CA 94303), modified for microbore ... Chemicals used were carbaxnazepine, ... grade (Spectrum. Chemi- cal Mfg. Co., Gardena,. CA 90248) with. 6.3 mL of a 0.4 ...
CLIN. CHEM. 34/9, 1863-1866
(1988)
Simultaneous Determination of Carbamazepine Plasma by Microbore Liquid Chromatography
and Its Epoxide and Transdiol
Metabolites
in
LillIan E. RIad and Ronald J. Sawchuk1 Simultaneous monitoring of carbamazepine and its metabolites is helpful in providing information on the induction process via the epoxide-diol route. The present method involves the liquid-chromatographic analysis of 1.0 mL of plasma for carbamazepine at concentrations of 0.025 to 4.0 mg/L and for both the 10,11-epoxide and 10,11-transdiol metabolites at concentrations of 0.01 to 1.0 mg/L. Weighted regression equations, expressing peak area ratios as a function of concentrations of carbamazepine and its epoxide and transdiol metabolites in the standards, were used to determine concentrations in plasma samples. AdditIonal tion
Keyphrases: anticonvulsants weighted least-squares regression
.
epilepsy
induc-
Carbamazepine (CBZ), an anticonvulsant, is effective in the treatment of grand mal and psychomotor epilepsy (1).2 In addition, it is the drug of choice in trigeminal neuralgia (2). CBZ is extensively metabolized (3-5). The most important metabolic pathway yields carbamazepine-10,11-epoxide (CBZE), an active metabolite shown to be as potent as the parent drug in animal models of epilepsy (6). CBZE is almost completely converted to the tmns-10,11-dihydroxy10,11-dihydrocarbamazepine (CBZ transdiol; CBZD), which is excreted in the urine (7, 8). From plasma concentration measurements, CBZE is more potent than CBZ in the treatment of trigeminal neuralgia (9), suggesting that monitoring both CBZ and CBZE may be of value. CBZ induces its own metabolism during chronic dosing (10, 11), resulting in enhanced formation of CBZE. Monitoring the ratios CBZE/CBZ and CBZD/CBZ during multiple dosing should provide needed information concerning the induction process occurring via the epoxide-diol route, as there are indications that not only epoxidation but also transdiol formation pathways are induced (8, 12). In addition, the coadministration of other anticonvulsants that interact with CBZ causes alterations in the relative concentrations of CBZE and CBZD. The clinical consequences of these changes are not yet clear. Here we describe a rapid and precise microbore HPLC procedure for simultaneous quantification of CBZ, CBZE, and CBZD in plasma in concentrations as low as 25 agfL for CBZ and 10 p.gfL for both CBZE and CBZD. This method also offers the advantages of using small-bore columns: solvent economy and increased mass sensitivity (13).
Clinical
Pharmacokinetics Laboratory, Dept. University of Minnesota,
of Pharmaceutics,
College of Pharmacy,
Minneapolis,
MN
55455.
‘Address 2Noptandard
correspondence to this author. abbreviations: CBZ, carbamazepine;
mazepine-1O,11-epoxide;
and CBZD,
CBZE, carbatmns-1O,11-dihydroxy-1O,11-
dihydrocarbamazepirie.
Received March
22, 1988; accepted May 6, 1988.
Materials
and Methods
This method is a modification of a procedure developed earlier in our laboratory (14). Instrumentation. For liquid chromatography, we used a high-pressure liquid chromatograph (Model 1081B; Hewlett-Packard, Palo Alto, CA 94303), modified for microbore chromatography and equipped with an automatic sampling system. The separation is done on a 20 cm X 2.1 mm (i. d.) prepacked, microparticulate (5-un average particle size), reversed-phase column (ODS, C18; Hewlett-Packard, Avondale, PA 19311). The flow rate of methanolldistilled water (45/55 by vol) is 0.5 mL/min, and the oven temperature is maintained at 36#{176}C. Column effluent is monitored at 212 nm with a photodiode-array detector (Model 1040A; Hewlett-Packard). Reagents. Chemicals used were carbaxnazepine, carbamazepine-10,11-epoxide, and trans-10,11-dihydroxy-10,11dihydrocarbamazepine (Ciba-Geigy Corp., Ardsley, NY 10502); cyheptamide (Supelco, Inc., Bellefonte, PA 16826); chloroform, distilled in glass (American Burdick and Jackson, Muskegon, MI 49442); methanol (“HPLC” grade; Mallinckrodt, Inc., Paris, KY 40361); and tert-butyl alcohol (An grade, Mallinckrodt). The phosphate buffer, pH 11.2, is prepared by mixing 50 mL of 0.2 moIJL disodium hydrogen phosphate heptahydrate, reagent grade (Spectrum Chemical Mfg. Co., Gardena, CA 90248) with 6.3 mL of a 0.4 molIL solution of sodium hydroxide (reagent grade; Fisher Scientific, Fair Lawn, NJ 07410). Sample extraction and chromatography. To nine 35-mL ground-glass-stoppered centrifuge tubes add CBZ in methanol to provide final concentrations in the plasma matrix of 0, 0.025, 0.05, 0.10, 0.25, 0.50, 1.0, 2.0, and 4.0 mg/L, respectively. To these same tubes, add CBZE and CBZD in methanol to provide respective concentrations of each of 0, 0.01, 0.02, 0.03, 0.05, 0.10, 0.25, 0.50, and 1.0 mgIL. To these tubes add 20 iL of cyheptamide in methanol (30 mgfL), and evaporate the methanol under reduced pressure. These tubes are used in preparing the standard curve. Add identical amounts (0.6 jtg) of cyheptamide in methanol to another series of 35-mL tubes for samples and allow the methanol to evaporate. To the standard-curve tubes, add 1 mL of drug-free (blank) plasma, and to the sample tubes, 1-mL aliquots of each of the plasma samples to be analyzed. To each tube, add 2 mL of phosphate buffer (pH 11.2) and 20 mL of tert-butyl alcohol in chloroform (50 mLIL), stopper, and shake horizontally at 180 cycles per minute on a mechanical shaker (Eberbach Corp., Ann Arbor, MI 48106) for 5 mm. Centrifuge the tubes at 750 X g for 5 mm and aspirate and discard the aqueous layer. Transfer the chloroform phase to clean 35-mL glass-stoppered centrifuge tubes and evaporate the solvent at 65 #{176}C under reduced pressure (we used the “Evapo Mix”; Buchler Instruments, Fort Lee, NJ 07024). Dissolve the residue in 50 pL of mobile phase by vortex mixing, and transfer the solution to a microvial for automatic injection. Inject 10 pL of the sample and use the aboveCLINICAL CHEMISTRY,
Vol. 34, No. 9, 1988
1863
recovery, expressed as a ratio that prepared, was calculated.
mentioned chromatographic conditions. Calculate the peak area ratios of CBZ, CBZE, and CBZD to cyheptamide. Cakulations. The concentrations of CBZ, CBZE, and CBZD in the samples are calculated by using standard curves obtained by applying weighted least-squares regression analysis to express peak area ratios as a function of CBZ, CBZE, and CBZD concentration of the standards. These are analyzed in parallel with each batch of plasma samples. Weighted regression is used in determining the lowest limit of reliable assay measurement (15), in which the weight used at each concentration is the reciprocal of the observed sample variance. Results Assay
and
data.
Physical recovery. Peak-area ratios of CBZ, CBZE, and CBZD measured in plasma extracts were compared with those measured in unextracted samples containing known amounts of CBZ, CBZE, and CBZD. To determine the physical recovery, we added the internal standard to the samples before injection. The mean (and SD) recoveries (%) of CBZ, CBZE, and CBZD were 87 (4.4), 73 (7.4), and 53 (4.4), respectively, and individual recovery values were not concentration dependent. Specificity and sensitivity. Figure 1 shows typical chromatograms from a blank, a standard sample, and a subject’s plasma sample. No interfering peaks are noted in blank plasma. The mean retention times, in minutes, for CBZD, CBZE, CBZ, and cyheptamide in a typical run, and their coefficients ofvariation, were, respectively, 2.25 (1.4%), 2.80 (1.3%), 5.78 (0.6%), and 8.92 (0.5%). The spectra of CBZ, CBZE, and CBZD were obtained on-line during chromatography of the compounds. Such spectra for these compounds, whether extracted from plasma or unextracted, were qualitatively similar. Acquiring spectra on-line affords rapid confirmation of peak identity and purity. The present method offers high sensitivity; CBZ can be quantified with acceptable precision at concentrations of 25 pg/L and both CBZE and CBZD at concentrations of 10 tgfL. This sensitivity is adequate for analyzing plasma of subjects who are receiving single doses of CBZ, for the purpose of bioequivalency studies. On the other hand, patients on chronic CBZ dosing exhibit CBZ, CBZE, and CBZD concentrations in plasma greater than those described here. For such patients, 0.2-mL plasma aliquots are diluted with blank plasma to give a total volume of 1.0 mL and subjected to the same analytical procedure. Interference. Plasma samples from a patient who was receiving two other commonly prescribed anticonvulsants, phenobarbital and phenytoin, were diluted with blank plas-
Discussion
Characteristics
and precision. Good linearity was exhibited for 25 jtg/L to 4.0 mg/L and also for CBZE and CBZD from 10 tg/L to 1.0 mgfL, where the coefficient of determination, r2, exceeded 0.98 for all standard curves. Within-run precision was determined by comparing the peak-area ratios for three standard curves extracted and irjected on the same day. Run-to-run precision of standard curves extracted on four different days was similarly determined. Table 1 summarizes these data. Run-to-run standard deviations are used in defining the observed sample variance for purposes of weighting data in calculating the line of best fit for the standard curve. The advantage of using weighted least squares is that it does not assume a constant variance over the entire concentration range of interest. Thus observations with a large variance are given less weight than observations with smaller variance, which is appropriate when nonconstant variance exists. Analytical recovery. Known amounts of CBZ, CBZE, and CBZD were added to plasma to provide five different concentrations. These served as quality-control samples and were analyzed in parallel with each batch of plasma samples whose concentrations were to be determined. Analytical Linearity CBZ from
Table
1. AnalytIcal
Precision
of the Analysis for Carbamazepine, Carbamazepine-1 0,11 -tranadlol
Carbamazepine Concn, mg/L
A. Within-run precision (n 3) 0.0 0.025 0.0564 (0.00104) 0.050 0.121 (0.0010) 0.10 0.217 (0.0055) 0.25 0.554 (0.018) 0.50 1.13 (0.031) 1.00 2.48 (0.031) 2.00 4.94 (0.10) 4.00 9.83 (0.13) Slope 2.47 (0.035) B. Run-to-run precision (n 4)
CV, %
Concn, mg/L
Carbamazeplne-1
Peak-area
mean
Concn,
ratio, (SD)
0,11 -epoxlde,
and
Carbamazepine-1O,11-tranadioi
1-spoxide
Carbamezepln.-1O,1
Peak-area ratio, mean (SD)
of concentration measured to Table 2 summarizes these
CV, %
mg/L
Peek-area ratio, mean (SO)
CV, %
=
1.8 0.8 2.5 3.3 2.7 1.2 2.0 1.3 1.4
0.0 0.010 0.020 0.030 0.050 0.10 0.25 0.50 1.0 Slope
17.2 6.6 4.5 2.7 2.9 5.5 5.7 4.7 5.1
0.0 0.010 0.020 0.030 0.050 0.10 0.25 0.50 1.0 Slope
-
-
-
0.0170(0.0032) 0.0448
(0.0034)
0.0616 0.106 0.211 0.536 1.13 2.38 2.37
(0.0019) (0.0023) (0.013) (0.0067)
(0.021) (0.046) (0.046)
19.0 7.6 3.1 2.2 6.2 1.2 1.9 1.9 1.9
0.0 0.010 0.020 0.030 0.050 0.10 0.25 0.50 1.0 Slope
-
0.0112 0.0250 0.0387 0.0609 0.125 0.313 0.630 1.19 1.20
-
(0.00097) (0.0020) (0.0026) (0.0016) (0.0066) (0.0061) (0.022) (0.015) (0.021)
8.7 8.0 6.7 2.6 5.3 1.9 3.5 1.3 1.8
=
o 0.025 0.050 0.10 0.25 0.50 1.0 2.0 4.0 Slope
1864
-
-
0.0609 0.126 0.227 0.550 1.15 2.38 4.75 9.13 2.34
(0.0105) (0.0083)
(0.0102) (0.015) (0.034) (0.13) (0.27) (0.44) (0.12)
CLINICAL CHEMISTRY,
Vol. 34, No. 9, 1988
-
0.0231 (0.0027) 0.0489
(0.0034) (0.00504)
0.0712 0.120 (0.0011) 0.215 (0.014) 0.569 (0.039) 1.15 (0.069) 2.35 (0.103) 2.34 (0.11)
-
11.7 6.9 7.1 8.9 6.5 6.8 6.0 4.4 4.7
0.0 0.010 0.020 0.030 0.050 0.10 0.25 0.50 1.0 Slope
-
0.0109 0.0209 0.0372 0.0592 0.121 0.298 0.604 1.18 1.19
-
(0.00103) (0.00075) (0.0031) (0.0048) (0.0096) (0.028) (0.042) (0.061) (0.066)
9.4 3.6 8.3 8.1 7.9 9.4 6.9 5.2 5.5
Table
2. A nalytical
Carbamazepine Added concn, mg/L
Mean measured
Recovery
of Quality-Control
Carbamazepine-10,1
(SD) concn,
Mean (SD) recovery, %
mg/L
0.05 0.10 0.25 0.50 1.00
0.0464 0.0958 0.245 0.495 0.981
(0.0044) (0.0049) (0.0084) (0.0098) (0.016)
2.00
1.97
(0.044)
93 96 98 99 98 98
(8.8) (4.9) (3.3) (2.0) (1.6) (2.2)
Added concn, mg/I
0.01 0.02 0.05 0.10 0.20 0.40
Mean (SD) measured concn, mg/L
0.0106 0.0201 0.0528 0,103 0.203 0.408
(0.00011) (0.0012) (0.0066) (0.0025) (0.0062) (0.028)
(n
=
5)
Carbamazeplne-10,1
Mean (SD)
Added
recovery,
concn, mg/I
106(11) 101 (6.2) 106 (13) 103 (2.5) 102 (3.3) 101 (6.2)
Mean
1-fransdiol
(SD)
Mean
concn, mg/I.
measured
0.01 0.02 0.05 0.10 0.20 0.40
0.00998 0.0190 0.0485
(0.00090) (0.00302) (0.0038)
0.101 0.197 0.418
(0.0078) (0.0077) (0.027)
(SD)
recovery, %
100 (9.0) 95 (15)
97 101 99 105
(7.7) (7.8) (3.8) (6.7)
3
3
A
Sa mpies
1-epoxide
4
B
Fig. 1. Representative chromatograms obtained in the analysis of 1.0 mL of plasma blank plasma; B, blank plasma supplemented withCBZD (0.10 mgfL), CBZE (0.10 mg/L). and CBZ (0.50 mg/I); C, plasma sample containing CBZD (0.102 mg/I), CBZE (0.0823 mg/I), and CBZ (1.2 mg/I). 1, CBZD; CBZE; 3, CBZ; 4, internai standard (cyheptamide) A,
ma and taken through the extraction and chromatographic procedure. Neither phenobarbital nor its p-hydroxy metabolite nor the 5-(4-hydroxyphenyl)-5-phenylhydantoin metabolite of phenytoin was extracted at the pH used. Thus they did not interfere with the analysis for CBZ or its metabolites. Retention times relative to CBZ were: CBZD, 0.38; CBZE, 0.48; and phenytoin, 0.82. Interference by clonazepam with CBZ assay was studied by extracting 1 mL of plasma supplemented with clonazepam and CBZ to give concentrations of 0.5 and 5.0 mg/L, respectively. The retention time of clonazepam relative to that of CBZ was 1.07, with incomplete resolution of peaks. However, this does not represent significant interference, because therapeutic concentrations of clonazepam are an order of magnitude lower than those of CBZ, and the detector response at these concentrations for clonazepam is much less than for CBZ.
a, E
z ‘C
z
w z cJ
TIME
(h)
Fig. 2. Mean concentrations (n 12) of CBZ, CBZE, and CBZD in plasma after a single 400-mg dose of Tegretol =
Measurement
of CBZ,
CBZE,
and CBZD
in Plasma
We used this method to measure CBZ, CBZE, and CBZD concentrations in plasma in a 2 x 2 crossover single-dose (400 mg) carbama.zepine bioequivalency study in which Tegretol was used as the reference formulation. The mean half-life values and their standard deviations for CBZ, CBZE, and CBZD in 12 healthy adult male volunteers were 36.3 (4.82), 32.2 (5.13), and 33.5 (5.38) h, respectively. The average concentrations of CBZ and its metabolites in plasma are shown in Figure 2. The precision and linearity offered by this method permit analysis for CBZ, CBZE, and CBZD over a wide range of plasma concentrations such as those observed after single doses of CBZ.
We thank Dr. Keith Chan, Pharmaceuticals Division, Ciba-Geigy Corp., Ardsley, NY 10502, for supplying pure carbamazepine, carbamazepine-10,11-epox.ide, and tnzns-10,11-dihydroxy-10,11-dihydrocarbamazepine. References
1. Jongmans JWM. Report on the antiepileptic Epilepsia 1964;5:74-82. 2. Blom S. Trigeminal neuralgia, its treatment vulsant drug (G-32883). 3. Morselli PL, Frigerio carbamazepine [Review].
action of Tegretol.
with a new anticonLancet 1962;i:839-40. A. Metabolism and pharmacokinetics of Drug Metab Rev 1975;4:97-113.
CLINICAL CHEMISTRY,
Vol. 34, No. 9, 1988
1865
L. Clinical pharmacokinetics of carbamazepine [RePharmacokinet 1978;3:128-43. 5. Faigle JW, Feldmann KF. Carbamazepine biotransformation. In: Woodbury DM, Penry JK, Pippenger CE, eda. Antiepileptic 4. Bertilsson viewl. Clin
drugs.
New York:
Raven
Press, 1982:483-95.
6. Faigle JW, Feldmann KF, Baltzer V. Anticonvulsant effect of carbamazepine. An attempt to distinguish between the potency of the parent drug and its epoxide metabolite. In: Gardner-Thorpe C, Jam D, Meinardi H, Pippenger CE, eds. Antiepileptic drug monitoring.Tunbridge Wells, U.K.: Pitman Med. Publ. Co., 1977:104-8. 7. Tomson T, Tybring G, Bertilsson L. Single-dose kinetics and metabolism of carbamazepine-10,11-epoxide. Clin Pharmacol Ther 1983;33:58-65. 8. Eichelbaum M, Tomson T, Tyrbing G, Bertilsson L. Carbamazepine metabolism in man, induction and pharmacogenetic aspects. Clin Pharmacokinet 1985;10:80-90. 9. Tomson T, Bertilsson L. Potent therapeutic effect of carbamazepine-10,11-epoxide in trigeminal neuralgia. Arch Neurol 1984;41:598-601.
CLIN. CHEM. 34/9, 1866-1867
Time-Dependence
cellular
Keyphrases:
Zlfe Krosl,
circadian
rhythmn
.
diseases
involving
Neopterin is a metabolite derived from guanosine triphosphate. On stimulation with interferon gamma secreted by Tlymphocytes, human macrophages release neopterin into the bloodstream (1). It is eliminated almost exclusively in urine, therefore urinary neopterin is considered to be a good reflection of its biological production. Neopterin has been found useful in the follow-up of diseases involving cellular immunity, e.g., cancer (2), AIDS (3), Crohn’s disease (4), and sarcoidosis (5), and in allograft recipients (6). Given recent descriptions of large circadian variations of lymphocytes (OKT3, OKT4, NK cells) (7), we wondered if there was also a rhythmicity in the urinary excretion of neopterin. Department of Biochemistry, Facult#{233} de M#{233}decine Pitie-Salpetri#{232}re, 91 Bd. de l’H#{244}pital, 75634 Paris Cedex 13, France.
1866
2, 1988; accepted May 3, 1988.
CLINICAL CHEMISTRY,
13. Scott RPW. An introduction to small-bore columns. J Chromatogr Sci 1985;23:233-7. 14. Sawchuk RJ, Cartier LL. Simultaneous liquid-chromatographic determination of carbamazepine and its epoxide met.abolite in plasma. Cliii Chem 1982;28:2127-30. 15. Oppenheimer L, Capizzi TP, Weppelman RM, Mehta H. Determining the lowest limit of reliable assay measurement. Anal Chem 1983;55:638-43.
Activity
and Yvan TouItou
immunity
Received March
12. Bourgeois BFD, Wad N. Carbamazepine-10,11.diol steady-state serum levels and renal excretion during carbamazepine therapy in adults and children. Ther Drug Momt 1984a;6:259-65.
a Marker of Cellular Immune
Neopterin, a marker of cellular immune system activation, is produced by human macrophages after induction by interferon gamma (secreted by T-Iymphocytes) and is eliminated mostly in urine. We have documented the circadian rhythm of urinary neopterin in five healthy young men (about 25 years old), using voidings collected during 48 h at fixed 4-h intervals. We repeated the experiment three times, one week apart. Neopterin was measured by high-performance liquid chromatography (HPLC). We clearly show a peak of the excretion of neopterin in the early morning (around 0630 hours ±2 h), with total variability (peak-trough difference) reaching 51%. Neopterin is commonly assayed in urinary fractions, so it is imperative to use urine specimens collected at the same time of day-e.g., the first morning urines-to avoid misinterpretation in follow-up of patients. AddItIonal
11. Bertilsson L, Hojer B, Tybring G, Osterloh J, Rane A. Autoinduction of carbamazepine metabolism in children examined by a stable isotopic technique. Clin Pharmacol Ther 1980;27:83-8.
(1988)
of Urinary Neopterin,
Andr#{233} Auz#{233}by, Andr#{233} Bogdan,
10. Eichelbaum M, Ekbom K, Bertilsson L, Ringberger VA, Rane A. Plasma kinetics of carbainazepine and its epoxide metabolite in man after single and multiple doses. Eur J Clin Pharmacol 1975;8:337-41.
Vol. 34, No. 9, 1988
Subjects,
MaterIals,
and
Methods
Subjects Five young men, medical students, ages 25 (SD 2) years, volunteered for the study. They were considered healthy on routine clinical and laboratory examinations. None took any medication before or during the study, and all had a regular social routine, with lights turned off at 2330 hours (±1 h) and on at 0730 hours (±1 h). Unrestricted meals were taken at 0730, 1230, and 2000 hours (all ±1 h). Protocol Each subject collected his total urine each 4 h during 48 h, beginning at 2030 hours, and this procedure was performed three times, once a week during late February and March. The total protocol therefore yielded 180 specimens. After collection, specimens were promptly stored in darkness at -20 #{176}C until assay. Analytical
Procedures
neopterin and creatinine were determined by reversed-phase HPLC as previously described (8). In brief: for HPLC we used an LC 5560 chromatograph (Varian, Walnut Creek, CA) monitored with a Vista 402 data system (Varian) and equipped with an autosampler (Varian 8085), an ultraviolet detector (Varian UV 200), a fluorescence detector (Varian Fluorichrom), and a 10-iL injection valve (Valco Instruments, Houston, TX). The ready-to-use C18 reversed-phase column, 4 x 125 mm, with 10-zm-diameter packing (E. Merck, Darmstadt, F.R.G.) had a 4 x 25 mm guard column (Merck). The mobile phase was potassium phosphate buffer (15 mmol/L, pH 6.45), the flow rate 0.8 mL/ mm. We diluted 100 1.i.L of urine in 1 mL of mobile phase and njected 10 L. Neopterin was detected by its native fluoresUrinary
