BIOPHARMACEUTICS & DRUG DISPOSITION, VOL. 12, 425434 (1991)
PHARMACOKINETICS A N D ORAL BIOAVAILABILITY OF CARBIMIDE IN MAN R. OBACH*, J . TORRENT+, H . COLOM*, J . PRUGONOSA*,
c. PEWRE*
AND
J. DOMENECH*
*Pharmacokineticsand Biopharmaceutics Unit, School of Pharmacy, University of Barcelona. Nucleo Universitario de Pedralbes, Barcelona 08028, Spain 'Clinical Pharmacology Unit. Hospital Sant Pau. School of Medecine U.A.B..Sani Antoni M" Claret 167, Barcelona 08025, Spain
ABSTRACT A pharmacokinetic study of carbimide, an inhibitor of aldehyde dehydrogenase, used as an adjuvant in the aversive therapy of chronic alcoholism, has been carried out in male human volunteers for intravenous and oral administration. Carbimide plasma concentrations were determined by a sensitive and specific high performance liquid chromatographic method. The intravenous doses administered were 0.1,0.3, 0.6, and 1 mg kg-I and linear pharmacokinetics were observed for this dose range. Elimination half-life and total plasma clearance values ranged from 42 to 52 min and from 14.4 to 205ml kg-l min-I, respectively. After oral administration of 1 and 1.5mgkg-l of carbimide, elimination half-life values were 75 and 61 min, respectively, being higher than the corresponding value obtained after 0.3 mg kg-' doses, i.e. 39 min. In all cases, rapid absorption was indicated by t,, values ranging from 10.5 to 15.5 min. Absorption was not complete, the oral bioavailability being 53 per cent and 70 per cent for the 0.3 and 1mg kg-I carbimide dose, respectively. The data indicate that there is a first-pass effect for carbimide. KEY WORDS
Carbimide
Cyanamide Pharmacokinetics Humans
Bioavailability
INTRODUCTION Carbimide (cyanamide) is an inhibitor of aldehyde dehydrogenase (EC. 1.2.1.3.; ALDH), and is used in the therapeutic management of chronic alcoholism. The drug acts by blocking ethanol biotransformation and consequently increasing hepatic and blood acetaldehyde concentrations.' The calcium carbimideethanol interaction has been widely studied in man. Brien et al.' characterized the time-course, intensity, and duration of this pharmacodynamic interaction, indicating that the cardiovascular response is correlated with blood acetaldehyde concentration. Further, they found that the clinical impact of this interaction is related to the ethanol dose and depends on the interval between calcium carbimide administration and the intake of alcohol. The same authors found a marked intersubject variability of this interaction in a previous r e p ~ r t . ~ Correspondenceto: R. Obach, Pharmacokineticsand BiopharmaceuticsUnit, School of Pharmacy. University of Barcelona, Nucleo Universitario de Pedralbes, Barcelona 08028, Spain.
0 142-278219 11060425- 10$05.OO 0 1991 by John Wiley & Sons, Ltd.
Received 13 August 1990 Revised 4 February I991
426
R . OBACH E T A L .
Tachycardia, hypotension, and facial flushing are correlated with elevated blood acetaldehyde concentration, and these cardiovascular effects appear to be the underlying pharmacologic basis for the use of this drug in the treatment of chronic alcoholi~rn.~ Although calcium carbimide has been clinically available for the past 50 years, there are few data regarding its pharmacokinetic behaviour in man and its oral bioavailability is as yet unknown. Nevertheless, it has been demonstrated that N-acetylcarbimide is the main urinary metabolite in rat, rabbit, dog and also in man.6 Therefore, the aim of the present work was to study in volunteers the pharmacokinetic profile and absolute oral bioavailability of carbimide for several doses given by intravenous infusion and oral administration. MATERIAL AND METHODS Subjects
Ten healthy paid male volunteers with a mean age of 27-4 years (range 20 to 35) and a mean body weight of 77.5 kg (range 70 to 89) took part in the study. They underwent physical examinations and extensive laboratory screening before entering the trial. The subjects were not consumers of alcohol or tobacco, and were instructed to abstain from all food and drink containing ethanol 1 week before, during, and 1 week after the study period. The intake of any drug other than the trial medication was not allowed within 4 weeks before and during the study. The protocol was approved by the Hospital's Ethical Committee and the study was performed in accordance with the rules and regulations for conducting clinical trials stated in the Declaration of Helsinki and revised by the World Medical Assembly at Tokyo and Venice. Informed written consent was obtained from all subjects. The volunteers were free to withdraw from the study at any time. Drug administration
Carbimide (Fluka, Buchs, Switzerland) was administered both orally and by intravenous infusion. Carbimide parenteral doses of 0.1, 0.3, 0.6, and 1 mg kg-' were prepared under sterile conditions and dissolved in a 50 ml saline solution. Drug administration was performed using a volumetric pump at a constant infusion rate of 150mlh-' for a fixed period of 20 min. Oral doses of carbimide of 0.3, 1, and 1.5mg kg-' were dissolved in water and prepared in 20 ml monodose vials. Study design and procedures
The trial was divided into two consecutive phases. The aim of the first phase was to assess tolerability of carbimide for intravenous administration. Therefore, four subjects participated in each of the four scaled intravenous carbimide
CARBIMIDE
427
doses. Subjects 1 to 4 received 0-1mg kg-' carbimide intravenous infusion; subjects 3 to 6 received 0.3 mg kg-'; subjects 5 to 8 received 0.6 mg kg-I; and, finally, subjects 7 to 10 received the highest carbimide infusion dose of 1 mg kg-I. The last four subjects, who received the highest intravenous dose, were included in the second study phase, which involved the oral administration of three carbimide doses in single-blind conditions. Because the duration of the inhibition of ALDH activity has not been described, especially after the i.v. administration of carbimide, the wash-out period between each drug administration was set at 15 days. In all cases drug administration was given at 8:OO am. The subjects received a light breakfast and standard lunch at 2.5 and 6 h after the drug intake. Heparinized blood samples were obtained for carbimide plasma concentration determinations; the samples were drawn by means of an indwelling catheter, inserted into a forearm vein, at the following times: at baseline (before drug infusion), at 2, 5, 15, and 20 min during the drug infusion period, and at 5 , 10, 15, 30, 60, 120, 240, 480, and 1440 min post-infusion. Samples after oral drug administration were collected at baseline and at 2, 5, 7, 10, 12, 15, 20, 40, 60, 90, 120, 240, 360, 480, and 1440 min post-treatment. Plasma was separated by centrifugation at 4" at 1000 g for 15 min and frozen at - 20" until analysed. Blood pressure, heart rate, and biochemical check-ups were monitored at baseline and at several times after each drug administration. Effects experienced by the subjects were also recorded on the individual case report form. Analytical method
Carbimide plasma concentrations were determined using a high pressure liquid chromatographic (HPLC) method with spectrofluorometric detection of the carbimide dansyl deri~ative.~ Chromatographic analysis was carried out using a Waters HPLC system (Waters Associated Inc., Mildford, MA, USA) equipped with two M-510 solvent delivery systems, a M-721 solvent programmer, a Wisp 710B automatic injector, and a M-420 AC fluorescence detector. In our hands, this method was accurate (maximum relative error was 5.6 per cent) and reproducible (maximum relative standard deviation was 9-8 per cent). The assay was linear within the range of 25 to 500 ng m1-I. The detection limit was l0ngml-'. Pharmacokinetic calculations
The experimental carbimide plasma concentration values obtained after i.v. drug infusion were fitted to compartmental open models using an extended non-linear least-squares regression method (ELSMOS program).* The model that best described the raw data was selected according to MAICE's M rite ria.^
428
R . OBACH ET A L .
The following polyexponential equations were applied to the observed carbimide plasma concentration-time data: i=m
C=
C, . (1 - e-K
'x)
. e-K
(l-fJ/(l - e-K
T)
[=I
where C is the drug concentration at time t , T is the infusion time, t , equals t during the infusion period and equals T during the post-infusion time and the Z is over m exponential terms, C, is the linear concentration coefficient, and K, the drug disposition constant. Based on the coefficients and exponents of the fitted exponential function, the following pharmacokinetic parameters were calculated using the equations proposed by Wagner:'O area under the curve from 0 to infinity (AUCo-,), total plasma clearance (CL,), elimination half-life ( t l & and volume of distribution at steady-state (Vd,,). The data obtained after oral carbimide administration were handled according to a non-compartmental approach (pkcalc program)," which allows the estimation of the following parameters:12 area under the concentration-time curve (AUCo- -), 8-elimination half-life (tl/J, experimental peak plasma concentration (Cmx),and time required to achieve the peak (tmaX). To evaluate whether carbimide displayed linear pharmacokinetic behaviour in man, C b , t,/2, Vd,, and AUCo-, normalized for drug dose were compared across doses. Statistical analysis
Comparisons of the pharmacokinetic parameters across drug doses were performed by a one-way ANOVA using the Scheffk test for assessing differences between groups. Homogeneity of variance was tested by applying the Bartlett test. When necessary, a non-parametric statistical analysis, such as KruskallWallis or Mann-Whitney test, was used. Linear regression between AUC values and administered drug doses was carried out, and the intercept value of the regression was compared to zero by means of a non-paired t-test. In all cases a p < 0.05 was considered significant. RESULTS Mean carbimide plasma levels obtained after i.v. administration of 0.1, 0-3, 0.6, and 1 mg kg-' are shown in Figure 1. After the infusion, carbimide concentration displayed a biphasic disposition profile and a parallelism with the monoexponential terminal slopes. According to the MAICE test, the experimental data were best fitted to a two-compartment open model. The main pharmacokinetic parameters (mean f SD) obtained for each i.v. dose of carbimide are presented in Table 1. Mean distribution constant values ( a )ranged between 0.15 and 0.24 min-I. The mean 8-elimination half-life values
429
CARBIMIDE
ioooo
i
i'
0
I
60
120
id0 TIME
240
300
360
420
(mid
Figure 1. Mean carbimide plasma level-time curves after i.v. carbimide infusion of 0.1 mg kg-l (a), 0.3 m g kg-I (q, 0.6 mg kg-' (A), and 1 mg kg-l(+) to healthy male subjects
were very close and only fluctuated between 42 and 52 min. AUC values increased proportionally with the administered dose according to the following linear regression equation: AUC = 72 498.4 D - 3268.7 (r = 0.997,p < 0.01) The intercept of this regression line (3268.7 f 2367.6) was not statistically different from zero (t = 1-38, p > 0.05). Total plasma clearance values ranged from 14.4 to 20.5 ml kg-' min-I, and no statistical differences were found when compared across the administered doses. Apparent volume of distribution values in the central compartment and for steady-state conditions ranged from 221 to 296 and 668 to 800ml kg-', respectively. Comparisons of the main pharmacokinetic parameters across the doses of carbimide studied suggest that carbimide displays linear pharmacokinetics after i.v. administration. Carbimide plasma concentration-time profiles after oral administration of 0.3, 1, and 1.5mg kg-' carbimide are depicted in Figure 2, and the main pharmacokinetic parameters (mean f SD) are shown in Table 2. AUC values increased proportionally with the administered dose of carbimide according to the following linear regression equation: AUC = 60 425.2 D - 9010.8 (r = 0.998, p < 0.05) The intercept of this regression line (9010.8 rf: 4248.1) was not statistically different from zero (t = 2.12, p > 0-05). The oral absorption of carbimide was rapid with tmaxvalues ranging from 10.5 to 15.5 min. Peak plasma carbimide
282.0 f 101.65 854.3 f 194.9
1 48625 f 3290 902 f 362 10.5 f 4.2 76.5 f 13.3* 0.70 f 0.16
10279 f 5159 197 f 134 15.5 f 5.3 39.9 f 7.5 0.53 f 0.24
Dose (mg kg-') 0.3
* p < 0.05 with respect to 0.3mgkg-I (ANOVA followed by Scheffe's test).
Area under curve (AUCO-,) Plasma peak concentration (CmaX) Time to peak (tmax) Elimination half-life (tI1*) Bioavailability (F)
Parameter
238.59 f 73.02 667.7 f 143.1
221.4 f 78.02 800.3 f 118.8
0.1827 f 1218 0.0136 f 0.0024 52.3 f 8.53 37 077.0 f 9809.3 17.06 f 4.44
0.6
295.9 f 141.1 799.3 f 108.2
*
0.1496 f 0.0619 0.0137 f 0.0024 51.75 f 8.81 70902.0 f 11844 14.43 2.74
1 .O
Units ng m1-I min ngml-l min min -
83 254 f 26 144 1706 f 1040 14.6 f 4.6 61.5 f 2.8* -
ml kg-l ml kg-l
min-l min-I min ngml-l rnin ml kg-' min-l
Units
1.5
Table 2. Pharmacokinetic Parameters (mean f SD) of carbimide in man after oral administration
0.2093 f 0.1003 0.0173 f 0.0060 43.5 f 13.79 18 838.8 f 2292.4 16.09 f 1.77
f 0.0600 5 0.0045 f 8.95 f 1187.3 f 5.32
0.2433 0.0173 41.9 5104.2 20.5
Rapid disposition constant ( a ) Slow disposition constant (/3) Elimination half-life ( t , / * ) Area Under Curve (AUC,-,) Total plasma Clearance (C1,) Volume of distribution central compartment ( V,) Volume of distribution (Vd,,)
Dose (mg kg-I) 0.1
Parameter 0.3
Table 1. Pharmacokinetic parameters (mean f SD) of carbimide in man after intravenous infusion
?
h Y b
P
0
w
P
43 1
CARBIMIDE
concentration increased in accordance with the administered drug dose. After oral administration of 1 and 1.5mg kg-' of carbimide /3-elimination half-life values were 76.5 and 61.5 min, respectively. These values were statistically higher than those obtained after oral administration of 0.3 mg kg-' (39-9 min). Oral bioavailability after 0-3mg kg-l carbimide was not complete, being only 53% (see Figure 3). The corresponding estimated value after 1 mg/kg was 70%. No relevant clinical changes in blood pressure and heart rate were observed during the study, and no abnormal findings were found in the biochemical and haematological analysis. Furthermore, no adverse effects were observed after intravenous infusion or oral administration of carbimide doses.
DISCUSSION
The results of this study represent the first pharmacokinetic investigation of carbimide in man involving parenteral and oral single-dose administration. In spite of the fact that this drug has been used clinically for more than 50 years, the lack of a sensitive and reproducible analytical method for measuring carbimide plasma levels after oral doses can account for the absence of human pharmacokinetic data. Nevertheless, Shirota et d 6have demonstrated that hepatic biotransformation of carbimide gives N-acetylcarbimide as the main
-0
60
120
160
240
300
360
420
TIUE (mid Figure 2. Mean carbimide plasma level-time-curves after single oral carbimide administration of 0.3 mg kg-I 0 , l mg kg-I (+),and 1.5 mg kg-I (0)to healthy male subjects
432
R. OBACH ET A L .
100Or
10' 0
I
30
90
60
TIME
120
150
hinl
Figure 3. Comparative mean carbimide plasma level-time curves after intravenous infusion (0) and oral administration (.) of single doses of 0.3 mg kg-' carbimide
metabolite in different species including man. There is no previous report concerning the pharmacokinetics of carbimide for i.v. drug administration, as carbimide is usually taken orally in the treatment of alcoholism. Therefore, the first part of this study dealt with tolerability and clinical safety of carbimide for i.v. administration of increasing doses of the drug. Also, the i.v. route is needed to obtain an unambiguous estimation of the main pharmacokinetic parameters and to calculate absolute bioavailability of a drug. In this study, it has been observed that the open two-compartment model accurately predicts the carbimide plasma level-time course after i.v. administration (MAICE test, p < 0.05). The elimination half-life of carbimide in man was short, being less than 1 h for all the doses administered. This is in agreement with the elimination half-life data reported by Obach et a l l 3 for rats (mean value of 33 min) and for dogs (values ranged between 39 to 61 min). Plasma clearance values did not change for the range of doses studied (144-20.5 ml kg-' min-I), and were similar to the data reported for dogs (12.G19.7ml kg-' min-I), but lower than the values observed in rats (1 17 ml kg-I This difference from the plasma clearance values for rats could be explained by the higher apparent volume of distribution (Vd,,) in this species. For the apparent volume of distribution, there is a close correspondence between the value for man (around 0.8 1kg-l) and the value reported for dogs (around 1 1kg-I), and a clear difference with the data obtained for rats (3.8 1 kg-'. The carbimide erythrocytes concentration for rats is much higher
CARBIMIDE
433
than the intra-erythrocyte carbimide concentration observed for dogs or man.14 This fact could explain the higher apparent volume of distribution of carbimide for rats. The pharmacokinetics of carbimide in man within the range of intravenously administered doses (0.1-1 mg kg-I) were linear. No statistical differences (p > 0.05) were found when comparing pharmacokinetic parameters, such as total plasma clearance, elimination half-life, central and steady-state volume of distribution, and area under the curve normalized by the administered dose (AUC/D), indicating that they were dose-independent. Conversely, statistical differences (p < 0.05) were obtained when the elimination half-life values after oral administration were compared across the doses. This fact suggests nonlinear pharmacokinetics after oral administration. Dose-dependent pharmacokinetics have been described in rats and rabbits with higher doses of carbimide,I53I6and in dogs within a dose range of 1 4 m gkg-’. I 3 The dose-dependent pharmacokinetics could be related to saturation of the carbimide N-acetylation process. In fact, Shirota et have pointed out the existence of inter-species differences regarding the amount and/or activity of the N-acetyltransferase enzyme, which, in turn, could explain the possibility of saturation at different carbimide doses for each species. It remains to be determined whether this phenomenon exists in the human. The absorption rate of carbimide, evaluated by the time required to achieve peak plasma drug concentration, was rapid. The values obtained were in the range of those observed in other animal species such as dog and rat.13 The fast absorption of carbimide could be due to its low molecular weight and its structure, which allow simultaneous diffusion, through gut epithelial membranes and pores.I7 The oral bioavailability of carbimide was not complete. The extent of absorption was also incomplete when carbimide was administered in other animal species; bioavailability values of 69 per cent and 65 per cent have been reported Conversely, Deitrich for rats (2 mg kg-I) and dogs (4 mg kg-I), respe~tive1y.l~ et aZ.,’* using I4C-carbimide, demonstrated complete absorption of the drug after oral administration. All these findings suggest that a first-pass effect takes place in man after oral administration of carbimide. Based on calcium carbimide-ethanol interaction studies reported mainly by and the carbimide pharmacokinetic data Brien et a1.1-3 and Peachey et obtained in this study, there appears to be no relationship between the time courses of the increased blood acetaldehyde and the plasma carbimide concentration. This lack of correlation has recently been corroborated using the inhibition of human erythrocyte ALDH as an index of the pharmacological effect of carbimide.19 After oral administration of 1 mg kg-’ carbimide, inhibition of erythrocyte ALDH lasted 6 days, whereas carbimide was only measurable in plasma for 6 h. This study provides the first preliminary data on the pharmacokinetics profile of carbimide in man after intravenous and oral administrations. Further studies
434
R. OBACH ET AL.
should be required to elucidate the complex pharmacokinetic and pharmacodynamic relationships of carbimide in man.
ACKNOWLEDGEMENTS This work was partially financed with a grant from the FISS (number 87/1404). The authors wish to thank Professor F. Jane for kindly authorizing the use of the Clinical Pharmacology Unit's facilities to conduct this study.
REFERENCES 1. J. F. Brien, J. E. Peachey, B. J. Rogers and C. W. Loomis, Eur. J. Clin. Pharmacol., 14, 133 (1978). 2. J. F. Brien, J. E. Peachey, C. W. Loomis and B. J. Rogers, Clin. Pharmacol. Ther., 25, 454 (1979). 3. J. F. Brien, J. E. Peacheyand C. W. Loomis, Clin.Pharmacol. Ther., 27,426 (1980). 4. J. E. Peachey, J. F. Brien, C. W. Loomis and B. J. Rogers, Alcohol Clin.Exp. Res., 4, 322 (1 980). 5 . J. E. Peachey and C. A. Naranjo, Drugs, 27,171 (1984). 6. F. N. Shirota, H. T. Nagasawa, C. H. Kwon and E. G. De Master, Drug Metab. Disposit., 12,337 (1984). 7. J. Pruiionosa, R. Obach and J. M. Vallts, J. Chrumatogr., 377,253 (1986). 8. R. J. Francis, Comp. Prog. Biumed., 18,43 (1984). 9. H. Akaike, Math. Sci., 14, 156,s (1976). 10. J. G. Wagner, Fundamentals of Clinical Pharmacokinetics, Drug Intelligence Publications, Illinois, 1985. 11. R. C. Shumaker, Drug Metab. Rev., 17,341 (1986). 12. D. Perrier and M. Mayersohn, J. Pharm. Sci., 71,372 (1982) 13. R. Obach, H. Colom, J. Arso, C. Peraire and J. Pruiionosa, J. Pharm. Pharmacol., 41, 624 (1989). 14. R. Obach, J. Vallts, J. Pruiionosa, J. Arso and J. M. Vallts, Drogodependencias, un reto multidisciplinar, 1, 29-35. Ed. Servicio Central de Publicaciones del Gobiemo Vasco, Vitoria, Spain, 1984. 15. R. Obach, J. Pruiionosa, A. Menargues, M. Nomen, J. M. Valltsand C. Peraire, Rev. Farmacol. Clin.Exp. 3, 186 (1 986). Ed. Prous, Barcelona, Spain. 16. R. Obach, J. Moreno, J. Domenech and J. M. Pla-Delfina, Actas Ze' Congreso Europe0 de Biofarmacia y Farmacocine'tica,Technique et Documentation, Clermont Ferrand, 2,367 (1981), France. 17. J. M. Pla-Delfinaand J. Moreno, J. Pharmacokinet. Biopharm., 9,191 (1981). 18. R. A. Deitrich, P. A. Troxell, W. Worth and G. V. Erwin, Biochem. Pharmacol., 25, 2733 (1976). 19. R. Obach, J. Pruiionosa, J. Torrent, H. Colom, I. Izquierdo, and J. Domenech. Eur. J. Drug Metab, Special Issue, 4th European Congress of Biopharmaceutics and Pharmacokinetics, 1990, Abstract 167, p. 45. Medecine et Higiene Publishers, GenCva.
PHARMACOKINETICS A N D ORAL BIOAVAILABILITY OF CARBIMIDE IN MAN R. OBACH*, J . TORRENT+, H . COLOM*, J . PRUGONOSA*,
c. PEWRE*
AND
J. DOMENECH*
*Pharmacokineticsand Biopharmaceutics Unit, School of Pharmacy, University of Barcelona. Nucleo Universitario de Pedralbes, Barcelona 08028, Spain 'Clinical Pharmacology Unit. Hospital Sant Pau. School of Medecine U.A.B..Sani Antoni M" Claret 167, Barcelona 08025, Spain
ABSTRACT A pharmacokinetic study of carbimide, an inhibitor of aldehyde dehydrogenase, used as an adjuvant in the aversive therapy of chronic alcoholism, has been carried out in male human volunteers for intravenous and oral administration. Carbimide plasma concentrations were determined by a sensitive and specific high performance liquid chromatographic method. The intravenous doses administered were 0.1,0.3, 0.6, and 1 mg kg-I and linear pharmacokinetics were observed for this dose range. Elimination half-life and total plasma clearance values ranged from 42 to 52 min and from 14.4 to 205ml kg-l min-I, respectively. After oral administration of 1 and 1.5mgkg-l of carbimide, elimination half-life values were 75 and 61 min, respectively, being higher than the corresponding value obtained after 0.3 mg kg-' doses, i.e. 39 min. In all cases, rapid absorption was indicated by t,, values ranging from 10.5 to 15.5 min. Absorption was not complete, the oral bioavailability being 53 per cent and 70 per cent for the 0.3 and 1mg kg-I carbimide dose, respectively. The data indicate that there is a first-pass effect for carbimide. KEY WORDS
Carbimide
Cyanamide Pharmacokinetics Humans
Bioavailability
INTRODUCTION Carbimide (cyanamide) is an inhibitor of aldehyde dehydrogenase (EC. 1.2.1.3.; ALDH), and is used in the therapeutic management of chronic alcoholism. The drug acts by blocking ethanol biotransformation and consequently increasing hepatic and blood acetaldehyde concentrations.' The calcium carbimideethanol interaction has been widely studied in man. Brien et al.' characterized the time-course, intensity, and duration of this pharmacodynamic interaction, indicating that the cardiovascular response is correlated with blood acetaldehyde concentration. Further, they found that the clinical impact of this interaction is related to the ethanol dose and depends on the interval between calcium carbimide administration and the intake of alcohol. The same authors found a marked intersubject variability of this interaction in a previous r e p ~ r t . ~ Correspondenceto: R. Obach, Pharmacokineticsand BiopharmaceuticsUnit, School of Pharmacy. University of Barcelona, Nucleo Universitario de Pedralbes, Barcelona 08028, Spain.
0 142-278219 11060425- 10$05.OO 0 1991 by John Wiley & Sons, Ltd.
Received 13 August 1990 Revised 4 February I991
426
R . OBACH E T A L .
Tachycardia, hypotension, and facial flushing are correlated with elevated blood acetaldehyde concentration, and these cardiovascular effects appear to be the underlying pharmacologic basis for the use of this drug in the treatment of chronic alcoholi~rn.~ Although calcium carbimide has been clinically available for the past 50 years, there are few data regarding its pharmacokinetic behaviour in man and its oral bioavailability is as yet unknown. Nevertheless, it has been demonstrated that N-acetylcarbimide is the main urinary metabolite in rat, rabbit, dog and also in man.6 Therefore, the aim of the present work was to study in volunteers the pharmacokinetic profile and absolute oral bioavailability of carbimide for several doses given by intravenous infusion and oral administration. MATERIAL AND METHODS Subjects
Ten healthy paid male volunteers with a mean age of 27-4 years (range 20 to 35) and a mean body weight of 77.5 kg (range 70 to 89) took part in the study. They underwent physical examinations and extensive laboratory screening before entering the trial. The subjects were not consumers of alcohol or tobacco, and were instructed to abstain from all food and drink containing ethanol 1 week before, during, and 1 week after the study period. The intake of any drug other than the trial medication was not allowed within 4 weeks before and during the study. The protocol was approved by the Hospital's Ethical Committee and the study was performed in accordance with the rules and regulations for conducting clinical trials stated in the Declaration of Helsinki and revised by the World Medical Assembly at Tokyo and Venice. Informed written consent was obtained from all subjects. The volunteers were free to withdraw from the study at any time. Drug administration
Carbimide (Fluka, Buchs, Switzerland) was administered both orally and by intravenous infusion. Carbimide parenteral doses of 0.1, 0.3, 0.6, and 1 mg kg-' were prepared under sterile conditions and dissolved in a 50 ml saline solution. Drug administration was performed using a volumetric pump at a constant infusion rate of 150mlh-' for a fixed period of 20 min. Oral doses of carbimide of 0.3, 1, and 1.5mg kg-' were dissolved in water and prepared in 20 ml monodose vials. Study design and procedures
The trial was divided into two consecutive phases. The aim of the first phase was to assess tolerability of carbimide for intravenous administration. Therefore, four subjects participated in each of the four scaled intravenous carbimide
CARBIMIDE
427
doses. Subjects 1 to 4 received 0-1mg kg-' carbimide intravenous infusion; subjects 3 to 6 received 0.3 mg kg-'; subjects 5 to 8 received 0.6 mg kg-I; and, finally, subjects 7 to 10 received the highest carbimide infusion dose of 1 mg kg-I. The last four subjects, who received the highest intravenous dose, were included in the second study phase, which involved the oral administration of three carbimide doses in single-blind conditions. Because the duration of the inhibition of ALDH activity has not been described, especially after the i.v. administration of carbimide, the wash-out period between each drug administration was set at 15 days. In all cases drug administration was given at 8:OO am. The subjects received a light breakfast and standard lunch at 2.5 and 6 h after the drug intake. Heparinized blood samples were obtained for carbimide plasma concentration determinations; the samples were drawn by means of an indwelling catheter, inserted into a forearm vein, at the following times: at baseline (before drug infusion), at 2, 5, 15, and 20 min during the drug infusion period, and at 5 , 10, 15, 30, 60, 120, 240, 480, and 1440 min post-infusion. Samples after oral drug administration were collected at baseline and at 2, 5, 7, 10, 12, 15, 20, 40, 60, 90, 120, 240, 360, 480, and 1440 min post-treatment. Plasma was separated by centrifugation at 4" at 1000 g for 15 min and frozen at - 20" until analysed. Blood pressure, heart rate, and biochemical check-ups were monitored at baseline and at several times after each drug administration. Effects experienced by the subjects were also recorded on the individual case report form. Analytical method
Carbimide plasma concentrations were determined using a high pressure liquid chromatographic (HPLC) method with spectrofluorometric detection of the carbimide dansyl deri~ative.~ Chromatographic analysis was carried out using a Waters HPLC system (Waters Associated Inc., Mildford, MA, USA) equipped with two M-510 solvent delivery systems, a M-721 solvent programmer, a Wisp 710B automatic injector, and a M-420 AC fluorescence detector. In our hands, this method was accurate (maximum relative error was 5.6 per cent) and reproducible (maximum relative standard deviation was 9-8 per cent). The assay was linear within the range of 25 to 500 ng m1-I. The detection limit was l0ngml-'. Pharmacokinetic calculations
The experimental carbimide plasma concentration values obtained after i.v. drug infusion were fitted to compartmental open models using an extended non-linear least-squares regression method (ELSMOS program).* The model that best described the raw data was selected according to MAICE's M rite ria.^
428
R . OBACH ET A L .
The following polyexponential equations were applied to the observed carbimide plasma concentration-time data: i=m
C=
C, . (1 - e-K
'x)
. e-K
(l-fJ/(l - e-K
T)
[=I
where C is the drug concentration at time t , T is the infusion time, t , equals t during the infusion period and equals T during the post-infusion time and the Z is over m exponential terms, C, is the linear concentration coefficient, and K, the drug disposition constant. Based on the coefficients and exponents of the fitted exponential function, the following pharmacokinetic parameters were calculated using the equations proposed by Wagner:'O area under the curve from 0 to infinity (AUCo-,), total plasma clearance (CL,), elimination half-life ( t l & and volume of distribution at steady-state (Vd,,). The data obtained after oral carbimide administration were handled according to a non-compartmental approach (pkcalc program)," which allows the estimation of the following parameters:12 area under the concentration-time curve (AUCo- -), 8-elimination half-life (tl/J, experimental peak plasma concentration (Cmx),and time required to achieve the peak (tmaX). To evaluate whether carbimide displayed linear pharmacokinetic behaviour in man, C b , t,/2, Vd,, and AUCo-, normalized for drug dose were compared across doses. Statistical analysis
Comparisons of the pharmacokinetic parameters across drug doses were performed by a one-way ANOVA using the Scheffk test for assessing differences between groups. Homogeneity of variance was tested by applying the Bartlett test. When necessary, a non-parametric statistical analysis, such as KruskallWallis or Mann-Whitney test, was used. Linear regression between AUC values and administered drug doses was carried out, and the intercept value of the regression was compared to zero by means of a non-paired t-test. In all cases a p < 0.05 was considered significant. RESULTS Mean carbimide plasma levels obtained after i.v. administration of 0.1, 0-3, 0.6, and 1 mg kg-' are shown in Figure 1. After the infusion, carbimide concentration displayed a biphasic disposition profile and a parallelism with the monoexponential terminal slopes. According to the MAICE test, the experimental data were best fitted to a two-compartment open model. The main pharmacokinetic parameters (mean f SD) obtained for each i.v. dose of carbimide are presented in Table 1. Mean distribution constant values ( a )ranged between 0.15 and 0.24 min-I. The mean 8-elimination half-life values
429
CARBIMIDE
ioooo
i
i'
0
I
60
120
id0 TIME
240
300
360
420
(mid
Figure 1. Mean carbimide plasma level-time curves after i.v. carbimide infusion of 0.1 mg kg-l (a), 0.3 m g kg-I (q, 0.6 mg kg-' (A), and 1 mg kg-l(+) to healthy male subjects
were very close and only fluctuated between 42 and 52 min. AUC values increased proportionally with the administered dose according to the following linear regression equation: AUC = 72 498.4 D - 3268.7 (r = 0.997,p < 0.01) The intercept of this regression line (3268.7 f 2367.6) was not statistically different from zero (t = 1-38, p > 0.05). Total plasma clearance values ranged from 14.4 to 20.5 ml kg-' min-I, and no statistical differences were found when compared across the administered doses. Apparent volume of distribution values in the central compartment and for steady-state conditions ranged from 221 to 296 and 668 to 800ml kg-', respectively. Comparisons of the main pharmacokinetic parameters across the doses of carbimide studied suggest that carbimide displays linear pharmacokinetics after i.v. administration. Carbimide plasma concentration-time profiles after oral administration of 0.3, 1, and 1.5mg kg-' carbimide are depicted in Figure 2, and the main pharmacokinetic parameters (mean f SD) are shown in Table 2. AUC values increased proportionally with the administered dose of carbimide according to the following linear regression equation: AUC = 60 425.2 D - 9010.8 (r = 0.998, p < 0.05) The intercept of this regression line (9010.8 rf: 4248.1) was not statistically different from zero (t = 2.12, p > 0-05). The oral absorption of carbimide was rapid with tmaxvalues ranging from 10.5 to 15.5 min. Peak plasma carbimide
282.0 f 101.65 854.3 f 194.9
1 48625 f 3290 902 f 362 10.5 f 4.2 76.5 f 13.3* 0.70 f 0.16
10279 f 5159 197 f 134 15.5 f 5.3 39.9 f 7.5 0.53 f 0.24
Dose (mg kg-') 0.3
* p < 0.05 with respect to 0.3mgkg-I (ANOVA followed by Scheffe's test).
Area under curve (AUCO-,) Plasma peak concentration (CmaX) Time to peak (tmax) Elimination half-life (tI1*) Bioavailability (F)
Parameter
238.59 f 73.02 667.7 f 143.1
221.4 f 78.02 800.3 f 118.8
0.1827 f 1218 0.0136 f 0.0024 52.3 f 8.53 37 077.0 f 9809.3 17.06 f 4.44
0.6
295.9 f 141.1 799.3 f 108.2
*
0.1496 f 0.0619 0.0137 f 0.0024 51.75 f 8.81 70902.0 f 11844 14.43 2.74
1 .O
Units ng m1-I min ngml-l min min -
83 254 f 26 144 1706 f 1040 14.6 f 4.6 61.5 f 2.8* -
ml kg-l ml kg-l
min-l min-I min ngml-l rnin ml kg-' min-l
Units
1.5
Table 2. Pharmacokinetic Parameters (mean f SD) of carbimide in man after oral administration
0.2093 f 0.1003 0.0173 f 0.0060 43.5 f 13.79 18 838.8 f 2292.4 16.09 f 1.77
f 0.0600 5 0.0045 f 8.95 f 1187.3 f 5.32
0.2433 0.0173 41.9 5104.2 20.5
Rapid disposition constant ( a ) Slow disposition constant (/3) Elimination half-life ( t , / * ) Area Under Curve (AUC,-,) Total plasma Clearance (C1,) Volume of distribution central compartment ( V,) Volume of distribution (Vd,,)
Dose (mg kg-I) 0.1
Parameter 0.3
Table 1. Pharmacokinetic parameters (mean f SD) of carbimide in man after intravenous infusion
?
h Y b
P
0
w
P
43 1
CARBIMIDE
concentration increased in accordance with the administered drug dose. After oral administration of 1 and 1.5mg kg-' of carbimide /3-elimination half-life values were 76.5 and 61.5 min, respectively. These values were statistically higher than those obtained after oral administration of 0.3 mg kg-' (39-9 min). Oral bioavailability after 0-3mg kg-l carbimide was not complete, being only 53% (see Figure 3). The corresponding estimated value after 1 mg/kg was 70%. No relevant clinical changes in blood pressure and heart rate were observed during the study, and no abnormal findings were found in the biochemical and haematological analysis. Furthermore, no adverse effects were observed after intravenous infusion or oral administration of carbimide doses.
DISCUSSION
The results of this study represent the first pharmacokinetic investigation of carbimide in man involving parenteral and oral single-dose administration. In spite of the fact that this drug has been used clinically for more than 50 years, the lack of a sensitive and reproducible analytical method for measuring carbimide plasma levels after oral doses can account for the absence of human pharmacokinetic data. Nevertheless, Shirota et d 6have demonstrated that hepatic biotransformation of carbimide gives N-acetylcarbimide as the main
-0
60
120
160
240
300
360
420
TIUE (mid Figure 2. Mean carbimide plasma level-time-curves after single oral carbimide administration of 0.3 mg kg-I 0 , l mg kg-I (+),and 1.5 mg kg-I (0)to healthy male subjects
432
R. OBACH ET A L .
100Or
10' 0
I
30
90
60
TIME
120
150
hinl
Figure 3. Comparative mean carbimide plasma level-time curves after intravenous infusion (0) and oral administration (.) of single doses of 0.3 mg kg-' carbimide
metabolite in different species including man. There is no previous report concerning the pharmacokinetics of carbimide for i.v. drug administration, as carbimide is usually taken orally in the treatment of alcoholism. Therefore, the first part of this study dealt with tolerability and clinical safety of carbimide for i.v. administration of increasing doses of the drug. Also, the i.v. route is needed to obtain an unambiguous estimation of the main pharmacokinetic parameters and to calculate absolute bioavailability of a drug. In this study, it has been observed that the open two-compartment model accurately predicts the carbimide plasma level-time course after i.v. administration (MAICE test, p < 0.05). The elimination half-life of carbimide in man was short, being less than 1 h for all the doses administered. This is in agreement with the elimination half-life data reported by Obach et a l l 3 for rats (mean value of 33 min) and for dogs (values ranged between 39 to 61 min). Plasma clearance values did not change for the range of doses studied (144-20.5 ml kg-' min-I), and were similar to the data reported for dogs (12.G19.7ml kg-' min-I), but lower than the values observed in rats (1 17 ml kg-I This difference from the plasma clearance values for rats could be explained by the higher apparent volume of distribution (Vd,,) in this species. For the apparent volume of distribution, there is a close correspondence between the value for man (around 0.8 1kg-l) and the value reported for dogs (around 1 1kg-I), and a clear difference with the data obtained for rats (3.8 1 kg-'. The carbimide erythrocytes concentration for rats is much higher
CARBIMIDE
433
than the intra-erythrocyte carbimide concentration observed for dogs or man.14 This fact could explain the higher apparent volume of distribution of carbimide for rats. The pharmacokinetics of carbimide in man within the range of intravenously administered doses (0.1-1 mg kg-I) were linear. No statistical differences (p > 0.05) were found when comparing pharmacokinetic parameters, such as total plasma clearance, elimination half-life, central and steady-state volume of distribution, and area under the curve normalized by the administered dose (AUC/D), indicating that they were dose-independent. Conversely, statistical differences (p < 0.05) were obtained when the elimination half-life values after oral administration were compared across the doses. This fact suggests nonlinear pharmacokinetics after oral administration. Dose-dependent pharmacokinetics have been described in rats and rabbits with higher doses of carbimide,I53I6and in dogs within a dose range of 1 4 m gkg-’. I 3 The dose-dependent pharmacokinetics could be related to saturation of the carbimide N-acetylation process. In fact, Shirota et have pointed out the existence of inter-species differences regarding the amount and/or activity of the N-acetyltransferase enzyme, which, in turn, could explain the possibility of saturation at different carbimide doses for each species. It remains to be determined whether this phenomenon exists in the human. The absorption rate of carbimide, evaluated by the time required to achieve peak plasma drug concentration, was rapid. The values obtained were in the range of those observed in other animal species such as dog and rat.13 The fast absorption of carbimide could be due to its low molecular weight and its structure, which allow simultaneous diffusion, through gut epithelial membranes and pores.I7 The oral bioavailability of carbimide was not complete. The extent of absorption was also incomplete when carbimide was administered in other animal species; bioavailability values of 69 per cent and 65 per cent have been reported Conversely, Deitrich for rats (2 mg kg-I) and dogs (4 mg kg-I), respe~tive1y.l~ et aZ.,’* using I4C-carbimide, demonstrated complete absorption of the drug after oral administration. All these findings suggest that a first-pass effect takes place in man after oral administration of carbimide. Based on calcium carbimide-ethanol interaction studies reported mainly by and the carbimide pharmacokinetic data Brien et a1.1-3 and Peachey et obtained in this study, there appears to be no relationship between the time courses of the increased blood acetaldehyde and the plasma carbimide concentration. This lack of correlation has recently been corroborated using the inhibition of human erythrocyte ALDH as an index of the pharmacological effect of carbimide.19 After oral administration of 1 mg kg-’ carbimide, inhibition of erythrocyte ALDH lasted 6 days, whereas carbimide was only measurable in plasma for 6 h. This study provides the first preliminary data on the pharmacokinetics profile of carbimide in man after intravenous and oral administrations. Further studies
434
R. OBACH ET AL.
should be required to elucidate the complex pharmacokinetic and pharmacodynamic relationships of carbimide in man.
ACKNOWLEDGEMENTS This work was partially financed with a grant from the FISS (number 87/1404). The authors wish to thank Professor F. Jane for kindly authorizing the use of the Clinical Pharmacology Unit's facilities to conduct this study.
REFERENCES 1. J. F. Brien, J. E. Peachey, B. J. Rogers and C. W. Loomis, Eur. J. Clin. Pharmacol., 14, 133 (1978). 2. J. F. Brien, J. E. Peachey, C. W. Loomis and B. J. Rogers, Clin. Pharmacol. Ther., 25, 454 (1979). 3. J. F. Brien, J. E. Peacheyand C. W. Loomis, Clin.Pharmacol. Ther., 27,426 (1980). 4. J. E. Peachey, J. F. Brien, C. W. Loomis and B. J. Rogers, Alcohol Clin.Exp. Res., 4, 322 (1 980). 5 . J. E. Peachey and C. A. Naranjo, Drugs, 27,171 (1984). 6. F. N. Shirota, H. T. Nagasawa, C. H. Kwon and E. G. De Master, Drug Metab. Disposit., 12,337 (1984). 7. J. Pruiionosa, R. Obach and J. M. Vallts, J. Chrumatogr., 377,253 (1986). 8. R. J. Francis, Comp. Prog. Biumed., 18,43 (1984). 9. H. Akaike, Math. Sci., 14, 156,s (1976). 10. J. G. Wagner, Fundamentals of Clinical Pharmacokinetics, Drug Intelligence Publications, Illinois, 1985. 11. R. C. Shumaker, Drug Metab. Rev., 17,341 (1986). 12. D. Perrier and M. Mayersohn, J. Pharm. Sci., 71,372 (1982) 13. R. Obach, H. Colom, J. Arso, C. Peraire and J. Pruiionosa, J. Pharm. Pharmacol., 41, 624 (1989). 14. R. Obach, J. Vallts, J. Pruiionosa, J. Arso and J. M. Vallts, Drogodependencias, un reto multidisciplinar, 1, 29-35. Ed. Servicio Central de Publicaciones del Gobiemo Vasco, Vitoria, Spain, 1984. 15. R. Obach, J. Pruiionosa, A. Menargues, M. Nomen, J. M. Valltsand C. Peraire, Rev. Farmacol. Clin.Exp. 3, 186 (1 986). Ed. Prous, Barcelona, Spain. 16. R. Obach, J. Moreno, J. Domenech and J. M. Pla-Delfina, Actas Ze' Congreso Europe0 de Biofarmacia y Farmacocine'tica,Technique et Documentation, Clermont Ferrand, 2,367 (1981), France. 17. J. M. Pla-Delfinaand J. Moreno, J. Pharmacokinet. Biopharm., 9,191 (1981). 18. R. A. Deitrich, P. A. Troxell, W. Worth and G. V. Erwin, Biochem. Pharmacol., 25, 2733 (1976). 19. R. Obach, J. Pruiionosa, J. Torrent, H. Colom, I. Izquierdo, and J. Domenech. Eur. J. Drug Metab, Special Issue, 4th European Congress of Biopharmaceutics and Pharmacokinetics, 1990, Abstract 167, p. 45. Medecine et Higiene Publishers, GenCva.
