CANADIAN JOURNAL OF PSYCHIATRY Vol. 33 Ottawa, Canada, October 1988 No.7 Carbamazepine in Psychiatry: A Review* M. ISRAEL, M.D. 1 AND P. BEAUDRY, M.D. 2 .patients reported in the literature were noted to show positive effects on mood and behavior during treatment of their seizure disorder with carbamazepine. Although most of the studies reviewed at the time were uncontrolled clinical trials, seven of the studies were double-blind placebo controlled. In the early 1970's two groups of Japanese investigators (2,3) began to publish the results of open clinical trials suggesting a positive effect of carbamazepine in the treatment of bipolar affective disorder. Takezaki and Hanaoka (2) noted moderate to marked improvement in 7 out of 10 bipolar patients and in 6 out of 10 patients with symptomatic mania. Okuma (3) and his collaborators described an anti manic effect in 55% of 90 cases treated with carbamazepine. These studies were followed by double-blind placebo controlled studies comparing carbamazepine to neuroleptics (4) and to placebo (5) which confirmed the antimanic effects of carbamazepine. In a multi-institutional cooperative study carried out in Japan carbamazepine was compared to chlorpromazine in a double-blind design on a series of 63 acutely manic bipolar patients (4). Moderate to marked improvement was noted in 70% of the patients treated with carbamazepine versus 60% in the chlorpromazine treated group. Ballenger and Post (5) in a double-blind placebo controlled study demonstrated a partial to marked response in 5 of 9 manic patients (4 of the 9 patients showed a marked response). A recent review of both American and Japanese studies has suggested that 153 of 254 patients (60%) reported in the literature have responded to carbamazepine when treated during an acute manic episode (6). Controlled studies also suggest that carbamazepine exerts a prophylactic effect against both manic and depressive episodes (3,7,8). Okuma et al. (3) in an open clinical trial with 51 bipolar patients found a prophylactic effect for carbamazepine of 71% for manic episodes and 64% for depressive episodes. In a later study by the same group of investigators (3) the prophylactic efficacy of carbamazepine was explored using a double-blind placebo controlled design. A group of 22 bipolar patients (12 in the carbamazepine group and 10 in the placebo group) were followed for one year. The results revealed 60% moderate to marked efficacy in the carbamazepine group versus 22.2% in the placebo group. Similarly, Post et al. (8) followed a group of 7 rapidly cycling patients and noted a significant reduction in the number of manic and depressive episodes with carbamazepine treatment. Carbamazepine is a tricyclic compound structurally related to imipramine that has been used since the 1960's for the treatment of trigeminal neuralgias and then approved as an anticonvulsant in the U.S. in 1974. In the last decade the use ofcarbamazepine has been expanded to include the treatment ofcertain disorders ofmood and behaviour. In this paper possible clinical applications of carbamazepine in psychiatry are reviewed along with its pharmacokinetic and pharmacodynamic profile. Also discussed are adverse reactions and drug interactions which are relevant to its use. arbamazepine is an imino stilbene with a tricyclic chemical structure similar to the antidepressant C imipramine. has been used by neurologists since the It early 1960's for the treatment of trigeminal neuralgia and epilepsy. It is especially effective in the treatment of clinical and experimental seizures of the temporal lobe and limbic system. Amelioration of affect, cognitive performance and initiative and social behavior was also observed by neurologists in 50% of their patients treated with carbamazepine. These findings prompted studies on the efficacy of carbamazepine in the treatment of mental disorders. This paper will: (i) review the possible clinical applications of carbamazepine in psychiatry; (ii) discuss its pharmacokinetic and pharmacodynamic profile; and (iii) describe the adverse reactions and drug interactions often reported with its use. 1. Clinical Applications of Carbamazepine in Psychiatry Use of Carbamazepine in Affective Disorders It is through its increasing usage in the treatment of temporal lobe epilepsy that the psychotropic effects of carbamazepine began to be recognized. In an extensive review summarizing 40 studies, Dalby (1) noted that most studies reporting on the anticonvulsant effects of carbamazepine also mentioned a psychotropic effect of this agent. Approximately 50 percent of the 2,500 *Manuscript received October 1986: revised July 1987. I Assistant Professor, Department of Psychiatry, McGill University; Staff Psychiatrist, Allan Memorial Institute, Montreal, Quebec 2Assistant Professor, Department of Psychiatry, McGill University; Staff Psychiatrist, Clinical Psychopharmacology Unit, Allan Memorial Institute, Montreal, Quebec. Address reprint requests to: Dr. M. Israel, Allan Memorial Institute, 1025 Pine Avenue West, Montreal, Quebec H3A lAI Can. J. Psychiatry Vol. 33, October 1988 577 578 CANADIAN JOURNAL OF PSYCHIATRY The authors also reported a reduction in the severity and duration of episodes. Finally, a recent study has provided some preliminary data suggesting that carbamazepine also possesses some acute antidepressant efficacy (9). In a double-blind offon-off study with 35 treatment resistant depressed patients, Post et al. (9) reported at least a mild improvement in 57% of their sample with more substantial improvement in 34%. Of those depressed patients who did respond to carbamazepine, those who were more severely depressed at the initiation of treatment tended to respond better. Also noted was a nonsignificant trend towards better response to carbamazepine in bipolar versus unipolar depressed patients. Although the design of this study adequately controlled for placebo responders, the authors themselves acknowledge that spontaneous remissions could not effectively be ruled out among the responders. Most of the responders in this study did not relapse upon discontinuation of carbamazepine and the investigators did not have the opportunity to rechallenge patients in a second drug trial. Clearly further studies need to be carried out before the antidepressant efficacy of carbamazepine can be adequately evaluated. It is of interest to note that to date most of the patients studied were initially poor or non-responders to lithium carbonate. This would suggest that lithium and carbamazepine act via different mechanisms of action. Furthermore, in several studies carbamazepine was added to other agents including neuroleptics, tricyclic antidepressants and lithium. Nevertheless, some studies have demonstrated clinical efficacy using carbamazepine alone (4,7,8). Finally, there is little data available on direct comparisons between lithium and carbamazepine. Two recent studies in fact report contradictory findings. Lerer et al. (10) reported preliminary findings suggesting that carbamazepine is not as effective as lithium in the treatment of acute mania. In their study 15 manic patients were assigned randomly to treatment with either carbamazepine or lithium for 4 weeks using a "double placebo" system (8 received carbamazepine + placebo and 7 lithium + placebo). Their results indicate that carbamazepine was not as effective as lithium in attenuating the manic syndrome and that overall there was little therapeutic benefit in the patients treated with carbamazepine. A possible explanation for their negative findings is that unlike previous researchers, they did not use patients who were previously refractory to lithium or who were rapid cyclers. In contrast, Placidi et al. (11) compared carbamazepine to lithium in the acute and prophylactic management of patients with major affective, schizo affective or schizophreniform psychoses in a three year prospective double-blind randomized trial. They studied 83 patients (42 received carbamazepine and 41 received lithium) and found that both drugs were effective in two-thirds of patients. Furthermore, the two drugs appeared comparable in most outcome measures except for a significantly higher dropout rate Vol. 33, No.7 for patients with mood-incongruent psychotic features who were treated with lithium. Both drugs also appeared more effective in preventing excited rather than depressive symptoms. Thus, their findings suggest that carbamazepine is as effective as lithium in the treatment of the full spectrum of psychoses with affective components. In general, when antimanic effects of carbamazepine are noted they tend to occur within the first week of treatment. The hyperactive, aggressive and excitable aspects of the manic syndrome seem to respond more rapidly than the psychotic components (6). The onset of antidepressant effects also seems to lag behind that of the antimanic effects, often requiring several weeks before improvement is noted. However, improvement in sleep is generally evident within the first week in both manic and depressed patients. Finally, there is general agreement that the more manic and psychotic the patient, the greater the likelihood of a therapeutic response to carbamazepine. In addition it appears that those patients with rapid or continuous manic depressive cycles are those most likely to respond to carbamazepine (6). Use of Carbamazepine in Aggressive and Non-Affective Psychosis There are several reports on the treatment of aggressive behavior and non-affective syndromes. Most of these are open studies demonstrating the efficacy of carbamazepine in treating aggression in schizophrenics (12-14), schizoaffectives (15), personality disorders (14) and organic brain syndromes (14,15). Hakola and Laulumaa (12) described 8 schizophrenic women in a maximum security ward with histories of violent episodic outbursts who responded to the addition of carbamazepine to their neuroleptic regimen. All women had non-specific EEG abnormalities as well as irritative activity with focal tendencies in 3 cases and without focal tendencies in 4 cases. Treatment with carbamazepine resulted in almost complete disappearance of violent behavior in all 8 patients. In 4 patients the specific schizophrenic symptoms were also markedly decreased. Luchins (13) described 7 patients (6 schizoaffective, 1 mixed personality disorder) with normal EEGs who received carbamazepine in an off-on-off design in addition to their neuroleptic regimen. A significant decrease in the number of aggressive episodes was noted during carbamazepine treatment as compared to before or after the treatment. The same author in a later study (14) found no difference when comparing the efficacy of carbamazepine in decreasing the frequency of aggressive episodes in 8 violent patients with abnormal EEGs versus 11 violent patients with normal EEGs. A few controlled studies have also been conducted in this area (16,17). Neppe (16) carried out a 15 week double-blind placebo controlled crossover study on 11 patients with mixed diagnoses (8 were schizophrenics). All patients had nonspecific temporal lobe abnormalities on EEG and in all cases carbamazepine was again added to the pre-existing October, 1988 CARBAMAZEPINE IN PSYCHIATRY neuroleptic regimen. Significant improvement was noted in 8 patients. Furthermore, a reduction in the frequency and severity of aggression was noted in the 6 patients who exhibited aggressiveness. Finally, Klein et al. (17) compared carbamazepine and haldol to placebo and haldol in patients with excited schizophrenic or affective psychosis in a controlled double-blind design. Twentythree patients received carbamazepine and haldol whereas 20 patients received placebo and haloperidol. Carbamazepine was found to be superior to placebo and equally effective in schizophrenic or affective psychoses as measured by percentage of improvement over baseline on the Brief Psychiatric Rating Scale (BPRS). All the patients who improved had normal EEGs. Furthermore, 15 of 18 BPRS symptoms improved, and items which did not depend on activity (for example, unusual thought content) improved as much as those that did. Some studies have specifically reported a beneficial effect of carbamazepine in the treatment of episodic dyscontrol syndromes (18-20). Some of the patients reported showed definite EEG abnormalities. For example, Tunks (18) described a 24 year old woman with episodic dyscontrol syndrome characterized by intermittent aggressive outbursts associated with dysphoria which disappeared when carbamazepine was added to her treatment regimen (primidone and chlorpromazine). The patient also had non-specific EEG abnormalities. Gardner and Cowdry (20) recently studied the effects of carbamazepine on behavioral dyscontrol in 11 women with borderline personality disorder. Using a double-blind crossover trial, they found significantly less severe behavioral dyscontrol with carbamazepine than with placebo. EEG studies were not carried out in their patients. Although the results of the above studies suggest a potential use for carbamazepine in non-affective psychiatric disorders, the small number of patients, the paucity of controlled studies and the tendency to combine patients with different psychiatric diagnoses make the interpretation of results difficult. In addition, in all studies mentioned carbamazepine was always used as an adjunct to other psychotropic drugs although many authors report a reduction in the dosage of other drugs when carbamazepine was added to the treatment regimen. Finally, the presence of EEG abnormalities in some of the patients studied adds another dimension that requires experimental clarification. For instance, in order to demonstrate a specific effect of carbamazepine on aggressivity unrelated to its anticonvulsant effect, more rigorous electrophysiological studies other than standard surface EEGs would need to be performed. Similarly, studies would have to differentiate between the proposed mood stabilizing effects of carbamazepine and specific effects on violent or psychotic behavior. The data presented so far deals with a small group of selected heterogeneous patients who mainly share the characteristic of being refractory to all conventional treatments and who therefore make such distinctions difficult. It is 579 therefore not surprising that some studies also report negative findings. For example, Dalby (21) noted a paradoxical worsening of aggression in response to carbamazepine in a small number of patients with brain damage and/ or mental retardation. Stevens (22) also reported on 2 schizophrenics who markedly worsened when carbamazepine was added to their treatment regimen. 2. Pharmacokinetic and Pharmacodynamic Profile Pharmacokinetics Carbamazepine is rapidly absorbed after oral intake. Peak plasma concentrations are reached within 2 to 6 hours with considerable individual variations. Carbamazepine is approximately 75% bound to plasma proteins whereas its active metabolite carbamazepine 1O,1l-epoxide is 50% bound. Carbamazepine is metabolized in the liver and may induce its own metabolism. The initial half-life of the parent compound ranges from 25 to 65 hours while the steady state half-life ranges from 12 to 17 hours. Thus, after 2 to 3 weeks of treatment serum levels usually decrease, as a result of induction of hepatic enzymes, and it may be necessary to increase the dose in order to maintain therapeutic drug levels. The half-life of the active metabolite is in the range of 5 to 8 hours. Given the half-life at steady states, carbamazepine must be used in divided doses (usually b.i.d. or t.i.d. depending on the overall dose). Carbamazepine is metabolized by the liver to its active metabolite carbamazepine 1O,1l-epoxide. There is then further metabolism to carbamazepine 10,Ll-dehydroxide and subsequent conjugation with glucoronic acid. Less than 1% is recovered in the urine as a parent compound or metabolite. Pharmacodynamics Mechanisms of Actions In both clinical and experimental settings, carbamazepine has been shown to have a relatively greater limbic than cortical anticonvulsant effect (23,24). Carbamazepine has been demonstrated to suppress after-discharges and spontaneously induced seizure discharges preferentially in limbic areas over other cortical areas (24-26). Carbamazepine also suppresses seizures kindled in limbic areas by repetitive subthreshold electrical stimulation (25-27). The postulated role of the limbic system in the regulation of emotion has led some researchers to suggest that the efficacy of carbamazepine in the treatment of affective disorders may well be mediated via its preferential effect on limbic structures (24). The effect of carbamazepine on various neurotransmitters has been extensively reviewed and will be summarized below. To date, it appears that carbamazepine's mechanism of action is largely nonspecific as it affects several neuronal systems. Noradrenergic Effects Carbamazepine has been found to be a noradrenergic reuptake blocker although it is relatively a weaker 580 CANADIAN JOURNAL OF PSYCHIATRY inhibitor of reuptake when compared to imipramine (28). The fact that unlike other antieonvulsants carbamazepine seems to exert an effect on catecholamine systems has led some investigators to suggest that the action is implicated in carbamazepine's psychotropic effect rather than mediating its anticonvulsant action (29,30). Other studies, however, have shown that depletion of norepinephrine with 6-hydroxydopamine decreases the anticonvulsant efficacy of carbamazepine (31). Effects on Dopaminergic Systems The effects of carbamazepine on the dopaminergic system have been extensively studied and reviewed (32). The data collected to date suggests that carbamazepine acts through mechanisms other than the blockade of dopaminergic receptors. In contrast to neuroleptics carbamazepine: (i) does not tend to produce extrapyramidal side effects such as parkinsonism or tardive dyskinesia; (ii) does not cause substantial increases in plasma prolactin; (iii) does not increase homovanillic acid (HV A) levels in human cerebrospinal fluid (CSF) although it decreases probenicid-induced accumulation of HVA; (iv) does not increase firing of dopaminergic neurons in the substantia nigra; and (v) does not block stimulant induced psychomotor activity. Furthermore, carbamazepine does not appear to block dopamine receptors as measured directly by displacement of [3H]spiroperidol. The data therefore suggest that carbamazepine may decrease dopamine turnover but unlike neuroleptics it does not act via blockade of dopamine receptors. Effects on GABA Carbamazepine has been reported to decrease GABA turnover (33). This effect could reflect a functional increase in GABA activity, a property shared by other anticonvulsants such as valproic acid. It is also of interest to note that lithium and propranolol also decrease GAB A turnover. Thus, the action of carbamazepine on GABA may be potentially related to both its psychotropic and anticonvulsant effects. It must be noted, however, that carbamazepine was not shown to affect CSF GABA levels in bipolar patients who responded to the drug (34). Effects on Benzodiazepine Receptor Sites There is now evidence to suggest that the "peripheraltype" benzodiazepine receptor is functionally involved in the anticonvulsant effect of carbamazepine (35). This receptor, although originally described in the periphery, is also known to be localized in central structures including olfactory bulb, hippocampus and cortex. Carbamazepine, however, does not seem to act at the central benzodiapezine site and is not associated with tolerance or withdrawal effects. The implication of these findings on the psychotropic effects of carbamazepine remain to be elucidated. Vol. 33, No.7 Adenosine Receptor Effects Carbamazepine is reported to be a potent agent in displacing binding of specific adenosine antagonists and agonists (36). Carbamazepine has been shown to share many properties with adenosine antagonists (37). The data suggests that carbamazepine's effects on the adenosine receptors is not critical to its anticonvulsant activity but may well be related to its other clinical effects. It has also been proposed that the effects of carbamazepine on the noradrenergic system may be linked to its action on adenosine (38). Serotonergic Effects Carbamazepine has not been found to alter basal levels of 5HlAA in the CSF of affectively ill patients (39). However, it has been found that carbamazepine, unlike other anticonvulsants, increased plasma levels of tryptophan. In the context of the postulated role of the serotonergic system in affective disorders this finding may reflect a relationship between carbamazepine's psychotropic action and the serotonergic system. Cholinergic Effects Researchers have found that carbamazepine increased acetylcholine levels in rat striatum while choline levels were decreased (40). Effects on Cyclic Nucleotides Carbamazepine inhibits accumulation of cyclic AMP induced by many stimuli (38). Basal levels of cyclic AMP or cyclic GMP remained unchanged in affectively ill patients treated with carbamazepine (41). However, it was noted that probenecid-induced accumulation of cyclic AMP and cyclic GMP were significantly decreased during carbamazepine treatment. These findings are of interest because of the postulated role of cyclic nucleotides in both affective illness and seizure disorders as well as the proposed role of lithium's effect on adenylate cyclase activity in its mechanism of action. Endocrine Effects Carbamazepine induces escape from dexamethasone suppression (42,43). Carbamazepine also increases urinary free cortisol in both normal volunteers (44) as well as in depressed patients with normal baseline levels of urinary free cortisol (43). It has been suggested that the effects of carbamazepine on cortisol may be mediated via its effects on somatostatin (38). Carbamazepine has been observed to significantly decrease somatostatin in CSF. Since somatostatin has been shown to inhibit CRFmediated secretion of ACTH, a decrease in somatostatin would result in removal of ACTH inhibition and therefore in cortisol hypersecretion. Somatostatin has also been reported to be decreased in untreated depressed patients. The implication of the above findings on the mechanism of action of carbamazepine awaits further research. October, 1988 CARBAMAZEPINE IN PSYCHIATRY In contrast to lithium, carbamazepine has been noted to have an antidiuretic effect. Several studies suggest that carbamazepine may actually act directly at the vasopressin receptor (reviewed in 38,45). Carbamazepine has been successfully used in the treatment of central diabetes insipidus (46). It is not yet clear, however, whether a relationship exists between carbamazepine's effects on vasopressin and its mechanism of action as an anticonvulsant or psychotropic agent. Finally, carbamazepine has been demonstrated to decrease circulating levels of T 3 and T 4 with altering levels of TSH. Furthermore, it has been found tht the decreases in peripheral thyroid hormones were significantly greater in carbamazepine responders when compared to nonresponders (47). This finding led investigators to suggest that carbamazepine's effects on thyroid hormones may be related to its therapeutic efficacy. The exact mechanism, however, remains to be elucidated and would represent a worthwhile goal for future research. 3. Adverse Reactions and Drug Interactions Adverse Reactions According to clinical experience approximately 25% of patients receiving carbamazepine experience adverse reactions (3,48). Most untoward effects subside within one week or after a reduction in dosage. Adverse reactions can be minimized by starting with small doses and increasing the amount gradually. Adverse reactions in psychiatric patients appear to be similar to those encountered in epileptic and pain patients. The most commonly reported adverse effects include nausea, drowsiness, dry mouth, vertigo, ataxia and diplopia. These effects tend to be mild and transient and if they persist, tend to remit when drug dosages are decreased. Dermatologic reactions are reported to occur in 8-15% of patients treated with carbamazepine (6,48,49). These range from mild exanthems to more serious lesions such as Stevens-Johnson syndrome, exfoliative dermatitis and toxic epidermic necrolysis. It is recommended that the drug be discontinued if dermatologic reactions occur as they may represent early signs of more serious toxicity. Although rare, hepatic toxicity has also been associated with carbamazepine treatment (49), cholestatic jaundice, hepatocellular jaundice and abnormal liver function tests accompanied by immune complexes have been reported (49). In some patients elevation of liver enzymes will be transient. However, it is recommended that baseline liver function tests (including serum bilirubin, alkaline phosphatase, SGOT, SGPT and lactate dehydrogenase) be obtained and that symptoms such as nausea, vomiting, anorexia and fever be immediately investigated. Routine monitoring ofliver function tests is also recommended with discontinuation of the drug if persistent elevation of liver enzymes is noted or if the patient becomes symptomatic. As a result of its antidiuretic action, hypnotremia has also been reported in both psychiatric and nonpsychiatric populations treated with carbamazepine (45). 581 In a recent study, a significant incidence of hyponatremia was reported in 12 patients treated with carbamazepine when compared to placebo (45). Furthermore, a higher risk was noted with older patients as well as in patients with higher serum drug levels. It is currently estimated that self-induced water intoxication can occur in 3 to 4% of hospitalized psychiatric patients (50). Thus, these patients would constitute a population at risk for the development of clinically significant water intoxication if treated with carbamazepine. It is therefore advisable to obtain a serum sodium determination in any patient who drinks large quantities of water or who develops unexplained alterations in central nervous sytem functioning while on carbamazepine. A wide range of hematological effects have been reported as a result of treatment with carbamazepine. These include transient leukopenia, eosinophelia, leucocytosis, reticulocytosis, thrombocytopenia purpura, agranulocytosis, macrocytic anemia and aplastic anemia. A 1983 review of carbamazepine induced hematological toxicity reported 22 cases of aplastic anemia and 15 cases of agranulocytosis some of which were lethal (51). It must be noted that the link between carbamazepine and the severe hematological effects in many of the cases was obscured by the use of polypharmacy as well as by the presence in some cases of severe underlying illnesses. Although most of the cases reviewed to date occurred in non-psychiatric populations, recent studies have noted the occurrence of fatal hematological effects in psychiatric patients (52). At present the estimated incidence of carbamazepine induced aplastic anemia and/ or agranulocytosis is 1 in 20,000 to 40,000 patients (51). A transient decrease in the white count can be expected in many patients during the first few weeks of treatment (53). In patients with low or below normal pretreatment white counts, leukopenia may result. There is general agreement in the literature that in most patients white blood cell counts will tend to return to baseline over time (usually within I month). In a recent review of carbamazepine-induced leukopenia it was found that 94% of the cases were over 45, suggesting that age may be a factor (54). Since most cases of aplastic anemia have occurred early in carbamazepine treatment it is generally recommended that a complete blood count with differential be obtained weekly for the first 2 to 3 months and then every 3 months for the next 2 to 3 years. Table I (6) summarizes current recommended guidelines for discontinuation of carbamazepine. In addition patients must also be advised to immediately contact their physician ifthey develop fever, sore throat, mouth ulcers, easy bruising, petechia, purpura or bleeding from mucous membranes. Drug Interactions With the increasing use of carbamazepine in the treatment of psychiatric disorders, reports have emerged 582 CANADIAN JOURNAL OF PSYCHIATRY Table I Recommended Guidelines for CBZ Discontinuation WBC total Neutrophils Erythrocytes Hematocrit Hemoglobin Platelets Reticulocytes Serum iron < < < < < < < > 4000jcu mm 1500jcu mm 4,000,000jcu mm 32% 11.0 gjdl 100,000jcu mm 0.3% 150ugjdl on the potential interactions between carbamazepine and other psychotropic drugs. Clinical reports have suggested that synergistic effect may exist between carbamazepine and lithium in the treatment of patients with bipolar disorders (55,56) and in violent behavior (57). Neurotoxicity has also been observed when carbamazepine is combined with lithium (58). This observation, however, was made on 5 patients, all of whom had at least one other risk factor such as concurrent medical or neurological illness. In all cases, signs of neurotoxicity remitted after the discontinuation of one of the drugs. Interactions have also been described between carbamazepine and neuroleptics. Specifically, delirium developed in one patient when carbamazepine was added to a regimen of haloperidol and thioridazine which remitted when the neuroleptics were discontinued (59). A similar syndrome was later described in a patient who was treated with only haloperidol and carbamazepine (60). Carbamazepine has also been reported to lower haloperidol plasma levels which may in some cases result in a deterioration of clinical status (61,62). This effect is thought to be mediated via a carbamazepine-induced increase in the metabolism of haloperidol (62). Carbamazepine induces the microsomal enzymes and can therefore accelerate the metabolism of any drug that is metabolized by this system, including phenytoin, warfarin and doxycycline (63). Similarly, carbamazepine induces its own metabolism resulting in decreased serum levels, usually 2 to 3 weeks after the initiation of treatment. There have also been several case reports of carbamazepine toxicity when it is co-administered with the monoamine oxidase inhibitor isoniazid (64-66). It is proposed that the interaction is mediated by isoniazidinduced inhibition of microsomal enzymes resulting in a rise in carbamazepine blood levels. However, no toxicity was reported when tranylcypromine was co-administered with carbamazepine in one patient (67). Until further studies are done documenting the potential interaction between monoamine oxidase inhibitors and carbamazepine, it is recommended that these drugs be administered concomitantly only at moderate doses and that blood levels of carbamazepine be monitored weekly in the first month of therapy. Vol. 33, No.7 Dosage and Blood Level Monitoring It is best to start with low doses with small increments in order to minimize adverse effects and maximize patient compliance. It is recommended to start treatment with doses of 200 mg daily or twice daily and then to gradually increase the dose by 200 mg increments every few days. Most patients will require an average of 1000 mg daily in divided doses, although many patients respond to 400 to 600 mg per day. A few patients have required up to 2200 mg per day. However, it is not recommended to use high doses unless some response is seen with lower doses. A close relationship does not exist between therapeutic response and carbamazepine blood levels. Therapeutic responses tend to occur in the range of 6 to 12 ugml which corresponds to the range used for the treatment of epilepsy. A closer correlation may exist between levels of carbamazepine 10,11- epoxide, in both plasma and CSF, and the degree of clinical response (68). Conclusions Although there is growing evidence for a role of carbamazepine in the treatment of certain psychiatric disorders, further research is required before specific indications for its use can be established. To date the evidence does suggest that carbamazepine may represent a potential alternative in the treatment of selected cases of manic-depressive illness. This evidence is of special significance in light of the growing recognition that as many as 20 to 30% of patients with manic- depressive illness fail to respond to lithium carbonate or are unable to tolerate its side effects. In fact, the literature suggests that rapid cyclers and those who are initially more manic and psychotic tend to respond most favorably to carbamazepine. Nevertheless, carbamazepine should not be used as a first line drug in the treatment of manicdepressive illness. One possible exception may be those patients with a history of seizures or paroxysmal discharges on EEG. In such cases carbamazepine may be preferable to lithium or neuroleptics because of the associated risk of lowering the seizure threshold. The data for the antidepressant effects of carbamazepine are less conclusive and additional controlled trials are necessary before its efficacy can be accurately estimated. However, carbamazepine should still be considered as an alternative for those patients who are refractory to more established antidepressants. In addition, carbamazepine may be useful in treating patients with violent outbursts whose aggressiveness cannot be controlled with neuroleptics. Finally, although much attention has been focused on the potentially lethal side effects of carbamazepine, these tend to be rare and it must be remembered that carbamazepine has been used relatively safely for many years in the treatment of nonpsychiatric disorders. It is, however, important to remember that carbamazepine has not yet been approved by the Food and Drug Administration (FDA) or the Health Protection Board October, 1988 CARBAMAZEPINE IN PSYCHIATRY (HPB) for its application in psychiatric disorders. As carbamazepine remains an investigational drug in psychiatry it is recommended that patients be informed of this fact with the concomitant provision of information regarding the benefits and anticipated risks associated with its use. Preferably the information provided to the patient and the patient's written consent should be well documented in the chart. Special care should also be taken to document any adverse reactions along with their course and treatment. References I. Dalby MA. Antiepileptic and psychotropic effect of carbamazepine (Tegretol) in the treatment of psychomotor epilepsy. Epilepsia 1971; 12: 325-334. 2. Takezaki H, Hanaoka M. The use of Carbamazepine (Tegretol) in the control of manic-depressive psychosis and other manic- depressive states. Clin Psychiatry 1971; 13: 173-183. 3. Okuma T. Therapeutic and prophylactic effects of carbamazepine in bipolar disorders. Psychiatr Clin North Amer 1983; 6: 157- 174. 4. Okuma T, Inanaga K, Otsuki S, et al. Comparison of the antimanic efficacy of carbamazepine and chlorpromazine: a double- blind controlled study. Psychopharmacology 1979; 66: 211-217. 5. Ballenger JC, Post RM. Carbamazepine in manicdepressive illness: a new treatment. Am J Psychiatry 1980; 137: 782-790. 6. Post RM. Clinical perspectives on the use of carbamazepine in manic-depressive illness. Fair Oaks Hospitals Psychiatry Letter 1985; 3(4). 7. Okuma T, Kazutoyo I, Otsuki S, et al. A preliminary double- blind study on the efficacy of carbamazepine in prophylaxis of manic-depressive illness. Psychopharmacology 1981; 73: 95-96. 8. Post RM, Uhde TW, Ballenger JC, et al. Prophylactic efficacy of carbamazepine in manic-depressive illness. Am J Psychiatry 1983; 140: 1602-1604. 9. Post RM, Uhde TW, Roy-Byrne PP, et al. Antidepressant effects of carbamazepine. Am J Psychiatry 1986; 143: 1-34. 10. Lerer B, Moore N, Meyendorff E, et al. Carbamazepine and lithium: different profiles in affective disorder? Psychopharmacol Bull 1985; 21: 18-22. II. Placidi GF, Lenzi A, Lazzerini F, et al. The comparative efficacy and safety of carbamazepine versus lithium: a randomized, double-blind 3-year trial in 83 patients. J Clin Psychiatry 1986; 47: 490-494. 12. Hakola HPA, Laulumaa VAO. Carbamazepine in treatment of violent schizophrenics. Lancet 1982; i: 1358. 13. Luchins DJ. Carbamazepine for the violent psychiatric patient. Lancet 1983; ii: 766. 14. Luchins DJ. Carbamazepine in violent non-epileptic schizophrenics. Psychopharmacol Bull 1984; 20: 569-571. 15. Folks DG, King LD, Dowdy SB, et aI. Carbamazepine treatment of selectively affectively disordered patients. Am J Psychiatry 1982; 139: 115-117. 16. Neppe VM. Carbamazepine for the violent psychiatric patient. Lancet 1982; ii: 334. 17. Klein E, Bental E, Lerer B, et al. Carbamazepine and haloperidol versus placebo and haloperidol in excited psychoses. Arch Gen Psychiatry 1984; 41: 165-170. 583 18. Tunks ER, Dermer SW. Carbamazepine in the dyscontrol syndrome associated with limbic system dysfunction. J Nerv Ment Dis 1977; 164: 56-63. 19. Garbutt JC, Loosen PT. Is carbamazepine helpful in paroxysmal behavior disorders? Am J Psychiatry 1983; 140: 1363-1364. 20. Gardner DL, Cowdry RW. Positive effects of carbamazepine on behavioral dyscontrol in borderline personality disorder. Am J Psychiatry 1986; 143: 519-522. 21. Dalby MA. Antiepileptic and psychotropic effects of carbamazepine (Tegretol) in the treatment of psychomotor epilepsy. Epilepsia 1975; 12: 325-334. 22. Stevens JR, Bigelow L, Denney D, et al. Telemetered EEGEOG during psychotic behaviors of schizophrenia. Arch Gen Psychiatry 1979; 36: 251-269. 23. Julien RM, Hollister RP. Carbamazepine: mechanisms of action. In: Penry JK, Daly DD, eds. Complex partial seizures and their treatment. Advances in neurology, Vol. II. New York: Raven Press, 1975: 263-277. 24. Albright PS, Burnham WM. Development of a new pharmacological seizures model: effects of anticonvulsants on cortical- and amygdala-kindled seizures in the rat. Epilepsia 1980; 21: 681- 689. 25. Post RM, Uhde TW, Putnam FW, et al. Kindling and carbamazepine in affective illness. J Nerv Ment Dis 1982; 170: 717-731. 26. Babington RG, Horovitz ZP. Neuropharmacology of SO 10,996, a compound with several therapeutic indications. Arch Int Pharm Ther 1973; 202: 106-118. 27. Wada JA. Pharmacological prophylaxis in the kindling model of epilepsy. Arch NeuroI1977; 34: 389 395. 28. Purdy RE, Julien RM, Fairhurst AS, et aI. Effect of carbamazepine on the in vitro uptake and release of norepinephrine in adrenergic nerves of rabbit aorta and in whole brain synaptosomes. Epilepsia 1977; 18: 251-257. 29. Waldmeier PC, Baumann PA, Fehr B, et al. Carbamazepine decreases catecholamine turnover in the rat brain. Pharm Exp Ther 1984; 231: 166-172. 30. Evans RW, Gualtieri CT. Carbamazepine: a neuropsychological and psychiatric profile. Clin Neuropsychopharmaco11985; 8: 221-241. 31. Quattrone A, Crunelli V, Saminin R. Seizures susceptibility and anticonvulsant activity of carbamazepine, diphenylhydantoin and phenobarbital in rats with selectivve depletion of brain monoamines. Neuropharmacology 1978; 17: 643-647. 32. Post. RM, Rubinow DR, Uhde TW, et al. Dopaminergic effects of carbamazepine relationship to clinical response in affective illness. Arch Gen Psychiatry 1986; 43: 392-396. 33. Post RM, Uhde TW, Rubinow DR, et al. Biochemical effects of carbamazepine: relationship to its mechanism of action in affective illness. Prog Neuropsychopharmacol Bioi Psychiatry 1983; 7(2-3): 263-271. 34. Post RM, Ballenger JC, Hare TA, et al. Lack of effect of carbamazepine on gamma-aminobutyric acid in cerebrospinal fluid. Neurology 1980; 30: 1008-1011. 35. Weiss SRB, Post RM, Patel J, et al. Differential mediation of the anticonvulsant effects of carbamazepine and diazepam. Life Sci 1985; 36: 2413-2419. 36. Marangos PJ, Post RM, Patel J, et al. Specific and potent interactions of carbamazepine with brain adenosine receptors. Eur J Pharmacol1983; 93: 175-182. 584 CANADIAN 10URNAL OF PSYCHIATRY 37. Skeritt lH, Davies LP, Johnson GAR. Interaction of the anticonvulsant carbamazepine with adenosine receptors. I: neurochemical studies. Epilepsia 1983; 24: 634-642. 38. Post RM, Rubinow DR, Uhde TW. Biochemical mechanisms of action of carbamazepine in affective illness and epilepsy. Psychopharmacol Bull 1984; 20: 585-589. 39. Pratt 1, lenner P, Showon SD, et al. Anticonvulsant drugs alter plasma tryptophan concentrations in epileptic patients: implications for antiepileptic action and mental function. 1 Neurol Neurosurg Psychiatry 1984; 47: 1131-1133. 40. Consolo S, BIanchi S, Ladinsky H. Effects of carbamazepine on cholinergic parameters in rat brain areas. Neuropharmacology 1976; 15: 653-657. 41. Post RM, Ballenger lC, Uhde TW, et al. Effect of carbamazepine on cyclic neuleotides in CSF of patients with affective illness. Biol Psychiatry 1982; 17: 1037-1045. '42. Privitena MR, Greden IF, Gardner KW, et al. Interference by carbamazepine with the dexamethasone suppression test. BioI Psychiatry 1982; 17: 611-620. 43. Post RM, Uhde TW, Ballenger lC, et al. Carbamazepine, temporal lobe epilepsy and manic-depressive illness. In: Koella WP, Tremble MR, eds. Advances in biological psychiatry, Vol. 8. Temporal lobe epilepsy, mania and schizophrenia and the limbic system. Basel: S. Karger 1982: 117-156. 44. Luhdorf K, Christiansen P, Hansen 1M, et al. The influence of phenytoin and carbamazepine on endocrine . function: preliminary results. In: Penny lK, ed. Epilepsy, the 8th international symposium. New York: Raven Press, 1977: 209-213. 45. Uhde TW. Effects of carbamazepine on serum electrolytes: clinical and theoretical implications. 1 Clin Psychopharmaco11983; 3: 103-106. 46. Wales lK. Treatment of diabetes insipidus with carbamazepine. Lancet 1975;i: 948-951. 47. Roy-Byrne PP, Joffe RT, Uhde TW, et al. Effects of carbamazepine on thyroid function in affectively ill patients: clinical and theoretical implications. Arch Gen Psychiatry 1984; 41: 1150-1153. 48. Livingston S, Pauli LL, Bernau W; Carbamazepine (Tegretol) in epilepsy: nine year follow-up study with special emphasis on untoward effects. Dis Nerv Syst 1974; 35: 103-107. 49. Moore NC, Lerer B, Meyendorff E, et al. Three cases of carbamazepine toxicity. Am 1 Psychiatry 1985; 142: 974-975. 50. Nixon PA, Rothman lS, Chen W. Demeclocycline in the prophylaxis of self-induced water intoxication. Am 1 Psychiatry 1982; 139: 828-829. 51. Pisciotta AV. Hematologic toxicity of carbamazepine. In: Penry lK, Daly DD, eds. Complex partial seizures and their treatment. Advances in neurology, Vol. 11. New York: Raven Press, 1975: 355-368. 52. Luchins Dl. Fatal aganulocytoses in a chronic schizophrenic patient treated with carbamazepine. Am 1 Psychiatry 1984; 141: 687-688. 53. Joffe RT, Post RM, Roy-Byrne PP, et al. Hematological effects of carbamazepine in patients with affective illness. Am 1 Psychiatry 1985; 142: 1196-1199. 54. Owens CWI, Parker NE, Nunn PP, et al. Agranulocytosis associated with carbamazepine, and a positive reaction with anti- lymphoid leukemia antiserum during recovery. Postgrad Med 11980; 56: 665-668. Vol. 33, No.7 55. Lipinski 1, Pope HG. Possible synergistic interaction between carbamazepine and lithium carbonate in the treatment of three acutely manic patients. Am 1 Psychiatry 1982; 139: 948-949. 56. Kwamie Y, Persad E, Stancer H. The use of carbamazepine as' an adjunctive medication in the treatment of affective disorders: a clinical report. Can 1 Psychiatry 1984; 29(7): 605-608. 57. Buck Of), Harvey P. Combined carbamazepine and lithium therapy for violent therapy (letter). Am 1 Psychiatry 1986, 143: 1487. 58. Shukla S, Godwin CD, Long LEB, et al. Lithiumcarbamazepine neurotoxicity and risk factors. Am 1 Psychiatry 1984; 141: 1604-1606. 59. Kanter GL, Yerevanian BI, Ciccone lR. Case report of a possible interaction between neuroleptics and carbamazepine. Am 1 Psychiatry 1984; 141: 1101-1102. 60. Yerevanian BI, Hodgman CH. A haloperidolcarbamazepine interaction in a patient with rapid cycling bipolar disorder (letter). Am 1 Psychiatry 1985; 142: 785-786. 61. lann MW, Ereshefsky L, Saklad SR, et al. Effects of carbamazepine on plasma haloperidol levels. 1 Clin Psychopharmacol1985; 5: 106-109. 62. Arana GW, Goff DC, Friedman H, et al. Does carbamazepine- induced reduction of plasma haloperidol levels worsen psychotic symptoms? Am 1 Psychiatry 1986; 143: 650-651. 63. Hansen 1M, Siersback-Nielsen K, Skousted L. Carbamazepine- induced acceleration of diphenylhydantoin and warfarin metabolism in man. Clin Pharmacol Ther 1971; 12: 539-543. 64. Wright 1M, Stokes EF, Sweeney VP. Isoniazid-induced carbamazepine toxicity and vice versa. N Engl 1 Med 1982; 18: 1325-1327. 65. Valsalan VC, Cooper GL. Carbamazepine intoxication caused by interaction with isoniazid. Br Med 1 1982; 283: 261-262. 66. Block SH. Carbamazepine-isoniazid interaction. Pediatrics 1982; 69: 494-495. 67. Joffe RT, Post RM, Uhde TW. Lack of pharmacokinetic interaction of carbamazepine with tranylcypromine. Arch Gen Psychiatry 1985; 42: 738. 68. Post RM, Uhde TW, Wolff EA. Profile of clinical efficacy and side effects of carbamazepine in psychiatric illness: relationship to blood and CSF levels of carbamazepine and its -10,11-epoxide metabolite. Acta Psychiatr Scand 1984; 69: 104-120. Resume La carbamazepine est un compose tricyclique relie par sa structure a l'imipramine, qui a ere utilise depuis les annees 60 pour Ie traitement de la nevralgie essentielle du trijumeau et qui a ete approuve aux Etats- Unis comme anticonvulsivant en 1974. Depuis la demiere decennie, on se sert aussi de ce medicament pour Ie traitement de certains troubles de l'humeur et du comportement. Dans Ie present article, les auteurs examinent certains usages possibles de la carbamazepine en psychiatrie ala lumiere de ses caracteristiques pharmacokinesiques et pharmacodynamiques. lis discutent enfin des reactions negatives et des interactions medicamenteuses reliees ace medicament.