In vitro inhibition and induction of human liver ......Dec 31, 2008 · water for CYP inhibition studies and in DMSO for enzyme induction studies. The following reagents were purchased

DMD 28274

1

In vitro inhibition and induction of human liver cytochrome P450 (CYP)

enzymes by milnacipran

Brandy L. Paris, Brian W. Ogilvie, Julie A. Scheinkoenig, Florence Ndikum-Moffor, Remi Gibson and

Andrew Parkinson

XenoTech, LLC, Lenexa, KS 66219 USA (B.L.P., B.W.O., J.A.S., F.M., R.G., A.P.)

DMD Fast Forward. Published on July 16, 2009 as doi:10.1124/dmd.109.028274

Copyright 2009 by the American Society for Pharmacology and Experimental Therapeutics.

This article has not been copyedited and formatted. The final version may differ from this version.DMD Fast Forward. Published on July 16, 2009 as DOI: 10.1124/dmd.109.028274

at ASPE

T Journals on M

arch 16, 2021dm

d.aspetjournals.orgD

ownloaded from

http://dmd.aspetjournals.org/

DMD 28274

2

Running title: CYP Inhibition and Induction by Milnacipran

Address correspondence to:

Andrew Parkinson

XenoTech LLC

16825 W. 116Th Street

Lenexa, Kansas 66219, USA

Tel: (913) 438-7450

Fax: (913) 227-7199

E-mail: [email protected]

Document Summary:

Number of Text Pages 19

Number of Tables 5

Number of Figures 5

Number of References 32

Number of Words in the Abstract 245

Number of Words in the Introduction 606

Number of Words in the Discussion 2170

Abbreviations used are: amu, atomic mass units; ANOVA, analysis of variance; CYP, cytochrome P450;

DMEM, Dulbecco’s Modified Eagle’s Medium; DMSO, dimethyl sulfoxide; EMs, extensive

metabolizers; ESI, electrospray ionization; FBS, fetal bovine serum; FDA, Food & Drug Administration;

IC50, inhibitor concentration that causes 50% inhibition; INR, international normalized ratio; ITS+,

insulin, human transferring, and selenous acid; LC/MS-MS, liquid chromatography/tandem mass

spectrometry; MCM, Modified Chee’s medium; MEM, minimum essential medium; PK,

pharmacokinetic(s); PMs, poor metabolizers; SNRI, selective serotonin-norepinephrine reuptake

inhibitor; SSRI, selective serotonin-reuptake inhibitor; TCA, tricyclic antidepressant; UGT, UDP-

glucuronosyltransferase.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

3

ABSTRACT:

Milnacipran (Savella™) inhibits both norepinephrine and serotonin reuptake and is distinguished

by a nearly three fold greater potency in inhibiting norepinephrine reuptake in vitro compared to

serotonin. We evaluated the ability of milnacipran to inhibit and induce human CYP enzymes in

vitro. In human liver microsomes, milnacipran did not inhibit CYP1A2, 2B6, 2C8, 2C9, 2C19 or

2D6 (IC50 ≥ 100 µM); whereas, a comparator with dual-reuptake properties (duloxetine

[Cymbalta®]) inhibited CYP2D6 (IC50 = 7 µM) and CYP2B6 (IC50 = 15 µM) with a relatively

high potency. Milnacipran inhibited CYP3A4/5 in a substrate-dependent manner (i.e.,

midazolam 1′-hydroxylation IC50 ≈30 µM; testosterone 6β-hydroxylation IC50 ≈100 µM);

whereas, duloxetine inhibited both CYP3A4/5 activities with equal potency (IC50 = 37 and

38 µM, respectively). Milnacipran produced no time-dependent inhibition (<10%) of CYP

activity, whereas duloxetine produced time-dependent inhibition of CYP1A2, 2B6, 2C19 and

3A4/5. To evaluate CYP induction, freshly isolated human hepatocytes (n = 3) were cultured

and treated once daily for three days with milnacipran (3, 10 and 30 µM), after which

microsomal CYP activities were measured. While positive controls (omeprazole, phenobarbital

and rifampin) caused anticipated CYP induction, milnacipran had minimal effect on CYP1A2,

2C8, 2C9 or 2C19 activity. The highest concentration of milnacipran (30 µM; >10x plasma

Cmax) produced a 2.6- and 2.2-fold increase in CYP2B6 and CYP3A4/5 activity (making it 26%

and 34% as effective as phenobarbital and rifampin, respectively). Given these results,

milnacipran is not expected to cause clinically significant CYP inhibition or induction.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

4

INTRODUCTION

Milnacipran (Savella™), which was recently approved for the treatment of fibromyalgia, is a

dual reuptake inhibitor of norepinephrine and serotonin, which is distinguished by an

approximately three fold greater potency in inhibiting norepinephrine reuptake in vitro compared

with serotonin reuptake (Vaishnavi et al., 2004). These two neurotransmitters have been shown

to exert significant modulatory effects on peripheral and central pain processing (Dubner and

Hargreaves, 1989). Selective serotonin-norepinephrine reuptake inhibitors (SNRIs) like

duloxetine and venlafaxine are more potent inhibitors of serotonin reuptake than norepinephrine

reuptake, whereas the converse is true of milnacipran (Vaishnavi et al., 2004).

Milnacipran is well absorbed (85-90%) after oral administration and has linear pharmacokinetics

(PK) over the therapeutic dose range (Delini-Stula, 2000). The terminal elimination half-life in

plasma is 6-8 h, and steady state levels can be predicted from single dose PK data indicating the

absence of auto-inhibition or auto-induction. Milnacipran is eliminated primarily by renal

excretion of the unchanged drug (50 - 60%), conjugation to form a carbamoyl glucuronide

(~20%) and N-dealkylation by cytochrome P450 (mainly CYP3A4) to N-desethyl milnacipran

(~8%) (Delini-Stula, 2000; Puozzo et al., 1996 and 2005; Tsuruta et al., 2000; Forest Research

Institute, personal communication). Metabolism by cytochrome P450 plays only a minor role in

the elimination of milnacipran (Caccia, 1998; Grzesiak et al., 2000; Sawada and Ohtani, 2001;

Puozzo et al., 2002). Consequently, genetic polymorphisms in CYP2D6 and inhibition of this

enzyme do not impact the pharmacokinetics of milnacipran (Puozzo et al., 2005), in contrast to

the situation with many SSRIs and TCAs (Bertilsson et al., 2002; Preskorn et al., 2007). The PK

profile of milnacipran is the same in both CYP2D6 poor metabolizers (PMs) and extensive

metabolizers (EMs), and the same holds for CYP2C19 PMs and EMs (Puozzo et al., 2005).


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

5

Limited to no modification of the pharmacokinetic profile is expected when milnacipran is co-

administered with fluoxetine (a strong CYP2D6 inhibitor) or carbamazepine (an inducer of

CYP2B6, CYP3A4 and several other enzymes) (Puozzo et al., 2002, 2005, 2006).

From the perspective of drug-drug interactions, drugs can be viewed as victims (objects) or

perpetrators (precipitants) (Ogilvie et al., 2008). Milnacipran has low victim potential because

its clearance is not heavily dependent on metabolism by a single drug-metabolizing enzyme;

hence, its PK profile is not significantly impacted by the genetic polymorphisms, CYP inhibitors

or CYP inducers that impact the disposition of other antidepressant drugs. More than half the

drug (50-60%) is eliminated unchanged in urine, which indicates kidney function is the primary

determinant of milnacipran’s elimination. The in vitro studies described in this report were

designed to evaluate the perpetrator potential of milnacipran. The enzyme-inducing potential of

milnacipran was evaluated in three preparations of freshly cultured human hepatocytes, and

focused on the major inducible human CYP enzymes, namely CYP1A2, 2B6, 2C8, 2C9, 2C19

and 3A4/5. CYP2D6 was not examined because this enzyme is recognized by the FDA as being

non-inducible (US FDA, 2006). The ability of milnacipran to function as a direct-acting and

metabolism-dependent inhibitor of CYP enzymes was evaluated with human liver microsomes.

The enzymes evaluated included CYP1A2, 2B6, 2C8, 2C9, 2C19, 2D6 and 3A4/5 (with two

substrates), as recommended by the FDA (US FDA, 2006; Huang et al., 2008). In the CYP

inhibition study, the SNRI duloxetine (Cymbalta®) was included as a comparator. The

structures of milnacipran and duloxetine are shown in Fig. 1. The in vitro studies described

herein were conducted in accordance with the FDA’s draft guidance document on the conduct of

in vitro metabolism studies (US FDA, 2006) and the principles promulgated by Tucker et al.

(2001), Bjornsson et al. (2003) and Huang et al. (2008).


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

6

MATERIALS AND METHODS

Chemicals and reagents

Milnacipran and duloxetine were provided by Forest Research Institute (Jersey City, NJ). Stock

solutions of milnacipran (10 and 50 mM) and duloxetine (10 mM) were prepared in high purity

water for CYP inhibition studies and in DMSO for enzyme induction studies.

The following reagents were purchased from Sigma-Aldrich (St. Louis, MO): bupropion HCl,

dextromethorphan, diclofenac, 4´-hydroxydiclofenac, (±)-4´-hydroxymephenytoin, 6β-hydroxy-

testosterone, midazolam, phenacetin and testosterone. Acetaminophen, N-desethylamodiaquine,

dextrorphan and 1´-hydroxymidazolam were purchased from Cerilliant (Round Rock, TX).

Amodiaquine was purchased from US Pharmacopeia (Rockville, MD). S-Mephenytoin was

purchased from Toronto Research Chemicals Inc. (North York, Ontario, Canada).

Hydroxybupropion, ITS+ and Matrigel were purchased from BD Biosciences (Bedford, MA).

Dulbecco’s Modified Eagle’s Medium (DMEM), GlutaMAX-1, insulin, MEM-non-essential

amino acids, modified Eagle’s Medium Dr. Chee's modification (MCM) and liquid penicillin-

streptomycin were purchased from Invitrogen (Grand Island, NY). PureCol was purchased from

Inamed BioMaterials (Fremont, CA). Fetal bovine serum (FBS) was purchased from SAFC

Biosciences (Lenexa, KS). Loctite 4013 was purchased from the Loctite Corporation (Rocky

Hill, CT). BCA Protein Assay Kit was purchased from Pierce Chemical Co. (Rockford, IL). All

other reagents were obtained from commercial sources, most of which have been described

elsewhere (Robertson et al., 2000; Madan et al., 2003; Ogilvie et al., 2006).

Test system.

Pooled human liver microsomes (n=16, mixed gender) were prepared and characterized at

XenoTech, LLC (Lenexa, KS). Human hepatocytes from non-transplantable livers were


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

7

prepared at XenoTech from three individual donors, all male Caucasians (ages 51, 74 and 77);

initial cell viability was 83, 93 and 77%, respectively.

In vitro CYP inhibition

The ability of milnacipran and duloxetine to inhibit the major drug-metabolizing CYP enzymes

in a direct and time-dependent manner was investigated with a pool of human liver microsomes

(pool of 16 individuals), as described by Ogilvie et al., (2006, 2008). Briefly, duplicate

incubations were conducted at 37±1°C in 400-μL incubation mixtures containing potassium

phosphate buffer (50 mM, pH 7.4), MgCl2 (3 mM), EDTA (1 mM, pH 7.4), an

NADPH-generating system (consisting of 1 mM NADP, 5 mM glucose-6-phosphate and 1

Unit/ml glucose-6-phosphate dehydrogenase) and CYP marker substrate as indicated in Table 1.

Reactions were initiated by the addition of the NADPH-generating system and terminated after

5 min by an equal volume of acetonitrile (v/v) containing an appropriate internal standard, as

summarized in Table 1. Precipitated protein was removed by centrifugation (920 × g for 10 min

at 10°C). Calibration and quality control (QC) metabolite standards were prepared in zero-time

incubations. The analytical procedures are summarized in Table 1.

To evaluate milnacipran and duloxetine as direct-acting inhibitors, pooled human liver

microsomes (≤ 0.1 mg/ml) were incubated with CYP marker substrates (at concentrations

approximately equal to Km, as shown in Table 1) in the presence and absence of milnacipran or

duloxetine (at concentrations ranging from 0.1 to 100 µM) to determine the IC50 value.

To examine their ability to act as time-dependent inhibitors, milnacipran or duloxetine (at the

same concentrations used to evaluate direct inhibition) were preincubated at 37 ± 1 °C, in

duplicate, with human liver microsomes and an NADPH-generating system for 30 minutes. After

the preincubation period, the marker substrate (at a concentration approximately equal to its Km)


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

8

was added, and the incubation continued for 5 min to measure residual CYP activity. Reactions

were terminated after 5 min by the addition of an equal volume of acetonitrile (containing the

appropriate internal standard, Table 1). Known direct-acting and metabolism-dependent

inhibitors were included as positive controls, most of which appear on the FDA’s list of

recommended or accepted in vitro inhibitors (Ogilvie et al. 2008; US FDA, 2006). Samples were

analyzed as described in the analytical methods section (see below).

In vitro CYP induction

The ability of milnacipran to induce or suppress the expression of CYP enzymes was

investigated in primary cultures of freshly isolated human hepatocytes with a Matrigel® overlay.

After a two-day adaptation period, three preparations of cultured human hepatocytes from three

separate human livers were treated once daily for three consecutive days with milnacipran (3, 10

and 30 µM) or one of three prototypical CYP inducers, namely omeprazole (100 µM),

phenobarbital (750 µM) and rifampin (10 µM), at the final concentrations indicated.

Milnacipran and the positive controls were dissolved in DMSO, and hepatocytes treated with

DMSO (final concentration 0.1%, v/v) served as negative controls. The isolation, culturing and

treatment procedures were performed essentially as described by Madan et al. (2003). Human

hepatocytes were harvested 24 h after the third treatment to prepare microsomes, as described by

Madan et al. (2003). Microsomes (0.004-0.02 mg/ml) were incubated with phenacetin (80 µM),

bupropion (500 µM), amodiaquine (20 µM), diclofenac (100 µM), S-mephenytoin (400 µM) and

testosterone (250 µM) for 10-30 min to measure CYP1A2, 2B6, 2C8, 2C9, 2C19 and 3A4/5

activity, respectively, essentially as described above (see in vitro CYP inhibition section).

Samples were analyzed as described in the analytical methods section (see below).


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

9

Analytical methods

All analyses of CYP enzyme activities were performed with validated HPLC-MS/MS methods.

The MS equipment was either an ABI Sciex (Applied Biosystems/MDS Sciex, Foster City, CA)

API 2000, API 3000 or API 4000 mass spectrometer with Shimadzu HPLC pumps and

autosampler systems. The HPLC columns used were as follows: Waters Atlantis C18 (5-µm

particle size, 50 mm × 2.1 mm) (Waters, Milford, MA) preceded by a Phenomenex Luna C-8

guard column (4.0 mm × 2.0 mm) (Phenomenex, Torrance, CA), or Waters Atlantis (5-µm

particle size, 100 mm × 2.1 mm) (Waters, Milford, MA) preceded by a Phenomenex Luna C-8

guard column (4.0 mm × 2.0 mm) (Phenomenex, Torrance, CA). Formic acid or ammonium

acetate based mobile phases were used for all sample analyses and flow rates ranged from

approximately 0.55 ml/min to 0.90 ml/min. All columns were maintained at ambient temperature

during analysis. Metabolites were quantified by back calculation of a weighted (1/x), linear,

least-squares regression. The regression fit was based on analyte/internal standard peak-area

ratios calculated from calibration standard samples, which were prepared from authentic

metabolite standards. Peak areas were integrated with Applied Biosystems/ MDS Sciex (Foster

City, CA) Analyst data system, version 1.4.1 or 1.4.2.

Statistical analyses

CYP inhibition data were processed with a validated, custom software program (DI IC50 LCMS

Template version 2.0.3) for the computer program Microsoft Excel (Office 2000 version 9.0,

Microsoft Inc., Redmond, WA), and IC50 values were determined by nonlinear regression with

XLfit3 (Version 3.0.5, ID Business Solutions Ltd., Guildford, Surrey, UK). XLfit is an Excel

add-in program that is a component of the validated DI IC50 LCMS Template version 2.0.3. This


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

10

software utilizes the Levenberg-Marquardt algorithm to perform non-linear regression fitting of

the data to the following 4-parameter sigmoidal-logistic IC50 equation:

( )

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛⎟⎠

⎞⎜⎝

⎛+

+=slope

50IC1

background - range background fit

x

Background was set to zero with a range up to 100 to express data as percent of control. This

software has been verified for its ability to calculate IC50 values that lie within the concentration

range of inhibitor studied. When less than 50% inhibition is observed, the data are not

extrapolated, hence, IC50 values are reported as being greater than the highest concentration of

inhibitor tested.

CYP induction data were processed with a validated, custom software program (EI Interim Data

Engine, version 1.2.1) for the computer program Microsoft Excel (Office 2003 version 11.0,

Microsoft Corporation, Redmond, WA). Statistically significant differences between group

means were calculated by equal variance and normality tests to determine if the data were

parametrically distributed. For parametrically distributed data sets, a one-way repeated measures

analysis of variance (ANOVA) was carried out to determine if there were significant differences

between the group means. For non-parametrically distributed data sets, a Kruskal-Wallis

ANOVA was performed. The ANOVA was followed by a Dunnett’s post hoc test to identify the

group means that were significantly different from the controls (p < 0.05 or 5% level of

significance). This statistical test is designed for multiple comparisons with a mean, such as

comparing multiple treatment groups with a control group. Statistical analyses were performed

with Sigma Stat Statistical Analysis System (version 2.03, Systat Software, Inc., Point

Richmond, CA). The enzyme-inducing effects of milnacipran and the prototypical inducers were

compared in terms of relative effectiveness, which was calculated as follows:


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

11

100rate) control - rate control (Positive

rate) control - rate an(Milnacipr esseffectiven relative Percent ×=

The positive control for CYP1A2 and CYP2B6 induction was omeprazole and phenobarbital,

respectively. For all other CYP enzymes the positive control was rifampin.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

12

RESULTS

In vitro CYP inhibition. Milnacipran and duloxetine (0.1 – 100 µM) were evaluated for their

abilities to inhibit CYP activity in pooled human liver microsomes with CYP-selective substrates

at concentrations approximately equal to Km. In the case of CYP3A4/5, two substrates were used

(testosterone and midazolam), as recommended by the US FDA (2006). Prior to measuring CYP

activity, the test articles were preincubated with NADPH-fortified human liver microsomes for

zero or 30 minutes to assess the potential for time-dependent inhibition. The results are

summarized in Figs. 2-3 and Table 2.

Milnacipran did not inhibit CYP1A2, 2B6, 2C8, 2C9, 2C19 or 2D6 (IC50 ≥ 100 µM), as shown

in Fig. 2. Duloxetine inhibited some of these enzymes more potently than milnacipran,

inhibiting CYP2D6 with an IC50 of 7 µM, and CYP2B6 with an IC50 of 15 µM. As shown in

Fig 2, preincubation of milnacipran with NADPH-fortified human liver microsomes did not

increase its inhibitory effect on CYP1A2, 2B6, 2C8, 2C9, 2C19 or 2D6. In contrast, duloxetine

produced time-dependent inhibition of CYP1A2, 2B6 and 2C19 as indicated by the leftward shift

in IC50 curves following the 30-min preincubation period. The 30-min preincubation period

slightly decreased the inhibitory effect of duloxetine on CYP2D6 (Fig. 2).

As shown in Fig. 3, milnacipran inhibited CYP3A4/5 in a substrate-dependent manner inasmuch

as it inhibited midazolam 1′-hydroxylation (IC50 = 28 µM) more potently than it inhibited

testosterone 6β-hydroxylation (IC50 = ~100 µM). Inhibition of midazolam 1´-hydroxylation was

also evaluated with a wider range of milnacipran concentrations (up to 500 µM), which indicated

an IC50 value of 31 µM, thereby confirming the original estimate of 28 µM (data not shown).

Duloxetine also directly inhibited CYP3A4/5, with IC50 values of 37 and 38 µM for the

1´-hydroxylation of midazolam and the 6β-hydroxylation of testosterone, respectively,


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

13

suggesting the inhibition is not dependent on substrate. Milnacipran did not produce

time-dependent inhibition (<10%) of CYP3A4/5 activity, whereas duloxetine did produce time-

dependent inhibition of CYP3A4/5 activity towards testosterone and midazolam (although the

time-dependent inhibition towards midazolam was to a lesser extent) (Fig. 3).

In vitro CYP induction. To evaluate the enzyme-inducing potential of milnacipran, three

preparations of freshly isolated human hepatocytes were cultured and treated once daily for three

consecutive days with milnacipran (3, 10 or 30 µM) or one of three prototypical enzyme

inducers, namely, omeprazole (100 µM), phenobarbital (750 µM) and rifampin (10 µM).

Microsomes were prepared 24 h after the final treatment and assayed for CYP1A2, 2B6, 2C8,

2C9, 2C19 and 3A4/5 activity. Under the conditions examined, milnacipran caused no cell

toxicity based on light-microscopic evaluation. Throughout the treatment period, the cultured

hepatocytes were free of detectable autophagic and lipid vesicles, were cuboidal in shape and

contained intact cell membranes and granular cytoplasm with one or two centrally located nuclei.

As shown in Table 3 and Fig. 4, all three preparations of human hepatocytes responded as

expected to treatment with prototypical CYP inducers: treatment with omeprazole produced a

marked increase in CYP1A2 (~37 fold), whereas treatment with phenobarbital or rifampin

produced an increase in CYP2B6 (6-10 fold), CYP2C8 (4-5 fold), CYP2C9 (~2 fold), CYP2C19

(3-6 fold), and CYP3A4 (~4 fold).

Treatment of human hepatocytes with up to 30 µM milnacipran for three consecutive days had

little or no effect on CYP1A2, CYP2C8, CYP2C9 or CYP2C19 activity (Table 3, Fig. 4).

Milnacipran produced a concentration-dependent increase in CYP2B6 activity with the highest

concentration tested (30 µM) effecting a statistically significant increase (2.59-fold: p < 0.05) in

CYP2B6 activity. At concentrations of 1, 10 and 30 µM, milnacipran was ~1%, 12% and 26%


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

14

as effective as phenobarbital as a CYP2B6 inducer (Fig. 5). In one of the three preparations of

hepatocytes treated with the highest concentration of milnacipran (30 µM), which is almost an

order of magnitude greater than the highest plasma Cmaxss value observed clinically (3.1 µM)

(Forest Research Institute, personal communication), the relative effectiveness for CYP2B6

induction exceeded the FDA’s cutoff value of 40% (measured value was 51.9%, individual data

not shown).

Milnacipran produced a concentration-dependent increase in CYP3A4/5 activity (Table 3,

Fig. 4). The highest concentration of milnacipran (30 µM) produced a 2.15-fold increase in

CYP3A4/5 activity, which was not statistically significant. At concentrations of 1, 10 and

30 µM, milnacipran was ~6%, 20% and 34% as effective as rifampin as a CYP3A4/5 inducer

(Fig. 5). In one of the three preparations of hepatocytes treated with the highest concentration of

milnacipran (30 µM), which is almost an order of magnitude greater than the highest plasma

Cmaxss value observed clinically (3.1 µM), the relative effectiveness for CYP3A4 induction

exceeded the FDA’s cutoff value of 40% (measured value was 42.4%, individual data not

shown).


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

15

DISCUSSION

The FDA provides the following guidance to design and interpret in vitro studies to evaluate the

victim and perpetrator potential of a new drug: (1) identify the role of CYP1A2, 2B6, 2C8, 2C9,

2C19, 2D6 and 3A4 in the metabolism of the drug and further clarify its victim potential by

identifying any other pathways that contribute 25% or more to the drug’s clearance; (2) evaluate

the potential for direct inhibition of CYP enzymes based on the ratio of [I], the plasma Cmaxss of

total (bound and free) drug, and the inhibition constant Ki, with a cutoff value of [I]/Ki = 0.1,

below which it is reasonable to assume a drug will not cause clinically significant CYP

inhibition; (3) evaluate the potential for time-dependent inhibition of CYP enzymes and conduct

clinical studies to assess the in vivo significance of positive in vitro findings, and (4) evaluate the

potential for enzyme induction in three preparations of human hepatocytes and conduct clinical

enzyme induction studies when, at pharmacologically relevant concentrations, a drug is 40% or

more as effective as a suitable positive control (US FDA, 2006; Huang et al., 2008).

Previous studies have established that milnacipran has low victim potential with respect to

metabolism by cytochrome P450. More than half (50-60%) of the drug is eliminated unchanged

in urine, 20% is conjugated to a carbamoyl-glucuronide and 8% is metabolized by cytochrome

P450 (Delini-Stula, 2000; Puozzo et al., 2005; Tsuruta et al., 2000; Forest Research Institute,

personal communication). The latter is most likely catalyzed by CYP3A4, which converts

milnacipran to one hydroxylated and two N-dealkylated metabolites (Puozzo et al., 1996; Tsuruta

et al., 2000). N-Desethyl-milnacipran is the major circulating oxidative metabolite of

milnacipran, accounting for approximately 10% of the dose excreted in urine (Puozzo et al.,

2005). In contrast, the SNRI duloxetine has high victim potential with respect to metabolism by

cytochrome P450. Duloxetine is extensively metabolized by CYP1A2 and CYP2D6, and clinical


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

16

studies with [14C]-duloxetine have shown that the parent drug accounts for only 3% of systemic

exposure (plasma AUC) to [14C]-duloxetine-derived radioactivity

(http://www.fda.gov/cder/foi/nda/2004/021427_s000_Cymbalta.htm and

http://www.fda.gov/cder/foi/label/2007/021427s015s017lbl.pdf). Inhibition of CYP1A2 by

fluvoxamine results in a 5- to 6-fold increase in duloxetine AUC in CYP2D6 PM subjects;

whereas, inhibition of CYP2D6 by paroxetine increases plasma AUC by 60% in EMs (Cymbalta

package insert, http://www.fda.gov/cder/foi/label/2007/021427s015s017lbl.pdf). Cigarette

smoking, which induces CYP1A2, decreases the plasma AUC of duloxetine by about one third

(http://www.fda.gov/cder/foi/nda/2004/021427_s000_Cymbalta.htm and

http://www.fda.gov/cder/foi/label/2007/021427s015s017lbl.pdf). In contrast to the situation with

duloxetine and several other antidepressants (Bertilsson et al., 2002; Preskorn et al., 2007),

genetic polymorphisms in CYP2D6 and CYP2C19 have no impact on the PK of milnacipran

(Puozzo et al., 2005), nor does inhibition of CYP2D6 and CYP3A4 by fluoxetine (DeVane et al.,

2004; Puozzo et al., 2006). Enzyme induction by carbamazepine is associated with a small

decrease (20%) in milnacipran plasma steady state concentrations (Puozzo et al., 2002).

The present in vitro study was designed to evaluate the perpetrator potential of milnacipran. At

the pharmacologically relevant concentration of 3 µM and even at 10 µM milnacipran, which is

4-5 times the mean steady-state plasma Cmaxss of 2.2 µM, milnacipran produced no significant

induction of CYP1A2, 2B6, 2C8, 2C9, 2C19 or CYP3A4 in cultured human hepatocytes under

conditions where the positive controls exerted their anticipated inductive effects. At 30 µM

(~14x Cmaxss), milnacipran was 26% as effective as phenobarbital at inducing CYP2B6 and was

34% as effective as rifampin at inducing CYP3A4. These in vitro results indicate that, at

pharmacologically relevant concentrations, milnacipran is not 40% or more as effective as


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

17

phenobarbital or rifampin at inducing CYP enzymes, which, based on the relevant FDA

Guidance for Industry (2006) suggests that milnacipran will not produce clinically significant

enzyme induction.

The potential for milnacipran to inhibit CYP enzymes in human liver microsomes was compared

with that of duloxetine. The only CYP enzyme potentially inhibited by milnacipran was

CYP3A4, the enzyme implicated in its N-dealkylation and hydroxylation, which are minor

pathways of milnacipran clearance (Tsuruta, 2000). Milnacipran inhibited the 1′-hydroxylation

of midazolam with an IC50 of approximately 30 µM. The concentration of midazolam was 4 µM,

which is approximately equal to its Km; hence, the Ki value would be ~15 µM or ~30 µM

depending on whether the inhibition of CYP3A4 by milnacipran was competitive or non-

competitive, respectively (Ogilvie et al., 2008). Following a dosage of 100 mg b.i.d., the mean

Cmax at steady state is 2.2 µM, and ranges from 1.3 to 3.1 µM (Forest Research Institute,

personal communication). Based on a conservative estimate of Ki (~15 µM) and the mean

Cmaxss of 2.2 µM, the value of [I]/Ki for CYP3A4 inhibition by milnacipran is ~0.15, which

slightly exceeds the FDA’s cut-off of 0.1. When testosterone was used to measure CYP3A4

activity, the Ki for milnacipran was conservatively estimated to be 51 µM (based on an IC50 of

~102 µM). Accordingly, [I]/Ki is 0.043 (based on mean Cmaxss of 2.2 µM), which falls below

the FDA’s cut-off value of 0.1.

Milnacipran has been reported to have no effect on the urinary excretion of 6β-hydroxycortisol

or the PK of carbamazepine, which, despite being imperfect markers of CYP3A4 activity,

provides some evidence that milnacipran does not cause clinically significant inhibition of

CYP3A4 (Table 4) (Puozzo et al., 2005). For all other CYP enzymes, estimates of [I]/Ki are less

than 0.1, and there was no evidence of metabolism-dependent inhibition. Based on this cut-off,


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

18

the lack of time-dependent inhibition and the rank order approach to extrapolating in vitro

findings to the in vivo situation (Obach et al., 2005; Obach et al., 2006), milnacipran would not

be expected to cause clinically significant inhibition of CYP1A2, 2B6, 2C8, 2C9, 2C19 or 2D6.

In clinical drug-drug interaction studies, (summarized in Table 4), milnacipran has been shown

to cause no inhibition (or induction) of CYP1A2 (with caffeine as the in vivo probe substrate),

CYP2C19 (racemic mephenytoin) and CYP2D6 (sparteine) (Puozzo et al., 2005). Overall, there

is good correspondence between the in vitro results and the available in vivo clinical findings.

Compared with milnacipran, duloxetine was a more potent inhibitor of all the CYP enzymes

examined, and showed evidence of time-dependent inhibition of CYP1A2, 2B6, 2C19 and

CYP3A4/5 (Figs. 2-3). This study did not establish the effects of NADPH, or whether the

metabolism-dependent inhibition of CYP enzymes by duloxetine was due to the formation of

metabolites that are more potent reversible inhibitors or are irreversible inhibitors of CYP1A2,

2B6, 2C19 and 3A4. Based on experiments with recombinant human CYP enzymes, Lobo et al.

(2008) reported that the hydroxylation of the naphthyl ring of duloxetine is catalyzed by

CYP1A2 (in the 4-, 5- and 6-positions) and by CYP2D6 (in the 4- and 5-positions). These

hydroxylated metabolites (as well as di-hydroxylated metabolites and conjugates) are major

duloxetine metabolites in human plasma and urine, whereas N-demethylated and O-dealkylated

metabolites are minor in vivo metabolites (Lantz et al., 2003). N-Demethylation of duloxetine

would be expected to produce a more potent direct-acting CYP inhibitor than the parent drug,

whereas di-hydroxylation of duloxetine to a catechol metabolite on the naphthyl ring may

potentially lead to irreversible inhibition of one or more CYP enzymes. In this regard, it is

interesting that CYP1A2 (which showed evidence of time-dependent inhibition) catalyzes the 5-

and 6-hydroxylation of duloxetine. Consequently, hydroxylation at both these sites (which


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

19

appear on the same ring and are adjacent to each other) would lead to catechol formation. In

contrast, CYP2D6 (which showed no evidence of time-dependent inhibition) catalyzes the 4- and

5-hydroxylation of duloxetine. Hydroxylation at both these sites would not produce a catechol

because these two sites are not adjacent to each other but appear on different rings of the

naphthyl moiety (Lantz et al., 2003; Lobo et al., 2008). Duloxetine also contains a thiophene

ring. Although metabolites involving thiophene oxidation have not been reported for duloxetine,

this particular functional group is associated with several cases of irreversible CYP inhibition, as

in the case of tienilic acid, ticlopidine and clopidogrel (Fontana et al., 2005; Ogilvie et al., 2008;

Parkinson and Ogilvie, 2008).

Duloxetine inhibited both of the enzymes implicated in its metabolism, namely CYP1A2 and

CYP2D6. Duloxetine inhibited CYP1A2 with an IC50 of 50 µM without any preincubation and

an IC50 of 18 µM with a 30-min preincubation. Lobo et al. (2008) also evaluated duloxetine as a

direct-acting inhibitor of CYP1A2 and reported that duloxetine causes competitive inhibition of

CYP1A2 with a Ki value of 18 µM. When incubated with marker substrate at a concentration

equal to Km, the Ki value for a competitive inhibitor is half its IC50 value, hence, the Ki of

18 µM reported by Lobo et al. (2008) translates to an IC50 value of 36 µM, which is comparable

to our value of 50 µM (determined without a preincubation period) and 18 µM (determined with

a preincubation period). Lobo et al. (2008) do not specifically report having evaluated duloxetine

as a time-dependent inhibitor of CYP1A2. Clinical interaction studies with theophylline, a

CYP1A2 substrate, established that duloxetine is a weak inhibitor of CYP1A2 in vivo; it

increased the plasma AUC of theophylline by 7% (1-15%) in one study and 20% (13-27%) in

another (Table 5). With no pre-incubation, the enzyme most potently inhibited by duloxetine was

CYP2D6 (IC50 = 7 µM). With a preincubation, the enzyme most potently inhibited by


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

20

duloxetine was CYP2B6 (IC50 = ~5 µM). Duloxetine produces clinically significant inhibition

of CYP2D6 based on its ability to cause up to a 3-fold increase in the plasma AUC of

desipramine, a sensitive CYP2D6 in vivo probe drug (Table 5). The enzyme most potently

inhibited by duloxetine in vitro was CYP2D6, and the most pronounced clinical drug-drug

interaction reported for duloxetine is its interaction with desipramine, a drug whose clearance is

largely dependent (80-90%) on metabolism by CYP2D6. However, the inhibition of CYP2D6

observed in vivo would not be predicted from the in vitro inhibition data based on [I]/ Ki. The

average maximum plasma concentrations of duloxetine at steady state (Cmaxss) is 20.7 ng/ml or

~0.07 µM (Skinner et al., 2003), and a conservative estimate of Ki for the inhibition of CYP2D6

by duloxetine is 3.5 µM (assuming the inhibition is competitive, such that Ki is half the IC50

value). Accordingly, the [I]/ Ki value for the direct inhibition of CYP2D6 by duloxetine (0.02) is

well below the FDA’s cut-off of 0.1. It is not clear why the extrapolation of the in vitro data to

the in vivo situation based on the [I]/ Ki value underestimates the clinical inhibition of CYP2D6

by duloxetine. Duloxetine undergoes rapid and extensive first-pass metabolism in the liver and

gut, for which reason parent drug accounts for only 3% or 9% of drug-related material in plasma

based on AUC and Cmax, respectively (Lantz et al., 2003). Accordingly, hepatic levels of

duloxetine may be considerably greater than those in plasma, hence, the underestimation may be

the result of basing the [I]/ Ki value on too low a value of [I].

Duloxetine produced metabolism-dependent inhibition of CYP2B6 and CYP2C19 in vitro;

however, the effect of duloxetine on the in vivo disposition of probe drugs for these enzymes has

not been investigated. Duloxetine did not produce direct or time-dependent inhibition of

CYP2C9, however, in a single case report, duloxetine was found to be the likely cause of an

increased INR in a patient who had been on a stable dose of warfarin for a year before starting


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

21

duloxetine (Glueck et al., 2006). Milnacipran did not inhibit CYP2C9 in vitro, and did not alter

the pharmacokinetics of S-warfarin (Forest Research Institute, personal communication).

Furthermore, milnacipran did not affect the pharmacodynamics of warfarin as indicated by the

International Normalized Ratio (INR) (Forest Research Institute, personal communication). At

steady state, the plasma Cmax of duloxetine ranges from 15-35 ng/ml (about 0.05 to 0.1 µM).

Based on a conservative estimate of 3.5 µM for the Ki value for inhibition of CYP2D6 (i.e., half

the IC50 value with [S] = Km) and a Cmaxss value of 0.1 µM, the [I]/Ki value is well below the

FDA cut-off of 0.1, and yet duloxetine causes clinically significant inhibition of CYP2D6.

Duloxetine is metabolized so extensively that parent drug accounts for only 3% of systemic

exposure to [14C]-duloxetine-derived radioactivity. Although metabolites of duloxetine might

account for the greater degree of CYP2D6 inhibition observed in vivo compared with that

predicted from in vitro studies, it is interesting that CYP2D6 was not among the enzymes that

duloxetine inhibited in a metabolism-dependent manner.

In summary, the results of this in vitro study established that duloxetine inhibits CYP2D6 and

other CYP enzymes, and has been shown to cause clinically significant inhibition of CYP2D6.

In contrast, the only human CYP enzyme inhibited by milnacipran is CYP3A4, which

milnacipran inhibited weakly and in a substrate-dependent manner (midazolam, but not

testosterone). Milnacipran would not be expected to produce clinically significant inhibition of

CYP enzymes, which is consistent with clinical data demonstrating a lack of interaction between

milnacipran and drugs metabolized by CYP1A2, 2C9, 2C19, 2D6 or 3A4. In addition, the

results of the present study suggest that milnacipran will not produce clinically significant

induction of CYP enzymes.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

22

REFERENCES

Bertilsson L, Dahl ML, Dalen P, and Al-Shurbaji A (2002) Molecular genetics of CYP2D6:

clinical relevance with focus on psychotropic drugs. Br J Clin Pharmacol 53:111-122.

Bjornsson TD, Callaghan JT, Einolf HJ, Fischer V, Gan L, Grimm S, Kao J, King SP, Miwa G,

Ni L, Kumar G, McLeod J, Obach RS, Roberts S, Roe A, Shah A, Snikeris F, Sullivan

JT, Tweedie D, Vega JM, Walsh J, and Wrighton SA (2003) The conduct of in vitro and

in vivo drug-drug interaction studies: A Pharmaceutical Research and Manufacturers of

America (PhRMA) perspective. Drug Metab Dispos 31:815-832.

Caccia S (1998) Metabolism of the newer antidepressants. An overview of the pharmacological

and pharmacokinetic implications. Clin Pharm 34:281-302.

DeVane CL, Donovan JL, Liston HL, Markowitz JS, Cheng KT, Risch SC, and Willard L (2004)

Comparative CYP3A4 inhibitory effects of venlafaxine, fluoxetine, sertraline, and

nefazodone in healthy volunteers. J Clin Psychopharmacol 24:4-10.

Delini-Stula A (2000) Milnacipran: An antidepressant with dual selectivity of action on

noradrenaline and serotonin uptake. Hum Psychopharmacol Clin Exp 15:255-260.

Dubner R and Hargreaves KM (1989) The neurobiology of pain and its modulation. Clin J Pain

5 Suppl 2:S1-4.

Fontana E, Dansette PM, and Poli SM (2005) Cytochrome p450 enzymes mechanism based

inhibitors: common sub-structures and reactivity. Current Drug Metab 6:413-454.

Grzesiak M, Beszlej JA, and Kiejna A (2000) [Pharmacokinetics of second generation

antidepressants]. Psychiatr Pol 34:577-594.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

23

Glueck CJ, Khalil Q, Winiarska M, and Wang P (2006) Interaction of duloxetine and warfarin

causing severe elevation of international normalized ratio. JAMA 295:1517-1518.

Huang S, Strong JM, Zhang L, Reynolds K, Nallani S, Temple R, Abraham S, Habet SA, Baweja

RK, Burckart GJ, Chung S, Colangelo P, Furcht D, Green MD, Hepp P, Karnaukhova E,

Ko HS, Lee JI, Marroum P, Norden J, Qiu W, Rahman A, Sobel S, Stifano T, Thummel

K, Wei X, Yasuda S, Zheng J, Zhao H, and Lesko L (2008) New era in drug interaction

evaluation: US Food and Drug Administration update on CYP enzymes, transporters, and

the guidance process. J Clin Pharmacol. 48:662-670.

Hua TC, Pan A, Chan C, Poo YK, Skinner MH, Knadler MP, Gonzales CR, and Wise SD (2004)

Effect of duloxetine on tolterodine pharmacokinetics in healthy volunteers. Br J Clin

Pharmacol 57:652-656.

Lantz RJ, Gillespie TA, Rash TJ, Kuo F, Skinner M, Kuan HY, and Knadler MP (2003)

Metabolism, excretion, and pharmacokinetics of duloxetine in healthy human subjects.

Drug Metab Dispos 31:1142-1150.

Lobo ED, Bergstrom RF, Reddy S, Quinlan T, Chappell J, Hong Q, Ring B, and Knadler MP

(2008) In vitro and in vivo evaluations of cytochrome P450 1A2 interactions with

duloxetine. Clin Pharmacokinet 47:191-202.

Madan A, Graham RA, Carroll KM, Mudra DM, Burton LA, Krueger LA, Downey AD,

Czerwinski M, Forster J, Ribadeneira MD, Gan L, LeCluyse EL, Zech K, Robertson P,

Jr. Koch P, Antonian L, Wagner G, Yu L, and Parkinson A (2003) Effects of prototypical

microsomal enzyme inducers on cytochrome P450 (CYP) expression in cultured human

hepatocytes. Drug Metab Dispos 31:421-431.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

24

Obach RS, Walsky RL, Venkatakrishnan K, Houston JB, and Tremaine LM (2005) In vitro

cytochrome P450 inhibition data and the prediction of drug-drug interactions: Qualitative

relationships, quantitative predictions, and the rank-order approach. Clin Pharmacol Ther

78:582-592.

Obach RS, Walsky RL, Venkatakrishnan K, Gaman EA, Houston JB, and Tremaine LM (2006)

The utility of in vitro cytochrome P450 inhibition data in the prediction of drug-drug

interactions. J Pharmacol Exp Ther 316:336-348.

Ogilvie BW, Zhang D, Li W, Rodrigues AD, Gipson AE, Holsapple J, Toren P, and Parkinson A

(2006) Glucuronidation converts gemfibrozil to a potent, metabolism-dependent inhibitor

of CYP2C8: implications for drug-drug interactions. Drug Metab Dispos 34:191-197.

Ogilvie BW, Usuki E, Yerino P, and Parkinson A (2008) In vitro approaches for studying the

inhibition of drug-metabolizing enzymes and identifying the drug-metabolizing enzymes

responsible for the metabolism of drugs (reaction phenotyping) with emphasis on

cytochrome P450, in Drug-Drug Interactions. Drugs and the Pharmaceutical Sciences,

2nd ed. (Rodrigues AD, ed.) pp 231-358, Informa Healthcare, New York.

Parkinson A and Ogilvie BW (2008) Chapter 6: Biotransformation of Xenobiotics, in Casarett &

Doull’s Toxicology. The Basic Science of Poisons, 7th ed. (Klaassen CD, ed.) pp 161-304,

McGraw-Hill, New York.

Patroneva A, Connolly S, Fatato P, Pedersen R, Jiang Q, Paul J, Guico-Pabia C, Isler JA,

Burczynski ME, and Nichols A (2008) An Assessment of Drug-Drug Interactions: The

Effect of Desvenlafaxine and Duloxetine on the Pharmacokinetics of the CYP2D6 Probe

Desipramine in Healthy Subjects. Drug Metab Dispos doi: 10.1124/dmd.108.021527.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

25

Preskorn SH, Greenblatt DJ, Flockhart D, Luo Y, Perloff ES, Harmatz JS, Baker B, Klick-Davis

A, Desta Z, and Burt T (2007) Comparison of duloxetine, escitalopram, and sertraline

effects on cytochrome P450 2D6 function in healthy volunteers. J Clin Psychopharmacol

27:28-34.

Puozzo C and Leonard BE (1996) Pharmacokinetics of milnacipran in comparison with other

antidepressants. Int Clin Psychopharmacol 11 Suppl 4:15-27.

Puozzo C, Panconi E, and Deprez D (2002) Pharmacology and pharmacokinetics of milnacipran.

Int Clin Psychopharmacol 17 Suppl 1:S25-35.

Puozzo C, Lens S, Reh C, Michaelis K, Rosillon D, Deroubaix X, and Deprez D (2005) Lack of

interaction of milnacipran with the cytochrome P450 isoenzymes frequently involved in

the metabolism of antidepressants. Clin Pharmacokinet 44:977-988.

Puozzo C, Hermann P, and Chassard D (2006) Lack of pharmacokinetic interaction when

switching from fluoxetine to milnacipran. Int Clin Psychopharmacol 21:153-158.

Robertson P, Decory HH, Madan A, and Parkinson A (2000) In vitro inhibition and induction of

human hepatic cytochrome P450 enzymes by modafinil. Drug Metab Dispos 28:664-

671.

Sawada Y and Ohtani H (2001) [Pharmacokinetics and drug interactions of antidepressive

agents]. Nippon Rinsho 59:1539-1545.

Skinner MH, Kuan HY, Pan A, Sathirakul K, Knadler MP, Gonzales CR, Yeo KP, Reddy S, Lim

M, Ayan-Oshodi M, and Wise SD (2003) Duloxetine is both an inhibitor and a substrate

of cytochrome P4502D6 in healthy volunteers. Clin Pharmacol Ther 73:170-177.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

26

US Food and Drug Administration. Draft Guidance for Industry: Drug Interaction Studies—

Study Design, Data Analysis and Implications for Dosing and Labeling, 2006.

http://www.fda.gov/cder/guidance/6695dft.pdf. Accessed November 4, 2008.

Tucker GT, Houston JB, and Huang SM (2001) Optimizing drug development: strategies to

assess drug metabolism/transporter interaction potential--toward a consensus. Pharm Res

18:1071-1080.

Tsuruta K, Tsurui K, Okazaki K, Ueda K, Kawasumi S, Shimada N, and Sawada Y (2000)

Examination of drug-drug interaction of milnacipran hydrochloride in the presence of

human P-450. Iyakuhin Kenkyu 31:659-667 (Japanese with English abstract).

Vaishnavi SN, Nemeroff CB, Plott SJ, Rao SG, Kranzler J, and Owens MJ (2004). Milnacipran:

a comparative analysis of human monoamine uptake and transporter binding affinity. Biol

Psychiatry 55:320-322.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

27

Footnotes:

This study was sponsored by Forest Research Institute, Jersey City, NJ; however, the views

expressed herein are those of the publishing authors alone.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

28

FIGURE LEGENDS

Fig. 1: Structures of the dual reuptake inhibitors milnacipran (Savella™) and duloxetine

(Cymbalta®).

Fig. 2: Effects of milnacipran and duloxetine on selected CYP activities with and without a 30-

min preincubation with NADPH-fortified human liver microsomes. For each CYP enzyme

assayed, the substrate concentration was approximately equal to Km (see Table 1 for details).

Fig. 3: Effects of milnacipran and duloxetine on CYP3A4/5 activity towards testosterone and

midazolam with and without a 30-min preincubation with NADPH-fortified human liver

microsomes. The concentration of testosterone and midazolam was approximately equal to Km

(see Table 1 for details).

Fig. 4: Effects of treating cultured human hepatocytes with milnacipran or prototypical inducers

on microsomal CYP activity. Three preparations of human hepatocytes were treated once daily

for three consecutive days with milnacipran or one of three prototypical enzyme inducers.

Microsomes were prepared 24 h after the last treatment and assayed for CYP activity as

described in Materials and Methods. Values are presented as fold increase over negative control

(microsomes from DMSO-treated hepatoctyes) based on the absolute values shown in Table 3. *,

significantly different from vehicle control (DMSO) p<0.05 when the positive control groups

were excluded from the statistical analysis; †, statistically significance found among treatment

groups according to Kruskal-Wallis One Way Analysis of Variance on ranks (p<0.05) and


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

29

Dunnett’s Method when the positive control groups (omeprazole, phenobarbital and rifampin)

were included in the statistical analysis; ††, significantly different from the vehicle control

(DMSO) according to the Dunnett’s Method (p<0.05) when the positive control groups

(omeprazole, phenobarbital and rifampin) were included in the statistical analysis.

Fig. 5: Comparison of the effectiveness of milnacipran as an enzyme inducer in human

hepatocytes relative to prototypical inducers. Percent relative effectiveness was calculated as

described in Materials and Methods, based on the CYP activities in Table 3. For CYP1A2 and

CYP2B6, the positive control was omeprazole and phenobarbital, respectively. For all other

CYP enzymes, the positive control was rifampin. ND, not detected; OME, omeprazole; PB,

phenobarbital; RIF, rifampin.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

30

Table 1: Experimental conditions for measuring microsomal CYP activity for enzyme inhibition studies

Enzyme CYP Activity Substrate

concentration (µM)

Protein (µg/ml)

Ionization mode a

Mass transition monitored

(amu b)

Internal standard

CYP1A2 Phenacetin O-dealkylation

40 100 ESI+ 152 � 110 d4-Acetaminophen

CYP2B6 Bupropion

hydroxylation 50 100 ESI+ 256 � 238 d6-Hydroxybupropion

CYP2C8 Amodiaquine N-

dealkylation 7 100 ESI+ 328 � 283 d5-N-Desethylamodiaquine

CYP2C9 Diclofenac 4´-hydroxylation

6 100 ESI– 310 � 266 d4-4´-Hydroxydiclofenac

CYP2C19 S-Mephenytoin 4´-

hydroxylation 40 100 ESI– 233 � 190 d3-4´-Hydroxymephenytoin

CYP2D6 Dextromethorphan O-demethylation

7.5 100 ESI+ 258 � 157 d3-Dextrorphan

CYP3A4/5 Testosterone 6β-

hydroxylation 100 100 ESI+ 305 � 269 d3-6β-Hydroxytestosterone

CYP3A4/5 Midazolam 1´-hydroxylation

4 50 ESI+ 342 � 324 d3-1´-Hydroxymidazolam

a Indicates the type of ionization (i.e., electrospray ionization [ESI]) and the polarity (+ or −).

b Atomic mass units


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

31

Table 2: Comparison of milnacipran and duloxetine as inhibitors of selected CYP enzymes with and without a 30-min preincubation with NADPH-fortified human liver microsomes

Enzyme CYP Activity

Direct inhibition (Zero-minute preincubation)

Time-dependent inhibition (30-minute preincubation)

Milnacipran IC50 (µM)

Duloxetine IC50 (µM)

Milnacipran IC50 (µM)

Duloxetine IC50 (µM)

CYP1A2 Phenacetin O-dealkylation >100 50 >100 18

CYP2B6 Bupropion hydroxylation 98 15 >100 4.7

CYP2C8 Amodiaquine N-dealkylation >100 60 >100 49

CYP2C9 Diclofenac 4´-hydroxylation >100 >100 >100 >100

CYP2C19 S-Mephenytoin 4´-hydroxylation >100 27 >100 8.6

CYP2D6 Dextromethorphan O-demethylation >100 7.0 >100 13

CYP3A4/5 Testosterone 6β-hydroxylation >100 a 38 >100 17

CYP3A4/5 Midazolam 1´-hydroxylation 28 37 24 26

a The IC50 value calculated was 102 µM, but is reported as >100 µM since concentrations beyond the concentration range studied should be considered estimates.


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

32

Table 3: Effects of treating cultured human hepatocytes with DMSO, milnacipran or prototypical inducers on microsomal CYP activity

Treatment Concentration Enzyme activity (pmol/mg microsomal protein/min) a

CYP1A2 CYP2B6 CYP2C8 CYP2C9 CYP2C19 CYP3A4/5

DMSO 0.1% (v/v) 63.5 ± 50.8 65.9 ± 34.4 236 ± 73 928 ± 320 6.41 ± 4.45 2980 ± 1040

Milnacipran 3 µM 64.6 ± 50.1 76.1 ± 48.9 259 ± 124 979 ± 348 6.25 ± 3.71 3470 ± 480

Milnacipran 10 µM 72.1 ± 56.1 111 ± 61 276 ± 81 997 ± 305 6.45 ± 3.44 4700 ± 540

Milnacipran 30 µM 73.4 ± 59.4 168 ± 90* 285 ± 74 1010 ± 280 5.80 ± 3.08 5840 ± 580

Omeprazole 100 µM 1630 ± 750† 419 ± 214† 904 ± 483 1320 ± 650 8.27 ± 6.30 4100 ± 2280

Phenobarbital 750 µM 124 ± 69 666 ± 399† 959 ± 257† 1680 ± 650† 20.8 ± 17.0 11400 ± 1300††

Rifampin 10 µM 114 ± 86 388 ± 262† 1130 ± 250† 1930 ± 570† 43.7 ± 33.1†† 11500 ± 1000††

a Values (rounded to three significant figures) are the mean ± standard deviation (rounded to the same degree of accuracy) of three determinations (i.e., three human hepatocyte preparations).

* Significantly different from the vehicle control (DMSO) according to the Dunnett’s Method (p<0.05) when the positive control groups (omeprazole, phenobarbital and rifampin) were excluded from the statistical analysis.

† Statistical significance found among the treatment groups according to Kruskal-Wallis One Way Analysis on ranks (p<0.05) and Dunnett’s Method when the positive control groups (omeprazole, phenobarbital and rifampin) were included in the statistical analysis.

†† Significantly different from the vehicle control (DMSO) according to the Dunnett’s Method (p<0.05) when the positive control groups (omeprazole, phenobarbital and rifampin) were included in the statistical analysis.

CYP1A2: Phenacetin O-dealkylation

CYP2B6: Bupropion hydroxylation

CYP2C8: Amodiaquine N-dealkylation

CYP2C9: Diclofenac 4´-hydroxylation

CYP2C19: S-Mephenytoin 4´-hydroxylation

CYP3A4/5: Testosterone 6β-hydroxylation


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

33

Table 4: Summary of clinical drug-drug interactions studies to evaluate the effect of milnacipran on the disposition of coadministered drugs.

Victim drug Na Enzymeb Treatment Change in

pharmacokinetics of the victim drug

Reference

Sparteine 25 CYP2D6

Milnacipran: 50 mg single dose (Day 1), then 50 mg b.i.d. (Days 2-8) Sparteine: 100 mg q.d. (Days -2, 1, 8, 20)

19.5% increase in sparteine/metabolite ratio

in CYP2C19 EMsc

Puozzo et al., 2005

Mephenytoin 25 CYP2C19

Milnacipran: 50 mg single dose (Day 1), then 50 mg b.i.d. (Days 2-8) Mephenytoin: 100 mg q.d. (Days -2, 1, 8, 20)

No increase in S/R-mephenytoin ratiod

Puozzo et al., 2005

Caffeine 25 CYP1A2

Milnacipran: 50 mg single dose (Day 1), then 50 mg b.i.d. (Days 2-8) Caffeine: 200 mg q.d., (Days -2, 1, 8, 20)

No increase in the caffeine/paraxanthine

AUC0-12 ratioe

Puozzo et al., 2005

Warfarin 25

CYP2C9 for S-warfarin and CYP1A2/3A4 for R-warfarinf

Milnacipran: 25 mg bid (Days 1-3), 50 mg bid (Days 4-6), and 100 mg bid (Days 7-11) Warfarin: 25 mg single dose on Day 11 and Day 25 Or Milnacipran: 25 mg bid (Days 15-17), 50 mg bid (Days 18-20), and 100 mg bid (Days 21-25) Warfarin: 25 mg single dose on Day 1 and Day 25

No change in S- or R-warfarin pharmaco-

kinetics or pharmaco-dynamics (INR)

FDA drug label and FRI personal communication

Carbamazepine 25 CYP3A4

Milnacipran: 50 mg bid (Days 1-4) and Days (29 – 35) Carbamazepine: 100 mg bid (Days 8-11); 200 mg bid (Days 12 - 35)

No change in the pharmacokinetics of carbamazepine or its epoxide metabolite

FDA drug label and FRI personal communication

a Number of subjects completing each study b Principal metabolizing enzyme c Prior to milnacipran treatment, the sparteine/metabolite ratio was 0.51 and 14.6 in CYP2C19 extensive

metabolizers (EMs) and poor metabolizers (PMs), respectively

d Prior to milnacipran treatment, the S/R-mephenytoin ratio was 0.066 and 1.56 in CYP2C19 extensive metabolizers (EMs) and poor metabolizers (PMs), respectively

e Prior to milnacipran treatment, the caffeine/paraxanthine AUC ratio was 2.3 f Two-sequence crossover design used

FDA, Food & Drug Administration; FRI, Forest Research Institute; INR, international normalized ratio (a measure of prothrombin time)


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


DMD 28274

34

Table 5 Summary of clinical drug-drug interaction studies to evaluate the effect of duloxetine on the disposition of coadministered drugs

Victim drug Na Enzymeb Treatmentc

Arithmetic mean AUC

change (%)

AUC Geometric mean ratio

Reference Mean 90% CI

Theophylline 28 CYP1A2

Duloxetine: 60 mg b.i.d. (Days 1-4), then 60 mg q.d. (Day 5) Theophylline: 197.5 mg (Day 5)

1.13 1.07 – 1.18d Lobo et al., 2008

Temazepam Unk. UGT

Duloxetine: 20 mg q.d. (Days 1-6) Temazepam: 30 mg q.d. (Days 1-6)

8.6 1.11 1.02-1.21 Duloxetine NDA (21-427)e

Lorazepam 16 UGT

Duloxetine: 60 mg b.i.d. (Days 1-6) Lorazepam: 2 mg b.i.d. (Days 4-6)

8.2 1.08 1.00-1.16 Duloxetine NDA (21-427)e

Desipramine 17 CYP2D6

Duloxetine: 30 mg b.i.d. (Days 6-15) Desipramine: 50 mg (Day 11)

107 2.22 1.95-2.51 Patroneva et al., 2008

Desipramine 13 CYP2D6

Duloxetine: 40 mg b.i.d. (Days 8-13), then 60 mg b.i.d. (Days 14-27) Desipramine: 50 mg (Day 21)

168 2.92 2.55-3.34 Skinner et al., 2003

Metoprolol 16 CYP2D6

Duloxetine: 30 mg q.d. (Day 1), then 60 mg q.d. (Days 2-17) Metoprolol: 100 mg q.d. (Day 17)

180f

94.5g

Preskorn et al., 2007

Tolterodine 16 CYP2D6,

3A4

Duloxetine: 40 mg b.i.d. (Days 1–4, then 40 mg q.d. (Day 5) Tolterodine: 2 mg b.i.d. (Days 1-4, then 2 mg q.d. (Day 5)

1.71 1.31-2.23h Hua et al., 2004

a Number of subjects completing each study b Principal metabolizing enzyme c q.d. and b.i.d. refer to once-a-day and twice-a-day dosing, respectively d Theophylline AUC increased 20% in women (n=18; statistically significant) but only 7% in men (n=10;

statistically insignificant). e Duloxetine NDA (21-427): http://www.fda.gov/cder/foi/nda/2004/021427_s000_Cymbalta.htm f Original value reported in Table 5 in Preskorn et al., 2007 g Recalculated value based on AUC values reported in Table 4 in Preskorn et al., 2007. The recalculated value

agrees with that reported in the University of Washington Metabolism and Transport Drug Interaction Database (MTDI database: http://www.druginteractioninfo.org).

h Value represents 95% CI (all other values are 90% CI) CI, Confidence interval; IV, intravenous; Unk, Unknown; UGT, UDP-glucuronosyltransferase


at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


N

O

NH2

O

S

NH

1R(S), 2S(R)-Milnacipran (S)-Duloxetine

Figure 1 This article has not been copyedited and form

atted. The final version m

ay differ from this version.

DM

D Fast Forw

ard. Published on July 16, 2009 as DO

I: 10.1124/dmd.109.028274

at ASPET Journals on March 16, 2021 dmd.aspetjournals.org Downloaded from


CYP1A2

[Drug concentration] (µM)0.1 1 10 100 1000

Phe

nac

etin

O-d

ealk

ylat

ion

(Per

cent

con

trol

)0

20

40

60

80

100

120

>100 µM

>100 µM

50 µM

18 µM

IC50 values

CYP1A2


Bu

pro

pio

n h

ydro

xyla

tio

n(P

erce

nt c

ontr

ol)

0

20

40

60

80

100

120

>100 µM

>100 µM

15 µM

4.7 µM

IC50 values

CYP2B6


Am

od

iaq

uin

e N

-dea

lkyl

atio

n

(Per

cent

con

trol

)

0

20

40

60

80

100

120

>100 µM

>100 µM

60 µM

49 µM

IC50 values

CYP2C8


Dic

lofe

nac

4´-

hyd

roxy

lati

on

(P

erce

nt c

ontr

ol)

0

20

40

60

80

100

120

>100 µM

>100 µM

>100 µM

>100 µM

IC50 values

CYP2C9


S-M

ephe

nyt

oin

4´-

hyd

roxy

latio

n

(Per

cent

con

trol

)

0

20

40

60

80

100

120

>100 µM

>100 µM

27 µM

8.6 µM

IC50 values

CYP2C19


Dex

tro

met

ho

rph

an O

-dem

eth

ylat

ion

(Per

cent

con

trol

)

0

20

40

60

80

100

120

>100 µM

>100 µM

7.0 µM

13 µM

IC50 values

CYP2D6

Milnacipran zero preincubation

Milnacipran 30-min preincubation

Duloxetine zero preincubation

Duloxetine 30-min preincubation

Figure 2

This article has not been copyedited and form



DM

D Fast Forw


I: 10.1124/dmd.109.028274



Milnacipran Duloxetine

Milnacipran zero preincubation

Milnacipran 30-min preincubation

Duloxetine zero preincubation

Duloxetine 30-min preincubation

[Milnacipran concentration] (µM)0.1 1 10 100 1000

Tes

tost

ero

ne

6 β-h

ydro

xyla

tio

n

(Per

cent

con

trol

)

0

20

40

60

80

100

120

>100 µM

>100 µM

IC50 values

CYP3A4/5: Testosterone

[Duloxetine concentration] (µM)0.1 1 10 100 1000

Tes

tost

eron

e 6 β

-hyd

roxy

latio

n (P

erce

nt c

ontr

ol)

0

20

40

60

80

100

120

38 µM

21 µM

IC50 values

CYP3A4/5: Testosterone

[Milnacipran concentration] (µM)0.1 1 10 100 1000

Mid

azo

lam

1´-

hydr

oxy

latio

n (P

erce

nt c

ontr

ol)

0

20

40

60

80

100

120

28 µM

24 µM

IC50 values

CYP3A4/5: Midazolam

[Duloxetine concentration] (µM)0.1 1 10 100 1000

Mid

azo

lam

1´-

hyd

roxy

latio

n

(Per

cent

con

trol

)

0

20

40

60

80

100

120

37 µM

26 µM

IC50 values

CYP3A4/5: Midazolam

Figure 3

This article has not been copyedited and form



DM

D Fast Forw


I: 10.1124/dmd.109.028274




at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from



at ASPE

T Journals on M

arch 16, 2021dm


ownloaded from


Documents

In vitro inhibition and induction of human liver ......Dec 31, 2008 · water for CYP inhibition studies and in DMSO for enzyme induction studies. The following reagents were purchased