~rr~ibe~ o{ harmar~!~ a r ~ b

Arch Pharm Res Vol 30, No 6, 778-784, 2006

http://apr.psk.or.kr

Nasal Absorption Studies of Granisetron in Rats Using a Validated High-performance Liquid Chromatographic Method with Mass Spectrometric Detection Jong Soo Woo Pharm. R&D Institute, Hanmi Pharm. Co., Ltd., Hwasung 445-913, Korea

(Received September 4, 2006) Granisetron is a selective 5-HT3 receptor antagonist that is used therapeutically for the prevention of vomiting and nausea associated with emetogenic cancer chemotherapy. Although forms of the drug are commercially available for intravenous and oral dosage, there is a need for intranasal delivery formulations in specific patient populations in which the use of these dosage forms may be unfeasible and/or inconvenient. A rapid and specific high-performance liquid chromatography method with mass spectrometric detection (LC-MS) was developed and validated for the analysis of granisetron in plasma after nasal administration in rats. Granisetron was separated in a reverse-phase C-18 column without interference from other components of plasma. This method involves a rapid assay for the determination of granisetron in a small volume of plasma with a run time of 12 min using ondansetron as an internal standard. Data were acquired in the electrospray ionization (ESI) mode with positive ion detection and application of single ion recording (SIR). Granisetron and ondansetron were detected at m/z values of 313.2 and 294.2, respectively. The method described was found to be suitable for the analysis of all samples collected during preclinical pharmacokinetic investigations of granisetron in rats after nasal administration. To date, the first pharmacokinetic study after intranasal administration of granisetron was performed and some pharmacokinetic parameters were presented in this paper. Granisetron was found to be well absorbed through nasal route and the bioavailability of this drug following nasal administration was comparable with that of intravenous administration.

Key words: Nasal administration, Granisetron, LC-MS

INTRODUCTION

CH3

_./

Granisetron (Fig. 1), 1-methyI-N-[(3-endo)-9-methyl-9azabicyclo[3.3.1]non-3-yl]-lH-indazole-3-carboxamide, is a selective 5-HT3 receptor antagonist that may have beneficial therapeutic effects in the treatment of vomiting and nausea associated with initial and repeat courses of emetogenic cancer chemotherapy (Huang et aL, 1998; Sanger and Nelson, 1989; Upward et aL, 1990; Carmichael et aL, 1989). Forms for intravenous and oral dosage of the drug are currently in use. However, these conventional dosage forms might not be appropriate for some patients who are Correspondence to: Jong Soo Woo, Pharrn. R & D Institute, Hanmi Pharm. Co., Ltd., 893-5 Hajeo-ri Paltan-myeon, Hwasung-si Kyounggi-do, 445-913, Korea Tel: 82-31-356-3311, Fax: 82-31-356-7139 E-mail: [email protected]

/

CH3

~~oNN ~ Fig. 1. Structure of granisetron either vomiting or, for some reason, cannot absorb orally administered drugs efficiently after oral administration. One modern strategy to cope with this problem is to develop new formulations for alternative routes of administration. Among these, the nasal delivery system finds convenient applications in specific patient populations that either have difficulty in taking drugs orally or have low oral bioavailability. For the determination of granisetron, a number of HPLC techniques coupled with fluorescence detection (Kudoh et

778

Nasal Administration of Granisetron

779

aL, 1994; Boppana, 1995; Pinguet et aL, 1996) and mass spectrometric analysis (Boppana et aL, 1996) have been reported. However, the sample clean-up procedures of most of these methods consume significant amounts of time. The present paper describes a specific, reliable, and sensitive isocratic reverse-phase HPLC method with mass spectrometric detection to quantify granisetron in rat plasma by employing a combination of published methods (Kudoh etaL, 1994; Boppana, 1995; Pinguet et aL, 1996). Ondansetron was used as an internal standard. This method was validated with respect to accuracy, precision, selectivity, and limits of quantitation (LOQ) and detection (LOD) according to Good Laboratory Practice Guidelines (USP, 1995; Bressolle et aL, 1996). The sample preparation involved simple extraction. The simple sample preparation also makes the method appropriate for analysis of the large number of samples obtained in pharmacokinetic studies for nasal administration in rats.

MATERIALS A N D M E T H O D S Materials and reagents Granisetron HCI and ondansetron HCI were kindly provided by Hanmi Pharmaceutical Company (Seoul, Republic of Korea). Yakuri Pure Chemicals Co., Ltd (Kyoto, Japan) was the supplier of ammonium acetate. Millex-JG 0.2 ~Lm PVDF filters (Millipore Corp., Bedford, MA, U.S.A.) were used for filtration of mobile phase. Water used for LC was obtained from Merck (Darmstadt, Germany). Other chemicals were of reagent grade or HPLC grade.

Animals Male Sprague-Dawley rats, weighing between 230 and 270 g, were used in this study. The animals were obtained from Samtako (Osan, Republic of Korea), housed at 23 + 2~ with a 12:12 h light : dark cycle, and given a SAM#31 pellet diet (Samtako, Osan, Republic of Korea). The rats were kept under standardized conditions, i.e. free access to food and water. Clean cages and fresh water were provided twice a week. The animals were acclimatized to laboratory conditions during the week prior to the experiments.

Animal experiments The surgical procedure and method of administration have been previously described (Hussain et aL, 1980). A diagram of this procedure is shown in Fig. 2. An incision was made in the neck of each rat, and the trachea was cannulated with polyethylene (PE) tubing (0.86 mm i.d., 1.52 mm o.d., Natsume, Tokyo, Japan). A closed-end tube was inserted through an incision in the esophagus and secured against the posterior part of the nasal cavity to prevent any loss of the treatment. The nasopalatine passage

Fig. 2. Diagram of the surgical procedure used for intranasal administrationof drugsto rats was also blocked with an adhesive agent to ensure that there was no drainage of the drug into the mouth. Administration of the drug started 30 min post-operation. A dose of 6 mg/kg granisetron HCI (48 mg/mL solution) was administered into the nasal cavity by means of a microsyringe. For intravenous administration, the same dose was injected into a cannulated jugular vein. Blood samples (500 pL) were collected from the femoral artery at 0, 5, 15, 30, 45, 60, 90, 120, 180, and 300 min after administration. After blood sampling, 500 /~L of saline was supplied via intravenous routes. After centrifugation, the plasma was separated and frozen (-70~ until analysis. The animals were divided into 2 groups with 3 animals in each group.

Preparation of test sample Granisetron HCI was dissolved in saline solution and water to concentrations of 48 mg/mL for intravenous and nasal administration, respectively. The solutions were used within 3 days of preparation in the 3 rats of the treatment group. The volume left in the polyethylene tubes after administration of solutions was considered to be negligible.

Calculations The areas under the concentration-time curves (AUC) were calculated using the linear trapezoidal rule. The calculations on bioavailability were based on data gathered from 0 to 300 min. The intraindividual variability in pharo macokinetic parameters between different groups was

780

estimated by calculating the coefficient of variation (C.V.) (Bjerre et aL, 1996).

Instrumentation

J.S. Woo

of 5, 10, 250, and 1000 ng/mL in plasma and stored at-20 ~ until analysis. Quantification of QC sample concentrations was obtained by interpolation from the equations of the regression lines of the respective calibration curves.

The Waters Alliance HT 2795 Chromatography System (Waters Corp., Milford, MA, U.S.A.) was utilized in this experiment. System control and data processing was performed by MassLynx 3.5 (Waters Corp., Milford, MA, U.S.A.). For separation, a Waters XTerra MS column (150 • 2.1 mm, 3 i~m, Waters Corp., Milford, MA, U.S.A.) was used. MS parameters were as follows; the quantitative determination of granisetron was performed with the Waters ZQ 4000 mass spectrometer (Waters Corp., Milford, MA, U.S.A.). Data were acquired in the electrospray ionization (ESI) mode with positive ion detection and application of single ion recording (SIR). A cone voltage of 36V and a capillary voltage of 3.00 kV were used. The desolvation temperature was maintained at 150~ Granisetron and ondansetron were detected at m/ z values of 313.2 [M+H]§ and 294.2 [M+H] § respectively.

To evaluate the specificity of the method, drug-free plasma was carried through the assay procedure, and the retention times of endogenous compounds in plasma were compared with those of granisetron and ondansetron.

Chromatographic conditions

Recovery

A mixture of water: acetonitrile : 10 mM NH4OAc (pH 3.5) = 27:23:50 (v/v%) was used as mobile phase at a flow rate of 0.2 mL/min. This mobile phase was filtered through a Millex-JG 0.2 pm PVDF filter (Millipore, Bedford, MA, U.S.A.), and was then ultrasonically deaerated prior to use. The column temperature was held at 40~ The autosampler temperature was held at 10~ The injection volume was 10 mL.

Attempts were made to calculate the overall recovery of analytes and the internal standard by separately injecting standard solutions and spiked plasma standards, which contained equivalent concentrations of analytes, onto columns according to the methods described previously. The peak areas of granisetron were lower for the aqueous standards than for the plasma standards, yielding unexpectedly high recoveries for all of the analytes (>100%). This did not vary significantly over the range of concentrations studied. These high recoveries may be the result of poorer retention or poorer recovery of the analytes by the columns from the aqueous samples than from the plasma samples. For blank plasma samples, no peak of the co-eluted biological material was detected; however, the possibility that enhanced ionization of the analyte was induced by interference cannot be ruled out. This was consistent with previously reported studies (Boppana et al., 1996).

Extraction procedure To each 0.2 mL aliquot of the rat plasma in an E-tube, 0.1 mL of 60% acetonitrile containing 500 ng/mL of I.S. (ondansetron) and 0.2 mL of acetonitrile were added in order to precipitate the protein, and the aliquots were then vortexed for 2 min. AfLer centrifuging at 2000 g at room temperature for 5 min, 0.2 mL of supernatants were transferred to an E-tube containing 0.2 mL of 10 mM NH4OAc (pH 3.5); the solutions were then vortexed and centrifuged at 2000 g at room temperature for 5 min. The supematants were each transferred to a clear autosampler vial containing a 0.25 mL insert. A 10 ~L volume of the sample was injected onto the LC-MS system.

Calibration curves and QC samples Stock solutions of granisetron (0.1 mg/mL) and ondansetron (0.1 mg/mL) were prepared in methanol and stored at -20~ Standard solutions were used to spike blank plasma in order to obtain calibration standards at the concentrations of 5, 10, 25, 50, 100, 250, 500, and 1000 ng/mL. QC samples for evaluating the accuracy and precision of the method were prepared at concentrations

Data analysis The peak area of granisetron was used as the assay parameter. Peak-area ratios were plotted against theoretical concentrations. Standard calibration curves were obtained from unweighted least-squares linear regression analysis of the data. The linearity of the method was confirmed using the classical statistical tests, specifically, comparison of intercepts with zero and correlation coefficients (Pinguet et al. , 1996).

Specificity

Precision and accuracy The precision and accuracy of this method were established by repetitive analyses of QC samples in plasma against a calibration curve. Each QC sample was analyzed six consecutive times within one day (n=6) to determine intra-day precision and accuracy, and once a day for six successive days (n=6) at four concentration levels to determine inter-day precision and accuracy. Accuracy was expressed as mean back-calculated concentrations/ nominal concentrations • 100, while the precision was provided by the intra- and inter-day relative standard deviations (RSDs).

Nasal Administration of Granisetron

781

and structure of a drug. Several advantages can be exploited using HPLC-MS: there is no need to derivatize analytes; using mass spectral data, impurities are easily detected, which is of utmost importance for quantification work; and fully resolved peaks are not required for mass spectral analysis. The analytical method presented in this study is based on electrospray ionization (ESI). Three instrument parameters of the LC/MSD's ESI can be tuned to a particular compound (class): cone voltage, which can influence the fragmentation behavior and ion transmission; capillary voltage (mass-dependent with little impact on sensitivity); and desolvation temperature (heat for drying aerosol). Granisetron sensitivity was maximized at a cone voltage setting of 36 V and capillary voltage of 3.00 kV. At a desolvation temperature of 150~ (250 L/h), the mobile phase was evaporated sufficiently without thermal degradation of the analyte. In Fig. 3, a mass spectrum extracted from a total ion chromatogram (TIC) of a standard solution (10 #g/mL) is presented. Fig. 4 shows SIR chromatograms (SIR, m/z 313.2) of granisetron and I.S.

Determination of the limits of quantitation and detection The sensitivity of the method was evaluated by determining the limit of detection (LOD) and the limit of quantitation (LOQ), According to ICH guideline (ICH Topic Q2B), LOD was defined as 3.3 • cdS and LOQ was defined as 10 x o/S based on the 'standard deviation of the response and slope' obtained using the calibration curve (ICH, 1996). The standard deviation of the y-intercepts of regression lines was used as c~(the standard deviation of the response) and S = the slope of the calibration curve.

Stability studies In stock solutions, granisetmn was stable over 5 days at 20~ 7 days at 10~ and 30 days at -20~ No significant deviation was found from the nominal value. It was previously reported that granisetron was stable in plasma when frozen at -20~ for two months (Pinguet eta/., 1996), which was consistent with this study. In this study, all plasma samples were analyzed less than one week after sampling. The stability of granisetmn in an autosampler vial during a 48-h period at room temperature and at 10~ was investigated. At all temperatures tested, no significant differences between nominal values and concentrations of granisetron after storage were found.

Method validation

RESULTS A N D DISCUSSION

The optimized methods were validated for granisetron determination using ondansetron as internal standard. The validation requires the assessment of reproducibility, detector response linearity with sample concentration, sensitivity, and accuracy.

LC-MS procedure

Linearity

HPLC-MS provides information on the molecular weight

Detector response linearity was determined by preparing

G10pprn4 155 (4 686) Cra (155 157) 3132

3. Scan ES+ 3 92e7

100-

%

/314 2

3353

153

006"3~ "~6' 3s 3~0 3~5 3~0 3~5 ~0 ~ .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

350 355 360 365 370 37~ 380 305 3o0 39~ 400 .

, .. . .. .

i

. . . .

Mass (m/z) Fig. 3. Massspectrum of granisetron

~ . . . .

J . . . .

i

. . . .

i

. . . .

i

. . . J

. . . .

f

. . . .

i

. . . .

i f'n/z

782

J.S. Woo

2 5 0 3 . 7 s i n ( M n . 3"43)

1: S I R o f 1 C h a n n e l

100-

4.60

ES+ TIC

(A)

agoe5

%

-0

.... , .... , .... 2 5 0 n ! , S m ( M n , 3",(3)

k

, ....

, ....

, ....

~ ....

, ....

, ....

i ....

, ....

100.

, ....

, ....

, ....

, ....

, ....

, ....

, ....

5.50

i ....

J ....

i ....

i .... , .... J .... , 2: S i R o f 1 C h a n n e l E S + TIC

(B)

1.o8e8

m%

2.42

_o4.m ...................................................................................................................... I O0

200

3.00

400

500

600

700

800

900

I0.00

Time

11 O0

E 4 ST ~ ' ( M n . 3 x 3 )

1200

1 SIR of 1 Channel

100

4.~3

ES+ TIC 1.04e6

(c) %

ol . . . .

,

. . . .

i

. . . .

,

. . . .

t

. . . .

,

. . . .

i

. . . .

,

. . . .

E4 St I (Mn, 3x3)

/ L i

. . . .

,

. . . .

i

. . . .

100

i

. . . .

,

. . . .

i

. . . .

~

. . . .

i

. . . .

i

. . . .

i

. . . .

,

. . . .

i

. . . .

,

. . . .

i

. . . .

2

=

. . . .

Of 1

SIR

i

. . . .

,

Channel

5.52

ES+ TIC 1.06e6

j

(D)

m %

2.42

0

' ' ' l . . . .

l . . . .

1.00

' . . . .

I

. . . .

2.00

}

. . . .

~

. . . .

3.00

1

. . . .

l . . . .

4.00

'

l . . . .

l

5,00

~

. . . .

6.00

l . . . .

}

. . . .

7.00

l . . . .

' . . . .

8.00

l . . . .

I . . . .

9 O0

l . . . .

l . . . .

10 00

l . . . .

~

. . . .

11.00

I

. . . .

,Time 12.00

Retention time(min) Fig. 4. Chromatograms of granisetron in rat plasma. (a) and (b) 250 ng/mL, (c) and (d) 45 rain after nasal administration. The retention times of granisetron and the internal standard are 4.6 and 5.5 rain, respectively. (a) and (c) ' m/z = 313.2 for granisetron, (b) and (d) ' m/z = 294.2 for

internal standard. five calibration samples covering the concentration range of 5-1000 ng/mL. Each sample was injected in triplicate. The coefficient of determination and calibration curves are shown in Table I. Linearity was obtained with an average coefficient of determination (R 2) higher than 0.999, over the calibration range of the analyte. The limit of detection (LOD) is

estimated to be 3.3 x ~/S. The LOD was determined by injecting standard solutions of various concentrations. The estimated LOD was 1.55 ng/mL, and the lower limit of quantitation (LOQ) was 4.71 ng/mL (Table I). Precision and a c c u r a c y

In Table II, the results for accuracy and intra-day and

Nasal Administration of Granisetron

783

Table I. Regressiondata for the calibration curves

1o~o.o

LC-MS Range (ng/mL) Line Coefficientof determination LOD (ng/mL) LOQ (ng/mL)

---iv

E

5-1000 y = 0.00249x+ 0.00039 0.9997 1.55 4.71

1000.0

e-

I~

,~_=

-~-- Nasal

100.0

4-

t0.0

Table II. Assessment of accuracy and precision Nominalconcentration Concentration Precision Accuracy (ng/mL) Found (rig/mE) (%) (%)

E m

1.0

m

Intra-day (n=6) 5 10 250 1000

5.06 10.83 245.7 1006.8

12.7 4.5 1.7 0.5

101.2 108.3 98.3 100.7

5.27 11.21 249.8 1011.7

13.5 8.2 3.1 0.8

105.4 112.1 99.9 101.2

t

a

t

2

3 Time

=

4

5

(hr)

Fig, 5. Plasma granisetronconcentration after nasal and intravenous administration of 6 mg/kg to rats (n=3). The points represent mean • S.D.

Inter-day (n=6) 5 10 250 1000

i

0.1

n = numberof replicates. inter-day precisions for QC samples are presented. The Jntra-day precision showed RSDs of 0.5 to 12.7%. The inter-day precision RSDs varied from 0.8 to 13.5%. The intra-day accuracy ranged from 98.3 to 108.3%. The interday accuracy ranged from 99.9 to 112.1%. The stability of test solutions was assessed immediately after preparation and 48 h later at room temperature and at 10~ The stabilities of granisetron and intemal standard at the tested concentrations were good at room temperature for at least 48 h, with no observed change in concentration.

In vivo study Fig. 5 shows the plasma granisetron concentrations versus time curves after nasal and intravenous administration of granisetron. The bioavailability of granisetron following nasal administration of 6 mg/kg appeared to be comparable to that following intravenous administration of the same dose (Table III). As shown in Fig. 5, granisetron

was readily and rapidly absorbed through the nasal mucosa of the rat. The peak plasma level was attained within 15 min of delivering the dose into the nasal cavity. The elimination half-life of granisetron after nasal administration was also similar to that observed after intravenous administration. The bioavailability of a drug following nasal administration is expected to increase because it evades hepatic first-pass metabolism. In healthy individuals, the time to reach peak concentration is 3.0 h after oral administration. Oral formulation of granisetron, thus, should be given to patients up to 1 h prior to administration of chemotherapy (Hussain et aL, 2000; Blum et aL, 2003; PDR, 2005). CONCLUSION Granisetron was found to be well absorbed through nasal route in this study. The bioavailability of this drug following nasal administration was comparable with that of intravenous administration. The LC-MS method presented in this study involves a rapid assay for the determination of granisetron in a small volume of plasma with a run time of 12 min. The volume injected is equivalent to as little as 2 ~L of plasma. The separation between granisetron and endogenous substances was satisfactory. Moreover, the

Table III. Pharmacokinetic parameters after intranasal (i.n.) administration of granisetron using intravenous (i.v.) administration as a reference Adm. Route

Dose (mg/kg)

Cma,(ng/mL)

tmax(h)

AUGust(ng h/mL)

t~2 (hr)

Fabs(%)

i.n. (n=3)

6

520.2 + 91.3

0.25 + 0.00

609.5 • 6.6

1.7 • 0.2

101.0 • 1.1

i.v. (n=3)

6

1352.1 + 223.5

603.2 _+47.2

1.0 _+0.0

Data are expressedas mean + S.D.

784

specificity from drugs that may be co-administered is good. The LC-MS method used in this study was found to be suitable for the analysis of all samples collected during preclinical pharmacokinetic investigations of granisetron in rats af{er nasal administration. In conclusion, the intranasal route of granisetron administration could overcome the limitations of oral delivery of granisetron to patients undergoing emetogenic cancer chemotherapy. The intranasal route of administration holds promise as an alternative method to prevent nausea and vomiting associated with such therapy.

REFERENCES Bjerre, C., BjSrk, E., and Camber, O., Bioavailability of the sedative propiomazine after nasal administration in rats. Int. J. Pharm., 144, 217-224 (1996). Blum, R. A., Majumdar, A., McCrea, J., Busillo, J., Orlowski, L. H., Panebianco, D., Hesney, M., Petty, K. J., Goldberg, M. R., Murphy, M. G, Gottesdiener, K. M., Hustad, C. M., Lates, C., Kraft, W. K., Van Buren, S., Waldman, S. A., and Greenberg, H. E., Effects of aprepitant on the pharmacokinetics of ondansetron and granisetron in healthy subjects. Clin. Ther., 25, 1407-1419 (2003). Boppana, V. K., Simultaneous determination of granisetron and its 7-hydroxy metabolite in human plasma by reversed-phase high-performance liquid chromatography utilizing fluorescence and electrochemical detection. J. Chromatogr. A., 692, 195202 (1995). Boppana, M K., Miller-Stein, C., and Schaefer, W. H., Direct plasma liquid chromatographic-tandem mass spectrometric analysis of granisetron and its 7-hydroxy metabolite utilizing internal surface reversed-phase guard columns and automated column switching devices. J. Chromatogr. B., 678, 227-236 (1996). Bressolle, F., Bromet-Petit, M., and Audran, M., Validation of liquid chromatographic and gas chromatographic methods Applications to pharmacokinetics. J. Chromatogr. B., 686, 310 (1996).

J.S. Woo

Carmichael, J., Cantwell, B. M., Edwards, C. M., Zussman, B. D., Thompson, S., Rapeport, W. G, and Harris, A. L., A pharmacokinetic study of granisetron (BRL 43694A), a selective 5-HT3 receptor antagonist: correlation with antiemetic response. Cancer Chemother. PharmacoL, 24, 45-49 (1989). Huang, C.-T., Chen, K.-C., Chen, C.-F., and Tsai, T.-H., Simultaneous measurement of blood and brain microdialysates of granisetron in rat by high-performance liquid chromatography with fluorescence detection. J. Chromatogr. B., 716,251-255 (1998). Hussain, A. A., Dakkuri, A., and Itoh, S., Nasal absorption of ondansetron in rats: an alternative route of drug delivery. Cancer Chemother. PharmacoL, 45, 432-434 (2000). Hussain, A. A., Hirai, S., and Bawarshi, R., Nasal absorption of propranolol from different dosage forms by rats and dogs. J. Pharm. Sci., 69, 1411-1413 (1980). International Conference on Harmonization, Note for Guidance on Validation of Analytical Procedures: Methodology, Committee for Proprietary Medical Products, CPMP/ICHI281195, Approval 18 December 1996. Kudoh, S., Sato T., Okada, H., Kumakura, H., and Nakamura, H., Simultaneous determination of granisetron and 7hydroxygranisetron in human plasma by high-performance liquid chromatography with fluorescence detection. J. Chromatogr. B., 660, 205-210 (1994). Physicians' Desk Reference, 59th ed (2005) Medical Economics, Montvale, New Jersey, pp. 2901-2903 Sanger, G J. and Nelson, D. R., Selective and functional 5hydroxytryptamine3 receptor antagonism by BRL 43694 (granisetron). Eur. J. Pharmacol., 159, 113-124 (1989). United States Pharmacopoeia 23, Section 1225, Validation of Compendial Methods, United States Pharmacopoeia Convention, Inc., Rockville, MD, 1995. Upward, J. W., Arnold, B. D., Link, C., Pierce, D. M., Allen, A., and Tasker, T. C., The clinical pharmacology of granisetron (BRL 43694), a novel specific 5-HT3 antagonist. Eur. J. Cancer, 26, $12-$15 (1990).