binding of the drugs varies between 50% and 60%.^47,64 CQ and HCQ are known to be enantioselective in their dispositions. Both medications have similar stereoselective patterns of protein binding. S(+)-isomers are more bound to plasma proteins than R(-)-isomers, suggesting that free plasma concentrations are higher for R(-)-forms.⁴⁵ Such differences in the plasma protein binding can be responsible in part for the variations in therapeutic response and toxicity of the isomers in COVID-19 patients treated with CQ or HCQ, as only free drug can interact with receptors and produce therapeutic and/or side effects.

Chloroquine and hydroxychloroquine have long half-lives. The long half-life can be attributed to extensive tissue uptake rather than decreased elimination. Chloroquine is N-desethylated into two major metabolites largely by CYP3A4 and CYP2C8, while hydroxychloroquine metabolizes into three metabolites primarily by CYP3A4.^69,70 Metabolism of both drugs is stereoselective. The higher blood and plasma concentrations of (R)-forms confirm the stereoselective metabolism of the medications, where S-enantiomers metabolize faster than R-enantiomers.^74,81 As a result, S(+)-isomers have shorter half-life than R(-)-isomers. Similar doses of both medications produce large (11-fold) variations in the blood concentrations in patients with rheumatoid arthritis and in healthy volunteers.^47,64,82,83 Comparable differences in drug levels may be expected in COVID-19 patients. The variability can be explained by the stereoselective metabolism as well as genetic polymorphism of the P-450 enzymes involved in the biotransformation of the medications.

Individual variation in drug response is a critical challenge in effective drug pharmacotherapy. Up to 70% of individuals that receive a drug, exhibit lack of efficacy or adverse drug reactions, at least partially, due to genetic polymorphisms.⁹⁹ Gene polymorphisms influence metabolism as well as active transport.

CYP2D6 the most extensively studied CYP gene metabolizes approximately 25% of all drugs. Genetic polymorphisms resulting in increased CYP2D6 metabolic capabilities have been linked to decreased treatment response with tricyclic antidepressants, increased occurrences of respiratory depression and opioid toxicity.^101,102 CYP2C9 deficiency is related to bleeding complications with warfarin and other anticoagulants treatment. Fifty percent of interindividual variability in dose requirements is observed in concert with age, body surface area and polymorphisms in VKORC1.¹⁰³

CYP2C8 is involved in the metabolism of many medications including non-steroidal anti-inflammatory drugs, thiazolidinediones, chemotherapy agents, chloroquine and hydroxychloroquine. CYP2C8 genotyping should be considered as a viable option in Africans and Europeans in which 19.2% and 17.2% of CYP2C8 alleles, respectively, exhibit reduced functionality.¹⁰⁴

The CYP3A subfamily is the most abundant of the P-450 enzymes. CYP3A4 and CYP3A5 metabolize more than half of the marketed drugs.^105,106 The most common variant of CYP3A4 enzyme, CYP3A4*1B has been associated with reduced CYP3A4 activity.¹⁰⁷ The CYP3A4*1B allelic frequency varies among different ethnic groups, ranging from 0% in Chinese to 67% in African Americans.^108–110 CYP3A5 is polymorphically expressed in 10–20% in Caucasians, 33% in Japanese and 55% in African Americans.¹⁰⁹ The primary variant is CYP3A5*3, which has been associated with low CYP3A5 protein expression and reduced metabolic activity. The CYP3A5*3 allele frequency varies from approximately 50% in African Americans to 90% in Caucasians.¹⁰⁹ Other allelic variants have been reported for both CYP3A4 and CYP3A5. However, the variants occur at relatively low allelic frequencies and their functional significance has not been verified and validated.

Determination of CYP3A, CYP2C8 and CYP2D6 polymorphism and, therefore, activity is important to establish safe and efficient dosing of chloroquine and hydroxychloroquine for treatment of COVID-19 patients. A recent study reported for the first time the influence of the CYP2C8 gene on clearance in patients with chloroquine/primaquine therapy.¹²¹ Wild-type CYP2C8 homozygous individuals achieved greater reduction in gametocytes as compared to individuals without this genotype.¹²¹ Another study demonstrated CYP2D6 polymorphisms in systemic lupus erythematosus patients treated with hydroxychloroquine. CYP2D6*10 allelic variants were found to be significantly associated with altered metabolism of HCQ.¹²⁴ A study with 200 lupus erythematosus patients treated with hydroxychloroquine demonstrated 39% of the patients failed to respond to the therapy or developed toxicity.¹²⁵ Similar trend was observed in a recent clinical trial with COVID-patients treated with hydroxychloroquine.⁴³ Additionally, the study with 200 lupus erythematosus patients showed a trend for CYP2C8 variants to be associated with better response.¹²⁵ Overall, the results suggest that CYP2C8, CYP2D6 and CYP3A genetic polymorphisms may influence chloroquine and hydroxychloroquine pharmacokinetics and COVID-19 patients treated with the same dose of CQ or HCQ may exhibit lack of efficacy or adverse reactions. The variability in therapeutic response may require dose adjustment of CQ/HCQ in treatment of COVID-19 patients.

Urinary excretion is the primary route of elimination for chloroquine and hydroxychloroquine. The 50% of a CQ dose is recovered in the urine as unchanged drug, while only 16–21% of an HCQ dose is renally excreted as unchanged drug.^16,47 However, a greater fraction of absorbed HCQ dose excretes in the feces compared to the fraction of CQ dose. The elimination half-lives of both medications are significantly longer in patients with chronic renal disease.^16,57 This finding recommends that the two drugs should be used with caution in patients with renal impairment, as kidney dysfunction may lead to greater drug retention and higher risk of adverse effects. Renal elimination of both compounds is stereoselective, (S)-isomers have a mean renal clearance approximately twice that of (R)-isomers.⁸¹

The main mechanism of renal elimination of the medications is tubular secretion.^57,90 However, the molecular mechanisms of the renal tubular secretion remain mostly unidentified. It was reported that chloroquine is a substrate and potent inhibitor of the MATE1 transporter.⁹¹ Given the similarity in structure between CQ and HCQ, it is possible to propose that HCQ is also a substrate for MATE1. MATE1 is a proton-substrate antiporter expressed in the kidney and liver that facilitates the export of organic cations, such as metformin, paraquat, and oxaliplatin into urine and bile.⁹² Genetic polymorphisms of MATE1 may alter the pharmacokinetics and pharmacodynamics of the medications, as drug transporters are key determinants of elimination of the drugs. MATE1, encoded by the SLC47A1 gene, has been identified as a major efflux transporter involved in the renal excretion of chloroquine.⁹¹ Functional SNP of MATE1 (rs2289669 G>A) was associated with increased glucose-lowering activity of metformin through slowing renal excretion of the anti-diabetic drug.^111,112 The allele frequency ranges from 10.4% in African Americans to 49% in Mexican Americans.¹¹³ Other SNPs may also alter transport activity of MATE1 and lead to changed elimination of the corresponding drugs.^114,115 Genetic polymorphisms of MATE1 can affect renal elimination of CQ and HCQ and, therefore, may require dose adjustment based on pharmacogenomic profiles of COVID-19 patients.

Indeed, the active transporters and CYP enzyme polymorphisms may explain the variations in blood concentrations, therapeutic responses and severity of adverse effects of chloroquine and hydroxychloroquine. Despite the evidence of the influence of genetic polymorphisms on the pharmacokinetics of chloroquine and hydroxychloroquine, no large pharmacogenomics studies have been conducted to provide guidance on the use, dosing, and duration of the therapy in COVID-19 patients.

Additionally, chloroquine and hydroxychloroquine are involved in several DDIs. Cimetidine and ketoconazole, CYP3A4 inhibitors increased serum concentrations of CQ.^71,84 Predictably, cimetidine and ketoconazole may also increase HCQ blood concentrations by inhibition metabolism of the drug. Co-administration of CQ and HCQ with moderate and strong CYP3A4 (boceprevir, cobicistat, azole anti-fungal agents, macrolide antibiotics, etc.), CYP2C8 (gemfibrozil, clopidogrel, deferasirox, teriflunomide) and MATE1 inhibitors may result in increased plasma concentrations, longer half-life, exaggerated therapeutic effect and the toxicity of chloroquine and hydroxychloroquine.⁸⁵ Significant drug interactions with chloroquine and hydroxychloroquine that should be avoided or require additional monitoring include digoxin, antiepileptics, antacids, cyclosporine, amiodarone, azithromycin, moxifloxacin, insulin and other antidiabetic agents, tamoxifen, and praziquantel.^32,33 With the currently known or potential DDIs, the use of chloroquine and hydroxychloroquine with other drug therapy requires consideration for patient safety in COVID-19 patients.

Conclusions

Limited pharmacogenomic studies have been performed investigating the inter-patient variability of chloroquine and hydroxychloroquine in both malaria and lupus patient populations. Moreover, data to support the use of hydroxychloroquine and chloroquine for COVID-19 are limited and inconclusive. The off-label use of chloroquine and hydroxychloroquine to treat COVID-19 must be used with caution given the toxicities: cardiac, retinal and cutaneous severe adverse effects. Well-designed randomized trials incorporating pharmacogenomics need to be performed in a timely manner to achieve safe and effective dosing and to reduce severity of adverse effects.