Professional Documents
Culture Documents
8/2010 (497-503)
Key words Abstract. Objectives: To explore if non- the same time places patients at a risk for drug
non-concurrent dosing concurrent amlodipine dosing results in less interaction, which can be sometimes fatal.
amlodipine simvastatin drug interaction, the pharmacokinetic pro- Drug interaction is more likely to occur when
drug interaction files, safety and efficacy endpoints were as-
cytochrome P-450 a drug inhibits the enzyme that metabolizes a
sessed following repeated doses of simva-
statin, co-administered concurrently or non- co-administered drug or two drugs compete
concurrently with amlodipine in patients with for the same metabolizing enzyme.
coexisting hypertension and hyperlipidemia. The latter is the case for simvastatin and am-
*This author contributed
equally to this work.
Methods: Seventeen patients randomly re- lodipine, both of which involve the cyto-
ceived daily doses of 20 mg simvastatin and
5 mg amlodipine for 6 weeks, either with both chrome P-450 (CYP) 3A4 as the main drug me-
drugs at 7:00 PM (concurrent) or with simva- tabolizing enzyme [9, 24]. For example, when
statin at 7:00 PM followed by amlodipine at simvastatin was combined with amlodipine, the
11:00 PM (non-concurrent). The maximum biological activity of simvastatin, as measured
plasma concentration (Cmax) and the area un- by the peak concentration and the area under
der the concentration-time curve up to the last
the concentration-time curve of 3-hydroxy-
quantifiable concentration (AUClast) were es-
timated at steady state. Lipid profiles and 3-methylglutaryl-coenzyme A (HMG-CoA)
blood pressure values were also compared reductase inhibitors, increased by 42.7%
between the concurrent and non-concurrent (p < 0.05) and 26.7% (p < 0.05), respectively
groups. Results: The Cmax and AUClast and of [20].
Received simvastatin acid in the non-concurrent amlo-
Simvastatin is an inactive lactone pro-
December 7, 2009; dipine dosing group were 63.2% and 66.0%,
accepted respectively, of the values obtained in the drug, undergoing hydrolysis by plasma ester-
February 24, 2010 concurrent group (1.2 1.0 vs. 1.9 0.9 ases (i.e., non-CYP process) to active simvas-
ng/ml and 10.3 8.3 vs. 15.6 7.5 h ng/ml, tatin acid to exert its lipid-lowering effect
Correspondence to respectively, mean standard deviation). [24]. Simvastatin acid has the greatest inhibi-
H. Lee, MD, PhD Changes from baseline in lipid profile and
Associate Adjunct tory effect on the HMG-CoA reductase com-
blood pressure were comparable between the
Professor, Director groups. Conclusions: Non-concurrent dosing pared to the other active CYP3A4 downstream
Center for Drug metabolites [8], and is further metabolized
may be a useful and safe therapeutic option
Development Science,
for patients who require two or more drugs mainly by CYP3A4 (i.e., 80%) and, to some
Department of Bio-
pharmaceutical
administered concomitantly, but who are extent, by CYP2C8 as well [21]. CYP3A4 in-
Sciences, School of likely to develop unwanted drug interactions. hibitors reduce the presystemic metabolism
Pharmacy, University of simvastatin more than the systemic metab-
of California,
olism, leading to increased plasma concentra-
San Francisco, UC
Washington Center, Introduction tions of both simvastatin and simvastatin acid
1608 Rhode Island [19]. Amlodipine, unlike verapamil or diltia-
Avenue, NW, Hypertension and hyperlipidemia often zem, does not significantly inhibit CYP3A4
Washington, DC 20036,
coexist, rendering combination of two or more in humans [14, 18, 19]. Therefore, in the pre-
USA
Howard.Lee@ucsf.edu drugs a rational treatment option. However, vious example by Nishio et al. [20] it is most
leehwd@gmail.com co-administration of more than one drug at likely the increased plasma concentrations of
Park, Lee, Choi et al. 498
simvastatin acid with some contributions from and other applicable laws and regulatory re-
the downstream active metabolites that aug- quirements in Korea, which conform to both
mented biological activity of simvastatin al- the Declaration of Helsinki and other interna-
though the actual mechanism is unclear. tional ethical standards.
When a combination of simvastatin and
amlodipine is indicated, one option may be to
take amlodipine several hours later after the Study design and treatments
time of simvastatin dosing. This non-concur-
rent dosing regimen is sensible because later This was a randomized, parallel, open-la-
presentation of amlodipine to the liver and, to bel study designed to compare the pharma-
some extent, the gut, may enable more effi- cokinetic profiles of simvastatin acid and vari-
cient metabolism of simvastatin without in- ous pharmacodynamic endpoints between two
terference from co-administered amlodipine. amlodipine dosing regimens. Eligible patients
Based on this understanding, it was hypo- randomly received daily oral doses of 20 mg
thesized that non-concurrent dosing of amlo- simvastatin (Zocor) and 5 mg of amlodipine
dipine may result in a reduced plasma level of (Norvasc), purchased in the open market, for
simvastatin acid, compared to concurrent (i.e., 6 weeks based on one of the following sched-
simultaneous) administration. The objective ules: 1) simvastatin and amlodipine at 7:00
of the present study was to compare the phar- PM (i.e., concurrent administration), or 2) sim-
macokinetic profiles of simvastatin acid fol- vastatin at 7:00 PM, followed by amlodipine at
lowing repeated doses of simvastatin, co-ad- 11:00 PM (i.e., non-concurrent administra-
ministered with amlodipine concurrently or tion). To increase drug compliance, a drug
non-concurrently, in comorbid hypertensive diary was distributed and patients were asked
and hyperlipidemic patients. In addition, sev- to record actual date and time of drug intake,
eral pharmacodynamic endpoints such as bl- followed by a weekly telephonic interview.
ood pressure and lipid profile were compared Other antihypertensive agents, lipid low-
between the different dosing regimens of am- ering drugs, and drugs that can affect CYP
lodipine. 3A4 activity (e.g., ketoconazole, itracona-
zole, erythromycin, clarithromycin, and pro-
tease inhibitors) were not allowed during the
Methods study. In addition, patients were instructed
not to take any food or beverages containing
Subjects grapefruit or pomelo at least within 1 week of
study commencement and during the study.
Patients with coexisting hypertension and
hyperlipidemia, aged 30 75 years, were
screened after a 1-week drug wash-out period, Pharmacokinetic sampling and
which was extended to 2 weeks in case of assay
amlodipine because of its long half-life. The
sitting morning systolic (SitSBP) and dia- Subjects were instructed to take the as-
stolic (SitDBP) blood pressure of eligible pa- signed medications at the scheduled times as
tients, measured at the screening and study closely as possible during the study and ad-
initiation visits, fell in the ranges of 140 179 mitted to the clinic on Day 42 around 5:00 PM,
mmHg and 90 114 mmHg, respectively. having an intravenous line inserted for phar-
Fasting blood low density lipoprotein (LDL) macokinetic plasma sampling. A regular meal
cholesterol level was also assessed, and was then provided, and the assigned medica-
should be 125 mg/dl. Patients with hyper- tions were administered under the supervi-
tension or hyperlipidemia of a secondary na- sion of the study personnel. Pharmacokinetic
ture, past medical history of stroke or myocar- samples were obtained at 0 (i.e., pre-dose), 2,
dial infarction, symptomatic arrhythmia or 3, 4, 6, 8, 10, and 15 h post simvastatin dose.
unstable angina, were excluded. This study Patients were discharged in the morning of
was approved by the Korea University Insti- the next day.
tutional Review Board and conducted ac- Plasma concentrations of simvastatin,
cording to the current Good Clinical Practices simvastatin acid, and amlodipine were as-
Attenuated drug interaction by non-concurrent dosing 499
sessed using the validated LC/MS/MS meth- which were compared between the concur-
ods [4, 17], which covered 0.1 20.0 ng/ml rent and non-concurrent amlodipine dosing
(simvastatin), 0.1 2.0 ng/ml (simvastatin groups. Changes from baseline in SitDBP,
acid), and 0.5 40.0 ng/ml (amlodipine), re- fasting total cholesterol, high density lipopro-
spectively. The inter- and intra-assay coeffi- tein (HDL) cholesterol and triglyceride were
cients of variation were less than 11.2%. In also evaluated. The frequency of adverse
addition, the accuracy of the inter- and intra- events and number of subjects experiencing
assay ranged between 88.2 and 109.7%. them were summarized by preferred term and
body system in each amlodipine dose group.
The two sample t-test and c2-test were used
Safety and efficacy assessments for continuous and categorical variables, re-
spectively, using SAS (version 9.1.3, SAS In-
Clinical and laboratory evaluations in- stitute Inc., Cary, NC, USA). A two-tailed
cluding SitSBP and SitDBP measurements, p value of 0.05 or less was regarded as statisti-
EKG, hematology, blood chemistry including cally significant, and no correction for the
lipids profile, and urinalysis were measured multiple comparisons was made due to the
at baseline and Week 6. In addition, compre- exploratory nature of this study.
hensive safety assessments, including physi-
cal examinations and monitoring of adverse
events, were performed throughout the study.
Patients were removed from the study if their Results
SitSBP or SitDBP values were 180 or 115
mmHg, respectively.
Subjects and baseline
characteristics
A total of 17 subjects, aged 42 74, were
Pharmacokinetic analysis randomly assigned to either the concurrent
(n = 9) or non-concurrent (n = 8) amlodipine
The following pharmacokinetic parame-
dosing group (Table 1). One subject from the
ters were estimated from the plasma concen-
non-concurrent group dropped out from the
trations of simvastatin and simvastatin acid:
study because of protocol violation. In addi-
Cmax (the maximum plasma concentration),
tion, 1 subject from the concurrent group did
tmax (the time at which Cmax occurs), and
not take the assigned medication just prior to
AUClast (the area under the plasma concentra-
the pharmacokinetic sampling on Day 42, and
tion-time curve to the last quantifiable con-
was therefore removed from the pharmaco-
centration). Since amlodipine was adminis-
kinetic analysis dataset. Approximately, 60%
tered concurrently or non-concurrently with
were female subjects, and there was no statis-
simvastatin depending on the dosing sched-
tically significant difference between the con-
ule, AUClast and partial AUC, i.e., AUC04
current and non-concurrent amlodipine dos-
(AUC up to 4 h post dose of simvastatin) were
ing groups for blood pressure and lipid profile
estimated to assess the effect of the initial ex-
values.
posure by amlodipine. WinNonlin Pro-
fessional (version 5.2.1, Pharsight Corp.,
Mountain View, CA, USA) was used and AUC
was estimated by the linear trapezoidal method. Pharmacokinetic analysis
Table 2 summarizes the pharmacokinetic
Statistical analysis parameters of simvastatin, simvastatin acid,
and amlodipine. Though not statistically sig-
Descriptive statistics (e.g., mean and stan- nificant, the Cmax and AUClast of simvastatin
dard deviation for continuous variables, fre- acid in the non-concurrent amlodipine dosing
quency and percentage for categorical vari- group were numerically lower than those in
ables) were used to summarize data. Changes the concurrent dosing group (1.2 1.0 vs. 1.9
from baseline at Week 6 in SitSBP and LDL 0.9 ng/ml and 10.3 8.3 vs. 15.6 7.5 h
cholesterol were primary efficacy endpoints, ng/ml, respectively, mean standard devia-
Park, Lee, Choi et al. 500
aMean standard deviation and frequency (%) are shown for continuous and
categorical variables, respectively. None of the variables were statistically sig- Discussion
nificant (i.e., p > 0.05) when compared between the concurrent and non-concur-
rent amlodipine dosing groups. SitSBP = sitting systolic blood pressure, SitDBP
= sitting diastolic blood pressure, LDL = low density lipoprotein, HDL = high density
The key finding of the present study is that
lipoprotein. the non-concurrent administration of amlo-
dipine (i.e., 4 h after simvastatin) resulted in a
numerically lower systemic level of simvas-
tatin acid while the efficacy of statin appeared
to be maintained. The Cmax and AUClast and
of simvastatin acid in the non-concurrent
amlodipine dosing group were 63.2% and
66.0%, respectively, of the values obtained in
the concurrent group (Table 2) (Figure 1), but
the changes in lipid profiles from baseline
were comparable between different amlodi-
pine dosing groups after 6 weeks of simvasta-
tin treatment (Table 3).
It is also interesting to note that the
AUClast and AUC04 for amlodipine were
similar between the concurrent and the non-
concurrent amlodipine dosing groups (Table
2). This observation may reflect amlodipines
flat (i.e., less fluctuating) pharmacokinetic
profile at steady state due to its long half-life
[25]. Therefore, the systemic level of amlo-
dipine during the post-absorption phase is un-
Figure 1. Mean plasma concentration-time pro-
files of simvastatin (left panel) and simvastatin acid likely to be the mechanism for the different
(right panel) on Day 42 for concurrent () and non- levels of simvastatin acid shown in the pres-
concurrent (l) dosing of amlodipine. Error bars ent study.
represent the standard deviation at each sampling
time.
The selective increases in the AUClast and
Cmax of simvastatin acid seen in the present
study support the notion that amlodipine does
tion). This reduced systemic level of simva- not significantly inhibit CYP3A4. If amlo-
statin acid in the non-concurrent amlodipine dipine inhibited CYP3A4 in the concurrent
dosing group is clearly shown in Figure 1 dosing group, which did not occur or was mit-
(right panel). In addition, the tmax of sim- igated in the non-concurrent group, both
vastatin acid in the concurrent amlodipine lactone and acid forms of simvastatin would
dosing group was delayed by 2.1 h. have reduced in the latter group. For example,
Attenuated drug interaction by non-concurrent dosing 501
aMean standard deviation are shown. None of the variables were statistically significant (i.e., p > 0.05) when compared between the
concurrent and non-concurrent amlodipine dosing groups.
but this occurs only after the majority of the metabolizing enzymes. We strongly believe
dosed simvastatin has already been absorbed, that the present study has answered this criti-
giving the drug sufficient time to be effec- cal question.
tively metabolized. The present study does have some draw-
Whatever the mechanism might be, the re- backs, one of which is its parallel design. The
sults of the present study have several impor- direction and extent of pharmacokinetic drug
tant clinical implications. HMG-CoA reduc- interaction can be best evaluated using the
tase inhibitors, collectively known as statins, crossover design, in which each subject serves
can cause serious adverse drug reactions such as his or her own control. However, we also
as elevation in liver enzymes and various wanted to examine pharmacodynamic inter-
forms of myopathy [10]. Since simvastatin action, in which long-term follow-up for each
acid has the greatest inhibitory effect on the individual is a mandate, rendering the parallel
HMG-CoA reductase compared to the other design logistically more feasible and practi-
active CYP3A4 downstream metabolites, sim- cal. Another drawback of the present study is
vastatin acid may be the major, if not the only, the fact that it did not have groups adminis-
moiety responsible for adverse drug reactions tered simvastatin or amlodipine only. How-
by simvastatin. ever, because eligible patients had to have co-
Therefore, the increased systemic levels existing hypertension and hyperlipidemia,
of simvastatin acid following a combined treatment only for one condition was not con-
treatment with amlodipine is likely to increase sidered ethically appropriate. Its small sam-
the risk of harmful side effects of simvastatin ple size, mainly due to the exploratory nature
[5]. However, it may be possible to reduce the of the present study, is another limitation.
likelihood of this kind of drug interaction, at Based on this concern, we estimated the post
least from the pharmacokinetic perspective, hoc power to detect the observed difference in
by staggering the administration times of each the logarithm transformed Cmax and AUClast
drug, as exemplified in the present study. of simvastatin acid between amlodipine dos-
Previously, staggered or non-concurrent ing groups, with a two-tailed a error of 0.05.
administrations of the antiulcer agent sucral- 33% and 35.4% were the post hoc powers for
fate with the quinolone antimicrobials such as Cmax and AUClast, respectively, which means
norfloxacin/ofloxacin [15] and sparfloxacin the present study was underpowered. All of
[12] were investigated. By administering the these issues, except for the evaluation of phar-
quinolone antimicrobials 2 h (norfloxa- macodynamic interaction, can be addressed in
cin/ofloxacin) or 4 h (sparfloxacin) before a pharmacokinetic drug interaction study in
sucralfate, the oral bioavailability of the healthy volunteers or patients with mild hy-
quinolone antimicrobials were comparable to pertension using a full randomized crossover
that seen with the quinolones administered design, in which simvastatin only, amlodipine
alone. In addition, a study in dogs reported a only, concurrent and non-concurrent amlodip-
mitigated pharmacokinetic drug interaction ine and simvastatin groups are examined.
between midazolam and diltiazem when dilti- This study has been planned and will be car-
azem was administered at least 1 h after ried out shortly.
midazolam [22]. Midazolam is regarded as In conclusion, the non-concurrent ad-
the most sensitive substrate of CYP3A4 [7], ministration of amlodipine with regard to
and diltiazem is one of the strong inhibitors the dosing time of simvastatin resulted in
of CYP3A4 [11]. Thus, the preventability of less pharmacokinetic drug interaction between
pharmacokinetic drug interaction by non-con- amlodipine and simvastatin, which may lead
current dosing might have been unduly em- to an improved safety profile compared to
phasized in this study. Similarly, only a single concurrent dosing regimen. Non-concurrent
administration of midazolam and diltiazem dosing may be a useful and safe therapeutic
was examined, leaving an important question option for patients who require two or more
unanswered; whether the reduced pharmaco- drugs administered concomitantly, but who
kinetic drug interaction by non-concurrent are likely to develop unwanted drug interac-
dosing could be reproduced after multiple tions.
doses even when neither drug is a sensitive
substrate or a strong inhibitor of certain drug
Attenuated drug interaction by non-concurrent dosing 503
The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate.