Determination of Pivaloylcarnitine in Human Plasma and Urine by High-Performance Liquid Chromatography with Fluorescence Detection MASAHARU KONISHI~AND HIROSHIHASHIMOTO Received May 14, 1991, from the Shionogi Research Laboratories, Shionogi & Company, Ltd., Fukushima-ku, Osaka 553, Japan. for publicahon March 11, 1992.
major urinary metabolite of the pivaloyloxymethyl ester derivative of methyldopa.5Furthermore, PC is the metabolite commonly recognized with this type of prodrug, and its formation from pivalic acid accompanies endogenous carnitine consumption.- Monitoring of PC in addition to that of endogenous carpitine in biological fluids must be done to confirm drug safety, especially in long-term therapy. Moreover, it is also meaningful to evaluate how exogenous pivalic acid is metabolized in the human body. Despite the need for a method to determine PC levels in biological fluids, none has been reported. Because carnitines lack chromophores, sensitive liquid chromatographic analysis is difficult, except when liquid chromatography-mass spectrometry is used.9 Highperformance liquid chromatography (HPLC) has been used to determine carnitine or acylcarnitines in biological fluids with UV detection &r tagging of the compounds with a phenacyl group as a chromophore to enhance detectability.loJ1 suggested that S-1108was almost quantitativelyconverted to pivalic acid This study focused on the determination of PC in plasma and then conjugated with carnitine. and urine to investigate the metabolism of pivalic acid after the administration of S-1108to humans. Our method involved solid-phase extraction of sample matrices with a cationexchange column and derivatization with a fluorescent rePivaloyloxymethyl ( + )46R,7R)-7-[(2)-2-(2-amino-4- agent, 3-bromomethyl-6,7-dimethoxy-l-methyl-2(1H)thiazolyl)-2-pentenamidol-3-carbamoyloxyethyl-8-oxo-5-quinoxalinone (Br-DMEQ),'z which yields a carboxyl ester thia-l-azabicyclo[4.2.0loct-2-ene-2-carboxylate hydrochlocompound for higher sensitivity and selectivity in HPLC ride hydrate (S-llOS), an oral cephem antibiotic, is hydroanalysis. The results show that our method is sensitive and lyzed by an esterase in the intestinal tract to produce S-1006, reliable for studying the metabolic fate of PC. pivalic acid, and formaldehyde (Scheme 11.1 This metabolic pathway is similar to those of other prodrugs having the Experimental Section pivaloyloxymethyl ester group in their chemical structures.- The metabolism of pivalic acid was studied by Reagents-LCarnitine hydrochloride was obtained from Sigma Vickers et al.,6 and pivaloylcarnitine (PC)was found to be the Chemical Company (St. Louis, MO), and Br-DMEQ was purchased
Abstract 0 A high-performance liquid chromatographic (HPLC)method was developed for the determination of pivaloylcarnitine,one of the major metabolites of pivaioyloxymethyl (+)-(6R,7g-7-[(.?)-2-(2-amlno-4thiazolyl)-2-pentenamido]-3-carbamoyloxymethyl-8-0~0-5-thia-lazabicyclo[4.2.0]oct-2-ene-2-carboxylate hydrochloride hydrate (S1108), an oral cephem antibiotic, in human plasma and urine. Fluorescence detection was done with 3-bromomethyl-6,7-dimethoxy-l-methyl2(1H)quinoxalinone as the labeling reagent. Pivaloylcarnitine and cyclopropylacetylcarnitine. the internal standard, were selectively fractionated from plasma or urine on a disposable cation-exchangecolumn. Derivatization was completed in 20 min at 40°C in the presence of ff,N-diisopropylethylamine as the catalyst. A column-switching device was used to remove the excess reagent for HPLC analysis.The recovery of pivaloylcarnitinewas >98%. and average within-dayand between-day coefficients of variation were 0.999 for both plasma and urine. The mean slopes, intercepts, and correlation coefficients (k standard deviations) of daily calibrations were 2.503 0.107 (pg/mL-l, 0.00677 5 0.00821, and >0.9999, respectively, for plasma (n = 10) and 0.02049 It 0.00029 (r*glmL)-', 0.00331 0.00848, and >0.9999, respectively, for urine (n = 10). Recovery-The absolute recoveries of PC and IS were evaluated by comparing the peak areas of these compounds obtained from the sample treatments of blank matrices spiked with the standards with those obtained from the solutions containing the derivatives of the standards. The
Journal of Pharmaceutical SciencesI 1039 Vol. 81, No. 10, October 1992
_ h l
10 20 30 40
10 20 30 40
10 20 30 40
10 20 30 40
Figure 1-Separation of a reaction mixture of standards (a)and a blank plasma sample (b)on the precolumn. For separation conditions,see text. 1, CPAC-DMEQ; 2. PC-DMEQ. recoveries were >95% for both compounds; these data indicate the satisfactory recoveries of PC and IS during the assay. The recoveries of the analytes from the matrices were calculated from the results of linearity studies based on the calibration curves of the aqueous solutions containing known amounts of derivatives and prepared on individual days. The analytical recoveries (n = 15) were 99.0% for both plasma and urine samples. Figure 2 shows typical chromatograms of blank samples, samples spiked with PC and IS, and actual samples. Precision-The precision of the method was determined by analyzing three pools of blank matrices with known PC concentration of 0.05,0.4, and 2 pg/mL for plasma and 5, 50, and 500 pg/mL for urine. As shown in Table I, the average between-day coefficients of variation (CVs) were 3.6% for plasma samples and 4.7% for urine samples. These values were slightly higher than within-day CVs (2.1 and 2.9%, respectively, for plasma and urine samples). Detection Limit-The detection limit for PC was investigated by analyzing blank plasma and blank urine samples obtained from 10 healthy volunteers. The detection limits calculated as recommended by the International Union of Pure and Applied Chemistry16 were 0.02 pglmL for plasma and 1 pg/mL for urine. Stability of Test Sample Solutions-The stability of the DMEQ derivatives in solution after derivatization was evaluated for a reaction solution containing 4.8 mL of 50 mM phosphate buffer (pH 3.0).Both PC and CPAC derivatives were stable for approximately 20 h a t room temperature, and no change was observed even after 1 month of preservation at 4 "C.The stability of PC in both plasma and urine for up to 2 months a t storage temperatures of -20 and -80 "C was also evaluated. Regardless of sample matrices, PC was stable for up to 2 months a t -80 "C.At -20 "C, PC in urine was stable over the storage period, but the content of PC in plasma decreased to 94% after 2 months. 1040 I Journal of Pharmaceutical Sciences Vol. 81, No. 10, October 1992
10 20 30 40
10 20 30 40
Retention time (m in)
Figure 2-Typical chromatogramsof blank plasma and urine samples (a and a', respectively), blank plasma and urine samples spiked with PC and IS (b and b', respectively), and plasma and urine samples rfom a volunteer who had received S:llo8 (c and c', respectivelyj. 1, CPACDMEQ; 2, PC-DMEQ. Table I-Preclslon of Assay for PC In Human Plasma and Urine Amount Measured, d m L Matrix
Amount Added, p.g/mL
0.051 0.407 2.02 5.01 50.1 501
0.055 (3.6) 0.407(15 ) 2.02 (1.1) 5.20 (3.7) 50.3 (2.8) 497 12.31
0.053(4.5) 0.407(3.3) 2.04 (3.1) 5.05 (6.0) 50.1 (4.4) 487 13.7)
* Expressed as the mean for five repetitive determinations. Mean percentages for plasma and urine were 2.1 and 2.9,respectively. Expressed as the mean for five repetttive determinations for 3 days. Mean percentages for plasma and urine were 3.6and 4.7,respectively. Applications of Method-The method was used to analyze plasma and urine samples from volunteers who had received 51108 in a clinical trial. Figure 3 shows the typical profile of PC concentrations in plasma after oral administration of 200 mg of S-1108. PC can be determined satisfactorily over the assay range established by this method. Urinary excretion of PC for the same volunteers is shown in Figure 4. More than
C a, 0 C 0 0
7 L I
Time after administration (hr)
Flgure SProfile of PC concentrations in plasma after oral administration of S-1108 (100 mg). The data points and the bars represent the
Time after administration (hr) Figure &Profile of urinary excretion of PC after oral administration of S-1108 (100 mg). The data points and the bars represent the average and the range of measurements, respectively, for two volunteers. Key: (m) recovery; (0)cumulative recovery.
90%of the pivalic acid liberated from 5-1108 was excreted as PC. Limitations of Method-The limitations of the method mostly depend on the life span of the precolumn. Our precolumn had a silicone film coating on the surface of the silica gel to make it resistant to a basic mobile phase.17 However, the precolumn tended to lose its efficiency with repetitive injections of samples because of damage from excess reagent and alkaline buffer. The precolumn could be used for at least 200 samples before its performance decreased. The analytical column was considered usable for at least 500 injections of samples.
4. Saikawa, I.; Nakajima, Y.; Tai, M.; Sakai, H.; Demachi, K.; Kajita, T.;Hayakawa, H.; Onoda, M.; Fukuda, H.; Sadaki, H. Yakugaku Zasshi 1986,106,476-490. 5. Vickers, S.;Duncan, C. A. H.; White, S. D.; Ram it, H: G:; Smith, J. L.; Walker, R. W.; Flynn, H.; Arison, B. H. denobrotux 1985, 15,453-458. 6. Melegh, B.; Kerner, J.; Bieber, L. L. Biochem. Pharmncol. 1987, 36,3405-3409. 7. Holme, E.;Greter, J.; Jacobson, C.-E.; Lindstedt, S.; Nordin, I.; Kristiansson, B.; Jodal, U. Lancet 1989,i i , 469-473. 8. Mele h, B.; Kerner, J.; Jaszai, V.; Bieber, L. L. Biochem. Med. Metaf. Bwl. 1990,43,30-38. 9. Minker, P. E.; Ingalls, S. T.; Kormos, L. S.; Weir, D. E.;Hoppel, C . L. J . Chromatogr. 1984,336,271-283. 10. Minker, P. E.;Ingalls, S. T.; Hoppel, C. L. Anal. Biochem. 1990,
average and the range of measurements, respectively, for two volunteers.
Conclusions The HPLC method described here for the determination of
PC in human plasma and urine is convenient and reliable. This method was satisfactorily used to analyze plasma and urine samples from volunteers who had received 5-1108 in a clinical trial. This method also should be applicable to the simultaneous determination of carnitine and acylcarnitines in biological fluids.
References and Notes 1. Toteuka, K.; Shimizu, K.; Konishi, M.; Yamamoto, S.Antimicrob. Agents Chemther. 1992,36,757-761. 2. Johansen, M.; Bundgaard, H.; Falch, E. Znt. J. Phurm. 1983,13, 89-98. 3. Bundgaard, H.; Klixbull, U. Znt. J. Phurm. 1985,27,175-183.
18.5. 29-35. 11. Yergey, A. L.;Liberato, D. J.; Millington, D. S. Anal. Biochern. 1984,139,278-283. 12. Yamaguchi, M.; Hara, S.; Matsunaga, R.; Nakamura, M. J . Chromutogr. 1985,346,227-236. 13. Bohmer, T.; Bremer, J. Biochim. Biophys. Actu 1968, 152, 559-567. 14. Konishi, M.; Hashimoto, H. J. Phurm. Sci. 1990,79,379-383. 15. Yamaguchi, M.; Matsunaga, R.; Hara, S.; Nakamura, M. J . Chromatogr. 1986,375,2735. 16. International Union of Pure and Applied Chemistry Pure Appl. Chem. 1976,45,99-103. 17. Ohtsu, Y.; Fukui, H.; Kanda, T.; Nakamura, K.; Nakano, M.; Fqjiyama, Y. Chromatogmphla 1987,24,38W84. Nakata, 0.; - - - I - -
Acknowledgments We thank Dr. Yoshio Hamashima of Shionogi Research Laboratories for synthesis of the acylcarnitines.
Journal of Pharmaceutical Sciences I 1041 Vol. 81, No. 10, October 1992