A Comprehensive Gas Chromatography Procedure for Measurement of Drugs in Biological Materials

JAMES R. SHIPE, B.S. AND JOHN SAVORY, Ph .D.

Division of Clinical Chemistry, Departments of Medicine, Biochemistry and Pathology and Hospital Laboratories, The University of North Carolina, Chapel Hill, NC 27514

ABSTRACT Procedures have been developed for the determination in serum and urine of 21 drugs which include all of the barbiturates, methyprylon, glutethimide, meprobamate, propoxyphene, methaqualone, primidone, diphenylhydantoin, methadone, codeine, morphine, amphetamine and methamphetamine. The method involves pipeting 1.0 ml of serum or urine into an extraction tube, adding an internal standard together with either a pH 4.9 or pH 8.3 buffer and extracting with chloroform. The drugs in the extract are concentrated by evaporation of the organic layer and identified and quantitated by gas-liquid chromatography (GLC) using a column packed with 3 percent OV 17. GLC analysis times are kept below seven minutes by using temperature program ming which allows good separations of drugs of widely varying volatilities. All calculations are based on peak areas determined by an electronic inte grator. Losses of drugs during extraction, evaporation and chromatography steps are compensated for by the use of internal standards. Actual extraction efficiencies of drugs from serum or urine range from 75 to 100 percent although recoveries, based on standards carried through the entire procedures, are quantitative. Precision (s.d.) based on replicate determinations is approx imately ± 0.2 mg per dl.

Introduction the final analytical measurement of drugs. Increasing emphasis is being placed on Resolution of different compounds is ex identification and quantitation of drugs in cellent, low limits of detection are possible patients suspected of overdose. In addition, and the technique affords high precision. the monitoring of therapeutic levels of a Numerous reports have been made of number of drugs, especially those in the the measurement of individual drugs or antiepileptic category, is of considerable groups of drugs using gas chromatographic importance.7’9’13 techniques.1’3’4’6’10’11’12 However, the clin For most clinical laboratories, gas chro ical laboratory scientist has the problem of matography offers the best technique for making a selection from this wide variety

57 5 8 SHIPE AND SAVORY

TABLE I chromatographic conditions are outlined in Gas Chromatographic Conditions table I.

Carrier gas N2 45 ml per min C o l u m n s Hydrogen flow rate 50 ml per min

Air flow rate 1.4 cubic ft per hr Glass U-shaped columns 3 ft X 2 mm in

Detector temperature 250° C ternal diameter are packed with 3 percent

Infector temperature 230° C OV-17 on Gas-Chrom Q 100-120 mesh. This packing is available commercially. § Column temperature Sedatives 140 to 240° C The columns are conditioned with normal program at 20° per min Amphetamines 150° C isothermal carrier gas flow by heating at 100° for one Alkaloids 240° C isothermal hour and increasing the temperature 2° per Electrometer range 200 on Model 2600 minute up to 240°. This final temperature Chart speed 1 in per min is maintained for 18 hours. Columns condi Recorder Fall scale range - 1 mV tioned in this manner will give excellent resolution for approximately one year. of methods to fulfill the needs of his labo S o l v e n t E v a p o r a t io n B a t h ratory. A fairly comprehensive approach to drug measurements was reported by Bar A water bath maintained at 75° is used rett1 and involved the determination of the to evaporate solvents following extractions. barbiturates, glutethimide, amphetamines, Nitrogen is used to aid the evaporation and alkaloids and the common tranquilizers. is supplied through a glass manifold and The present study uses most of the basic directed into the evaporation tubes through methodology described by Barrett but also Pasteur pipets. involves several modifications which in clude temperature programming for drug R eagen ts screening purposes and for monitoring B arbiturates , N e u t r a l s , || D il a n t i n therapeutic levels of the antiepileptic drugs. Sodium acetate buffer ( 1.0 M, pH 4.8). Precisely 410 ml of 1.0 N acetic acid and Special Apparatus 590 ml of 1.0 N sodium acetate are mixed and the pH adjusted to 4.8 with either G a s C hromatograph acetic acid or sodium acetate. A dual-column instrument* equipped Barbital internal standard. A stock solu with dual hydrogen flame ionization de tion of 100 mg per 100 ml is made up in tectors, dual electrometers and linear tem methanol. perature programming is employed. This Buffer (containing internal standard). system is coupled to two strip chart re Two ml of the barbital internal standard corders f and an electronic digital inte stock solution are diluted to 100 ml with grator equipped with a printer, j The gas acetate buffer. Diphenylhydanoin internal standard. A ° Model 2600 Bendix Corporation, Process In struments Div., Drawer 477, Ronceverte, WV stock solution is prepared by dissolving 100 24970. f Model SRG, Sargent-Welch Scientific Co., § Applied Science Laboratories, Inc., P.O. Box 7300 N. Linder Ave., Skokie, IL 60076. 440, State College, PA 16801. | Model 6230, Vidar Autolab, 77 Ortega Ave., || Neutrals are those drugs which extract at any Mountain View, CA 94040; NOW Div. of Spectra- pH. Those commonly encountered in the clinical Physics, 655 Clyde Ave., Mountain View, CA laboratory are: methyprylon, gluthethimide, and «4040. meprobamate. DRUG MEASUREMENT IN BIOLOGICAL MATERIALS BY GAS CHROMATOGRAPHY 5 9 mg 5-(p-methylphenyl)-5-phenyl hydan- S t a n d a r d S o l u t io n s toinfi in 100 ml of chloroform. Sedative slock standards. The pure drugs Chloroform (containing Dilantin inter were obtained from the appropriate manu nal standard). Two ml of the stock Dilan facturer. A stock standard of each of the tin internal standard are diluted to one following drugs was prepared by dissolv liter with redistilled chloroform. ing 100 mg (if present as sodium salt, Trimethylanilinium hydroxide (TMAH) volume weight for free acid must be in methanol ( 0.2 M ). Thirty ml of N, N- corrected) of each compound in 100 ml dimethyl aniline* are dissolved in 200 ml of methanol: methyprylon, aprobarbital, of ethyl acetate and added to 25 ml of butabarbital, amobarbital, pentobarbital, methyl iodide. The reaction mixture is al secobarbital, glutethimide, propoxyphene, lowed to stand overnight at room tempera methaqualone, phenobarbital, primidone, ture and the product, N,N, N-trimethylani- diphenylhydantoin, flurazepam. line iodide, recovered by filtration followed Sedative working standards. One ml of by recrystallization from hot absolute eth each of the stock sedative standards is anol. Exactly 5.2 g of the iodide are dis combined and diluted to 100 ml with solved in 100 ml of absolute methanol to water. The final concentration of each drug which are added 3.5 g of powdered silver is then one mg per 100 ml. oxide. The slurry is mixed at room temper Stock amphetamine standards (2.5 mg ature for one hour and the silver iodide per 100 ml). Exactly 42.75 mg of ampheta removed by filtration. The TMAH reagent mine sulfate and 31.10 mg of methamphet- is then stored in a 120 ml ground-glass amine HC1 are dissolved in one liter of stoppered bottle. water. Stock alkaloid standards (100 mg per 100 m l). Each of the following compounds is A l k a l o id s a n d A mphetamines dissolved in 100 ml of water: 131.6 mg Borate buffer (pH 8.3). A saturated so codeine phosphate, 133.6 mg morphine sul lution of sodium borate is adjusted to pH fate and 100 mg methadone. 8.3 with 2 N NaOH. Alkaloid and amphetamine working stan Amphetamine internal standard. Five mg dards. One ml of each of the stock alka N-propylamphetamine HClf are dissolved loids is combined with 20 ml of the am in 100 ml of water. phetamine stock standard and diluted to 100 ml with water. The final concentration Alkaloid internal standard. Five hundred of each alkaloid is then one mg per 100 ml mg of nalorphine hydrochloride are dis and 0.5 mg per 100 ml for the ampheta solved in 100 ml of water. mines. 'Nalorphine buffer. One ml of the stock nalorphine internal standard is diluted to Procedures 100 ml with borate buffer. S e d a t iv e a n d A nticonvulsant D r u g s Re-distilled acetic anhydride. Reagent grade acetic anhydride is distilled through One ml of serum or standard, one ml of a 12 inch Vigreaux column. acetate buffer and 10 ml of chloroform are added to a 15 ml ground-glass stoppered If Aldrich Chemical Co., Inc. centrifuge tube. For diphenylhydantoin de * Eastman Kodak. terminations, the chloroform contains the f Generously provided by J. W. Brackett, San Mateo County Coroner Toxicology Laboratory, diphenylhydantoin internal standard. The San Mateo, CA 94403. tubes are shaken for one minute to extract 60 SHIPE AND SAVORY

phetamine and 20 ml of chloroform are combined in a 50 ml ground-glass stop pered centrifuge tube. The tubes are shaken vigorously for one minute, centri fuged and the upper aqueous layer re

9 moved by aspiration and discarded. The 14 organic phase is dried by shaking with one g of anhydrous sodium sulfate. A 10 ml aliquot of the chloroform is combined with 0.2 ml acetic anhydride in a 15 ml conical centrifuge tube and the contents evaporated to dryness. The res idue is re-dissolved in 25 /xl of chloroform and 3 /*1 of the extract are injected onto the gas chromatograph column at 240° for the detection of morphine, codeine and meth adone. The oven temperature is then low

J------1______I______1______I______1______L ered to 150° and 3 /xl of the extract are 2 4 6 TIME(minutes) again injected on the column for the mea surement of amphetamine and metham- F i g u r e 1. Typical chromatogram from analysis of an aqueous drug standard at a concentration of phetamine. one mg per dl: (1) Barbital, (2) methyprylon, (3) aprobarbital, (4) butabarbital, (5) amobar- Calculation bital, (6) pentobarbital, (7) secobarbital, (8) glutethimide, (9) phenobarbital, (10) propox Step 1. From the integration printout, yphene, (11) methaqualone, (12) primidone, the ratio of the peak area of the specific (13) diphenylhydantoin, (14) diphenylhyantoin drug is calculated to that of the respective internal standard and (15) flurazepam. internal standard. the drugs into the chloroform layer. Gentle shaking is essential for the serum to avoid (n s S s ii)s‘d- E“fe”> emulsion formation. The tubes are then A = number of integrator counts centrifuged for two minutes and the tipper peak area* aqueous layer is aspirated and discarded. Step 2. Similarly, the ratio is calculated Nine ml of the chloroform layer are pi- for the drug suspect in the Test (Patient’s) peted into 15 ml conical centrifuge tubes printout. and the contents evaporated to dryness in the 75° water bath by directing a gentle stream of nitrogen over the chloroform. (AissSfsis:)Twt" Eatio™ The residue after evaporation is re-dis Step 3. From the calculations in steps 1 solved in 25 /xl of chloroform and a 3 /xl and 2, the calculation is made of the serum aliquot is injected onto the column of the or urine level of the drug in the patient. gas chromatograph and temperature pro Drug level (mg per 100 ml) gramming carried out from 140° to 240° at _ Ratio Test a rate of 20° per minute. Ratio Std. X concentration of standard A l k a l o id s a n d A mphetamines * Peak areas can be determined by other means Ten ml of urine or aqueous standard, 5 such as planimetry or triangulation if an integrator ml of nalorphine buffer, 0.5 n-propylam- is not available. DRUG MEASUREMENT IN BIOLOGICAL MATERIALS BY GAS CHROMATOGRAPHY 61

J TIME(minutes)

F ig u r e 2. Chromatogram obtained from serum toxicology control, all drugs present in 2 mg per dl concentrations: (1) Barbital, (2) methyprylon, (3) amobarbital, (4) glutethimide, (5) meprobamate, (6) phénobarbital, (7) methaqualone, (8) diphenylhydantoin and (9) diphenylhydantoin internal standard. F ig u r e 3. Chromatogram of urine alkaloid screen with each drug being present at one mg per dl concentrations: (1) Methadone, (2) codeine, (3) morphine and (4) nalorphine.

Results and Discussion shown in figure 1. The use of an internal standard, barbital, greatly increased the S e r u m S e d a t iv e D r u g P r o c e d u r e precision of the method by negating the In the procedure for the measurement of effect of sample loss during extraction, the sedative drugs, both isothermal and evaporation and gas chromatographic steps. temperature programming conditions for Internal standards, therefore, were used for operation of the gas chromatograph were all drug measurements described in the investigated. It was found that tempera present study. ture programming offered advantages of The precision of the method was evalu shortening the gas chromatograph analysis ated for phenobarbital and diphenylhydan- time to seven minutes and also minimized tion by performing 12 replicate analyses on diffusion of higher boiling components, different aliquots of a commercial lypho- such as phénobarbital thus giving sharp lyzed toxicology control serum, f Phenobar- symmetrical peaks. A typical chromato gram of the standard mixture of drugs is f Lederle Diagnostics, Pearl River, NY 10965. 62 SHIPE AND SAVORY

- 0 I 2 3 2 4 6 8 TIME(minutes) TIME (minutes)

F ig u r e 4 . Chromatogram of urine amphetamine screen, the amphetamines being present at a level of 0.5 mg per <31: (1) Amphetamine, (2) methamphetamine and (3) N-propylamphetamine. F i g u r e 5. Chromatogram of a standard mixture of glutethimide and meprobamate at a con centration of one mg per dl: (1) Barbital, (2) glutethimide and meprobamate and (3) diphenylhydantoin internal standard. F i g u r e 6. Chromatogram of methyl derivatives of glutethimide and meprobamate: (1) Bar bital, (2) glutethimide, (3) meprobamate and (4) diphenylhydantoin internal standard. bital gave a mean of 1.99 mg per dl with with the measurement of methprylon. To a standard deviation of ±0.14 mg per dl illustrate the performance of the sedative and a relative standard deviation (RSD) drug method, a chromatogram obtained of 7.37 percent. Corresponding data for during the analysis of the lypholyzed com diphenylhydantoin was 2.00 ± 0.05 mg per mercial control serum is shown in figure 2. dl with a RSD of 2.49 percent. These This chromatogram demonstrates excellent values approximate the day-to-day preci resolution with minimal peak tailing. sion which was evaluated by plotting stan dard curves for phenobarbital and diphen- U r i n e A l k a l o id a n d A m p h e t a m i n e ylhydantion daily over a period of four P r o c e d u r e s months. As in the method for the sedative drugs, The major differences in the present pro internal standards were employed for both cedure from that described by Barrett1 are alkaloid and amphetamine measurements. that a single extraction has been used for Nalorphine and N-propylamphetamine, both the anticonvulsant and the sedative which is a homologous derivative of am drugs, and only one set of instrumental phetamine, were incorporated into the conditions, aside from column temperature, combined extraction procedure. N-acetyl have been employed. In addition, barbital derivatives of the alkaloids and ampheta has been substituted for Sedulon® as an mines were formed prior to the gas chro internal standard since the latter interferes matographic analysis of the urine extracts. DRUG MEASUREMENT IN BIOLOGICAL MATERIALS BY GAS CHROMATOGRAPHY 63

In figure 3 is demonstrated a typical sample can be re-extracted with buffer not chromatogram from the analysis of a urine containing barbital. Quantitation can be toxicology control* which provides excel accomplished by using aqueous barbital lent sensitivity for methadone and codeine standards and another internal standard. shown in this analysis at levels of one mg For example, the diphenylhydantion inter per dl. Morphine levels as low as 0.5 mg nal standard may be used in such a situa per dl can be detected in standard solu tion. In the urine amphetamine screen, if tions. In urine samples, however, 85 per the level of methamphetamine is much cent of morphine exists in the conjugated higher than that of amphetamine, then form which will not be measured in the most of the amphetamine probably came previously mentioned procedure without a from the metabolism of methamphetamine. prior hydrolysis step such as has been de Some drugs with retention times similar scribed by Kullberg and Gorodetzky.8 For to those included in the present study may a sensitive reliable method for morphine, also be extracted by the proposed proce such a hydrolysis step should be included dure. To confirm the presence and identity in the procedure. of a drug, the serum extracts are “spiked” In figure 4 is shown a chromatogram of with chloroform solutions of the drug in the amphetamine method using the same question. Thus, a drug eluting close to urine toxicology control. Different column secobarbital will not be reported as seco temperatures were required for the am barbital when the “spiked” chromatogram phetamines and alkaloids owing to the has two peaks in that vicinity. great differences between the volatilities The problem of different drugs having of their N-acetyl derivatives. identical retention times is less easily re solved. Several such interferences have Linearity Studies been found: methohexital and pentobarbi Standard solutions of the sedative and tal, glutethimide and meprobamate, with anticonvulsant drugs at concentrations of others being possible. Positive identifica 10, 5, 2.5, 1.0 and 0.5 mg per dl were tion of these unresolved compounds can carried through the entire analysis proce best be accomplished by utilizing a mass dures. In all cases, the ratio of peak area spectrometer as the detector for the gas of drug to that of the internal standard was chromatograph. In some instances, separa linearily related to the concentration. tion can be achieved by using another sta tionary phase or by forming derivatives Sources of Interference and re-chromatographing. Glutethimide and meprobamate can be resolved by add Sera from normal individuals taking no ing 25 /xl of TMAH reagent and re-chro medication were checked for possible in matographing. In figure 5 is shown the terferences in the proposed methods. No single peak obtained in the screening pro endogenous gas chromatography peaks cedure which can be separated by the were observed in the extracts from any of methylation of glutethimide and mepro these samples. Patients who have injested bamate as shown in figure 6. barbital or a drug which metabolizes to The method described was specifically barbital, e.g. metharbital,5 will give falsely designed to detect drug overdoses and has high internal standard responses. This been used effectively for over a year on a problem is easily detected by the abnor 24-hour a day basis by medical technol mally large internal standard peak and the ogists having no highly specialized training * Lederle Diagnostics. in gas chromatography. The results seda 6 4 SHIPE AND SAVORY tive screen are available in less than 30 5. G o o d m a n , L. S. a n d G i l m a n , A.: The Phar minutes after receiving the sample, and macological Basis of Therapeutics. Macmillan, New York, p. 214, 1970. one technologist could routinely analyze 6. G r i f f i t h s , W . C., O l e k s y k , S. K ., D e x t r a z e , up to 48 samples per day. P., a n d D ia m o n d , I.: The gas chromato The sensitivity of the method has been graphic determination of anticonvulsant drugs- in serum: Ann. Clin. Lab. Sci. 3:369-373, adequate for such overdose cases and has 1973. also been applied with considerable suc 7. K o c h -W e s e r , J.: Drug therapy: Serum drug cess to monitoring therapeutic levels of the concentrations as therapeutic guides. New Eng. J. Med. 287:227-231, 1972. anticonvulsant drugs phenobarbital, primi 8. K u l l b e r g , M. P. a n d G o r o d e t z k y , C. W.: done and diphenylhydantoin. Studies on the use of XAD-2 resin for detec tion of abused drugs in urine. Clin. Chem. References 20:177-183, 1974. 9. M e l l ic h a p , J. G .: Treatment of convulsive 1. B a r r e t t , M. J.: An integrated gas chromato disorders. New Eng. J. Med. 287:105, 1972. graphic program for drug screening in serum and urine. Clin. Chem. Newsletter 3:1-10, 10. P r o e l e s , H. F . a n d L o h m a n n , H. J.: Profile 1971. of sedatives and tranquilizers in serum as measured by gas-liquid chromatography. Clin. 2. B l o o m e r , H. A ., M a d d o c k , R. K., S h e e b e , Chem. 17:222-228, 1971. J. B ., a n d A d a m s , E. J.: Rapid diagnosis of sedative intoxication by gas chromatography. 11. R e id , R . W ., K a t z e n , R ., a n d C l i x c e r , J. M.: Ann. Int. Med. 72:223-228, 1970. Analysis of blood and other body fluids by 3. E v e n s o n , M. S., J o n e s , P., a n d D a r c e y , B.: gas chromatography: Simplified rapid separa Simultaneous measurement of diphenylhydan tion and quantitation of toxicologic sub toin and primidone in serum by gas-liquid stances. Amer. J. Clin. Pathol. 53:462-467, chromatography. Clin. Chem. 16:107-110, 1970. 1970. 12. S i n e , H. E., M c K e n n a , M . J., L a w , M . R ., 4. F l a n a g a n , R. J. a n d W it h e r s , G.: Rapid a n d M u r r a y , M . H.: Emergency drug anal micro method for the screening and measure ysis. J. Chromatog. Sci. 10:297-302, 1971. ment of barbiturates and related compounds 13. W in e k , C. L.: Laboratory criteria for the ade in plasma by gas-liquid chromatography. J. quacy of treatment and significance of blood Clin. Path. 25:899-904, 1972. levels. Clin. Toxicol. 3:541-549, 1970.