[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.  Simulated Average Unbound Hydroxychloroquine Concentrations in Plasma for 120 Hours in Children and Adults
Simulated Average Unbound Hydroxychloroquine Concentrations in Plasma for 120 Hours in Children and Adults

Average unbound hydroxychloroquine concentrations in plasma for 120 hours in children and adults under the proposed dosing algorithm. Age classifications are based on postnatal age and are defined as follows: neonate (0 to <30 days), young infant (1 to <6 months), infant (6 to <24 months), young child (2 to <6 years), child (6 to <12 years), adolescent (12-18 years), and adult (20-50 years). Horizontal lines of the box correspond to the first quartile, median, and third quartiles. Upper whiskers extend from the third quartile to the largest observation to a maximum length of 1.5 times the interquartile range (ie, third quartile to first quartile). Lower whiskers extend from the first quartile to the lowest observation to a maximum length of 1.5 times the interquartile range (ie, third quartile to first quartile).

Figure 2.  Hydroxychloroquine Adult Population Physiologically Based Pharmacokinetic (PBPK) Model Simulations
Hydroxychloroquine Adult Population Physiologically Based Pharmacokinetic (PBPK) Model Simulations

Hydroxychloroquine adult population PBPK model simulations depicting a dosage regimen consisting of 400 mg every 12 hours × 2 doses followed by 200 mg every 12 hours × 8 doses. Linear scale (A) and semilog scale (B). Model-generated median and 95% prediction intervals are displayed as dashed lines and colored areas, respectively. An in vitro–derived effective concentration (50% EC = 0.72uM [242 ng/mL]) value for hydroxychloroquine for treatment of severe acute respiratory syndrome coronavirus 2–infected Vero cells is displayed for reference (dotted line).

Figure 3.  Simulated Plasma Exposures for Remdesivir in Children and Adults
Simulated Plasma Exposures for Remdesivir in Children and Adults

Simulated plasma exposures for remdesivir (area under the plasma concentration time curve from time 0 to infinity) in children and adults for the proposed dosing algorithm. Exposures representative of loading dose administration. Age classifications are based on postnatal age and are defined as follows: neonate (0 to <30 days), young infant (1 to <6 months), infant (6 to <24 months), young child (2 to <6 years), child (6 to <12 years), adolescent (12-18 years), and adult (20-50 years). Horizontal lines of the box correspond to the first quartile, median, and third quartiles. Upper whiskers extend from the third quartile to the largest observation to a maximum length of 1.5 times the interquartile range (ie, third quartile to first quartile). Lower whiskers extend from the first quartile to the lowest observation to a maximum length of 1.5 times the interquartile range (ie, third quartile to first quartile).

Table 1.  Hydroxychloroquine: Suggested Oral Pediatric Dosing Algorithma,b
Hydroxychloroquine: Suggested Oral Pediatric Dosing Algorithma,b
Table 2.  Pediatric Dosing Algorithm for Remdesivira
Pediatric Dosing Algorithm for Remdesivira
Supplement.

eMethods. Adult PBPK Model Evaluation

eTable 1. Hydroxychloroquine adult PBPK model parameterization

eTable 2. Adult PBPK Model Evaluation

eTable 3. Average Unbound Hydroxychloroquine Concentrations (Cavg) in Plasma Over 120 Hours in Children and Adults Using the Proposed Dosing Scheme*

eTable 4. Simulated Remdesivir Plasma Exposures (AUC from time 0 to ∞) Following Loading Dose Administration in Children and Adults under the Proposed Dosing Algorithm (Day 1)*

eTable 5. Simulated remdesivir plasma exposures (area under the concentration time curve [AUC] from time 0 to ∞) following loading dose administration in children and adults based on dosages recommended for Ebola virus diseasea

eTable 6. Average unbound hydroxychloroquine concentrations (Cavg) in plasma over 120 hours in children and adults using the proposed dosing scheme. Simulations computed using an alternative PBPK model that assumed proportional contributions of CYP2C8, CYP2D6 and CYP3A4 to hepatic intrinsic clearance were 59%, 16%, and 25%, respectively, based on in-vitro chloroquine metabolism data17

eFigure 1. Evaluation of the Developed Adult PBPK Model in Whole Blood and Plasma

eFigure 2. Comparison of Plasma Unbound Hydroxychloroquine Concentrations

eFigure 3. Comparison of Unbound Hydroxychloroquine Concentrations in Plasma to Lung Interstitial Fluid in Adults and Different Pediatric Age Groups

eFigure 4. Simulated Plasma Exposures for Remdesivir

eFigure 5. Comparison Simulated Hydroxychloroquine Concentrations in (A) Plasma/Serum and (B) Unbound Plasma/Serum Between Published Adult Models

eFigure 6. Comparison of Mean Plasma Remdesivir Exposures Versus Dose in Healthy Adult Volunteers

eFigure 7. Mean GS441524 Peripheral Blood Mononuclear Cell Exposures Versus Mean Plasma Remdesivir Exposures

1.
Dong  Y, Mo  X, Hu  Y,  et al.  Epidemiology of COVID-19 among children in China.   Pediatrics. 2020;e20200702. doi:10.1542/peds.2020-0702PubMedGoogle Scholar
2.
Lu  X, Zhang  L, Du  H,  et al; Chinese Pediatric Novel Coronavirus Study Team.  SARS-CoV-2 infection in children.   N Engl J Med. 2020;382(17):1663-1665. doi:10.1056/NEJMc2005073PubMedGoogle ScholarCrossref
3.
Li  Q, Guan  X, Wu  P,  et al.  Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia.   N Engl J Med. 2020;382(13):1199-1207. doi:10.1056/NEJMoa2001316PubMedGoogle ScholarCrossref
4.
Huang  C, Wang  Y, Li  X,  et al.  Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China.   Lancet. 2020;395(10223):497-506. doi:10.1016/S0140-6736(20)30183-5PubMedGoogle ScholarCrossref
5.
Li  G, De Clercq  E.  Therapeutic options for the 2019 novel coronavirus (2019-nCoV).   Nat Rev Drug Discov. 2020;19(3):149-150. doi:10.1038/d41573-020-00016-0PubMedGoogle ScholarCrossref
6.
Ko  WC, Rolain  JM, Lee  NY,  et al.  Arguments in favour of remdesivir for treating SARS-CoV-2 infections.   Int J Antimicrob Agents. 2020;55(4):105933. doi:10.1016/j.ijantimicag.2020.105933PubMedGoogle Scholar
7.
Colson  P, Rolain  JM, Raoult  D.  Chloroquine for the 2019 novel coronavirus SARS-CoV-2.   Int J Antimicrob Agents. 2020;55(3):105923. doi:10.1016/j.ijantimicag.2020.105923PubMedGoogle Scholar
8.
Raoult  D, Houpikian  P, Tissot Dupont  H, Riss  JM, Arditi-Djiane  J, Brouqui  P.  Treatment of Q fever endocarditis: comparison of 2 regimens containing doxycycline and ofloxacin or hydroxychloroquine.   Arch Intern Med. 1999;159(2):167-173. doi:10.1001/archinte.159.2.167PubMedGoogle ScholarCrossref
9.
Biot  C, Daher  W, Chavain  N,  et al.  Design and synthesis of hydroxyferroquine derivatives with antimalarial and antiviral activities.   J Med Chem. 2006;49(9):2845-2849. doi:10.1021/jm0601856PubMedGoogle ScholarCrossref
10.
Yao  X, Ye  F, Zhang  M,  et al.  In vitro antiviral activity and projection of optimized dosing design of hydroxychloroquine for the treatment of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).   Clin Infect Dis. 2020:ciaa237. doi:10.1093/cid/ciaa237PubMedGoogle Scholar
11.
Tett  SE, Cutler  DJ, Day  RO, Brown  KF.  A dose-ranging study of the pharmacokinetics of hydroxy-chloroquine following intravenous administration to healthy volunteers.   Br J Clin Pharmacol. 1988;26(3):303-313. doi:10.1111/j.1365-2125.1988.tb05281.xPubMedGoogle ScholarCrossref
12.
Furst  DE.  Pharmacokinetics of hydroxychloroquine and chloroquine during treatment of rheumatic diseases.   Lupus. 1996;5(suppl 1):S11-S15. doi:10.1177/0961203396005001041PubMedGoogle ScholarCrossref
13.
Lim  HS, Im  JS, Cho  JY,  et al.  Pharmacokinetics of hydroxychloroquine and its clinical implications in chemoprophylaxis against malaria caused by Plasmodium vivax.   Antimicrob Agents Chemother. 2009;53(4):1468-1475. doi:10.1128/AAC.00339-08PubMedGoogle ScholarCrossref
14.
Martinez  MA.  Compounds with therapeutic potential against novel respiratory 2019 coronavirus.   Antimicrob Agents Chemother. 2020;64(5):e00399-20. doi:10.1128/AAC.00399-20PubMedGoogle Scholar
15.
Warren  TK, Jordan  R, Lo  MK,  et al.  Therapeutic efficacy of the small molecule GS-5734 against Ebola virus in rhesus monkeys.   Nature. 2016;531(7594):381-385. doi:10.1038/nature17180PubMedGoogle ScholarCrossref
16.
Agostini  ML, Andres  EL, Sims  AC,  et al.  Coronavirus susceptibility to the antiviral remdesivir (GS-5734) is mediated by the viral polymerase and the proofreading exoribonuclease.   mBio. 2018;9(2):e00221-e18. doi:10.1128/mBio.00221-18PubMedGoogle ScholarCrossref
17.
European Medicines Agency (EMA). Summary on compassionate use: Remdesivir Gilead (Procedure No. EMEA/H/K/5622/CU). Published April 3, 2020. Accessed April 14, 2020. https://www.ema.europa.eu/en/documents/other/summary-compassionate-use-remdesivir-gilead_en.pdf
18.
Wang  M, Cao  R, Zhang  L,  et al.  Remdesivir and chloroquine effectively inhibit the recently emerged novel coronavirus (2019-nCoV) in vitro.   Cell Res. 2020;30(3):269-271. doi:10.1038/s41422-020-0282-0PubMedGoogle ScholarCrossref
19.
Maharaj  AR, Barrett  JS, Edginton  AN.  A workflow example of PBPK modeling to support pediatric research and development: case study with lorazepam.   AAPS J. 2013;15(2):455-464. doi:10.1208/s12248-013-9451-0PubMedGoogle ScholarCrossref
20.
Maharaj  AR, Edginton  AN.  Physiologically based pharmacokinetic modeling and simulation in pediatric drug development.   CPT Pharmacometrics Syst Pharmacol. 2014;3:e150. doi:10.1038/psp.2014.45PubMedGoogle Scholar
21.
Grimstein  M, Yang  Y, Zhang  X,  et al.  Physiologically based pharmacokinetic modeling in regulatory science: an update from the U.S. Food and Drug Administration’s Office of Clinical Pharmacology.   J Pharm Sci. 2019;108(1):21-25. doi:10.1016/j.xphs.2018.10.033PubMedGoogle ScholarCrossref
22.
Mahmood  I, Tegenge  MA.  A comparative study between allometric scaling and physiologically based pharmacokinetic modeling for the prediction of drug clearance from neonates to adolescents.   J Clin Pharmacol. 2019;59(2):189-197. doi:10.1002/jcph.1310PubMedGoogle ScholarCrossref
23.
Dykstra  K, Mehrotra  N, Tornøe  CW,  et al.  Reporting guidelines for population pharmacokinetic analyses.   J Pharmacokinet Pharmacodyn. 2015;42(3):301-314. doi:10.1007/s10928-015-9417-1PubMedGoogle ScholarCrossref
24.
Wickham  H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag; 2009.
25.
Wilke  CO. cowplot: Streamlined Plot Theme and Plot Annotations for 'ggplot2'. rdrr.io website. Accessed April 14, 2020. https://rdrr.io/cran/cowplot/
26.
Denney  W, Duvvuri  S, Buckeridge  C.  Simple, automatic noncompartmental analysis: the PKNCA R package.   J Pharmacokinet Pharmacodyn. 2015;42:S65-S65.Google Scholar
27.
Grothendieck  G, Zeileis  A.  zoo: S3 infrastructure for regular and irregular time series.   J Stat Softw. 2005;14(i06).Google Scholar
28.
Leong  R, Vieira  ML, Zhao  P,  et al.  Regulatory experience with physiologically based pharmacokinetic modeling for pediatric drug trials.   Clin Pharmacol Ther. 2012;91(5):926-931. doi:10.1038/clpt.2012.19PubMedGoogle ScholarCrossref
29.
Brocks  DR, Skeith  KJ, Johnston  C,  et al.  Hematologic disposition of hydroxychloroquine enantiomers.   J Clin Pharmacol. 1994;34(11):1088-1097. doi:10.1002/j.1552-4604.1994.tb01986.xPubMedGoogle ScholarCrossref
30.
McLachlan  AJ, Cutler  DJ, Tett  SE.  Plasma protein binding of the enantiomers of hydroxychloroquine and metabolites.   Eur J Clin Pharmacol. 1993;44(5):481-484. doi:10.1007/BF00315548PubMedGoogle ScholarCrossref
31.
Müller  F, König  J, Glaeser  H,  et al.  Molecular mechanism of renal tubular secretion of the antimalarial drug chloroquine.   Antimicrob Agents Chemother. 2011;55(7):3091-3098. doi:10.1128/AAC.01835-10PubMedGoogle ScholarCrossref
32.
Fan  HW, Ma  ZX, Chen  J, Yang  XY, Cheng  JL, Li  YB.  Pharmacokinetics and bioequivalence study of hydroxychloroquine sulfate tablets in Chinese healthy volunteers by LC-MS/MS.   Rheumatol Ther. 2015;2(2):183-195. doi:10.1007/s40744-015-0012-0PubMedGoogle ScholarCrossref
33.
Tett  SE, Cutler  DJ, Day  RO.  Bioavailability of hydroxychloroquine tablets assessed with deconvolution techniques.   J Pharm Sci. 1992;81(2):155-159. doi:10.1002/jps.2600810211PubMedGoogle ScholarCrossref
34.
Li  XQ, Björkman  A, Andersson  TB, Gustafsson  LL, Masimirembwa  CM.  Identification of human cytochrome P(450)s that metabolise anti-parasitic drugs and predictions of in vivo drug hepatic clearance from in vitro data.   Eur J Clin Pharmacol. 2003;59(5-6):429-442. doi:10.1007/s00228-003-0636-9PubMedGoogle ScholarCrossref
35.
McLachlan  AJ, Tett  SE, Cutler  DJ, Day  RO.  Bioavailability of hydroxychloroquine tablets in patients with rheumatoid arthritis.   Br J Rheumatol. 1994;33(3):235-239. doi:10.1093/rheumatology/33.3.235PubMedGoogle ScholarCrossref
36.
McLachlan  AJ, Tett  SE, Cutler  DJ, Day  RO.  Absorption and in vivo dissolution of hydroxycholoroquine in fed subjects assessed using deconvolution techniques.   Br J Clin Pharmacol. 1993;36(5):405-411. doi:10.1111/j.1365-2125.1993.tb00388.xPubMedGoogle ScholarCrossref
37.
Tett  SE, Cutler  DJ, Day  RO, Brown  KF.  Bioavailability of hydroxychloroquine tablets in healthy volunteers.   Br J Clin Pharmacol. 1989;27(6):771-779. doi:10.1111/j.1365-2125.1989.tb03439.xPubMedGoogle ScholarCrossref
38.
Cheung  KWK, van Groen  BD, Spaans  E,  et al.  A comprehensive analysis of ontogeny of renal drug transporters: mRNA analyses, quantitative proteomics, and localization.   Clin Pharmacol Ther. 2019;106(5):1083-1092. doi:10.1002/cpt.1516PubMedGoogle ScholarCrossref
39.
Willmann  S, Höhn  K, Edginton  A,  et al.  Development of a physiology-based whole-body population model for assessing the influence of individual variability on the pharmacokinetics of drugs.   J Pharmacokinet Pharmacodyn. 2007;34(3):401-431. doi:10.1007/s10928-007-9053-5PubMedGoogle ScholarCrossref
40.
Sciensano Epidemiology of Infectious Disease. Interim clinical guidance for adults with suspected or confirmed COVID-19 in Belgium, Version 7. sciensano website. Published April 7, 2020. Accessed April 14, 2020. https://epidemio.wiv-isp.be/ID/Documents/Covid19/COVID-19_InterimGuidelines_Treatment_ENG.pdf
41.
Michigan Medicine, University of Michigan. Inpatient guidance for treatment of COVID-19 in adults and children. Michigan Medicine, University of Michigan website. Accessed April 14, 2020. http://www.med.umich.edu/asp/pdf/adult_guidelines/COVID-19-treatment.pdf
42.
Kramer  NI, Krismartina  M, Rico-Rico  A, Blaauboer  BJ, Hermens  JL.  Quantifying processes determining the free concentration of phenanthrene in Basal cytotoxicity assays.   Chem Res Toxicol. 2012;25(2):436-445. doi:10.1021/tx200479kPubMedGoogle ScholarCrossref
43.
Postnikova  E, Cong  Y, DeWald  LE,  et al.  Testing therapeutics in cell-based assays: factors that influence the apparent potency of drugs.   PLoS One. 2018;13(3):e0194880. doi:10.1371/journal.pone.0194880PubMedGoogle Scholar
44.
Liu  J, Cao  R, Xu  M,  et al.  Hydroxychloroquine, a less toxic derivative of chloroquine, is effective in inhibiting SARS-CoV-2 infection in vitro.   Cell Discov. 2020;6:16. doi:10.1038/s41421-020-0156-0PubMedGoogle ScholarCrossref
45.
Zhou  D, Dai  SM, Tong  Q.  COVID-19: a recommendation to examine the effect of hydroxychloroquine in preventing infection and progression.   J Antimicrob Chemother. 2020:dkaa114. doi:10.1093/jac/dkaa114PubMedGoogle Scholar
46.
Arnold  SLM, Buckner  F.  Hydroxychloroquine for treatment of SARS-CoV-2 infection? improving our confidence in a model-based approach to dose selection.   Clin Transl Sci. 2020;. doi:10.1111/cts.12797PubMedGoogle Scholar
47.
Balevic  SJ, Green  TP, Clowse  MEB, Eudy  AM, Schanberg  LE, Cohen-Wolkowiez  M.  Pharmacokinetics of hydroxychloroquine in pregnancies with rheumatic diseases.   Clin Pharmacokinet. 2019;58(4):525-533. doi:10.1007/s40262-018-0712-zPubMedGoogle ScholarCrossref
48.
Morita  S, Takahashi  T, Yoshida  Y, Yokota  N.  Population pharmacokinetics of hydroxychloroquine in Japanese patients with cutaneous or systemic lupus erythematosus.   Ther Drug Monit. 2016;38(2):259-267. doi:10.1097/FTD.0000000000000261PubMedGoogle ScholarCrossref
49.
Tegenge  MA, Mahmood  I.  Age- and bodyweight-dependent allometric exponent model for scaling clearance and maintenance dose of theophylline from neonates to adults.   Ther Drug Monit. 2018;40(5):635-641. doi:10.1097/FTD.0000000000000543PubMedGoogle ScholarCrossref
50.
Lee  JY, Vinayagamoorthy  N, Han  K,  et al.  Association of polymorphisms of cytochrome P450 2D6 with blood hydroxychloroquine levels in patients with systemic lupus erythematosus.   Arthritis Rheumatol. 2016;68(1):184-190. doi:10.1002/art.39402PubMedGoogle ScholarCrossref
51.
Projean  D, Baune  B, Farinotti  R,  et al.  In vitro metabolism of chloroquine: identification of CYP2C8, CYP3A4, and CYP2D6 as the main isoforms catalyzing N-desethylchloroquine formation.   Drug Metab Dispos. 2003;31(6):748-754. doi:10.1124/dmd.31.6.748PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Views 6,863
    Citations 0
    Original Investigation
    Caring for the Critically Ill Patient
    June 5, 2020

    Simulated Assessment of Pharmacokinetically Guided Dosing for Investigational Treatments of Pediatric Patients With Coronavirus Disease 2019

    Author Affiliations
    • 1Duke Clinical Research Institute, Durham, North Carolina
    • 2Department of Pediatrics, Duke University School of Medicine, Durham, North Carolina
    • 3Department of Pharmacy, Duke University Medical Center, Durham, North Carolina
    • 4University of North Carolina Eshelman School of Pharmacy, Division of Pharmacotherapy and Experimental Therapeutics, University of North Carolina at Chapel Hill
    JAMA Pediatr. Published online June 5, 2020. doi:10.1001/jamapediatrics.2020.2422
    Key Points

    Question  What are appropriate dosing strategies for hydroxychloroquine and remdesivir in children with coronavirus disease 2019?

    Findings  In this simulation-based dose-ranging study, pediatric dosing strategies were devised that provided similar exposures between children within different developmental stages and adults. However, the analysis raised concerns regarding hydroxychloroquine use for coronavirus disease 2019 treatment because unbound plasma concentrations were less than those postulated to mediate an antiviral effect.

    Meaning  To confirm the appropriateness of the proposed dosing schemes, prospective pharmacokinetic, safety, and efficacy studies in children are required.

    Abstract

    Importance  Children of all ages appear susceptible to severe acute respiratory syndrome coronavirus 2 infection. To support pediatric clinical studies for investigational treatments of coronavirus disease 2019 (COVID-19), pediatric-specific dosing is required.

    Objective  To define pediatric-specific dosing regimens for hydroxychloroquine and remdesivir for COVID-19 treatment.

    Design, Setting, and Participants  Pharmacokinetic modeling and simulation were used to extrapolate investigated adult dosages toward children (March 2020-April 2020). Physiologically based pharmacokinetic modeling was used to inform pediatric dosing for hydroxychloroquine. For remdesivir, pediatric dosages were derived using allometric-scaling with age-dependent exponents. Dosing simulations were conducted using simulated pediatric and adult participants based on the demographics of a white US population.

    Interventions  Simulated drug exposures following a 5-day course of hydroxychloroquine (400 mg every 12 hours × 2 doses followed by 200 mg every 12 hours × 8 doses) and a single 200-mg intravenous dose of remdesivir were computed for simulated adult participants. A simulation-based dose-ranging study was conducted in simulated children exploring different absolute and weight-normalized dosing strategies.

    Main Outcomes and Measures  The primary outcome for hydroxychloroquine was average unbound plasma concentrations for 5 treatment days. Additionally, unbound interstitial lung concentrations were simulated. For remdesivir, the primary outcome was plasma exposure (area under the curve, 0 to infinity) following single-dose administration.

    Results  For hydroxychloroquine, the physiologically based pharmacokinetic model analysis included 500 and 600 simulated white adult and pediatric participants, respectively, and supported weight-normalized dosing for children weighing less than 50 kg. Geometric mean-simulated average unbound plasma concentration values among children within different developmental age groups (32-35 ng/mL) were congruent to adults (32 ng/mL). Simulated unbound hydroxychloroquine concentrations in lung interstitial fluid mirrored those in unbound plasma and were notably lower than in vitro concentrations needed to mediate antiviral activity. For remdesivir, the analysis included 1000 and 6000 simulated adult and pediatric participants, respectively. The proposed pediatric dosing strategy supported weight-normalized dosing for participants weighing less than 60 kg. Geometric mean-simulated plasma area under the time curve 0 to infinity values among children within different developmental age-groups (4315-5027 ng × h/mL) were similar to adults (4398 ng × h/mL).

    Conclusions and Relevance  This analysis provides pediatric-specific dosing suggestions for hydroxychloroquine and remdesivir and raises concerns regarding hydroxychloroquine use for COVID-19 treatment because concentrations were less than those needed to mediate an antiviral effect.

    Introduction

    Children of all ages appear susceptible to infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the novel coronavirus responsible for coronavirus disease 2019 (COVID-19).1 Epidemiologic and clinical data in both adults and children remain limited as COVID-19 continues to spread worldwide.1-4 While most pediatric cases have been asymptomatic or mild, cases of critically ill infants and children have been reported. No drugs have been proven efficacious in clinical trials for treating SARS-CoV-2, but hydroxychloroquine and the antiviral remdesivir (GS-5743) are undergoing clinical studies in adults.5-7

    Hydroxychloroquine has immunomodulating properties and is widely used in adults and children for the prophylaxis and treatment of malaria, other intracellular infections, and rheumatologic diseases.8 Hydroxychloroquine has shown in vitro activity against the related severe acute respiratory syndrome coronavirus 1,9 and in March 2020 demonstrated in vitro activity against SARS-CoV-2.10 Adult pharmacokinetic (PK) data indicate hydroxychloroquine exhibits extensive tissue and red blood cell distribution (volume of distribution in plasma >44 000 L).11 Clearance is primarily mediated through hepatic metabolism, with kidney excretion representing a minor eliminatory pathway.12,13 Based on a previously published physiologically based PK (PBPK) modeling analysis,10 400 mg twice daily for 1 day, followed by 200 mg twice daily for 4 additional days was postulated to provide optimal lung tissue concentrations in adults for treatment of SARS-CoV-2 infections.

    Remdesivir (GS-5473) is a phosphoramidate prodrug of an adenine nucleotide analogue that exhibits broad-spectrum antiviral activity against coronaviruses.14 Antiviral activity is mediated via intracellular conversion to its active triphosphate metabolite, GS-443902, which inhibits RNA-dependent RNA polymerase.15,16 Following intravenous administration in healthy adults, plasma concentrations of remdesivir rapidly decrease (half-life of approximately 1 hour).17 Pathways of remdesivir clearance in humans are incompletely characterized. Remdesivir is not yet approved for clinical use, but a February 2020 in vitro study demonstrated its activity against SARS-CoV-2.18

    Clinical trial efforts are rapidly being mobilized to combat COVID-19, with a focus on adult patients. To support clinical research efforts in pediatric populations, we sought to provide dosing suggestions for hydroxychloroquine and remdesivir in children using PK modeling and simulation.

    Methods

    Based on the availability of physicochemical and drug-specific absorption, distribution, metabolism, and excretion information, different methods were adopted to scale currently investigated adult dosages to children. For hydroxychloroquine, the availability of human PK studies permitted the development of a pediatric PBPK model.19 Physiologically based PK models are mathematical constructs combining organism-specific and drug-specific parameters to provide a priori predictions of drug PK.20 This modeling strategy is routinely incorporated into regulatory submissions to support initial dosing for pediatric clinical studies.21 Limited published human PK studies were available for remdesivir, so pediatric dosages were derived using a modified allometric scaling approach.22 This analysis follows applicable sections of the Enhancing the Quality and Transparency of Health Research reporting guidelines for population PK studies,23 was conducted from March 2020 to April 2020, and was exempt from the Duke University Health System institutional review board review because the research did not involve human participants.

    Software

    Physiologically based PK modeling was performed in PK-Sim, version 8 (Open Systems Pharmacology). Postprocessing of PBPK model simulations and allometry-based dose scaling were conducted in R, version 3.4.3 (R Foundation for Statistical Computing), and RStudio, version 1.1.383 (RStudio; Integrated Development for R), with the ggplot2, cowplot, zoo, and PKNCA packages.24-27

    Hydroxychloroquine

    Pediatric PBPK models were developed using a US Food and Drug Administration–supported workflow that leverages the abundance of adult PK information to construct an adult PBPK model prior to scaling toward children.19,28

    Adult PBPK Model Development

    An adult PBPK model for hydroxychloroquine was developed using physicochemical and absorption, distribution, metabolism, and excretion data obtained from the literature (eTable 1 in the Supplement). Hydroxychloroquine displays affinity for albumin and α1-acid glycoprotein12; however, for model parameterization, plasma protein binding was attributed toward α1-acid glycoprotein.29 Hepatic metabolism and kidney excretion represent major and minor pathways of hydroxychloroquine clearance, respectively.12,13 Approximately 26% of hydroxychloroquine appears unchanged in urine after intravenous dosing.11 Because the product of the glomerular filtration rate (120 mL/min) and fraction unbound in plasma (0.4830) is less than the estimated plasma kidney clearance for hydroxychloroquine (211 mL/min11), both glomerular filtration and tubular secretion were assumed to contribute toward kidney excretion. The specific mechanism of hydroxychloroquine tubular secretion is unknown; therefore, tubular secretion was attributed to multiantimicrobial extrusion protein-1 (MATE-1), an apical kidney transporter. This assumption was implemented based on an in vitro drug transport study that identified the antimalarial chloroquine, a structurally similar aminoquinoline, as a substrate for MATE-1.31

    Literature identifying the specific isozymes and their contribution toward hydroxychloroquine metabolism is lacking, so an initial adult PBPK model was developed using a single first-order hepatic intrinsic clearance process. Using blood and plasma concentration time data digitized from published adult PK studies, the following parameters were optimized in the initial adult model using PK-Sim’s Parameter Identification module: logarithm of the octanol-water partition coefficient (a direct predictor of drug distribution), intestinal (transcellular) permeability, and hepatic intrinsic clearance.11,32,33 Additionally, tubular secretion intrinsic clearance was optimized to maintain a fraction excreted unchanged in the urine of 26%.11 Following this parameter optimization process, the single hepatic intrinsic clearance process was segregated to proportionally account for the contributions of 3 hepatic isozymes (cytochrome P450 [CYP] 2C8, 2D6, and 3A4), defined based on in vitro metabolism data for chloroquine using human liver microsomes and recombinant cytochrome P450 enzymes (45%, 44%, and 11%, respectively).34 The developed adult PBPK model was evaluated against PK data from 3 studies not used for model development35-37 (eMethods in the Supplement).

    Pediatric PBPK Model Development

    Pediatric PBPK model development consisted of scaling anatomical (eg, organ sizes) and physiological (eg, blood flows) model parameters from the evaluated adult model toward children (birth to 18 years postnatal age). Age-specific anatomic and physiologic parameters are defined within PK-Sim. Ontogeny functions defined within PK-Sim for α1-acid glycoprotein, hepatic CYP2C8, CYP2D6, and CYP3A4 were used to facilitate pediatric PK predictions. For the apical kidney transporter, MATE-1, a previous investigation evaluating the ontogeny of mRNA expression and protein abundances of select kidney transporters failed to demonstrate an age-dependent association for MATE-1.38 Therefore, no ontogeny function was included for this transporter, but interpatient variability in MATE-1 expression was included in the model; variability was defined using digitized protein abundance data across all analyzed samples (geometric SD, 1.48).38

    Pediatric Dosing Simulations

    The developed pediatric PBPK model was used to inform oral hydroxychloroquine dosing for children ranging from term neonates to adolescents. We aimed to devise a pediatric dosing regimen to provide similar unbound plasma concentrations as predicted for adults receiving 400 mg of hydroxychloroquine sulfate every 12 hours × 2 doses, followed by 200 mg every 12 hours × 8 doses. This dose was recommended based on a previously published PBPK model analysis,10 where simulated lung tissue concentrations (adjusted by the fraction unbound in plasma) in adults were reported to exceed an in vitro–determined effective concentration for 50% inhibition of viral replication (50% EC; 0.72μM; 242 ng/mL).10 Age-specific pediatric dosages were optimized to achieve similar average unbound hydroxychloroquine concentrations over 120 hours in plasma during a 5-day period as those estimated for adults. Comparability of exposures was asserted for instances where the geometric mean of average unbound hydroxychloroquine concentrations for children fell within 80% to 125% of the simulated value for adults. The dosing scheme was developed based on model simulations for a simulated population consisting of 600 children (birth to 18 years postnatal age). Adult reference simulations were generated for 500 simulated participants ranging from 20 to 50 years postnatal age. Simulated participants were generated in PK-Sim based on the demographics of a white US population.39 The male-to-female ratio of the generated population was 50:50. To demonstrate the relevance of hydroxychloroquine exposures in unbound plasma toward the target organ space of interest for SARS-CoV-2 infections, unbound lung interstitial fluid simulations were generated to compare concentrations between the 2 matrices.

    Remdesivir

    A modified allometric-based scaling approach was used to inform remdesivir dosing in children ranging from term neonates to adolescents.22

    Allometry With Age-Dependent Exponents

    Drug clearance is a primary PK parameter that, in conjunction with dose, can provide estimates of drug exposure. To estimate age-specific clearance values for remdesivir in children, we used an allometric approach with age-dependent scalers.22 Using this approach, pediatric clearance values were estimated from adult values using Equation 1:

    Image description not available.

    where CLi represents the individual clearance; CLstd, clearance of remdesivir in a healthy 70-kg adult; WTi, individual weight; and b, an age-dependent exponent with values of 1.1, 1.0, 0.9, and 0.75 for children 3 months or younger, older than 3 months to 2 years, older than 2 years to 5 years, and older than 5 years, respectively.

    Pediatric Dosing Simulations

    In clinical trials, the adult remdesivir dose for SARS-CoV-2 infection is a 200-mg intravenous load (day 1) followed by 100 mg intravenous daily (≥day 2).40,41 Because pediatric dosing is not yet established for SARS-CoV-2, pediatric dosing for Ebola virus disease (EVD) may provide a logical starting point. For pediatric patients weighing less than 40 kg, EVD dosing consists of 5 mg/kg intravenous load (day 1) followed by 2.5 mg/kg daily (≥day 2).

    In this analysis, we aimed to (1) estimate plasma remdesivir exposure in children following administration of EVD recommended dosages and (2) devise a pediatric dosing regimen to provide similar plasma exposures as observed in adults receiving currently recommend doses. Clearance of remdesivir in adults was calculated from a dose-ranging study (10-225 mg) where areas under the plasma concentration time curve (AUC) from 0 to infinity were defined.17 For the evaluated dosing cohorts, each composed of 8 healthy adult participants, clearance was defined as the quotient of dose and AUC time from 0 to infinity. Mean clearance for all dosing cohorts was assumed to represent the value for a healthy 70-kg adult.

    Pharmacokinetic simulations were conducted for simulated populations consisting of 6000 simulated children (birth to 18 years postnatal age) and 1000 simulated adults (age 20-50 years). Simulated participants were generated in PK-Sim based on a white US population.39 The male-to-female ratio of the generated population was 50:50. Simulated AUC 0 to infinity was calculated for each simulated participant according to Equation 2:

    Image description not available.

    where Dosei is the individual dose in mg; CLi, individual clearance (liters per hour) calculated based on Equation 1; and ni, a random variable used to introduce interindividual variability into PK simulations. We empirically parameterized ni as N(0,0.3) to provide an approximate interindividual variability of 30% on remdesivir clearance. Considering the plasma half-life of remdesivir in adults is approximately 1 hour and a dosing frequency of every 24 hours, it was inferred that remdesivir would not accumulate in plasma using daily dosing.17 Therefore, PK simulations were computed after single doses. Pediatric doses were considered equivalent to adults if the geometric mean of plasma AUC 0 to infinity in children was within 80% to 125% of adult values.

    Results
    Hydroxychloroquine
    Adult PBPK Model Evaluation

    The developed adult PBPK model adequately characterized hydroxychloroquine disposition in plasma and blood among healthy volunteers and participants with rheumatoid arthritis following oral dosing (eTable 2 and eFigure 1 in the Supplement). Ratios of model predicted vs observed maximum concentration values ranged between 0.81 to 0.96 in blood and plasma. Physiologically based PK model-predicted blood and plasma exposures computed over 100 hours (ie, AUC, 0-100 hours) were similar to estimates derived from previously published empirically based PK models.37 Ratios of PBPK model predicted vs empirically derived AUC from 0 to 100 hour values were 1.03 for plasma and 1.17 for blood. Additionally, empirically derived concentration time estimates in blood and plasma were adequately described by the developed adult PBPK model (eFigure 1F and G in the Supplement).

    Age-Specific Dosing Simulations

    The PBPK model–informed pediatric dosing scheme classifies patients based on weight and uses weight-normalized dosing for children less than 50 kg (Table 1). Using the suggested dosing regimen, geometric mean-simulated average unbound plasma concentrations values among children within different developmental age groups (32-35 ng/mL) were congruent to adult values (32 ng/mL; Figure 1; eTable 3 in the Supplement). Additionally, the trajectory of simulated plasma unbound hydroxychloroquine concentrations within different developmental age groups were similar to those depicted in adults over 120 hours (eFigure 2 in the Supplement). Furthermore, PBPK model simulations indicated that plasma unbound concentrations approximated unbound lung interstitial fluid concentrations (eFigure 3 in the Supplement). For example, geometric mean-simulated average concentrations values in unbound plasma and unbound lung interstitial fluid were both 35 ng/mL in neonates. Lastly, to demonstrate the differences in hydroxychloroquine exposure between sampling matrices, we simulated blood, plasma, and unbound plasma exposures in adults following the administration of a previously defined 5-day dosing regimen (400 mg every 12 hours × 2 doses, followed by 200 mg every 12 hours ×8 doses; Figure 2). Geometric mean-simulated average concentration values in whole blood, plasma, and unbound plasma were 375 ng/mL, 66 ng/mL, and 32 ng/mL, respectively. Importantly, during the 5 days of dosing, unbound plasma hydroxychloroquine concentrations were less than the reference in vitro 50% EC value for viral inhibition (0.72μM; 242 ng/mL).10

    Age-Specific Dosing Simulations: Remdesivir

    Table 2 displays the suggested pediatric dosing regimen for remdesivir, based on allometry with age-dependent exponents. The dosing scheme classifies patients by weight and provides weight-normalized dosages for patients weighing less than 60 kg. Geometric mean-simulated plasma AUC 0 to infinity values among children within different developmental age groups (4315-5027 ng × h/mL) were similar to adults (4398 ng × h/mL; Figure 3; eTable 4 in the Supplement). In contrast, for children younger than 12 years, geometric mean plasma AUC 0 to infinity values were 147% to 256% of the adult value for simulations based on dosing recommendations for EVD (eFigure 4 and eTable 5 in the Supplement).

    Discussion

    We sought to provide pediatric-specific dosing regimens for investigational therapies against COVID-19 in children. Although simulated hydroxychloroquine concentrations in unbound plasma were similar between children and adults using the model-derived dosing scheme (eTable 3 in the Supplement), a significant finding from our analysis was that unbound plasma concentrations failed to exceed the reference 50% EC value (Figure 2). This in vitro–determined value was derived based on 48-hour incubations with SARS-CoV-2-infected Vero cells (0.72μM; 242 ng/mL).10 Factors such as drug binding in incubation, virus input, time of infection, and cell passage number can influence drug potency in cell-based assays, so variability in in vitro–determined 50% EC values is expected.42,43 Nevertheless, separately published in vitro–derived 50% EC values for hydroxychloroquine were considerably higher, ranging from 4.06 to 17.31μM (1364-5814 ng/mL), concentrations that are unachievable in unbound plasma with current evaluated dosing.44 Because simulated hydroxychloroquine concentrations in unbound plasma parallel those in unbound lung interstitial fluid, our analysis indicates that target site concentrations of hydroxychloroquine for COVID-19 will be lower than those required to exert an antiviral effect; however, our findings do not deny the potential utility of hydroxychloroquine for COVID-19 treatment based on other mechanisms of action, such as immunomodulation.45 Additionally, because viruses are intrinsically intracellular pathogens, the reference concentration of interest may lie within the intracellular space. In contrast to a competing PBPK model-based analysis for hydroxychloroquine, we chose not to simulate lung intracellular or lung tissue concentrations for 2 reasons.10 First, such simulations are highly dependent on the defined lung tissue-to-plasma partition coefficient. Depending on the source (preclinical species or experimental conditions), lung tissue-to-plasma partition coefficient values can vary, influencing the outcome of model simulations.46 Second, in vitro studies assessing hydroxychloroquine’s antiviral effects were based on drug added to incubations containing Vero cells,10,44 where 50% EC values were defined based on drug concentrations in incubation (ie, extracellular concentration), not intracellular or tissue concentrations. Therefore, in this analysis, we simulated unbound concentrations in the lung interstitial fluid (ie, extracellular fluid), which represents the most relevant compartment for assessing hydroxychloroquine’s in vivo antiviral efficacy based on in vitro–derived 50% EC values.

    Simulated hydroxychloroquine plasma concentrations for a typical 70-kg adult participant from our PBPK model were similar to estimates generated by another published PBPK model (400 mg every 12 hours × 2 doses followed by 200 mg every 12 hours × 8 doses; eFigure 5A in the Supplement).10 However, simulated (typical) concentration estimates from competing population PK models were relatively higher (eFigure 5A in the Supplement).13,47,48 For example, the maximum plasma concentration value on dosing day 5 was 7.5 times higher for the Lim et al13 population PK model compared with our PBPK model13; despite these discrepancies, anticipated unbound plasma concentrations are expected to be less than 242 ng/mL (ie, the lowest reported 50% EC value) for all models except for the Lim et al13 model (eFigure 5B in the Supplement). As such, simulations based on different modeling approaches may provide alternate interpretations of the utility of hydroxychloroquine for COVID-19 treatment.

    Our understanding of clearance pathways for remdesivir and its metabolites in humans are limited, so we used an empirical-based strategy, allometry with age-dependent exponents, to estimate pediatric dosing.22 Compared with traditional fixed-exponent allometry, the use of age-dependent exponents provides more conservative dosage estimates for younger children (<5 years), where key physiologic processes that mediate drug clearance have yet to reach maturity.49 In an analysis of 73 drugs, which included compounds cleared through various pathways (ie, hepatic and kidney), allometry with age-dependent exponents provided a similar predictive capacity to pediatric PBPK modeling, with 90.6% and 91.1% of clearance or exposure estimates falling within 0.5-fold to 2-fold of observed pediatric values, respectively.22 In our analysis, dose extrapolations for remdesivir were performed assuming the presence of dose-proportional PK (ie, linear PK) in both children and adults. Although pediatric PK information for remdesivir is currently unavailable, preliminary data from adults indicate the presence of dose-proportional PK (eFigure 6 in the Supplement).

    In this analysis, simulations were based on white populations, which were used to define anatomic and physiologic parameters for the hydroxychloroquine PBPK model and the age and weight distribution of participants in the remdesivir allometric-based analysis. Owing to hydroxychloroquine’s extensive tissue distribution, we do not anticipate weight-normalized volume of distribution values to vary substantially among different races. However, CYP2D6 may contribute to hydroxychloroquine’s hepatic metabolism, so racial/ethnic differences in enzyme polymorphisms might play a role in perpetuating variability in drug clearance. For example, an analysis of 194 Korean adult patients with systemic lupus erythematosus indicated that CYP2D6 polymorphism (rs1065852) was associated with differences in whole-blood concentrations of the hydroxychloroquine metabolite, N-desethyl hydroxychloroquine.50 Nevertheless, no difference in whole-blood concentrations of hydroxychloroquine was observed between the different polymorphisms. For our analysis of remdesivir, dosing was segregated using weight-based cohorts. Consequently, differences in body size between different races/ethnicities were not anticipated to impart a substantial influence.

    Limitations

    This analysis has some limitations. The proposed pediatric dosing regimens for both compounds were based solely on PK considerations, which assumes that the concentration-effect association (ie, pharmacodynamic) is preserved between children and adults. Additionally, owing to knowledge gaps associated with hydroxychloroquine metabolism, the identity and proportional contribution of the specific isozymes responsible for hepatic metabolism were assumed based on in vitro drug metabolism data for chloroquine.34 To assess our model’s sensitivity to these assumptions, we formulated an alternative PBPK model based on a separate in vitro analysis of chloroquine metabolism where proportional contributions of CYP2C8, CYP2D6, and CYP3A4 toward hepatic intrinsic clearance were 59%, 16%, and 25%, respectively.51 Using this alternative model in conjunction with the previously proposed pediatric dosing scheme, congruent unbound plasma concentrations between children and adults were observed (eTable 6 in the Supplement). Furthermore, the proposed pediatric dosing scheme for remdesivir was based on identifying dosages that would provide equivalent plasma exposures between children and adults; however, because remdesivir is a nonactive prodrug, extrapolating pediatric dosages based on intracellular exposure of the active triphosphate moiety (GS-443902) would be more appropriate. Unfortunately, PK information defining the association between remdesivir dose, plasma exposure, and intracellular GS-443902 exposure in children is lacking. Based on a preliminary analysis of remdesivir PK studies in adults, there appears to be a linear association between remdesivir plasma exposure and intracellular peripheral blood mononuclear cell exposure of GS-441524, a nucleoside metabolite that may reflect the total amount of dephosphorylated as well as monophosphate, diphosphate, and triphosphate metabolites of remdesivir (eFigure 7 in the Supplement).17 To define pediatric-specific dosages, our analysis assumed the presence of a similar association in children. Lastly, owing to limited pediatric PK information for both compounds, we could not evaluate the predictive performance of the developed PK models used to inform pediatric dosing.

    Conclusions

    We used PK modeling and simulation to identify pediatric-specific dosages for hydroxychloroquine and remdesivir, which highlights how PK modeling and simulation can support the rapid development of pediatric clinical studies. While our study focused on investigational COVID-19 treatments, the described methods are generalizable and can be applied to other drugs and disease states. Importantly, our findings raise concerns about hydroxychloroquine for COVID-19 treatment because unbound plasma exposures were less than those needed to mediate an antiviral effect. To confirm the appropriateness of the proposed dosing schemes, prospective PK, safety, and efficacy studies in children are required.

    Back to top
    Article Information

    Corresponding Author: Michael Cohen-Wolkowiez, MD, PhD, Department of Pediatrics, Duke University School of Medicine, Durham, NC 27705 (michael.cohenwolkowiez@duke.edu).

    Accepted for Publication: May 13, 2020.

    Published Online: June 5, 2020. doi:10.1001/jamapediatrics.2020.2422

    Author Contributions: Dr Cohen-Wolkowiez had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Maharaj, C. P. Hornik, Balevic, Smith, Benjamin, Cohen-Wolkowiez.

    Acquisition, analysis, or interpretation of data: Maharaj, Wu, C. P. Hornik, Balevic, C. D. Hornik, Gonzalez, Zimmerman, Cohen-Wolkowiez.

    Drafting of the manuscript: Maharaj, Wu, C. P. Hornik, C. D. Hornik.

    Critical revision of the manuscript for important intellectual content: Maharaj, C. P. Hornik, Balevic, C. D. Hornik, Smith, Gonzalez, Zimmerman, Benjamin, Cohen-Wolkowiez.

    Statistical analysis: Maharaj, Wu, C. P. Hornik, Balevic.

    Obtained funding: Benjamin, Cohen-Wolkowiez.

    Administrative, technical, or material support: C. D. Hornik, Cohen-Wolkowiez.

    Supervision: C. P. Hornik, Benjamin, Cohen-Wolkowiez.

    Conflict of Interest Disclosures: Dr C. P. Hornik reported personal fees from Anavex Pharmaceuticals and grants from Pfizer outside the submitted work. Dr Balevic reported grants, nonfinancial support, and other support from the US Food and Drug Administration, grants from the National Institutes of Health, Patient-Centered Outcomes Research Institute, Rheumatology Research Foundation, Thrasher Research Fund, and Childhood Arthritis and Research Alliance/Arthritis Foundation; and personal fees from UCB outside the submitted work. Dr Gonzalez reported a travel grant through the University of North Carolina at Chapel Hill to give a presentation at Boehringer Ingelheim outside the submitted work and support for research from the Eunice Kennedy Shriver National Institute of Child Health and Human Development. Dr Zimmerman reported grants from the National Institutes of Health during the conduct of the study. Dr Maharaj receives research support from the Thrasher Research Fund (www.thrasherresearch.org). Dr Smith receives funding from the National Institutes of Health. Dr Benjamin receives support from the National Institutes of Health, National Institute of Child Health and Human Development, the National Center for Advancing Translational Sciences, and Food and Drug Administration and personal fees from Astellas Pharma, Cidara Therapeutics, Allergan, and Lediant outside the submitted work. Dr Cohen-Wolkowiez receives support from the National Institutes of Health, National Institute of Child Health and Human Development, the National Center for Advancing Translational Sciences, and the US Food and Drug Administration; he also receives research support from industry for neonatal and pediatric drug development. No other disclosures were reported.

    Funding/Support: This work was funded through support from the National Institutes of Health grant 1K24-AI143971 (Dr Cohen-Wolkowiez). This work was also funded under the National Institute of Child Health and Human Development contract (HHSN275201000003I) for the Pediatric Trials Network (Principal Investigator, Dr Benjamin).

    Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Best Pharmaceuticals for Children Act–Pediatric Trials Network Steering Committee Members: Daniel K. Benjamin Jr, MD, PhD, MPH, Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina; Christoph P. Hornik, MD, PhD, MPH, Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina; Kanecia O. Zimmerman, MD, Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina; Phyllis Kennel, MS, RD, LDN, PMP, Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina; Rose Beci, BS, Duke Clinical Research Institute, Duke University School of Medicine, Durham, North Carolina; Chi Dang Hornik, PharmD, Duke University Medical Center, Durham, North Carolina; Gregory L. Kearns, BSc (Pharm), PharmD, PhD, TCU-UNTHSC School of Medicine, Fort Worth, Texas; Matthew Laughon, MD, MPH, University of North Carolina at Chapel Hill; Ian M. Paul, MD, MSc, Penn State College of Medicine, Hershey, Pennsylvania; Janice E. Sullivan, MD: University of Louisville, Louisville, Kentucky; Kelly Wade, MD, PhD, MSCE, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania; Paula Delmore, MSM, Wichita Medical Research and Education Foundation, Wichita, Kansas; Perdita Taylor-Zapata, MD, The Eunice Kennedy Shriver National Institute of Child Health and Human Development; June Lee, MD, PhD, The Eunice Kennedy Shriver National Institute of Child Health and Human Development. PTN Publications Committee: Chaired by Thomas P. Green, MD, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, Illinois.

    Disclaimer: All information and materials in the manuscript are original. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

    Additional Contributions: We thank Erin Campbell, MS, for her editorial assistance. Ms Campbell did not receive compensation for her assistance, apart from her employment at Duke Clinical Research Instititue.

    References
    1.
    Dong  Y, Mo  X, Hu  Y,  et al.  Epidemiology of COVID-19 among children in China.   Pediatrics. 2020;e20200702. doi:10.1542/peds.2020-0702PubMedGoogle Scholar
    2.
    Lu  X, Zhang  L, Du  H,  et al; Chinese Pediatric Novel Coronavirus Study Team.  SARS-CoV-2 infection in children.   N Engl J Med. 2020;382(17):1663-1665. doi:10.1056/NEJMc2005073PubMedGoogle ScholarCrossref
    3.
    Li  Q, Guan  X, Wu  P,  et al.  Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia.   N Engl J Med. 2020;382(13):1199-1207. doi:10.1056/NEJMoa2001316PubMedGoogle ScholarCrossref
    4.
    Huang  C, Wang  Y, Li  X,  et al.  Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China.   Lancet. 2020;395(10223):497-506. doi:10.1016/S0140-6736(20)30183-5PubMedGoogle ScholarCrossref
    5.
    Li  G, De Clercq  E.  Therapeutic options for the 2019 novel coronavirus (2019-nCoV).   Nat Rev Drug Discov. 2020;19(3):149-150. doi:10.1038/d41573-020-00016-0PubMedGoogle ScholarCrossref
    6.
    Ko  WC, Rolain  JM, Lee  NY,  et al.  Arguments in favour of remdesivir for treating SARS-CoV-2 infections.   Int J Antimicrob Agents. 2020;55(4):105933. doi:10.1016/j.ijantimicag.2020.105933PubMedGoogle Scholar
    7.
    Colson  P, Rolain  JM, Raoult  D.  Chloroquine for the 2019 novel coronavirus SARS-CoV-2.   Int J Antimicrob Agents. 2020;55(3):105923. doi:10.1016/j.ijantimicag.2020.105923PubMedGoogle Scholar
    8.
    Raoult  D, Houpikian  P, Tissot Dupont  H, Riss  JM, Arditi-Djiane  J, Brouqui  P.  Treatment of Q fever endocarditis: comparison of 2 regimens containing doxycycline and ofloxacin or hydroxychloroquine.   Arch Intern Med. 1999;159(2):167-173. doi:10.1001/archinte.159.2.167PubMedGoogle ScholarCrossref
    9.
    Biot  C, Daher  W, Chavain  N,  et al.  Design and synthesis of hydroxyferroquine derivatives with antimalarial and antiviral activities.   J Med Chem. 2006;49(9):2845-2849. doi:10.1021/jm0601856PubMedGoogle ScholarCrossref
    10.
    Yao  X, Ye  F, Zhang  M,  et al.  In vitro antiviral activity and projection of optimized dosing design of hydroxychloroquine for the treatment of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).   Clin Infect Dis. 2020:ciaa237. doi:10.1093/cid/ciaa237PubMedGoogle Scholar
    11.
    Tett  SE, Cutler  DJ, Day  RO, Brown  KF.  A dose-ranging study of the pharmacokinetics of hydroxy-chloroquine following intravenous administration to healthy volunteers.   Br J Clin Pharmacol. 1988;26(3):303-313. doi:10.1111/j.1365-2125.1988.tb05281.xPubMedGoogle ScholarCrossref
    12.
    Furst  DE.  Pharmacokinetics of hydroxychloroquine and chloroquine during treatment of rheumatic diseases.   Lupus. 1996;5(suppl 1):S11-S15. doi:10.1177/0961203396005001041PubMedGoogle ScholarCrossref
    13.
    Lim  HS, Im  JS, Cho  JY,  et al.  Pharmacokinetics of hydroxychloroquine and its clinical implications in chemoprophylaxis against malaria caused by Plasmodium vivax.   Antimicrob Agents Chemother. 2009;53(4):1468-1475. doi:10.1128/AAC.00339-08PubMedGoogle ScholarCrossref
    14.
    Martinez  MA.  Compounds with therapeutic potential against novel respiratory 2019 coronavirus.   Antimicrob Agents Chemother. 2020;64(5):e00399-20. doi:10.1128/AAC.00399-20PubMedGoogle Scholar
    15.
    Warren  TK, Jordan  R, Lo  MK,  et al.  Therapeutic efficacy of the small molecule GS-5734 against Ebola virus in rhesus monkeys.   Nature. 2016;531(7594):381-385. doi:10.1038/nature17180PubMedGoogle ScholarCrossref
    16.
    Agostini  ML, Andres  EL, Sims  AC,  et al.  Coronavirus susceptibility to the antiviral remdesivir (GS-5734) is mediated by the viral polymerase and the proofreading exoribonuclease.   mBio. 2018;9(2):e00221-e18. doi:10.1128/mBio.00221-18PubMedGoogle ScholarCrossref
    17.
    European Medicines Agency (EMA). Summary on compassionate use: Remdesivir Gilead (Procedure No. EMEA/H/K/5622/CU). Published April 3, 2020. Accessed April 14, 2020. https://www.ema.europa.eu/en/documents/other/summary-compassionate-use-remdesivir-gilead_en.pdf
    18.
    Wang  M, Cao  R, Zhang  L,  et al.  Remdesivir and chloroquine effectively inhibit the recently emerged novel coronavirus (2019-nCoV) in vitro.   Cell Res. 2020;30(3):269-271. doi:10.1038/s41422-020-0282-0PubMedGoogle ScholarCrossref
    19.
    Maharaj  AR, Barrett  JS, Edginton  AN.  A workflow example of PBPK modeling to support pediatric research and development: case study with lorazepam.   AAPS J. 2013;15(2):455-464. doi:10.1208/s12248-013-9451-0PubMedGoogle ScholarCrossref
    20.
    Maharaj  AR, Edginton  AN.  Physiologically based pharmacokinetic modeling and simulation in pediatric drug development.   CPT Pharmacometrics Syst Pharmacol. 2014;3:e150. doi:10.1038/psp.2014.45PubMedGoogle Scholar
    21.
    Grimstein  M, Yang  Y, Zhang  X,  et al.  Physiologically based pharmacokinetic modeling in regulatory science: an update from the U.S. Food and Drug Administration’s Office of Clinical Pharmacology.   J Pharm Sci. 2019;108(1):21-25. doi:10.1016/j.xphs.2018.10.033PubMedGoogle ScholarCrossref
    22.
    Mahmood  I, Tegenge  MA.  A comparative study between allometric scaling and physiologically based pharmacokinetic modeling for the prediction of drug clearance from neonates to adolescents.   J Clin Pharmacol. 2019;59(2):189-197. doi:10.1002/jcph.1310PubMedGoogle ScholarCrossref
    23.
    Dykstra  K, Mehrotra  N, Tornøe  CW,  et al.  Reporting guidelines for population pharmacokinetic analyses.   J Pharmacokinet Pharmacodyn. 2015;42(3):301-314. doi:10.1007/s10928-015-9417-1PubMedGoogle ScholarCrossref
    24.
    Wickham  H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag; 2009.
    25.
    Wilke  CO. cowplot: Streamlined Plot Theme and Plot Annotations for 'ggplot2'. rdrr.io website. Accessed April 14, 2020. https://rdrr.io/cran/cowplot/
    26.
    Denney  W, Duvvuri  S, Buckeridge  C.  Simple, automatic noncompartmental analysis: the PKNCA R package.   J Pharmacokinet Pharmacodyn. 2015;42:S65-S65.Google Scholar
    27.
    Grothendieck  G, Zeileis  A.  zoo: S3 infrastructure for regular and irregular time series.   J Stat Softw. 2005;14(i06).Google Scholar
    28.
    Leong  R, Vieira  ML, Zhao  P,  et al.  Regulatory experience with physiologically based pharmacokinetic modeling for pediatric drug trials.   Clin Pharmacol Ther. 2012;91(5):926-931. doi:10.1038/clpt.2012.19PubMedGoogle ScholarCrossref
    29.
    Brocks  DR, Skeith  KJ, Johnston  C,  et al.  Hematologic disposition of hydroxychloroquine enantiomers.   J Clin Pharmacol. 1994;34(11):1088-1097. doi:10.1002/j.1552-4604.1994.tb01986.xPubMedGoogle ScholarCrossref
    30.
    McLachlan  AJ, Cutler  DJ, Tett  SE.  Plasma protein binding of the enantiomers of hydroxychloroquine and metabolites.   Eur J Clin Pharmacol. 1993;44(5):481-484. doi:10.1007/BF00315548PubMedGoogle ScholarCrossref
    31.
    Müller  F, König  J, Glaeser  H,  et al.  Molecular mechanism of renal tubular secretion of the antimalarial drug chloroquine.   Antimicrob Agents Chemother. 2011;55(7):3091-3098. doi:10.1128/AAC.01835-10PubMedGoogle ScholarCrossref
    32.
    Fan  HW, Ma  ZX, Chen  J, Yang  XY, Cheng  JL, Li  YB.  Pharmacokinetics and bioequivalence study of hydroxychloroquine sulfate tablets in Chinese healthy volunteers by LC-MS/MS.   Rheumatol Ther. 2015;2(2):183-195. doi:10.1007/s40744-015-0012-0PubMedGoogle ScholarCrossref
    33.
    Tett  SE, Cutler  DJ, Day  RO.  Bioavailability of hydroxychloroquine tablets assessed with deconvolution techniques.   J Pharm Sci. 1992;81(2):155-159. doi:10.1002/jps.2600810211PubMedGoogle ScholarCrossref
    34.
    Li  XQ, Björkman  A, Andersson  TB, Gustafsson  LL, Masimirembwa  CM.  Identification of human cytochrome P(450)s that metabolise anti-parasitic drugs and predictions of in vivo drug hepatic clearance from in vitro data.   Eur J Clin Pharmacol. 2003;59(5-6):429-442. doi:10.1007/s00228-003-0636-9PubMedGoogle ScholarCrossref
    35.
    McLachlan  AJ, Tett  SE, Cutler  DJ, Day  RO.  Bioavailability of hydroxychloroquine tablets in patients with rheumatoid arthritis.   Br J Rheumatol. 1994;33(3):235-239. doi:10.1093/rheumatology/33.3.235PubMedGoogle ScholarCrossref
    36.
    McLachlan  AJ, Tett  SE, Cutler  DJ, Day  RO.  Absorption and in vivo dissolution of hydroxycholoroquine in fed subjects assessed using deconvolution techniques.   Br J Clin Pharmacol. 1993;36(5):405-411. doi:10.1111/j.1365-2125.1993.tb00388.xPubMedGoogle ScholarCrossref
    37.
    Tett  SE, Cutler  DJ, Day  RO, Brown  KF.  Bioavailability of hydroxychloroquine tablets in healthy volunteers.   Br J Clin Pharmacol. 1989;27(6):771-779. doi:10.1111/j.1365-2125.1989.tb03439.xPubMedGoogle ScholarCrossref
    38.
    Cheung  KWK, van Groen  BD, Spaans  E,  et al.  A comprehensive analysis of ontogeny of renal drug transporters: mRNA analyses, quantitative proteomics, and localization.   Clin Pharmacol Ther. 2019;106(5):1083-1092. doi:10.1002/cpt.1516PubMedGoogle ScholarCrossref
    39.
    Willmann  S, Höhn  K, Edginton  A,  et al.  Development of a physiology-based whole-body population model for assessing the influence of individual variability on the pharmacokinetics of drugs.   J Pharmacokinet Pharmacodyn. 2007;34(3):401-431. doi:10.1007/s10928-007-9053-5PubMedGoogle ScholarCrossref
    40.
    Sciensano Epidemiology of Infectious Disease. Interim clinical guidance for adults with suspected or confirmed COVID-19 in Belgium, Version 7. sciensano website. Published April 7, 2020. Accessed April 14, 2020. https://epidemio.wiv-isp.be/ID/Documents/Covid19/COVID-19_InterimGuidelines_Treatment_ENG.pdf
    41.
    Michigan Medicine, University of Michigan. Inpatient guidance for treatment of COVID-19 in adults and children. Michigan Medicine, University of Michigan website. Accessed April 14, 2020. http://www.med.umich.edu/asp/pdf/adult_guidelines/COVID-19-treatment.pdf
    42.
    Kramer  NI, Krismartina  M, Rico-Rico  A, Blaauboer  BJ, Hermens  JL.  Quantifying processes determining the free concentration of phenanthrene in Basal cytotoxicity assays.   Chem Res Toxicol. 2012;25(2):436-445. doi:10.1021/tx200479kPubMedGoogle ScholarCrossref
    43.
    Postnikova  E, Cong  Y, DeWald  LE,  et al.  Testing therapeutics in cell-based assays: factors that influence the apparent potency of drugs.   PLoS One. 2018;13(3):e0194880. doi:10.1371/journal.pone.0194880PubMedGoogle Scholar
    44.
    Liu  J, Cao  R, Xu  M,  et al.  Hydroxychloroquine, a less toxic derivative of chloroquine, is effective in inhibiting SARS-CoV-2 infection in vitro.   Cell Discov. 2020;6:16. doi:10.1038/s41421-020-0156-0PubMedGoogle ScholarCrossref
    45.
    Zhou  D, Dai  SM, Tong  Q.  COVID-19: a recommendation to examine the effect of hydroxychloroquine in preventing infection and progression.   J Antimicrob Chemother. 2020:dkaa114. doi:10.1093/jac/dkaa114PubMedGoogle Scholar
    46.
    Arnold  SLM, Buckner  F.  Hydroxychloroquine for treatment of SARS-CoV-2 infection? improving our confidence in a model-based approach to dose selection.   Clin Transl Sci. 2020;. doi:10.1111/cts.12797PubMedGoogle Scholar
    47.
    Balevic  SJ, Green  TP, Clowse  MEB, Eudy  AM, Schanberg  LE, Cohen-Wolkowiez  M.  Pharmacokinetics of hydroxychloroquine in pregnancies with rheumatic diseases.   Clin Pharmacokinet. 2019;58(4):525-533. doi:10.1007/s40262-018-0712-zPubMedGoogle ScholarCrossref
    48.
    Morita  S, Takahashi  T, Yoshida  Y, Yokota  N.  Population pharmacokinetics of hydroxychloroquine in Japanese patients with cutaneous or systemic lupus erythematosus.   Ther Drug Monit. 2016;38(2):259-267. doi:10.1097/FTD.0000000000000261PubMedGoogle ScholarCrossref
    49.
    Tegenge  MA, Mahmood  I.  Age- and bodyweight-dependent allometric exponent model for scaling clearance and maintenance dose of theophylline from neonates to adults.   Ther Drug Monit. 2018;40(5):635-641. doi:10.1097/FTD.0000000000000543PubMedGoogle ScholarCrossref
    50.
    Lee  JY, Vinayagamoorthy  N, Han  K,  et al.  Association of polymorphisms of cytochrome P450 2D6 with blood hydroxychloroquine levels in patients with systemic lupus erythematosus.   Arthritis Rheumatol. 2016;68(1):184-190. doi:10.1002/art.39402PubMedGoogle ScholarCrossref
    51.
    Projean  D, Baune  B, Farinotti  R,  et al.  In vitro metabolism of chloroquine: identification of CYP2C8, CYP3A4, and CYP2D6 as the main isoforms catalyzing N-desethylchloroquine formation.   Drug Metab Dispos. 2003;31(6):748-754. doi:10.1124/dmd.31.6.748PubMedGoogle ScholarCrossref
    ×