Assessment of inhibitory effects on major human cytochrome P450 enzymes by spasmolytics used in the treatment of overactive bladder syndrome

Abstract

Background:

The objective of this study was to examine the inhibitory potential of darifenacin, fesoterodine, oxybutynin, propiverine, solifenacin, tolterodine and trospium chloride on the seven major human cytochrome P450 enzymes (CYP) by using a standardized and validated seven-in-one cytochrome P450 cocktail inhibition assay.

Methods:

An in vitro cocktail of seven highly selective probe substrates was incubated with human liver microsomes and varying concentrations of the seven test compounds. The major metabolites of the probe substrates were simultaneously analysed using a validated liquid chromatography tandem mass spectrometry (LC-MS/MS) method. Enzyme kinetics were estimated by determining IC₅₀ and K_i values via nonlinear regression. Obtained K_i values were used for predictions of potential clinical impact of the inhibition using a static mechanistic prediction model.

Results:

In this study, 49 IC₅₀ experiments were conducted. In six cases, IC₅₀ values lower than the calculated threshold for drug–drug interactions (DDIs) in the gut wall were observed. In these cases, no increase in inhibition was determined after a 30 min preincubation. Considering a typical dosing regimen and applying the obtained K_i values of 0.72 µM (darifenacin, 15 mg daily) and 7.2 µM [propiverine, 30 mg daily, immediate release (IR)] for the inhibition of CYP2D6 yielded a predicted 1.9-fold and 1.4-fold increase in the area under the curve (AUC) of debrisoquine (CYP2D6 substrate), respectively. Due to the inhibition of the particular intestinal CYP3A4, the obtained K_i values of 14 µM of propiverine (30 mg daily, IR) resulted in a predicted doubling of the AUC for midazolam (CYP3A4 substrate).

Conclusions:

In vitro/in vivo extrapolation based on pharmacokinetic data and the conducted screening experiments yielded similar effects of darifenacin on CYP2D6 and propiverine on CYP3A4 as obtained in separately conducted in vivo DDI studies. As a novel finding, propiverine was identified to potentially inhibit CYP2D6 at clinically occurring concentrations.

Keywords

overactive bladder syndrome spasmolytics antimuscarinic agents cytochrome P450 drug-drug interactions N-in-one cocktail assay inhibition

Background

Approximately one in six adults suffers from overactive bladder syndrome (OAB) with the prevalence increasing steadily at advancing age.¹^,² The major pharmacological target in the alleviation of OAB symptoms is the modulation of muscarinic receptors, in particular M₃ receptors which have been identified to be primarily responsible for detrusor contraction.³^–8 Antimuscarinic agents, which are also referred to as anticholinergics or spasmolytics, are the cornerstone of current symptomatic treatment of OAB symptoms and are advocated by all international expert associations.⁴^,⁹ Long-term treatment with antimuscarinics leads to adverse drug reactions (ADRs) linked to the mechanism of action, that is, dry mouth, constipation, blurred vision, tachycardia and confusion.⁷^,⁹^,¹⁰ The severity of these ADRs is potentially increased by altered pharmacokinetics of the antimuscarinic agent.¹⁰^,¹¹ All antimuscarinics used for the treatment of OAB, with the exception of trospium chloride, are either metabolized by cytochrome P450 enzymes CYP2D6 or CYP3A4.⁹^,¹²^,¹³ It is therefore well known that caution needs to be applied when antimuscarinic drugs such as darifenacin, fesoterodine, oxybutynin, propiverine, solifenacin and tolterodine are coadministered with CYP2D6 and CYP3A4 inhibitors, such as quinidine, paroxetine or ketoconazole.¹⁴^–17 However, there is limited information regarding the inhibitory effect of antimuscarinics on the seven major human CYPs that mediate 90% of all CYP-mediated drug metabolism.¹⁸

Thus, the primary objective of this study was to examine the direct and time-dependent inhibitory potential of the antimuscarinics darifenacin, fesoterodine, oxybutynin, propiverine, solifenacin, tolterodine and trospium chloride on major human CYPs using a previously published seven-in-one inhibition assay.¹⁹ Based on the screening results, the most potent in vitro interactions were characterized in detail (i.e. mode of inhibition) and mechanistic static models were applied in order to quantitatively estimate the in vivo drug–drug interaction (DDI) effect.

Materials and methods

The development and validation of the seven-in-one human CYP inhibition assay and a detailed description of data analysis has recently been published.¹⁹

Chemicals

The tested antimuscarinic drugs were obtained from Cayman Chemical (Ann Arbor, MI, USA). Pooled human liver microsomes (HLMs; 20-donor pool) and nicotinamide adenine dinucleotide phosphate (NADPH)-regenerating system were obtained from BD Biosciences (Woburn, MA, USA). All other chemicals were of analytical grade or higher. All stock solutions were prepared in water, acetonitrile (ACN) or a mixture of both, depending on the individual solubility of the compound, and afterwards diluted to appropriate working solutions in 80/20 (v/v) water/ACN.

Microsomal incubation conditions

General incubation conditions and sample preparation

All microsomal incubations were performed at least in triplicate. For the K_i experiments the ACN content was set to 1% to minimize the effect of the organic solvent on the enzyme activity. In all other experiments, the organic solvent content was standardized at 2% ACN in order to make poorly soluble, lipophilic test compounds available for the screening tool used. Incubation mixtures contained 0.1 M phosphate buffer (pH 7.4), HLM (0.25 mg protein/ml), NADPH-regenerating system containing NADP⁺ (1.3 mm), glucose 6 phosphate (3.3 mm), glucose 6 phosphate dehydrogenase (0.4 U/ml), magnesium chloride (3.3 mm) and varying concentrations of the seven tested antimuscarinic drugs.

Direct inhibition assay and time-dependent inhibition assay

Direct inhibition of major human CYPs by the tested antimuscarinic drugs was assessed by incubating five to six concentrations of the test compound with the cocktail of seven substrates for 10 min (for substrate concentrations, see Table 1).

Table 1.

Probe substrates, enzyme reactions, monitored major metabolite and substrate concentrations used in the cocktail screening assay.

Probe substrate	CYP	Substrate concentration (µM)	Detected major metabolite
Caffeine	1A2	220	Paraxanthine
Bupropion	2B6	10	2-Hydroxybupropion
Rosiglitazone	2C8	5	N-Desmethylrosiglitazone
Tolbutamide	2C9	67	4-Hydroxytolbutamide
Omeprazole	2C19	17	5-Hydroxyomeprazole
Dextromethorphan	2D6	2	Dextrorphan
Midazolam	3A4	3	1-Hydroxymidazolam

CYP, cytochrome P450.

The incubation concentrations of the antimuscarinics were chosen according to their solubility and their calculated threshold for potential in vivo DDIs (see Table 2). The obtained IC₅₀ values were subsequently converted to K_i values via the Cheng–Prusoff equation for competitive inhibition. These calculated K_i values were then compared with the concentrations of the respective drug at the enzyme site with the maximum exposure, that is, gut wall (see Table 2). This was done to minimize the risk of missing clinically relevant DDIs and due to the fact that the in vivo plasma concentration of antimuscarinics is scarce. If the IC₅₀ values obtained during the direct inhibition experiments did not exclude an in vivo DDI, single point inactivation experiments were performed to identify time-dependent inhibition of the tested antimuscarinics. A single concentration of the test compound sensitive to detect time-dependent inhibition (TDI) (approximately around the IC₂₅; actual concentrations are stated in Table 3) was preincubated for 30 min in the presence and absence of NADPH.²⁰ Remaining CYP activity was determined by subsequently adding substrates and, if applicable, NADPH to the preincubation mixtures for an additional 10 min. To account for any unspecific loss of activity over the total incubation time (e.g. due to possible enzyme degradation), controls (no inhibitor) in the presence and absence of NADPH were run in parallel.

Table 2.

IC₅₀ and calculated K_i values of the tested antimuscarinics for the seven major human CYP enzymes determined via the cocktail assay, as well as the threshold for potential in vivo DDIs in the gut wall to trigger detailed examinations. IC₅₀ values below the threshold are indicated in bold.

	DDI threshold	CYP	IC₅₀ (µM)	95% CI (µM)	Calculated K_i (µM)
Darifenacin	14 µM	1A2	>50
		2B6	–
		2C8	>50
		2C9	>50
		2C19	–
		2D6	0.37	(0.23–0.51)	0.29
		3A4	>50
Fesoterodine	6.1 µM	1A2	>1000
		2B6	–
		2C8	>1000
		2C9	–
		2C19	>1000
		2D6	110	(60–150)	86
		3A4	190	(60–320)	120
Oxybutynin	10 µM	1A2	200	(130–270)	140
		2B6	21	(11–31)	20
		2C8	36	(22–50)	23
		2C9	46	(33–59)	36
		2C19	66	(45–88)	46
		2D6	120	(67–170)	94
		3A4	180	(110–250)	120
Propiverine	45 µM	1A2	>1000
		2B6	5.8	(3.7–7.9)	5.5
		2C8	77	(25–130)	49
		2C9	59	(44–74)	46
		2C19	25	(15–35)	18
		2D6	2.4	(1.8–2.9)	1.9
		3A4	30	(13–48)	19
Solifenacin	8.3 µM	1A2	>1000
		2B6	34	(26–43)	32
		2C8	84	(17–150)	53
		2C9	52	(31–73)	41
		2C19	140	(87–200)	98
		2D6	13	(11–16)	10
		3A4	100	(0.37–200)	65
Tolterodine	3.4 µM	1A2	>1000
		2B6	98	(28–170)	92
		2C8	120	(59–180)	76
		2C9	>1000
		2C19	200	(140–260)	140
		2D6	2.9	(1.9–4.0)	2.3
		3A4	34	(20–48)	22
Trospium chloride	42 µM	1A2	–
		2B6	>1000
		2C8	–
		2C9	–
		2C19	–
		2D6	55	(37–74)	43
		3A4	70	(24–120)	45

Determination of K_i values.

CI, confidence interval; CYP, cytochrome P450; IC50, half maximal inhibitory concentration; Ki, inhibitory constant; DDI, drug–drug interaction.

Table 3.

Percentage time-dependent inhibition and associated 95% confidence intervals of darifenacin and tolterodine on CYP2D6 and propiverine on CYP2B6, CYP2C19, CYP2D6 and CYP3A4.

Test compound	CYP	Test compound concentration	% TDI	95% CI
Darifenacin	2D6	0.10 µM	5.1	–16 to 26
Propiverine	2B6	2.0 µM	–0.65	–21 to 19
	2C19	20 µM	4.5	–6.0 to 15
	2D6	2.0 µM	1.5	–8.2 to 11
	3A4	20 µM	9.2	–2.9 to 21
Tolterodine	2D6	1.0 µM	–1.3	–16 to 13

CI, confidence interval; CYP, cytochrome 450; TDI, time-dependent inhibition.

K_i values were determined to assess the mode of inhibition. Five concentrations of the inhibitor were incubated for 10 min with five different concentrations of the relevant substrate. In order to ensure optimized conditions for the K_i experiments, the inhibitor and substrate concentrations were chosen according to the prior direct inhibition screening results and K_m determinations (see section ‘Direct inhibition screening’, Appendices 1 and 2). The inhibitor concentrations generally covered a concentration range up to eightfold of the prior determined IC₅₀ values, whereas the substrate concentrations covered a range of 0.5-fold to eightfold of the expected K_m values.

LC-MS/MS conditions

The formation of the major metabolites and the internal standard, labetalol, were analysed using an API 5000 with QJET^TM Ion Guide (Applied Biosystems, Foster City, CA, USA), linked to a binary Agilent 1200 (Agilent Technologies Inc., Santa Clara, CA, USA). For further details regarding storage conditions and liquid chromatography tandem mass spectrometry (LC-MS/MS) parameters, please refer to Dahlinger et al.¹⁹ The analytical method was validated in accordance with the European Medicines Agency guideline on bioanalytical method validation.²¹

Data analysis

IC₅₀ estimation

For the direct inhibition experiments, the IC₅₀ values were calculated via nonlinear least squares regression using the R built-in PORT optimization routines. A detailed description can be found in Dahlinger et al.¹⁹

Time-dependent inhibition

Percentage time-dependent inhibition (%TDI) following preincubation was estimated according to the formula developed by Atkinson et al.²²:

% T D I = 100 \times (1 - \frac{\frac{A_{t e s t c m p d + N A D P H} - b_{+ N A D P H}}{A_{c o n t r o l + N A D P H} - b_{+ N A D P H}}}{\frac{A_{t e s t c m p d - N A D P H} - b_{- N A D P H}}{A_{c o n t r o l + N A D P H} - b_{- N A D P H}}})

where A is the triplicate-averaged (arithmetic) activity after preincubation at the previously estimated IC₂₅ or no test compound concentration and in the presence and absence of the cofactor NADPH, respectively; b is the respective baseline activity after preincubation in the presence and absence of NADPH. Ninety five percent confidence intervals (95% CIs) of %TDI were estimated using ordinary nonparametric (resampling of original response data with replacement) stratified bootstrapping.

Based on theoretical considerations and subsequent validation, a test compound was classified as a time-dependent inhibitor if the %TDI was greater than 10%. The derivation and validation of the 10% cutoff is described in Dahlinger et al.¹⁹

K_i values

To determine the type of reversible inhibition and to quantitatively describe the binding affinity of the inhibitor to the enzyme or to the enzyme–substrate complex, the dissociation constant K_i was calculated according to the Michaelis–Menten equation.²³^,²⁴ The mode of inhibition was determined based on the statistical evaluation of the Michaelis–Menten plots by Akaike’s information criterion, and a visual inspection of the Dixon, Hanes–Woolf and Lineweaver–Burk plots.

Prediction of potential in vivo DDIs

In a first step, a basic DDI model was used to assess whether further more detailed in vitro inhibition experiments were needed (i.e. time-dependent inhibition, mode of inhibition experiments). This model applies a worst case scenario which is based on a constant exposure of the investigational drug at the enzyme site. The basic model states that enzyme inhibition in the intestinal wall, that is, at the site with the highest concentrations of a potential perpetrator in vivo, cannot be excluded if

\frac{[I]}{K_{i}} ⩾ 10

where [I] is the maximum dose taken on one occasion divided by a volume of 250 ml.¹⁵^,²⁵ The applied doses were based on the relevant prescribing information and are shown in Appendix 3. This threshold is particularly relevant for orally administered drugs and particularly important in cases when the enzyme has a manifest abundance in the enterocyte (e.g. CYP3A4). This basic model was applied as the threshold for further in vitro characterization (TDI and K_i determination) due to the lack of clinical data.

In a second step, obtained K_i values and collected pharmacokinetic data (see Appendices 4 and 5) were used to estimate the clinical net effect of the observed reversible in vitro inhibition by applying the model of Fahmi et al.¹⁵^,²⁵^,²⁶

\begin{array}{l} A U C r a t i o = \\ (\frac{1}{[A_{h}] * f_{m} + (1 - f_{m})}) * (\frac{1}{[A_{g}] * (1 - F_{g}) + F_{g}}) \end{array}

Here, A describes the reversible inhibition occurring in the liver (subscript h) and gut wall (subscript g), f_m represents the fraction of the victim drug which is metabolized by the inhibited CYP, and F_g describes the fraction of the absorbed drug escaping gut wall extraction. A more detailed description can be found in the original manuscript by Fahmi et al.²⁶

Results

Direct inhibition screening

In 6 out of the 49 possible substrate and antimuscarinic agent combinations, the IC₅₀ values observed led to K_i values lower than the threshold for potential in vivo DDIs. Notably, CYP2D6 was inhibited by darifenacin in the higher nanomolar range and by propiverine in the single-digit micromolar range [CYP2D6: IC₅₀ (darifenacin) was 0.37 µM and IC₅₀ (propiverine) was 2.4 µM]. A summary of all direct inhibition results is presented in Table 2.

Time-dependent inhibition screening

Time-dependent inhibition of CYP2D6 by darifenacin and tolterodine as well as inhibition of CYP2B6, CYP2C19, CYP2D6 and CYP3A4 by propiverine were studied according to the results from the direct screening experiments and potential risks for in vivo DDIs of these compounds. Detailed results are presented in Table 2. By definition, for a time-dependent inhibitor, inhibition increases with longer incubation time. By applying the predefined criteria (point estimate of %TDI > 10%), no TDI was observed in any of the seven cases. The results from the TDI experiments are summarized in Table 3.

Detailed examination of the mode of inhibition

The same substrate and antimuscarinic combinations as for determining the TDI were also assessed for mode of inhibition. The most potent inhibition was observed for darifenacin. Darifenacin competitively inhibited CYP2D6 with a K_i of 0.72 µM. A summary of the experiments for examining the mode of inhibition can be found in Table 4. The Michaelis–Menten plot for inhibition of CYP2D6 by propiverine is depicted in Figure 1.

Table 4.

Summary of the K_i values and most likely mode of inhibition of the tested spasmolytics.

Test compound	CYP	Inhibition type	K_i ± SD (µM)
Darifenacin	2D6	Competitive	0.72 ± 0.27
Propiverine	2B6	Competitive	14 ± 4.0
	2C19	Mixed type	51 ± 32
	2D6	Competitive	7.2 ± 1.6
	3A4	Mixed type	15 ± 9.9
Tolterodine	2D6	Competitive	3.7 ± 0.98

CYP, cytochrome P450; Ki, inhibitory constant; SD, standard deviation.

Figure 1.

Michaelis–Menten plot for the inhibition of cytochrome P450 (CYP)-2D6 by propiverine. The varying concentrations of propiverine are stated on the right-hand side of each individual plot. Shown are the arithmetic means and associated standard deviations of three replicates.

In vivo:in vitro extrapolation

Based on the screening results and the detailed examination of the mode of inhibition presented above, reversible inhibition of hepatic CYP2B6 and CYP2D6 and hepatic and intestinal inhibition of CYP3A4 were identified as the key contributors for potential in vivo DDIs of darifenacin and propiverine.

Prediction of the inhibition effect of darifenacin on CYP2D6. Applying the obtained K_i value of 0.72 µM for the inhibition of CYP2D6 by darifenacin yielded predicted fold increases in the area under the curve (AUC) of debrisoquine of 1.6 and 1.9 for the typical dosing of darifenacin at 7.5 mg and 15 mg, respectively. Debrisoquine is a probe drug almost exclusively metabolized by CYP2D6 [f_m(debrisoquine) = 0.98]. For dextromethorphan, another commonly used CYP2D6 substrate [(f_m(dextromethorphan) = 0.8], the predicted fold increase in AUC was 1.4 (7.5 mg darifenacin) and 1.6 (15 mg darifenacin). All other input parameters for the substrates and darifenacin can be found in Appendices 4 and 6.

Prediction of the inhibition effect of propiverine on CYP2B6, CYP2D6 and CYP3A4. Screening experiments identified propiverine to potentially cause clinically relevant in vivo interactions due to reversible CYP inhibition when using substrates specific for CYP2B6, CYP2C19, CYP2D6 and CYP3A4 (see above). The detailed mode of inhibition experiments confirmed K_i values below the calculated threshold and, thus, potential in vivo interaction risks for CYP2B6, CYP2D6 and CYP3A4.

For bupropion, a clinically important CYP2B6 substrate, only a minor increase in the expected AUC was predicted for all common dosage regimens (15–45 mg) and the two formulations used [extended release (ER) and immediate release (IR)].

In contrast, metabolism of debrisoquine was predicted to be potently inhibited by propiverine. For the IR formulations of propiverine, a dose-dependent increase (15–45 mg propiverine) in debrisoquine AUC of 1.17- to 1.59-fold was predicted. When assuming that the major metabolite propiverine-N-oxide is as potent as the parent drug, which to date has not yet been evaluated, the effect is even more pronounced and resulted in a debrisoquine AUC increase of 1.82-fold for the 45 mg propiverine IR formulation.

The predicted increase in AUC for the inhibition of the CYP3A4 substrate midazolam was driven by the inhibition of the intestinal 3A4-mediated metabolism of midazolam by propiverine. A dose of 45 mg IR propiverine would result in a predicted doubling of the AUC for midazolam (AUCR_hepatic = 1.25; AUC_intestinal = 1.61).

A conclusive list of predictions and available information on in vivo DDI trials can be found in Table 5.

Table 5.

Summary of the predicted increase in AUC due to the inhibition of major CYPs by darifenacin and propiverine and the major metabolite, as well as observed AUC ratios from selected clinical trials.

Antimuscarinic drug	Dose	Victim drug	Predicted AUC ratio(parent alone)	Predicted AUC ratio(parent and metabolite)	Observed AUC ratio
Darifenacin	7.5 mg	Debrisoquine	1.6^$	n/a	2.2⁸ 3.2⁸
	15 mg	(CYP2D6 substrate)	1.9^$
Darifenacin	7.5 mg	Dextromethorphan	1.4^$	n/a	n/a
	15 mg	(CYP2D6 substrate)	1.6^$
Propiverine	15 mg ER	Bupropion(CYP2B6 substrate)	1.0	1.0	n/a
	30 mg ER		1.0	1.0
	45 mg ER		1.0	1.1
	15 mg IR		1.1	1.1
	30 mg IR		1.1	1.2
	45 mg IR		1.2	1.2*
Propiverine	15 mg ER	Debrisoquine(CYP2D6 substrate)	1.0	1.1	n/a
	30 mg ER		1.1	1.1
	45 mg ER		1.1	1.2
	15 mg IR		1.2	1.2^$
	30 mg IR		1.4^$	1.6^$
	45 mg IR		1.6*^,$	1.8*^,$
Propiverine	15 mg ER	Dextromethorphan(CYP2D6 substrate)	1.0	1.0	n/a
	30 mg ER		1.1	1.1
	45 mg ER		1.1	1.1
	15 mg IR		1.1	1.2
	30 mg IR		1.3^$	1.4^$
	45 mg IR		1.4*^,$	1.6*^,$
Propiverine	15 mg ER	Midazolam(CYP3A4 substrate)	1.2	1.3^$	1.5 (15mg IR)²⁷
	30 mg ER		1.4^$	1.4^$
	45 mg ER		1.5^$	1.5^$
	15 mg IR		1.6^$	1.6^$
	30 mg IR		1.9^$	1.9^$
	45 mg IR		2.0*^,$	2.1*^,$

AUC, area under the curve; CYP, cytochrome P450; ER, extended release table; IR, immediate release tablet; n/a, not applicable.

Based on a linear relationship between dose and response. No clinical data available.

Predicted AUC ratio is above 1.25, the established threshold which necessitates an in vivo DDI study. The input parameters of propiverine, propiverine N oxide and the CYP3A4 substrate midazolam are found in Appendices 4 –6.

Discussion

Antimuscarinics represent the first-line pharmacotherapy in the treatment of OAB symptoms. With prescriptions of antimuscarinics increasing globally, the clinical need for systematic information on DDIs caused by antimuscarinic drugs becomes evident.⁴^,⁹

Among the seven tested antimuscarinics, darifenacin and propiverine inhibited major human CYPs in vitro at concentrations obtained during clinical use. Fesoterodine, oxybutynin solifenacin, tolterodine and trospium chloride were shown to not inhibit major human CYPs at clinically relevant concentrations.

The prescribing information for fesoterodine and solifenacin states that both drugs have a minimal potential for DDIs. Indeed, the direct inhibition screening for fesoterodine and solifenacin showed only a slight inhibition of the major human CYPs at concentrations above those obtained during clinical use.¹⁰^,¹³^,²⁸^,²⁹

Lukkari et al. observed more than a 50% reduction in CYP3A4 activity (IC₅₀ = 7.7 µM) at oxybutynin concentrations that might theoretically occur in the gut wall (threshold: 20 µM).³⁰ However, the direct inhibition screening results did not confirm a clinically relevant inhibition of CYP3A4 or any of the other major human CYP enzymes. Possible explanations for this discrepancy might be differences in experimental conditions [substrate: testosterone versus midazolam; enzyme source: in-house prepared HLM (three donors) versus commercially purchased pooled HLMs (20 donors)].

CYP enzymes were shown to only play a minor role in the metabolism of trospium chloride.³¹ Yet, Beckman-Knopp et al. showed that trospium chloride inhibited CYP2D6 in the lower double-digit micromolar range (IC₅₀ = 27–44 µM).³² This is consistent with the direct inhibition screening which gave an IC₅₀ value of 55 µM for CYP2D6. It should be noted that these values are approximately 1000-fold higher than the therapeutically obtained trospium chloride peak plasma concentrations of around 50 nm.³²^,³³

The in vitro screening experiments for tolterodine detected inhibition of CYP2D6 (IC₅₀ = 2.9 µM, threshold_intestinal 3.4 µM) below the highest theoretical tolterodine concentration in the intestinal wall. In the follow-up in vitro studies, tolterodine showed no time-dependent inhibition and competitively inhibited CYP2D6 with a K_i of 3.7 µM. As CYP2D6 is only marginally expressed in the gut wall³⁴ and the highest reported systemic concentration of tolterodine is more than 30-fold lower than the obtained K_i value,⁹ clinically relevant in vivo DDIs due to inhibition of major human CYPs by tolterodine can be excluded.

In our in vitro studies, darifenacin potently inhibited the CYP2D6-mediated dextromethorphan O demethylation (IC₅₀ = 0.37 µM). This finding was confirmed in the subsequent, more detailed mode of inhibition experiment (K_i = 0.72 µM, competitive inhibition, no TDI). In a separate study, a slightly higher IC₅₀ value of 1.7 µM was determined for the inhibition of CYP2D6 by darifenacin.⁸ For all other tested CYP-selective model reactions, less than a 50% decrease in the activity was observed at 50 µM, that is, the highest darifenacin concentration used in the incubation mixture. Opposed to this finding, Skerjanec et al. reported IC₅₀ values between 5.3 and 43 µM for the inhibition of CYP3A4 by darifenacin. Interestingly, in a clinical study it was shown that repetitive supratherapeutic levels of darifenacin only mildly increased the AUC of the selective CYP3A4 substrate midazolam (AUC ratio of 1.17) and had no effect on midazolam’s c_max (0.99)_.⁸ Afterwards, a mechanistic static model was applied in order to quantitatively estimate the in vivo DDI effect. The predictions of the increase in AUC ratio were based on the obtained in vitro reversible dissociation constants and published pharmacokinetic data. In the case of the inhibition of CYP2D6 by darifenacin, the K_i value of 0.72 µM obtained from our study translated to an increase in AUC of debrisoquine, a sensitive CYP2D6 substrate, of 1.6- and 1.9-fold when considering the common darifenacin doses of 7.5 mg and 15 mg, respectively. For comparison, in a clinical DDI trial, the effect of a 7.5 mg and 15 mg darifenacin tablet on the AUC of debrisoquine was evaluated. During this interaction trial AUC increases of 2.2- and 3.25-fold, respectively, were observed. A potential explanation for the underestimation of the inhibitory effect of darifenacin in this study would be inhibition of CYP2D6 by darifenacin metabolites. Numerous darifenacin metabolites have been identified. However, to our knowledge, there is no information available regarding their potential inhibitory properties. In addition, only pharmacokinetic data for hydroxydarifenacin, one of the various darifenacin metabolites, is available, which makes even hypothetical calculations irrelevant.⁸^,³⁵

The screening of propiverine with the seven-in-one CYP inhibition assay provided IC₅₀ values which relate to K_i values below the calculated threshold for potential in vivo DDIs for CYP2B6, CYP2C19, CYP2D6 and CYP3A4 in the direct screening experiments. Follow-up in vitro studies showed no time-dependent inhibition of propiverine on any of the tested CYPs. The reversible mode of inhibition experiments yielded similar K_i values to those calculated from the previous direct inhibition screening experiments (maximum threefold difference) and confirmed the potentially clinically relevant interaction of propiverine with CYP2B6, CYP2D6 and CYP3A4, but not CYP2C19. Applying a mechanistic static model, no clinically relevant increase in AUC of bupropion, a sensitive CYP2B6 substrate, was observed at any of the commonly used propiverine dosing regimens.²⁶ For sensitive CYP2D6 substrates, such as debrisoquine 30 mg and 45 mg IR, propiverine was estimated to increase the exposure of debrisoquine by 1.4- and 1.6-fold, respectively. In particular, when assuming that the major metabolite propiverine N oxide, which is circulating at manifold higher concentrations than the parent drug, inhibits CYP2D6 activity to a similar degree, an AUC ratio of 1.6 and 1.8 for 30 mg IR and 45 mg IR, respectively, is calculated.³⁶ Interestingly, for all commonly administered ER propiverine doses (15–45 mg), only minor effects on CYP2D6 were estimated (AUC ratio ⩽ 1.25). This is due to the fact that the c_max for propiverine and propiverine N oxide is approximately twofold and threefold higher for the IR tablet than the ER tablet. Administration of 15 mg and 30 mg IR propiverine tablets was estimated to increase the AUC of the CYP3A4 substrate midazolam by 1.6- and 1.9-fold. In comparison, respective concentrations of propiverine ER tablets were predicted to increase the midazolam exposure by only 1.2- and 1.4-fold, respectively. In 2005, Tomalik-Scharte et al. reported on an in vivo cocktail DDI study under continuous propiverine treatment (15 mg IR propiverine (test), placebo (reference) twice daily) a reduced hepatic and intestinal CYP3A4 activity. In accordance with our predictions, a moderate increase of 1.5-fold in the AUC of oral midazolam, when compared to placebo treatment, was observed (AUC Ratio_predicted of 1.6). Furthermore, Tomalik-Scharte et al. showed that the increase in AUC ratio was driven by inhibition of intestinal CYP3A4. In accordance with the screening results, the clinical DDI cocktail study revealed that propiverine treatment had no relevant effect on CYP1A2, CYP2C9 and CYP2C19. For CYP2B6, no model substrate was included in the in vivo substrate cocktail and to date the results of propiverine on CYP2D6 have not been reported.²⁷

Based on our in vitro results, clinical interactions of darifenacin with CYP2D6 substrates and propiverine with CYP2D6 and CYP3A4 substrates need to be considered. Major drug classes containing a number of drugs which are cleared by CYP2D6 are antidepressants, neuroleptics, β blockers and antiarrhythmics. CYP3A4 accounts for at least one third of all CYP-mediated drug metabolism, and many drugs belonging to classes of benzodiazepines, calcium channel blockers, immunosuppressives, statins, macrolides and steroids (i.e. oral contraceptives) are metabolized by CYP3A4. Alertness is probably indicated when darifenacin or propiverine are simultaneously administered with tricyclic antidepressants, such as imipramine. Imipramine is a nonselective CYP2D6 substrate and therefore prone to pharmacokinetic interactions with propiverine and darifenacin. Additionally, imipramine shows antimuscarinic activity which potentially results in pharmacodynamic interactions with these antimuscarinics.⁸^,³⁷^,¹⁸

Conclusion

In conclusion, the screening of the seven antimuscarinic drugs essentially confirmed the respective declarations on the DDI potential as described in the prescribing information by the manufacturers. In vivo predictions based on pharmacokinetic data and the conducted in vitro experiments yielded similar effects of darifenacin on CYP2D6 and propiverine on CYP3A4 as obtained in separately conducted in vivo DDI studies. As a novel finding, propiverine was identified to potentially inhibit CYP2D6 at clinically occurring concentrations, depending on the pharmaceutical preparation. None of the tested antimuscarinics was found to be a time-dependent CYP inhibitor. Our findings illustrate the importance of a clinical DDI study to elucidate and quantify the effect of propiverine on CYP2D6. These clinical results would then also allow for modelling the effects of different propiverine dosing regimens (once daily versus twice daily, IR versus ER) on various CYP2D6 substrates.

Footnotes

Appendix

Appendix 6.

Input parameters of substrates for CYP2B6, CYP2D6 and CYP3A4 for the prediction of in vivo CYP inhibition.

Victim drug	Category	f_m	F_g
Bupropion	CYP2B6 substrate	0.60³⁸	1³⁴^,³⁹
Debrisoquine	CYP2D6 substrate	0.98⁴⁰	1³⁴^,³⁹
Dextromethorphan	CYP2D6 substrate	0.80²⁰	1³⁴^,³⁹
Midazolam	CYP3A4 substrate	0.94⁴¹	0.6⁴¹

CYP, cytochrome P450; f_m, fraction of systemic clearance of substrate mediated by certain CYP enzyme; F_g, fraction available after intestinal metabolism.

Acknowledgements

We thank Cristina Grigore for encouragement, helpful discussions and proofreading during the preparation of the manuscript.

Funding

This work was supported by the Koeln Fortune Program/Faculty of Medicine, University of Cologne [grant number 161/2014].

Conflict of interest statement

The authors declare that there is no conflict of interest.

References

Milsom

Abrams

Cardozo

. How widespread are the symptoms of an overactive bladder and how are they managed? A population-based prevalence study. BJU Int 2001; 87: 760–766.

Stewart

Van Rooyen

Cundiff

. Prevalence and burden of overactive bladder in the United States. World J Urol 2003; 2: 327–336.

Abrams

Andersson

K-E

Buccafusco

. Muscarinic receptors: their distribution and function in body systems, and the implications for treating overactive bladder. Br J Pharmacol 2006; 148: 565–578.

Caulfield

Birdsall

. International Union of Pharmacology. XVII. Classification of muscarinic acetylcholine receptors. Pharmacol Rev 1998; 50: 279–290.

Doroshyenko

Fuhr

. Clinical pharmacokinetics and pharmacodynamics of solifenacin. Clin Pharmacokinet 2009; 48: 281–302.

Eglen

Hegde

Watson

. Muscarinic receptor subtypes and smooth muscle function. Pharmacol Rev 1996; 48: 531–565.

Hashim

Abrams

. Drug treatment of overactive bladder: efficacy, cost and quality-of-life considerations. Drugs 2004; 64: 1643–1656.

Skerjanec

. The clinical pharmacokinetics of darifenacin. Clinical Pharmacokinet 2006; 45: 325–350.

Guay

DRP

. Clinical pharmacokinetics of drugs used to treat urge incontinence. Clin Pharmacokinet 2003; 42: 1243–1285.

10.

Witte

LPW

Mulder

WMC

de la Rosette

JJMCH

. Muscarinic receptor antagonists for overactive bladder treatment: does one fit all? Curr Opin Urol 2009; 19: 13–19.

11.

Zanger

Raimundo

Eichelbaum

. Cytochrome P450 2D6: overview and update on pharmacology, genetics, biochemistry. Naunyn Schmiedebergs Arch Pharmacol 2004; 369: 23–37.

12.

Chapple

Khullar

Gabriel

. The effects of antimuscarinic treatments in overactive bladder: an update of a systematic review and meta-analysis. Eur Urol 2008; 54: 543–562.

13.

Ohtake

Saitoh

Yuyama

. Pharmacological characterization of a new antimuscarinic agent, solifenacin succinate, in comparison with other antimuscarinic agents. Biol Pharm Bull 2007; 30: 54–58.

14.

Crewe

Lennard

Tucker

. The effect of selective serotonin re-uptake inhibitors on cytochrome P4502D6 (CYP2D6) activity in human liver microsomes. 1992/09/01 ed. Br J Clin Pharmacol 1992; 34: 262–265.

15.

European Medicines Agency. Guideline on the investigation of drug interactions, http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2012/07/WC500129606.pdf (2012, accessed 4 May 2017).

16.

Kobayashi

Chiba

Yagi

. Identification of cytochrome P450 isoforms involved in citalopram N-demethylation by human liver microsomes. J Pharmacol Exp Ther 1997; 280: 927–933.

17.

Yaïch

Popon

Médard

. In-vitro cytochrome P450 dependent metabolism of oxybutynin to N-deethyloxybutynin in humans. Pharmacogenetics 1998; 8(5): 449–451.

18.

Zanger

Schwab

. Cytochrome P450 enzymes in drug metabolism: regulation of gene expression, enzyme activities, and impact of genetic variation. Pharmacol Ther 2013; 138: 103–141.

19.

Dahlinger

Düchting

Sydow

. Development and validation of an in vitro, seven-in-one human cytochrome P450 assay for evaluation of both direct and time-dependent inhibition. J Pharmacol Toxicol Methods 2015; 77: 66–75.

20.

Obach

Walsky

Venkatakrishnan

. The utility of in vitro cytochrome P450 inhibition data in the prediction of drug-drug interactions. J Pharmacol Exp Ther 2006; 316: 336–348.

21.

European Medicines Agency. Guideline on bioanalytical method validation (EMEA/CHMP/EWP/192217/2009), http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2011/08/WC500109686.pdf (2011).

22.

Atkinson

Kenny

Grime

. Automated assessment of time-dependent inhibition of human cytochrome P450 enzymes using liquid chromatography-tandem mass spectrometry analysis. Drug Metab Dispos 2005; 33: 1637–1647.

23.

Copeland

. Enzymes: a practical introduction to structure, mechanism, and data analysis. New York: John Wiley & Sons, 2000.

24.

Copeland

. Evaluation of enzyme inhibitors in drug discovery: a guide for medicinal chemists and pharmacologists. 2nd ed. Hoboken, NJ: John Wiley & Sons, 2013.

25.

Food and Drug Administration. Drug interaction studies – study design, data analysis, implication for dosing and labeling recommendations, http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/ucm292362.pdf (2012, accessed 4 May 2017).

26.

Fahmi

Hurst

Plowchalk

. Comparison of different algorithms for predicting clinical drug-drug interactions, based on the use of CYP3A4 in vitro data: predictions of compounds as precipitants of interaction. Drug Metab Dispos 2009; 37: 1658–1666.

27.

Tomalik-Scharte

Jetter

Kinzig-Schippers

. Effect of propiverine on cytochrome P450 enzymes: a cocktail interaction study in healthy volunteers. Drug Metab Dispos 2005; 33: 1859–1866.

28.

Malhotra

Sachse

Wood

. Evaluation of drug-drug interactions with fesoterodine. Eur J Clin Pharmacol 2009; 65: 551–560.

29.

Staskin

. Overactive bladder in the elderly: a guide to pharmacological management. Drugs & Aging 2005; 22: 1013–1028.

30.

Lukkari

Taavitsainen

Juhakoskr

. Cytochrome P450 specificity of metabolism and interactions of oxybutynin in human liver microsomes. Pharmacol Toxicol 1998; 82: 161–166.

31.

Sanctura

(Trospium Chloride): Drug Approval Package. Indevus Pharmaceuticals, Inc., https://www.accessdata.fda.gov/drugsatfda_docs/nda/2007/022103s000toc.cfm (2007, accessed 4 May 2017).

32.

Beckmann-Knopp

Rietbrock

Weyhenmeyer

. Inhibitory effects of trospium chloride on cytochrome P450 enzymes in human liver microsomes. Pharmacol Toxicol 1999; 85: 299–304.

33.

Doroshyenko

Jetter

Odenthal

. Clinical pharmacokinetics of trospium chloride. Clin Pharmacokinet 2005; 44: 701–720.

34.

Paine

Hart

Ludington

. The human intestinal cytochrome P450 ‘pie.’ Drug Metab Dispos 2006; 34: 880–886.

35.

Novartis. Enablex: Clinical Pharmacology and Biopharmaceutics Review, https://www.accessdata.fda.gov/drugsatfda_docs/nda/2004/21-513_Enablex_biopharmr.pdf (2003, accessed 4 May 2017).

36.

May

Giessmann

Wegner

. Oral absorption of propiverine solution and of the immediate and extended release dosage forms: influence of regioselective intestinal elimination. Eur J Clin Pharmacol 2008; 64: 1085–1092.

37.

Dahl

Iselius

Alm

. Polymorphic 2-hydroxylation of desipramine. A population and family study. Eur J Clin Pharmacol 1993; 44: 445–450.

38.

Barter

Tucker

Rowland-Yeo

. Differences in cytochrome p450-mediated pharmacokinetics between chinese and caucasian populations predicted by mechanistic physiologically based pharmacokinetic modelling. Clin Pharmacokinet 2013; 52: 1085–1100.

39.

Thelen

Dressman

. Cytochrome P450-mediated metabolism in the human gut wall. Pharm Pharmacol 2009; 61: 541–558.

40.

Frank

Jaehde

Fuhr

. Evaluation of probe drugs and pharmacokinetic metrics for CYP2D6 phenotyping. Eur J Clin Pharmacol 2007; 63: 321–333.

41.

Galetin

Burt

Gibbons

. Prediction of time-dependent CYP3A4 drug-drug interactions: impact of enzyme degradation, parallel elimination pathways, and intestinal inhibition. Drug Metab Dispos 2006; 34: 166–175.

Assessment of inhibitory effects on major human cytochrome P450 enzymes by spasmolytics used in the treatment of overactive bladder syndrome

Abstract

Background:

Methods:

Results:

Conclusions:

Keywords

Background

Materials and methods

Chemicals

Microsomal incubation conditions

General incubation conditions and sample preparation

Direct inhibition assay and time-dependent inhibition assay

LC-MS/MS conditions

Data analysis

IC50 estimation

Time-dependent inhibition

Ki values

Prediction of potential in vivo DDIs

Results

Direct inhibition screening

Time-dependent inhibition screening

Detailed examination of the mode of inhibition

In vivo:in vitro extrapolation

Discussion

Conclusion

Footnotes

Appendix

Acknowledgements

Funding

Conflict of interest statement

References

IC₅₀ estimation

K_i values