Single and Multiple Ion Recording Techniques for the Analysis of Diphenylhydantoin and its Major Metabolite in Plasma? J. D. BatyS and P. R. Robinson8 Department of Medicine, University of Liverpool, Liverpool L69 3BX, England

A method has been developed for single ion monitoring of diphenylhydantoin and its major metabolite 5-(p-hydroxyphenyl)-5-phenylhy~ntoin in plasma. A plasma extract is reacted with N,Obis(trimethylsily1)acetamideand single ion recording is carried out using a gas chromatographmass spectrometer system. The mass value selected, m/e 254,is common to the TMS ethers of diphenylhydantoinand its principal metabolite 5-(p-hydroxyphenyl)-5-phenylhydantoin.The results indicate that one cause of an adverse reaction to diphenylhydantoin could be a reduced ab&ty to hydroxylatethe drug. Quantitativemethods for the analysisof the drug and its major metabolite have also been developed. Diphenylhydantoin and 5-(p-hydroxyphenyl)-5phenylhydantoin can be analysed in plasma after addition of deuterium labelled internal standards and conversion to volatile derivatives for mass fragmentographic analysis. Diphenylhydantoin and its internal standard are analysed as the N,N-dimethyl derivative, and the hydroxylated metabolite and its internal standard are converted to a pertrimethylsilylcompound by reaction with N,O-bis-(trimethylsily1)acetamide.

INTRODUCTION Diphenylhydantoin (DPH, I), has been used for many years in the treatment of epilepsy. A serious problem in the use of this drug is the narrow range of plasma concentrations over which the drug is effective (1020 pg ml-I).' At lower concentrations it may be ineffective while at high plasma concentrations (>25 pg mi-') adverse side-effects may occur. The situation is further complicated by the fact that large inter-individual variations in DPH plasma levels have been reported in subjects given the same dose of the drug.2 Genetic factors influence the elimination and metabolism of the and in any individual, genetic and environmental factors together will determine the overall pharmacokinetics and metabolic profile of DPH.4

H

1

Many methods are available for the analysis of DPH. The older colorimetric have been replaced by more sensitive and specific methods such as t Abbreviations: DPH = diphenylhydantoin; HPPH = 5-(phydroxyphenyl)-5-phenylhydantoin; Th4AH = trimethylanilinium hydroxide; BSA = N, 0-bis-(trimethylsi1yI)acetamide. j:Present address: Department of Biochemical Medicine, Ninewells Hospital, Dundee, Scotland. 0 Present address: StirlingWinthrop Ltd, Newcastle-upon-Tyne, England.

0Heyden & Son Ltd, 1977 36 BIOMEDICAL MASS SPECTROMETRY, VOL. 4,

NO. 1, 1977

fl~orimetry,~ high pressure liquid chromatography,8 enzyme immunoassayg and gas liquid chromatography."," Little attention has been paid to measuring 5(p-hydroxyphenyl)-5-phenylhydantoin (HPPH, 2) in

H 2

plasma. A method based on h.p.1.c. and scintillation counting appears to be very time consurning,l2whilst a spectrophotometric method did not completely separate DPH from HPPH. l 3 Gas-liquid chromatographic methods for the analysis of DPH and HPPH have been Multiple ion recording has been used to study the low concentrations of DPH found in breast milk and in neonatal Recently, such a procedure has been described for the analysis of DPH and HPPH in plasma using an extractive alkylation technique." We have developed single and multiple ion recording methods to study metabolic profiles of subjects who have reached a 'steady state' plasma level of DPH or whose plasma level is falling from a steady state level. We have used a single ion recording technique to obtain a rapid and semiquantitative trace of an ion present in the mass spectrum of the trimethylsilyl ether of DPH and its major metabolite HPPH. For quantitative studies on the concentration of DPH and HPPH in plasma we have developed a multiple ion recording assay.

ANALYSIS OF DIPHENYLHYDANTOIN

EXPERIMENTAL

Mass spectral analysis indicated that the pentadeuteriobenzophenone had a molecular ion at m/e

Apparatus Gas chromatography was carried out using a Varian Aerograph Series 1800 instrument with a hydro en flame ionization detector. Glass columns (5 ft x gF.in) packed with 3% OV-17 on Gas Chrom. Q were conditioned overnight at 270 "C with a nitrogen carrier gas flow. Gas chromatography mass spectroscopy (g.c.m.s.) analyses were performed using a Pye 104 Gas Chromatograph coupled to an MS- 12 mass spectrometer (Associated Electrical Industries Ltd, Manchester, England). A silicone rubber membrane molecular separator was used as an interface. The mass spectrometer was operated at 8 kV, with a trap current of 300 p A and an ionizing voltage of 25eV. The analogue data were analysed by a PDP 8/I computer (Digital Equipment Corporation, Maynard, Massachusetts, U.S.A.) with a 4 K memory and a 256 K disk store. A commercial data system (A.E.I. DS-30) was modified for this work. Multiple ion recording was carried out using a multiple-peak monitoring device manufactured by A.E.I. This allowed up to six separate mass values to be monitored by ra id switchingof the accelerating voltage (Sweeley et al.).'30 The signal from each mass was displayed on a Rikadenki multichannel chart recorder. Reference compounds and materials All solvents were of 'Analar' standard and distilled twice before use. Hexadeuteriobenzene (E2H6]benzene) 99% was obtained from Prochem-BOC. 0.5 M trimethylanilinium hydroxide (TMAH) was synthesized according to the method of Brochmann-Hanssen et al." Trimethylsilyl ethers were prepared using N,O-bis(trimethylsily1)acetamide (BSA) supplied by B.D.H. Chemicals, Poole, Dorset. Diphenylhydantoin and 5-(p-hydroxyphenyl)-5phenylhydantoin Gold Label) obtained from Aldrich Chemical Co., Gillingham, Dorset were both recrystallized from ethanol + water before use. Benzoyl chloride, anisoyl chloride, acetamide and pmethoxybenzophenone were obtained from Aldrich. Benzophenone and p-hydroxybenzophenone were obtained from Koch-Light Labs., Ltd, Colnbrook, Buckinghamshire. P-Glucuronidase (Type H2) was supplied by the Sigma Chemical Company, Kingstonupon-Thames, Surrey. The activity of 1ml of this preparation was 134 000 Fishman units.

H

3

187 compared with m/e 182 for the protio form. Major fragments corresponding to [C6'H5]+ and [C$H5CO]+ were observed. Thus five deuterium atoms had been incorporated into one phenyl ring of benzo henone. Pentadeuterio DPH was synthesized from [ H5]benzophenone using a modification of the Bucherer-Berge synthesis." This involved heating 2 g of the labelled benzophenone with 1.3 g of potassium cyanide and 3.4 g of ammonium carbonate in ammoniacal ethanol at 60 "C for 5 days. The product was extracted into alkali and filtered. Acidification of the filtrate gave a white precipitate of pentadeuterio DPH which was separated by filtration. The pentadeuterio DPH had a melting point of 292294 "C. Thin-layer chromatography of the product was performed using a chloroform +n-butanol + ammonia (140 :80: 10) solvent system. Single spots were seen for both pentadeuterio DPH and a DPH standard when visualized under U.V. light. The Rfvalues of both spots were identical (0.37). Conversion of pentadeuterio DPH to the 1,3dimethyl derivative was carried out by flash-heater methylation in the g.c.m.s. using trimethylanilinium hydroxide. The mass spectrum of the dimethyl derivative is shown in Fig. 1. Reaction of the pentadeuterio DPH with diazomethane produced only the monomethyl derivative. 5-(p-hydroxyphenyi)-5-pentadeuteriophenylhydantoin (4). Hexadeuteriobenzene (10 ml) was refluxed for 5 h with 2.3 ml of anisoyl chloride in the presence of 2.1 g of anhydrous aluminium chloride, to give pentadeuteriop-methoxybenzophenone. 2.5 g of this compound was refluxed at 120 "C for 4 h with 3 g of aluminium chloride to convert it to the phenol. The pentadeuterio phydroxybenzophenone was isolated by extraction into alkali, acidification and re-extraction into dichloromethane. Comparison of the mass spectrum of labelled and non-labelled p-hydroxybenzophenone showed that 5 deuterium atoms were incorporated into the nonsubstituted phenyl ring. However the [M+ l]"/[M]+'

12

Synthesis of deuteriated internal standards D

Two deuterio labelled compounds were synthesized for the present study. Pentadeuterio-5,5 diphenylhydantoin (3). A mixture of 10 ml of hexadeuteriobenzene and 2 ml of benzoyl chloride was refluxed for 5 h with 2 g of anhydrous aluminium chloride. The deuteriobenzophenone produced (70% yield) had identical g.c. characteristics to protiobenzophenone. The meiting point was 46-48 T.

H 4

BIOMEDICAL MASS SPECTROMETRY, VOL. 4, NO. 1, 1977 37

J. D. BATY AND P. R. ROBINSON

100,

50

I00

I50 m/e

200

250

Figure 1. The mass spectrum of N,N-dimethylpentadeuteriodiphenylhydantoin.

ratio in this compound was markedly increased compared with the protio form of p-hydroxybenzophenone. Anisic acid, a by-product of the Friedel-Craft reaction, was isolated and also showed an increased [M+ 1]+:[MI' ratio. Thus, some isotopic enrichment of the anisoyl chloride had occurred by deuterium exchange with hexadeuteriobenzene during the FriedelCraft reaction. Because this effect was not observed when benzoyl chloride was substituted for anisoyl chloride, the exchange must have been promoted by the para substituent. The deuteriated hydantoin was synthesized by heating 2 g of deuteriated hydroxybenzophenone with 0.7 g of potassium cyanide, 4.7 g of ammonium carbonate and 25 g of acetamide in a sealed tube at 110 "C for 50 h.23 Alkaline extraction of the product gave a light brown solid. The 5-(p-hydroxyphenyl)-5-pentadeuteriophenylhydantoin was recrystallized twice from aqueous ethanol plus activated charcoal. The melting point was 315-318 "C (decomposition). The overall yield of this reaction was 35%, based on anisoyl chloride. Attempts to methylate the hydantoin gave variable results. TMAH produced a permethylated derivative only when used in 100-fold molar excess and with a g.c. injection port temperature approaching 300 "C. These conditions caused rapid deterioration of the column stationary phase near the injection heater. Failure to completely methylate the compound in the g.c.m.s. system produced a marked 'memory' effect, which gave rise

to high blank values. Reaction with diazomethane produced a mixture of the N-methyl and N,N'-dimethyl derivatives. Reaction of the HPPH and its deuterium analogue with BSA proceeded without any of the problems described above. Complete conversion into the persilyl compound was achieved by reaction with BSA at 80 "C for 30 min. The mass spectra of the TMS ethers of compounds 2 and 4 are shown in Figs. 2 and 3. The mass spectrum of compound 4 confirms the incorporation of five deuterium atoms into the monosubstituted benzene ring and indicates that a small amount of deuterium incorporation has occurred into the phenolic ring. It was shown that deuterium atoms were stable under the analysis conditions. Extraction and analysis Single ion recording of plasma profiles. Plasma (1ml), 0.2 ml of phosphate buffer (3 M NaH2P04) and 0.1 ml of ,G-glucuronidase were incubated at 37 "C for 18 h. Sampleswere then extracted by vortexing for 2 min with 10 ml of 1,2-dichloroethane. After centrifugation at 2500 rev min-' X 10 min the organic layer was separated and dried with anhydrous sodium sulphate. It was evaporated to dryness at room temperature under a stream of nitrogen. BSA (50 PI) was added and the resulting solution was heated at 80 "C for 30 min. This procedure converted

TMSi 0

OTMSi

m/e

Figure 2. The mass spectrum of the persilyl TMS ether of 5-(p-hydroxyphenyI)d-phenylhydantoin. 38 BIOMEDICAL MASS SPECTROMETRY, VOL. 4, NO. 1, 1977

ANALYSIS OF DIPHENYLHYDANTOIN 100-

80-

8 D 0

5 60-

D 0

.- 40G)

-

2

20-

Figure 3. The mass spectrum of the persilyl T M S ether of 5-(p-hydroxyphenyl)-5-pentadeuteriophenylhydantoin.

the hydantoins into their trimethylsilylethers and 1-2 pl of the solution was injected into the g.c.m.s. For single ion recording at m/e 254 the g.1.c. column was held at 200 "C for 3 min, then programmed at 6 "C min-' up to 250 "C. The helium flow rate was 40 ml min-l. Quantification of DPH and HPPH by mass fragmentography. After addition of known amounts of pentadeuterio DPH and HPPH (40 pg and 5pg) as internal standards, 1 m l of plasma was treated with p-glucuronidase as described above. 10 ml of cyclohexane was added and the mixture vortexed for 2 min. The cyclohexane layer was separated and discarded. This procedure removes endogenous plasma lipids which interfere with the analysis. The plasma sample was extracted with 1,2-dichloroethane and then the extract divided into two parts. One-fifth was used for the DPH assay and the remainder for the HPPH assay. DPH analysis Trimethylanilinium hydroxide (100 p1, in 50% methanol) was added to the 1,2-dichloroethane layer and the mixture vortexed for 1min. This extracted the acidic hydantoins into the strongly basic aqueous layer. On standing, this upper layer separated out and 1-2 p l was injected into the g.c.m.s. The injection port temperature of 280 "C ensured complete methylation of DPH and the pentadeuterio-DPH internal standard. A column temperature of 230°C and a flow rate of 40 ml min-' gave a retention time of 3 min for dimethyl DPH and pentadeuterio DPH. The multi-peak monitor was tuned so that the accelerating voltage was set to monitor masses 280 and 285 (the molecular ions of dimethyl DPH and pentadeuterio DPH). The signals produced from these ions were displayed on separate channels of the recorder. Methylation produced sharp peaks with no tailing. HPPH analysis

The organic extract was evaporated to dryness under a stream of nitrogen, the residue heated for 30min at 80 "C with 50 p1 BSA and 1-2 pl of the resulting solution was injected into the g.c.m.s. The column temperature was adjusted to 220 "C to give a retention time of 3 min for persilyi HPPH and pentadeuterio HPPH. The

molecular ions of the persilyl derivatives (m/e 484 and m/e 489) were monitored. Production of g.c.m.s. calibration curves A fixed amount of the appropriate deuteriated internal standard was added to blank plasma samples to which known amounts of DPH and HPPH had been added. For DPH the lower limit was 1000 ng ml-' and for HPPH 100 ng rn1-I. The 'normal' range of DPH levels in plasma is 5-20 pg ml-'. After hydrolysis and extraction as described above a calibration curve was produced relating peak height of the labelled and non-labelled material to their concentration in the plasma samples. A regression line of peak height on concentration was calculated using a least squares fit. The 95% confidence intervals (standard deviation of the concentration data times Students t (a = 14) were 360 ng ml-' for DPH, and 54 ng ml-' for HPPH.

RESULTS AND DISCUSSION ~~

M/e 254 is the base peak in the spectrum of persilyl HPPH, and its probable origin is shown in Fig. 2. In TMS-DPH it is present as an isotope peak of the ion at m/e 253 (Fig. 4). As a result, even though the concentration of HPPH in plasma is normally much less than DPH no alteration in instrumental parameters is required to obtain a significant HPPH peak height. Figure 5 shows a single ion recording of m/e from a volunteer ;t a steady state DPH plasma level of 10.5 pgml- , and the profile obtained from this subject's plasma prior to drug ingestion. The peaks due to the TMS ethers of both DPH and HPPH are clearly visible. The compounds giving rise to the other signals at m/e 254 have not been identified at this time although the later peak is almost certainly TMS cholesterol. One volunteer was taken off the drug after one week because of adverse side reactions including nystagmus and lethargy. The profile of this subject is shown in Fig. 6. It shows a pronounced reduction in the amount of HPPH produced from DPH. Subsequent measurement revealed a DPH plasma level of 29 pg m1-l. This is within the range where intoxication is known to occur. The voiunceer reported taking no other medication prior BIOMEDICAL MASS SPECTROMETRY, VOL. 4, NO. 1, 1977 39

J. D. BATY AND P. R. ROBINSON 1000

TMSi

80 a,

N-TMSi

0 c

P

60-

n a,

.-

;j -

40-

d

20-

, l ' 1 # 1 ,

[ I

I I , I

I

I / ,I ',pII,,

I

, /

I

to or during the test. Therefore, the presence of high levels of DPH in this subject could possibly be due to a reduced ability to para -hydroxylate the drug. A number of cases of limited capacity to form HPPH have been described and are presumed to be of genetic rigi in.^^,^^ These semiquantitative profiles illustrate the usefulness of the TMS ethers for the derivatization of DPH and its major metabolite. We were unable to detect any other qualitative differences between drug and blank plasma profiles. The blank plasma profiles from all subjects were very similar. No significant interference was found at the retention times corresponding to the TMS ethers of both DPH and HPPH. The mass spectrum of the TMS ether of 5-(4-hydroxy3-methoxyphenyl)-5-phenylhydantoinhas an ion at m/e 254. This compound has been isolated as a metabolite of DPH in rat urine.26Its presence could not be detected on our profiles.

,

I

I

I,

, ,/ ,

I,

I

, , Jj / , , , , I , I , , , / , I , , , I , I

i., ,

11,

Quantitative studies on plasma samples from both steady state and steady state decay experiments indicate that the steady state DPH values were approximately 515 pg m1-I. The HPPH levels were in the range 0.53 pgml-I. Steady state decay studies showed drug and metabolite to have nearly equal plasma concentrations after 60 h. In these studies subjects received a DPH sodium dose of 5 mg kg-I three times daily.27The sensitivity of the HPPH assay is reduced because of the low intensity of the ion at m/e 484. It is not possible to use the ion at m/e 254 (and mle 259 in the deuterated analogue) because of a substantial peak at m/e 254 in this latter compound. Since the internal standard would be added to plasma at 10-100 times the concentration of HPPH it would produce extremely high blank values. The intense molecular ion of the permethylated HPPH and its internal standard would provide a more sensitive assay. However, because of the difficulties

b)

d m N

d

m

?

*F

N

F

J

1

1

1

l

1

1

1

1

1

1

1

1

1

1

1

Time (min)

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

Time (min)

Figure 5. Single ion recording of m/e254from (a) pre-drug and (b) post-drug plasma of a subject receiving diphenylhydantoin sodium. The plasma extract was reacted with BSA prior to anaiysis.

40 BIOMEDICAL MASS SPECTROMETRY, VOL. 4, NO. 1, 1977

ANALYSIS OF DIPHENYL,HYDANTOIN

described in using trimethylanilinium hydroxide we were unable to produce a satisfactory standard line with these derivatives. Other workers have measured HPPH in plasma by selected ion monitoring of the permethylated derivative.” However, they found the conditions required for extractive alkylation to be critical. Under these conditions degradation of the catechol metabolite of DPH occurred. Our failure to achieve reproducible results with permethylated derivatives of HPPH could be due to the design of the interface and separator in our instrument, but we feel it is largely caused by extensive column deterioration, giving rise to extraneous peaks. The range of DPH concentration found in these volunteers was 5-15 pg m1-I. Obviously multiple ion recording is not essential to assay this drug, although the sensitivityof the method would permit measurements to be made on 100 p1 samples of blood. The coefficient of variation for the analysis of 100 pl of plasma containing 5 pg ml-I of DPH was 4.9% (n = 6).

U

U

9

F

I

v‘

DPH v

I

l

l

I

HDPH I

I

I

1

I

I

I

I

I

l

Acknowledgements

Time (min)

Figure 6. Single ion recording of mle254from post-drug plasma of a subject who developed adverse side reactions after diphenylhydantoin sodium. The plasma extractwas reactedwith BSA prior to analysis.

We thank Dr J. L. Cunninghamfor provision of blood samples and the organizationof drug tests. We thank Miss M. F. Bullen, S.R.N. for the collection of blood samples. Financial support is acknowledged from the Nuffield Foundation and the M.R.C. (studentship to P.R.R.).

REFERENCES 1. H. Kuttand F. McDowell, J. Am. Med. Assoc. 203,167 (1968). 2. P. T. Lascelles, R. S. Kocen and E. G. Reynolds, J. Neurol. Neurosurg. Psychiatry 33,501 (1970). 3. P. Buch Andreasen, A. Freland, L. Skovsted, S. Aas Andersen and M. Hauge, Acta Med. Scand. 193,561 (1973). 4. H. Kutt, Ann. N.Y. Acad. Sci. 179,705 (1971). 5. W. A. Dill, A. Kazendo, L. M. Wolf and A. J. Glazko, J. Pharmacol. Exp. Ther. 118,270 (1956). 6. J. Wallace, J. Biggs and E. V. Dahl.Ana1. Chem.37,410 (1965). 7. W. A. Dill and A. J. Glazko, Clin. Chem. 18,675 (1972). 8. S. H. Atwell, V. A. Green and W. G. Haney, J. Pharm. Sci. 64, 806 (1975). 9. R. F. Schneider, Clin. Chem. 20,869 (1974). 10. A. Berlin, S. Agurell, 0. BorgB, L. Lund and F. Sjoqvist, Scand. J. Clin. Lab. Invest. 29,281 (1972). 11. D. Sampson, I. Harasymiv and W. J. Hensley, Clin. Chem. 17, 382 (1971). 12. K. S. Albert, M. R. Hallmark, E. Sakmar, D. J. Weidlerand J. G. Wagner, Res. Common. Chem. Pathol. Pharmacol. 9, 463 (1974). 13. W. A. Dill, J. Baukema, T. Chang and A. J. Glazko, Proc. SOC. Exp. Biol. Med. 137, 674 (1971). 14. T. Chang and A. J. Glazko, J. Lab. Clin. Med. 75, 145 (1970). 15. A. Estas and P. A. Dumont, J. Chromatogr.82,307 (t973).

16. B. Karlen, M. Garle, A. Rane, M. Gutova and B. Lindborg,Eur. J. Clin. Pharmacol. 8,359 (1975). 17. M. G. Horning, J. Nowlin, K. Lertratanangkoon, R. N. Stillwell, W. G. Stillwell and R. M. Hill, Clin. Chem., 19, 845 (1973). 18. A. Rane, M. Garle, 0. BorgB and F. Sjoqvist, Clin. Pharmacol. Therap. 15,39 (1974). 19. C. Hoppel, M. Garle and M. Elander, J. Chromatogr. 116,53 (1976). 20. C. C. Sweeley, W. H. Elliot, I. Fries and R. Ryhage,Anal. Chem. 38,1549 (1968). 21. E. Brochmann-Hanssen and T. Oke-Olawuji, J. Pharm. Sci. 58,370 (1969). 22. H. R. Henze, US. Pat. 2,409,754 (1946). 23. H. R. Henze and A. F. Isbell, J. Am. Chem. SOC.76,4152 (1954). 24. H. Kutt, M. Wolk, R. Scherman and F. McDowell, Neurology 14,542 (1964). 25. N. Gerber, R. Lynn and J. Oates, Ann. Intern. Med. 77, 765 (1972). 26. T. Chang, R. A. Okerholm and A. J. Glazko, Res. Commun. Chem. Pathol. Pharmacol. 4,13 (1972). 27. J. L. Cunningham, personal communication.

Received 26 February 1976 @ Heyden & Son Ltd, 1977

BIOMEDICAL MASS SPECTROMETRY, VOL. 4, NO. 1, 1977 41