Determination of chlordiazepoxide hydrochloride (librium) and its

Wade J. Adams , Ursula M. Rykert , Paul A. Bombardt. Analytical Letters 1980 ... W. Franklin Smyth , T. Jenkins , J. Siekiera , A. Baydar. Analytica C...
0 downloads 0 Views 825KB Size
Determination of Chlordiazepoxide Hydrochlor ide (Librium) and Its.Major Metabolites in Plasma by Differential Pulse Polarography M. R. Hackman, M. A. Brooks and J. A. F. d e Silva Department of Biochemistry and Drug Metabolism, Hoffmann-La Roche lnc., Nutley, N.J. 077 70

1.S. M a Department of Chemistry, City University of New York, Brooklyn, N . Y . 11270

A sensitive and specific differential pulse polarographic assay was developed for chtordiazepoxide and its two major metabolites in plasma. The assay involves the selective extraction of the compounds into diethyl ether from plasma buffered to pH 9.0, followed by thin-layer chromatographic separation of the three compounds. The compounds are eluted from the thin-layer chromatogram with methanol, the residue of which is dissolved in 0.1N H2S04 and analyzed by differential pulse polarography. The overall recovery of chlordiazepoxide, N-desmethylchlordiazepoxide, and demoxepam are 61.9% f 3.1 (std de;), 62.0% f 3.7 (std dev) and 62.7% f 4.3 (std dev), respectively. The sensitivity limit for chlordiazepoxide and its desmethyl metabolite is approximately 0.05 pg/ml of plasma and that for demoxepam is 0.10 pg/ml of plasma using a 2-ml specimen per assay. The assay was applied to the determination of chlordiazepoxide and its metabolites following a single 30-mg dose by i.v. and oral routes and chronic administration of Librium.

Chlordiazepoxide hydrochloride, 7-chloro-2-methylamino-5-phenyl-3H-1,4-benzodiazepine-4-oxidehydrochloride, marketed as the active ingredient in Librium is extensively used as a psychotherapeutic drug. Studies (1-3) have shown that the compound is metabolized in man to form two majpr metibolites, the N-desmethylchlordiazepoxide and the lactam (demoxepam), Figure 1. The determination of the parent compound in plasma by colorimetry employing the Bratton-Marshall chromophore formed as a result of hydrolytic cleavage of the compound to the amine followed by diazotization and coupling with N-(1-naphthyl)-ethylenediamine has been reported (4-8). A toxicological spectrophotometric procedure to measure plasma levels of the compound by its ultraviolet absorbance has also been reported recently ( 9 ) . Spectrofluorometric methods ( I , 3 ) employing selective extraction, hydrolysis, and photochemical rearrangement to yield fluorescent derivatives have been used to determine chlordiazepoxide and its two major metabolites. These methods have enabled all three compounds to be determined with a sensitivity of 0.1-0.2 kg/ml using a 2-ml specimen of plasma ( 3 ) . B. A. Koechlin and L. D'Arconte, Anal. E!ochem.. 5, 195 (1963) B. A. Koechlin, M . A. Schwartz, G. Krol. and W. Oberhansli, J . Pharmacol. Exp. Ther.. 148, 399 (1965) M. A. Schwartz and E. Postma, J. Pharm. Sci.. 55, 1358 (1966) L. 0. Randall, Dis. Nerv. Syst.. 22 (Supplement 7 ) , 1 (1961). J. Baumler and S. Hippstein, Helv. Chim. Acta.. 44, 2208 (1961) G. Kamm and R. Baier, Arzneim.-Forsch.. 19, 213 (1969) C. S. Frings and P. S. Cohen, Amer. d. Clin. Pathol.. 56, 216 (1971). K. Besserer, S. Henzler. E. Kohler, and H . J. Mallach, Arzneim.Forsch.. 21, 2003 (1971) P. Jatlow, clin. Chem., 18, 516 (1972)

CHLORDIAZEPOXIDE

DES 54 E T r Y L CHLORDIAZEPOXIDE

I

Tr

Figure 1. Metabolic Schwartz et a / (3)

DEMOXEPAfE

m

pathways of I in man as reported by

An electron-capture gas-liquid chromatographic (ECGLC) procedure (10) for the determination of diazepam and its N-desmethyl metabolite as their benzophenones can also be used for the determination of chlordiazepoxide. It would, however, require a thin-layer chromatographic separation of the parent compound from its metabolites followed by conversion to the benzophenone prior to EC-GLC analysis. Recently, another EC-GLC method was reported ( I I ) , for the determination of intact chlordiazepoxide a t levels as low as 0.5 ng/ml plasma. This method employs selective extraction of chlordiazepoxide into heptane containing 1.5% isoamyl alcohol from alkalinized plasma and can measure chlordiazepoxide levels following ingestion of a single 5-mg dose. However, it does not measure the two major metabolites. The polarographic behavior of chlordiazepoxide was first reported by Oelschlager (12), who reported three reduction waves for the compound in Britton-Robinson buffer ( p H 7.8). Studies dealing with the effect of pH, substituent groups, and concentration on the diffusion current produced by the polarographic reduction of chlordiazepoxide and other benzodiazepines, were reported by Senkowski et al. (13). Further studies on the mechanism of the polarographic reduction of chlordiazepoxide have also been reported (14-17). Direct current polarographic analysis has been used to analyze chlordiazepoxide in pharmaceutical formulations (18), for toxicological analysis (19), and to determine levels in horse serum (16). None of these assays for the com(10) J. A. F. de Silva, M . A. Schwartz, V . Stefanovic. J. Kaplan, and L. D'Arconte, Anal. Chem., 36, 2099 (1964) (11) I. A. Zingales,J. Chromatogr.. 61, 237 (1971). (12) H. Oelschlager, Arch. Pharm. Eerl.. 296, 396 (1963). (13) 6 .2 . Senkowski, M . S. Levin. J. R . Urbigkit, and E. G. Wollish, Anal. Chem., 36, 1991 (1964). (14) H. Oelschlager and H. Hoffmann, Arch. Pharm. lweinheimi. 300, 817 (1967). (15) ti. Oelschlager, J. Volke, H. Hoffmann, and E. Kurek, Arch. Pharm. i Weinheimi. 300, 250 (1967). (16) E. Jacobsen and T. V. Jacobsen, Anal. Chim. Acta. 55,293 (1971). (17) J. M . Clifford and W. F. Smyth. Fresenius' 2. Anal. Chem.. 264, 149 (1973). (18) G. Caille, J. Braun, and J. A. Mockle, Can. J . Pharm. Sci., 5, 78 (1970) (19) G. Cimbura and R . C. Gupta, J. Forensic Sci.. IO, 282 (1965). ANALYTICAL

CHEMISTRY, VOL. 46, NO. 8 , J U L Y 1974

1075

Polarographic Analyzer equipped with a PAR Model 112A Drop Timer and electrode assembly and a Houston Omnigraphic X-Yrecorder (Model 2200-3-3) was used. Electrode Assembly. A three-electrode semi-micro cell with a working volume of 2.0 ml containing a dropping mercury electrode (DME; capillary tube of inside diameter = 0.05-0.08 mm from E. H. Sargent No. S-29419) as the indicator electrode, a saturated calomel electrode (SCE; fiber junction calomel from Beckman, No. 39178) as the reference electrode and a platinum wire as the auxiliary electrode was used, (see Figure 2). Instrumental Parameters. Polarography was performed in the differential pulse polarographic mode with a -50mV pulse being applied using a 1.0-second drop. The flow rate was 2.74 mg/sec (m2I3 = 1.9581). Scans were performed from -0.450 to -0.800 V us. SCE for all three compounds a t a scan rate of 2 mV/sec with a full scan range of 1.5 V. The usual current sensitivity required was W 0.5, 1,or 2 pA full scale deflection. Procedures. Determination of I and Its Metabolites in Figure 2. Polarographic semimicro cell containing the three opPlasma. Into a 15-ml conical glass-stoppered centrifuge erational electrodes for analysis tube, add 2 ml of plasma sample, 1 ml of distilled water, DME = dropping mercury electrode; Pt. aux. = platinum auxiliary elec2 ml of 1M phosphate buffer (pH 9.0), and 7.0 ml of ditrode; SCE = saturated calomel electrode ethyl ether. Stopper the tube with a Teflon stopper, shake pound in biological fluids (16, 19) is capable of determinon a reciprocating shaker for 10 minutes a t a moderate ing the drug following a therapeutic dose, unless excesspeed and then centrifuge for 5 minutes at 2000 rpm. sively large volumes of sample are used. Along with the samples, process a 2-ml specimen of conThe differential pulse polarographic method presented trol plasma, and separate 2-ml specimen of control plaspossesses high sensitivity and is capable of measuring levma; one containing 0.5 pg of I, 11, and I11 as internal stanels of the parent compound and its metabolites of 0.05dards, one containing 1.0 pg of the three compounds, and 0.10 pg/ml plasma using a 2-ml sample. It is also specific one containing 2.0 pg of the three compounds. These are by virtue of the thin-layer chromatographic separation prepared by adding 0.05, 0.1, and 0.2 ml of the working and by the fact that each compound may be identified by standards into 15-ml conical centrifuge tubes, evaporating its distinctive peak potential value ( E p ) .Since the polarothe methanolic solution to dryness in a 65 "C water bath graphic assay measures these compounds as their intact under a stream of nitrogen, and then adding 2 ml of conentities, it has a distinct advantage over the spectrofluotrol specimen to the residues. Carefully remove the dirometric assay (3) in that the photochemical conversion of ethyl ether layer using a 10-ml serological pipet, and the compounds to fluorescent derivatives is eliminated. transfer this layer into another 15-ml conical glass stoppered centrifuge tube. Evaporate the extract to dryness in EXPERIMENTAL a 45 "C water bath (Evapomix, Buchler Instruments, Fort Assay in Plasma. Reagents. All reagents were of analytLee, N.J.) while reextracting the sample with a second 7.0 ical reagent grade (ACS) purity and were used without ml of diethyl ether as described above. Combine the secfurther purification. The buffer used in the initial extracond extract with the residue of the first extract in the tion was 1M phosphate buffer (pH 9.0). The Teagent was 15-ml tube and evaporate to dryness at 45 "C. made by mixing 900 ml of 1M K2HP04 and 100 ml 1M Dissolve the residue from the combined extract in 100 KH2P04. The pH should be checked and adjusted to pH p l diethyl ether and transfer quantitatively onto a 20- X 9.0 with 1M KH2P04 or 1M K2HP04 as necessary. The 20-cm Quantum Q4F silica gel thin-layer chromatographic supporting electrolyte was 0.1N HzS04 prepared from U1plete (Quantum Industries, Fairfield, N.J.), Rinse the trex (J. T. Baker) grade concentrated sulfuric acid. Other tube with two successive 50-pl aliquots of ether and transreagents included chloroform (Fisher), methanol (J. T. fer to the thin-layer chromatoplate. Develop the plate in a Baker), diethyl ether (anhydrous; Mallinckrodt) and acevapor saturated chamber using ch1oroform:acetone (60:40) tone (J. T. Baker). until the solvent front has ascended 15 cm. The plate is Standard Solutions. Dissolve 25.0 mg each of I [Chlordiexamined under short wave UV, and the areas on the siliazepoxide; 7-ch~oro-2-methylamino-5-phenyl-3H-1,4-benca gel corresponding to I ( R f = 0.20), I1 (Rf = 0.05), and zodiazepine-4-oxide, C16H14N30C1 (mol wt = 299.71; mp I11 ( R f = 0.33) are identified by the R f of 5 pg of authentic = 236-236.5 "C)]; II [N-desmethylchlordiazepoxide;2standards run alongside the sample extracts. These areas amino-7-chloro-5-phenyl-3H-1,4-benzodiazepine-4-oxide,are scraped off and transferred into a 15-ml centrifuge C15H12N30Cl (mol wt = 285.73, mp = 255-256 "C)], and tube to which 5 ml of methanol is added. The tubes are I11 [Demoxepam, Lactam; 7-chloro-1,3-dihydro-2H-1,4- slurried on a Vortex mixer for approximately 30 seconds and centrifuged for 5 minutes at 2000 rpm to spin down benzodiazepine-2-one-4-oxide,Cl5HllNzOzCl (mol wt = 286.72, mp = 235-236 "C)] in 25 ml of methanol to give the silica gel. The methanol supernatant is transferred stock solutions each containing 1 mg/ml. Dilute 1 ml of into another 15-ml conical centrifuge tube and evaporated each stock solution to 100 ml with methanol to, give workto dryness in a 65 "C water bath under a stream of nitrogen. To the residues, add 2.0 m'l of 0.1N sulfuric acid 1 ing solutions containing, 10.0 pg/ml. Fresh working solutions should be prepared every two weeks. Suitable aliminute prior to DPP analysis. Mix well on the Vortex quots of these solutions are added to blood as internal super mixer and deoxygenate the samples for 5 minutes standards. with nitrogen gas bubbled in through a microfilter stick Parameters for Polarographic Analysis. Instrument. [Scientific Glass Apparatus Inc., Bloomfield, N.J. (No. Princeton Applied Research Corporation (PAR) Model 174 JD-5385-coarse porosity)]. Transfer the deoxygenated 1076

ANALYTICAL CHEMISTRY, VOL. 46, NO. 8, JULY 1974

I

-0.600 V.

o

5r

f

O.l,uA

L

O(1

CURREN

+ z w

a

a

v 3

0 2-

01-

-6.450 -b.550 -0.650 POTENTIAL (VOLTS VI. S.C.El

Figure 3. Polarograms and sample calculations of resultant current for chlordiazepoxide ( I ) ( A ) Plasma, 3-hour post-oral administration of 30-mg dose. ( E ) Plasma con t ro I

sample into the polarographic cell and analyze the samples for I, 11, and I11 by scanning between -0.450 and -0.800 ‘V us. SCE using the differential pulse mode of operation and the polarographic parameters previously described. The polarograms are recorded on the X-Y recorder and show analytical peaks a t E, = -0.600 V, -0.590 V, and -0.645 V us. SCE for I, 11, and 111, respectively. Calculations. Calculate the resultant current due to the polarographic reduction of I, 11, and I11 on the basis of their peak heights (h,) measured a t -0.600 V, -0.590 V and -0.645 V us. SCE, respectively, see Figure 3. The resultant current is calculated as follows: h , (in cm) X 25.4 cm (full scale deflection) current scale @A) = sample @A) (1) The concentration of I, 11, or I11 is calculated on the basis of the resultant current of the internal standard, thus: @A) sample Concn of Int. std @g) X @A) Int. Std ml of plasma p g of compound/ml of plasma (2) Alternately, the concentrations of I, 11, or I11 can be determined from internal standard curves (Figure 4 ) . I t should be noted that I11 which is reduced at the highest potential ( E , = -0.645 V us. SCE) shows a sloping base line at low concentrations. This is caused by backgr6und interferences from reducible impurities in the supporting electrolyte. This results in a non-zero intercept for I11 (Figure 4), and therefore it is recommended that for highest accuracy the graphical technique be used for the measurement of I11 at low concentration (10.2pglml). The current (PA) per kg of the internal standards is then compared directly to that of the external standards to obtain per cent recovery, thus: (pA/pg/ml) Int. std x 100 = YO Recovery (pA/pg/ml) Ext. std

(3)

RESULTS AND DISCUSSION Polarographic Parameters. Polarographic analysis of I and its metabolites I1 and 111 following therapeutic doses of Librium is possible only because of new instrumenta-

1

1

0 0 5 01

1

I

0 25

05

2 IO

PLASMA CONCENTRATION ( p g / m l I

Figure 4. Polarographic calibration curves of I , I I , and I I I recovered from plasma as internal standards 0-0 Chlordiazepoxide ( I ) , x and O-*-O Demoxepam ( I I I)

--x

N-desmethyl chlordiazepoxide ( I I ) ,

tion which utilizes differential pulse polarography (DPP). Classical direct current (dc) polarography cannot attain sufficient sensitivity to measure these levels. DPP analysis yields sufficient sensitivity to measure these levels because of an increased output signal, which is a direct result of the current sampling procedure of DPP and an instrumental gain factor (10-fold) which is incorporated into the PAR equipment (20). A comparison of these two techniques is illustrated in Figure 5. A sample of 100 pg of I dissolved in 3.0 ml 0.1N HzS04 was scanned a t a rate of 5 mV/sec from 0.000 to -1.300 volts us. SCE with a 0.5-second drop time, employing both techniques. A comparison of the sensitivity employed shows that the output signal (PA) for the reduction of the 4,5-azomethine functional group at -0.600 V us. SCE is over 20-fold higher for DPP us. dc polarography. The calculation of diffusion current for quantitation is greatly simplified (see calculation section for details) employing DPP in comparison to dc polarography. In DPP, the standard saw tooth “s” shaped display of dc polarography is replaced by Gaussian shaped curves with readily discernible peaks. As a result, the time-consuming extrapolations to find the limiting and residual currents necessary in dc polarography are eliminated. The nature of the DPP display enables the measurement of the E , values to be performed far more easily and reproducibly than the measurement of the corresponding Ell2 values of dc polarography. The high degree of resolution attainable employing DPP is shown in the analysis of 50 kg/ml of I, 11, and I11 in 0.1N HzSO4 supporting electrolyte (Figure 6). This resolution of DPP enables the analyst to discern changes in structure such as those encountered in metabolite identification. Although all three compounds possess an N4oxide and a 4,5-azomethine functional group, the peaks due to these reducible groups in each compound are all distinguishable because of the peak shifts caused by structural differences a t the N 1 - c ~position. Compounds I (20) Princeton Applied Research Corporation, Instruction Manual for PAR Model 174 ( I X - Z ) , Princeton, N.J. (1973).

ANALYTICAL CHEMISTRY, VOL. 46, NO. 8, JULY 1974

1077

t'

--

ED .-0.6OOV

D.PP

-0.OOO

I -0.200

-0.400

.VI

-Loo0

-0.800

-0.600

POTENTIAL (VOLTS

-1.200

Sc.a

Figure 5. Comparison of dc polarography vs. DPP for I

T 2.OgA I

A CURRENT

1

-0.oOO

I

-0.200

1

-0.400

I

POTENTIAL (VOLTS .VI

Figure 6. Polarographic scans of I, I I , and I I I in 0.1 N Hzs0.1 1078

ANALYTICAL CHEMISTRY, VOL. 46, NO. 8, JULY 1974

I

-moo

-0.600 SC.Q

I

I

-Loo0

I

-1.200

I

Table 11. C u r r e n t (PA) as a Function of Scan Rate and Drop Time.

Table I. Polarographic D a t a f o r Chlordiazepoxide and Its Metabolitesa Compound

Chlordiazepoxide (I) Desmethylchlordiazepoxide (11) Demoxepam (111)

E, us. SCE, V

-0,275 -0.600 - 1.135 -0,275 -0.590 - 1.020 -0.315 -0.645

Bond reduced

3 Na+O

>C s=N4-Nl=Cz


C srN4-

-NI=CZ 3 N,-0 >C s=N


0

-0 E

2

5 z &w

0

-0 4

C

I

I

I

1

I

I

I

)

I

3

5

7

9

II

13

14

PH

Figure 7. Effect of pH on peak

potentials of I , I I, and I I I

A N A L Y T I C A L C H E M I S T R Y , VOL. 46, NO. 8, J U L Y 1974

1079

Y

t

m i N

1

I

1

I

I O

O

I O

O

O

m

c

O

n

t

(Q$l

I

lo

I

I O N

O

"

-. O

lN3MMn3

-1. w 0

aii

I m

I

0 c

I 0

I 0

I 0

I 0

I 0

LD

0

0

n

N

( V d l lN3MMn3

1080

-

:

P

0

The choice of the proper supporting electrolyte for polarographic determination was based on studies of peak potential and current us. pH. Ten-pg samples of each compound were analyzed by DPP in 0.1N H2SO4, in 1M phosphate buffers ranging in pH from 3 t o 12.8, in 0.1N NaOH (pH 13), and in 1N NaOH (pH 14). Graphs of E,, (in volts us. SCE) us. pH indicated that a linear relationship exists for both the 4,5-azomethine and N4-oxide functional groups in compounds I, 11, and III, while the 1,2azomethine functional group of compounds I and I1 does not show this linear relationship (Figure 7 ) . The N4-oxide and 4,5-azomethine functional group showed a change in E p of 81.4 mV and 65.4 mV, respectively, per pH unit. Plots of current (FA) us. pH for each functional group showed that the reduction of the 4,5-azomethine functional group yielded the highest currents at all pH values (Figures &4, 8B, and 8C). Inspection of the current us. pH plots indicates that a complex polarographic reduction mechanism exists above pH 9. This is in agreement with the results of Clifford and Smyth ( 1 7 ) . However, it is apparent that optimal sensitivity is obtained at lower pH values, hence 0.1N H2SO4 was selected as the supporting electrolyte. Assay Parameters. Extraction of the biological specimen was carried out as described in the spectrofluorometric assay ( 3 ) with the following modifications. Compounds I, 11, and I11 in the plasma (buffered to pH 9.0 with phosphate buffer) were extracted twice into diethyl ether and the extracts were combined. Separation of the three compounds in the diethyl ether extract from each,other prior to DPP analysis could be effected only by TLC. A solvent back-extraction into alkali to separate compounds I and I1 from compound I11 is pos= sible (3), but, unfortunately, compound I11 undergoes - rearrangement with the resultant loss of polarographic ac: tivity. For this reason, thin-layer chromatography using ch1oroform:acetone (60:40) in a vapor saturated chamber 2 as the developing system was employed to effectively sep= arate I, 11, and I11 from each other and from co-extracted 2 -m biological materials. Cellulose, alumina, and silica gel ad2 sorbants were also examined for the separation of the compounds. Of the three adsorbants, only silica gel gave good resolution. Chromatoplates manufactured by E. Merck [silica gel SG-60 F254, Cat. No. 5765, Brinkmann 2! 5 Instruments, Westbury, N.Y.] and Quantum Industries a x [Q4F silica gel plates Cat. No. 2020, Quantum Industries, % Fairfield, N.J.] were evaluated for elution efficiency. $ Higher elution efficiencies were obtained from the Quantum plates than the Merck plates, probably due to the smaller amount of gypsum (CaS04.2H20) binder used on - the Quantum plate. n A number of different solvents were examined to increase elution efficiency. The highest elution recovery 5 from silica gel was obtained with solvents of high polarity. E These solvents were prepared with water and/or aqueous 0.W sulfuric acid solutions and methanol in varying proportions, The most efficient recovery from silica gel was 5 obtained with methanol-0.1N sulfuric acid, the recovery 5 decreasing with methanol-water, and was lowest with Z methanol itself. However, the use of 0.1N sulfuric acid + 8 was precluded because of the acid hydrolysis of compound E I1 when left in this medium for a period of time greater 6 than one hour. Also, zinc silicate, the fluorescent indicaE tor on the silica gel plates, was eluted with these aqueous .g methanolic solvents. As a result, an interfering polaroY graphic peak due to the reduction of ZnZ+ was obtained approximately -1.000 volts us. SCE. Since the concentration of Zn2+ was in great excess over that of I, 11, and 111, its presence made the measurements of the three com-

I

-

0

lo

ANALYTICAL CHEMISTRY, VOL. 46, NO. 8 , JULY 1974

Table IV. Plasma Levels of Chlordiazepoxide and Its Metabolites Following a Single 30-mg Oral Dose Time, hr

Chlordiazepoxide, I/ml

1.55 2.39 2.12 1.07 1.28 1.19 0.97 0.75 0.16

1

1.5 2 3 4 6 8

12 24 30 48 72

0.11

0.06 n.m.

N-Desmethylchlordiazepoxide, Pg I I / d

Demoxepam, pg III/ml

0.08 0.18 0.29 0.15 0.26 0.39 0.56 0.39 0.33 0.32 0.13 n.m.

n.m.a n.m. n.m. n.m. n.m. n.m. n.m. n.m. n.m. n.m. 0 . 05b 0.08b

a n.m.