391

Clinica Chimica Acta, 93 (1979) 391-399 0 Elsevier/North-Holland Biomedical Press

CCA 10177

ROUTINE IDENTIFICATION AND DETERMINATION 11 BARBITURATES IN BIOLOGICAL SAMPLES

F. VINCENT

a**, C. FEUERSTEIN

b, M. GAVEND

OF

a and J. FAURE

c

a Laboratoire de Pharmacologic, C.H.U. Grenoble, b Exploration fonctionnelle, Section neurophysiologie, Pavilion de neurologie, C.H.U. Grenoble and c Service de mkdecine interne et toxicologic, C.H. U. Grenoble, Grenoble (France) (Received

November

21st, 1978)

Summary A very easy, reliable and specific gas-chromatographic method for identification and dosage of 11 barbiturates in plasma is presented. Methylene unit values determined on two types of column (one polar, one non-polar) allows very accurate identification with a minimum risk of error due to fluctuations in operating conditions. Sensitivity of the method is 0.5 gg/ml plasma with a flame ionization detector. Derivatization procedure is complete, without any degradation phenomenon as tested by mass spectrometry. Such a method can be easily used as routine procedure for toxicological or pharmacological applications.

Introduction Quantitative determination and qualitative identification of barbiturates in biological samples (blood, urine, gastric juice, etc.) are of particular interest in the fields of chemical toxicology and pharmacology. Techniques involving UV spectrophotometry [l] or thin-layer chromatography [ 2-31 are inadequate against gas-chromatographic (GC) methods. Therefore, we have developed and improved GC techniques for routine evaluation of usual barbiturates for qualitative and quantitative investigation. This technique has adequate specificity and sensitivity for all determinations involved in the different cases encountered in human pharmacology and toxicology. It is also very easy to carry out. * TO whom correspondence

should be addressed.

392

Materials and methods Reagents Trimethyl anilinium hydroxide 0.2 M in methanol = Meth Elute@ (Pierce Chemical, Rockford, IL, U.S.A.). n-Alkanes: Cl,+, Czz (Applied Science Laboratories, PA, U.S.A.). Distilled diethyl ether, n-hexane, carbon disulfide and the other reagents are of analytical grade (Merck, Darmstadt, F.R.G.). Phosphate buffer, pH 8 (9.5 ml of 11.8 g/l Na,HPO, . 12 Hz0 and 0.5 ml of 8.8 g/l KH,PO,). Carbonate buffer, pH 11 (8.4 g/l NaHCO, and 3.6 g/l NaOH). Stock solutions are made of sodium salts of each barbiturate (butobarbital, butalbital, amobarbital, secobarbital, pentobarbital, vinbarbital, cyclobarbital, hexobarbital, heptabarbital, phenobarbital and methohexital) in distilled water up to 100 mg/l. Methohexital sodium salt is used as internal standard and diluted up to 300 mg/l. These stock solutions can be kept refrigerated at +4”C during one month. Apparatus Gas chromatography is performed with a Girdel 3000 (Giravion-Dorand, Suresnes, France) or a P.E. 900 (Perkin Elmer, Norwalk, CT, U.S.A.). Both chromatographs have a dual column system, two flame ionisation detectors (FID) and a linear temperature programmer. Mass spectrometry (MS) is performed with a combined GLC-MS system Finnigan (Finnigan, Basel, Switzerland) 3200 and interactive data system 6100 using electron impact ionisation. Column preparation Two types of phase are used: OV-17 and OV-101 adsorbed on Chromosorb Q WHP 100-120 Mesh (Applied Sciences Laboratories, State College Pennsylvania, PA, U.S.A.) by a filtration technique. Glass columns, 3 m long, 3 mm id., 6 mm o.d., are filled with these dried phases and conditioned before use by slow heating (30°C to 290°C at a rate of 0.5”C/min) followed by a 24 h heating at 290°C. A gentle nitrogen flow is maintained through the columns during the whole procedure. Working conditions GLC Nitrogen (as carrier gas) flow rate: 30 ml/min Air flow rate: 350 ml/min Hydrogen flow rate: 22 ml/min Injector heater: 260°C Detector heater: 280°C Oven temperature: either 150°C 12 min then l”C/min up to 280°C for OV-17 columns or 150°C up to 280°C (at a rate of B”C/min) for OV-101 columns.

393

GLC-MS Helium (as carrier gas) flow rate: 25 ml/min Injector heater: 260°C Interface: 250°C Transfer line: 250°C Manifold: 60°C Electron energy: 50 eV Preamplifier: lo-’ A/V.

Procedure 60 ml stoppered centrifuge glass tubes containing 1 ml plasma, 100 ~1 of a 300 mg/l Methohexital solution, 2 ml phosphate buffer pH 8, and 20 ml ether/ hexane (1 : 1, v/v) are agitated for 10 min. After slow centrifugation the organic phase is transferred to another set of identical tubes containing 10 ml of distilled water. To the remaining initial aqueous phase is added 20 ml of ether/hexane mixture and after agitation and slow centrifugation, the etherlhexane phase is combined to the previous 20 ml organic phase. This new set of tubes is also agitated and centrifuged and the aqueous phase is discarded. The remaining organic phase is extracted twice with 5 ml carbonate buffer, pH 11. The combined 5-ml aqueous phases are acidified with 2.5 ml 1 M HCl and re-extracted twice with 20 ml ether/hexane as previously. The 40-ml organic phase is filtered on anhydrous sodium sulfate maintained on a 41 Whatman paper and evaporated to dryness under N2 in a Reacti-Vial. The residue is taken up in 50 ~1 carbon disulfide. After vortexing, 1 ~1 is pipetted in a lo-p1 Hamilton syringe containing 1 ~1 Meth Elute@ and 1 ~1 of a mixture of adequate alkanes: Cl6 and Czz for OV-17 3% columns, C,, and C!,,, TABLE

I

QUALITATIVE

IDENTIFICATION

OF ELEVEN

BARBITURATES

Mean end S.D. methylem unit values are determined on ten different chromatographic recovery of each barbiturate. Barbiturates

OV_l7.3% No.

Butalbital Butabarbitei Amoberbital Pentoberbital Secoberbital Vinberbitai Methohexital Hexoberbital Cycloberbital Phenobarbital Heptabarbital

1 2 3 4 5 6 7 a 9 10 11

CV_lOl. Methyiene unit values Mean

S.D.

17.08 17.25 17.55 17.88 18.29 18.44 19.36 20.87 21.15 21.30 22.02

0.04 0.05 0.05 0.02 0.06 0.05 0.06 0.04 0.04 0.04 0.05

No.

1 2 3 4 6 5 7 8 10 9 11

rime. Extraction

Extraction recovery

3% Methylene unit values Mean

S.D.

15.33 15.33 15.79 16.07 16.52 16.30 17.07 18.43 17.92 16.24 19.37

0.02 0.02 0.04 0.09 0.05 0.03 0.06 0.02 0.03 0.03 0.03

(%I

91 89 93 95 90 92 60 82 85 87 80

394

for OV-101 3% columns. chromatograph, allowing injection chamber.

The content of the syringe is then injected in the immediate methylation of barbiturates within the

Results Qualitative analysis Peak identification is based on methylene method of Dagliesh et al. [41 (Table I).

unit (M.U.) values according

to the

‘I i!

e!

L2

O"17 lo9

A %I OQ7

3%

I

L.

1

‘-

__*_-

n-

:IL

Fig. 1. A. Chromatogram obtained on OV-1’7 after extracting a pooled rates; each peak corresponds to 0.3 ~g of barbiturate. B. Chromatogram pooled barbiturate free plasma.

plasma containing 11 barbituobtained on OV-17 3% from

395

On OV-101, butobarbital and butalbital are not separated, but heptabarbital is well isolated from plasma impurities. On the other hand, butobarbital and butalbital are differentiated on OV-17 but the impurities and heptabarbital have the same M.U. (Figs. 1 and 2). Identification and specificity of each chromatograph peak in plasma have been confirmed by gas chromatography-mass spectrometry (GC-MS) (Fig. 3).

40

“TIME

”

IO MINUTES

Fig. 2. A. Chromatogram obtained on OV-101 3% after extracting urates; each peak represents 0.5 pg of barbiturate. B. Chromatogram barbiturate free plasma.

a pooled plasma containing 11 barbitobtained on OV-101 3% from pooled

DIMETHYL

PHENORARRITAL

CL,

Q

PM s-260

L

I

1,

200

B

DIMETHYL

I

“’

-J,.,..d 1

250

SEGOBARBXTAL

PM ~266

i

Ali. 50

100

Fig. 3. A. Phenobarbital

pJ&_&_+

150 MS. B. Secobarbital MS.

iLw

2i10

M+

I

Quan ti tatiue analysis A standard plasma pool charged with known increasing amounts of each barbiturate (5 to 50 mgfl) and a constant amount of methohexital (30 mgfl) is used for c~ibration curve determination. The respective ratio of the rectangular areas of each barbiturate to interna standard (obtained as follows: L - E/L’ . 1’, where L and L’ correspond to the respective distances from solvent front to each barbiturate peak (L) or to internal standard peak (L’), 1 and 1’ correspond to the respective heights of each barbiturate peak (E) or of the internal standard peak (I’)) is then plotted against respective barbiturate concentrations (Fig. 4).

397

Fig. 4. Calibration curve of phenobarbital, secobarbital and pentobarbital; the concentrations barbiturates extracted from plasma are plotted against respective peak areas.

of the three

The concentration range used here for curve calibration is that most used for quantitative determinations in the fields of toxicology and pharmacology. Discussion In our hands extraction recovery is the most favourable at pH 8 with disodium potassium phosphate buffer according to Truhaut method [5] and ether/hexane (1 : 1, v/v) as organic solvent. The triplet step extraction procedure used here eliminates other impurities such as salicylates, hydantoines [6], etc. from the organic acid phase subjected to GC. In fact, such artefact drugs are recovered only in the first organic acid phase as confirmed by UV spectrophotometry at each step. Garle and Petters [7] perform an only one-step extraction procedure; however, this requires further purification by adsorption on charcoal. Since GC of native barbiturates is generally inadequate because of peak tailing and of low sensitivity [6-S] it became more feasible to transform barbiturates into less polar derivatives. Methyl silylation seems to give unstable products [9]. Alkylation of the molecules appears to be the most satisfying derivatization procedure as confirmed by Greeley [lo]. This author is able to separate 14 barbiturates after butylation whereas Garle and Petters [ 71 use capillary column to separate butyl derivatives of 15, barbiturates. As hexyl and heptyl derivatives Giovanniello and Pecci [ll] separate 17 barbiturates but it is necessary to use both derivatives for such a separation. In our hands methylation remains the most adequate derivatization: (1) Dimethyl sulfate or diazomethane as methylating agents have numerous disadvantages in comparison to flash methylation which allows an instant synthesis without any danger [ 12,131. (2) Based on the inconsistent results obtained by Davis et al. [14] with tetramethyl ammonium hydroxide, trimethyl anilinium hydroxide (TMAH) in methanol has been chosen as methylation agent. Methylation is completely per-

formed within the injection port of the chromatograph [ 151. When methylation is performed by dissolving the dry extract of barbiturates with TMAH itself apart of the chromatograph, degradation of the derivatives, due to too long an incubation, can occur, particularly with phenobarbital. Degradation of methyl derivatives can also occur if the TMAH solution is too concentrated [ 14,16,17]. Therefore, a 0.2 M solution is used. Inadequate temperature of the injection port, too basic a pH and inadequate solvents have also a great influence on these degradation phenomena [l&19]. Characterization of methyl derivatives has been assessed by means of coupled GC-MS using our usual volumns OV-17 and OV-101. The mass spectra then confirm the quality of synthesis on each NH group of the different barbiturates (Fig. 3). Furthermore such mass spectra performed on different extracts from patients’ plasma confirm the absence of interfering substances within the different barbiturate peaks. In our procedure, two types of column are usually used, one polar, one nonpolar, which gives some advantages such as: (1) good separation of butalbital from butobarbital on OV-17 only (Fig. 1); (2) good separation of heptabarbital from natural plasma interfering impurities on OV-101 only (Fig. 2); (3) differences in the elution time of some barbiturates, allowing a very accurate identification of unknown barbiturate (Table I). M.U. values are of better accuracy for identification of unknown peaks than are retention times only because fluctuations in oven temperature, carrier gas flow rate, reproducibility of column packing, type of GC apparatus etc. do not affect M.U. values of a given phase. Each barbiturate is then identified with two M.U. values corresponding to the OV-101 and OV-17 ones (Table I). Methohexital has been chosen as internal standard for different reasons [ 121: (1) this compound is not used as therapeutic barbiturate in our country, which leads to assume this barbiturate has little chance of being found in the plasma of our patients, (2) as barbiturate, its chemical structure is very close to the barbiturates to be determined, particularly considering its extraction and derivatization characteristics, (3) it is eluted at approximately half the temperature programme. The sensitivity of the method (0.5 to 1 mg/l depending on the barbiturate tested) is adequate for the different cases encountered in pharmacology or toxicology. It allows determination from only 1 ml plasma. Internal standard is added to plasma before extraction allowing then a good precision and a very good reproducibility in recovery (Table I). Fidelity has been tested by calculating, on the different chromatograms the amount found for 10 identical levels of barbiturates added to 10 separate pools of plasma. There is less than 3% variation. Conclusion The procedure presented here represents an easy method with adequate sensitivity, precision, fidelity and specificity. The derivatization is rapid, complete and no degradation of derivatives has been yet observed (as confirmed mass spectrome try). In our hands, this method appears very accurate for toxicology and pharmacology, such as:

399

identification and specific dosage of barbiturates responsible for acute intoxication, therapeutic control of plasma levels of phenobarbital on treated epileptic patients, determination of plasma levels of pentobarbital during cranial trauma. Acknowledgements Pure barbiturates were generous gifts from: Lilly Research Centre (Methohexital), Specia Laboratories, Paris (Butobarbital and phenobarbital), Sandoz Laboratories, Rueil-Malmaison (Butalbital), Roussel Laboratories UCLAF (Amobarbital and secobarbital), Abbott Laboratories, Saint Remy sur Avre (Pentobarbital), Bottu Laboratories, Paris (Vinbarbital), Bayer Laboratories, Sens (Cyclobarbital and hexobarbital), Geigy Laboratories, Rueil-Malmaison (Heptabarbital). We are particularly grateful to Bernard Karl Huber from Finnigan Inc. (Basel) for giving us free access to the GC-MS system used in this study. We are very indebted to Anne Marie Renard, for her technical assistance and competence and to Brigitte Grivaud for typing. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1’1 18 19

Pehr. F. (1975) Clin. Chem. 21.1609-1611 Roerig, D.L., Lewand. D.L., Mueller, M.A. and Wang, R.I.H. (1975) Clin. Chem. 21. 672-675 Polcaro, C. (1976) J. Chromatogr. 125.431-434 Dagliesh, C.E.. Horning, E.C. and Horning, M.G. (1966) Biochem. J. 792-810 Truhaut, R., Boiteau. H. and Puisieux, F. (1964) Arch. Mal. Prof. 25,315.-322 Berry. D.J. (1973) J. Chromatogr. 86. 89-105 Garle, M. and Petters, I. (1977) J. Chromatogr. 140,165-169 Cooper, R.G.. Greaves, MS. and Owen, G. (1972) Clin. Chem. 18,1343-1349 Chang. T. and Glazko. A.J. (1970) J. Lab. Clin. Med. 75.145 Greeley. R.H. (1974) Clin. Chem. 20,192-194 Giovanniello, T.J. and Pecci, J. (1977) Clin. Chem. 23, 2154-2155 Cabic, A.P. and Allain, P. (1973) Therapic 28.951-967 Brachet-Liermain, A., Ferrus. L., Clerc, Y. and Michon, D. (1972) Ann. Biol. Clin. 30, 243-252 Davis, H.L.. Falk. K.J. and Bailey, D.G. (1975) J. Chromatogr. 107, 61-66 Skinner, R.F., Gallaher. E.G. and Predmore, D.B. (1973) Ann. Chem. 45, 574-576 Osiewicz, R.. Aggarwal, V.. Young, R.H. and Sunshine. I. (1974) J. Chromatogr. 88, 157-164 Gallery, P.S. and Leslie, J. (1976) Clin. Chem. 22,926 Gallery, P.S. and Leslie, J. (1977) J. Pharm. Sci. 66. 578-580 Kelly, R.. Valentour. J. and Sunshine, I. (1977) J. Chromatogr. 138. 413-422