Professional Documents
Culture Documents
Effect of High-Dose Metronidazole On Pharmacokinetics of Oral Budesonide and Vice Versa
Effect of High-Dose Metronidazole On Pharmacokinetics of Oral Budesonide and Vice Versa
Pharmacology
http://www.jclinpharm.org
Published by:
http://www.sagepublications.com
On behalf of:
American College of Clinical Pharmacology
Additional services and information for The Journal of Clinical Pharmacology can be found at:
Subscriptions: http://www.jclinpharm.org/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Recent case reports suggest that addition of high-dose budesonide (90% confidence interval, 79%-115%) nor on the
metronidazole might be associated with elevated plasma AUC ratios of 6β-hydroxybudesonide/budesonide and 16α-
concentrations of substrates of cytochrome P450 (CYP) 3A. hydroxyprednisolone/budesonide. In parallel, metronidazole
Because patients with fistulizing Crohn’s disease benefit by did not significantly alter formation of 6β-hydroxycortisol.
using high doses of metronidazole for prolonged periods, this Vice versa, budesonide did not affect the AUC of metronida-
study’s primary aim was to evaluate the effect of high-dose zole (90% confidence interval, 91%-100%). The authors
metronidazole on the pharmacokinetics of oral budesonide, conclude that in contrast to concomitant intake of other imi-
a sensitive substrate of CYP3A commonly prescribed in acute dazoles such as ketoconazole, concomitant intake of metron-
inflammatory bowel disease. Twelve healthy adults received idazole may not lead to serious safety concerns due to
1.5 g metronidazole per day over 1 week. The CYP3A-depen- elevated systemic concentrations of the glucocorticoid
dent metabolic profile of an oral dose of budesonide (3 mg) budesonide.
and that of endogenous cortisol were compared intraindivid-
ually before and after administration of metronidazole. There Keywords: Budesonide; cortisol; CYP3A; metronidazole
was neither a significant effect of high-dose metronidazole on Journal of Clinical Pharmacology, 2007;47:1532-1539
the area under the plasma concentration-time curve (AUC) of © 2007 the American College of Clinical Pharmacology
plasma concentration-time curve (AUC) of CYP3A disease by coadministration of the older synthetic
substrates.6 Metronidazole may inhibit CYP3A glucocorticoid, prednisone.20
according to in vitro data,7 but it has been shown
that low-dose metronidazole (400 mg twice daily)
did not affect the pharmacokinetics of the CYP3A METHODS
substrate midazolam.8 However, 3 recent case
reports describing (1) the elevation of tacrolimus Subjects
trough concentrations (dual substrate of CYP3A and
P-glycoprotein [P-gp]) by addition of 2 g metronida- Twelve healthy male Caucasians (30.3 ± 5.6 years,
zole per day, (2) the occurrence of Torsades de 74.9 ± 9.8 kg, 23.2 ± 2.6 kg/m2; mean ± SD) were
Pointes due to coadministration of 1.5 g metronida- enrolled in the drug interaction study during May
zole per day and amiodarone (substrate of CYP3A), 2006. All subjects were nonsmokers. Intake of any
and (3) an indication for an unexpected drug inter- medication within the 2 weeks before or during the
action between metronidazole (25 mg/kg per day) conduct of the trial precluded participation. Further
and oral budesonide in a patient with Crohn’s dis- exclusion criteria were administration of any gluco-
ease substantiated our decision to test the effect of corticoid within 6 weeks before the first study day,
high-dose metronidazole on the pharmacokinetics of use of drugs during the 4 weeks before the first study
oral budesonide.9-11 day that might influence CYP3A activity,21 and
In a recent guidance document on drug interaction intake of grapefruit within 1 week before the first
studies by the US Food and Drug Administration study day or during the trial.
defining the best practices, budesonide is given as
sensitive substrate of CYP3A.12 The drug is a newer Study Design
synthetic glucocorticoid with high glucocorticoid
receptor binding affinity and a low rate of systemic The investigation was conducted as a fixed-order
side effects because of low systemic bioavailabil- study. In the morning on study days 1 and 9, a sin-
ity.13 Absolute bioavailability of oral budesonide is gle oral dose of 3 mg budesonide was administered.
10% due to extensive first-pass metabolism in gut On study days 2 to 8, each subject received 1.5 g
and liver by CYP3A enzymes.14 The main metabo- metronidazole per day (divided into a morning and
lites formed via CYP3A are 6β-hydroxybudesonide an evening dose). On study day 9, last metronidazole
and 16α-hydroxyprednisolone.15 Moreover, budes- (750 mg) was given simultaneously with oral budes-
onide is a substrate of the intestinal drug efflux pump, onide. To ensure compliance, each subject had to
P-gp.16 Oral budesonide is approved for the treat- report to the study site for each drug intake (mouth
ment of mild to moderate exacerbations of Crohn’s check). Subjects were hospitalized the evening
disease, the incidence of which has been shown to before study days 1, 8, and 9. After an overnight fast,
be increasing rapidly in Western countries.17,18 a standardized lunch was served not until 4 hours
For safety reasons, it must be clarified if patients after ingestion of the study drug(s), and intake of fluid
with Crohn’s disease who are using the standard was standardized over 8 hours. Pharmacokinetic
dosage of oral budesonide are at an increased risk of profiling of budesonide was performed identically
Cushingoid symptoms and other steroid-related side on study days 1 and 9 by determination of budes-
effects during treatment with metronidazole because onide and the 2 CYP3A-dependent metabolites (6β-
of elevated systemic budesonide concentrations hydroxybudesonide, 16α-hydroxyprednisolone) in
resulting from inhibition of biotransformation via plasma before and during 24 hours after drug intake.
CYP3A. Therefore, the primary aim of the study was In detail, blood samples were collected just before
to investigate the impact of 1.5 g metronidazole per and 2, 2.5, 3, 3.5, 4, 4.5, 5, 5.5, 6, 7, 8, 10, 12, and
day, given over 1 week, on the CYP3A-dependent 24 hours after drug administration. Pharmacokinetic
metabolic profile of budesonide and that of endoge- profiling of metronidazole was performed on study
nous cortisol. CYP3A catalyzes the transformation of days 8 and 9 by measuring metronidazole in plasma;
cortisol to 6β-hydroxycortisol.19 The secondary aim blood samples were collected just before and 0.5, 1,
was to study vice versa a possible effect of budes- 1.5, 2, 3, 4, 6, 8, and 12 hours after drug administra-
onide on the disposition of metronidazole due to a tion. To prove that the dosing strategy was adequate
report on a significant reduction of metronidazole to achieve steady state during the interaction, blood
plasma concentrations in 6 patients with Crohn’s samples for trough concentrations of metronidazole
were collected before each morning dosing on study oral clearance (CL/f = dose/AUC0-∞) and apparent
days 6 to 8. On study days 1 and 9, urine was col- volume of distribution (Vd/f = dose/[AUC0-∞⋅λ]) were
lected for 12 hours after dosing to determine urinary normalized for body weight. Peak trough fluctuation
excretion of cortisol and 6β-hydroxycortisol. Plasma (PTF) of metronidazole plasma concentrations at
and urine were stored at –20°C until batch analysis. steady state was derived from the equation PTF =
100⋅(Css,max – Css,min)/Css,av with Css,av = AUCss,0-12 h/12.
Analytical Methods Ratios of metabolite formation (AUCMet/ AUCBudesonide,
where Met is the CYP3A-dependent metabolite), such
Concentrations of budesonide, 6β-hydroxybudesonide, as AUC0-24 h of 6β-hydroxybudesonide to AUC0-24 h of
and 16α-hydroxyprednisolone in plasma were deter- budesonide, were used as indices of CYP3A activity.
mined by validated liquid chromatography tandem In addition, possible induction or inhibition of
mass spectrometry as described previously.22 The CYP3A enzymes by metronidazole was evaluated by
mass-to-charge transition (m/z) was monitored at measuring the cumulative amount of cortisol and
489.2 to 339.2 for budesonide, 505.0 to 373.1 for 6β- 6β-hydroxycortisol excreted into the urine during the
hydroxybudesonide, 374.9 to 327.1 for 16α-hydrox- 12-hour collection period (Ae0-12 h); metabolic ratios
yprednisolone, and 492.8 to 375.2 for the internal (Ae6β-OH-cortisol/Aecortisol) were calculated in each subject.
standard (flunisolide). The lower limits of quantifi-
cation in plasma were 0.1 ng/mL for budesonide and Statistical Analysis
6β-hydroxybudesonide and 0.5 ng/mL for 16α-
hydroxyprednisolone. Between-day and within-day The study had a 94% power to detect a 20% differ-
coefficients of variation of quality controls were less ence in the plasma AUC of budesonide between
than 15%. Concentrations of cortisol and 6β-hydrox- study day 1 and study day 9 with a P value of less
ycortisol in urine were determined by validated liq- than .05. Sample size calculation was based on the
uid chromatography tandem mass spectrometry. The coefficient of variation of the budesonide AUC in a
mass-to-charge transition (m/z) was monitored at previous study.22 Statistical analysis was performed
421.1 to 331.2 for cortisol, 437.1 to 347.0 for 6β- by use of the software package GraphPad Instat
hydroxycortisol, and 451.2 to 361.1 for the internal (Version 3.05, GraphPad Software, Inc., San Diego,
standard (dexamethasone). The lower limit of quan- California). Wilcoxon matched pairs test was per-
tification was 1 ng/mL. Between-day and within-day formed to compare pharmacokinetic/pharmacody-
coefficients of variation of quality controls were less namic parameters between 2 different study days
than 10%. Concentrations of metronidazole in (eg, day 1 vs day 9). Two-sided P < .05 was regarded
plasma were determined by standard high-perfor- as statistically significant. According to the guide-
mance liquid chromatography.23 The lower limit of lines by the US Food and Drug Administration and
quantification was 0.03 μg/mL. Between-day and the European Medicines Agency, the ratio of means
within-day coefficients of variation of quality con- with confidence intervals (CIs) of AUC and Cmax are
trols were less than 7%. the characteristics to determine equivalence for orally
administered drugs; parametric testing using analy-
Pharmacokinetic and sis of variance (ANOVA) on log-transformed data is
Pharmacodynamic Analyses the rule. Therefore, 90% CIs of the log-transformed
parameters AUC0-24 h [AUCss,0-12 h] and Cmax [Css,max] of
Standard model-independent methods were used to budesonide [metronidazole] before and during
determine the pharmacokinetic parameters of interest metronidazole [budesonide] administration were
(Kinetica Version 4.0, Thermo Electron Corporation, derived from the residual variance in multifactor
Philadelphia, Pennsylvania). Peak plasma concentra- ANOVA. The 90% CIs were based on the ratios of
tion (Cmax) and time of Cmax (tmax) were taken directly the population means (during/before comedication).
from the plasma concentration-time curves. Areas
under the plasma concentration-time curves covering Ethics
different intervals (eg, AUC0-24 h) were determined by
a combination of linear and logarithmic trapezoidal The study protocol was approved by the ethics com-
methods with extrapolation to infinity (AUC0-∞). mittee of the University Hospital Tübingen. The
Terminal elimination half-life (t1/2) was calculated from study was conducted in accordance with the ethical
the final slope of the log-linear concentration-time guidelines of the Declaration of Helsinki and the Inter-
curve by least squares linear regression. Apparent national Conference on Harmonization guidelines
for good clinical practice. Written informed consent Table I Pharmacokinetic Parameters of
was obtained from each participant. The project is Budesonide and 2 CYP3A-Dependent Metabolites
registered in the public trials registry sponsored by the
US National Library of Medicine (www.clinicaltrials Baseline With Metronidazole
.gov, #NCT00338910). Budesonide
Cmax, ng/mL 1.00 ± 0.35 1.04 ± 0.40
RESULTS tmax, h 4.5 (4.1−6.2) 4.5 (3.9−4.8)
tlag, h 3.0 (2.8−3.8) 2.8 (2.0−3.4)
All subjects completed the study according to the AUC0−∞, h⋅ng/mL 4.88 ± 2.04 4.77 ± 2.11
protocol with excellent compliance. The study t1/2, h 2.6 (2.1−5.1) 3.6 (2.6−6.0)
drugs were well tolerated. There was no serious CL/f, L/min/kg 0.17 ± 0.08 0.18 ± 0.09
Vd/f, L/kg 43.8 ± 15.2 60.6 ± 38.1
adverse drug reaction. Metronidazole trough con-
6β-OH-Budesonide
centrations in plasma did not differ significantly
Cmax, ng/mL 1.36 ± 0.56 1.36 ± 0.45
between study days 6 to 8 (ANOVA), which means tmax, h 4.8 (4.1−6.9) 4.5 (4.0−5.0)
that steady state was achieved on study day 8. tlag, h 3.0 (2.1−3.8) 2.8 (2.0−3.3)
AUC0−∞, h⋅ng/mL 9.11 ± 3.64 8.52 ± 2.64
Effect of Metronidazole t1/2, h 4.8 (3.5−6.0) 3.8 (3.4−7.5)
on Budesonide and Cortisol 16α-OH-Prednisolone
Cmax, ng/mL 8.94 ± 3.76 9.92 ± 2.70
Pharmacokinetic parameters of budesonide and both tmax, h 5.3 (4.5−6.8) 4.5 (4.0−5.1)
tlag, h 2.8 (1.9−3.6) 2.5 (1.8−3.1)
CYP3A-dependent metabolites (mean ± SD or median
AUC0−∞, h⋅ng/mL 33.27 ± 6.29 33.69 ± 6.19
with 95% CI) are given in Table I for comparison of t1/2, h 1.6 (1.3−2.7) 2.0 (1.6−2.5)
baseline (study day 1) with the effect of high-dose
metronidazole (study day 9). Wilcoxon testing did not Cmax, peak plasma concentration; tmax, time of peak plasma concentra-
tion; tlag, lag time; AUC0−∞, area under the plasma concentration-time
reveal a significant difference between study day 1 curve extrapolated to infinity; t1/2, terminal elimination half-life; CL/f,
and study day 9 in any of the parameters displayed apparent oral clearance; Vd/f, apparent volume of distribution.
(eg, AUC0-∞ of budesonide: 4.88 ± 2.04 h⋅ng/mL vs 4.77
± 2.11 h⋅ng/mL). Mean plasma concentration-time
1.5
curves of budesonide, 6β-hydroxybudesonide, and
Day 1
Plasma Budesonide (ng/mL)
15
Day 1
1.5
Day 9
AUCMet/AUCBudesonide
10
1.0
0.5
5
0
0 2 4 6 8 10 12 14 16 18 20 22 24
0
12.0 20
Day 1
10.0
Day 9 15
8.0
AeMet/AeCortisol
6.0
10
4.0
2.0 5
0
0 2 4 6 8 10 12 14 16 18 20 22 24 0
Time (h) Day 1 Day 9
concentrations (Css,min) of metronidazole were signif- We observed no increase in the plasma AUC of
icantly lower on study day 9 than on study day 8 budesonide (point estimate 0.95) upon coadminis-
(P < .05). The 90% CIs of both AUC and the Cmax ratio tration with metronidazole. Daily intake of 1.5 g
of metronidazole during budesonide administration metronidazole over 1 week did not significantly
relative to baseline were narrow (AUC: 0.91-1.00, affect the biotransformation of oral budesonide via
Cmax: 0.90-1.08), with the ratios of means being near CYP3A as shown by extensive analysis of metabolite
to 1.00 (AUC: 0.96, Cmax: 0.99). kinetics in plasma; AUC ratios of 6β-hydroxybudes-
onide/budesonide and 16α-hydroxyprednisolone/
DISCUSSION budesonide have been validated previously as a marker
of CYP3A activity using the prototype inducer
This human study shows that high-dose metronidazole rifampicin.24 Likewise, we found no significant
does not affect the pharmacokinetics of oral budesonide. effect by high-dose metronidazole on the formation
Table II Pharmacokinetic Parameters of treatment period longer than 3 days for our system-
Metronidazole atic evaluation. However, no such drug interaction
was found.
Baseline With Budesonide Interestingly, there are case reports describing an
Css,max, μg/mL 27.6 ± 5.1 26.9 ± 2.9 increase in plasma concentrations of dual substrates
tss,max, h 0.5 (0.5−0.8) 0.5 (0.5−0.8) of CYP3A and P-gp with a narrow therapeutic range
Css,min, μg/mL 11.4 ± 2.5 10.6 ± 2.1* (tacrolimus, cyclosporine, carbamazepine, quini-
AUCss,0-12 h, h⋅μg/mL 193.6 ± 28.0 184.6 ± 23.6* dine) by coadministration of metronidazole9,26-30; for
PTF, % 101.4 ± 22.1 107.9 ± 23.3 example, 2 case reports document marked elevation
in tacrolimus concentrations with the addition of
Css,max, peak plasma concentration at steady state; tss,max, time of peak
plasma concentration at steady state; Css,min, trough plasma concentra- metronidazole,9,26 and 3 case reports describe a
tion at steady state; AUCss,0-12 h, area under the plasma concentration- harmful increase in cyclosporine concentrations in
time curve at steady state during the dosing interval of 12 hours; PTF, patients due to concurrent metronidazole.26,29,30
peak trough fluctuation at steady state.
*P < .05 vs baseline (Wilcoxon test). Some of the authors suggested that metronidazole
should be added to the list of CYP inhibitors that can
produce clinically important drug interactions.
The publication on the concurrent use of amio-
35 darone and metronidazole highlighted a serious phar-
macodynamic effect on the conduction system of the
Plasma Metronidazole (µg/mL)
30 Day 8
heart with positive dechallenge but did not provide
25 Day 9 information on drug concentrations in plasma for a
20 precise interpretation.10 A study patient who had
15
taken metronidazole (1.2 g per day during 12 days)
until the evening before the first administration of
10 oral budesonide showed remarkably delayed absorp-
5 tion of budesonide, which was no more observed 1
week later following withdrawal of metronidazole.
0
0 2 4 6 8 10 12 14 16 18 20 22 24 Rechallenge was impossible in that subject because of
resection of the ileum a short time later. The results of
Time (h) our study do not confirm the previous observation: lag
time of budesonide was not prolonged by metron-
Figure 6. Plasma metronidazole concentration-time curves in idazole. Altered intraluminal pH, delayed gastric
12 subjects during steady-state (twice-daily dosing of metronida-
zole, total 1.5 g/day) following an oral morning dose of 750 mg emptying, or intestinal transit might explain delayed
metronidazole on study day 8 without (solid circle) and on study absorption of the enteric-coated budesonide prepa-
day 9 together with (open circle) a single oral dose of 3 mg budes- ration in an isolated case.11
onide. Data are presented as mean and SD.
We started our trial knowing that there is no
simple relationship between in vitro inhibition
of 6β-hydroxycortisol, which is an older validated potency of a substance and magnitude of drug inter-
test for evaluating induction and inhibition of action in the clinical setting. Some CYP3A
CYP3A enzymes.19,25 inhibitors that did not demonstrate high in vitro
A previous clinical trial in 10 healthy volunteers potency had been shown to cause drug interactions
examining the CYP3A-altering properties of low- of more than a 2-fold increase in AUC. These
dose metronidazole (400 mg twice daily during 3 include CYP3A inhibitors causing mechanism-based
days) found no effect (midazolam plasma AUC ratio irreversible inactivation (eg, clarithromycin, dilti-
0.93).8 As that study evaluated only the effect of low- azem).31 Unlike other human CYP enzymes, CYP3A
dose metronidazole during a short course of treatment, offers the complicating factor that there appear to be
that result does not exclude the possibility of a sig- 3 different substrate-binding types.32 In addition, the
nificant effect of metronidazole on CYP3A activity at magnitude of an interaction on a CYP3A probe will
a higher dosage or a longer duration of treatment as depend on the amount of intestinal versus hepatic
suggested in the newer case reports. Because testing extraction with those drugs that have a greater intesti-
for a possible drug interaction should maximize the nal extraction with larger interactions.31 Thus, very
possibility of finding an interaction, we used the low bioavailability of oral budesonide at baseline might
maximum approved dose of metronidazole with a have increased several-fold in the face of CYP3A
16. Dilger K, Schwab M, Fromm MF. Identification of budesonide 28. Spina E, Pisani F, Perucca E. Clinically significant pharmaco-
and prednisone as substrates of the intestinal drug efflux pump kinetic drug interactions with carbamazepine: an update. Clin
P-glycoprotein. Inflamm Bowel Dis. 2004;10:578-583. Pharmacokinet. 1996;31:198-214.
17. Bar-Meir S, Chowers Y, Lavy A, et al. Budesonide versus pred- 29. Vincent F, Glotz D, Kreft-Jais C, Boudjeltia S, Duboust A,
nisone in the treatment of active Crohn’s disease. The Israeli Bariety J. Acute kidney failure in a patient treated with
Budesonide Study Group. Gastroenterology. 1998;115:835-840. cyclosporine A and metronidazole following kidney transplanta-
18. Armitage EL, Aldhous MC, Anderson N, et al. Incidence of tion. Therapie. 1994;49:155.
juvenile-onset Crohn’s disease in Scotland: association with north- 30. Zylber-Katz E, Rubinger D, Berlatzky Y. Cyclosporine interac-
ern latitude and affluence. Gastroenterology. 2004;127:1051-1057. tions with metronidazole and cimetidine. Drug Intell Clin Pharm.
19. Ged C, Rouillon JM, Pichard L, et al. The increase in urinary 1988;22:504-505.
excretion of 6 beta-hydroxycortisol as a marker of human hepatic 31. Obach RS, Walsky RL, Venkatakrishnan K, Houston JB,
cytochrome P450IIIA induction. Br J Clin Pharmacol. 1989;28: Tremaine LM. In vitro cytochrome P450 inhibition data and the
373-387. prediction of drug-drug interactions: qualitative relationships,
20. Eradiri O, Jamali F, Thomson AB. Interaction of metronidazole quantitative predictions, and the rank-order approach. Clin
with phenobarbital, cimetidine, prednisone, and sulfasalazine in Pharmacol Ther. 2005;78:582-592.
Crohn’s disease. Biopharm Drug Dispos. 1988;9:219-227. 32. Kenworthy KE, Bloomer JC, Clarke SE, Houston JB. CYP3A4
21. Flockhart DA. Drug Interactions—Cytochrome P450 System. drug interactions: correlation of 10 in vitro probe substrates. Br J
Available at: http://medicine.iupui.edu/flockhart/table.htm. Clin Pharmacol. 1999;48:716-727.
Accessed July 3, 2007. 33. Bressler B, Sands BE. Review article: Medical therapy for fis-
22. Hempfling W, Grunhage F, Dilger K, Reichel C, Beuers U, tulizing Crohn’s disease. Aliment Pharmacol Ther. 2006;24:
Sauerbruch T. Pharmacokinetics and pharmacodynamic action of 1283-1293.
budesonide in early- and late-stage primary biliary cirrhosis. 34. Perencevich M, Burakoff R. Use of antibiotics in the treatment of
Hepatology. 2003;38:196-202. inflammatory bowel disease. Inflamm Bowel Dis. 2006;12:651-664.
23. Sirowej H, Bussemas HH, Harhoff F. Determination of drugs 35. Rutgeerts P, Hiele M, Geboes K, et al. Controlled trial of
by HPLC with simultaneous UV-detection at two wavelengths. metronidazole treatment for prevention of Crohn’s recurrence
J Clin Chem Clin Biochem. 1985;23:560. after ileal resection. Gastroenterology. 1995;108:1617-1621.
24. Dilger K, Denk A, Heeg MH, Beuers U. No relevant effect of 36. Sutherland L, Singleton J, Sessions J, et al. Double blind,
ursodeoxycholic acid on cytochrome P450 3A metabolism in pri- placebo controlled trial of metronidazole in Crohn’s disease. Gut.
mary biliary cirrhosis. Hepatology. 2005;41:595-602. 1991;32:1071-1075.
25. Galteau MM, Shamsa F. Urinary 6beta-hydroxycortisol: a val- 37. Cudmore SL, Delgaty KL, Hayward-McClelland SF, Petrin DP,
idated test for evaluating drug induction or drug inhibition medi- Garber GE. Treatment of infections caused by metronidazole-resistant
ated through CYP3A in humans and in animals. Eur J Clin Trichomonas vaginalis. Clin Microbiol Rev. 2004;17:783-793.
Pharmacol. 2003;59:713-733. 38. Lamp KC, Freeman CD, Klutman NE, Lacy MK.
26. Herzig K, Johnson DW. Marked elevation of blood cyclosporin Pharmacokinetics and pharmacodynamics of the nitroimidazole
and tacrolimus levels due to concurrent metronidazole therapy. antimicrobials. Clin Pharmacokinet. 1999;36:353-373.
Nephrol Dial Transplant. 1999;14:521-523.
27. Cooke CE, Sklar GE, Nappi JM. Possible pharmacokinetic
interaction with quinidine: ciprofloxacin or metronidazole? Ann
Pharmacother. 1996;30:364-366.