EUROPEAN JOURNAL OF DRUG METABOUSM AND PHARMACOKINETICS, 1992, VoL 17, No.4, pp. 301-308

Pharmacokinetics of nefazodone in the dog following single oral administration U.A. SHUKLA, P.H. MARATHE, I.A. LABUDDE, K.A. PITTMAN and R.H.BARBHAIYA Department ofMetabolism and Pharmacokinetics, Pharmaceutical Research Institute, Bristol-Myers Squibb Company, Syracuse, New York, USA

Receivedfor publication: May 26, 1992

Keywords: Nefazodone, pharmacokinetics, dog, single dose, oral

SUMMARY The pharmacokinetics of nefazodone (NEF) and two of its pharmacologically active metabolites viz hydroxynefazodone (HO-NEF) and m-chlorophenylpiperazine (mCPP) were determined following single oral administration of 100, 200 and 400 mg NEF to 6 beagle dogs in a three-way crossover study. Blood samples were collected for 48 h and plasma was analyzed for NEF, HO-NEF and mCPP by a validated HPLC assay. NEF was rapidly absorbed after oral administration. C IIIlIlt values for all three compounds and AUCinf values for HO-NEF and mCPP were dose-proportional; AUCinf values for NEF were dose-linear but not dose-proportional. The Tm values for NEF and HO-NEF following the 400 mg dose were significantly greater than those for the 100 mg dose. No differences in mCPP Tm were observed among the doses. The CIIIlIlt and AUCinf ratios for metabolite:NEF were about 2-fold lower for the 200 and 400 mg doses than those observed for the 100 mg dose. However. due to extensive variability. the ratios for three doses were not significantly different based on statistical analysis. Overall. these data suggest the pharmacokinetics of NEF are dose-dependent in the beagle dog. Statistical significance for dose-dependency for many of the pharmacokinetic parameters could not be demonstrated due to high variability associated with the plasma concentration vs time profiles.

INTRODUCTION Nefazodone (NEF), 2-[3-[4-(3-chlorophenyl)-I-piperazinyl]propyl]-5-ethyl-2,4-dihydro-4-(2-phenoxyethyl) -3H-l,2,4-triazol-3-one hydrochloride is a new compound under development as an antidepressant and is currently undergoing phase Illb trials. It is chemically distinct from the tricyclics and monoamine oxidase inhibitors (Fig. 1). Early studies suggest that NEF may

Please send reprint requests to: Dr Rashmi H. Barbhaiya, Department of Metabolism and Pharmacokinetics. Pharmaceutical Research Institute. Bristol-Myers Squibb Company. P.O. Box 4755, Syracuse. NY 13221, USA

be a clinically effective antidepressant drug that has little or no anticholinergic side effects (1, 2). NEF is an effective inhibitor of serotonin reuptake with negligible effects on norepinephrine uptake and it is also known to function as an antagonist of 5-HT2 receptors (3-5). Two metabolites of NEF were identified in the early phase of development viz hydroxynefazodone (HO-NEF) and m-chlorophenyl- piperazine (mCPP) (Fig. 1) (6). Both metabolites possess significant phannacologicaI activity as seen from in vivo and in vitro tests for antidepressant activity in mice and rats. The pharmacologic profile of HO-NEF parallels that of NEF while the profile of mCPP is complex with some synergistic and some antagonistic effects to NEF.

Eur. J. Drug Metab. Pharmacokinet., 1992, No.4

302

Study design

(1)

This was a single-dose study in which 6 beagle dogs received oral doses of 100, 200, and 400 mg of NEF HCl in a three-period crossover design according to the dosing schedule in Table I. The three treatments were separated by a l-week washout period. Table I: Dosing schedule for dogs receiving 100,200, and 400 mg nefazodone hydrochloride.

(2)

CI

Dog No.

Weight (kg)

Session 1 Session 2 Session 3

1

12.1

l00mg

200mg

400mg

2

11.1

200mg

l00mg

400mg

3

10.2

400mg

200mg

l00mg

4 5

11.3

l00mg

200mg

400mg

11.8

200mg

l00mg

400mg

6

12.2

400mg

200mg

l00mg

H-U-@

(3)

Fig. 1 : Chemical structures of: (1) nefazodone (2) hydroxynefazodone (3) m-chlorophenylpiperazine.

The beagle dog has been utilized as one of the primary species for toxicologic evaluation of NEF. The present study was designed to determine the phannacokinetics and to assess the dose- proportionality of NEF and the metabolites HO-NEF and mCPP after single oral doses of 100, 200 and 400 mg of NEF hydrochloride to the beagle dog.

MATERIALS AND METHODS Animals Six healthy adult male beagle dogs (mean weight, 11.5 kg) were used in this study. The animals were identified by ear tatoos and randomly assigned numbers 1 to 6. The dogs were individually housed in stainless steel metabolism cages. The animals were fasted overnight (18 h) before administration of NEF and for 4 h after dosing. During the washout period of the study, food was provided once a day between 7:00 and 10:00 a.m. and fresh drinking water was offered ad libitum.

Immediately prior to dosing, 100 ml of water was administered orally. 1,2, or 4 capsules containing 100 mg NEF HCl (Bristol-Myers Squibb Pharmaceutical Research Institute, Evansville, IN, USA) were administered orally to the animals followed by 30 ml of water to ensure the capsules were swallowed. Dosing was in the morning at approximately 8:00 am. after an 18 h overnight fast Serial blood samples (8 ml each) were collected during each treatment in labeled Vacutainer~ tubes containing K:3EDTA just prior to dose (predose), and at 0.5, 1, 1.5, 2, 3, 5, 7, 12, 24, and 48 h after drug administration. After mixing with the anticoagulant, plasma was separated by centrifugation and stored at -20·C until analyzed forNEF, HQ-NEF and mCPP.

Assay Plasma samples were analyzed for NEF, HQ-NEF and mCPP by a validated reverse-phase HPLC/UV method (7) with a modification. Plasma samples were extracted manually instead of employing a robotic extraction as described previously. Spiked quality control (QC) samples were prepared in control dog plasma using reference standards and stored with the study samples. QC samples were analyzed with study samples to establish sample stability, assay accuracy and precision. The standard curves were linear (correlation coefficient> 0.996) and reproducible « 10% RSD for slope except for mCPP where %RSD was 29%) over the

U. A. Shukla et 01•• Nefazodone & single oral admin (dog) concentration range of 10-1000 ng/ml for NEF and HO-NEF. and 5-250 ng/ml for mCPP. Samples with predicted concentrations greater than the highest standard were reanalyzed after appropriate dilution. The mean observed concentrations of the QC samples deviated < 9% from nominal values indicating the assay method was accurate. The small between-day and within-day errors « 8%) for the QC samples indicated that the assay method was precise and reproducible.

Pharmacokinetic analysis Plasma concentration (C) versus time (t) data for NEF. HO-NEF and mCPP were analyzed by non- compartmental methods (8, 9). The terminal log-linear phase of the plasma concentration-time curve was identified by least-squares linear regression of data points which yielded a minimum mean square error. The area under the plasma concentration-time curve from zero to infinity, AUCinf, was determined by a combination of trapezoidal and log-trapezoidal methods plus the extrapolated area. The extrapolated area was determined by dividing the predicted concentration at the time of last non-zero plasma concentration by the slope (B) of the terminal log-linear phase. The halflife, TII2. of the terminal log-linear phase was calculated as 0.693 divided by the absolute value of B. The peak plasma concentration, Cmax, and the time at which Cmax occurred, T max. were obtained from the observed data. The exposure to HO-NEF and mCPP in relation to NEF was determined by calculating the Cmax ratio for metabolite:NEF using units of J.1MIl and the AUCinf ratio for metabolite:NEF using units of J,lM.h/l.

Statistical methods Weighted linear regression was used to determine the dose-linearity and dose-proportionality of Cmax and AUCinf for each compound. These analyses were weighted by the reciprocal of the variance for each dosage level to correct for increasing variance with increasing dose. Dose-linearity was tested by the t-test for a linear trend across doses. The test for nonlinearity was perfonned using the lack-of-fit t-statistic. Dose-proportionality was evaluated after dose-linearity was observed by a t-statistic for the significance of the intercept In these analyses. dog was used as a blocking factor; significance of the intercept was based on the average intercept for all dogs. Analyses were con-

303

ducted separately for NEF. HC-NEF and mCPP. The Cmax and AUCinf ratios for metabolite:NEF and TII2 values were analyzed by a randomized block design. Although this study was conducted as a threeway crossover design, the small sample size precluded the evaluation of a sequence effect Since no dogs received the 400 mg dose in session 2 and none received the 200 mg dose in session 3, the effect of study period also could not be clearly identified. Therefore. the randomized block model was used instead of the three-way crossover model and considered the effect of dog. the treatment effect, and the residual error. If the effect of treatments was statistically significant, then Tukey's procedure was used to compare treatment means. Levene's test was used to verify homogeneity of variance in this analysis. Analyses were conducted separately for all three compounds. Except for Levene's test (P = 0.(01), all statistical tests were conducted at the P = 0.05 level.

RESULTS The mean (SO) pharmacokinetic parameters for NEF, HO-NEF and mCPP are presented in Tables IT, ill and IV respectively. Plasma concentration vs time profiles of NEF. HO-NEF and mCPP in a representative dog (dog 1) are shown in Figures 2, 3 and 4 respectively. Mean Cmax for NEF at 100, 200 and 400 mg doses were 194, 629 and 1187 ng/ml respectively and were observed at mean Tmax of:5 1.4 h. Mean TII2 of NEF ranged from 2.23 to 5.24 h with the TII2 following the 400 mg dose (5.24 h) being significantly greater than that following the 100 mg dose (2.23 h). Figure 5 shows plots of Cmax and AUCinf vs dose for NEF in individual dogs. Mean Cmax values were in a 1:3:6 ratio for the 1:2:4 administered dose indicating that the Cmax values increased disproportionately with dose. However, statistical analysis showed that Cmax of NEF was linearly related to the dose in the range examined in this study with R2 = 0.64. The intercept of this regression was not significantly different from zero (P > 0.05) indicating Cmax was also dose-proportional. For doses in the 1:2:4 proportion, mean AUCinf values for NEF were in the ratio of 1:4:7. Statistical analysis indicated that AUCinf values increased in a dose-linear but not dose-proportional manner (R2 = 0.81) with the intercept significantly different from zero (P < 0.05). Irrespective of dose, mean C max values for HONEF were observed at a mean T max of:5 1.4 h. Mean Cmax values for the 100,200 and 400 mg doses were 364, 702 and 1109 ng/ml respectively. Figure 6

Eur. J. DrugMetab. Pharmacokinet., 1992, No.4

304

Table II : Mean (SO) pharmaeokinetic parameters of NEF following oral administration of various doses of NEF HCl to beagle dogs (n .. 6).

Table III : Mean (SO) pharmacokinetic parameters of HO-NEF following oral administration of various doses of NEF HCl to beagle dogs (n .. 6).

Dose (mg)

Parameter

]00

200

Dose (mg)

400

Parameter

]00

C max (nglml)

364

702

1109

(180)

(236)

(492)

200

400

Cmax (nglml)

194

629

1187

(133)

(435)

(886)

AUCinf

754·

2866·

5246

AUCinf

1229·

2430

4286

(ng.hlml)

(507)

(1467)

(3678)

(ng.hlml)

(653)

(1383)

(1979)

2.23·

3.79·

5.24#

Tlfl

1.36·

2.34

3.41#

(0.93)

(1.58)

(2.10)

(0.55)

(1.07)

(1.99)

Tlfl

(h)

Tmax (h)

1.2

1.3

1.8

(0.7)

(0.8)

(1.1)

• n - S. AUCInf and T1I2 were not detennincd for dog 6 at 100 mg and dog 2 at 200 mg due to insufficientdata # Significantly longer than TI/2 at 100 mg (P < 0.05)

Table N : Mean (SO) pharmaeokinetic parameters of mCPP following oral administration ofvarious doses ofNEF HCl to beagle dogs (n » 6).

(h)

T max (h)

C max ratio··

AUCinf ratio••

1.2

1.3

1.4

(0.7)

(0.8)

(1.0)

2.16

1.91

1.30

(0.643)

(1.72)

(0.85)

1.79·

1.11

1.11

(0.442)

(0.415)

(0.605)

• n - S. AUC'mt and TI/2 were oot determined for dog 6 at 100 mg due to insufI"ICient data .. Metabolite to parent drug ratio; no statisticallysignificant differences were observed among doses # SignifICantly longer than TI/2 at 100 mg (P < 0.05)

Dose (mg) Parameter

]00

200

400

C max

58.2

92.3

184

(nglml)

(30.5)

(40.9)

(80.3)

AUCinf

452

854

1520

(ng.hlml)

(226)

(403)

(641)

TIn (h)#

T max (h)

C max ratio·

AUCinf ratio·

4.49

5.28

7.35

(1.16)

(1.34)

(4.56)

3.0

2.8

2.7

(1.6)

(1.2)

(2.0)

0.989

0.634

0.560

(0.536)

(0.733)

(0.425)

2.60

0.794

1.17

(2.03)

(0.203)

(1.07)

• Metabolite to parent drug ratio; no statistically significantdifferences were observed among doses # No statisticallysignifICant differences were observed among doses

shows plots of Cmax and AUCinf vs dose for HO-NEF in individual dogs. Mean Cmax values for HQ..NEF were in the ratio of 1:2:3 for the 1:2:4 dose proportion and were shown to be dose-proportional with R2 = 0.73 by statistical analysis. The intercept of the weighted linear regression was not significantly different from zero (P > 0.05). Mean HQ..NEF AUCinf values were in the ratio of 1:2:3 for the 1:2:4 dose proportion. The AUCinf data were also found to be dose-proportional with R2 = 0.72. The mean Cmu ratios for HO-NEF:NEF decreased from 2.16 at 100 mg to 1.30 at 400 mg dose but were not significantly different with respect to dose (P > 0.05). The mean AUCinfratios for HO-NEF:NEF exhibited significant differences by ANOVA with respect to dose, however, none of the pairwise comparisons were significant (P > 0.05). Mean TI/2 values of HO-NEF ranged from 1.36 h to 3.41 h at the 3 doses and the TI/2 following the 400 mg dose (3.41 h) was significantly longer than that following the 100 mg dose (1.36 h) (P < 0.05). The TI/2 of HO-NEF was either shorter or comparable to that of NEF at each dose level.

U. A. Shukla et al., Ne[azodone &: single oral admin (dog) 10000 , . - - - - - - - - - - - - - - - - - - ,

--

-E

C)

c:

C)

c: u,

z

-

1000

o

........

c:

,-. 1\

u,

w z, 0 ::z:

as

c:

\

100

0

~. ~ ~

o

E

rn

as a.

1000

.....0

'.

o

c: CD o c: o o as

10000 C

E

w

305

-.

~

......as

c:

CD 0 C

0

0

as

E 0

10 ' - - - - , - - - - , - - . , - - - r - - r - - - . . . J o 4 8 12 16 20 24

as a. 10 0

4

8

12

18

20

Time (hour) Time (hour) Fig. 2 : Plasma concentrations of NEF following oral administration of 100 (open circles), 200 (filled circles) and 400 (open squares) mg dose of NEF in dog I.

Fig. 3 : Plasma concentrations of HO-NEF following oral administration of 100 (open circles), 200 (filled circles) and 400 (open squares) mg dose of NEF in dog 1.

1000 . . - - - - - - - - - - - - - - - - - - ,

::::E

--

The mean Cmax of mCPP occurred at a later time (Tmax > 2 h) when compared to NEF and HQ-NEF. Mean rnCPP emu values were 58.2, 92.3 and 184

C)

c:

00-

o E .....o c:

o

:;...

-

100

.\

c:

CD

o c: o o as E rn as

fo~"0'-----

o

0-

10

'------r---r---"""'T""----,....--"'"'T"'"----'

o

4

8

12

18

20

24

nglml respectively for the 3 doses. Cmu: and AUCinf values at each dose level for individual dogs are plotted in Figure 7. Mean Cmax and mean AUCinf data were in the ratio of 1:2:3 for the dosage levels in the 1:2:4 proportion. Both Cmax and AUCinf for mCPP were fouod to be linearly related to dose with R2 = 0.77 and R2 = 0.75 respectively. The intercepts of both lines were not significantly different from zero (P > 0.05) indicating dose-proportionality of both Cmax and AUCinf for mCPP. No statistically significant differences wereobserved betweenthe doses with respect to Cmax and AUCinf ratios for mCPP:NEF (P > 0.05). Mean T1I2 values for mCPP ranged from 4.49 to 7.35 h and were statistically not different between doses. The Till of mCPP was Ioager than that of NEF at each dose level.

Time (hour) Fig. 4 : Plasma concentrations of mCPP following oral administration of 100 (open circles), 200 (filled circles) and 400 (open squares) mg dose ofNEF in dog I.

DISCUSSION The results of this study indicated that NEF is rapidly

24

306

Eur. J. DrugMetab. Pharmacokinet., 1992, No.4 2500 , . . . - - - - - - - - - - - - - - - - - - - ,

3000

-.s :::::E ClI

>C

•

E 0 u..

......

.s E

Cmax-3.82xdoae·183

2500

Cmax-2.87xdoae+117

ClI

2000

•E

1600

u..

1000

6

600

)(

e

0

1500

W Z

1000

W Z

:r::

600

0

0 0

100

200

300

400

0

500

100

10

-...

.....

A

AUC(INFl-19.9xdoae·1625

8

.c

ClI

-

E .c::CII

e

•

iL~ zl; 0 .... ..c:

gt:.

ez

8

.... (J

4

400

8000,..------------------,

AUC(INF)-11.5xdos.·135 5000

4000

::J

-e

-e

u, W Z

300

oS

r;~

~.

200

Do•• emg)

DOl. emg)

~

•

2000

u.. W Z

2

2000

6

::t:

0 0

100

200

300

400

600

DOl. emg)

0'"""""----1-------,,...----,-------,,...---.....

o

100

200

300

400

600

008. (mg)

Fig. 5 : Plot of NEF Cmu (upper panel) and AUCinf (lower panel) vs NEF dose in individual dogs : dog 1 (open triangles), dog 2 (open circles), dog 3 (filled triangles), dog 4 (filled circles), dog 5 (open squares), dog 6 (filled diamonds), mean (filled squares). The line represents weighted linear regression on the data. AUCinf for dog 2 at 200 mg and dog 6 at 100 mg were not determined and thus not plotted.

Fig. 6 : Plot of HO-NEF Cmu: (upper panel) and AUCinf (lower panel) vs NEF dose in individual dogs : dog 1 (open triangles), dog 2 (open circles), dog 3 (filled triangles), dog 4 (filled circles), dog 5 (open squares), dog 6 (filled diamonds), mean (filled squares. The line represents weighted linear regression on the data. AUCinf for dog 6 at 100 mg was not determined and thus not plotted.

absorbed following oral administration of 100, 200 and 400 mg doses to the dog with maximum plasma levels occurring within 2 h after dosing. Statistically the Cmax data for NEF were accepted as doseproportional. However, mean Cmax values for NEF increased in a 1:3:6 ratio for a 1:2:4 increase in dose. Individual NEF Cmax values also increased disproportionately. However, the high variability in the data as indicated by a relatively low correlation precluded demonstration of statistical significance for dosedependency. Increasing the dose of NEF from 100 to 400 mg decreased the elimination rate of NEF, as

indicated by NEF AUCinf, which increased disproportionately with respect to dose. The AUCinf for NEF increased in a 1:4:7 ratio for a 1:2:4 increase in dose. The more than proportional increase in Cmax and AUCinf was apparent when the dose increased from 100 to 200 mg. As the dose increased from 200 to 400 mg, the increase appeared proportional or even less than proportional to dose. Experiments at less than 100 mg NEF dose would be useful to characterize the apparent nonlinearity in pharmacokinetics. However, due to limited assay sensitivity these experiments were not feasible. One dog (dog 2)

U. A. Shukla et al., Ne[azodone & single oral admin (dog) 400

--E

Cmax-O.405xdo.e+15.11 320

Dl

.s

= e

0

lL lL

240

180

0

e

10

oL.----r----r----,----r---.....J 100 200 400 100 o

Dos8 (mg)

!.z::.--

3000 . . . - - - - - - - - - - - - - - - - - - - ,

AUC(INF)-3.63xdose+98.2 2500

.scit

2000

~

1500

-c

1000

u:0' :::>

lL

a..

o

E

500

100

200

400

500

307

AUCinf data for the two metabolites were dose-proportional. The moderately low correlations (R2 s O.n) indicated marked variability in the data. The AUCinf ratios for metabolite:NEF decreased (although not by statistical comparison) at the two higher doses suggesting the relative exposure to the two metabolites in comparison to NEF was reduced. At the low dose, the exposure to the metabolites was approximately twice that of NEF. However, at the two higher doses, the exposure to metabolites was similar to that of the parent compound. This is consistent with saturation of the first pass metabolism of NEF at the higher doses. When compared to NEF Tll2, the overall Tll2 datafor HO-NEF and mCPP suggest that HO-NEF displays formation rate-limited kinetics while mCPP exhibits elimination rate-limited kinetics. In general, the data in this study suggest that pharmacokinetics of NEF are dose-dependent in the dog. The best evidence for dose-depeodent kinetics came from the more than proportional increase in AUCinf for NEF when the dose was increased from 100 to 400 mg. Other parameters such as C max for NEF and the Cmax and AUCinf ratios for metabolite:NEF appeared to show dose-dependency when individual data were examined. However, no statistical significance for dose-