Edoardo Spina MD, PhD
Associate Professor, Section of Pharmacology, Department of Clinical and Experimental Medicine and Pharmacology, University of Messina, Policlinico Universitario, Messina, Italy
Carbamazepine (CBZ) is an iminodibenzyl derivative with anticonvulsant properties that is structurally related to the tricyclic antidepressants. It is one of the most commonly prescribed drugs for epilepsy, but it also is used in the treatment of other neurologic conditions, such as chronic pain syndromes and trigeminal neuralgia, as well as in a variety of psychiatric disorders (1,2).
CHEMISTRY AND METABOLIC SCHEME
The chemical name of CBZ is 5H-dibenz[b,f]azepine-5-carboxamide or 5-carbamoil-5H-dibenz[b,f]azepine. It is a white or yellowish-white, crystalline, almost odorless powder, tasteless or with a slightly bitter taste. CBZ has a molecular weight of 236.3 and a melting point between 189°C and 193°C. Unlike tricyclic antidepressants, which are basic substances, CBZ is a neutral compound because its carbamoyl side chain is part of a nonionized urea moiety (3). CBZ is a highly lipophilic agent, as indicated by a partition coefficient of (P) 58 in the system n-octanol/aqueous buffer pH = 7.4, which is soluble in various organic solvents, such as alcohol and acetone, but practically insoluble in water. CBZ is prepared by chemical synthesis and exists in several crystal modifications, with the α and β forms being of primary interest. Solid dosage formulations usually contain the β-modification, whereas in aqueous syrup formulations CBZ is present as the dihydrate. Crystalline CBZ is chemically stable under normal storage conditions at room temperature.
CBZ is extensively metabolized in the body, and several metabolites are formed by parallel or sequential reactions catalyzed by different enzymes (Figure 21.1). The pharmacokinetic parameters of CBZ are summarized in Table 21.1.
ABSORPTION
Absorption of CBZ from the gastrointestinal tract is rather slow and extremely variable (4). Single-dose studies have shown that peak plasma CBZ concentrations usually occur between 4 and 8 hours after oral administration of regular tablets, but peaks as late as 24 hours have been reported, depending on the formulation (4, 5, 6, 7, 8, 9). After reaching a maximum, plasma concentrations typically plateau for 10 to 30 hours before declining (5,7). Peak plasma CBZ concentrations are reached more rapidly (within an average of 3 hours) in patients on long-term treatment owing to the autoinduction process (see section on Biotransformation) (4). The delayed and irregular absorption of CBZ from conventional tablets probably is related to its very slow dissolution in the gastrointestinal fluid or to its anticholinergic properties, which may modify the gastrointestinal transit time (4,5). Earlier kinetic studies found evidence of dose-independent absorption of CBZ in the range of 50 to 600 mg (5,7,8). More recent investigations, however, indicate that the rate and extent of CBZ absorption appear to be dose dependent and that the time to reach maximal concentration is prolonged when the dose is increased (4,10). It has been suggested that CBZ undergoes a simultaneous first-order and zero-order absorption, with approximately 35% of the available dose absorbed at a zero-order rate (11). Oral solutions of CBZ are absorbed more rapidly and produce higher peaks than tablets (12,13). On the other hand, controlled-release formulations ensure a smoother absorption profile and allow less frequent administration (14).
Bioavailability
Because of the lack of an injectable formulation, the absolute bioavailability of CBZ in humans has not been determined. However, based on the recovery of radiolabeled CBZ in urine and feces after single-dose administration of 14C-CBZ in a gelatin capsule, the oral bioavailability has been estimated to range from 75% to 85% (15). In a pharmacokinetic study with a parenteral formulation for experimental use, comparison of the oral and intravenous routes of CBZ administration in two healthy subjects showed complete bioavailability of the oral dose (16). Data from different studies indicate that the oral bioavailability of CBZ is similar whether given as conventional tablets, solutions, suspensions, syrups, or newly developed chewable or sustained-release formulations (12, 13, 14,17, 18, 19, 20, 21). The relative bioavailability of a rectally administered suspension of CBZ was found to be similar to that of an orally administered tablet (22,23). However, in a study of children with epilepsy, the relative bioavailability of a suppository for rectal administration was found to be 80% compared with slow-release tablets (24). The bioavailability of CBZ was slightly, but not significantly, increased when tablets were taken with meals, probably because of enhanced solubilization of the drug by the bile secreted after food ingestion (5).
P.237
|
FIGURE 21.1. Biotransformation of carbamazepine in humans, and the enzymes catalyzing the major metabolic reactions. A, carbamazepine; B, carbamazepine-10,11-epoxide; C, trans-10,11-dihydroxy-10,11-dihydrocarbamazepine; D, 9-hydroxymethyl-10-carbamoyl acridan; E, 1-, 2-, 3-, and 4-hydroxycarbamazepine; F, carbamazepine N-glucuronide; CYP, cytochrome P450; mEH, microsomal epoxide hydrolase; UDPGT, uridine diphosphate glucuronosyltransferase. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
TABLE 21.1. PHARMACOKINETIC PARAMETERS OF CARBAMAZEPINE |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
P.238
Formulations and Routes of Administration
For human use, CBZ is commercially available for oral and rectal administration (1). Oral preparations consist of various solid dosage forms, including controlled-release formulations, and a syrup. The marketed formulations include tablets of 100, 200, or 400 mg, chewable tablets of 100 or 200 mg, controlled-release tablets of 100, 200, or 400 mg, and a suspension of 100 mg/5 mL. Solid-dose preparations with modified-release characteristics have been developed to reduce peak-related toxic effects and to decrease clinically troublesome fluctuations in plasma CBZ concentrations during the dosage interval (14). In addition, sustained-release formulations may be administered only once or twice daily, compared with three or four daily doses of conventional formulations, thus improving patient compliance. CBZ also is supplied as suppositories of 125 or 250 mg for rectal administration. Because of the low solubility of CBZ in water, no parenteral preparation has been marketed, although an injectable formulation has been used in a pharmacokinetic study in humans (16) and in animal models (25).
The major metabolite of CBZ, carbamazepine-10,11-epoxide (CBZ epoxide), has been directly administered orally to humans in the form of a suspension, a solution, and an enteric-coated tablet (26, 27, 28). Because it is unstable in an acid environment, it is advisable to coadminister the suspension or solution formulations with antacids.
DISTRIBUTION
CBZ is a neutral and fairly lipophilic compound that easily crosses the blood-brain barrier and other biologic membranes of the body and rapidly distributes to various organs and tissues, without any preferential affinity for specific regions. After administration of single oral doses to healthy volunteers and patients, the apparent volume of distribution of CBZ has been found to range between 0.79 and 1.86 L/kg (6,7). These values have been calculated assuming complete bioavailability of the drug, and the real volumes therefore might be slightly lower and less variable. The apparent volume of distribution of CBZ epoxide, calculated after direct administration to healthy subjects, was found to range between 0.59 and 1.57 L/kg (26,28).
Plasma Protein Binding
Both CBZ and its major metabolite are bound to plasma proteins, primarily to albumin and, to a lesser extent, to α1-acid glycoprotein. The plasma protein binding of CBZ ranges between 70% and 80%, as determined in vitro by equilibrium dialysis and ultrafiltration and in vivo by estimating the concentration of the drug simultaneously in plasma and in cerebrospinal fluid or saliva (7,29, 30, 31). The protein binding is independent of total plasma concentrations over the therapeutic range, but may be reduced at supratherapeutic levels (29). Plasma protein binding of CBZ epoxide has been reported to range between 50% and 60%, and the epoxide metabolite seems not to bind to α1-acid glycoprotein (31,32).
Protein binding of CBZ appears to be slightly lower in newborns and children than in adults, with a free fraction of 30% to 35% for parent drug and 55% to 60% for the epoxide metabolite (33, 34, 35, 36). One study has reported an elevation of free CBZ concentrations in plasma of elderly patients, probably due to an age-related decrease in levels of nonglycated albumin, which appears to be the major ligand of CBZ in serum (37). No significant modifications in CBZ protein binding have been observed during pregnancy (34,38). However, the unbound fractions of CBZ and CBZ epoxide have been reported to increase slightly throughout pregnancy from 21% to 27% for parent drug and from 42% to 52% for the metabolite (39). Patients with hepatic diseases were found to have slightly lower plasma protein binding compared with normal subjects, whereas there were no significant differences in binding capacity between patients with renal disease and healthy individuals (29). The free CBZ fraction was reported to be moderately decreased in patients with disease states associated with an increased α1-acid glycoprotein concentration, such as inflammation, myocardial infarction, cancer, and trauma (40). Unlike phenytoin and valproic acid, the plasma protein binding of CBZ shows very little interindividual variation, suggesting that there is no need to monitor free rather than total plasma concentrations (41).
Cerebrospinal Fluid, Brain, Saliva, and Other Tissues
Both CBZ and CBZ epoxide readily pass into the central nervous system, and their concentrations in cerebrospinal fluid reflect the free fraction of the drug. The cerebrospinal
P.239
fluid concentrations of CBZ have been reported to range from 17% to 31% of the total plasma concentrations, whereas those of CBZ epoxide were found to be approximately 45% to 55% of the corresponding total plasma levels (30,42,43). In patients undergoing brain surgery for tumor removal, the ratio between brain and plasma concentrations ranged from 0.8 to 1.6 for CBZ and from 0.5 to 1.5 for CBZ epoxide (44,45). The postmortem concentrations of CBZ and its epoxide metabolite in frontal cortex were approximately 1.4 and 1.1 times higher, respectively, than those simultaneously present in serum (46).
Salivary concentrations of CBZ and CBZ epoxide in humans are similar to the unbound concentrations in plasma, and have been reported to range from 20% to 30% of plasma concentrations for parent drug and from 30% and 40% for the metabolite (31,47,48). Determination of salivary CBZ concentrations may represent a useful and easy tool for measuring unbound drug (47). CBZ concentrations in tears also reflect the free fraction of the drug in plasma (49). On the other hand, CBZ appears to have a limited penetration into red blood cells, as indicated by erythrocyte-to-plasma ratios of 0.14 to 0.38 (29,50).
Transplacental Passage
CBZ penetrates the placenta extensively and rapidly and distributes to different tissues and organs of the fetus homogeneously (51). Fetal plasma concentrations of CBZ, determined in human umbilical cord, were found to range between 50% and 80% of maternal levels (51, 52, 53). There is in vitro evidence that human fetal liver during weeks 15 to 21 of gestation is able to metabolize CBZ to its 10,11-epoxide (54). Accordingly, CBZ epoxide also has been detected in fetal tissues and amniotic fluid (51). Concentrations of CBZ and its epoxide metabolite in amniotic fluid were found to be 2 to 2.5 times higher than corresponding free concentrations in maternal serum (55). Such a difference might be related to a higher presence of maternal proteins.
Breast Milk
Both CBZ and the epoxide metabolite are transferred to breast milk (51). Concentrations of CBZ in breast milk have been reported to be approximately 30% to 40% of those in maternal plasma, whereas the corresponding value for CBZ epoxide was approximately 50% (56,57). The possible daily amount of CBZ transferred to the newborn during breast-feeding has been estimated to range between 2 and 5 mg (57). The effect of such a dose on the newborn has yet to be evaluated.
ROUTES OF ELIMINATION
CBZ is eliminated by biotransformation followed by urinary and biliary excretion of the parent drug and the formed metabolites. After administration of a single oral dose of 14C-labeled CBZ, 72% of the radioactivity was excreted in the urine, and the remaining 28% was recovered in feces (15).
Biotransformation
CBZ is extensively metabolized in the body, with less than 2% of an oral dose excreted unchanged in urine (3,15). Biotransformation of CBZ appears to occur mainly in the liver, although animal studies suggest that the lung also contributes, approximately by 5%, to the total body clearance of the drug (58). Several metabolites are formed by parallel or consecutive reactions, as indicated in Figure 21.1. The major pathways of CBZ biotransformation include the epoxide-diol pathway, aromatic hydroxylation, and conjugation reactions. It has been estimated that metabolites from these three major routes account for 80% to 90% of total urinary radioactivity (3,15)
The epoxide-diol pathway is quantitatively the most important in CBZ biotransformation and it consists of an initial oxidation of the 10,11 double bond of the sevenmembered azepine ring followed by a hydrolysis reaction. CBZ is first oxidized to the chemically stable CBZ epoxide, which is pharmacologically active. Except for a small amount excreted as such in urine (~1% of the dose), this primary metabolite is extensively hydrolyzed (hydrated) to trans-10,11-dihydroxy-10,11-dihydrocarbamazepine (trans-CBZ-diol). The diol, which is present in human plasma mainly as its S,S-enantiomer (59), is then excreted into the urine, where it accounts for approximately 35% of a carbamazepine dose, partly as unconjugated and partly as its mono-O-glucuronide (60). A minor metabolite of the epoxide-diol pathway is 9-hydroxymethyl-10-carbamoyl acridan (9-HM-10-CA), which is almost completely conjugated with glucuronic acid before excretion in urine (61). However, this metabolite appears to be formed mainly from CBZ itself and only in small part through the epoxide-diol pathway (62).
An additional, but quantitatively less important, pathway of CBZ metabolism is represented by hydroxylation at different positions of the six-membered aromatic rings, with the formation of four possible phenolic products, 1-, 2-, 3-, and 4-hydroxycarbamazepine (3). Animal studies indicate that reactive arene oxides are intermediates of these reactions (63). Two other metabolites of this pathway carry a hydroxyl group in position 2 and a methoxy group in position 1 or 3. These phenolic metabolites are then conjugated with glucuronic acid and, to a lesser extent, with sulfuric acid and excreted in urine. Only traces of these metabolites are excreted unconjugated.
Conjugation reactions usually are regarded as the third most important route of CBZ biotransformation (3). The drug may be directly conjugated with glucuronic acid. Glucuronidation occurs at the amino group of the carbamoyl
P.240
side chain. Unlike most other glucuronides, this N-glucuronide conjugate cannot be hydrolyzed with β-glucuronidase. Glucuronidation also occurs as a secondary metabolic process, and almost all CBZ metabolites carrying free hydroxyl groups are converted to their O-glucuronides. These conjugated metabolites are susceptible to cleavage with β-glucuronidase. As an additional secondary conjugation process, CBZ and its phenolic metabolites can be conjugated with sulfuric acid.
With regard to the pharmacologic activity of CBZ metabolites, it is well established that CBZ epoxide has anticonvulsant effects in animal models of epilepsy and might therefore contribute to the clinical effects of CBZ in humans (64). Conversely, trans-CBZ-diol is devoid of anticonvulsant activity, whereas 9-HM-10-CA and 2- and 3-hydroxy-CBZ have shown only little or no activity in animal tests (3).
It is well documented that CBZ induces its own metabolism during long-term therapy (autoinduction) (6,62,65). Combination therapy with other anticonvulsants (e.g., phenytoin or phenobarbital) further induces CBZ metabolism (heteroinduction) (62,66,67). The autoinduction process involves the epoxide-diol pathway, and it has been demonstrated that both the CBZ epoxidation and the subsequent epoxide hydrolysis are induced, although the latter reaction to a lesser extent (60,62). There is conflicting evidence concerning the susceptibility to autoinduction of the formation of 9-HM-10-CA and aromatic hydroxylation (62,68). On the other hand, glucuronidation of diol, phenolic, or acridan metabolites appears not to be autoinducible (68). The course of the autoinduction process appears to be complex, discontinuous, and prolonged (69). There is evidence that it may start within 24 hours of first exposure to CBZ (70) and seems to be complete during the first 3 to 5 weeks of treatment (71). According to further studies (72,73), the process of autoinduction is complete within 1 week of starting CBZ therapy or dose change, and it appears to be dose dependent.
There has been a recent increase in knowledge of the specific enzymes involved in the major pathways of CBZ biotransformation, particularly cytochrome P450 (CYP) isoenzymes. Studies with purified or cloned CYP enzymes have demonstrated that the epoxidation of CBZ is mediated by CYP3A4 and CYP2C8, with CYP3A4 playing the most important role (74). These isoforms are induced by repeated administration of CBZ (75,76) and by phenobarbital (77). The subsequent hydrolysis of the epoxide to form trans-diol is catalyzed by a microsomal epoxide hydrolase, which is responsible for the inactivation of epoxides derived from the oxidative metabolism of xenobiotics (78). This enzyme also may be induced by repeated doses of CBZ (60,62,75) and by phenobarbital (79), although to a lesser extent than the CYP isoforms responsible for the formation of CBZ epoxide. The CYP1A2 isoform appears to be the major enzyme responsible for the aromatic hydroxylation of CBZ (80). Direct N-glucuronidation of CBZ as well as glucuronidation of its metabolites are catalyzed by microsomal uridine diphosphate glucuronosyltransferase (UDPGT) (3). The specific UDPGT isoform responsible for these reactions is yet unidentified.
With regard to the genetic aspects of CBZ metabolism, of the different CYP isoenzymes involved in its oxidative reactions—namely, CYP1A2, CYP2C8, and CYP3A4—only the CYP2C8 isoform is polymorphically expressed (81). Wide interindividual variability has been reported in the expression of CYP1A2 and CYP3A4 in humans, but to date no specific polymorphisms have been identified (3). Microsomal epoxide hydrolase is universally expressed, with no evidence for enzyme-deficient phenotypes (82). Polymorphisms have been described for the UDPGT variant catalyzing glucuronidation of bilirubin, but it is not known if it is involved in CBZ glucuronidation (3).
Biliary and Renal Excretion
As previously stated, approximately 28% of an oral dose of 14C-labeled CBZ was found in the feces, suggesting both incomplete absorption and a nonnegligible biliary excretion of unidentified metabolites (15). In this respect, only 1% of a single 400-mg oral dose of CBZ was eliminated by bile within 72 hours, indicating no significant enterohepatic circulation (83). However, these results do not exclude the possibility of a more consistent biliary excretion of conjugated metabolites.
CBZ is eliminated in the urine mainly as metabolites, and elimination is faster in patients receiving other antiepileptic drugs. In epileptic patients stabilized on CBZ treatment, 20% to 60% of the daily dose is excreted as the trans-CBZ-diol, 5% to 11% as 9-HM-10-CA, 5% to 10% as phenolic metabolites, 1% to 2% as CBZ epoxide, and 0.5% as unchanged CBZ (60,62,84).
CLEARANCE AND HALF-LIFE
The elimination of CBZ is well described by a one-compartment open model and follows an apparent first-order process, as demonstrated by the log-linear terminal phase of the typical plasma concentration-time curve after a single dose to a normal subject (6,8). However, because of the autoinduction process, the plasma elimination of CBZ is considerably more rapid during maintenance therapy than after single-dose administration to healthy volunteers, leading to a condition defined as time-dependent kinetics (69,85). In this situation, clearance values increase and half-lives decrease with time, with the result that steady-state CBZ levels are lower than those predicted from single-dose kinetic studies. As a consequence, higher doses are needed to maintain the same plasma concentrations. Therefore, the values of the elimination parameters of CBZ depend on whether the drug has been studied
P.241
after administration of its first dose or after there has been time for autoinduction to occur. The apparent plasma half-life and total-body clearance usually exhibit large interindividual differences, not only in relation to various pathophysiologic factors and coadministration of other drugs, but because the autoinduction process reaches different levels in different individuals.
Healthy Subjects
In healthy subjects, after administration of a single dose of CBZ, the plasma half-life has been reported to range from 18 to 55 hours and the corresponding clearance values from 0.013 to 0.061 L/hr/kg or from 1.14 to 1.80 L/hr (5, 6, 7). In a study of epileptic patients, the plasma half-life of CBZ declined from 35.6 ± 15.3 hours after the initial single dose to 20.9 ± 5.0 hours after repeated administration (6).
The renal clearance values for CBZ epoxide and trans-CBZ-diol in epileptic patients are (mean ± standard deviation) 9.7 ± 3.9 and 70.9 ± 13.9 mL/min, respectively (60). After massive CBZ overdosage, renal clearances of CBZ and CBZ epoxide were low (1 and 8 mL/min, respectively) and flow dependent, whereas renal clearance of trans-CBZ-diol was approximately 160 to 350 mL/min and independent of the urine flow (86).
After direct administration of CBZ epoxide to healthy volunteers, the half-life of the epoxide metabolite was found to range from 5 to 11 hours, and corresponding clearance values ranged from 0.063 to 0.136 L/hr/kg (26,28).
Comedicated Epileptic Patients
The elimination of CBZ is increased in patients receiving enzyme-inducing anticonvulsants (62,66,67). Eichelbaum et al. (85) studied CBZ kinetics in three groups of subjects: (a) healthy volunteers, (b) epileptic patients on CBZ monotherapy, and (c) epileptic patients comedicated with other anticonvulsants. The plasma clearances were 0.020 ± 0.003, 0.055 ± 0.007 and 0.113 ± 0.033 L/hr/kg in the three groups, respectively, and the corresponding half-life values were 26.2 ± 6.1, 12.3 ± 0.8, and 8.2 ± 3.3 hours. The increased clearance in the patients was due mainly to an induction of the epoxide-diol pathway, as reflected by an increased urinary excretion of the trans-CBZ-diol metabolite. The effect of other drugs on CBZ elimination is discussed in Chapter 23.
Children
The clearance of CBZ appears to be age dependent, with the highest disposition rate during infancy and early childhood (87,88). A within-subject study, conducted over several years, indicated that the major modifications in clearance occur between 9 and 13 years of age and that adult values are reached at 15 to 17 years of age (89). The apparent plasma half-life of CBZ is shorter in children than in adults and has been reported to range from 3 to 32 hours (65,90). In a study of children 10 to 13 years of age, the half-life of CBZ declined from 25 to 32 hours after the first single dose, to 18 to 22 hours after 4 to 6 days of treatment, and to 10 to 14 hours after 4 weeks of therapy (65). In the same children, the mean first dose clearance was 0.028 ± 0.031 L/hr/kg, which doubled to 0.056 ± 0.011 L/hr/kg after 17 to 32 days of treatment. Because of the increased clearance, children need relatively larger total daily doses of CBZ, on a milligram per kilogram basis, than adults (88,91).
Elderly
The pharmacokinetics of a single 400-mg oral dose of CBZ were compared in a group of six young healthy volunteers aged between 20 and 25 years and in a group of five elderly volunteers aged between 66 and 84 years. No age-dependent modifications in elimination parameters were detected (92). However, an age-related decrease in the apparent oral clearance of CBZ was reported in a more recent population pharmacokinetic analysis (93).
Pregnancy
Data on the elimination of CBZ during pregnancy are conflicting. Some studies found no significant changes in the total plasma concentrations and in the intrinsic clearance of CBZ during pregnancy (38,94). The observation of an increased CBZ epoxide/CBZ ratio, associated with a decrease in levels of trans-CBZ-diol, was explained as the result of inhibition of the epoxide hydrolase rather than increased epoxidation (38). Other studies reported a decrease in total plasma concentrations of CBZ, whereas unbound concentrations remained constant (39,95). An increase in total plasma clearance of CBZ, largely dependent on an increased epoxide formation, has been documented during pregnancy in women comedicated with an anticonvulsant (96)
Other Possible Determinants of Intersubject Variability
There are no specific studies available on the effect of liver, renal, or cardiac diseases on the elimination parameters of CBZ (2,4). Moreover, no significant interethnic differences in CBZ kinetics and metabolism are to be expected, based on observations in Asian or African patients compared with whites (4).
RELATIONSHIP BETWEEN SERUM CONCENTRATION AND DOSE
Most studies have reported a poor or weak correlation between the dose and plasma concentration of either CBZ
P.242
or CBZ epoxide (30,42,97,98). In this respect, plasma levels of the metabolite were found to correlate more closely with CBZ dose than levels of the parent drug (60,99). In patients taking the same daily dose, there is a threefold to eightfold interindividual variability in steady-state plasma concentrations of CBZ and its epoxide metabolite, largely dependent on the different factors affecting the elimination of CBZ (30,42,60,97, 98, 99, 100). When CBZ is used as monotherapy, the plasma CBZ epoxide concentrations usually are 10% to 50% of those of the parent drug, without a clear relationship between concentrations of both compounds (99,101). Apparently, the ratio of CBZ epoxide to CBZ in plasma shows wide interindividual and intraindividual variability according to a number of factors, including age, physiologic status, concurrent drug therapy, dose regimens, and dosage schedule (30,36,42,60,91,98,99, 102). Furthermore, it has been clearly documented that steady-state plasma concentrations of both CBZ and CBZ epoxide oscillate markedly during the dosing intervals. Fluctuations of 40% to 150% for CBZ and 40% to 500% for the epoxide have been described (103, 104, 105). The magnitude of these interdosage fluctuations is influenced both by the absorption characteristics of the CBZ dosage form in use and by its elimination half-life. To minimize drug fluctuations, slow-release formulations have been commercialized (14).
In individual patients receiving CBZ monotherapy, it has been shown that steady-state plasma CBZ concentrations increase linearly with the dose over the range from 200 to 1,300 mg/day (106). However, at high doses, plasma levels of CBZ were found to increase less than proportionately in relation to the magnitude of drug dosage increases (68,107, 108, 109). Such a curvilinear relationship between dose and plasma CBZ concentration is due mainly to ongoing dose-dependent autoinduction of its metabolism (72,73); alternatively, the nonlinearity may be explained by a dose dependency in the absorption of CBZ (10).
RELATIONSHIP BETWEEN SERUM CONCENTRATION AND EFFECT
Several studies have investigated the relationship between plasma CBZ concentrations and clinical effects in epilepsy. The therapeutic range of plasma concentrations of CBZ for treating epilepsy has been estimated at 4 to 12 µg/mL or 17 to 51 µmol/L (110), although different intervals have been proposed, such as 4 to 8 µg/mL (111), 4 to 10 µg/mL (97), 6 to 10 µg/mL (112), 6 to 12 µg/mL (113), or 8 to 12 µg/mL (114). The lower limit of the therapeutic range is not clearly defined, and good seizure control has been observed over a wide range of plasma concentrations. Although earlier studies (112,114), conducted mostly in patients with intractable seizures, indicated that an effective response could not be achieved at concentrations lower than 4 to 6 µg/mL, other investigations in a broader range of epileptic patients showed that lower levels also could be effective. In this respect, Callaghan et al. (100) reported plasma CBZ concentrations from 1.2 to 8.0 µg/mL in 10 of 13 patients with complete seizure control. Shorvon et al. (111) treated 25 newly diagnosed patients with generalized tonic-clonic or partial seizures with CBZ as monotherapy, using a target concentration range of 4 to 8 µg/mL. Five patients achieved good control of seizures in spite of plasma CBZ concentrations less than 4 µg/mL. Seizures were completely controlled in 17 patients with levels within the targeted range, although they remained uncontrolled in 3 patients despite serum concentrations of 4 to 8 µg/mL. The upper limit of the therapeutic range also is poorly defined. However, although side effects have been reported to occur over a wide range of CBZ concentrations (100,114), they are more likely to appear at concentrations exceeding 10 to 12 µg/mL (110). In massive CBZ poisoning, plasma concentrations above 40 µg/mL have been associated with an increased risk of coma, seizures, respiratory failure, and cardiac conduction defects (115). During recovery from CBZ overdosage, plasma CBZ concentrations above 25 µg/mL were associated with coma and seizures, levels in the range 15 to 25 µg/mL with combativeness, hallucinations, and choreiform movements, and levels of 11 to 15 µg/mL with ataxia and drowsiness (116).
Some methodologic shortcomings may explain the discrepancies among studies (101,117). In particular, when trying to establish a therapeutic range, correlations should be studied at different dosages in the same patients or by randomization of groups of subjects to predefined dosages or plasma concentrations. It is remarkable that such data are available in only one study (111).
The concentration-response relationship for CBZ may be complicated by pharmacokinetic factors, such as the autoinduction of CBZ metabolism, the presence of an active metabolite, and variability in plasma protein binding. As previously stated, CBZ undergoes autoinduction of metabolism with a decrease in the plasma level-dose relationship and in half-life values during maintenance therapy, with a further decrease in the presence of concomitant antiepileptic drugs. Consequently, considerable interdosage fluctuations may occur in relation to the dosing interval. Therefore, collection of samples should be standardized with respect to time of dosing. A correlation between interdosage fluctuations in plasma CBZ concentrations and occurrence of intermittent side effects has been described (103,118,119). No conclusive evidence is available regarding the contribution of the active metabolite CBZ epoxide to the overall therapeutic and toxic effects (117). It is likely that, under usual conditions, CBZ epoxide does not contribute appreciably to clinical effects. However, in situations associated with abnormally elevated CBZ epoxide concentrations, such as in patients treated with inhibitors of epoxide hydrolase (Chapter 23), development of CBZ toxicity
P.243
may be attributable almost completely to markedly increased concentrations of CBZ epoxide in plasma. Another factor that may complicate the concentration-response relationship for CBZ is represented by differences in plasma protein binding. However, in the case of CBZ, the extent of drug binding is relatively modest (70% to 80%) and shows limited interindividual variability, so that monitoring free instead of total plasma CBZ concentrations has not proved advantageous (41).
Pharmacodynamic factors also may contribute to explain the variability in the CBZ level-response relationship. In particular, the therapeutic response of individual patients to a given plasma CBZ concentration appears to be considerably conditioned by the type and severity of epilepsy (120). In addition, the possibility of pharmacodynamic interactions with other anticonvulsants needs to be considered. When CBZ is coadministered with enzyme-inducing anticonvulsants, therapeutic and toxic effects usually occur at low plasma drug concentrations (121,122).
A limited number of studies have investigated the relationship between plasma CBZ levels and clinical response in clinical indications other than epilepsy. In a controlled study of patients with trigeminal neuralgia, optimal pain control was achieved at the concentration range of 5.7 to 10.1 µg/mL (106). In another investigation, involving patients with neuralgic pain, plasma CBZ concentrations in the range 2 to 7 µg/mL were found to be associated with a 25% to 75% reduction in pain in 50% of patients (123). On the other hand, no correlation between plasma concentration of CBZ and clinical effects was found in 18 affectively ill patients treated with CBZ as monotherapy (43).
CONVERSION
Conversion factor:
Conversion:
(µg/mL) × 4.23 = µmol/L
(µmol/L) ÷ 4.23 = µg/mL
REFERENCES
P.244
P.245
P.246