Artemisinin antimalarials moderately affect cytochrome P450 enzyme activity in healthy subjects

doi: 10.1111/j.1472-8206.2007.00471.x

O R I G I N A L

A R T I C L E

Artemisinin antimalarials moderately affectcytochrome P450 enzyme activity in healthysubjects

Sara Asimusa, Doaa Elsherbinyb, Trinh N. Haic, Britt Janssonb,Nguyen V. Huongc, Max G. Petzoldd, Ulrika S.H. Simonssonb,Michael Ashtona*aUnit for Pharmacokinetics and Drug Metabolism, Department of Pharmacology, Sahlgrenska Academy at Goteborg

University, Box 431, 405 30 Goteborg, SwedenbDivision of Pharmacokinetics and Drug Therapy, Department of Pharmaceutical Biosciences, Uppsala University,

Box 591, 751 24 Uppsala, SwedencNational Institute of Malariology, Parasitology and Entomology, B.C. 10 200 Tu-liem, Hanoi, VietnamdNordic School of Public Health, Box 121 33, 402 42 Goteborg, Sweden

Keywords

artemisinin,

cytochrome P450,

induction,

inhibition,

malaria,

metabolism

Received 11 September 2006;

revised 30 October 2006;

accepted 14 December 2006

*Correspondence and reprints:

[email protected]

A B S T R A C T

The aim of this study was to investigate which principal human cytochrome P450

(CYP450) enzymes are affected by artemisinin and to what degree the artemisinin

derivatives differ with respect to their respective induction and inhibition capacity.

Seventy-five healthy adults were randomized to receive therapeutic oral doses of

artemisinin, dihydroartemisinin, arteether, artemether or artesunate for 5 days

(days 1–5). A six-drug cocktail consisting of caffeine, coumarin, mephenytoin,

metoprolol, chlorzoxazone and midazolam was administered orally on days )6, 1,

5 and 10 to assess the activities of CYP1A2, CYP2A6, CYP2C19, CYP2D6,

CYP2E1 and CYP3A, respectively. Four-hour plasma concentrations of parent

drugs and corresponding metabolites and 7-hydroxycoumarin urine concentrations

were quantified by liquid chromatography-tandem mass spectrometry. The 1-

hydroxymidazolam/midazolam 4-h plasma concentration ratio (CYP3A) was

increased on day 5 by artemisinin [2.66-fold (98.75% CI: 2.10–3.36)], artemether

[1.54 (1.14–2.09)] and dihydroartemisinin [1.25 (1.06–1.47)] compared with day

)6. The S-4¢-hydroxymephenytoin/S-mephenytoin ratio (CYP2C19) was increased

on day 5 by artemisinin [1.69 (1.47–1.94)] and arteether [1.33 (1.15–1.55)]

compared with day )6. The paraxanthine/caffeine ratio (CYP1A2) was decreased

on day 1 after administration of artemisinin [0.27 (0.18–0.39)], arteether [0.70

(0.55–0.89)] and dihydroartemisinin [0.73 (0.59–0.90)] compared with day )6.

The a-hydroxymetoprolol/metoprolol ratio (CYP2D6) was lower on day 1

compared with day )6 in the artemisinin [0.82 (0.70–0.96)] and dihydroartem-

isinin [0.83 (0.71–0.96)] groups, respectively. In the artemisinin-treated subjects

this decrease was followed by a 1.34-fold (1.14–1.58) increase from day 1 to day

5. These results show that intake of artemisinin antimalarials affect the activities of

several principal human drug metabolizing CYP450 enzymes. Even though not

significant in all treatment groups, changes in the individual metrics were of the

same direction for all the artemisinin drugs, suggesting a class effect that needs to

be considered in the development of new artemisinin derivatives and combination

treatments of malaria.

ª 2007 The Authors Journal compilation ª 2007 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 21 (2007) 307–316 307

I N T R O D U C T I O N

The artemisinin class of endoperoxides represents a

relatively new, important class of antimalarial drugs.

Repeated oral and rectal administration of the class

parent, artemisinin, has shown remarkable time-

dependent pharmacokinetics in both healthy subjects

and patients [1–3]. Declining plasma concentrations

during multiple dosing have also been reported for the

derivatives artemether and, less convincingly, artesu-

nate [4,5]. A pronounced, unusual capacity for auto-

induction of drug metabolism has been suggested to be

the explanation of this time-dependency [6]. In addi-

tion to the auto-induction phenomenon, artemisinin

has also demonstrated a capacity to increase the

metabolism of other drugs mediated by several different

cytochrome P450 (CYP450) enzymes. A twofold

increase in omeprazole metabolism to 5-hydroxyomep-

razole, mediated by a CYP2C19 pathway, has been

reported when co-administered with artemisinin [7].

An increased metabolism of S-mephenytoin to S-

nirvanol by CYP2B6 has also been reported [8]. The

induction mechanism appears to be by activation of

nuclear receptors PXR and/or CAR [9]. In addition,

artemisinin is suggested to be an inhibitor of drug

metabolism. In an in vitro screening study, both

artemisinin and dihydroartemisinin were found to be

potent inhibitors of CYP1A2 and CYP2C19, while no

inhibitory effect of artesunate was observed [10]. The

inhibitory effect of artemisinin on CYP1A2 activity has

recently been confirmed in healthy subjects [11]. New

synthetic endoperoxides, such as the trioxolanes, have

also shown potential for CYP450 inhibition. Results

from in vitro studies indicate that at least CYP3A4,

CYP2C9 and CYP2C19 are inhibited by these com-

pounds [12].

It is not known to what extent a similar capacity for

induction or inhibition also applies to the other deriva-

tives in clinical use (dihydroartemisinin, its pro-drug

artesunate, artemether and arteether). The effect of these

drugs on the activity of the principal CYP450 enzymes

involved in drug metabolism has not yet been studied.

With a potential for induction and inhibition of drug

metabolism and the introduction of artemisinin-based

combination treatment as first-line therapy of falciparum

malaria, the risk for drug–drug interactions resulting in

a potentiated or diminished activity of other drugs

substantially increases. It is therefore highly important

to elucidate which principal CYP450 enzymes are

affected by these drugs.

Administration of a probe compound has been a

common way of estimating in vivo metabolic activity in

humans. A probe drug is selected on the basis that a

quantifiable route of its metabolism is mostly or exclu-

sively mediated by one individual enzyme.

A cocktail approach, which involves the administra-

tion of a number of probe substances simultaneously,

offers the possibility of obtaining information about

several enzymes in one study. By using metabolic ratios

such as the concentration of a metabolite to that of the

parent drug in urine, plasma or saliva as metrics, the

activity of multiple enzymes can be estimated concur-

rently [13–16].

The aim of this study was to investigate the ability of

the artemisinin class of endoperoxides to either induce

and/or inhibit principal CYP450 enzymes. A secondary

objective was to compare the potential for drug–drug

interactions within the class, as a means to select the

most suitable artemisinin derivative to be a partner in

combination treatment. A probe cocktail consisting of

caffeine, coumarin, mephenytoin, metoprolol, chlorzoxa-

zone and midazolam was administered as single oral

doses to simultaneously assess the in vivo activities of

CYP1A2, CYP2A6, CYP2C19, CYP2D6, CYP2E1 and

CYP3A, respectively, before and after repeated adminis-

tration of artemisinin, dihydroartemisinin, arteether,

artemether or artesunate.

M A T E R I A L S A N D M E T H O D S

Ethics

The study was reviewed and approved by the Ministry of

Health Hanoi, Vietnam and separately approved by the

ethics committee of Goteborg University, Goteborg,

Sweden and the Swedish Medical Products Agency,

Uppsala, Sweden. Investigations were conducted in

accordance with the Helsinki Declaration and conformed

to the standard established for Good Clinical Practices.

Participants were given a detailed description of the

study and their written informed consent was obtained

prior to study inclusion. The study was carried out at the

Institute of Malariology, Parasitology and Entomology,

Hanoi, Vietnam in September 2002.

Subjects

Seventy-five healthy volunteers, 51 men and 24 women,

were included in this open randomized parallel group

study. Seventy-two of the subjects belonged to the

predominant Kinh ethnic group, the other three subjects

were of Thai ethnicity. The participants were all judged

308 S. Asimus et al.

ª 2007 The Authors Journal compilation ª 2007 Blackwell Publishing Ltd. Fundamental & Clinical Pharmacology 21 (2007) 307–316

to be healthy on the basis of medical history, physical

examination and routine clinical laboratory determina-

tions. Subjects who had taken any antimalarial drug

within 1 month or any other drug within 2 weeks before

the start of the study and those with a history of alcohol

abuse were excluded from the study. Oral contraceptive

use was not recorded. Thirty-six of the subjects were

smokers of no more than 10 cigarettes per day. All

smokers were men.

Study design

The volunteers were randomized to treatment with one

of five different artemisinin study drugs; artemisinin

(2 · 250 mg capsules; Mediplantex, Hanoi, Vietnam),

dihydroartemisinin (60 mg tablets; Mediplantex), ar-

temether (Artenam�, 2 · 50 mg tablets; Arenco Phar-

maceutica nv, Geel, Belgium), arteether (Artecef�,

2 · 1 mL ampoules of the intramuscular preparation

containing 50 mg arteether in sesame oil; Artecef BV,

Maarsen, The Netherlands) or artesunate (Plasmo-

trim�, 2 · 50 mg tablets; Mepha Ltd, Aesch, Switzer-

land). Repeated oral doses were administered in the

morning for 5 days (day 1–5). The cocktail of probe

drugs, caffeine (Koffein Recip�, 100 mg tablets, Stock-

holm, Sweden), coumarin (5 mg capsules, Laboratory

of the National Corporation of Swedish Pharmacies,

Goteborg, Sweden), midazolam (Dormicum�, 7.5 mg

tablets; Roche a/s, Hvidovre, Denmark), mephenytoin

(Epilan Gerot�, 100 mg tablets; Gerot Pharmazeutica,

Vienna, Austria), metoprolol (Seloken�, 100 mg tab-

lets; AstraZeneca Sverige, Molndal, Sweden) and

chlorzoxazone (Paraflex�, 250 mg tablets, AstraZeneca

Sverige) were given orally 1 week before (day )6)

administration of the artemisinin drugs. On day 1 and

day 5, the administration of the probe drug cocktail

was repeated, 1 h after intake of the artemisinin

drugs. After a washout period of 5 days (day 10),

the probe drug cocktail was given again. Study

drug and the cocktail were each taken with 200 mL

water. All drug intakes were monitored by one of

the study investigators. The subjects fasted from the

evening before blood and urine sampling (days )6, 1,

5 and 10) and until 2 h after last drug intake (total

time of fasting approximately 12–14 h). No caffeine-

containing drinks were permitted during these days.

Subjects were allowed to smoke less than 10 cigarettes

and to drink less than 2 non-spirituous alcoholic

beverages per day 1 week before and during the whole

study. No other alcoholic drinks were permitted during

the study.

Clinical assessments

A physical examination was performed on days )11 and

15. Blood for biochemical analysis (total bilirubin,

creatinine, AST and ALT) and hematological tests

(hemoglobin, erythrocytes and leucocytes) was taken

on day )11. On this day an electrocardiogram was

performed on each subject and female subjects were

tested for pregnancy (Quickstick, Pharmatech, San

Diego, CA, USA). Subjects were interviewed about

adverse events on days 5 and 15. The interviews were

initially open, without any leading questions, and were

followed by specific questions according to a checklist.

Blood and urine collection

Probe compounds were measured in samples obtained

pre-dose and at 4 h after intake of the cocktail drugs on

days )6, 1, 5 and 10. Ten milliliters of blood was

collected into a Li-heparinized Vacutainer vial (Becton

Dickinson, Plymouth, UK) by venepuncture, and after

5 min centrifuged at 3000 g for 5 min after which the

plasma was transferred into cryotubes (Nunc, Roskilde,

Denmark). All plasma samples were immediately frozen

and kept at )70 �C, until transport on dry ice to Sweden.

On days )6, 1, 5 and 10, total voided urine was collected

for 8 h after administration of the cocktail drugs into

receptacles containing 1 g ascorbic acid. Total weight of

each urine sample voided was recorded and an aliquot

was kept at )70 �C, until transportation on dry ice to

Sweden.

Plasma assay

Blank plasma was obtained from the University Hospital

Blood Bank, Uppsala, Sweden. The plasma was spiked

with a mix of the analytes from stock solutions to get

seven to nine standard concentrations and three or four

quality control (QC) levels for each compound. Calibra-

tion curves were constructed using linear regression of

the analyte peak area (y) vs. the added concentration (x)

with a weighting factor of 1/y2. The curves were not

forced through the origin. The ranges of the standard

concentrations were 30–6000 ng/mL for caffeine

and paraxanthine, 40–8000 ng/mL for chlorzoxazone,

30–3000 ng/mL for 6-hydroxychlorzoxazone (6-OH-

chlorzoxazone), 1–100 ng/mL for 7-hydroxycoumarin

(7-OH-coumarin), 3–1500 ng/mL for S-mephenytoin,

1–500 ng/mL for S-4¢-hydroxymephenytoin (S-4¢-OH-

mephenytoin), 5–500 ng/mL for metoprolol and 2–

200 ng/mL for a-hydroxymetoprolol (a-OH-metoprolol),

midazolam and 1-hydroxymidazolam (1-OH-mida-

zolam).

Artemisinin drugs affect in vivo CYP450 activity 309


Plasma concentrations of caffeine, paraxanthine,

chlorzoxazone, 6-OH-chlorzoxazone, 7-OH-coumarin,

metoprolol, a-OH-metoprolol, midazolam and 1-OH-mid-

azolam were measured in b-glucuronidase-treated sam-

ples by use of a liquid chromatography-tandem mass

spectrometry (LC/MS/MS) method according to Scott

et al. [17] with some modifications. The system consisted

of two pumps (LC-10AD; Shimadzu, Kyoto, Japan) with a

high-pressure gradient mixer, an auto-sampler (Triath-

lon; Spark, Holland, The Netherlands) equipped with a

10 lL loop, a reversed phase column (HyPurity C18;

Thermo Hypersil-Keystone, Bellefonte, PA, USA), 3 lm

particle size, 50 · 4.6 mm protected by a 10 · 4 mm

guard column of the same material, and a triple

quadrupole mass spectrometer (Quattro Ultima; Micro-

mass, Manchester, UK). MS control and spectral process-

ing were performed using MassLynx software, version 4.0

(Micromass). The analytes were separated using a gra-

dient at a flow rate of 800 lL/min with the mobile phase

changing from 95% A (A; 0.05% formic acid in water) to

95% B (B; 70% acetonitrile in 0.05% formic acid) during

4.5 min and then 0.5 min later changing back to 95% A.

The flow from the column was split to 200 lL/min before

entering the source of the mass spectrometer.

The unknown plasma samples, standards and QCs

(250 lL) were incubated with b-glucuronidase solution

(250 lL) and sodium acetate buffer pH 4.75 (500 lL) at

37 �C for 3 h. The incubated plasma (200 lL) was

precipitated with acetonitrile (400 lL) by vortex mixing.

After centrifugation, 150 lL of the supernatant was

evaporated under nitrogen at 40 �C. The residue

was dissolved in 250 lL mobile phase A and 10 lL

was injected onto the column.

The plasma concentrations of S-mephenytoin and

S-4¢-OH-mephenytoin were determined by a separate LC/

MS/MS method [18]. The enantiomers were separated

on a chiral a1-acid glycoprotein column, 150 · 4 mm

(Chromtech, Hagersten, Sweden). The mobile phase

consisted of 2% acetonitrile in 5 mM ammonium acetate

and the flow rate of 900 lL/min was split to 230 lL/min

before entering the mass spectrometer.

Intra-day precision (coefficient of variation, CV),

determined by analyzing six replicates of the QCs and

five to six replicates of the lowest standard concentration

(lower limit of quantification) according to each of the

two described methods, were below 20% for all com-

pounds and levels. The inter-day precision was deter-

mined by analyzing the QCs (duplicates of ‡3 levels)

interspersed with the study samples on each day during

the routine analysis. The median value of the inter-day

precision of all QC levels was 6.5% (n ¼ 20 or 21 per

compound and level). For none of the analytes was the

CV above 16%.

Urine assay

Blank urine, from five healthy volunteers, was spiked

with 7-OH-coumarin to get seven standard concentra-

tions in the range 50–10 000 ng/mL and three QC

levels. The urine concentrations were determined with

the same LC/MS/MS method as for 7-OH-coumarin in

plasma with some modification. The analyte was

retained on the column with a mobile phase containing

20% acetonitrile in 0.05% formic acid. No gradient was

used. The unknown urine samples, standards and QCs

(100 lL) were incubated with b-glucuronidase solution

(100 lL) and sodium acetate buffer, pH 4.75 (200 lL) at

37 �C for 3 h. The incubated urine was diluted 10 times

with mobile phase and 10 lL of the dilution was injected

directly onto the column.

The intra- and inter-day precision (CV) were deter-

mined as described for the two plasma assays. The intra-

day precision were below 4.1% and inter-day precision

below 5.3% for the three QC levels (n ¼ 6).

Data analysis

The plasma concentration ratio of paraxanthine to

caffeine at 4 h was used to assess CYP1A2 activity

[19]. The total amount of 7-OH-coumarin excreted in

0- to 8-h urine was used as an index of CYP2A6

activity [20]. The ability to hydroxylate S-mephenytoin

(CYP2C19) was estimated by the S-4¢-OH-mephenytoin

to S-mephenytoin plasma ratio at 4 h post-dose. The 4-h

a-OH-metoprolol to metoprolol ratio and 6-OH-chlorz-

oxazone to chlorzoxazone ratio were used to evaluate the

activities of CYP2D6 and CYP2E1, respectively [13,21].

The concentration of 1-OH-midazolam divided by that of

midazolam in the 4-h plasma sample was used to

indicate CYP3A activity [22].

The metrics for the individual enzyme activities were

calculated from metabolite and drug concentrations

obtained on days )6, 1, 5 and 10. For comparison of

enzyme activity between study days the following four

contrasts were estimated in the statistical evaluation;

day 1 vs. day )6 (day 1/day )6), day 5 vs. day )6 (day

5/day )6), day 5 vs. day 1 (day 5/day 1) and day 10 vs.

day )6 (day 10/day )6).

Statistical analysis

A repeated ANOVA model with Gaussian random effects

was applied to the log-transformed (natural base) data



with subsequent model-diagnostics by checking residual

plots. An overall test level of 5% for the multiple (four)

tests per treatment group was selected. In accordance

with the Bonferroni-method for multiple testing, 98.75%

confidence intervals (CI) are presented and P-values are

compared to 0.0125 in the sequel. The Proc Mixed in

SAS 8.2 (SAS Company Inc., Cary, NC, USA) software

was used for the analysis. Results are presented as

quotients (mean, 98.75% CI) based on anti-logarithms of

the contrasts for the different occasions. Quotients larger,

or smaller, than unity may be associated with enzyme

induction or inhibition, respectively.

R E S U L T S

Subject demographic characteristics are summarized in

Table I. Seventy-four subjects completed the study. One

subject discontinued the study due to nausea on the first

day of cocktail drug intake. Apart from a brief period of

sleepiness and dizziness after administration of the

cocktail drugs in many subjects, the cocktail procedure

was well tolerated. On the fifth day of artemether intake,

one subject reported a skin reaction (rash) of moderate

severity, which gradually cleared within 3 weeks.

The effects of the artemisinin antimalarials on the

major CYP450 enzymes are summarized in Table II. Five

days intake of artemisinin, artemether and dihydro-

artemisinin significantly increased the index for CYP3A

activity. Day 5 compared with day )6, the mean 1-OH-

midazolam/midazolam concentration ratio at 4 h

increased 2.66-fold (CI: 2.10–3.36), 1.54 (1.14–2.09)

and 1.25 (1.06–1.47) by artemisinin, artemether and

dihydroartemisinin, respectively. In the artemisinin

group a 1.60-fold (1.26–2.02) increase was seen in the

same ratio on day 1 compared with day )6, and in the

artesunate group the ratio was significantly increased on

day 10 compared with day )6 (1.26, 1.01–1.57).

A significant increase in the mean 4-h S-4¢-OH-

mephenytoin/S-mephenytoin plasma concentration

ratio (CYP2C19) was observed from day )6 to day 5 in

two of the five treatment groups; artemisinin 1.69-fold

(1.47–1.94) and arteether 1.33 (1.15–1.55) (Figure 1).

Nine individuals had no measurable S-4¢-OH-mephenyt-

oin concentrations and were considered to be poor

metabolizers of CYP2C19, and hence excluded from the

data analysis.

Intake of artemisinin, dihydroartemisinin and arte-

ether significantly decreased the mean 4-h paraxan-

thine/caffeine plasma concentration ratio. The mean

ratios decreased to 0.27 (0.18–0.39), 0.70 (0.55–0.89)

and 0.73 (0.59–0.90) on day 1 compared with day )6,

in subjects receiving artemisinin, arteether and dihyd-

roartemisinin groups, respectively. A significant 2.22-

fold increase (1.54–3.21) was observed in the CYP1A2

metric day 5 compared with day 1 after multiple

administration of artemisinin. Residual concentrations

of caffeine and paraxanthine were found in the pre-dose

samples from many subjects.

The marker for CYP2D6 activity, a-OH-metoprolol/

metoprolol concentration ratio at 4 h, significantly

decreased to 0.82 (0.70–0.96) and 0.83 (0.71–0.96)

day 1 compared with day )6 in the artemisinin and

dihydroartemisinin groups, respectively. A significant

1.34-fold increase (1.14–1.58) was found in the same

ratio from day 1 to day 5 in the artemisinin group.

Two subjects, who had undetectable a-OH-metoprolol

concentrations, were considered to be poor metabolizers

of CYP2D6 and excluded from the data analysis.

Table I Subject demographics.

Artemisinin

(n ¼ 15)

Dihydroartemisinin

(n ¼ 14)

Arteether

(n ¼ 15)

Artemether

(n ¼ 15)

Artesunate

(n ¼ 15)

Gender

Male/female 7/8 13/1 13/2 8/7 10/5

Agea (years) 28 ± 8 (20–44) 30 ± 9 (20–45) 25 ± 6 (18–40) 31 ± 8 (21–45) 30 ± 9 (18–45)

Body weighta (kg) 52 ± 7 (41–62) 57 ± 9 (47–71) 53 ± 6 (45–65) 55 ± 7 (43–68) 56 ± 7 (45–69)

Number of smokers (n) 5 8 8 8 7

Average number of daily cigarettes among smokers 4 6 4 5 6

Ethnic group

Kinhb/Thai 15/0 13/1 14/1 15/0 14/1

Number of poor metabolizersc of CYP2C19/CYP2D6 2/0 1/1 1/0 2/1 3/0

aMean values ± SD (range).bThe most dominant ethnic group in Vietnam.cDefined as subjects with no measurable metabolite concentrations.



The 6-OH-chlorzoxazone/chlorzoxazone concentra-

tion ratio (CYP2E1) at 4 h was not affected apart from

a decrease to 0.68 (0.54–0.86) and 0.74 (0.58–0.94)

from day )6 to day 1 and day )6 to day 5 in the

artemisinin group, respectively. No significant change in

the amount of 7-OH-coumarin excreted in 0–8 h urine

was observed in any group. Coumarin plasma concen-

trations were below the detection limit, whereas 7-OH-

coumarin plasma concentrations exhibited no significant

changes between days (data not shown).

D I S C U S S I O N

This study was designed to investigate whether the

capacity of several principal drug metabolizing CYP450

enzymes are affected by artemisinin and its major

derivatives. Artemisinin has previously been reported to

induce CYP2C19 but not CYP3A4 activity in healthy

subjects [23]. Total urinary recovery of S-4¢-OH-meph-

enytoin and a S/R-mephenytoin ratio in urine have been

Table II Phenotyping metrics in the five different treatment groups [artemisinin (ART), dihydroartemisinin (DHA), arteether (ARE),

artemether (ARM) and artesunate (AS)]. The presented quotients (mean, 98.75% CI) are based on anti-logarithms of the contrast for the

different occasions.

Enzyme Phenotyping metric

Quotients of metric for

different occasionsb ART DHA ARE ARM AS

CYP1A2 Paraxanthine/caffeine

4-h concentration

ratio

day 1/day )6 0.27 (0.18–0.39)a 0.73 (0.59–0.90)a 0.70 (0.55–0.89)a 0.83 (0.69–1.02) 0.87 (0.69–1.09)

day 5/day )6 0.59 (0.41–0.85)a 0.85 (0.69–1.06) 0.70 (0.55–0.89)a 0.81 (0.67–0.98)a 1.00 (0.80–1.26)

day 5/day 1 2.22 (1.54–3.21)a 1.17 (0.95–1.45) 1.00 (0.78–1.27) 0.97 (0.80–1.18) 1.16 (0.92–1.45)

day 10/day )6 1.26 (0.88–1.81) 0.94 (0.76–1.16) 0.84 (0.66–1.06) 1.06 (0.87–1.30) 1.10 (0.88–1.38)

CYP2A6 7-OH-coumarin excreted

in 0- to 8-h urine

day 1/day )6 0.74 (0.40–1.40) 1.17 (0.73–1.88) 0.81 (0.38–1.71) 1.01 (0.63–1.62) 0.73 (0.38–1.44)

day 5/day )6 0.87 (0.48–1.60) 1.34 (0.84–2.14) 0.95 (0.45–2.02) 0.91 (0.57–1.45) 0.60 (0.30–1.17)

day 5/day 1 1.17 (0.62–2.23) 1.15 (0.71–1.85) 1.18 (0.56–2.51) 0.90 (0.56–1.44) 0.81 (0.41–1.61)

day 10/day )6 0.96 (0.53–1.74) 1.38 (0.87–2.19) 1.17 (0.55–2.47) 1.22 (0.77–1.94) 0.86 (0.44–1.68)

CYP2C19 S-4¢-OH-mephenytoin/

S-mephenytoin 4-h

concentration ratio

day 1/day )6 0.95 (0.83–1.09) 0.97 (0.78–1.21) 0.93 (0.80–1.08) 0.95 (0.79–1.14) 0.91 (0.73–1.14)

day 5/day )6 1.69 (1.47–1.94)a 1.16 (0.93–1.44) 1.33 (1.15–1.55)a 1.20 (1.00–1.44) 1.12 (0.89–1.40)

day 5/day 1 1.77 (1.54–2.04)a 1.19 (0.96–1.49) 1.44 (1.24–1.67)a 1.26 (1.05–1.52)a 1.22 (0.98–1.53)

day 10/day )6 1.65 (1.44–1.88)a 1.13 (0.91–1.41) 1.26 (1.08–1.46)a 1.14 (0.94–1.38) 1.18 (0.94–1.49)

CYP2D6 a)OH-metoprolol/

metoprolol 4-h

concentration ratio

day 1/day )6 0.82 (0.70–0.96)a 0.83 (0.71–0.96)a 0.89 (0.75–1.05) 0.90 (0.76–1.05) 0.90 (0.79–1.04)

day 5/day )6 1.10 (0.94–1.29) 0.95 (0.81–1.10) 1.02 (0.86–1.21) 0.97 (0.82–1.13) 1.02 (0.89–1.18)

day 5/day 1 1.34 (1.14–1.58)a 1.14 (0.99–1.33) 1.15 (0.97–1.37) 1.08 (0.92–1.27) 1.13 (0.99–1.30)

day 10/day )6 1.15 (0.98–1.34) 0.93 (0.80–1.08) 0.98 (0.83–1.17) 0.92 (0.78–1.09) 1.07 (0.93–1.24)

CYP2E1 6-OH-chlorzoxazone/

chlorzoxazone 4-h

concentration ratio

day 1/day )6 0.68 (0.54–0.86)a 0.93 (0.66–1.31) 1.13 (0.84–1.51) 1.06 (0.85–1.33) 0.96 (0.73–1.26)

day 5/day )6 0.74 (0.58–0.94)a 1.00 (0.70–1.41) 0.99 (0.74–1.32) 1.08 (0.86–1.35) 1.09 (0.83–1.43)

day 5/day 1 1.08 (0.85–1.38) 1.07 (0.76–1.52) 0.88 (0.66–1.17) 1.02 (0.81–1.28) 1.13 (0.86–1.48)

day 10/day )6 0.90 (0.71–1.14) 0.83 (0.59–1.17) 1.05 (0.78–1.42) 1.07 (0.85–1.35) 1.03 (0.79–1.36)

CYP3A 1-OH-midazolam/

midazolam 4-h

concentration ratio

day 1/day )6 1.60 (1.26–2.02)a 1.11 (0.94–1.30) 0.97 (0.79–1.20) 1.22 (0.90–1.65) 1.17 (0.94–1.47)

day 5/day )6 2.66 (2.10–3.36)a 1.25 (1.06–1.47)a 1.16 (0.94–1.43) 1.54 (1.14–2.09)a 1.25 (1.00–1.56

day 5/day 1 1.67 (1.31–2.12)a 1.13 (0.96–1.33) 1.19 (0.97–1.47) 1.27 (0.93–1.72) 1.06 (0.85–1.33)

day 10/day )6 1.25 (0.99–1.58) 1.16 (0.98–1.36) 1.12 (0.90–1.38) 1.15 (0.84–1.57) 1.26 (1.01–1.57)a

aP < 0.0125 (a adjusted for multiple testing).bQuotients >1 indicate increased enzyme activity, quotients <1 indicate decreased enzyme activity.

0

2

4

6

8

10

12

DayS-4

'-OH

-mep

hen

yto

in/S

-mep

hen

yto

in 4

hp

lasm

a co

nce

ntr

atio

n r

atio

–6 –4 –2 0 2 4 6 8 10

ARE

Figure 1 S-4¢-OH-mephenytoin/S-mephenytoin plasma concentra-

tion ratio measured on days )6, 1, 5 and 10 four hours after oral

intake of a six-drug cocktail in 14 healthy subjects (extensive

metabolizers of CYP2C19) treated with 100 mg arteether orally for

5 days (days 1–5). The significant (P < 0.0001) increase in the ratio

between days 1 and 5 indicating induction of CYP2C19 by arteether.



widely used to assess CYP2C19 activity [13–16].

Artemisinin has previously been reported to increase

oral clearance of both R- and S-mephenytoin [8],

confounding their ratio as a metric for CYP2C19 activity.

The S/R-mephenytoin ratios on day 1, 5 and 10 would

have been further confused by the presence of R-

mephenytoin concentrations remaining from the previ-

ous administrations, such as was observed in pre-dose

samples on these days. Therefore the S-4¢-OH-mephenyt-

oin/S-mephenytoin concentration ratio, although not a

pre-validated metric, was considered the best choice to

estimate CYP2C19 activity. Moreover, a metabolite/drug

plasma concentration ratio is a more specific metric for

enzyme activity than total recovery of a metabolite in

urine. In the present study a moderate but significant

increase in the S-4¢-OH-mephenytoin/S-mephenytoin

concentration ratio was found after artemisinin and

arteether intake. In these two groups the elevated

CYP2C19 activity was not normalized after 4 days

wash-out and was still significantly increased on day 10.

The 1-OH-midazolam/midazolam ratio increased in the

artemisinin, dihydroartemisinin and artemether groups

on day 5 compared with day )6. This suggests, in contrast

to previous findings with omeprazole as a probe drug and

urinary excretion ratio of endogenous 6b-hydroxycorti-

sol/cortisol as an additional marker for CYP3A4, that the

artemisinin antimalarials are inducers of CYP3A. Mida-

zolam is probably the best available probe to test CYP3A4,

even though it is not selective for CYP3A4 vs. CYP3A5

[24]. The results from the present study are supported by

the recent finding that artemisinin induces the expression

of CYP3A4 in primary human hepatocytes and in a

human intestinal cell line [9]. Interestingly, the increased

CYP3A activity by artemisinin administration was

observed even after one single dose (day 1 vs. day )6).

An interaction between chlorzoxazone and midazolam

has been reported when the two probes were used in the

same cocktail [25]. The interaction was thought to result

from chlorzoxazone inhibiting first-pass metabolism of

midazolam by CYP3A in the gut. No effect was seen on

hepatic CYP3A activity. Both gut wall and liver CYP3A

have been suggested to contribute substantially to the

metabolism of midazolam [26]. If the artemisinin drugs

induce intestinal CYP3A activity, this effect might not be

reflected in the present study. The increase in the 1-OH-

midazolam/midazolam concentration ratio would then be

a result of an induction of hepatic CYP3A activity only and

hence total induction could be underestimated.

The inhibitory effect on CYP1A2 by artemisinin, dihyd-

roartemisinin and arteether, indicated by a decrease in the

paraxanthine/caffeine ratio between day )6 and day 1, is

consistent with previously reported in vitro findings and

results from a recent study in healthy volunteers [10,11].

In the present study, a significant increase was observed in

the CYP1A2 index from day 1 to day 5 in the artemisinin

group, suggesting an induced activity of CYP1A2, an effect

not seen for any of the other derivatives. However, residual

concentrations of caffeine and paraxanthine in the pre-

dose samples in the present study confound the CYP1A2

metric in many subjects. Similar problems were found in

the Karolinska cocktail such that the authors state that

16 h of caffeine abstinence may be too short to avoid

measuring paraxanthine remaining in the body from

previous caffeine intake and suggest this period to be

extended to 36 h in future studies [27].

As CYP2D6 is considered to be a non-inducible enzyme,

the increase in the a-OH-metoprolol/metoprolol concen-

tration ratio between days 1 and 5 in the artemisinin

group was unexpected. The effect might be explained by an

induction of an enzyme normally not of great importance

for the formation of a-OH-metoprolol from metoprolol.

There were no indications of an induction of CYP2A6

in any of the treatment groups. Coumarin is the most

common probe for studying CYP2A6 activity and the

total amount of 7-OH-coumarin excreted in urine has

usually been used as a metric. However, coumarin might

not be an ideal probe for studying enzyme induction. It

undergoes extensive first-pass metabolism with systemic

availability of only 4% and 7-OH-coumarin, the major

metabolite formed, is excreted to 95% as the glucuronide

in urine within 4 h [20]. Consequently, the amount of

7-OH-coumarin excreted in a 0- to 8-h urine sample

would probably not reflect a possible induction of

CYP2A6. Increasing the coumarin dose, in order to

obtain full pharmacokinetic profiles of parent drug and

metabolite would appear to be a better choice.

The cocktail approach offers several advantages.

Multiple enzymes can be tested within one study and

the effect of intra-subject variability over time is

minimized [28]. Limitations include the risk of inter-

actions between drugs and the requirement of high

selective and sensitive analytical methods in order to

analyze several drugs and metabolites in the same

sample [28,29]. A number of studies have used the

cocktail methodology to investigate the activity of the

most important CYP450 enzymes (CYP1A2, CYP2C19,

CYP2D6, CYP2E1 and CYP3A) [13–16]. Zhu et al. have

previously validated the simultaneous oral administra-

tion of caffeine, mephenytoin, metoprolol, chlorzoxazone

and midazolam [13]. In the present study it was of



interest to include probes for CYP2A6 and CYP2B6, as

these enzymes have been shown to metabolize artem-

isinin in vitro [30]. Coumarin is a well studied probe for

CYP2A6 activity [31], but has not been used in a probe

cocktail before. To the best of our knowledge there are

no reports of coumarin affecting any of the enzymes

studied. CYP2B6 is induced by artemisinin in vivo and

considered likely to be involved in the auto-induction

phenomenon of this drug [8]. The suggested CYP2B6

probe bupropion [24], was considered unsuitable due to

the limited sampling scheme employed in this study.

That new synthetic endoperoxides may moderately

inhibit CYP2C9 in vitro [12] was not known when

the present study was designed, and this enzyme was

therefore not studied.

Extrinsic factors such as smoking and diet are known to

influence CYP450 activity. Smoking has been reported to

induce and inhibit CYP1A2 and CYP2A6 activities,

respectively, wherefore subjects ideally should abstain from

cigarette smoking [32,33]. As in the cultural settings of the

present study this was not deemed possible an approach of

allowing low regular smoking, monitored by questioning

was taken. As the daily number of cigarettes was monitored

and kept constant and the data analysis was based on intra-

individual changes this approach was judged feasible.

However, if CYP450 activities are induced by smokers,

baseline levels could be increased in these subjects and

hence further induction by the artemisinin drugs might not

occur to the same extent as in non-smokers.

The randomization of subjects to each treatment

group was not stratified for gender, explaining the

poorly matched number of females in the different

groups. Although some clinical studies suggest that

gender differences might exist in CYP450 enzyme

activity, gender does not appear to influence the indu-

cibility of CYP450 enzymes in freshly cultured human

hepatocytes [34]. Intake of oral contraceptives could

possibly alter the baseline level of enzyme activity, and

female participants were not specifically asked about

their use of oral contraceptives in the present study. The

different proportions of females/males might therefore

have an impact on the basal level of enzyme activity in

the treatment groups, but probably no effect on the

induction or inhibition observed.

Artesunate is a pro-drug and rapidly converted to

dihydroartemisinin by hydrolysis, wherefore the changes

in enzyme activities observed after artesunate intake is

likely to be caused by dihydroartemisinin.

Overall, the observed mean changes in activity of the

studied enzymes were moderate in most treatment groups.

Even though not significant in all treatment groups,

changes in the individual metrics pointed in the same

direction for all the artemisinin drugs, suggesting a class

effect. The clinical implications are not yet known but

could be of importance in situations of drug–drug inter-

actions involving drugs with narrow therapeutic win-

dows. Further, due to the individual distribution of the

change in enzymatic activity, for some individuals in the

population, the metabolic changes can be clinically

significant and warrants consideration in combination

treatment. The class parent, artemisinin, appeared to be

associated with the strongest capacity for induction and

inhibition. Therapeutic doses of artemisinin and its

derivatives were given, and thus the amounts adminis-

tered of each drug differ, with the highest dose for

artemisinin. When comparing the observed mean

changes in relation to molar doses of each artemisinin

drug given, the changes for the derivatives were in most

cases at least threefold higher than what was seen for

artemisinin indicating this compound to be the least

potent inducer.

C O N C L U S I O N

Intake of artemisinin and its principal semi-synthetic

derivatives affected the activities of several principal

human drug metabolizing CYP450s. CYP3A activity was

induced after multiple administrations of artemisinin,

dihydroartemisinin and artemether whereas CYP2C19

activity was induced after artemisinin and arteether

intake. CYP1A2 was inhibited by artemisinin, dihydro-

artemisinin and arteether and was induced by artemis-

inin. The CYP2D6 metric decreased on the first day of co-

administration with artemisinin and dihydroartemisinin

and was later increased following 5 days of artemisinin

intake. The magnitudes of the mean changes were

generally low and are not expected to present a problem

in the general patient population. Based on the present

results, which reflect phase I metabolic reactions only,

the most suitable artemisinin drug to be used in

combination treatment could not be selected. However,

the suggested class effect needs to be taken into account

in the development of new artemisinin derivatives and

drug combinations for the treatment of malaria.

A C K N O W L E D G E M E N T S

This study was supported by a grant from the Swedish

International Development Cooperation Agency (SIDA/

SAREC).



R E F E R E N C E S

1 Ashton M., Hai T.N., Sy N.D. et al. Artemisinin pharmaco-

kinetics is time-dependent during repeated oral administration

in healthy male adults. Drug Metab. Dispos. (1998) 26 25–27.

2 Ashton M., Nguyen D.S., Nguyen V.H. et al. Artemisinin

kinetics and dynamics during oral and rectal treatment of

uncomplicated malaria. Clin. Pharmacol. Ther. (1998) 63 482–

493.

3 Hassan Alin M., Ashton M., Kihamia C.M., Mtey G.J., Bjorkman

A. Multiple dose pharmacokinetics of oral artemisinin and

comparison of its efficacy with that of oral artesunate in

falciparum malaria patients. Trans. R. Soc. Trop. Med. Hyg.

(1996) 90 61–65.

4 Khanh N.X., de Vries P.J., Ha L.D., van Boxtel C.J., Koopmans

R., Kager P.A. Declining concentrations of dihydroartemisinin

in plasma during 5-day oral treatment with artesunate for

Falciparum malaria. Antimicrob. Agents Chemother. (1999) 43

690–692.

5 van Agtmael M.A., Cheng-Qi S., Qing J.X., Mull R., van Boxtel

C.J. Multiple dose pharmacokinetics of artemether in Chinese

patients with uncomplicated falciparum malaria. Int. J. Anti-

microb. Agents (1999) 12 151–158.

6 Ashton M., Sy N.D., Gordi T. et al. Evidence for time-dependent

artemisinin kinetics in adults with uncomplicated malaria.

Pharm. Pharmacol. Lett. (1996) 6 127–130.

7 Mihara K., Svensson U.S.H., Tybring G., Hai T.N., Bertilsson L.,

Ashton M. Stereospecific analysis of omeprazole supports

artemisinin as a potent inducer of CYP2C19. Fundam. Clin.

Pharmacol. (1999) 13 671–675.

8 Simonsson U.S.H., Jansson B., Hai T.N., Huong D.X., Tybring

G., Ashton M. Artemisinin autoinduction is caused by involve-

ment of cytochrome P450 2B6 but not 2C9. Clin. Pharmacol.

Ther. (2003) 74 32–43.

9 Burk O., Arnold K.A., Nussler A.K. et al. Antimalarial artem-

isinin drugs induce cytochrome P450 and MDR1 expression by

activation of xenosensors pregnane X receptor and constitutive

androstane receptor. Mol. Pharmacol. (2005) 67 1954–1965.

10 Bapiro T.E., Egnell A.C., Hasler J.A., Masimirembwa C.M.

Application of higher throughput screening (HTS) inhibition

assays to evaluate the interaction of antiparasitic drugs with

cytochrome P450s. Drug Metab. Dispos. (2001) 29 30–35.

11 Bapiro T.E., Sayi J., Hasler J.A., et al. Artemisinin and

thiabendazole are potent inhibitors of cytochrome P450 1A2

(CYP1A2) activity in humans. Eur. J. Clin. Pharmacol. (2005)

61 755–761.

12 Vennerstrom J.L., Arbe-Barnes S., Brun R. et al. Identification

of an antimalarial synthetic trioxolane drug development

candidate. Nature (2004) 430 900–904.

13 Zhu B., Ou-Yang D.S., Chen X.P. et al. Assessment of cyto-

chrome P450 activity by a five-drug cocktail approach. Clin.

Pharmacol. Ther. (2001) 70 455–461.

14 Frye R.F., Matzke G.R., Adedoyin A., Porter J.A., Branch R.A.

Validation of the five-drug ‘‘Pittsburgh cocktail’’ approach for

assessment of selective regulation of drug-metabolizing

enzymes. Clin. Pharmacol. Ther. (1997) 62 365–376.

15 Chen X.P., Tan Z.R., Huang S.L., Huang Z., Ou-Yang D.S.,

Zhou H.H. Isozyme-specific induction of low-dose aspirin on

cytochrome P450 in healthy subjects. Clin. Pharmacol. Ther.

(2003) 73 264–271.

16 Adedoyin A., Frye R.F., Mauro K., Branch R.A. Chloroquine

modulation of specific metabolizing enzymes activities: investi-

gation with selective five drug cocktail. Br. J. Clin. Pharmacol.

(1998) 46 215–219.

17 Scott R.J., Palmer J., Lewis I.A., Pleasance S. Determination of a

‘GW cocktail’ of cytochrome P450 probe substrates and their

metabolites in plasma and urine using automated solid phase

extraction and fast gradient liquid chromatography tandem

mass spectrometry. Rapid Commun. Mass Spectrom. (1999) 13

2305–2319.

18 Jansson B., Elsherbiny D., Simonsson U.S. Enantiospecific

separation and quantitation of mephenytoin and its metabolites

nirvanol and 4¢-hydroxymephenytoin in human plasma and

urine by liquid chromatography/tandem mass spectrometry.

Rapid Commun. Mass Spectrom. (2006) 20 463–472.

19 Fuhr U., Rost K.L. Simple and reliable CYP1A2 phenotyping

by the paraxanthine/caffeine ratio in plasma and in saliva.

Pharmacogenetics (1994) 4 109–116.

20 Pasanen M., Rannala Z., Tooming A., Sotaniemi E.A., Pelkonen

O., Rautio A. Hepatitis A impairs the function of human hepatic

CYP2A6 in vivo. Toxicology (1997) 123 177–184.

21 Tamminga W.J., Wemer J., Oosterhuis B. et al. An optimized

methodology for combined phenotyping and genotyping on

CYP2D6 and CYP2C19. Eur. J. Clin. Pharmacol. (2001) 57

143–146.

22 Lin Y.S., Lockwood G.F., Graham M.A. et al. In-vivo pheno-

typing for CYP3A by a single-point determination of midazolam

plasma concentration. Pharmacogenetics (2001) 11 781–791.

23 Svensson U.S.H., Ashton M., Trinh N.H. et al. Artemisinin

induces omeprazole metabolism in human beings. Clin. Phar-

macol. Ther. (1998) 64 160–167.

24 Tucker G.T., Houston J.B., Huang S.M. Optimizing drug

development: strategies to assess drug metabolism/transporter

interaction potential-toward a consensus. Clin. Pharmacol.

Ther. (2001) 70 103–114.

25 Palmer J.L., Scott R.J., Gibson A., Dickins M., Pleasance S.

An interaction between the cytochrome P450 probe substrates

chlorzoxazone (CYP2E1) and midazolam (CYP3A). Br. J. Clin.

Pharmacol. (2001) 52 555–561.

26 Thummel K.E., O’Shea D., Paine M.F. et al. Oral first-pass

elimination of midazolam involves both gastrointestinal and

hepatic CYP3A-mediated metabolism. Clin. Pharmacol. Ther.

(1996) 59 491–502.

27 Christensen M., Andersson K., Dalen P. et al. The Karolinska

cocktail for phenotyping of five human cytochrome P450

enzymes. Clin. Pharmacol. Ther. (2003) 73 517–528.

28 Breimer D.D., Schellens J.H. A ‘cocktail’ strategy to assess

in vivo oxidative drug metabolism in humans. Trends

Pharmacol. Sci. (1990) 11 223–225.

29 Paolini M., Biagi G.L., Bauer C., Cantelli-Forti G. Cocktail

strategy: complications and limitations. J. Clin. Pharmacol.

(1993) 33 1011–1012.



30 Svensson U.S.H., Ashton M. Identification of the human

cytochrome P450 enzymes involved in the in vitro metabolism

of artemisinin. Br. J. Clin. Pharmacol. (1999) 48 528–535.

31 Pelkonen O., Rautio A., Raunio H., Pasanen M. CYP2A6:

a human coumarin 7-hydroxylase. Toxicology (2000) 144

139–147.

32 Fuhr U. Tobacco smoking, alcohol, and special diet in drug

trials: exclusion criteria or target of specific investigations?

Int. J. Clin. Pharmacol. Ther. (1998) 36 604–606.

33 Poland R.E., Pechnick R.N., Cloak C.C., Wan Y.J., Nuccio I.,

Lin K.M. Effect of cigarette smoking on coumarin metabolism in

humans. Nicotine Tob. Res. (2000) 2 351–354.

34 Parkinson A., Mudra D.R., Johnson C., Dwyer A., Carroll K.M.

The effects of gender, age, ethnicity, and liver cirrhosis on

cytochrome P450 enzyme activity in human liver microsomes

and inducibility in cultured human hepatocytes. Toxicol. Appl.

Pharmacol. (2004) 199 193–209.



Artemisinin antimalarials moderately affect cytochrome P450 enzyme activity in healthy subjects

Documents

Transcript of Artemisinin antimalarials moderately affect cytochrome P450 enzyme activity in healthy subjects