CHIRALITY 484-90 (1992)
The Disposition of Venlafaxine Enantiomers in Dogs, Rats, and Humans Receiving Venlafaxine C. PAUL WANG, STANLEY R. HOWELL, J O A " SCATINA, AND SAMUEL F.S I S E " E Drug Metabohm Lhiion, Wyeth-AyerstResearch, Princeton, New Jersey
ABSTRACT A stereospecific high-performance liquid chromatographic (HPLC) method was developed for the quantitation of the enantiomers of venlafaxine, an antidepressant, in dog, rat, and human plasma. The procedure involves derivatization of venlafaxine with the chiral reagent, ( + )-S-naproxenchloride, and a postderivatization procedure. The method was linear in the range of 50 to 5,000 ng of each enantiomer per ml of plasma. No interference by endogenous substances or known metabolites of venlafaxine occurred. Studies to characterize the disposition of the enantiomers of venlafaxine were conducted in dog, rat, and human, area under the curve (AUC)and (S)/(R) followingoral administration of venlafaxine. The Gax, concentration ratios of the (R)- and (S)-enantiomerswere compared. In rats, the mean plasma ratio of (S)-venlafaxineto that of (R)-venlafaxineover 0.5 to 6.0 h varied from 2.97 to 8.50 with ,AUC&, ,and tlh values of the (R)-and (S)-enantiomers a mean value of 5.51 f 2.45. The G,, in dogs were not significantly different from one another (P> 0.1). The mean ratios [(S)/(R)]of enantiomers of venlafaxine in human over a 2 to 6 h interval ranged from 1.33to 1.35 with an overall ratio of 1.34 f 0.26 (n= 12). These ratios of the enantiomers [(S)/(R)]were not statistically different from unity (P> 0.1) indicating that the disposition of venlafaxine enantiomers in humans is not stereoselective and is more similar to that in dogs than that in rats. KEY WORDS: high-performanceliquid chromatography (HPLC),chiral assay, naproxen chloride, pharmacokinetics, racemate, stereoselective INTRODUCTION
methylamino)-l-(4-methoxyphenyl)ethyllcyl], (S)Venlafaxine, 1- [2 - (dimethylamino)- 1- (4- methoxyphenyl)- venlafaxine [wy-45,655,(S)-l-[2-(dimethylamino)-l-(4-methoxyethyllcyclohexanol(Fig. 1)is currently under clinical investiga- phenyl)ethyllcyclohexanol] and metabolites of venlafaxine tion as an antidepressant. The compound is a racemate, and [wy-45,233 (4-[2-dimethylamino)-1-(1-hydroxycyclohexyl)ethethboth of its enantiomers, (R)-venlafaxine and (S)-venlafaxine, yllphenol), Wy-45,494 (1-[2-(methylamino)-l-(4-methoxyphenhave been isolated and found to be biologically active.' Each yl)ethyl]cyclohexanol), Wy-45,821 (1-[2-(amino-l-(4-methoxyresembles tricyclic antidepressants in its inhibition of [3H]imi- phenyl)ethyl]cyclohexanol), Wy-46,689 (4-[2-(methylamino)-lpramine binding and ability to inhibit norepinephrine and (l-hydroxycyclohexyl)ethyllphenol), Wy-46,965 (4-[2-amino-lserotonin synaptosomal uptake. During in vivo screening, (1-hydroxycyclohexyl)ethyllphenol), Wy-47,877 (1-[2-(dimevenlafaxine was found to antagonize reserpine-induced hypo- thylamino)-l-(4-methoxyphenyl)ethyl]-cis-1,4-cyclohexanediol), thermia, potentiate L-DOPA response, and antagonize hista- Wy-47,894 (1-[2-(dimethylamino)-l-(4-methoxyphenyl)ethyl]mine-induced ACTH release. This pharmacological profile is truns-1,4-cyclohexanediol),Wy-47,961(5-[4-(methoxyphenyl)3(Fig. 1) were similar to that of desipramine, a classic antidepressant. How- methyl-l-oxa-3-azaspir~5,5]undecan-9-oll)] obtained from D r . Morris Husbands, Wyeth-Ayerst Research, ever, venlafaxine appears to have a more rapid onset of action with favorable side-effect profile. Because of the potential Princeton, NJ. Sodium borate-10-hydrate(crystal), chloroform differences in the metabolic disposition and pharmacokinetics (hydrocarbon stabilized, Baker Analyzed HPLC Reagent), and of the enantiomers in laboratory animals and man, the mea- aluminum oxide (neutral, Brockmann Activity Grade 1)were surement of venlafaxine enantiomers in biological specimens is purchased from J. T. Baker, while 0.1 N hydrochloric acid necessary. The purpose of the present study was to develop an (Certified Reagent), potassium phosphate (monobasic, HPLC HPLC method for the quantitation of venlafaxine enantiomers grade), O-phosphoric acid (85%, HPLC grade), and triethylain biological fluid and to investigate the disposition of ven- mine (HPLC reagent grade) were obtained from Fisher Scientific (Pittsburgh, PA). Isopropyl ether was purchased from lafaxine's enantiomers in human, rat, and dog. F l u b Chemical (Ronkonkoma, NY).Isopropyl ether was purified before each analysis by passing it through neutral aluMATERIALS AND METHODS mina. Purity was checked by HPLC as described in the HPLC Chemicals and Reagents Venlafaxine, as the hydrochloride salt, was obtained from Wyeth-Ayerst Laboratories, West Chester, PA. The internal for publication July 22, 1991; accepted October 8, 1991. standard (Wy-45,818, l-[2-(dimethylamino)-l-(2-chlorophenyl) Received Address reprint requests to C. Paul Wang, Drug Metabolism Division, Wyethethyllcyclohexanol), (R)-venlafaxine my-45,651, (R)-l-[2-(di- Ayerst Research, CN 8000, Princeton, NJ 08.543. ly2
0
1992 Wiley-Liss, Inc.
DISPOSITION OF VENLAFAXINE ENANTIOMERS
85
g oxalyl chloride at ambient temperature for 22 h. Following
Compound
R3 R4 R5 -
Venlafaxine
OCH3
H
H
WY-45,233
OH
H
H
WY-45,818
H
CI
H
WY-45,821
OCH3
H
H
WY-45,494
OCH3
H
H
WY-46,689
OH
H
H
WY-47,877
OCH3
H
OH
WY-47.894
OCH3
H
OH
WY-46.965
OH
H
H
* Chiral Center WY-47.961
Fig. 1. Chemical structures of venlafaxine, venlafaxine metabolites, and the internal standard (Wy-45,818).
analysis section before use. Sodium carbonate, anhydrous, AR was obtained from Mallinckrodt (Paris,KY). Naproxen [( + )-6methoxy-a-methyl-2-naphthalene acetic acid, 98%], and oxalyl chloride (99+ %) were purchased from Aldrich (Milwaukee, WI).Methylene chloride was purchased from EM Science (Gibbstown, NJ). Acetonitrile, UV, was obtained from Baxter (McGaw Park, IL). Deionized water was prepared by passing tap water through a Millipore Milli-Q reagent water system. Instrumentation
The HPLC system consisted of two Waters Model 6OOOA pumps (Waters Associates, Milford, MA), a Waters Model 720 HPLC system controller, a Waters Model M730 Data Module, a Waters WISP Model 712 autosampler,a Waters Lambda-max model 481 LC spectrophotometerset at 229 nm, and a Hewlett Packard model 339OA Integrator (Hewlett-Packard,Avondale, PA). Separation of the diastereomers was achieved using a Supelcosil LCS-DB column (150 x 4.6 mm i.d., 5 pm particle size) purchased from Supelco (Bellefonte, PA) and Supelguard LC-8-DBprecolumn with 2 cm x 4.6 mm Supelcosil cartridge. Characterization of the naproxen diastereomers of venlafaxine and internal standard Wy-45,818 was done by a Thermospray/ LC/MS system incorporating a Waters 600multisolvent delivery system, a Waters 490 MS programmable multiwavelength detector and a Finnigan 4600 TSQ with Vestec 701C Thermospray Interface mass spectrometer. The LC/MS analysis was performed by Oneida Research Services, Inc. (Whitesboro,NY). Preparation of Naproxen Chloride Naproxen was activated to the corresponding acyl chloride by reacting 8.1 g naproxen in 40 ml methylene chloride with 25
the evaporation of methylene chloride and the excess oxalyl chloride using a rotary evaporator at 42"C, the dry powder was dissolved in 35 ml of methylene chloride. n-Hexane (1120 ml) was then added slowly to the methylene chloride solution until the solution became slightly cloudy. The solution was then chilled over dry ice to allow crystal growth. Solvent was decanted carefully and rapidly, then nitrogen gas was applied immediately to dry the crystals. The needles were dried further over anhydrous calcium chloride in a vacuum dessicator. The melting point of the naproxen chloride was 9597°C. Biological Sampk Preparation
Four male beagle dogs, obtained from the animal colony of Wyeth-Ayerst Research, Princeton, NJ, were kept under observation for 3 days before dosing. On the first day of treatment, the body weights ranged from 8.30 to 17.05 kg. The dogs were allowed food and water ad libitum except for 16 h (overnight) prior to and 4 h after dosing on day 1 and day 14. The dogs received 22 mg venlafaxine as the hydrochloride salt/kg/day for 14 consecutive days. Gelatin macro capsules (Size No. 13, Torpac Ltd.) were hand-filled according to the weight of each dog. On day 14, blood samples were drawn into heparin-rinsed tubes from the cephalic vein at 0 (predose),0.5,1,2,4,6,8,12, 15,and 24 h after dosing. The plasma was separated and frozen at - 20°C until analysis. Sixty-fivemale albino SpragueDawley rats (CharlesRiver), weighing 188to 296 g, were subjected to standard 12 h nightday cycles and allowed food and water ad libitum, except for the last day of the study when the rats were fasted overnight prior to and 4 h after the last dose. The animals received a single ig dose of venlafaxine as the hydrochloride salt (120 mg of the free base/kg) daily for 14 consecutive days. Venlafaxine hydrochloride was dissolved in sterile water. On day 14, the rats were slightly anesthetized with a mixture of O2/C02(95:5) and halothane gas in groups of five at 0 (predose),0.5, 1,1.5,2, 4, 6, 8, 12, 15, 24, 32, and 48 h after dosing. As much blood as possible was collected from the descending aorta into heparinrinsed syringes. The plasma was separated and frozen at - 20°C until analysis. After giving their informed consent, nine male subjects received a single 50 mg capsule dose of venlafaxine, as the hydrochloride salt, after a 10 h overnight fast. Blood samples were obtained at pre-dose, 0.33, 0.66, 1, 1.5,2, 3, 4,6, 8, 10, 12, 16, 24 and 32 h after dosing. The plasma was separated and frozen prior to analysis. One ml, or the available volume, of plasma collected from 9 subjects at 2,3,4 and 6 h after dosing were pooled and analyzed. Plasmas were pooled due to low concentration of total racemic venlafaxine in the specimens and limited sensitivity of the assay. Extraction and Lkriuatization of Biological Specimens
A stock solution of venlafaxine was prepared by dissolving venlafaxine hydrochloride in water to a concentration of 1 mg/ml of the free base. Standard solutions were made by diluting the stock solution with deionized water to obtain venlafaxine concentrations of 200,100, 50, 20, 10, 4.0, and 2.0 pg/ml. The stock internal standard solution was prepared by dissolving Wy-45,818 as the hydrochloride salt in water. Fifty microliters of internal standard aqueous solution (Wy-45,818)
86
WANG ET AL.
at a concentration of 22.5 pg/ml and 0.2 ml of saturated sodium borate solution were added to 1.0 ml of sample (spiked control plasma or plasma from experimental animals) in a 16 x 125 mm disposable screw-cap culture tube and mixed by vortex. Five milliliters of purified isopropyl ether was added and the tubes were shaken for 15rnin and then centrifuged at 2,500 rpm for 10 min. A 4.5 ml aliquot of the isopropyl ether phase was transferred to a clean conical screw-captube containing 0.4 ml of 0.01 N hydrochloric acid. The tubes were shaken for 15 rnin and then centrifuged at 2,500 rpm for 10 min. The ether phase was aspirated and discarded. One milliliter of saturated sodium borate solution was added to the aqueous phase, and the tubes were vortexed. Following the addition of 2.0 ml of purified isopropyl ether, the tubes were shaken for 15 rnin and then centrifuged at 2,500 rpm for 5 min. The organic phase was transferred to a clean conical screw-captube and evaporated to dryness under nitrogen. Five to 10 mg of anhydrous sodium carbonate powder and then 0.2 ml of naproxen chloride solution (15 mg per 100 ml methylene chloride) were added to each tube and vortexed. The solutions were covered with aluminum foil and allowed to stand at room temperature to react for 20 h. The methylene chloride was then evaporated to dryness under a mild stream of nitrogen gas. To the residue, 1ml of deionized water was added and vortexed to completely dissolve the sodium carbonate. Two milliliters of chloroform was then added and the tubes were shaken for 10 min and centrifuged at 2,500 rpm for 5 min. The chloroformphase was transferred to a clean screw-cap conical tube and washed with 2 ml deionized water by shaking for 10 rnin and centrifuged at 2,500 rpm for 5 min. The final chloroform extract was then transferred to a clean mew-cap conical tube and evaporated to dryness under nitrogen gas. The diastereomeric esters were separated by HPLC. HPLC Analysis
of mobile phase A/B (50:50) and 11 pl was injected onto the HPLC column. The column was maintained at ambient temperature and the detection wavelength was 229 nm. Enantiomer concentrations in ng (free base)/ml were obtained from the regression line relating compound/internal standard (the derivative peak of the enantiomer of Wy-45,818 with longer retention time) peak area ratios to the compound concentrations calculated using the program “REG 1SD” in a Hewlett-Packard model 85 calculator. The racemate of venlafaxine was assumed to have an enantiomer ratio of R:S = 1:1 due to the lack of optical rotation. The specificity of the method was evaluated with regard to interference due to the presence of endogenous substances in the extracts of dog plasma and for interference from known metabolites. Three random control plasmas were processed by the method described and one control plasma spiked with solutions of known venlafaxine metabolites was analyzed. The retention times of the diastereomers of metabolites were evaluated with respect to the retention times of the diastereomers of venlafaxine and Wy-45,818. The amount of each metabolite added was Wy-45,233 (800 ng), Wy-45,030 (800 ng), Wy-45,818 (800 ng), Wy-4!5,494(800ng),Wy-47,877(800ng and 5,000 ng), Wy-47,894 (800 ng and 5,000 ng), Wy-45,821 (800 ng), Wy46,689 (800ng), Wy-46,965 (800 ng and 5,000 ng), and Wy47,961 (800 ng and 5,000 ng) per ml of dog plasma. The intraday precision of the method was assessed by calculating the coefficient of variation for replicate samples (n= 5) at a concentration of 300 ng of each enantiomer/ml. Identification of the Individual Enantiomers
To distinguish the respective chromatographic peaks of (R)-venlafaxine hydrochloride and (S)-venlafaxine hydrochloride, each was extracted, derivatized, and analyzed by HPLC individually.
The mobile phase was composed of solution A (acetonitrile) In Vivo Study of the Chiral Inversion o f the (R)- and and solution B, which was prepared by adjusting the pH of 0.1 (S)-Enantiomers of Venlafaxine in Rat M KH2P04 solution to 3.0 with 85% 0-phosphoric acid and Nineteen male Sprague-Dawley rats (Charles River), weighthen adding 0.07% (v/v) triethylamine to achieve a final pH of ing 187-217 g, were fasted overnight prior to dosing. The 3.25. An HPLC gradient profile was used to analyze the sample animals received a single 60 mg/kg ig dose of either (R)- or (Table 1).The derivatized extract was reconstituted in 0.4 ml (S)-venlafaxineas the hydrochloride salt. The rats were slightly anesthetized with a mixture of 02/C02 (95:5) and halothane gas in groups of three at 0.5,1.5, and 2.5 h after dosing. One rat, TABLE 1. Gradient program for the analysis of which was used to collect control blood, was not dosed. As venlafaxine enantiomersO much blood as possible was collected by cardiac puncture into Mobile phase (%) the heparinized tubes. The plasma was separated and frozen at Time Flow rate - 20°C until analysis. The individual venlafaxine enantiomer Ab Bc (min) (ml/min) concentrations were determined by HPLC. The ratios of the enantiomer concentrations of venlafaxine in the plasma samInitial 50 50 0.8 ples were calculated. 30 54 46 0.8 36 50 53 56
2.0 2.0 2.0 0.8
55 57 57 50
45 43 43 50
‘Linear gradient was used with all time intervals. bMobilephase A acetonitrile. ‘Mobile phase B 0.1 M KH@, (pH 4.42) adjusted to pH 3.0 with phosphoric acid followed by the addition of 0.07% (v/v) triethylamine to pH 3.25.
Calculation o f Pharmacokinetic Parameters
Pharmacokinetic parameters were calculated from plasma concentration data by standard noncompartmental methods. Zero-to-t h areas under the plasma concentration-time curve (AUCst) were calculated by the trapezoidal rule. Peak plasma and time to peak (&.J were obtained by concentration (C,,,=) data inspection. Terminal elimination half-lives (hh)were cal-
87
DISPOSITION OF VENLAFAXINE ENANTIOMERS
culated as tlA = In 2/&, where I(el was the slope of regression line of the terminal log-linear data points. AUCO_~ was obtained as AUCo - + CJ&.
Identification of the Individuul Enantiomers of Venlafaxine
(R)-Venlafaxine, when analyzed individually, appeared in the chromatogram as a single peak at 16.73min, while (S)-venStatistics lafaxine, when analyzed alone, appeared in the chromatogram A paired t- test was used for statistical comparison of the as a major peak at 19.34 min. This demonstrated that the peak G,,, h ~AUC&, , , or AUCG15 values of (R)- and 6)-venlafax- of venlafaxine-naproxen diastereomer with shorter retention ine in dog. The degree of significance was set at 01 = 0.05. A time is derived from (R)-venlafaxine,while the one with longer single factor, repeated measures analysis of variance, was used retention is derived from (S)-venlafaxine. to analyze venlafaxine enantiomer ratios in human. m L C Analysis RESULTS Typical HPLC chromatograms for control dog plasma, dog LCI Mass Spectrometric Characterization of the plasma spiked with venlafaxine and Wy-45,818, and plasma Naproxen Derivatives of Venlafaxine and Wy45,818 samples from a dog, a rat, and human given venlafaxine are The molecular weights of the naproxyl derivatives of (R)- presented in Figure 2. Under the conditions specified,the retenand (S)-venlafaxineand Wy-45,818 were 489 amu and 493 amu, tion times for the derivatives of (R)- and (S)-venlafaxineand respectively. The molecular weights confirmed the formation of Wy-45,818 enantiomerswere 16.53,18.03,19.74, and 22.38 min, diastereomers which are the esterification product of the terti- respectively. No endogenous interfering substances in the exary hydroxyl group of venlafaxine and Wy-45,818 and acyl tracts of three random control dog plasma samples were detected. The chromatogram of the extract of venlafaxine, Wygroup of naproxen chloride.
A
2 3
D 10 20 30 4 Retention Time (IT
I
10
20 30 40 50 Retention Time (min)
I
I
10
I
I
20 30 40 50 Retention Time (min)
3
E
C
10
L
20 30 40 Retention Time (min)
50
1
10
20 30 40 Retention Time (min)
50
Fig. 2. Chromatograms of control dog plasma (A), control dog plasma spiked with venlafaxine and Wy-45,818(B), and plasma samples from a dog (C), a rat (D), and human given venlafaxine (E). 1,(R)-venlafaxine;2, (S)-venlafaxine; 3, (2)-Wy-45,818(the enantiomer of Wy-45,818 with longer retention); 4, (R)-Wy-45,233;5, (S)-Wy-45,233.
88
WANG ET AL.
-
10.0 45,818, and known metabolites reacted with chiral reagent ( + )-(S)-naproxenchloride is shown in Figure 3. The retention -0- (S)-Venlafaxine times for the diastereomers of Wy-45,821,(R)- and (Shenlafaxine, (1)and (2) Wy-45,818, Wy-45,494, (R)- and (S)-Wy-45,233 and Wy-46,689 were 14.58, 16.30, and 17.79, 19.47 and 22.10, 20.01, 43.56, and 45.40, and 43.56 min, respectively. The dias- C tereomers of Wy-46,689 coeluted with the derivative of (R)-Wy- .-0 45,233. No peaks of the derivatives of Wy-47,877, Wy-47,894, 2 C Wy-47,961, and Wy-46,965, which are metabolites, can be de- 8 .1 tected in this chromatographic system even at concentrations 0 of 5,000 ng/ml. The linearity of the peak area ratio to enantiomer concentration was calculated using linear regression analysis. The 0 2 4 6 8 10 12 14 16 method was linear for concentrations of each enantiomer of Time (hr) between 50 and 5,000 ng/ml in dog plasma with correlation coefficients of 0.9999 for both (R)- and (S)-enantiomers. The Fig. 4. Mean plasma concentrationsof (S)-venlafaxine,(R)-venlafaxine,and intraday precision of the method was assessed by calculating ( =k)-venlafaxine in dog after daily administration of 22 mg/kg venlafaxine for the coefficient of variation for replicate samples (n= 5) at 300 14 consecutive days. Values are the mean of four subjects. ng/ml for each enantiomer of venlafaxine in dog plasma. The % CV for (R)- and (S)-venlafaxinewere 3.1 and 0.6, respectively. value of 1.51 f 0.11. The mean f SD LX values of (R)-and (S)-venlafaxinewere 1.78 f 0.21 and 2.47 f 1.00 pg/ml, reIn Vivo Study of the Chiral Inversion of the (R)- and spectively. The & was 1.13 f 0.63 h for both enantiomers. (S)-Enantiomersof Venbfaxine in Rat The AUC@, and tlh of (R)- and (S)-venlafaxinewere 10.70 f No (S)-venlafaxine appeared in either the (R)-venlafaxine 2.33 and 15.77 f 7.74 pgh/ml, and 4.10 f 0.57 and 4.08 f dose or in the plasma of rats receiving (R)-venlafaxine,indicat- 0.59 h, respectively. There was no significant difference in the ing that chiral inversion of (R)-venlafaxine did not occur. A C;nax, hh, AUC&, ,or AUCs15 values for (R)-and (S)-venlafaxsmall (R)-venlafaxinepeak was detected in the (S)-venlafaxine ine (P>O.l), which indicates an absence of a significant stedosage form and as well as in the plasma of rats receiving reoselective disposition of venlafaxine in dogs. (S)-venlafaxine.The mean f SD of the (R)/(S)ratio in the (S)-venlafaxinedosing material was 0.035 f 0.012. The mean Analysis of Rat Plusma (R)/(S)ratio in the plasma of rats receiving (S)-venlafaxine The pharmacokinetic data for (R)-venlafaxineand (S)-venranged from 0.028 to 0.030 over the 0.5 to 2.5 h interval. Comlafaxine in rat are presented in Table 3. The mean plasma parison of the (R)/(S) ratios in the dosing material and plasma concentrations of total racemic, (S)-and (Rkvenlafaxine are of the rats receiving (S)-venlafaxineindicates that no formation depicted in Figure 5. The ratios of the mean plasma concentraof (R)-venlafaxineoccurred in vivo. tions of (S)-venlafaxineto that of (R)-venlafaxineover the 0.5 to 6.0 h interval varied from 2.97 to 8.50 with a mean value of 5.51 Analysis of Dog Plasma f 2.45. The GX of (R)- and (S)-venlafaxinewere 0.88 f 0.59 Mean concentrations of (R)-venlafaxineand (S)-venlafaxine in dog plasma are depicted in Figure 4 and the pharmacokinetic data are presented in Table 2. The ratios of the mean plasma TABLE 2. Pharmacokinetics of venlafaxine enantiomers concentrationsof (S)-venlafaxineto that of (R)-venlafaxineover in the plasma of dogs receiving venlafaxine (22 the 0.5 to 15 h interval varied from 1.35 to 1.68. with a mean mg/kg/day) for 14 consecutive days
c c
-
~~
R
1
S
10
20
30
h
40
Retention Time (min)
50
Fig. 3. Chromatogram of control dog plasma spiked with venlafaxine, Wy45,818(internal standard), and known metabolites of venlafaxine: 1,Wy-45,821; 2, (R)-venlafaxine;3, (S)-venlafaxine;4, (l)-Wy45,818; 5, Wy-45,494; 6, (2tWy45,818 (internal standard); 7, (R)-Wy-45,233; 8, Wy-46.689; 9, (S)-Wy-45,233.
2
R
3
R
S
S
R
4
S
Mean f
Mean f
R
SD SD
Total f SD
S
2.09 3.96 1.62 1.96 1.69 2.06 1.73 1.87 1.78 0.21 2.47 1.00 4.25 1.21
0.5 0.5 2.0 2.0 1.0 1.0 1.0 1.0 1.13 0.63 1.13 0.63 1.13 0.63
4.71 4.23 4.20 4.50 4.13 4.38 3.34 3.21 4.10 0.57 4.08 0.59 3.97 0.58
11.63 23.79 10.06 13.04 10.54 13.17 7.07 7.30 9.82 1.95 14.32 6.88 24.15 8.64
12.80 26.25 11.08 14.63 11.54 14.62 7.38 7.58 10.70 2.33 15.77
7.74 26.84 10.00
89
DISPOSITION OF VENLAFAXINE ENANTIOMERS
TABLE 3. Pharmacokinetics of venlafaxine enantiomers in the plasma of rats after administrationof racemic venlafaxine (120 mg/kg/day) for 14 consecutive days Compound ~
~~
“Mean
0.5 0.5 0.5
0.88 f 0.59 2.61 f 1.15 3.49 f 1.72
(R)-Venlafaxine (S)-Venlafaxine Total venlafaxine
1.27 5.56 6.77
1.21 5.20
2.00 1.69 1.50
-
* SD.
and 2.61 f 1.15 pg/ml, respectively. The corresponding & values were 0.5 and 0.5 h, respectively. The AUC&, and tlh values of (R)- and (S)-venlafaxinewere 1.27 and 5.56 pgh/ml and 2.00 and 1.69 h, respectively, indicating that the disposition of venlafaxine in rats was stereoselective. Analysis of Human Plasma
The plasma ratios of the (R) and (S)enantiomersof venlafaxine in pooled samples are given in Table 4. The mean ratios [(S)/(R)]of the enantiomers of venlafaxine over a 2 to 6 h interval ranged from 1.33 to 1.35 with an overall ratio of 1.34 f 0.26 (n= 12). The plasma ratios of the enantiomers [(S)/(R)] were not significantly different from unity at each time point and overall ( P L 0.14), indicating no stereoselective disposition of venlafaxine enantiomers in man. The plasma [(S)/(R)] ratios of the enantiomers were not significantly different among time points (P= 0.97). DISCUSSION
An HF’LC method has been developed using naproxen chloride as the chiral reagent for the quantitation of venlafaxine enantiomers in dog, rat, and human plasma. The method is specific, precise, and linear at least in the range of 50 to 5,000 ng/ml. The minimum quantifiableconcentration was 50 ng/ml. The detection limit could be lowered to 25 ng/ml by increasing the injection volume, since at 22 p1 injection volume no interfering peaks were found in the control plasma. Spahn et al.495used naproxen chloride synthesizedfrom naproxen and thionyl chlo-
10.0
-
ride to quantify primary and secondary amines and alcohols. In this study, the derivatization condition has been modified and a postderivatizationprocess has been developed to separate the enantiomersof venlafaxine, a tertiary alcohol chiral compound. This method is applicable to the chiral assay of other compounds with chiral tertiary alcohol and a phenolic hydroxy functional group as demonstrated in the derivatization of the major human metabolite of venlafaxine, Wy-45,233. This method is practical for the routine chiral analysis of venlafaxine enantiomers in biological specimens, including plasma and urine. The method can be modified by using fluorescencedetection and/or a normal phase HPLC system. The pharmacokinetic disposition of venlafaxine in rats receiving multiple oral doses of drug has been demonstrated to be stereoselective.The ratio of the mean plasma concentrations of (S)-venlafaxineto that of (R)-venlafaxineover the 0.5 to 6.0 h interval varied from 2.97 to 8.50 with a mean value of 5.51 f 2.45. Because no in vivo chiral inversion of venlafaxine enantiomers in rat was found, the stereoselectivedisposition of (R)- and (S)-venlafaxinein rat can be assumed to arise from differences in drug absorption, distribution, metabolism, or excretion of the individual enantiomers. However, the elimination half-lives of both enantiomers were similar. The disposition of venlafaxine in dog receiving multiple oral doses of venlafaxine was not stereoselective. The ratio of the mean plasma concentration of (S)-venlafaxineto that of (R)-venlafaxine over the 0.5-15 h interval varied from 1.35 to 1.68,with a mean value of 1.51 f 0.11. The mean ratio of the plasma concentrationsof (S)-venlafaxine to that of (R)-venlafaxine over the 2-6 h time interval in human subjects receiving 50 mg oral doses of venlafaxine
-0- (S)-Venlafaxine -D-
E
(R)-Venlafaxine
TABLE 4. Ratios of (R)-and (S)-venlafaxinein pooled plasma of human subjects administered 50 mg of venlafaxine orally
\
8 E
1.0
Ratio of enantiomers (S)/(R)
I I
I
I
I
0
1
2
3 Time (hr)
.01
\ I
I
I
4
5
6
Fig. 5. Mean plasma concentrationsof (Stvenlafaxine,(R)-venlafaxine,and (*)-venlafaxine in rat after daily administration of 120 mg/kg venlafaxine for 14 consecutive days. Values are the mean of five replicates.
Pool No:
2 hb
3h
4h
6h
1 2 3
1.31 1.59 1.16
1.25 1.61 1.12
1.24 1.64 1.14
1.22 1.83 1.01
Mean f SD
1.35 0.22
1.33 0.25
1.34 0.27
1.35 0.42
OThree subjects per pool. *Hours after dosing.
90
WANG ET A L
varied from 1.33 to 1.35.The plasma ratios of the enantiomem [@/@)I were not significantlydifferentfrom unity at each time point and overall. It is concluded that the disposition of venlafaxine enantiomers in humans is not stereoselective and is similar to that in dogs. LITERATURE CITED 1. Muth, E., Haskins, J., Moyer, J. et al. Antidepressant biochemical profile of the novel bicyclic compound Wy-45,030,an ethyl cyclohexanol derivative. Biochem. Pharmacol. 54493-4497, 1986.
2. Fabre, L., Putman, €HI. I. An ascending single-dose tolerance study of w Y 4 , o a , a bicyclic a n t i d e P m t r in healthy men. cm. RE,. 42: 901-909,1987. 3, Gibaldi, M,, perrier, D. pharmacokinetics, 2nd ed. New York: Marcel &kker, 1982: 409417, ug. 4. Spahn, H., Kraup, D., Mutschler, E. Enantiospecific high-perfomance liquid chromatographic(HPLC)determination of baclofen and its fluoro analogue in biological material. Pharm. Res. 5107-112, 1988. 5. Spahn, H. S-(+)-Naproxen chloride as acylating agent for separating the enantiomers of chiral amines and alcohols. Arch. Pharm. (Weinheim) 321: 847-850,1988.
The Disposition of Venlafaxine Enantiomers in Dogs, Rats, and Humans Receiving Venlafaxine C. PAUL WANG, STANLEY R. HOWELL, J O A " SCATINA, AND SAMUEL F.S I S E " E Drug Metabohm Lhiion, Wyeth-AyerstResearch, Princeton, New Jersey
ABSTRACT A stereospecific high-performance liquid chromatographic (HPLC) method was developed for the quantitation of the enantiomers of venlafaxine, an antidepressant, in dog, rat, and human plasma. The procedure involves derivatization of venlafaxine with the chiral reagent, ( + )-S-naproxenchloride, and a postderivatization procedure. The method was linear in the range of 50 to 5,000 ng of each enantiomer per ml of plasma. No interference by endogenous substances or known metabolites of venlafaxine occurred. Studies to characterize the disposition of the enantiomers of venlafaxine were conducted in dog, rat, and human, area under the curve (AUC)and (S)/(R) followingoral administration of venlafaxine. The Gax, concentration ratios of the (R)- and (S)-enantiomerswere compared. In rats, the mean plasma ratio of (S)-venlafaxineto that of (R)-venlafaxineover 0.5 to 6.0 h varied from 2.97 to 8.50 with ,AUC&, ,and tlh values of the (R)-and (S)-enantiomers a mean value of 5.51 f 2.45. The G,, in dogs were not significantly different from one another (P> 0.1). The mean ratios [(S)/(R)]of enantiomers of venlafaxine in human over a 2 to 6 h interval ranged from 1.33to 1.35 with an overall ratio of 1.34 f 0.26 (n= 12). These ratios of the enantiomers [(S)/(R)]were not statistically different from unity (P> 0.1) indicating that the disposition of venlafaxine enantiomers in humans is not stereoselective and is more similar to that in dogs than that in rats. KEY WORDS: high-performanceliquid chromatography (HPLC),chiral assay, naproxen chloride, pharmacokinetics, racemate, stereoselective INTRODUCTION
methylamino)-l-(4-methoxyphenyl)ethyllcyl], (S)Venlafaxine, 1- [2 - (dimethylamino)- 1- (4- methoxyphenyl)- venlafaxine [wy-45,655,(S)-l-[2-(dimethylamino)-l-(4-methoxyethyllcyclohexanol(Fig. 1)is currently under clinical investiga- phenyl)ethyllcyclohexanol] and metabolites of venlafaxine tion as an antidepressant. The compound is a racemate, and [wy-45,233 (4-[2-dimethylamino)-1-(1-hydroxycyclohexyl)ethethboth of its enantiomers, (R)-venlafaxine and (S)-venlafaxine, yllphenol), Wy-45,494 (1-[2-(methylamino)-l-(4-methoxyphenhave been isolated and found to be biologically active.' Each yl)ethyl]cyclohexanol), Wy-45,821 (1-[2-(amino-l-(4-methoxyresembles tricyclic antidepressants in its inhibition of [3H]imi- phenyl)ethyl]cyclohexanol), Wy-46,689 (4-[2-(methylamino)-lpramine binding and ability to inhibit norepinephrine and (l-hydroxycyclohexyl)ethyllphenol), Wy-46,965 (4-[2-amino-lserotonin synaptosomal uptake. During in vivo screening, (1-hydroxycyclohexyl)ethyllphenol), Wy-47,877 (1-[2-(dimevenlafaxine was found to antagonize reserpine-induced hypo- thylamino)-l-(4-methoxyphenyl)ethyl]-cis-1,4-cyclohexanediol), thermia, potentiate L-DOPA response, and antagonize hista- Wy-47,894 (1-[2-(dimethylamino)-l-(4-methoxyphenyl)ethyl]mine-induced ACTH release. This pharmacological profile is truns-1,4-cyclohexanediol),Wy-47,961(5-[4-(methoxyphenyl)3(Fig. 1) were similar to that of desipramine, a classic antidepressant. How- methyl-l-oxa-3-azaspir~5,5]undecan-9-oll)] obtained from D r . Morris Husbands, Wyeth-Ayerst Research, ever, venlafaxine appears to have a more rapid onset of action with favorable side-effect profile. Because of the potential Princeton, NJ. Sodium borate-10-hydrate(crystal), chloroform differences in the metabolic disposition and pharmacokinetics (hydrocarbon stabilized, Baker Analyzed HPLC Reagent), and of the enantiomers in laboratory animals and man, the mea- aluminum oxide (neutral, Brockmann Activity Grade 1)were surement of venlafaxine enantiomers in biological specimens is purchased from J. T. Baker, while 0.1 N hydrochloric acid necessary. The purpose of the present study was to develop an (Certified Reagent), potassium phosphate (monobasic, HPLC HPLC method for the quantitation of venlafaxine enantiomers grade), O-phosphoric acid (85%, HPLC grade), and triethylain biological fluid and to investigate the disposition of ven- mine (HPLC reagent grade) were obtained from Fisher Scientific (Pittsburgh, PA). Isopropyl ether was purchased from lafaxine's enantiomers in human, rat, and dog. F l u b Chemical (Ronkonkoma, NY).Isopropyl ether was purified before each analysis by passing it through neutral aluMATERIALS AND METHODS mina. Purity was checked by HPLC as described in the HPLC Chemicals and Reagents Venlafaxine, as the hydrochloride salt, was obtained from Wyeth-Ayerst Laboratories, West Chester, PA. The internal for publication July 22, 1991; accepted October 8, 1991. standard (Wy-45,818, l-[2-(dimethylamino)-l-(2-chlorophenyl) Received Address reprint requests to C. Paul Wang, Drug Metabolism Division, Wyethethyllcyclohexanol), (R)-venlafaxine my-45,651, (R)-l-[2-(di- Ayerst Research, CN 8000, Princeton, NJ 08.543. ly2
0
1992 Wiley-Liss, Inc.
DISPOSITION OF VENLAFAXINE ENANTIOMERS
85
g oxalyl chloride at ambient temperature for 22 h. Following
Compound
R3 R4 R5 -
Venlafaxine
OCH3
H
H
WY-45,233
OH
H
H
WY-45,818
H
CI
H
WY-45,821
OCH3
H
H
WY-45,494
OCH3
H
H
WY-46,689
OH
H
H
WY-47,877
OCH3
H
OH
WY-47.894
OCH3
H
OH
WY-46.965
OH
H
H
* Chiral Center WY-47.961
Fig. 1. Chemical structures of venlafaxine, venlafaxine metabolites, and the internal standard (Wy-45,818).
analysis section before use. Sodium carbonate, anhydrous, AR was obtained from Mallinckrodt (Paris,KY). Naproxen [( + )-6methoxy-a-methyl-2-naphthalene acetic acid, 98%], and oxalyl chloride (99+ %) were purchased from Aldrich (Milwaukee, WI).Methylene chloride was purchased from EM Science (Gibbstown, NJ). Acetonitrile, UV, was obtained from Baxter (McGaw Park, IL). Deionized water was prepared by passing tap water through a Millipore Milli-Q reagent water system. Instrumentation
The HPLC system consisted of two Waters Model 6OOOA pumps (Waters Associates, Milford, MA), a Waters Model 720 HPLC system controller, a Waters Model M730 Data Module, a Waters WISP Model 712 autosampler,a Waters Lambda-max model 481 LC spectrophotometerset at 229 nm, and a Hewlett Packard model 339OA Integrator (Hewlett-Packard,Avondale, PA). Separation of the diastereomers was achieved using a Supelcosil LCS-DB column (150 x 4.6 mm i.d., 5 pm particle size) purchased from Supelco (Bellefonte, PA) and Supelguard LC-8-DBprecolumn with 2 cm x 4.6 mm Supelcosil cartridge. Characterization of the naproxen diastereomers of venlafaxine and internal standard Wy-45,818 was done by a Thermospray/ LC/MS system incorporating a Waters 600multisolvent delivery system, a Waters 490 MS programmable multiwavelength detector and a Finnigan 4600 TSQ with Vestec 701C Thermospray Interface mass spectrometer. The LC/MS analysis was performed by Oneida Research Services, Inc. (Whitesboro,NY). Preparation of Naproxen Chloride Naproxen was activated to the corresponding acyl chloride by reacting 8.1 g naproxen in 40 ml methylene chloride with 25
the evaporation of methylene chloride and the excess oxalyl chloride using a rotary evaporator at 42"C, the dry powder was dissolved in 35 ml of methylene chloride. n-Hexane (1120 ml) was then added slowly to the methylene chloride solution until the solution became slightly cloudy. The solution was then chilled over dry ice to allow crystal growth. Solvent was decanted carefully and rapidly, then nitrogen gas was applied immediately to dry the crystals. The needles were dried further over anhydrous calcium chloride in a vacuum dessicator. The melting point of the naproxen chloride was 9597°C. Biological Sampk Preparation
Four male beagle dogs, obtained from the animal colony of Wyeth-Ayerst Research, Princeton, NJ, were kept under observation for 3 days before dosing. On the first day of treatment, the body weights ranged from 8.30 to 17.05 kg. The dogs were allowed food and water ad libitum except for 16 h (overnight) prior to and 4 h after dosing on day 1 and day 14. The dogs received 22 mg venlafaxine as the hydrochloride salt/kg/day for 14 consecutive days. Gelatin macro capsules (Size No. 13, Torpac Ltd.) were hand-filled according to the weight of each dog. On day 14, blood samples were drawn into heparin-rinsed tubes from the cephalic vein at 0 (predose),0.5,1,2,4,6,8,12, 15,and 24 h after dosing. The plasma was separated and frozen at - 20°C until analysis. Sixty-fivemale albino SpragueDawley rats (CharlesRiver), weighing 188to 296 g, were subjected to standard 12 h nightday cycles and allowed food and water ad libitum, except for the last day of the study when the rats were fasted overnight prior to and 4 h after the last dose. The animals received a single ig dose of venlafaxine as the hydrochloride salt (120 mg of the free base/kg) daily for 14 consecutive days. Venlafaxine hydrochloride was dissolved in sterile water. On day 14, the rats were slightly anesthetized with a mixture of O2/C02(95:5) and halothane gas in groups of five at 0 (predose),0.5, 1,1.5,2, 4, 6, 8, 12, 15, 24, 32, and 48 h after dosing. As much blood as possible was collected from the descending aorta into heparinrinsed syringes. The plasma was separated and frozen at - 20°C until analysis. After giving their informed consent, nine male subjects received a single 50 mg capsule dose of venlafaxine, as the hydrochloride salt, after a 10 h overnight fast. Blood samples were obtained at pre-dose, 0.33, 0.66, 1, 1.5,2, 3, 4,6, 8, 10, 12, 16, 24 and 32 h after dosing. The plasma was separated and frozen prior to analysis. One ml, or the available volume, of plasma collected from 9 subjects at 2,3,4 and 6 h after dosing were pooled and analyzed. Plasmas were pooled due to low concentration of total racemic venlafaxine in the specimens and limited sensitivity of the assay. Extraction and Lkriuatization of Biological Specimens
A stock solution of venlafaxine was prepared by dissolving venlafaxine hydrochloride in water to a concentration of 1 mg/ml of the free base. Standard solutions were made by diluting the stock solution with deionized water to obtain venlafaxine concentrations of 200,100, 50, 20, 10, 4.0, and 2.0 pg/ml. The stock internal standard solution was prepared by dissolving Wy-45,818 as the hydrochloride salt in water. Fifty microliters of internal standard aqueous solution (Wy-45,818)
86
WANG ET AL.
at a concentration of 22.5 pg/ml and 0.2 ml of saturated sodium borate solution were added to 1.0 ml of sample (spiked control plasma or plasma from experimental animals) in a 16 x 125 mm disposable screw-cap culture tube and mixed by vortex. Five milliliters of purified isopropyl ether was added and the tubes were shaken for 15rnin and then centrifuged at 2,500 rpm for 10 min. A 4.5 ml aliquot of the isopropyl ether phase was transferred to a clean conical screw-captube containing 0.4 ml of 0.01 N hydrochloric acid. The tubes were shaken for 15 rnin and then centrifuged at 2,500 rpm for 10 min. The ether phase was aspirated and discarded. One milliliter of saturated sodium borate solution was added to the aqueous phase, and the tubes were vortexed. Following the addition of 2.0 ml of purified isopropyl ether, the tubes were shaken for 15 rnin and then centrifuged at 2,500 rpm for 5 min. The organic phase was transferred to a clean conical screw-captube and evaporated to dryness under nitrogen. Five to 10 mg of anhydrous sodium carbonate powder and then 0.2 ml of naproxen chloride solution (15 mg per 100 ml methylene chloride) were added to each tube and vortexed. The solutions were covered with aluminum foil and allowed to stand at room temperature to react for 20 h. The methylene chloride was then evaporated to dryness under a mild stream of nitrogen gas. To the residue, 1ml of deionized water was added and vortexed to completely dissolve the sodium carbonate. Two milliliters of chloroform was then added and the tubes were shaken for 10 min and centrifuged at 2,500 rpm for 5 min. The chloroformphase was transferred to a clean screw-cap conical tube and washed with 2 ml deionized water by shaking for 10 rnin and centrifuged at 2,500 rpm for 5 min. The final chloroform extract was then transferred to a clean mew-cap conical tube and evaporated to dryness under nitrogen gas. The diastereomeric esters were separated by HPLC. HPLC Analysis
of mobile phase A/B (50:50) and 11 pl was injected onto the HPLC column. The column was maintained at ambient temperature and the detection wavelength was 229 nm. Enantiomer concentrations in ng (free base)/ml were obtained from the regression line relating compound/internal standard (the derivative peak of the enantiomer of Wy-45,818 with longer retention time) peak area ratios to the compound concentrations calculated using the program “REG 1SD” in a Hewlett-Packard model 85 calculator. The racemate of venlafaxine was assumed to have an enantiomer ratio of R:S = 1:1 due to the lack of optical rotation. The specificity of the method was evaluated with regard to interference due to the presence of endogenous substances in the extracts of dog plasma and for interference from known metabolites. Three random control plasmas were processed by the method described and one control plasma spiked with solutions of known venlafaxine metabolites was analyzed. The retention times of the diastereomers of metabolites were evaluated with respect to the retention times of the diastereomers of venlafaxine and Wy-45,818. The amount of each metabolite added was Wy-45,233 (800 ng), Wy-45,030 (800 ng), Wy-45,818 (800 ng), Wy-4!5,494(800ng),Wy-47,877(800ng and 5,000 ng), Wy-47,894 (800 ng and 5,000 ng), Wy-45,821 (800 ng), Wy46,689 (800ng), Wy-46,965 (800 ng and 5,000 ng), and Wy47,961 (800 ng and 5,000 ng) per ml of dog plasma. The intraday precision of the method was assessed by calculating the coefficient of variation for replicate samples (n= 5) at a concentration of 300 ng of each enantiomer/ml. Identification of the Individual Enantiomers
To distinguish the respective chromatographic peaks of (R)-venlafaxine hydrochloride and (S)-venlafaxine hydrochloride, each was extracted, derivatized, and analyzed by HPLC individually.
The mobile phase was composed of solution A (acetonitrile) In Vivo Study of the Chiral Inversion o f the (R)- and and solution B, which was prepared by adjusting the pH of 0.1 (S)-Enantiomers of Venlafaxine in Rat M KH2P04 solution to 3.0 with 85% 0-phosphoric acid and Nineteen male Sprague-Dawley rats (Charles River), weighthen adding 0.07% (v/v) triethylamine to achieve a final pH of ing 187-217 g, were fasted overnight prior to dosing. The 3.25. An HPLC gradient profile was used to analyze the sample animals received a single 60 mg/kg ig dose of either (R)- or (Table 1).The derivatized extract was reconstituted in 0.4 ml (S)-venlafaxineas the hydrochloride salt. The rats were slightly anesthetized with a mixture of 02/C02 (95:5) and halothane gas in groups of three at 0.5,1.5, and 2.5 h after dosing. One rat, TABLE 1. Gradient program for the analysis of which was used to collect control blood, was not dosed. As venlafaxine enantiomersO much blood as possible was collected by cardiac puncture into Mobile phase (%) the heparinized tubes. The plasma was separated and frozen at Time Flow rate - 20°C until analysis. The individual venlafaxine enantiomer Ab Bc (min) (ml/min) concentrations were determined by HPLC. The ratios of the enantiomer concentrations of venlafaxine in the plasma samInitial 50 50 0.8 ples were calculated. 30 54 46 0.8 36 50 53 56
2.0 2.0 2.0 0.8
55 57 57 50
45 43 43 50
‘Linear gradient was used with all time intervals. bMobilephase A acetonitrile. ‘Mobile phase B 0.1 M KH@, (pH 4.42) adjusted to pH 3.0 with phosphoric acid followed by the addition of 0.07% (v/v) triethylamine to pH 3.25.
Calculation o f Pharmacokinetic Parameters
Pharmacokinetic parameters were calculated from plasma concentration data by standard noncompartmental methods. Zero-to-t h areas under the plasma concentration-time curve (AUCst) were calculated by the trapezoidal rule. Peak plasma and time to peak (&.J were obtained by concentration (C,,,=) data inspection. Terminal elimination half-lives (hh)were cal-
87
DISPOSITION OF VENLAFAXINE ENANTIOMERS
culated as tlA = In 2/&, where I(el was the slope of regression line of the terminal log-linear data points. AUCO_~ was obtained as AUCo - + CJ&.
Identification of the Individuul Enantiomers of Venlafaxine
(R)-Venlafaxine, when analyzed individually, appeared in the chromatogram as a single peak at 16.73min, while (S)-venStatistics lafaxine, when analyzed alone, appeared in the chromatogram A paired t- test was used for statistical comparison of the as a major peak at 19.34 min. This demonstrated that the peak G,,, h ~AUC&, , , or AUCG15 values of (R)- and 6)-venlafax- of venlafaxine-naproxen diastereomer with shorter retention ine in dog. The degree of significance was set at 01 = 0.05. A time is derived from (R)-venlafaxine,while the one with longer single factor, repeated measures analysis of variance, was used retention is derived from (S)-venlafaxine. to analyze venlafaxine enantiomer ratios in human. m L C Analysis RESULTS Typical HPLC chromatograms for control dog plasma, dog LCI Mass Spectrometric Characterization of the plasma spiked with venlafaxine and Wy-45,818, and plasma Naproxen Derivatives of Venlafaxine and Wy45,818 samples from a dog, a rat, and human given venlafaxine are The molecular weights of the naproxyl derivatives of (R)- presented in Figure 2. Under the conditions specified,the retenand (S)-venlafaxineand Wy-45,818 were 489 amu and 493 amu, tion times for the derivatives of (R)- and (S)-venlafaxineand respectively. The molecular weights confirmed the formation of Wy-45,818 enantiomerswere 16.53,18.03,19.74, and 22.38 min, diastereomers which are the esterification product of the terti- respectively. No endogenous interfering substances in the exary hydroxyl group of venlafaxine and Wy-45,818 and acyl tracts of three random control dog plasma samples were detected. The chromatogram of the extract of venlafaxine, Wygroup of naproxen chloride.
A
2 3
D 10 20 30 4 Retention Time (IT
I
10
20 30 40 50 Retention Time (min)
I
I
10
I
I
20 30 40 50 Retention Time (min)
3
E
C
10
L
20 30 40 Retention Time (min)
50
1
10
20 30 40 Retention Time (min)
50
Fig. 2. Chromatograms of control dog plasma (A), control dog plasma spiked with venlafaxine and Wy-45,818(B), and plasma samples from a dog (C), a rat (D), and human given venlafaxine (E). 1,(R)-venlafaxine;2, (S)-venlafaxine; 3, (2)-Wy-45,818(the enantiomer of Wy-45,818 with longer retention); 4, (R)-Wy-45,233;5, (S)-Wy-45,233.
88
WANG ET AL.
-
10.0 45,818, and known metabolites reacted with chiral reagent ( + )-(S)-naproxenchloride is shown in Figure 3. The retention -0- (S)-Venlafaxine times for the diastereomers of Wy-45,821,(R)- and (Shenlafaxine, (1)and (2) Wy-45,818, Wy-45,494, (R)- and (S)-Wy-45,233 and Wy-46,689 were 14.58, 16.30, and 17.79, 19.47 and 22.10, 20.01, 43.56, and 45.40, and 43.56 min, respectively. The dias- C tereomers of Wy-46,689 coeluted with the derivative of (R)-Wy- .-0 45,233. No peaks of the derivatives of Wy-47,877, Wy-47,894, 2 C Wy-47,961, and Wy-46,965, which are metabolites, can be de- 8 .1 tected in this chromatographic system even at concentrations 0 of 5,000 ng/ml. The linearity of the peak area ratio to enantiomer concentration was calculated using linear regression analysis. The 0 2 4 6 8 10 12 14 16 method was linear for concentrations of each enantiomer of Time (hr) between 50 and 5,000 ng/ml in dog plasma with correlation coefficients of 0.9999 for both (R)- and (S)-enantiomers. The Fig. 4. Mean plasma concentrationsof (S)-venlafaxine,(R)-venlafaxine,and intraday precision of the method was assessed by calculating ( =k)-venlafaxine in dog after daily administration of 22 mg/kg venlafaxine for the coefficient of variation for replicate samples (n= 5) at 300 14 consecutive days. Values are the mean of four subjects. ng/ml for each enantiomer of venlafaxine in dog plasma. The % CV for (R)- and (S)-venlafaxinewere 3.1 and 0.6, respectively. value of 1.51 f 0.11. The mean f SD LX values of (R)-and (S)-venlafaxinewere 1.78 f 0.21 and 2.47 f 1.00 pg/ml, reIn Vivo Study of the Chiral Inversion of the (R)- and spectively. The & was 1.13 f 0.63 h for both enantiomers. (S)-Enantiomersof Venbfaxine in Rat The AUC@, and tlh of (R)- and (S)-venlafaxinewere 10.70 f No (S)-venlafaxine appeared in either the (R)-venlafaxine 2.33 and 15.77 f 7.74 pgh/ml, and 4.10 f 0.57 and 4.08 f dose or in the plasma of rats receiving (R)-venlafaxine,indicat- 0.59 h, respectively. There was no significant difference in the ing that chiral inversion of (R)-venlafaxine did not occur. A C;nax, hh, AUC&, ,or AUCs15 values for (R)-and (S)-venlafaxsmall (R)-venlafaxinepeak was detected in the (S)-venlafaxine ine (P>O.l), which indicates an absence of a significant stedosage form and as well as in the plasma of rats receiving reoselective disposition of venlafaxine in dogs. (S)-venlafaxine.The mean f SD of the (R)/(S)ratio in the (S)-venlafaxinedosing material was 0.035 f 0.012. The mean Analysis of Rat Plusma (R)/(S)ratio in the plasma of rats receiving (S)-venlafaxine The pharmacokinetic data for (R)-venlafaxineand (S)-venranged from 0.028 to 0.030 over the 0.5 to 2.5 h interval. Comlafaxine in rat are presented in Table 3. The mean plasma parison of the (R)/(S) ratios in the dosing material and plasma concentrations of total racemic, (S)-and (Rkvenlafaxine are of the rats receiving (S)-venlafaxineindicates that no formation depicted in Figure 5. The ratios of the mean plasma concentraof (R)-venlafaxineoccurred in vivo. tions of (S)-venlafaxineto that of (R)-venlafaxineover the 0.5 to 6.0 h interval varied from 2.97 to 8.50 with a mean value of 5.51 Analysis of Dog Plasma f 2.45. The GX of (R)- and (S)-venlafaxinewere 0.88 f 0.59 Mean concentrations of (R)-venlafaxineand (S)-venlafaxine in dog plasma are depicted in Figure 4 and the pharmacokinetic data are presented in Table 2. The ratios of the mean plasma TABLE 2. Pharmacokinetics of venlafaxine enantiomers concentrationsof (S)-venlafaxineto that of (R)-venlafaxineover in the plasma of dogs receiving venlafaxine (22 the 0.5 to 15 h interval varied from 1.35 to 1.68. with a mean mg/kg/day) for 14 consecutive days
c c
-
~~
R
1
S
10
20
30
h
40
Retention Time (min)
50
Fig. 3. Chromatogram of control dog plasma spiked with venlafaxine, Wy45,818(internal standard), and known metabolites of venlafaxine: 1,Wy-45,821; 2, (R)-venlafaxine;3, (S)-venlafaxine;4, (l)-Wy45,818; 5, Wy-45,494; 6, (2tWy45,818 (internal standard); 7, (R)-Wy-45,233; 8, Wy-46.689; 9, (S)-Wy-45,233.
2
R
3
R
S
S
R
4
S
Mean f
Mean f
R
SD SD
Total f SD
S
2.09 3.96 1.62 1.96 1.69 2.06 1.73 1.87 1.78 0.21 2.47 1.00 4.25 1.21
0.5 0.5 2.0 2.0 1.0 1.0 1.0 1.0 1.13 0.63 1.13 0.63 1.13 0.63
4.71 4.23 4.20 4.50 4.13 4.38 3.34 3.21 4.10 0.57 4.08 0.59 3.97 0.58
11.63 23.79 10.06 13.04 10.54 13.17 7.07 7.30 9.82 1.95 14.32 6.88 24.15 8.64
12.80 26.25 11.08 14.63 11.54 14.62 7.38 7.58 10.70 2.33 15.77
7.74 26.84 10.00
89
DISPOSITION OF VENLAFAXINE ENANTIOMERS
TABLE 3. Pharmacokinetics of venlafaxine enantiomers in the plasma of rats after administrationof racemic venlafaxine (120 mg/kg/day) for 14 consecutive days Compound ~
~~
“Mean
0.5 0.5 0.5
0.88 f 0.59 2.61 f 1.15 3.49 f 1.72
(R)-Venlafaxine (S)-Venlafaxine Total venlafaxine
1.27 5.56 6.77
1.21 5.20
2.00 1.69 1.50
-
* SD.
and 2.61 f 1.15 pg/ml, respectively. The corresponding & values were 0.5 and 0.5 h, respectively. The AUC&, and tlh values of (R)- and (S)-venlafaxinewere 1.27 and 5.56 pgh/ml and 2.00 and 1.69 h, respectively, indicating that the disposition of venlafaxine in rats was stereoselective. Analysis of Human Plasma
The plasma ratios of the (R) and (S)enantiomersof venlafaxine in pooled samples are given in Table 4. The mean ratios [(S)/(R)]of the enantiomers of venlafaxine over a 2 to 6 h interval ranged from 1.33 to 1.35 with an overall ratio of 1.34 f 0.26 (n= 12). The plasma ratios of the enantiomers [(S)/(R)] were not significantly different from unity at each time point and overall ( P L 0.14), indicating no stereoselective disposition of venlafaxine enantiomers in man. The plasma [(S)/(R)] ratios of the enantiomers were not significantly different among time points (P= 0.97). DISCUSSION
An HF’LC method has been developed using naproxen chloride as the chiral reagent for the quantitation of venlafaxine enantiomers in dog, rat, and human plasma. The method is specific, precise, and linear at least in the range of 50 to 5,000 ng/ml. The minimum quantifiableconcentration was 50 ng/ml. The detection limit could be lowered to 25 ng/ml by increasing the injection volume, since at 22 p1 injection volume no interfering peaks were found in the control plasma. Spahn et al.495used naproxen chloride synthesizedfrom naproxen and thionyl chlo-
10.0
-
ride to quantify primary and secondary amines and alcohols. In this study, the derivatization condition has been modified and a postderivatizationprocess has been developed to separate the enantiomersof venlafaxine, a tertiary alcohol chiral compound. This method is applicable to the chiral assay of other compounds with chiral tertiary alcohol and a phenolic hydroxy functional group as demonstrated in the derivatization of the major human metabolite of venlafaxine, Wy-45,233. This method is practical for the routine chiral analysis of venlafaxine enantiomers in biological specimens, including plasma and urine. The method can be modified by using fluorescencedetection and/or a normal phase HPLC system. The pharmacokinetic disposition of venlafaxine in rats receiving multiple oral doses of drug has been demonstrated to be stereoselective.The ratio of the mean plasma concentrations of (S)-venlafaxineto that of (R)-venlafaxineover the 0.5 to 6.0 h interval varied from 2.97 to 8.50 with a mean value of 5.51 f 2.45. Because no in vivo chiral inversion of venlafaxine enantiomers in rat was found, the stereoselectivedisposition of (R)- and (S)-venlafaxinein rat can be assumed to arise from differences in drug absorption, distribution, metabolism, or excretion of the individual enantiomers. However, the elimination half-lives of both enantiomers were similar. The disposition of venlafaxine in dog receiving multiple oral doses of venlafaxine was not stereoselective. The ratio of the mean plasma concentration of (S)-venlafaxineto that of (R)-venlafaxine over the 0.5-15 h interval varied from 1.35 to 1.68,with a mean value of 1.51 f 0.11. The mean ratio of the plasma concentrationsof (S)-venlafaxine to that of (R)-venlafaxine over the 2-6 h time interval in human subjects receiving 50 mg oral doses of venlafaxine
-0- (S)-Venlafaxine -D-
E
(R)-Venlafaxine
TABLE 4. Ratios of (R)-and (S)-venlafaxinein pooled plasma of human subjects administered 50 mg of venlafaxine orally
\
8 E
1.0
Ratio of enantiomers (S)/(R)
I I
I
I
I
0
1
2
3 Time (hr)
.01
\ I
I
I
4
5
6
Fig. 5. Mean plasma concentrationsof (Stvenlafaxine,(R)-venlafaxine,and (*)-venlafaxine in rat after daily administration of 120 mg/kg venlafaxine for 14 consecutive days. Values are the mean of five replicates.
Pool No:
2 hb
3h
4h
6h
1 2 3
1.31 1.59 1.16
1.25 1.61 1.12
1.24 1.64 1.14
1.22 1.83 1.01
Mean f SD
1.35 0.22
1.33 0.25
1.34 0.27
1.35 0.42
OThree subjects per pool. *Hours after dosing.
90
WANG ET A L
varied from 1.33 to 1.35.The plasma ratios of the enantiomem [@/@)I were not significantlydifferentfrom unity at each time point and overall. It is concluded that the disposition of venlafaxine enantiomers in humans is not stereoselective and is similar to that in dogs. LITERATURE CITED 1. Muth, E., Haskins, J., Moyer, J. et al. Antidepressant biochemical profile of the novel bicyclic compound Wy-45,030,an ethyl cyclohexanol derivative. Biochem. Pharmacol. 54493-4497, 1986.
2. Fabre, L., Putman, €HI. I. An ascending single-dose tolerance study of w Y 4 , o a , a bicyclic a n t i d e P m t r in healthy men. cm. RE,. 42: 901-909,1987. 3, Gibaldi, M,, perrier, D. pharmacokinetics, 2nd ed. New York: Marcel &kker, 1982: 409417, ug. 4. Spahn, H., Kraup, D., Mutschler, E. Enantiospecific high-perfomance liquid chromatographic(HPLC)determination of baclofen and its fluoro analogue in biological material. Pharm. Res. 5107-112, 1988. 5. Spahn, H. S-(+)-Naproxen chloride as acylating agent for separating the enantiomers of chiral amines and alcohols. Arch. Pharm. (Weinheim) 321: 847-850,1988.
