Development of a Liquid Chromatographic Method for Bioanalytical Applications with Sildenafil

(1)

www.elsevier.com / locate / chromb

D

evelopment of a liquid chromatographic method for bioanalytical

applications with sildenafil

a a b a a ,

_*

Ming-Thau Sheu , An-Bang Wu , Geng-Cheng Yeh , Angel Hsia , Hsiu-O Ho

a

Graduate Institute of Pharmaceutical Sciences, Taipei Medical University, 250 Wu-Hsing Street, Taipei 110, Taiwan, PR China

b

Department of Pediatrics, Taipei Medical University, 250 Wu-Hsing Street, Taipei 110, Taiwan, PR China Received 6 January 2003; accepted 11 March 2003

Abstract

An improved HPLC method was developed for the determination of sildenafil concentrations in plasma. Analysis of sildenafil in plasma samples was simplified by utilizing a one-step liquid–liquid extraction after alkaline treatment of only 1 ml of plasma. The lower limit of quantitation was 10 ng / ml with a coefficient of variation of less than 20%. A linear range was found to exist from 10 to 1000 ng / ml. This HPLC method was validated with precisions (coefficient of variation, C.V.) for inter- and intra-day runs of 0.41–11.15% and 0.36–8.05%, respectively, and accuracies (the relative error of the mean, REM) for inter- and intra-day runs of 28.72–6.81% and 0.41–11.15%, respectively. A bioavailability study of sildenafil was performed on one normal healthy male volunteer by analyzing sildenafil plasma concentrations with this validated HPLC method. Results demonstrated that this HPLC method is appropriate for applications to bioavailability studies of sildenafil. In addition, an example of the influence of the co-administration of grapefruit juice on sildenafil pharmacokinetics in a healthy volunteer is presented.

Keywords: Sildenafil

1

. Introduction after oral administration, with absolute

bioavail-ability of about 40%. Its pharmacokinetics is dose-Sildenafil, ((1-[4-ethoxy-3-(6,7-dihydro-1-methyl- proportional over the recommended dose range. It is 7-oxo-3-propyl-1H-pyrazolo-[4,3-d] pyrimidin-5-yl)- eliminated predominantly by hepatic metabolism phenylsulphonyl]-4-methylpiperazine) (Fig. 1), is a (mainly by cytochrome P450 3A4) after absorption potent and selective inhibitor of cGMP (type V)- and is converted to an active metabolite with prop-specific phosphodiesterase which is capable of en- erties similar to itself. Sildenafil and this metabolite hancing relaxation of the penile corpus cavernosum have terminal half-lives of approximately 4–5 h. The and therefore has the potential to improve penile maximum observed plasma concentration of sil-erectile function [1]. Sildenafil is rapidly absorbed denafil is reached within 30–120 (median, 60) min of an oral dosing in a fasting state. Less than 2% of the administered dose is excreted in the urine as *Corresponding author. Tel.: 2377-1942; fax:

1886-2-sildenafil or its active metabolite [2,3]. 2377-1942.

E-mail address: [email protected](H.-O Ho). CYP 3A is the largest subfamily of CYP enzymes

(2)

on-line dialysis as well as a switching valve for trace enrichment, has also been used [13]. Plasma con-centrations of sildenafil and N-demethylsildenafil can be determined up to 24 h post-administration by use of liquid chromatography in tandem with mass spectrometry [9]. Radiolabeled sildenafil has also been used to monitor the pharmacokinetics of plasma and urine in the mouse, rat, rabbit, dog, and human [14]. However, normal laboratories with limited financial resources are less likely to have access to special techniques and complicated procedures, so a simpler assay method must be developed.

A reversed-phase HPLC system was recently developed by Liaw et al. [15] for the determination of sildenafil concentrations in an in vitro transdermal

Fig. 1. Structure of sildenafil. _{permeation study. The lower limit of a quantifiable}

concentration was reported to be 5 ng / ml for analyz-ing sildenafil samples collected in transdermal per-meation studies. But difficulties may arise in analyz-expressed in the human liver and gastrointestinal ing biological samples such as plasma. Losses may tract and is involved in the metabolism of many occur especially during treatments of extracts for clinically used drugs and other chemicals. Concomi- various sample preparation procedures. Therefore, tant intake of grapefruit juice causes inhibition of there is still a demand for development of a simpler CYP3A4-mediated first-pass metabolism of many and more-sensitive HPLC method for analysis of drugs in the gut wall and thereby increases the oral sildenafil in biological fluids such as plasma. The bioavailability of these agents[4–6].Recently many purpose of the present study was to compare and lines of evidence point to the intestine, rather than improve the assay method for sildenafil in plasma the liver, as the major site of this interaction in vivo using a reversed-phase HPLC system. This validated [7]. Lee and Min [8] observed that grapefruit juice HPLC method was then applied in a bioavailability appeared to increase the Cmax of sildenafil by 42% study of sildenafil in a human following a single oral without a significant change in the AUC from a administration of sildenafil with water or grapefruit single elderly male subject. Jetter et al. [9] reported juice.

that the AUC0 – ` of sildenafil increased 1.23-fold, and a trend toward prolongation of Tmax after grapefruit juice intake was observed. Cmax did not

differ significantly. Pharmacokinetic profiling of 2 . Experimental volunteers taking sildenafil with grapefruit juice was

inconsistent. The interaction of grapefruit juice with 2 .1. Materials and methods sildenafil is still worth examining.

There are several studies in the literature reporting 2 .1.1. Drug and methods

the determination of sildenafil in plasma samples Standard sildenafil citrate was obtained from Trans using an HPLC method [10,11]. An automated American Chemicals (San Diego, CA, USA). The HPLC method using narrow-bore column switching internal standard, butylparaben, was purchased from has also been employed for the simultaneous de- Sigma (St. Louis, MO, USA). Methanol and acetoni-termination of sildenafil and its active metabolite trile for liquid chromatography were HPLC grade from human plasma samples [12]. In addition, the and were obtained from Merck (Darmstadt, Ger-automated sequential trace enrichment of a dialysate many). All other reagents used were reagent grade or (ASTED) system, incorporating a dialysis system for better.

(3)

2

.2. Instrumentation plasma samples containing sildenafil at concentra-tions of 10, 20, 50, 100, 200, 500, and 1000 ng / ml A high-performance liquid chromatographic were supplemented with 20 mg butylparaben and (HPLC) system equipped with a pump (Jasco PU- extracted and analyzed as described above. The peak 980 Intelligent HPLC pump; Jasco, Tokyo, Japan) area ratio (PAR) of sildenafil to butylparaben was and an autosampler (Jasco AS-950-10 Intelligent measured, and a calibration curve was obtained from Sampler) were used. The eluent was detected with a the least-squares linear regression of the PAR versus Jasco UV-975 UV–Vis detector at 230 nm. A 4.63 spiked concentrations. The regression line was used 150-mm (I.D.) Inertsil 5 ODS-2 column to calculate concentrations of sildenafil in the

un-

known plasma samples based on the PAR. (Veracopak ) with a particle size of 5 mm and a

mobile phase consisting of acetonitrile and a 30-mM 2

.5. Validation of the assay method potassium dihydrogen phosphate buffer solution (pH

6.0 adjusted with 1 N NaOH) at a 55:45 (v / v) ratio

Several pre-dose human plasma samples from was used. The flow-rate was set at 0.5 ml / min. This

different subjects were tested for the presence of method was found to be selective, precise, and linear

interfering compounds. The coefficients of variation over a concentration range of 10–1000 ng / ml. All

and relative errors of the mean were used to validate mobile phase solutions were filtered and degassed

the precision and accuracy of the intra- and inter-day ultrasonically before use. The HPLC system was

assays by determining standard samples of sildenafil controlled by a PC workstation using

Chromatog-in plasma. For Chromatog-inter-day validation, five sets of raphy Data Station software (SISC, Taiwan) installed

control samples at seven different concentrations on it.

(10–1000 ng / ml) were evaluated on five different days (with seven standard curves being constructed). 2

.3. Internal standard solution and sample

For intra-day validation, five sets of controls at seven preparation

different drug concentrations were assayed with one standard curve on the same run.

Plasma sample preparation and the extraction method are described step by step as follows. The

2

.6. Blood sample collection and processing plasma sample (1 ml) was spiked with 0.1 ml of the

internal standard (butylparaben, 20 mg / ml in

metha-A bioavailability study (approved by the Ethics nol) solution and 0.1 ml of a NaOH solution (1 N).

Committee of Taipei Medical University Hospital) After vortex-mixing thoroughly for 5 s, the mixture

was conducted on one subject who received a single was extracted with 3 ml of ethyl acetate by

vortex-25-mg dose of sildenafil citrate with 250 ml of either mixing for 5 min, and then centrifuged at 2950 g for

water or grapefruit juice. After dosing, heparinized another 10 min. The supernatant (organic phase) was

venous blood samples (|10 ml) were collected by transferred to another clean glass tube and

evapo-means of an indwelling venous cannula of the cubital rated under a stream of nitrogen gas at 40 8C until

vein according to a predetermined time schedule, completely dry. Then, 0.2 ml of the mobile phase

which included a blank sample just prior to dosing was added to dissolve the residue, and 0.1 ml was

and then at 0.5, 1, 1.5, 2, 2.5, 3, 4, 6, and 8 h after automatically injected into the HPLC system for

dosing. Plasma was separated immediately by cen-analysis.

trifugation at 1690 g for 10 min, then transferred to suitably labeled tubes, and stored at 225 8C until the 2

.4. Quantification and calibration curve assay. preparation

2

.7. Analysis of pharmacokinetic data To examine the linearity of the assay, calibration

curves for sildenafil at concentrations ranging from The following parameters were assessed for the 10 to 1000 ng / ml in plasma were prepared. Standard period of 0–8 h: the area under the plasma

(4)

con-centration–time curves from time zero to the last with low background were observed. The peaks of measurable sildenafil sample time and to infinity interest were well resolved, and there was no inter-(AUC0 – last and AUC0 – `), maximum concentration ference from endogenous plasma substances. Also, (Cmax), and time to maximum concentration (Tmax). the symmetry of both peaks (sildenafil and All pharmacokinetic variables were calculated by butylparaben) was clearly indicated.

non-compartmental methods. Cmax and Tmax were The intra- and inter-day validations for assaying obtained directly from the concentration–time curve sildenafil in plasma samples are shown in Tables 1 data. The area under the concentration–time curve and 2. These data show good precision and high from time zero (predose) to time of last quantifiable accuracy of the analysis. The linearity of the

cali-2

concentration (AUC0 – last) was calculated using the bration curve of sildenafil was well correlated (r . linear trapezoidal method. 0.999) within a range of 10–1000 ng / ml for intra-and inter-day assays. All data show the excellent reproducibility of the sample analysis. Since the

3

. Results and discussion coefficient of variation for 10 ng / ml was less than 20%, it was set as the lower limit of quantitation The HPLC method carried out in this study was an (LLOQ). The coefficients of variation for the re-attempt at developing a simple chromatographic maining standard concentrations were all less than system capable of eluting and resolving sildenafil 15%, which complies with the requirements of assay from human plasma and which complies with gener- validation. Preferably, only 1 ml of plasma sample is al requirements for system suitability. Preliminary necessary to achieve such a low limit of quantitation. investigations were directed towards the effect of As expected, the limit of detection (LOD) was even certain variables on the suitability of the method. lower than one-third of the LLOQ.

The parameters assessed include the type and quanti- The method was applied to a bioavailability study ty of the organic modifier, the column, the salt for examining the influence of the co-administration concentration, and the pH of the mobile phase. The of grapefruit juice on the absorption of sildenafil. results show that the Inertsil 5 ODS-2 column system The absolute relative errors of the mean in each was suitable for the determination of sildenafil from sample analysis run for QC samples were between human plasma because of its excellent resolution, 0.09 and 4.45%, indicating that the stability of appropriate retention time, and good sensitivity (Fig. sildenafil in plasma during storage periods was

2). acceptable. Typical chromatographs of analyzed

After several trials in this study, it was found that plasma samples are shown in Fig. 4.The suitability the present method using one-step liquid–liquid of this method was proven in this bioavailability extraction of sildenafil in plasma after alkaline study of sildenafil after oral administration of sil-treatment was satisfactory. Using the peak area of denafil (25 mg) in one healthy male volunteer with the sildenafil sample at the same injection amount either water or grapefruit juice.

without extraction as 100%, it was found that the Fig. 5 displays the sildenafil plasma concentra-recovery of sildenafil added to human plasma was tion–time profiles in the volunteer after orally ad-almost quantitative (at least 75%) with minimal ministration of the sildenafil formulation with 250 ml interference. This indicates that the method is re- of either water or grapefruit juice. Grapefruit juice producible and suitable for the analysis of plasma changed the area under the sildenafil plasma con-samples. The method provides a simple and practical centration–time curve from time zero to infinity way to process plasma samples containing sildenafil [AUC0 – `] from 63.06 to 169.14 ng?h / ml, corre-with almost quantitative recovery. sponding to a 168% increase. A trend toward Fig. 3 shows typical HPLC chromatograms of prolongation of the Tmax of sildenafil after grapefruit sample analysis. No interfering peaks were observed juice intake from 0.50 to 0.75 h was observed. for drug-free human plasma. The retention times of Grapefruit juice increased the Cmax of sildenafil by sildenafil and butylparaben were around 7.6 and 11.3 approximately 35% from 50.53 to 68.02 ng / ml. min, respectively. Good separation and baselines These results indicate that co-administration of

(5)

sil-Fig. 2. HPLC chromatograms using the Inertsil ODS column. (A) Blank plasma; (B) the internal standard, butylparaben; (C) sildenafil (1000 ng / ml) and butylparaben (2 mg / ml).

(6)

enzyme [7]. Therefore grapefruit juice can increase sildenafil bioavailability and thus tends to delay sildenafil absorption.

4

. Conclusions

In conclusion, the HPLC assay method developed in this study using a reversed-phase system was proven to be acceptable for assaying sildenafil concentrations in plasma. High precision and

accura-Fig. 3. HPLC chromatograms of standard plasma concentrations

of sildenafil (from 10 to 1000 ng / ml) and butylparaben (2 mg / cy with minimal interference and peaks with high

ml). _{symmetry were demonstrated. A one-step liquid–}

denafil with grapefruit juice seems to increase the liquid extraction further provides a simple and Tmax, Cmax, and AUC. The grapefruit juice effect has practical way to process plasma samples containing been postulated to be the result of competitive sildenafil with almost quantitative recovery. This inhibition of CYP3A4 in the enterocytes lining the method is appropriate for applications to bioavail-small intestine, because grapefruit juice has been ability studies of the oral administration of sildenafil shown to contain 69,79-dihydroxybergamottin (DHB) and to examine the influence of grapefruit juice on which is capable of competitively inhibiting this the pharmacokinetic profiling of sildenafil.

T able 1

Precision and accuracy of intra-day validation

a

Concentration Y1 Y2 Y3 Y4 Y5 Mean SD C.V. REM

(ng / ml) (%) (%) 10 9.74 11.13 11.07 9.30 9.93 10.23 0.82 8.05 2.33 20 20.37 19.19 20.63 22.92 22.11 21.04 1.48 7.03 5.21 50 51.95 48.21 51.43 57.91 56.82 53.26 4.03 7.56 6.53 100 107.07 114.63 93.61 102.83 103.63 104.35 7.60 7.29 4.35 200 198.05 192.59 203.91 194.46 182.15 194.23 8.02 4.13 2.88 500 486.99 489.15 499.30 485.18 505.92 493.31 8.92 1.81 1.34 1000 1005.83 1005.13 1000.05 1007.41 999.45 1003.57 3.59 0.36 0.36

SD, standard deviation; C.V., coefficient of variation; REM, relative error of the mean.

a

Number of the replication.

T able 2

Precision and accuracy of inter-day validation

a

Concentration Y1 Y2 Y3 Y4 Y5 Mean SD C.V. REM

(ng / ml) (%) (%) 10 8.39 10.49 8.62 8.21 9.93 9.13 1.02 11.15 28.72 20 19.00 18.43 19.16 23.47 22.11 20.43 2.22 10.86 2.17 50 51.35 49.75 54.95 49.07 56.82 52.39 3.36 6.42 4.78 100 106.67 104.99 109.99 108.79 103.63 106.81 2.62 2.45 6.81 200 188.51 192.07 196.04 186.66 182.15 189.09 5.28 2.79 25.46 500 509.51 506.56 484.20 504.63 505.92 502.16 10.20 2.03 0.43 1000 996.56 997.71 1007.03 999.18 999.45 999.99 4.11 0.41 0.00

SD, standard deviation; C.V., coefficient of variation; REM, relative error of the mean.

a

(7)

Fig. 4. Representative results of HPLC analysis of plasma samples from the bioavailability study.

R

eferences

[1] M . Boolell, M.J. Allen, S.A. Ballard, S. Geti-Attee, G.J. Muirhead, A.M. Naylor, I.H. Osterloh, C. Jingell, Int. J. Impot. Res. 8 (1996) 47.

[2] C .J. Muirhead, M.J. Allen, G.C. James, J. Pearson, D.J. Rance, A.C. Houston, P.M. Dewland, Br. J. Clin. Pharmacol. 42 (1996) 268.

[3] D .J. Rance, F.M. Atkinson, G.C. James, F.M. Maclntyre, D.K. Walker, P. Wastall, Br. J. Clin. Pharmacol. 42 (1996) 267.

[4] D .G. Baily, J.M. Arnold, C. Muno, Clin. Pharmacol. Ther. 53 (1993) 637.

[5] M .P. Ducharme, L.H. Warbasse, D.J. Edwards, Clin. Phar-macol. Ther. 57 (1995) 485.

[6] Y .M. Ku, D.I. Min, M. Flanigan, Clin. Pharmacol. Ther. 38 (1998) 959.

[7] P . Schmiedlin-Ren, D.J. Edwards, M.E. Fitzsimmons, K. He, Fig. 5. Sildenafil plasma concentration–time profile of one vol- K.S. Lown, P.B. Watkins, Drug Metab. Dispos. 25 (1997)

(8)

[8] M . Lee, D.I. Min, Ther. Drug Monit. 21 (2001) 21. [13] D .J. Rance, F.M. Atkinson, G.C. James, F.M. Maclntyre, [9] A . Jetter, M. Kinzig-Schippers, M. Walchner-Bonjean, U. D.K. Walker, P. Wastall, Br. J. Clin. Pharmacol. 42 (1996)

¨

Hering, J. Bulitta, P. Schreiner, F. Sorgel, U. Fuhr, Clin. 267.

Pharmacol. Ther. 58 (1995) 127. [14] D .K. Walker, M.J. Ackland, G.C. James, G.J. Muirhead, D.J. [10] J .D.H. Cooper, D.C. Muirhead, J.E. Taylor, R.P. Baker, J. Ragnce, P. Waastall, P.A. Wright, Xenobiotica 29 (1999) 297.

Chromatogr. B 701 (1997) 87. [15] J . Liaw, T.W. Chang, J. Chromatogr. B 765 (2001) 161. [11] C .K. Jeong, H.Y. Lee, M.S. Jang, W.B. Kim, H.S. Lee, J.

Chromatogr. B 752 (2001) 141.

[12] G .J. Muirhead, M.J. Allen, G.C. James, J. Pearson, D.J. Rance, A.C. Houston, P.M. Dewland, Br. J. Clin. Pharmacol. 42 (1996) 268.