A Rapid HPLC Assay for the Simultaneous Determination of

ANALYTICAL SCIENCES SEPTEMBER 2004, VOL. 20 2004 © The Japan Society for Analytical Chemistry

1307

A Rapid HPLC Assay for the Simultaneous Determination of Propafenone and Its Major Metabolites in Human Serum Minoo AFSHAR and Mohammadreza ROUINI† Biopharmaceutics and Pharmacokinetics Division, Department of Pharmaceutics, Faculty of Pharmacy, Tehran University of Medical Sciences, 14155-6451, Tehran, Iran

A rapid and specific HPLC method has been developed and validated for the simultaneous determination of propafenone, an antiarrhythmic agent, and its major metabolites in human serum. The sample preparation was a simple deproteinization with a mixture of ZnSO4 and methanol, yielding almost 100% recoveries of three compounds. Separation was developed on a reverse-phase tracer excel C18 column (25 × 0.46 cm i.d., 5 µm), using an acetonitrile– phosphate buffer gradient at a flow rate of 1.7 ml min–1, and UV detection of 210 nm. The calibration curves were linear (r2 > 0.999) in the concentration range of 10 – 750 ng ml–1. The lower limit of quantification was 10 ng ml–1 for all of the compounds studied. The within and between day precisions in the measurement of QC samples at four tested concentrations were in the range of 1.4 – 8.1% and 4.2 – 11.5% RSD, respectively. The developed procedure was applied to assess the pharmacokinetics of propafenone and its major metabolites following administration of a single 300 mg oral dose of propafenone hydrochloride to three healthy male volunteers. (Received April 6, 2004; Accepted May 31, 2004)

Introduction Propafenone is a potent antiarrhythmic drug, which is widely used in the treatment of ventricular and supraventricular arrhythmias.1 It blocks the fast inward sodium current in all cardiac and other excitable tissues, like the central nervous system. Propafenone also has some β-blocking action, much less than that of propranolol, and a weak calcium-channelblocking effect.2 Propafenone undergoes extensive first-pass metabolism by the liver to form several metabolites. The main metabolic pathway of propafenone is ring hydroxylation to 5-OH-propafenone, being primarily mediated by CYP2D6. This pathway is polymorphically expressed in humans and is under genetic control, co-segregating with a well-described debrisoquine oxidation polymorphism,3 such that poor metabolizers can be distinguished from extensive metabolizers.4 An additional route of propafenone metabolism is N-demethylation to N-despropylpropafenone, which is primarily mediated via CYP3A4 and CYP1A2.5 Five hydroxy-propafenone has been shown to exert pharmacologic activity comparable to that of the Although N-despropylpropafenone exerted parent drug.6 antiarrhythmic activity in animal experiments,7 relatively little is known about its pharmacokinetics and pharmacodynamics in humans. The metabolic pathway of propafenone is illustrated in Fig. 1. The specific metabolism of propafenone, its potential drug-drug interactions and its widespread use have stimulated efforts to develop routine assays for this drug and its metabolites in human serum. Several chromatographic methods have been reported for the determination of propafenone and † To whom correspondence should be addressed. E-mail: [email protected]

5-OH-propafenone in biological matrices, including highperformance liquid chromatography (HPLC)8–15,18 and gas chromatography.16,17 For HPLC assays, different modes of detection have been employed, including UV8–15 and fluorescence with precolumn derivatization.18 Most of these methods have only been validated for propafenone, itself, or propafenone and 5-OH-propafenone. The resolution of N-despropylpropafenone, which is probably co-eluted under reported procedures, was rarely investigated. Among the methods described so far, only a few allow for the determination of the serum or plasma concentrations of propafenone and its main metabolites in a single run.19–21 The

Fig. 1

Metabolic pathway of propafenone.

1308

ANALYTICAL SCIENCES SEPTEMBER 2004, VOL. 20

procedure, based on high-performance liquid chromatography with mass spectrometry, is highly sensitive and specific, but involves equipments not generally available in a clinical set.21 The major drawbacks of the mentioned methods are using complicated labor- and time-intensive sample-processing methods with several liquid-liquid extraction steps or solidphase extraction as sample preparation procedures,19–21 long elution times,19 poor sensitivity and recovery specially for N-despropylpropafenone.20 This paper describes a rapid and specific HPLC method with UV detection for the simultaneous analysis of propafenone and its two major metabolites using direct protein precipitation. The developed method was successfully applied to assess the pharmacokinetics of propafenone and its two major metabolites following the oral administration of a single 300 mg dose of propafenone hydrochloride to three healthy male volunteers.

centrifuge. A clear supernatant was transferred into another eppendrof tube, and 200 µl of this was injected on to an HPLC column.

Experimental Materials Propafenone, 5-OH-propafenone hydrochloride and N-despropylpropafenone fumarate (Knoll, Ludwigshafen, Germany) were kindly donated by Dr. Ute Hofmann (Dr. Margarete Fischer-Bosch-Institut fur Klinische Pharmakologie, Stuttgart, Germany). Acetonitrile (HPLC-grade), potassium dihydrogenphosphate, and orthophosphoric acid (85%), zinc sulfate and methanol (HPLC-grade) were purchased from Merck (Darmstadt, Germany). Double-distilled water was used throughout the study. Preparation of standard solutions Standard stock solutions (100 µg ml–1 as free base) of each product were prepared separately in methanol and stored at –20˚C. Working solutions were prepared daily from these stock solutions by dilution with distilled water. Apparatus and chromatographic condition The chromatographic apparatus consisted of a low-pressure gradient HPLC pump, a UV variable-wavelength detector and an online degasser, all from Knauer (Berlin, Germany). A Reodyne Model 7725i injector with a 200 µl loop was used. The data were acquired and processed by means of Urochrome chromatography software (Knauer, Berlin, Germany). Chromatographic separation was achieved on a tracer excel reversed-phase column (C18, 25 × 0.46 cm i.d., 5 µm particle size) from Teknokroma (Spain) protected with a C18 guard column (Teknokroma, Spain). A gradient separation was used with solvent A being a phosphate buffer (pH = 3, 0.01 M) and acetonitrile (72:28) and solvent B being a phosphate buffer (pH = 3, 0.01 M) and acetonitrile (67:33). The gradient steps were as follows: 0 – 7 min, isocratic at 100% solvent A; changed to solvent B over 2 min; isocratic at 100% solvent B up to 17 min; changed to solvent A over 1 min, kept constant for 5 min to re-equilibrate. Total analysis time was 23 min at a flow rate of 1.7 ml min–1 at room temperature (25˚C). The eluate was monitored with a UV detector set at wavelength of 210 nm and the compounds were quantified using their peak height. Sample preparation An aliquot of 250 µl serum was precipitated with 50 µl of a zinc sulfate solution (35%) and 150 µl methanol, vortexed for 1 min and centrifuged for 10 min at 15800g in a 5415C eppendrof

Preparation of calibration standards During validation, eight point calibration standards were freshly prepared in serum for each run, covering a concentration range of 10 – 750 ng ml–1 for each compound. Calibration data were acquired by plotting the peak height of the analytes against the concentration of the calibration standards, followed by a linear-regression analysis. Selectivity and specificity Control human serum, obtained from ten healthy subjects, was assessed by the procedure as described above and compared with respective spiked serum samples to evaluate selectivity of the method. Several cardiovascular agents like amiodarone, verapamil, losartan, propranolol, atenolol, enalapril and captopril were also assessed for potential interferences. Accuracy, precision, limit of quantification (LOQ) and recovery The accuracy, between-day and within-day, precisions of the method were determined for each compound according to FDA guidance for bioanalytical method validation.22 Seven and five replicate spiked serum samples were assayed between-day and within-day, respectively, at four different concentrations (20, 100, 500 and 750 ng ml–1) for each analyte. The concentrations were calculated using calibration curves prepared and analyzed in the same run. The accuracy was calculated as the deviation of the mean from the nominal concentration. The between-day and within-day precisions were expressed as the relative standard deviation of each calculated concentration. For the concentration to be accepted as LOQ, the percent deviation from the nominal concentration (accuracy) and the relative standard deviation must be ±20% and less than 20%, respectively, considering at least five-times the response compared to the blank response. The average recovery of each compound was determined by comparing the peak-height obtained after injection of the processed QC samples with those achieved by direct injection of the same amount of drug in distilled water at different concentrations (ten samples for each concentration level). Application of the method Three male volunteers were included in this study. The study protocol was approved by the Ethics Committee of Tehran University of Medical Sciences and written informed consent was obtained from the volunteers. Volunteers were not allowed to take any other medication for two weeks before and throughout the study. Those with a history of cardiovascular disorders were excluded from participitation in the study. The volunteers received 300 mg of propafenone hydrochloride tablets (Knoll) as a single oral dose after an overnight fast. The intake of food was delayed for 3 h after medication. Peripheral venous blood samples were taken from each volunteer immediately before, and at 0.5, 1, 1.25, 1.5, 2, 2.5, 3, 4, 5, 6, 8, 10, 12 and 24 h after administration of propafenone. Serum samples were stored at –20˚C until analysis for a maximum of 1 month. Calculation of pharmacokinetic parameters Serum concentration–time curves of propafenone and its metabolites were evaluated by non-compartmental analysis. The maximum plasma concentration, Cmax, and the time to Cmax


1309 Table 1 Limit of quantitation (LOQ) for propafenone, 5-OH propafenone, and N-despropylpropafenone (n = 6) Concentration/ Between-day Accuracy, ng ml–1 RSD, % % Propafenone 5-OH-propafenone N-Despropylpropafenone

Fig. 2 Chromatograms of (A) blank human serum, (B) serum of the same person 1 h after oral administration of 300 mg propafenone hydrochloride, (C) plasma spiked with 200 ng/ml of 5-OHpropafenone (1), N-depropylpropafenone (2), and propafenone (3).

(Tmax) were obtained directly from the individual plasma concentration vs. time curves. The terminal half-life (t1/2), was obtained from a log-linear regression analysis of the plasma concentration time curves in the terminal phase. The area under the plasma concentration–time curve up to the last quantifiable plasma concentration (AUClqc) was determined according to the linear trapezoidal method.

Results and Discussion Selectivity and chromatography Various pH values of aqueous buffer, proportions of acetonitrile and phosphate buffer and flow rates were examined to obtain the best separation within the analytes, both from each other and from endogenous compounds. Although a mobile phase containing phosphate buffer (0.01 M, pH = 3): acetonitrile (72:28) met the above requirement, it was still not satisfactory due to the resulting broad and short propafenone peaks, which yielded quantification limits of higher than 50 ng/ml, not suitable for pharmacokinetic studies. Although the propafenone peak shape, and in consequence its quantification limit, was highly affected by the composition of acetonitrile in mobile phase, it was not possible to add more acetonitrile to the mentioned mobile phase without losing resolution between 5OH and N-despropylpropafenone. Therefore, a gradient elution was chosen to have sharper peaks, and in consequence a lower limit of quantification. Separation achieved using the experimental conditions of the present assay for propafenone and its main metabolites are presented in Fig. 2. Selectivity was indicated by the absence of any endogenous interference at retention times of the peak of interest, as evaluated by chromatograms of control human serum and serum spiked with three compounds. The retention times for 5-OH-propafenone, N-despropylpropafenone and propafenone were 7.18, 7.22, and 14.85 min, respectively. An unidentified peak (a) was detected in the serum of the volunteers having received propafenone. As such, the unidentified peak is likely to correspond to another propafenone hydroxylated derivate, 5-hydroxy-4 methoxypropafenone, which has been reported previously.19,23 The system suitability parameters for the method were as follows: the theoretical plates for 5-OH and N-despropyl-

10 10 10

7.8 14.1 14.1

90.5 92.8 87.1

propafenone >11800 and for propafenone >40900; the resolutions between each two consecutive peak of analytes as well as the unidentified peak (a) were >1.5. None of the drug mentioned above interfered with the analytes peaks. No change in the column efficiency and back pressure was observed over the entire study time. Accordingly, at least three hundred samples/column could be analyzed without any significant loss of resolution. Linearity Eight point calibration curves for propafenone and its metabolites on separate days were linear over the concentration range of 10 – 750 ng ml–1. The equations for the means (n = 6) of six standard curves are: for propafenone, y = 0.025x + 0.13 (r2 = 0.999); for 5-OH-propafenone, y = 0.023x + 0.13 (r2 = 0.999) and for N-despropylpropafenone, y = 0.041x + 0.19 (r2 = 0.999). RSD(%) values (slopes, intercepts) were (3.49, 6.7), (2.74, 12.5) and (5.22, 10.4) for propafenone, 5-OHpropafenone and N-despropylpropafenone, respectively. Limit of quantification LOQs, as previously defined, were 10 ng ml–1 for each compound. The LOQ values for three analytes are reported in Table 1. Recovery, accuracy and precision The recoveries of propafenone and its metabolites were evaluated by different pretreatment methods, like different volumes of methanol, acetonitrile, acetone, trichloroacetic acid, perchloric acid, zinc sulfate and a mixture of them for deproteinization. Although the addition of three volumes of methanol to one volume of serum yielded recoveries of higher than 70%, it was not satisfactory because of higher LOQs. Lower LOQs and higher recoveries were achieved by adding zinc sulfate or perchloric acid to methanol. However, the former was better due to harder precipitates. The results from the validation of the method in human serum are listed in Table 2. The method proved to be accurate and precise; the accuracy at four concentration levels ranged from 91.5 – 111.3% for all compounds. The within-day and betweenday precision ranged from 1.4 – 8.1% and 4.2 – 11.5%, respectively, for all analytes. For this reason, an internal standard was not used. The absolute recoveries ranged from 98.3 – 103.2%. Stability Stability was demonstrated for spiked serum samples stored at –20˚C for up to two months, at ambient temperature for at least 10 h and four freeze-thaw cycles. The stability of processed samples was also demonstrated over 10 h. The stability of stock solutions stored at –20˚C was determined for up to 1 month by injecting appropriate dilutions of stocks in distilled water on day 1, 15 and 30 and comparing their peak heights with fresh stock prepared on the day of analysis. Samples were considered to be

1310


Table 2 Between- and within-day variability, accuracy, and recovery for determination of propafenone, 5-OH-propafenone, and N-despropylpropafenone Concentration/ ng ml–1 Propafenone

5-OH-Propafenone

N-Despropylpropafenone

20 100 500 750 20 100 500 750 20 100 500 750

Between-day variability (n = 7) RSD, % Accuracy, % 8.4 6.2 8.5 4.5 5.4 4.2 5.7 4.4 11.5 10.9 6.6 6.3

Within-day variability (n = 5) RSD, % Accuracy, %

96.9 111.3 104.8 102.3 99.5 107.8 101.9 100.2 91.5 105.9 98.0 100.5

8.1 5.1 3.1 3.6 5.5 5.3 3.6 5.0 6.1 5.7 1.4 3.8

98.1 104.6 107.5 94.0 93.3 108.8 103.9 103.1 93.9 110.7 109.0 98.1

Recovery (n = 10) RSD, % % 105.8 103.4 102.5 98.6 99.5 99.4 102.3 100.2 98.3 99.8 103.2 103.3

3.3 2.9 5.0 5.1 2.4 5.5 3.2 4.0 6.4 5.6 6.1 5.8

Table 3 Individual pharmacokinetic data obtained from three healthy male volunteers following oral administration of 300 mg of propafenone hydrochloride

Propafenone

5-OHPropafenone Fig. 3 Serum concentration–time profile of propafenone and its two metabolites after the administration of a single 300 mg oral dose of propafenone to a healthy volunteer.

stable, if the assay values were within the acceptable limits of accuracy and precision. Under all conditions the results met the above criterion. Therefore, the samples were stable during the mentioned periods. Application of the method To apply the developed and validated method, the pharmacokinetics of propafenone and its metabolites was assessed in three healthy volunteers. The subjects were treated with 300 mg of propafenone hydrochloride. Plots of the propafenone, 5-OH-propafenone and N-despropylpropafenone serum concentrations of one volunteer as a function of time following oral dosing is shown in Fig. 3. The pharmacokinetic parameters of propafenone and its metabolites derived by non-compartmental analysis are summarized in Table 3.

N-Despropylpropafenone

Volunteer

Tmax/ h

Cmax/ ng ml–1

AUClqc/ ng h ml–1

t1/2/ h

1 2 3 1 2 3 1 2 3

1.5 2.5 1.5 1.5 1.5 1.5 1.5 1.25 1.5

279.49 210.12 523.68 135.22 90.96 139.48 77.72 92.48 90.79

816.04 943.68 1102.47 455.75 389.02 330.65 174.15 221.46 154.78

2.03 4.33 1.78 1.93 1.48 1.93 0.83 1.28 0.52

propafenone quantification limit to an amount suitable for pharmacokinetic studies, besides preserving the selectivity of the analytes. A small volume of serum is needed for sample preparation, and still the LOQ values (10 ng ml–1) are lower than the previously published methods, which were able to measure propafenone and its two metabolites.19,20 The method also applies a simple and rapid sample treatment instead of rather laborious liquid-liquid or solid-phase extraction as part of the chromatographic procedure, yielding better recoveries, especially for N-despropylpropafenone, thereby saving time and expense for sample preparation. The method was found to be suitable to be currently used for pharmacokinetic studies in humans.

Acknowledgements

Conclusion The method presented here describes a simple, specific and reproducible human serum assay for the determination of propafenone and its major metabolites in a single run. The possibility of measuring N-despropylpropafenone in addition to 5-hydroxypropafenone and propafenone extends the method to enzymes other than CYP2D6. It allows for the evaluation of Ndemethylation rates, which are considered to be indicators of CYP3A4 activity. Moreover, it further helps to gain insights into the pharmacologic aspects of N-despropylpropafenone in vivo. The gradient elution used in this procedure decreased the

This work was fully supported by a grant from Tehran University of Medical Sciences. The authors are grateful to Dr. Ute Hofmann for her kind donation of 5-OH-propafenone and N-despropylpropafenone.

References 1. D. W. G. Harron and R. N. Brogden, Drugs, 1987, 34, 617. 2. C. Dollery, “Therapeuthic drugs”, 1999, Churchil Livingestone, UK, 243. 3. J. T. Y. Hii, H. J. Duff, and E. D. Burgess, Clin.


1311

Pharmacokinet., 1991, 21, 1. 4. J. E. Axelson, G. L. Y. Chan, E. B. Kirsten, W. D. Mason, R. C. Lanman, and C. R. Kerr, Br. J. Clin. Pharmacol., 1987, 23, 735. 5. S. Botsch, J. C. Gautier, P. Beaune, M. Eichelbaum, and H. K. Kroemer, Mol. Pharmocol., 1993, 43, 120. 6. W. E. Haefeli, S. Vozeh, H. R. Ha, W. Taeschner, and F. Follath, Am. J. Cardiol., 1991, 67, 1022. 7. S. A. Cahill and G. J. Gross, J. Pharmacol. Exp. Ther., 2004, 308, 59. 8. S. R. Harapat and R. E. Kates, J. Chromatogr., 1982, 230, 448. 9. E. Brode, R. Sachse, and H. D. Hoffmann, Arzneim.Forsch., 1982, 32, 1. 10. R. Kannan, D. Tidwell, and B. Singh, J. Chromatogr., 1983, 272, 428. 11. P. K. Kunicki and D. Sitkiewicz, J. Liq. Chromatogr. Rel. Technol., 1996, 9, 1169. 12. G. Mazzi, Chromatographia, 1987, 24, 313. 13. P. K. Kunicki and D. Sitkiewicz., Ther. Drug Monit., 1995, 17, 394.

14. P. Kubalec and E. Brandsteterova, J. Chromatogr., B, 1999, 726, 211. 15. P. K. Kunicki, D. Paczkowski, and D. Sitkiewicz, Pol. J. Pharmacol. Pharm., 1992, 44, 161. 16. B. Marchesini, S. Boschi, and M. B. Mantovani, J. Chromatogr., 1982, 232, 435. 17. G. L. Chan, J. E. Axelson, F. S. Abbbott, C. R. Kerr, and K. M. McErlane, J. Chromatogr., 1987, 417, 295. 18. E. Brode, U. Kripp, and M. Hollmann, Arzneim.-Forsch., 1984, 34, 1455. 19. E. R. Kates, Y. G. Yee, and R. A. Winkle, Clin. Pharmacol. Ther., 1985, 37, 610. 20. R. Latini, A. Sica, S. Marchi, Z. M. Chen, M. Gavinelli, and E. Benfenati, J. Chromatogr., B, 1988, 424, 211. 21. U. Hofmann, M. Pecia, G. Heinkele, K. Dilger, H. K. Kroemer, and M. Eichelbam, J. Chromatogr., B, 2000, 748, 113. 22. FDA Guidance for bioanalytical method validation, 2001. 23. S. Vozeh, W. Haefeli, H. R. Ha, J. Vlecek, and F. Follath, Eur. J. Clin. Pharmacol., 1990, 38, 509.