Determination of a Cephalosporin Antibiotic, Ceftibuten, in Human Plasma with Column-Switching High-Performance Liquid Chromatography with Ultraviolet Detection HWAI-TZONG PAN", PRAMILA KUMARI*, JOSEPHINE LIM*, AND CHIN-CHUNG

LIN'

Received June 6, 1991, from the 'National Yen -Ming Medial College, Department of Pharmacol y,, Taipei 11221, Taiwan, Republic of China, and *Schering-PlWgh Research, Drug Metabol!m Department, Bloomfield, NJ. Accepted for p%cation August 22,1991,

and analysis of complex biological sample matrixes. These methods not only require less time and labor for sample pretreatment but also decrease sample exposure to the atmosphere and, therefore, decrease the potential for decomposilosporin, in human plasma. Plasma samples were directly injected into the first chromatographic column for sample cleanup and extraction. tion. In addition, column-switching, two-dimensional HPLC Thereafter, a switching valve, located at the junction of the first cleanup analysis can offer more selectivity than column-switching, column and the second analytical column, opened during a e m l n unidimensional HPLC (with a short-trap column without interval after injection to transfer the segment containing ceftibuten into separation) when samples contain matrix interferences. Rosthe analytical column for quantitation. The method was used routinely in ton and Wijayaratnes reported a column-switching, twopharmacokinetic studies and h a s the advantages of simplicity,improved dimensional HPLC method to separate prostaglandin enansensithrity, and selectMty. Concentrations of ceftibuten in plasma over a tiomers. Column-switching, two-dimensional HPLC analysis range of 0.1-20 WmL can be determined with high preclsion and can also accommodate large sample injection volumes and reproducibility. A single, oral, 200-mg dose of ceftibuten in humans provide trace enrichment of the analyte, thereby resulting in resulted in a maximum concentration of 9.79 MmL, an area under the plot of concentration of drug in plasma versus time (0-12 h) of 47.77 narrow peaks and better sensitivity. A column-switching, pg h/rnL, and a half-life of 3.2h. These results document the utility of twodimensional HPLC assay combined with a heart-cutting the LCRCAJV method in clinical pharmacokinetic studies. technique (LC/LC/UV) was then developed to determine ceftibuten in human plasma to study the pharmacokinetics of ceftibuten in patients. The LC/LC/UV assay reported in this paper is straightforward, rapid, and precise, and its use in CeRibuten [(+)-(6R,7R)-7-[(2)-2-(2-amin0-4-thiazolyl)-4clinical pharmacokinetic studies is demonstrated. carboxycrotonamidol-8-oxo-5-thia-1-azabicyclo[4.2.Oloct-2ene-2-carboxylic acid; SCH 397201 is a new oral cephalosporin Experimental Section antibiotic that was discovered by Shionogi Research LaboraReagenteHPLC-grade acetonitrile was obtained from Fisher tories, LM., Tokyo, Japan. It is highly active against a variety Scientific (Pittaburgh, PA). Deionized water was collected from a of gram-negative bacteria (e.g., Escherichia coli, Klebsiella Milli-Q water purification system (Millipore Corporation, Milford, species, Proteus species, and Haemphilus influenme) and MA). Ammonium acetate (ACS grade) was from Sigma Chemical moderately active against gram-positive bacteria (e.g., EnCompany (St. Louis, MO). Drug-& plasma for preparation of tembcrcter, Citrobcrcter,and Sermtia species and Streptococcus standard solutions was obtained from the North Jersey Blood Center (Orange, NJ). pyogenes). Apparatus-The column-switching chromatographic system conA conventional high-performance liquid chromatographic sisted of two solvent delivery pumps. Pump A (Waters model 590)was (HPLC) method was originally developed by Shionogi Reused to deliver the "weak" mobile phase for analyte cleanup and search Laboratories to determine concentrations of ceftibuten preconcentration, and pump B (Waters 600-MSmultisolvent delivery in plasma. The HPLC method, however, needs some sample system) was used to deliver the "strong" mobile phase. Two standard pretreatment. Although a robotic system was commercially HPLC six-port valves (Rheodyne model 7000) were used; one was available for circumventing the labor-intensive liquid-liquid used for plasma sample iqjection, and the second was used to switch or solid-phase extraction in the conventional HPLC method, the analyte-containing fraction that was eluting from the first the potential for sample decomposition at mom temperature (cleanup) column into the second (analytical) column for quantitaand crossantamination over time was a major concern. tion. A diagram of the equipment setup is shown in Scheme I. The W detector (Waters Lambda-Max, model 481) waa operated at 263 nm, Therefore, a method involving direct injection of plasma and the chromatograms were recorded on a Waters SE-120strip chart samples into the HPLC system was necessary to simplify the recorder. assay and to decrease the chances for sample contamination Chromatography-The first cleanup column used on-line for and decomposition. Recently, column-switching techniques sample extraction and cleanup was a Waters aondapak phenyl for either unidimensionall-( or two-dimensional- HPLC packing in a 3.9-mm i.d. x 150-mmstainless steel column.The second analysis have gained increasing attention for direct injection (analytical) column was a pBondapak phenyl packing in a 4.6-mm i.d. Abrtract 0 A column-switching high-performance iiquld chromatographic assay combined with a heartcutting technique and UV detectlon (LC/LC/UV) was developed to deternine ceftibuten, a new oral cepha-

-

Ceftibuten (Sch 39720) 0022-3549/92/0700-&363$02.663$02.50/0 7992, American Pharmaceuticel Associetlon

(0

x 300-mm stainless steel column. Using the same stationary-phase packing in both columns is advantageous because chromatographic behavior is similar.9 The weak ionic mobile phase used in the first column was 0.1 M ammonium acetate (pH 6.51, and the strong ionic mobile phase used in the second column was 2% acetonitrile in 0.1M ammonium acetate (pH 6.5). Using similar buffer system for both extraction and analyte elution usually minimizes baseline disruption when column switching occurs. The flow rates in both columns were the same (1 d m i n ) . The time window to elute the analyte (cefiibuten) from the first column into the second column was 4-6 min from

Journal of Pharmaceutical Sciences I 683 Vol. 81, No. 7, July 1992

Sample extraction & clean -up

LCRCNV

Waste

Scheme I-Diagram

of equipment for column-switching HPLC system.

the time of injection of cis-ceftibuten standard (1 pg/mL)with a small amount of tmns-ceftibuten into the firat column, which was eluted by the weak mobile phase. The retention time of cis-ceftibuten in the first column was -5 min (Figure 1). Therefore, the time window for column switching was set at 4-6 min after injection to ensure quantitative recovery of the analyte fraction. The 1-pg/mLstandards were then injected into the two-column HPLC system, and the desired analyte fraction was switched between 4 and 6 min into the second column for the final separation. Injection of the ceftibuten standard (1 pg/mL) showed that the overall retention time through the two-column system was -13 min (Figure 2). A guard column was inserted before the first column to prevent plasma lipoproteins from depositing on the head of the column. The guard column was changed each night, and the first column was flushed with 2%acetonitrile in 0.1 M ammonium acetate overnight to elute any matrix residues that had adsorbed on the column. Sample Processing-Blood samples were drawn from 12 patienta with pulmonary disease immediately prior to the oral administration of a single, 200-mg capsule of ceftibuten (0h) and thereafter at 1,1.5, 2,3,4,6,8,10,and 12 h postdosing. Whole blood (10 mL) was collected separately in heparinized vials (Vacutainer) at the designated times, and the samples were immediately centifuged for 15 rnin. The samples of separated plasma were transferred into labeled plastic tubes and immediately frozen at -70°C.On the day of analysis, samples were thawed and diluted with 0.1 M ammonium acetate (l:l, v/v). A 100-pL aliquot was injected directly into the two-column H P E system. Drug-& control plasma was used to prepare the

LClUV

-

0

4

6

Minutes

Flgure l-chromatographic profile from the first (cleanup) column for determining the analyte window interval (with a 100-4 injection volume for a l-pg/mL cisceftibuten standard with a small amount of trans ceftibuten). The mobile phase was 0.1 M ammonium acetate, and the R, of cisceftibuten was 5 min. 604 / Journal of Pharmaceutical Sciences Vol. 81, No. 7, July 1992

14 12

6 4

0

Minutes

Figure 2-Typical chromatogram of a ceftibuten standard (1 M m L ) obtained with the column-switching HPLC system (the switching valve was opened only during the 4-6-min interval). The overall R,of ceftibuten was 13 min.

drug-spiked standardsthat contained 0.10,0.50,1.00,5.00,10.00, and 20.00 pg of ceftibutedml of plasma and that were used to prepare the daily calibration curve. Control samples of plasma that were run through the entire procedure were free of interfering peaks in the overall retention area of cis-ceftibuten [retention time (Rt),13min]. Data Acquisition-After each sample was analyzed, the concentration of drug was calculated by using the calibration curve. The regression line of the linear relationship between peak height of ceftibuten and theoretical concentration was defined as y = mx + b, where y is the peak height, x is the concentration of ceftibuten (pg/mL), m is the slope of the fitted line, and b is the y intercept of the regression line. No weighting factors were used to calculate the regression line.

Results and Discussion Posluszny and Weinbergel.6reported two configurations for column-switching, two-dimensional HPLC analysis. In the first configuration, the cleanup column was connected across the switching valve, and the desired portion was eluted by a strong mobile phase into the second column. In the second configuration, the cleanup column was connected before the switching valve, and the desired portion was eluted by a weak mobile phase into the second column. The second configuration was adopted in this study, because the retention volume of the desired analyte (ceftibuten) fraction was reduced by using a weak mobile phase to strip the drug from the cleanup column onto the head of the second analytical column. This focusing-enrichment step is more efficient if the second configuration is used. Assay Sensitivity, Specificity, and Linearity-The limit of quantitation (0.10 pg/mL) was set a s the lowest standard. Any unknown sample concentration 0.999. The only potential interference of cefiibuten would be the trans isomer (metabolite). However, the trans isomer has a longer R, (6-8 min) in the first preparative column and would not be eluted in the 4-6-min segment containing cis-

Table CConccHltretlons of Ceftlbuten Calculated from Conwntratlowllme Curves for Standard8' Experimentally Determined Concentration (pg/mL) Corresponding to a Nominal Concentration (pg/mL) of:

Day

0.10

0.50

1.00

5.00

10.00

20.00

-

0.09 -

0.61 0.52 0.46 0.41 0.39 0.44 0.55 0.60 0.53 0.42 0.47 0.51 0.51 0.56 0.51

1.04 1 .00 0.85 1.01 0.94 1.01 0.94 0.93 1.10 1.11 1.03 1.03 0.94 0.98 1.06

4.80 4.70 4.39 4.81 4.89 4.88 5.03 5.01 4.94 5.14 5.21 5.09 5.09 5.36 4.99

10.00 10.02 10.21 9.90 10.04 9.11 9.85 10.48 9.88 9.85 9.88 9.28 10.16 10.06 9.91

19.32 19.18 17.94 19.44 19.44 20.22 19.67 18.28 19.07 19.09 20.12 19.41 17.74 18.29 18.31

0.12 0.014 11.6

0.50 0.067 13.4

1 .00 0.07 7.0

4.96 0.23 4.6

9.91 0.34 3.4

19.04 0.76 4.0

0.11 0.11

-

,

14 12

,

,

,

I

6 4

I

0

Minutes

flgure &Typical chromatogram at the detection limit (0.05 pg/mL) for the column-switching HPLC system.

n

A

n

B

0.09 0.14 0.11 0.11 0.12 0.11

6 7 8 9 10 11 12 13 14 15

0.11

Mean SDc cv, Yo

Determined over time period of the study with LC/LC/UV method. '-, No O.lO-pg/mLstandard was run in these runs, and results 0.1 a / m L were repeated in subsequent runs. standard deviation.

Table ICData for Pmclrlon and Accuracy of Callbratlon Curve Standard8 Concentration of Mean Calibration n Determined Value, &mL Standard, pg/mL

0.10 0.50 1 .00 5.00 10.00 20.00 Flgure -A) Chromatogram taken immediately after sampling of control plasma from a drug-freedonor showing the absence of interfering peaks in the retention area of ceftibuten (13 min). (6) A typical chromatogram of a plasma sample taken 1 h after dose.

ceftibuten that ia switched into the second analytical column (Figure 1). Reproducibility, Accuracy, and Precision-The calibration curve data and interaesay statistical analysis (Table I) indicate that the obtained away values were reproducible, with a mean average coefficient of variation (CV) of 7.3% over the concentration range of the calibration curve. Interday precision (CV)and accuracy of the assay for the determination of ceftibuten standards (Table 11) indicate that the assay is both accurate and precise with a CV of 13.4% or less. The overall accuracy was 102.3%, with a precision of 8.4%. Pharmacokineties of Ceftibuten in Patients with Pulmonary Disease-In vivo application of the two-column LC/LC/UV method was demonstrated by determining the concentrations of ceftibuten in human plasma &r a single, 200-mg oral capsule dose in 12 patients with pulmonary disease. The mean ( 2 standard deviation) profile of concentration of drug in plasma-time (concentration-time curve; Figure 6)indicates a maximum concentration of 9.8 p g h L and a n area under the concentration-time curve (0-12 h) of 47.8 pg*h/mL.The mean elimination half-life (3.2 h) waa calculated from the individual plasma concentration-time curves. The two-column HPLC assay system used herein was

10 15 15 15 15 15

0.12 0.50 1.00 4.96 9.91 19.04

Mean SD" Accuracy, % Precision c v , %OC of Nominalb

0.014 0.097 0.070 0.23 0.34 0.76

120.0 100.0 100.0 99.2 99.1 95.2

11.6 19.5 7.0 4.6 3.4 4.0

*Standard deviation. bAverage, 102.3% (n = 6). CAverage, 8.4% (n = 6).

nme (hour) Flgure S M e a n plasma concentration-time curve for ceftibuten in patients with pulmonarydisease (n = 12)after a single, 200-mgoral dose (vertical bars represent the standard error of the mean).

sensitive, reliable, and selective for the quantitation of cisceftibuten, the mqjor component in ceftibuten capsules, in pharmacokinetic studies in humans.

References and Notes 1. Alfredeon, T. V. J . Chromtogr. 1981,218,715-728. 2. Kuo,B. S.; Mandagere, A.; Osborne, D. R.; Hwang, K. K.Pharm. Reg. 1990, 7, 1257-1261.

Joumel of Phannaceufical Sciences/ 665 Vol. 81, No. 7, July 1992

3. Lim, J. M.; Lin, C. C. Antimicrub. Agents Chemother. 1986, 29, 977-979.

4. Veals, J. W.; Lin, C. C. Am. Lab. 1988,20,42-47. 5. Berkowitz, S. A. Anal. Biochem. 1987,164,254-260. 6. Posluezny, J. V.; Weinberger, R. A d . Chem. 1968, 60,19631958. 7. Po8lu=nY, J. v.; Weinberger, R.J . Chmmtogr. 1990,507, 267276. 8. Roston, D. A.; Wijayaratne, R. Anal. Chem. 1988, 60,960-956.

866 I Journal of Pharmaceutical Sciences Vol. 81, No. 7, July 1992

9. Sonnefeld, W. J.; Zoller, W. H.; May, W. E.; Wise, Chem. 1982,54,723-727.

S. A. A d .

Acknowledgments We thank Dr. J. Arthur F. de Silva for his valuable advice and careful review of this report and Drs. Mitchell Cayen and Samson Symchowicz for their support and encouragement.