Isozyme-specific induction of low-dose aspirin on cytochrome P450 in healthy subjects
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Isozyme-specific induction of low-dose
aspirin on cytochrome P450 in healthy
subjects
Objective: This study was designed to define the effect of low-dose aspirin administration on the activity of
cytochrome P450 (CYP) in normal human subjects.
Methods: Aspirin, 50 mg daily, was given for 14 days to 18 nonsmoking healthy male volunteers. A modified
5-drug cocktail procedure consisting of caffeine, mephenytoin, metoprolol, chlorzoxazone, and midazolam
was performed to simultaneously assess in vivo activity of CYP1A2, CYP2C19, CYP2D6, CYP2E1, and
CYP3A, respectively. The activities were assessed on 4 occasions including at baseline, after 7 and 14 daily
doses of aspirin, and at 7 days after discontinuation of aspirin. Concentrations of parent drugs and corre-
sponding metabolites in biologic samples were assayed by reversed-phase HPLC.
Results: Both 7-day and 14-day aspirin intake increased the activity of CYP2C19 significantly, as indicated by
4-hydroxymephenytoin urinary recovery (P < .001). Induction of low-dose aspirin on CYP2C19 was time-
dependent. CYP3A activity indices increased moderately but significantly by both 7-day and 14-day aspirin
treatment (P < .05), but the percentage changes in CYP3A activity indices were not significant. Low-dose
aspirin had no effect on CYP1A2, CYP2D6, and CYP2E1 in vivo activity by either 7-day or 14-day treatment.
Conclusions: The effect of low-dose aspirin on CYPs was enzyme-specific. Both 7-day and 14-day low-dose
aspirin induced the in vivo activities of CYP2C19 but did not affect the activities of CYP1A2, CYP2D6, and
CYP2E1. The effect of low-dose aspirin on CYP3A activity awaits further confirmation. When low-dose
aspirin is used in combination with drugs that are substrates of CYP2C19, doses of the latter should be
adjusted to ensure their efficacy. (Clin Pharmacol Ther 2003;73:264-71.)
Xiao-Ping Chen, PhD, Zhi-Rong Tan, BS, Song-Lin Huang, BS, Zheng Huang, MD,
Dong-Sheng Ou-Yang, PhD, and Hong-Hao Zhou, MD Changsha, Hunan, China
Acetylsalicylic acid (aspirin) has been widely used as disease, and heart failure.1 The prolonged low-dose
a nonsteroidal anti-inflammatory agent for decades. A strategy has also been implicated in prophylaxis of
new use of the century-old drug is that low-dose aspirin digestive tract cancer in recent years.2 Apart from these
(50-100 mg/d) is frequently prescribed as an effective, therapeutic indications, low-dose aspirin usage was
safe, and economical agent for primary and secondary shown to be associated with reduced risk of dementia
thromboembolic prevention, especially in patients with such as Alzheimer’s disease.3 Acetylsalicylic acid is
cardiovascular diseases such as hypertension, coronary rapidly and completely deacetylated to salicylate by
esterases present in the gastrointestinal tract, liver, and
From the Pharmacogenetics Research Institute, Central South Uni- serum after oral or parenteral administration. Previous
versity. studies have indicated that aspirin may affect the phar-
Supported by the China Medical Board of New York (grants 99-697 macokinetics of coadministered drugs through induc-
and 01-755) and the National Natural Science Foundation of China tion of drug metabolism when used in relatively high
(grant F30130210).
Received for publication Feb 21, 2002; accepted Sept 25, 2002.
doses for anti-inflammatory and analgesic purposes.4,5
Reprint requests: Hong-Hao Zhou, MD, Pharmacogenetics Research As an important superfamily that plays a major role in
Institute, South Central University, Changsha, Hunan 410078, PR the metabolism of many clinically important drugs and
China. some other exogenous compounds, cytochrome P450
E-mail: hhzhou@public.cs.hn.cn (CYP) has become a target of study for drug-drug
Copyright © 2003 by the American Society for Clinical Pharmacol-
ogy & Therapeutics.
interactions in recent years. Much attention has been
0009-9236/2003/$30.00 ⫹ 0 paid to the induction and inhibition of CYP and sub-
doi:10.1067/mcp.2003.14 sequent drug efficacy or toxicity, as well as possible
264CLINICAL PHARMACOLOGY & THERAPEUTICS
VOLUME 73, NUMBER 3 Chen et al 265
associations with cancer and other diseases.6-8 Our in- given 100-mg caffeine tablets (Hunan Pharmaceutical
terest in acetylsalicylic acid was aroused by reported Inc), 200-mg chlorzoxazone tablets (Hongqi Pharma-
effects of acetylsalicylic acid or salicylate on CYP ceutical Inc, Shanghai, China), 100-mg metoprolol
activity (in particular, CYP2E1) when used in relatively tablets (Astra, Wuxi, China), 100-mg racemic mephe-
high doses.9-13 However, no data are available as to the nytoin tablets (Mesantoin; Sandoz Inc, East Hanover,
effect of aspirin or salicylate on CYP activity in hu- NJ), and 7.5-mg midazolam tablets (Dormicum;
mans. Because of the frequent use of low-dose aspirin, Hoffmann-La Roche Ltd, Basel, Switzerland) with 150
it is of interest to determine whether low-dose aspirin mL of water at 7:30 AM after an overnight fast. On day
administration modulates activities of CYP isozymes in 8, the cocktail drugs were given concurrently with 50
vivo in humans. mg aspirin. Food was given 1.5 hours after drug ad-
Multiple families and subfamilies of CYP have been ministration. For metoprolol ␣-hydroxylation and
identified and can be detected in adult human liver, the S-mephenytoin 4-hydroxylation phenotyping, each sub-
major organ involved in CYP-mediated metabolism.14 ject emptied the bladder before taking the probe drugs.
To investigate the effect of low-dose aspirin on CYP, a Urine was collected from 0 to 8 hours after drug intake
cocktail approach with isozyme-specific probe drugs in a container with ascorbic acid. The urine volume was
for simultaneous determination of the activity of recorded, and a 5-mL aliquot was stored for detecting
CYP1A2, CYP2C19, CYP2E1, CYP2D6, and CYP3A CYP2D6 and CYP2C19 activity. Blood samples (8
was used.15 These 5 isozymes consist of about 70% of mL) were collected at 1, 4, and 6 hours after adminis-
the total microsomal CYP proteins in adult human tration of probe drugs with ethylenediaminetetraacetic
liver14 and are responsible for the metabolism of nearly acid as the anticoagulant and were centrifuged, and
90% of drugs in clinical use. Therefore any modifica- plasma was harvested for determining CYP3A,
tion in activity of these enzymes may result in de- CYP2E1, and CYP1A2 activity, respectively. All sam-
creased efficacy or increased toxicity of medications. ples were stored frozen at ⫺20°C until analyzed.
The aim of this study was to investigate the effect of
both 7-day and 14-day daily low-dose aspirin on in vivo
activity of these enzymes in humans. Analytic technique
All drugs and corresponding metabolites in plasma
METHODS or urine were measured by reversed-phase HPLC tech-
Subjects niques with an ultraviolet detector. The within- and
Eighteen normal healthy Chinese male volunteers between-day coefficients of variation for each of these
(age, 21.3 ⫾ 1.5 years [mean ⫾ SD]; weight, 60.27 ⫾ assays were ⬍10%.
4.86 kg) participated in the study. The experimental Plasma assay for caffeine and paraxanthine. Para-
protocol was approved by the Ethics Committee of xanthine and caffeine in plasma were determined by
Xiangya School of Medicine (Changsha, China), and HPLC according to the method established by Ou-Yang
all subjects gave written informed consent before the et al.16 A 300-L aliquot of plasma was mixed with
study began. All subjects were nonsmokers and were in 100 L of the IS (100-mol/L -hydroxyethyl-
good health as indicated by medical history, routine theophylline) and 300 mg of ammonium sulfate. The
physical examination, and biochemical testing. All sub- sample was extracted with 5 mL of chloroform and
jects were asked to abstain from alcohol or caffeine- isopropanol (9:1). All components were separated on a
containing products or any other medications during the reversed-phase column (Spherisorb octadecylsilane 2;
study and a week before the study. Waters Corporation, Milford, Mass) (250 mm ⫻ 4 mm
internal diameter; 5 m particle size). The solvent used
Study protocol for elution was acetic acid, 0.05%; acetonitrile; and
An enteric-coated aspirin tablet (Hunan Pharmaceu- methanol. Typical conditions for elution were 81.5%
tical Inc, Changsha, China), 50 mg/d, was given to all acetic acid, 0.05%; 8.5% acetonitrile; and 10% metha-
subjects from day 2 to day 15 at 7:50 AM after breakfast. nol for 0 to 5 minutes and 81.5% to 72% acetic acid,
A modified 5-drug cocktail procedure15 for simulta- 0.05%; 8.5% to 18% acetonitrile; and 10% methanol
neously determining the in vivo activity of CYP1A2, for 5 to 13 minutes. The flow rate was 0.7 mL/min. The
CYP2C19, CYP2D6, CYP2E1, and CYP3A recently detection wavelength was 282 nm. The coefficients of
developed by us was applied to detect the activity of both intra-assay and interassay variations were ⬍8%
these isozymes on 4 separate occasions— day 1, day 8, for paraxanthine and ⬍10% for caffeine. The ranges of
day 16, and day 23. On each occasion, each subject was concentrations in the calibration curves for both para-CLINICAL PHARMACOLOGY & THERAPEUTICS 266 Chen et al MARCH 2003 xanthine and caffeine were 1.0 to 100 mol/L. The a 0.5-mL plasma sample was deconjugated with 1000 limit of detection was 0.1 mol/L for both compounds. U of -glucuronidase. Fifty microliters of phenacetin Urine assay for 4ⴕ-hydroxymephenytoin. Urine ali- (20 mol/L; used as IS) was added to the sample and quots were assayed for concentration of 4⬘- extracted by 5 mL of diethyl ether. The organic phase hydroxymephenytoin by a modification of a procedure was evaporated to dryness under a stream of nitrogen at developed in our laboratory.17 In brief, a mixture of 37°C. The residue was dissolved in 50 L of mobile urine containing total 4⬘-hydroxymephenytoin after en- phase, and a 20-L aliquot was injected into the HPLC zymatic deconjugation by 1000 U of -gluconidase and system. All components were separated on a C8 col- phenobarbital as the IS was extracted with diethyl umn (Hewlett-Packard) (4.6 mm ⫻ 150 mm; 5 par- ether. The residue remained after evaporation was re- ticle size). The mobile phase was composed of 25- constituted in 50 L of mobile phase, and 20 L was mmol/L sodium dihydrogen phosphate, acetonitrile, injected into the HPLC system. Separation of 4⬘- and methanol at a ratio of 66:24:10 (vol/vol/vol), and hydroxymephenytoin was achieved with a C18 column the pH of the mobile phase was finally adjusted to 3.5 (Hewlett-Packard, Palo Alto, Calif) (4.6 mm ⫻ 125 with orthophosphoric acid. The flow rate was 0.7 mL/ mm; 5 m particle size). The sample was eluted with a min for 0 to 5 minutes and then 1.4 mL/min afterward. mobile phase of acetonitrile and water (15:85 [vol/vol]) The ultraviolet detection was set at 282 nm for 0 to 6 pumped at a flow rate of 1.0 mL/min. The eluent was minutes and then shifted to 264 nm thereafter. The monitored at 204 nm. The retention times for 4⬘- column temperature was maintained at 37°C. The re- hydroxymephenytoin and the IS were within 6 minutes. tention times for 6-hydroxychlorzoxazone, IS, and The coefficients of both intra-assay and interassay vari- chlorzoxazone were within 8.5 minutes. The coeffi- ations were ⬍10%. The range of concentrations in the cients of both intra-assay and interassay variations were calibration curve was 12.5 to 1000 mol/L. The limit of ⬍10% for chlorzoxazone and ⬍9% for 6-hydroxy- detection was 0.1 mol/L. chlorzoxazone. A good linear relationship was obtained Urine assay for metoprolol and in the range of 0.25 to 80 mol/L for both chlorzoxa- ␣-hydroxymetoprolol. Metoprolol and zone and 6-hydroxychlorzoxazone. The limit of detec- ␣-hydroxymetoprolol in urine samples were deter- tion for both compounds in plasma was 0.125 mol/L. mined by a modified HPLC method according to Huang Plasma assay for midazolam and 1ⴕ- et al.18 Urine, 100 L, was alkalinized by sodium hydroxymidazolam. Midazolam and 1⬘- hydroxide (pH 12.5) and extracted with 5 mL dichlo- hydroxymidazolam were also determined by HPLC. In romethane after 250 ng propranolol was added as the brief, 1.0 mL of plasma was deconjugated with 200 U IS. The residue that remained after evaporation of the of -glucuronidase in sodium acetic acid buffer (pH organic phase was dissolved in 50 L of mobile phase, 4.75) at 37°C. After addition of 50 L of phenacetin and 20 L was injected into the HPLC system. All (IS, 20 mol/L), the mixture was extracted by 5 mL of chemicals were separated on a reversed-phase column diethyl ether. After centrifugation, the upper organic (Spherisorb octadecylsilane 2) (250 mm ⫻ 4 mm in- phase was transferred and evaporated to dryness. The ternal diameter; 5 m particle size) with the use of a residue was dissolved in 50 L of mobile phase, and a mixture of 700 mL of methanol, 300 mL of water, 1.2 20-L aliquot was injected into the HPLC system. All mL of glacial acetic acid, and 300 L of triethylamine components were separated isocratically on a C8 col- as the mobile phase pumped at a flow rate of 1.0 umn (Hewlett-Packard) (4.6 mm ⫻ 150 mm; 5 m mL/min. The retention times for metoprolol, ␣-hydroxy- particle size). The mobile phase was composed of 25- metoprolol, and the IS were within 5 minutes. The mmol/L sodium dihydrogen phosphate, acetonitrile, coefficients of both intra-assay and interassay varia- and methanol at a ratio of 55:35:10 (vol/vol/vol) and tions were ⬍5% for both metoprolol and ␣-hydroxy- was delivered at a flow rate of 1.0 mL/min; the pH of metoprolol. The ranges of concentrations in the cali- the eluent was finally adjusted to 3.5 with orthophos- bration curves were 0.1 to 15 g/mL for both phoric acid. The detection wavelength was 234 nm. The compounds. The limits of detection were 0.05 g/mL column temperature was maintained at 37°C. Retention and 0.01 g/mL for metoprolol and ␣-hydroxy- times for IS, 1⬘-hydroxymidazolam, and midazolam metoprolol, respectively. were within 8 minutes. The coefficients of both intra- Plasma assay for chlorzoxazone and assay and interassay variations were ⬍10% for both 6-hydroxychlorzoxazone. The concentrations of chlor- midazolam and 1⬘-hydroxymidazolam. A good linear zoxazone and 6-hydroxychlorzoxazone in plasma were relationship was obtained in the range of 12.5 to 800 assayed by a modification of Frye and Stiff.19 In brief, nmol/L for both midazolam and 1⬘-hydroxymidazolam.
CLINICAL PHARMACOLOGY & THERAPEUTICS
VOLUME 73, NUMBER 3 Chen et al 267
Table I. Comparison of phenotypic measures of specific metabolizing enzymes at baseline, during 7 and 14 days
of treatment with aspirin, and at 7 days after aspirin dosing
Activity index Day 1 Day 8 Day 16 Day 23 P value
4⬘-Hydroxymephenytoin recovery in urine (%) 44.5 (38.3, 54.0) 50.0 (45.8, 56.8)*† 60.0 (54.0, 69.8)‡§㛳 46.5 (42.0, 52.0) ⬍.001
Plasma 1⬘-hydroxymidazolam/midazolam ratio 2.42 (1.54, 3.32) 2.99 (1.65, 3.90)¶ 3.05 (2.05, 3.73)†¶ 2.29 (1.71, 2.86) .014
Plasma paraxanthine/caffeine ratio 0.38 (0.34, 0.47) 0.35 (0.28, 0.46) 0.38 (0.30, 0.50) 0.42 (0.28, 0.57) .978
Urinary metoprolol/␣-hydroxymetoprolol ratio 0.36 (0.24, 0.70) 0.49 (0.28, 0.70) 0.43 (0.20, 0.63) 0.36 (0.21, 0.82) .849
Plasma 6-hydroxychlorzoxazone/chlorozoxazone 0.85 (0.57, 1.03) 0.76 (0.59, 0.93) 0.78 (0.47, 1.21) 0.78 (0.63, 1.10) .372
ratio
Data are expressed as median with 25th percentile and 75th percentile in parentheses.
*P ⬍ .01, compared with day 1.
†P ⬍ .01, compared with day 23.
‡P ⬍ .001, compared with day 1.
§P ⬍ .001, compared with day 23.
㛳P ⬍ .01, compared with day 8.
¶P ⬍ .05, compared with day 1.
The limit of detection for both compounds in plasma tidrug intake) during cocktail phenotyping procedures.
was 6.5 nmol/L. No interfering peaks were observed in chromatograms
after cocktail administration. In addition, for each spe-
Data analysis cific HPLC assay, no interfering peaks were observed
The molar concentration of paraxanthine divided by in blood or urine samples after aspirin dosage alone. No
that of caffeine (paraxanthine/caffeine) in the 6-hour poor metabolizer of either CYP2C19 or CYP2D6 was
plasma sample was used to assess CYP1A2 activity. identified among the 18 subjects.
The percentage of the dose excreted as 4⬘- As shown in Table I and Fig 1, A, both 7-day and
hydroxymephenytoin in 0- to 8-hour urine was used as 14-day aspirin administration significantly increased
the phenotypic index of CYP2C19 activity. The 0- to (P ⬍ .001) the urinary recovery of 4⬘-
8-hour urinary metoprolol/␣-hydroxymetoprolol ratio hydroxymephenytoin, a measure of CYP2C19 activity.
was used to indicate CYP2D6 activity. The plasma This induction in CYP2C19 activity was progressive
molar concentration ratio of 6-hydroxychlorzoxazone from the 7-day aspirin treatment to the 14-day aspirin
to chlorzoxazone at 4 hours was used to assess CYP2E1 treatment. This increase in S-mephenytoin
activity. The molar concentration of 4-hydroxylase activity was reversed, and the 0- to
1-hydroxymidazolam divided by that of midazolam in
8-hour urinary recovery of 4⬘-hydroxymephenytoin re-
the 1-hour plasma sample was used to assess CYP3A
turned to the baseline level 7 days after discontinuation
activity.
of aspirin.
Statistical analysis As shown in Table I and Fig 1, B, both 7-day and
14-day aspirin intake produced moderate induction of
All analyses for comparing the phenotypic index
were performed by SPSS software for Windows (ver- CYP3A activity, as indicated by midazolam
sion 10.0; SPSS Inc, Chicago, Ill). Results were ex- 1-hydroxylation activity (day 8 or day 16 versus day 1,
pressed as median with 25th percentile and 75th per- P ⬍ .05; day 16 versus day 23, P ⬍ .01). Furthermore,
centile or as mean ⫾ SD. The activity indices of this induction in CYP3A activity was reversed and
individual isozymes on different occasions were ana- returned to the baseline level 7 days after aspirin dis-
lyzed by use of the Friedman rank sum test. Differences continuation. However, when changes in CYP3A ac-
in the change in CYP2C19 activity indices were also tivity indices were analyzed as percentages, no differ-
analyzed by the Friedman test. A significant difference ence was observed among the changes for day 8 versus
was regarded as P ⬍ .05. day 1, day 16 versus day 1, and day 23 versus day 1
(P ⫽ .076), which were 24.36% ⫾ 48.24%, 31.65% ⫾
RESULTS 33.25%, and 9.75% ⫾ 40.71%, respectively.
All 18 subjects completed the trial. No side effects There were no significant differences in any of the
occurred during the 14-day aspirin administration. No phenotypic indices determined for CYP1A2, CYP2D6,
other side effects were observed except for a brief and CYP2E1 activity among the different occasions
period of sleepiness (within about 1-3 hours after mul- including before aspirin, at 7 days and 14 days afterCLINICAL PHARMACOLOGY & THERAPEUTICS
268 Chen et al MARCH 2003
Fig 1. Comparison of percentage of 0- to 8-hour urinary
Fig 2. Comparison of paraxanthine (17X) to caffeine (137X)
4⬘-hydroxymephenytoin recovery (A) and 1⬘-
plasma ratio at 6 hours (A) and metoprolol to
hydroxymidazolam (1⬘-OH-MDZ) to midazolam (MDZ)
␣-hydroxymetoprolol ratio in 0- to 8-hour urine (B) from
plasma ratio at 1 hour (B) from individuals at baseline, at 7
individuals at baseline, at 7 and 14 days after aspirin treat-
and 14 days after aspirin treatment, and at 7 days after
ment, and at 7 days after discontinuation of aspirin.
discontinuation of aspirin.
resulting from induction or inhibition of drug-
aspirin intake, and at 7 days after aspirin discontinua- metabolizing enzymes in recent years. Aspirin, a com-
tion (Table I; Figs 2 and 3). monly used drug for various clinical purposes, dis-
placed some coadministered drugs from plasma protein
DISCUSSION and increased the clearance of drugs such as fenopro-
Patients are often taking multiple medications simul- fen4 and phenytoin,5 presumably through induction of
taneously, and consequently the problem of drug-drug drug metabolism, when used in relatively high doses
interactions among comedications occurs. To optimize for anti-inflammatory and analgesic purposes. As men-
drug efficacy and minimize toxicity, intensive attention tioned in the context above, low-dose aspirin is cur-
has been paid to pharmacokinetic drug interactions rently prescribed together with other medications moreCLINICAL PHARMACOLOGY & THERAPEUTICS VOLUME 73, NUMBER 3 Chen et al 269 frequently to patients with cardiovascular disease. There are studies that focus on the effect of low-dose aspirin on the efficacy of drugs such as angiotensin- converting enzyme inhibitors through pharmacody- namic mechanisms.20 However, little is known about the pharmacokinetic effects of low-dose aspirin on con- currently used drugs, although metabolic drug interac- tions involving aspirin are also possible. Therefore it is of interest to investigate whether low-dose aspirin af- fects the metabolism of comedications. As an important peroxisome proliferator, a 1.6-fold induction of total CYP content was observed in mouse liver after expo- sure to a diet containing acetylsalicylic acid, 1.0%, for 10 days; at the same time, aspirin increased the activity of CYP4A significantly and was used as a tool to induce CYP4A expression in some animal studies.10,21 However, a study carried out by Jagota9 suggested that aspirin caused marked depression of both total CYP and cytochrome b5 in liver microsomes in mice; the Fig 3. Comparison of 6-hydroxychlorzoxazone (HCZX) to latter is an important coenzyme of CYP isozymes such chlorzoxazone (CZX) plasma ratio at 4 hours from individu- as CYP2E1.22 In this study, we investigated the effect als at baseline, at 7 and 14 days after aspirin treatment, and at of 7-day or 14-day low-dose aspirin administration on 7 days after discontinuation of aspirin. the in vivo activity of major CYP isozymes involved in drug metabolism, including CYP1A2, CYP2C19, CYP2D6, CYP2E1, and CYP3A, by a multiprobe cock- low-dose aspirin treatment was significantly higher tail procedure. The results showed that low-dose aspirin than that by 7-day treatment. This time-dependent in- affected CYP activity and that the effect was isozyme- duction of CYP2C19 was in agreement with the obser- selective. Both 7-day and 14-day low-dose aspirin sig- vation of Branch et al,25 who also reported that the nificantly induced the activities of CYP2C19 and induction of CYP2C19 by 28 daily doses of rifampin CYP3A, as measured by mephenytoin and midazolam (INN, rifampicin) was greater than that by 3 daily as the probe drugs. The activities of CYP1A2, doses. One possible explanation for this is that it takes CYP2D6, and CYP2E1 as measured by the probe drugs time to reach maximal induction through synthesis of caffeine, metoprolol, and chlorzoxazone, respectively, new enzyme molecules by these inducers. Although a were not affected. recent report has shown that a steady state of induction The human CYP2Cs are an important subfamily of for the CYP2C family occurred at about 50 hours after CYP enzymes that metabolize approximately 20% of treatment by inducers in primary cultures of human clinically used drugs and some endogenous com- hepatocytes,26 the time interval needed to reach the pounds, such as arachidonic acid. There are 4 members steady state of induction for these enzymes in vivo in in this subfamily—CYP2C8, CYP2C9, CYP2C19, and humans is still uncertain. Furthermore, the concentra- CYP2C18.23 However, because of the limitations of the tion of inducers used in vitro may be different from that cocktail procedure applied at present,15 only one mem- used in vivo. ber of this subfamily, CYP2C19, was investigated in CYP3A is the predominant isoform of the CYP su- this study. With S-mephenytoin 4-hydroxylation activ- perfamily in both the intestinal epithelium and liver, ity as the activity index, our results showed that both which contributes about 30% of the total CYP proteins 7-day and 14-day aspirin intake increased the activity in adult liver. Despite its role in the biotransformation of CYP2C19 significantly. This induction of CYP2C19 of almost 50% of clinically used drugs,27 CYP3A also by aspirin may partially explain the increased clearance contributes to remarkable first-pass elimination of its of phenytoin, a substrate of CYP2C19, when used substrates after their oral administration. Therefore elu- concurrently with aspirin.5,24 Of course, the exact un- cidation of the modulation of CYP3A becomes impor- derlying mechanism for this induction is unclear and tant. The results of our study showed that both 7-day requires further investigation. The results also showed and 14-day low-dose aspirin treatment significantly in- that the degree of induction in CYP2C19 by 14-day creased the activity of CYP3A. However, when
CLINICAL PHARMACOLOGY & THERAPEUTICS
270 Chen et al MARCH 2003
changes in CYP3A activity indices were further ana- 2. Taketo MM. Cyclooxygenase-2 inhibitors in tumorigen-
lyzed as percentages, no difference was observed. Fur- esis (part II). J Natl Cancer Inst 1998;90:1609-20.
thermore, marked interindividual variations in the 3. Broe GA, Grayson DA, Creasey HM, Waite LM, Casey
changes were observed. Therefore it is not clear BJ, Bennett HP, et al. Anti-inflammatory drugs protect
whether the differences in CYP3A activity on different against Alzheimer disease at low doses. Arch Neurol
occasions resulted from aspirin administration or were 2000;57:1586-91.
4. Miners JO. Drug interactions involving aspirin (acetyl-
driven by a few individuals with greater variability. In
salicylic acid) and salicylic acid. Clin Pharmacokinet
view of the widespread and increasing comedications
1989;17:327-44.
of low-dose aspirin with drugs that are substrates of
5. Paxton JW. Effect of aspirin on salivary and serum
CYP3A, such as calcium channel blockers (eg, dihy- phenytoin kinetics in healthy subjects. Clin Pharmacol
dropyridine) and lipid-lowering drugs (eg, statins), fur- Ther 1980;27:170-8.
ther study of the effect of low-dose aspirin on the 6. Fuhr U. Induction of drug metabolising enzymes: phar-
activity of CYP3A would be important. macokinetics and toxicological consequences in human.
In contrast to the effect on CYP2C19, our results Clin Pharmacokinet 2000;38:493-504.
showed that neither 7-day nor 14-day low-dose aspirin 7. Han XM, Zhou HH. Polymorphism of CYP450 and can-
treatment affected the activity of CYP1A2, CYP2E1, cer susceptibility. Acta Pharmacol Sin 2000;21:673-9.
and CYP2D6. The negative effect of aspirin on 8. Riedl AG, Watts PM, Jenner P, Marsden CD. P450
CYP2E1 in this study is in disagreement with the enzymes and Parkinson’s disease: the story so far. Mov
results of studies carried out on both experimental Disord 1998;13:212-20.
animals and cultured cells as previously reported.11-13 9. Jagota SK. Depression of cytochrome P450-dependent
Pretreatment of rats with aspirin increased the hepatic drug biotransformation by poly(rI.rC) and aspirin. Bio-
microsomal p-nitrophenol hydroxylation activity, a re- chem Med Metab Biol 1989;41:212-6.
action mediated primarily by CYP2E1.11 An investiga- 10. Cai Y, Sohlenius AK, Andersson K, Sundberg C, DePi-
erre JW. Effects of acetylsalicylic acid on parameters
tion performed by Damme et al12 further confirmed this
related to peroxisome proliferation in mouse liver. Bio-
finding and indicated that induction of CYP2E1 by
chem Pharmacol 1994;47:2213-9.
aspirin was associated with elevation of CYP2E1 mes- 11. Pankow D, Damme B, Schror K. Acetylsalicylic acid—
senger ribonucleic acid. Because salicylate is a sub- inducer of cytochrome P-450 2E1? Arch Toxicol 1994;
strate of CYP2E1 and undergoes 5-hydroxylation in the 68:261-5.
5⬘-position to produce 2,5-dihydroxybenzoic acid me- 12. Damme B, Darmer D, Pankow D. Induction of hepatic
diated partly by CYP2E1,28 aspirin was also reported to cytochrome P4502E1 in rats by acetylsalicylic acid or
induce CYP2E1 by enzyme stabilization as other sub- sodium salicylate. Toxicology 1996;106:99-103.
strates of CYP2E1 did.13,29 However, it is not unex- 13. Wu D, Cederbaum AI. Sodium salicylate increases
pected that low-dose aspirin did not affect CYP2E1 CYP2E1 levels and enhances arachidonic acid toxicity in
activity as observed in this study, because the dosage of HepG2 cells and cultured rat hepatocytes. Mol Pharma-
aspirin used was far less than that used in other studies. col 2001;59:795-805.
In addition, the great interspecies differences may also 14. Hakkola J, Tanaka E, Pelkonen O. Developmental ex-
contribute to the disagreement. According to the results pression of Cytochrome P450 enzymes in human liver.
of this study, when drugs metabolized mainly by Pharmacol Toxicol 1998;82:209-17.
CYP2E1, CYP1A2, and CYP2D6 are used in combi- 15. Zhu B, Ou-Yang DS, Chen XP, Huang SL, Tan ZR, He
nation with low-dose aspirin, the pharmacokinetic ef- N, et al. Assessment of cytochrome P450 activity by a
five-drug cocktail approach. Clin Pharmacol Ther 2001;
fect of low-dose aspirin on these drugs can be ignored.
70:455-61.
This study has indicated the possibility of a drug-
16. Ou-Yang DS, Huang SL, Xie HG, Wang CY, Zhou HH.
drug interaction between low-dose aspirin and sub- Use of caffeine as a probe for rapid determination of
strates of CYP2C19. Studies of low-dose aspirin in cytochrome P-450 CYP1A2 activity in humans. Acta
combination with drugs that are substrates of CYP2C19 Pharmacol Sin 1998;19:44-6.
should be performed to determine whether the dose of 17. Xie HG, Huang SL, Zhou HH. High-performance liquid
the latter should be adjusted. chromatographic determination of urinary 4⬘-
hydroxymephenytoin, a metabolic marker for the hepatic
enzyme CYP2C19, in humans. J Chromatogr B Biomed
References Appl 1995;668:125-31.
1. Patrono C. Aspirin: new cardiovascular uses for an old 18. Huang SL, Xie HG, Wang J, Zhou HH. Determination of
drug. Am J Med 2001;110:62S-65S. metoprolol and ␣-hydroxymetoprolol in human urine byCLINICAL PHARMACOLOGY & THERAPEUTICS
VOLUME 73, NUMBER 3 Chen et al 271
high performance liquid chromatography with ultraviolet lective phenytoin hydroxylation in Japanese: difference
detection. Chin J Chromatogr 1996;14:408-9. in chiral preference of CYP2C9 and CYP2C19. Biochem
19. Frye RF, Stiff DD. Determination of chlorzoxazone and Pharmacol 1999;57:1297-303.
6-hydroxychlorzoxazone in human plasma and urine by 25. Branch RA, Adedoyin A, Frye RF, Wilson JW,
high-performance liquid chromatography. J Chromatogr Romkes M. In vivo modulation of CYP enzymes by
B Biomed Appl 1996;686:291-6. quinidine and rifampin. Clin Pharmacol Ther 2000;68:
20. Meune C, Mahe I, Mourad JJ, Simoneau G, Knellwolf 401-11.
AL, Bergmann JF, et al. Interaction between angiotensin- 26. Gerbal-Chaloin S, Pascussi JM, Pichard-Garcia L, Daujat
converting enzyme inhibitors and aspirin: a review. Eur M, Waechter F, Fabre JM, et al. Induction of CYP2C
J Clin Pharmacol 2000;56:609-20. genes in human hepatocytes in primary culture. Drug
21. Okita JR, Johnson SB, Castle PJ, Dezellem SC, Okita Metab Dispos 2001;29:242-51.
RT. Improved separation and immunodetection of rat 27. de Wildt SN, Kearns GL, Leeder JS, van den Anker JN.
cytochrome P450 4A forms in liver and kidney. Drug Cytochrome P450 3A: ontogeny and drug disposition.
Metab Dispos 1997;25:1008-12. Clin Pharmacokinet 1999;37:485-505.
22. Lieber CS. Cytochrome P-4502E1: its physiological and 28. Dupont I, Berthou F, Bodenez P, Bardou L, Guirriec C,
pathological role. Physiol Rew 1997;77:517-44. Stephan N, et al. Involvement of cytochromes P-450 2E1
23. Goldstein JA. Clinical relevance of genetic polymor- and 3A4 in the 5-hydroxylation of salicylate in humans.
phisms in the human CYP2C subfamily. Br J Clin Phar- Drug Metab Dispos 1999;27:322-6.
macol 2001;52:349-55. 29. Chien JY, Thummel KE, Slattery JT. Pharmacokinetic
24. Yasumori T, Chen LS, Li QH, Ueda M, Tsuzuki T, consequences of induction of CYP2E1 by ligand stabili-
Goldstein JA, et al. Human CYP2C-mediated stereose- zation. Drug Metab Dispos 1997;25:1165-75.
CORRECTION
In the article “Pharmacodynamic modeling of the electroencephalographic effects of flumazenil in
healthy volunteers sedated with midazolam” (Fiset P, Lemmens HLM, Egan TE, Shafer SL, Stanski DR.
Clin Pharmacol Ther 1995;58:567-82), T. E. Egan should have been T. D. Egan.You can also read
