The Gas Chromatographic Determination of Anticonvulsant Drugs in Serum

Home , Metharbital, Paramethadione, Phensuximide

WILLIAM C. GRIFFITHS, Ph.D., STEVEN K. OLEKSYK, PAUL DEXTRAZE, AND ISRAEL DIAMOND, M.D.

Department of Laboratory Medicine, Roger Williams General Hospital, Providence, RZ 02908

ABSTRACT A simple and accurate method is described for the quantitative determina tion of trimethadione, paramethadione, ethosuximide, metharbital, methsuximide, phensuximide, mephenytoin, ethotoin, primidone and diazepam in serum. Gas chromatography, with temperature programming, is employed and each of the drugs or any combination of them may be assayed on a single specimen during a single rapid determination on 3 percent OV-17 following chloroform extraction. Retention times relative to an internal standard (methyl myristate) are given. The recovery of the drugs is from 77 to 100 percent, the instrument response is linear for each drug, and the coefficients of variation are from 3 to 9 percent.

Introduction system, trimethadione, paramethadione, A recurring problem in the management ethosuximide, metharbital, methsuximide, of seizure disorders is the adjustment of phensuximide, mephenytoin, ethotoin, pri drug dosage to achieve a balance between midone and diazepam. Barbital appears seizure control and toxicity. It is recog following metharbital, and methyl myris nized that this is facilitated when the levels tate, the internal standard, follows barbital. of the involved drugs in the patient’s serum In figure 1 are shown the structures of the can be determined. 5 ’ 7 The two most com main classes of antiseizure drugs. Dia monly used seizure control drugs are di- zepam, a benzodiazepine derivative, shows phenylhydantoin and phénobarbital. Meth little resemblance to these in a flat draw ods for determining serum levels of these ing, but in three dimensional conformation agents are well established. 1 ’3 ’ 6 ’9 a proce bears a startling resemblance to diphenyl- dure is reported for the quantitative anal hydantoin. ysis for ten other seizure control drugs in Individual gas chromatographic methods serum, using a single extraction and injec have been reported for several of these tion on a vapor phase chromatograph. drugs. 2 ’ 6 ’ 8 However, the capability of han The ten drugs are, in increasing order dling all such analyses with a single system of retention time in our chromatographic obviates maintaining multiple columns and

3 6 9 3 7 0 GRIFFITHS, OLEKSYK, DEXTRAZE AND DIAMOND

HYDANTOINS Rx Rz Rz Ri DIPHENYLHYDANTOIN —NH— C6H5 C6H5 H ETHOTOIN —NH— CeHs H C2H: MEPHENYTOIN —NH— C6Hs C2H5 CH3

BARBITURATES O

PHENOBARBITAL —C—NH— C6H6 c 2h 5 H

O II BARBITAL —C—NH— c 2h 5 c 2h 6 H

O II PRIMIDONE (Carbonyl oxygen —C—NH— c 6h 5 c 2h 5 H at (2) is replaced by H2)

O II METHARBITAL —C—NH— C 2H 5 c 2h 5 CH;

OXAZOLADINEDIONES TRIMETHADIONE —O— CHa c h 3 CH. PARAMETHADIONE —O— C 2H 5 CHs CH;

SUCCINIMIDES

ETHOSUXIMIDE C2H5 c h 3 H METHSUXIMIDE c 6h 6 c h 3 CH; 1 1 1 1 OOO WWW PHENSUXIMIDE 1 1 1 c 6h 5 H CH:

F i g u r e 1. Structure of classes of antiseizure drugs. instrument conditions. Further, the savings silanized 80/100 Chromosorb W was em in time when several specimens are to be ployed.* analyzed or, as is often the case, when Chloroform and methanol were reagent more than one of the drugs is to be deter grade and used without further purifica mined, are obvious. Diphenylhydantoin tion. and phénobarbital are strongly column ab Methyl myristate was purchased.* sorbed and will produce no peak under the Acetate buffer, 0.2 M, pH 5.0. stated conditions if present even up to a level of 5 mg per dl. Standard Solutions There is also a more subtle advantage in The drugs used in the standard were this multiple drug analysis in that certain pure samples and free of filler. of these drugs, while being useful agents The standard solution contained 1 mg in their own right, are metabolites of other per ml of each of the drugs, and 0.25 mg drugs. Thus, primidone metabolizes to phé per ml of methyl myristate in methanol. nobarbital10 and metharbital metabolizes This is used for calibration of peak area, almost completely to barbital. 4 retention time and in the instrument linear R eagents ity study. It was also used to spike sera for the recovery studies. A six foot X 2 mm ID silanized glass column packed with 3 percent OV-17 on * Supelco, Bellefonte, PA 16823. GAS CHROMATOGRAPHIC DETERMINATION OF ANTICONVULSANT DRUGS IN SERUM 3 7 1

Apparatus TABLE I A Bendix model 2500 four column, four R e t e n t i o n T i m e s R e l a t i v e t o flame ionization detector gas chromato M e t h y l M y e i s t a t e graph was employed, f Hydrogen was sup Trimethadione 0.066 plied by an Elhygen hydrogen generator.! Paramethadione 0.103 The instrument readout was a Honeywell Ethosuximide 0.318 Metharbital 0.730 Model 194 dual pen recorder^ and an Barbital 0.925 Autolab Model 6300 digital integrator. || Air Methyl myristate 1.000 and helium carrier gas were from Airco'f Methsuximide 1.16 Phensuximide 1.25 and were used without further drying. Mephenytoin 1.51 Ethotoin 1.66 Procedures Primidone 2.23 Diazepam 2.35 To one ml of serum in a 50 ml conical tube are added one ml of acetate buffer and 10 ml of chloroform. The resulting each drug and plotting the peak areas ob mixture is shaken thoroughly and centri tained vs the amount injected. fuged. Of the chloroform extract, 8 ml are Recovery studies were carried out at the recovered and evaporated to a volume of 3 mg per dl and 5 mg per dl level of drug 0.2 to 0.5 ml under a stream of nitrogen at from spiked human serum samples shown room temperature, after which it is quanti to be originally free of drugs. tatively transferred to a 2 ml disposable culture tube and evaporated to dryness. Calculation The residue is taken up in 25 /xl of 0.25 For each drug: mg/ml solution of methyl myristate in peak area of drug in sample methanol, thus yielding a solution ready peak area of methyl myristate mg per dl = for injection. peak area of drug in std. The volatility of trimethadione, para- peak area of methyl myristate methadione, and ethosuximide precludes 25 yul the application of heat during the solvent X C X X 0.1 0 . 8 ml evaporation procedure. Injections of 2 to 4 ¡A of sample are where C = concentration of standard in made. The injection port temperature is jug per Ml- maintained at 260°, with the transfer zone R esults at 275° and the detector block at 300°. No component of normal, drug-free se The carrier gas flow rate is 50 ml per min rum interferes with any of the drugs deter (uncorrected). mined in the above method, although a The oven temperature is 100° at injec final serum peak with a long retention time tion, and is maintained for ten minutes, appears after the final drug. Retention during which trimethadione, parametha- times of the ten antiseizure drugs, plus dione, and ethosuximide are eluted. It is barbital, are shown in table I relative to then programmed at 3° per min up to 225°. methyl myristate as the internal standard. Linearity studies were performed by Separation is complete, and all peaks are making injections of 1, 2, 3, and 5 //I of symmetrical with the exception of that t Bendix Corporation, Ronceverte, WV 24970. from ethosuximide, which shows very | Milton Roy Co., St. Petersburg, FL 33733. slight tailing, and barbital, which shows § Honeywell, Fort Washington, PA 19034. || Vidar, Mountain View, CA 94040. moderate tailing. In figure 2 are results of ft Air Reduction Co., New York, NY. an injection of chloroform standard con- 3 7 2 GRIFFITHS, OLEKSYK, DEXTRAZE AND DIAMOND

TABLE II

R e c o v e r i e s o f A n t i s e i z u r e D r u g s f r o m S p i k e d S e r u m a n d C o e f f i c i e n t o f V a r i a t i o n o f E a c h A s s a y

% Recovery CV*

1. Trimethadione 75.0 9.3 2. Paramethadione 82.7 6.2 3. Ethosuximide 84.6 4.8 4. Metharbital 87.6 3.9 5. Methsuximide 90.2 3.4 6. Phensuximide 90.2 4.2 7. Mephenytoin 93.6 2.8 8. Ethotoin 91.7 4.2 9. Primidone 84.3 4.1 F i g u b e 2. Chromatographic scan of standard 10. Diazepam 100.7 6.5 plus added internal standard and barbital. Peaks represent 2 fig of each substance, except 5 jug of * Based on eight serial determinations. barbital. Figure represents photograph of actual recorder at each of two concentrations in serum, 3 output from two columns operating simulta neously, one pen scale having been traced over mg per dl and 5 mg per dl. They are all in india ink for clarity. One set of peaks has been quite close to quantitative. In our labora numbered to correspond with listing in table I. tory, tests results using the above proce Peak number 7 is methyl myristate, the internal standard. Total time span of figure is 50 minutes. dure are used uncorrected. The results of the instrument linearity taining an internal standard and barbital. study are shown in figures 4 and 5. The In figure 3 are results of analysis of a hu man serum spiked with each drug at a level of 5 mg per dl. The results of the recovery studies are shown in table II. The recoveries shown are averages of at least four studies done E E H X o LU X if: < I d CL

F i g u b e 3. Chromatographic scan resulting from analysis of a clean human serum sample to which all drugs have been added at a level of 3 mg per dl, except barbital at 8 mg per dl. Figures represent photograph of actual recorder output, traced over in india ink for clarity. Final 0 1 2 3 4 5 peak is a normal human serum component. Peaks have been numbered to correspond with listing in AMOUNT INJECTED (>ig) table I. Peak number 7 is methyl myristate, the internal standard. Peak number 13 is the long F i g u r e 4. Peak area vs micrograms injected for retention time serum component. Total time span methsuximide, paramethadione, ethotoin, ethosux- of figure is 80 minutes. imide, and primidone. GAS CHROMATOGRAPHIC DETERMINATION OF ANTICONVULSANT DRUGS IN SERUM 3 7 3

prudent to be aware of the extraction and chromatographic behavior of all drugs which the monitored patient is taking. The therapeutic serum level of valium is one to two orders of magnitude less than that of the other drugs in the scheme. However, valium shows very little irrever sible column absorption, has good sensitiv ity in the flame ionization detector and can be successfully determined by the proce dure described. The procedure described is a rapid, spe cific and precise method for the analysis for antiseizure drugs in serum. It provides a measure of practicality unobtainable when individual assays must be maintained and employed to cover all the drugs used in epilepsy control.

References

1. B a r r e t t , M. J.: An integrated gas chromato graphic program for drug screening in serum AMOUNT INJECTED (jig) and urine. Clin. Chem. Newsletter 3:1-10, 1971. F i g u r e 5. Peak area vs micrograms injected for 2. B o o k e r , H. E. a n d D a r c e y , B.: Simultaneous trimethadione, phensuximide, mephenytoin, dia determination of trimethadione and its metab zepam, and metharbital. olite, dimethadione, by gas-liquid chromatog raphy. Clin. Chem. 17:607-609, 1971. response of the instrument to each of the 3. G o l d b a u m , L . R.: Analytical determination of drugs is linear at least to 5 /¿g, and each of barbiturates. Anal. Chem. 24:1604-1607, 1952. *the linearity curves intercepts quite close 4. G o o d m a n , L. S. a n d G i l m a n , A.: The Phar to the origin, indicating minimal column macological Basis of Therapeutics, Macmillan, absorption. New York, p. 214, 1970. 5. K o c h -W e s e r , J.: Drug therapy: serum drug The coefficients of variation, as shown in concentrations as therapeutic guides. New table II, are based on at least eight deter Eng. J. Med. 287:227-231, 1972. minations of each drug in a serum pool, 6. K u p f e r b e r g , H. J.: Quantitative estimation of diphenylhydantoin, primidone and phenobar with assays performed by two people over bital in plasma by gas-liquid chromatography. •.a period of four days. Clin. Chem. Acta 29:283-288, 1970. 7. M i l l i c h a p , J. G.: Treatment of convulsive Sources of Interference disorders. New Eng. J. Med. 257:105, 1972. 8. P i p p e n g e r , C. E. a n d G i l l e n , H. W.: G a s Phenobarbital and diphenylhydantoin do chromatographic analysis for anticonvulsant not interfere (vida supra). Other barbit drugs in biologic fluids. Clin. Chem. 15:582- urates, glutethimide, methyprylon and 590, 1969. 9. W a l l a c e , J. E.: Microdetermination of di methaqualone, will give peaks within the phenylhydantoin in biological specimens by retention time of the drugs covered in the ultraviolet spectrophotometry. Anal. Chem. procedure discussed. However, it is rare for 40:978-980, 1968. 10. W o o d b u r y , D. M., P e n r y , J. K., a n d S c h m i d t , these to be used in eases of seizure dis R. P.: Antiepileptic Drugs, Raven, New York, order. As with any such procedure, it is pp. 361-366, 1972.