2c

CLIN. CHEM. 29/6, 1097-1100 (1983)

Reversed-Phase Liquid-Chromatographic Simultaneous Analysis for Thiopental and Pentobarbital in Serum Michael Kelner and David N. Bailey1 We describe a reversed-phase liquid-chromatographic assay suitable for therapeutic monitoring of thiopental and pentobarbital simultaneously in human serum. The drugs are extracted from serum at pH 6.6 into n-butyl chloride containing thiamylal and barbital as the respective internal stan-

dards. The compounds are back-extracted into dilute sodium hydroxide, an aliquot of which is submitted to chromatography. The lowest measurable concentrations are 1.0 mg/L for thiopental and 2.0 mg/L for pentobarbital. The standard curve is linear from 0 to 100 mg/L for both. Between-run CVs are: at 25 mg/L, 4.1% (thiopental) and 3.2% (pentobarbital); at 50 mg/L, 2.8% (thiopental) and 3.4% (pentobarbital). Data on patients receiving thiopental and pentobarbital illustrate use

HNNH

I

,2c

and short-acting barbiturates, respectively, the drugs are rapidly metabolized (4); yet relatively high concentrations in serum must be maintained for adequate treatment of intracranial hypertension (1-3). As a result, monitoring of the serum concentrations is important in this patient population. Thiopental and pentobarbital have been measured by spectrophotofluorometry (5), ultraviolet spectrophotometry (6, 7), gas-liquid chromatography (8,9), and “high-pressure” liquid chromatography (10-16). Spectrophotometric techniques are nonspecific and are relatively nonsensitive (6, 7).

methods require laborious de-

rivatization or the use of special detectors,or both (8, 9). In addition, thiopental may undergo desulfuration(oxidation) at the high temperatures required (10,11). Published liquidchromatographic procedures suffer either from use of internal standards that themselves are prescribed drugs (10), no sample “clean-up” (10-14), evaporation steps that are lengthy and risk oxidation and drug absorption onto glassware (10, 11, 14), or lack of validation of the technique on patient serum (11, 15, 16). Finally, in our experience, both thiopentaland pentobarbital may be administered simultaneously in the treatment of intracranial hypertension, and pentobarbital is also present as a metabolite of thiopental through hepatic oxidation (3,6). However, to our knowledge, no method for measuring the two compounds simultaneously has been described. We report here a reversed-phase “high-pressure” liquidchromatographic analysis for thiopental and pentobarbital simultaneously in serum, which avoids evaporation steps.

Division of Clinical Pathology, University

of California Medical

Center, 225 Dickinson Street, San Diego, CA 92103. Author to whom correspondence should be addressed. Received Jan. 31, 1983; accepted Mar. 17, 1983. 1

HN’NH

I

R1

0 R1

A2

(A)

Thiopental and pentobarbital (Figure 1) are widely used in the management of intracranial hypertension associated with metabolic coma and head injury (1-3). As ultra-short

chromatographic

I’

0

/\

of the method.

Gas-liquid

S

0

R2

(B)

Fig. 1. Structures of: (A) pentobarbital [R1

= -CH2CH3, R2 = -CH(CH3)CH2CH2CH3J and barbital [A1 = -CH2CH3, A2 = -CH2CH3J; (B) thiopental [A1 = -CH2CH3, A2 = -CH(Cl-13)CH2CH2CH3I and thiamylal [R1 = -CH2CH=CH2, R2 = -CH(CH3)CH2CI-12CH3]

Materials and Methods Apparatus We used a dual-pump high-pressure liquid chromatograph (Series 2) equipped with a variable-wavelength ultraviolet detector (ModelLC-85), a recorder (Model PE-024), and a 6-L loop injector(all from Perkin-Elmer Corp., Norwalk, CT 06582). The column was octadecyl sulfate (C18), 5-.zm particlesize, 4.6mm diameter, and 12.5cm length (Perkin-Elmer Corp.). The detector was set at 240 nm.

Reagents and Standards All reagents were analytical

(A.a)

grade unless otherwise

specified. n-Butyl chloride, distilled in glass (Burdick & Jackson Labs, Inc., Muskegon, MI 49442). Methanol, HPLC grade (J. T. Baker Chemical Co., Phillipsburg, NJ 08865). Tetrahydrofuran, HPLC grade (Fisher Scientific Co., Pustin, CA 92680). Disodium hydrogen phosphate (Naj-IPO..). Potassium dihydrogen phosphate (KHPO4). Sodium hydroxide, 0.45 mollL, in water. Prepare freshly each month, temperature.

and store in a polyethylene

bottle at room

Phosphate buffer, 0.16 moliL, pH 6.6. The water used to prepare this reagent was ifitered through a Teflon filter of 45-pin pore size (Gelman Sciences, Ann Arbor, MI 48166). Dissolve 22 g of KH2PO4 in water and dilute to 1000 mL with filtered water. Dissolve 18 g of Na2HPO4 in water and dilute to 1000 mL with ifitered water. Mix equal volumes of the two solutions. Check the pH. It should be 6.6 ± 0.2. Filter the buffer through the Teflon filter. Store at room temperature. Prepare freshly each month. CLINICAL CHEMISTRY, Vol. 29, No. 6, 1983

1097

Aqueous mobile phase (phosphate buffer/tetrahydrofuran, 86/14 by vol.). Prepare freshly each day. Stock standards in methanol. Store all standards at -20 #{176}C. Prepare freshly each month. Barbital (internal standard), 5 gIL. Dissolve 20 mg of barbital (Merck Sharp & Dohme, West Point, PA 19486) in 4 mL of methanol. Pentobarbital, 12.5 gIL. Dissolve 27.4mg of sodium pentobarbital (Applied Science Laboratories, State College, PA 16801), which is equivalent to 25 mg of pentobarbital free acid, in 2 mL of methanol. Thiopental, 12.5 gIL. Dissolve 54.5 mg of sodium thiopental (Abbott Laboratories, North Chicago, IL 60064), which is equivalent to 50 mg of thiopental free acid, in 4 mL of methanol. Thiamylal, 2 gIL. Dissolve 22.1 mg of sodium thiamylal (Parke-Davis, Tustin, CA 92680), which is equivalent to 20 mg of thiamylal free acid, in 10 mL of methanol. Dilute internal standards. Dissolve 200 L (1000 tg) of the barbital stock standard and 200 L (400 pg) of the thiamylal stock standard in each 100 mL of n-butyl chloride used for extraction. This provides 50 tg of barbital and 20 g of thiamylal per milliliter of serum extracted. Working standards in serum. Dilute the thiopental and pentobarbital stock standards appropriately with drug-free serum to yield concentrations in the range desired (usually 0-100 mg/L). Prepare freshly each day. Serum control, 50 mg/L. Dissolve 200 L each of the thiopental and pentobarbital stock standards in drug-free serum and dilute to 50 mL with drug-free serum. Store in small aliquots at -20 #{176}C. Prepare freshly every two weeks.

Procedure All glassware was soaked in 1 molIL HC1 at least overnight, rinsed with distilled water, and dried. Blood samples were centrifuged, and the serum was stored at -20 #{176}C until analysis no more than a week later. Pipet 2 mL of serum (samples, working standards, and control) into 15-mL glass culture tubes equipped with Teflon-lined screw caps. Add 1 mL of phosphate buffer to each tube and vortex-mix. Pipet 10 mL of n-butyl chloride (containing the internal standards) into each tube and extract vigorously for 3 mm. Centrifuge (3000 x g for about 5 mm) and transfer the solvent (upper) layer to 12-mL glass centrifuge tubes equipped with Teflon-lined screw caps. Add 100 L of sodium hydroxide, 0.45 mol/L, to each tube, and extract vigorously for 3 mm. Centrifuge until the aqueous (lower) phase is clear (about 10 mm). Carefully remove 15 tL of the aqueous (lower) layer and submit it to chromatography in the mixed-phase mode using the following mobilephase composition at a flow rate of 2.0 mLlmin: pump A, methanol (50%); pump B, aqueous mobile phase (50%). Identifr thiopental and pentobarbital from the retention times relative to those of their respective internal standards (RRT), thiamylal and barbital: thiopental, RRT 0.81; pentobarbital, RRT 2.1. Calculate the peak-height ratio (ratio of the peak height of the analyte to that of its internal standard) for thiopental and pentobarbital. Determine the concentration of each in serum by comparison of the peakheight ratios to those of the extracted serum standards.

Results A standard curve was prepared by adding thiopental and pentobarbital to drug-free serum. The plot of concentration against peak-height ratio was linear for each drug over the concentration range studied (0-100 mg/L). The “leastsquares’S regression equations were: thiopental, y = 21.784x 0.689 (r = 0.999); pentobarbital,y = 70.434x - 1.300 (r = 1098

CLINICAL CHEMISTRY, Vol. 29, No. 6, 1983

A

B

C

D

C

0

0

A A

A

lB II

C

B I 0246

I

I

0246

I 024

TIME

I

I B

024

(MINUTES)

Fig. 2. Liquid chromatogramsof extracts of (A) drug-free serum; ( serum reference, 25 rng/L;(C) serum from patientno. 3, who was receMng thiopental (pentobarbital,4.0 mg/L;thiopental, 4.4 mg/L); and (C) serum from patientno. 2, who was receiving thiopental (pentobarbital, 12.1 rng/L; thiopental, 2.9 mg/L). Peak identification: a, barbital internal standard; b, pentobarbital; C, thiopental; and d, thiamylal 0.998). Although

we could detect at least 100 ng of each drug as the pure compound, the serum “background” limited the working analytical sensitivity to 1.0 mgfL for thiopental and 2.0 mg/L for pentobarbital.

On analysis of sera collected in commercial glass evacuated blood-collection tubes (Vacutainer Tubes; Becton-Dickinson, Rutherford, NJ 7070) from 10 drug-free volunteers we observed no interfering chromatographic peaks (Figure 2). Analytical recoveries were calculated by comparison of peak heights from extracted serum standards with those from non-extracted standards of the compounds prepared in methanol. The uncorrected analytical recovery for thiopental ranged from 70 to 80% and for pentobarbital, from 60 to 65%, over the concentration range studied (0- 100 mg/L). Error introduced by these incomplete recoveries was compensated for by the use of simultaneously extracted standards prepared in serum. The analysis of 20 aliquots of drug-free serum supplemented with thiopental and pentobarbital revealed the following coefficients of variation (CVs) within run: thiopental 5.5% and pentobarbital 6.0% at a concentration of 25 mgI L; thiopental 5.7% and pentobarbital 6.8% at 50 mgIL. Twenty measurements of a pooled serum on different days gave the following between-run CVs: thiopental 4.1% and pentobarbital 3.2% at 25 mg/L; thiopental 2.8% and pentobarbital 3.4% at 50 mgIL. The concentrations of the two compounds in serum showed no change with time when the serum was stored at -20 #{176}C over the two weeks studied. The sodium hydroxide extracts were stable when re-analyzed at 4 h; however, they showed 60-70% deterioration of pentobarbital when stored overnight at 4 #{176}C. We evaluated 25 compounds (Table 1) for potential interference in the proposed assay. The compounds were initially chromatographed as their pure standards under the conditions of the assay. Compounds that chromatographed within 0.4 mm of thiopental, pentobarbital, or the thiamylal and barbital internal standards were added to drug-free serum at a concentration of 100 mg/L, and they were submitted to the complete extraction procedure. No interferences with thiopental or its internal standard, thiamylal, were encountered. However, amobarbital was extracted with and cochromatographed with pentobarbital. Acetaminophen, theophylline, and ethosuximide were extracted with and cechromatographed with the barbital internal standard, although we were often able to resolve acetaminophen and theophylline from barbital when the former were present in

Table 1. Compounds Studied for Chromatographic Interference Retention time, minutes

Compound

a

Ampicillin

a

Penicillin G

Vaiproic acid Ascorbic acid

Discussion

6

0.2 0.2 0.2 0.2

Chlorothiazide

Dicloxacillin Oxacillin Cefazolin Sulfamethoxazole

To our knowledge, no procedures as yet published are suitable for the simultaneous measurement of thiopental and pentobarbital in serum. Unlike other methods, our analysis does not require evaporation steps, which are

0.3

Acetazolamide

notoriously lengthy and risk oxidation of thiopental as well as adsorption onto glassware. The use of a back-extraction step both eliminates bases and lipids and permits direct injection of the aqueous phase, because reversed-phase chromatography is utilized. The mobile phase is sufficiently well buffered at pH 6.6 that introduction of dilute alkali is tolerated by the column for several hundred injections.

04 0.4

Acetylsalicylic acid Furosemide

0.4 0.4

Hydrochlorothiazide Salicylic acid Acetaminophen

addition, we analyzed sera collected in Vacutamers from patients who were receiving continuous intravenous infusion of pentobarbital for management of head injury. Table 2 summarizes our data from these analyses of serum. Figure 2 illustrates representative chromatograms.

0.4

0.6

Theophylline Barbital

Ethosuximide Phenobarbital

0:8 1.2 1.2

Phenylbutazone

Butabarbital Chlorzoxazone Ibuprofen Phenytoin

1.4

1.6 20

Arnobarbital Pentobarbital

2:4C

2.4 3.0 4.1 5.2

Secobarbital

Thiopental Thiamylal a These compounds

We selected thiamylal and barbital as internal standards because they are structural analogs of thiopental and pentobarbital, respectively (Figure 1). Thiamylal is a surgical anesthetic,but it is not co-administered with thiopental. Although widely available as a laboratory reagent, barbital

showed no absorbance at 240 nm.

0These compoundsextract and co-chromatographwiththe barbital internal

standard. Clhis compound extracts and co-chromatographswith pentobarbital. -

Table 2. Concentrations of Thiopental and Pentobarbital in Serum of Patientsa Patient no.

Concn In serum, Intravenous

dose

Other drugs receIved

mg/L

1

TP, 6 mglkg bolusc

TP, 5.3; PB, 2.8

1,

2 3 4 5

TP, 5 mg/kg bolusc TP, 5 mg/kg bolusc TP, 6 mg/kg bolusc PB, 175 mg/h continuous

TP, 2.9; PB, 12.1 TP, 4.4; PB, 4.0 TP, 3.8; PB,