BIOPHARMACEUTICS & DRUG DISPOSITION, VOL. 13, 41 1-416 (1992)
DISTRIBUTION OF CARBAMAZEPINE AND ITS EPOXIDE IN BLOOD COMPARTMENTS IN ADOLESCENT AND ADULT EPILEPTIC PATIENTS J. BONNETON, P. GENTON AND E. MESDJIAN
U 278 INSERM Faculte de Pharmacie, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 5, France and Centre Saint Paul, 300 Boulevard Sainte Marguerite, 13009 Murseille, France
ABSTRACT Carbamazepine (CBZ) used in the treatment of epilepsy, is metabolized in man to an active metabolite: carbamazepine 10-1 1 epoxide (CBZ-E). CBZ and CBZ-E concentrations in the different blood compartments (plasma ultrafiltrate, plasma proteins, and red blood cells (RBC)) of epileptic patients on CBZ monotherapy were assessed using HPLC. The highest CBZ and CBZ-E level to dose ratios were observed in the RBC. For CBZ and CBZ-E significant correlations were found between some blood compartments particularly between RBC and plasma ultrafiltrate. The measurement of CBZ and CBZ-E in all blood compartments may be interesting in uncontrolled patients and in patients presenting sideeffects which cannot be explained by total plasma concentration values. KEY WORDS
Carbamazepine Carbamazepine 10- 1 1 epoxide Red blood cells (RBC) Plasma proteins Plasma ultrafiltrate
INTRODUCTION Carbamazepine (CBZ) or 5-carbamoyl-5H-dibenzo (b,f) azepine, a tricyclic antiepileptic drug used in various seizure types, is metabolized by hepatic oxidation in humans to the principal metabolite 10,ll -epoxycarbamazepine (CBZ-E) which is changed by hepatic epoxide hydrolase into trans-l0,ll-dihydroxycarbamazepine. Total plasma concentrations of CBZ are often monitored for optimization and individualization of antiepileptic therapy. However, several factors can complicate the correlation between these concentrations and therapeutic effects. One of these factors is the presence of CBZ-E which possesses anticonvulsant and toxic properties similar to those of the parent drug,] and may contribute Correspondence to: Dr E. Mesdjian, U 278 INSERM Faculte de pharmacie, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 5, France.
0142-2782/92/06O411-06$5.OO 01992 by John Wiley & Sons, Ltd.
Received 6 June 1991 Revised 29 January 1992
412
J. BONNETON, P. GENTON AND E. MESDJIAN
to efficacy and side effects of CBZ treatment. Furthermore, it has been assumed that only the free drug in the blood is able to cross the blood-brain barrier and to reach the receptor sites. It has been suggested that the plasma protein-bound2 and red blood cells (RBC)-bound314anticonvulsant drugs are able to traverse capillary endothelia and to have access to the brain. Consequently, it is possible that the therapeutic effects of an antiepileptic drug are not well correlated with its free concentration in plasma ultrafiltrate, which, in turn, reflects imperfectly the drug disposition in the brain. This study was performed in epileptic patients on chronic CBZ monotherapy to determine the distribution of CBZ and its primary metabolite in the different blood compartments: plasma ultrafiltrate (protein free plasma), plasma proteins, and RBC. MATERIALS AND METHODS
Chemicals Carbamazepine, carbamazepine 10,ll-epoxide and 10-methoxycarbamazepine (CBZ-M) were supplied by Ciba-Geigy (Basel, Switzerland). Acetonitrile (HPLC grade) was purchased from BDH Chemicals Ltd (Poole, England), chloroform (RP NormapuP) from Prolabo (Paris, France), methanol from Carlo Erba (Milan, Italy), and phosphate buffer (pH= 7-00) from Merck (Darmstadt, FR Germany).
Liquid chromatography A Beckman System Gold Programmable Solvent Module 116 was used, fitted with an automatic programmable multidimensional injection system (Promis Spark, Holland). The detector was a Beckman 163 Variable Wavelength Detector, operated at 215 nm associated with an Integrator Stang ST 3A (Stang Instrument, France). The column was a Beckman Ultrasphere ODS ( 0 5 ~ ) 4.6 mm*250 mm. The mobile phase was acetonitrile-water (35:65); the flow rate was 1-8ml min-l.
Patients Eleven adolescent and adult epileptic patients aged from 16 to 76 years (mean+SD=37.5f21.4) on CBZ monotherapy (3.6 to 15-9mg kg-1 day-') undergoing treatment at the Centre Saint Paul (Marseille, France) were included in this study. All patients had symptomatic or cryptogenic partial epilep~ies.~ The patients received no slow release preparations of CBZ but conventional TegretoP 200mg tablets (Ciba Geigy France) and had been on stable CBZ therapy for at least 8 weeks. The patients' compliance and timing of drug intake
CARBAMAZEPINE DISTRIBUTION
41 3
were clinically assessed in all cases. Blood samples were obtained during the steady-state phase and after autoinduction could be assumed to have reached a stable state.
Blood sampling and separation According to the procedure used in the Centre Saint Paul, blood was sampled into heparin-rinsed syringes from patients 3 h after the morning oral dose. Immediately after the blood collection, plasma and RBC were separated by centrifugation (10 min at 5000 x g). The plasma was transferred into a glass tube; the residual plasma and the buffy coat were aspirated and discarded from the remaining RBC fraction. The RBC fraction was frozen at -20". Protein-free plasma fractions were obtained by ultrafiltration using Centriflo ultrafiltration cones (Amicon, Danvers, MA, USA), as described previously by Cramer et aL6s7 and by Hooper et aL8
Extraction procedure and HPLC analysis Concentrations of analyte in RBC were estimated by direct measurements on the RBC fractionsgJOrather than by calculation from data on total blood, plasma, and the hematocrit or other indirect approaches. A correction factor (2 per cent) for entrapped plasma was a ~ p l i e dAfter . ~ hemolysis by freezing and thawing, the analytes (CBZ and CBZ-E) were extracted directly from the unwashed concentrated RBC fraction to avoid the loss of drug ligated to the RBC by w a ~ h i n g . ~The J ~ drug and metabolite concentrations were determined using a HPLC method described by Elyas et al.I2 Before doing any analyses, all the methodology, especially preparation of the RBC and protein-free plasma fractions and the recovery, linearity, and sensitivity of the HPLC method, were checked. Briefly, internal standard (CBZM), phosphate buffer (0*3M), and chloroform were added to plasma or plasma ultrafiltrate or RBC contained in a glass tube. The mixture was stirred and centrifuged; the aqueous upper layer was discarded and the organic phase was transferred to a conical tube and evaporated to dryness in a stream of nitrogen at 40". The residue was dissolved in acetonitrile and analysed by HPLC. Standard curves for the determination of CBZ and its epoxide in total plasma, plasma ultrafiltrate, and RBC were prepared by treating blank blood samples in the same way as blood samples from patients and by adding known amounts of CBZ and CBZ-E to the different blood compartments. To exclude day-today variability, the standards and patient samples were measured in the same series.
Analysis of data All data are expressed as the mean and standard deviation in Tables 1 and 2. Non-parametric statistics were used for statistical evaluation: the Spearman
414
J. BONNETON, P. GENTON AND E. MESDJIAN
Table 1. Levels (L) (pg ml-I), level to dose ratios (LDR) (pg ml-I kg-' day-') of carbamazepine (CBZ) and CBZ-epoxide (CBZ-E), and CBZ-E/CBZ ratios in different blood compartments in epileptic patients on CBZ monotherapy Total plasma
Plasma ultrafiltrate 1.43f0.60 0.151f0.066
CBZ
L. LDR
10-31k3.20 1-07kO-25
CBZ-E
LDRL
1-55k1.24 0 14 k 0 -08
CBZ-E/CBZ
0.14k 0.11
-
Plasma proteins
RBC
8.88k2.74 10-90+3*57 0.919kO-196 1.13kO-30
0.49k0.39 1.07k0.99 2*09+1-27 0.044 k 0 -024 0.097 k 0.068 0 * 197 k 0 082 0.31 k0. 18
0.12k0.11
0.18 k0.08
Means (k SD). Table 2. Ratios of plasma ultrafiltrate, plasma proteins, and RBC to total plasma CBZ and CBZ-E concentrations in epileptic patients on CBZ monotheraDy
CBZ CBZ-E
Plasma ultrafiltrate/ total plasma
Plasma proteins/ RBC/ total plasma total plasma
0.14k0.03 0*34+0*02
0.86k0.03 0.67kO.13
1*06+0.21 1.53-FO.45
Means (k SD) rank correlation coefficients were calculated using the BMDP 3s program. A level of p < 0.05 was considered statistically significant.
RESULTS The data concerning the plasma protein-bound drugs, were obtained by subtracting the plasma ultrafiltrate levels from the total plasma concentrations. The results are expressed in pg ml-I, as the ratio of the level of bound CBZ or CBZ-E to CBZ dose (LDR) (pg ml-I.mg kg day-'), and as the CBZ-EKBZ ratio (Table 1). The highest LDRs for CBZ and CBZ-E were found in the RBC; LDRs for the total plasma and plasma proteins were lower and the lowest LDR values were obtained for the plasma ultrafiltrate. In RBC, the CBZ-EKBZ ratio is higher than in total plasma or plasma proteins, but lower than in the plasma ultrafiltrate. The ratios of analytes in plasma ultrafiltrate, plasma proteins and RBC to total plasma CBZ and CBZ-E concentrations are shown in Table 2. The correlations between CBZ and CBZ-E levels in the various blood compartments were calculated. We found significant correlations between the LDRs for the CBZ or the CBZ-E in (1) the plasma ultrafiltrate and total plasma (p
DISTRIBUTION OF CARBAMAZEPINE AND ITS EPOXIDE IN BLOOD COMPARTMENTS IN ADOLESCENT AND ADULT EPILEPTIC PATIENTS J. BONNETON, P. GENTON AND E. MESDJIAN
U 278 INSERM Faculte de Pharmacie, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 5, France and Centre Saint Paul, 300 Boulevard Sainte Marguerite, 13009 Murseille, France
ABSTRACT Carbamazepine (CBZ) used in the treatment of epilepsy, is metabolized in man to an active metabolite: carbamazepine 10-1 1 epoxide (CBZ-E). CBZ and CBZ-E concentrations in the different blood compartments (plasma ultrafiltrate, plasma proteins, and red blood cells (RBC)) of epileptic patients on CBZ monotherapy were assessed using HPLC. The highest CBZ and CBZ-E level to dose ratios were observed in the RBC. For CBZ and CBZ-E significant correlations were found between some blood compartments particularly between RBC and plasma ultrafiltrate. The measurement of CBZ and CBZ-E in all blood compartments may be interesting in uncontrolled patients and in patients presenting sideeffects which cannot be explained by total plasma concentration values. KEY WORDS
Carbamazepine Carbamazepine 10- 1 1 epoxide Red blood cells (RBC) Plasma proteins Plasma ultrafiltrate
INTRODUCTION Carbamazepine (CBZ) or 5-carbamoyl-5H-dibenzo (b,f) azepine, a tricyclic antiepileptic drug used in various seizure types, is metabolized by hepatic oxidation in humans to the principal metabolite 10,ll -epoxycarbamazepine (CBZ-E) which is changed by hepatic epoxide hydrolase into trans-l0,ll-dihydroxycarbamazepine. Total plasma concentrations of CBZ are often monitored for optimization and individualization of antiepileptic therapy. However, several factors can complicate the correlation between these concentrations and therapeutic effects. One of these factors is the presence of CBZ-E which possesses anticonvulsant and toxic properties similar to those of the parent drug,] and may contribute Correspondence to: Dr E. Mesdjian, U 278 INSERM Faculte de pharmacie, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 5, France.
0142-2782/92/06O411-06$5.OO 01992 by John Wiley & Sons, Ltd.
Received 6 June 1991 Revised 29 January 1992
412
J. BONNETON, P. GENTON AND E. MESDJIAN
to efficacy and side effects of CBZ treatment. Furthermore, it has been assumed that only the free drug in the blood is able to cross the blood-brain barrier and to reach the receptor sites. It has been suggested that the plasma protein-bound2 and red blood cells (RBC)-bound314anticonvulsant drugs are able to traverse capillary endothelia and to have access to the brain. Consequently, it is possible that the therapeutic effects of an antiepileptic drug are not well correlated with its free concentration in plasma ultrafiltrate, which, in turn, reflects imperfectly the drug disposition in the brain. This study was performed in epileptic patients on chronic CBZ monotherapy to determine the distribution of CBZ and its primary metabolite in the different blood compartments: plasma ultrafiltrate (protein free plasma), plasma proteins, and RBC. MATERIALS AND METHODS
Chemicals Carbamazepine, carbamazepine 10,ll-epoxide and 10-methoxycarbamazepine (CBZ-M) were supplied by Ciba-Geigy (Basel, Switzerland). Acetonitrile (HPLC grade) was purchased from BDH Chemicals Ltd (Poole, England), chloroform (RP NormapuP) from Prolabo (Paris, France), methanol from Carlo Erba (Milan, Italy), and phosphate buffer (pH= 7-00) from Merck (Darmstadt, FR Germany).
Liquid chromatography A Beckman System Gold Programmable Solvent Module 116 was used, fitted with an automatic programmable multidimensional injection system (Promis Spark, Holland). The detector was a Beckman 163 Variable Wavelength Detector, operated at 215 nm associated with an Integrator Stang ST 3A (Stang Instrument, France). The column was a Beckman Ultrasphere ODS ( 0 5 ~ ) 4.6 mm*250 mm. The mobile phase was acetonitrile-water (35:65); the flow rate was 1-8ml min-l.
Patients Eleven adolescent and adult epileptic patients aged from 16 to 76 years (mean+SD=37.5f21.4) on CBZ monotherapy (3.6 to 15-9mg kg-1 day-') undergoing treatment at the Centre Saint Paul (Marseille, France) were included in this study. All patients had symptomatic or cryptogenic partial epilep~ies.~ The patients received no slow release preparations of CBZ but conventional TegretoP 200mg tablets (Ciba Geigy France) and had been on stable CBZ therapy for at least 8 weeks. The patients' compliance and timing of drug intake
CARBAMAZEPINE DISTRIBUTION
41 3
were clinically assessed in all cases. Blood samples were obtained during the steady-state phase and after autoinduction could be assumed to have reached a stable state.
Blood sampling and separation According to the procedure used in the Centre Saint Paul, blood was sampled into heparin-rinsed syringes from patients 3 h after the morning oral dose. Immediately after the blood collection, plasma and RBC were separated by centrifugation (10 min at 5000 x g). The plasma was transferred into a glass tube; the residual plasma and the buffy coat were aspirated and discarded from the remaining RBC fraction. The RBC fraction was frozen at -20". Protein-free plasma fractions were obtained by ultrafiltration using Centriflo ultrafiltration cones (Amicon, Danvers, MA, USA), as described previously by Cramer et aL6s7 and by Hooper et aL8
Extraction procedure and HPLC analysis Concentrations of analyte in RBC were estimated by direct measurements on the RBC fractionsgJOrather than by calculation from data on total blood, plasma, and the hematocrit or other indirect approaches. A correction factor (2 per cent) for entrapped plasma was a ~ p l i e dAfter . ~ hemolysis by freezing and thawing, the analytes (CBZ and CBZ-E) were extracted directly from the unwashed concentrated RBC fraction to avoid the loss of drug ligated to the RBC by w a ~ h i n g . ~The J ~ drug and metabolite concentrations were determined using a HPLC method described by Elyas et al.I2 Before doing any analyses, all the methodology, especially preparation of the RBC and protein-free plasma fractions and the recovery, linearity, and sensitivity of the HPLC method, were checked. Briefly, internal standard (CBZM), phosphate buffer (0*3M), and chloroform were added to plasma or plasma ultrafiltrate or RBC contained in a glass tube. The mixture was stirred and centrifuged; the aqueous upper layer was discarded and the organic phase was transferred to a conical tube and evaporated to dryness in a stream of nitrogen at 40". The residue was dissolved in acetonitrile and analysed by HPLC. Standard curves for the determination of CBZ and its epoxide in total plasma, plasma ultrafiltrate, and RBC were prepared by treating blank blood samples in the same way as blood samples from patients and by adding known amounts of CBZ and CBZ-E to the different blood compartments. To exclude day-today variability, the standards and patient samples were measured in the same series.
Analysis of data All data are expressed as the mean and standard deviation in Tables 1 and 2. Non-parametric statistics were used for statistical evaluation: the Spearman
414
J. BONNETON, P. GENTON AND E. MESDJIAN
Table 1. Levels (L) (pg ml-I), level to dose ratios (LDR) (pg ml-I kg-' day-') of carbamazepine (CBZ) and CBZ-epoxide (CBZ-E), and CBZ-E/CBZ ratios in different blood compartments in epileptic patients on CBZ monotherapy Total plasma
Plasma ultrafiltrate 1.43f0.60 0.151f0.066
CBZ
L. LDR
10-31k3.20 1-07kO-25
CBZ-E
LDRL
1-55k1.24 0 14 k 0 -08
CBZ-E/CBZ
0.14k 0.11
-
Plasma proteins
RBC
8.88k2.74 10-90+3*57 0.919kO-196 1.13kO-30
0.49k0.39 1.07k0.99 2*09+1-27 0.044 k 0 -024 0.097 k 0.068 0 * 197 k 0 082 0.31 k0. 18
0.12k0.11
0.18 k0.08
Means (k SD). Table 2. Ratios of plasma ultrafiltrate, plasma proteins, and RBC to total plasma CBZ and CBZ-E concentrations in epileptic patients on CBZ monotheraDy
CBZ CBZ-E
Plasma ultrafiltrate/ total plasma
Plasma proteins/ RBC/ total plasma total plasma
0.14k0.03 0*34+0*02
0.86k0.03 0.67kO.13
1*06+0.21 1.53-FO.45
Means (k SD) rank correlation coefficients were calculated using the BMDP 3s program. A level of p < 0.05 was considered statistically significant.
RESULTS The data concerning the plasma protein-bound drugs, were obtained by subtracting the plasma ultrafiltrate levels from the total plasma concentrations. The results are expressed in pg ml-I, as the ratio of the level of bound CBZ or CBZ-E to CBZ dose (LDR) (pg ml-I.mg kg day-'), and as the CBZ-EKBZ ratio (Table 1). The highest LDRs for CBZ and CBZ-E were found in the RBC; LDRs for the total plasma and plasma proteins were lower and the lowest LDR values were obtained for the plasma ultrafiltrate. In RBC, the CBZ-EKBZ ratio is higher than in total plasma or plasma proteins, but lower than in the plasma ultrafiltrate. The ratios of analytes in plasma ultrafiltrate, plasma proteins and RBC to total plasma CBZ and CBZ-E concentrations are shown in Table 2. The correlations between CBZ and CBZ-E levels in the various blood compartments were calculated. We found significant correlations between the LDRs for the CBZ or the CBZ-E in (1) the plasma ultrafiltrate and total plasma (p
