Psychotropic Properties of Antiepileptic Drugs

Norman Sussman

Alan B. Ettinger

Interest in the comorbidity between neurological and psychiatric diseases, as well as the concomitant administration of antiepileptic drugs (AEDs) and psychotropic drugs, has grown as the technology for studying the brain and the variety of new drugs has expanded. There are common mechanisms at play in both epilepsy and psychiatric disorders. The anatomical and biochemical interrelationships of brain areas implicated in the manifestations of both epilepsy and psychiatric disorders are of theoretical interest and may one day provide a better basis for developing new treatments. Newer imaging and genetic technology has already yielded new insights into these disorders.

There has been considerable therapeutic crossover of AEDs into the treatment of psychiatric disorders. This chapter provides a clinical perspective on the use of AEDs in psychiatry, which could potentially have important implications for the use of AED in patients with epilepsy. One inherent limitation of this discussion is the absence of replicated controlled studies of psychotropic agents in seizure disorders.

Because the fields of psychiatry and epilepsy each has a different focus, clinicians within each field rarely have a comprehensive understanding of the nuances involved in the other area. Just as their training and clinical experience make epileptologists experts in the use of AEDs, psychiatrists who specialize in psychopharmacology have a more complete understanding of the benefits and risks of psychotropic agents. Care is fragmented because few psychiatrists become involved in treating patients with epilepsy, and the same can be said about epileptologists treating psychosis and complicated mood disorders.

Historical Perspective

In the early 1970s, carbamazepine (CBZ) began to be utilized as a treatment for acute mania in patients with rapid-cycling bipolar disorders, following observations that patients with seizure disorder and manic-depression showed an unmistakable improvement in their mood disorder (1). Subsequently, following similar observations, the U.S. Food and Drug Administration (FDA) approved divalproex sodium for the treatment of acute mania and, more recently, lamotrigine (LTG) as maintenance treatment for bipolar disorder (2). Although
these three AEDs are the only ones with an FDA indication for the treatment of a psychiatric disorder, virtually every other new AED has received claims of efficacy for some psychiatric symptoms or disorders.

The inadequacy of current treatment options for psychiatric disorders necessitates taking a chance on unproved treatments. Even with scores of newly available antidepressants, anxiolytics, and antipsychotics, existing psychotropic drugs are not fully effective or well tolerated by most patients. Large-scale studies show remission rates of less than 30% with a commonly used selective serotonin reuptake inhibitor (SSRI) after 8 weeks of treatment (STAR*D) (3,4) and both poor response and high discontinuation rates with the atypical neuroleptics (Clinical Antipsychotic Trials of Intervention Effectiveness [CATIE]) (5). Lithium, the gold standard for treatment of bipolar disorder, proves ineffective in or is poorly tolerated by more than 50% of patients. The use of AEDs in psychiatric disorders is also based on the belief that there are shared biological mechanisms involved in epilepsy and these disorders.

Considerable research, and theorizing, has focused on the pathology underlying different types of seizure disorders and the mechanisms of action of the AEDs. Similar efforts have been made to understand psychiatric disorders and their treatment with psychotropic agents. Serotonin (5-HT), dopamine (DA), and norepinephrine (NE) are the known targets of antidepressants and antipsychotics, whereas γ-aminobutyric acid (GABA) appears to mediate the effects of benzodiazepine anxiolysis. In recent years, an increasing body of literature has shown the pathogenic role played by these three neurotransmitters in both mood disorders and epilepsy, particularly serotonin and NE. For example, the role of 5-HT in human epilepsy has been recently identified in a positron emission tomography (PET) study of patients with temporal lobe epilepsy using the 5-HT_1A receptor antagonist ([18F] trans-4-fluro-N-2-[4-(2-methoxyphenyl)piperazin-1-yl]ethyl-N-(2- pyridyl) cyclohexanecarboxamide). These patients were found to have a reduced 5-HT_1A binding in mesial temporal structures ipsilateral to the seizure focus in patients with and without hippocampal atrophy (6). Likewise, Sargent et al. demonstrated reduced 5-HT_1A receptors binding potential of values in frontal, temporal, and limbic cortex with PET studies using [11C]WAY-100635 in both unmedicated and medicated depressed patients compared with healthy volunteers (7). (This topic is reviewed in great detail in Chapters 3 and 6.) It is therefore not surprising that drugs that work as mood stabilizers, that is, prevent the recurrence of depressive and manic episodes, often are the same drugs that act as AEDs.

Likewise, glutamate, glycine, GABA, and ion channels are increasingly of interest to psychiatrists. Particular attention is directed at the antidepressant-like effects of brain-derived neurotrophic factor (BDNF). Upregulation of BDNF messenger ribonucleic acid (mRNA) and its receptor in rats exposed to electroconvulsive seizure (ECS) and antidepressants suggests involvement of this protein in depression and stress (8,9). Elevations in BDNF protein in brain are consistent with antidepressant efficacy (10). Long-term electroconvulsive therapy (ECT), like antidepressant drug therapy, increases BDNF (11). Long-term ECT or antidepressant medication blocks restraint-induced stress downregulation of BDNF mRNA in the hippocampus (9,12). This effect is not observed when other types of psychiatric drugs are used. In addition, long-term ECT and antidepressant drug therapy enhance induction and prolong expression of BDNF. Whether this results in neuroprotection against stress-related degeneration remains to be proved conclusively, but the evidence is highly suggestive that it does.

Psychotropic Effects of Antiepileptic Drugs

AEDs are “dirty,” or alternatively “broad spectrum,” drugs in the sense that they target multiple systems in the central nervous
system (CNS). Since even experts are unable to fully explain why they produce anticonvulsive effects, it is even more difficult to identify the mechanisms involved in their psychoactive effects. Several animal models of epilepsy have provided evidence of the impact of AEDs on several neurotransmitter systems. For example, AEDs with established psychotropic effects (CBZ, valproic acid [VPA], and LTG) have been found to cause an increase in 5-HT (13,14,15). In fact, the anticonvulsant protection of CBZ can be blocked with 5-HT depleting drugs in genetically epilepsy-prone rats (GEPRs) (16). Likewise, Clinckers et al. investigated the impact of oxcarbazepine (OXC) infusion on the extracellular hippocampal concentration of 5-HT and DA in the focal pilocarpine model for limbic seizures (17). When OXC was administered together with verapamil or probenecid (so as to ensure its passage through the blood-brain barrier), complete seizure remission was obtained, associated with an increase in 5-HT and DA extracellular concentrations. Furthermore, it has been suggested that the anticonvulsant effect of the vagal nerve stimulator in the rat could be mediated by noradrenergic and serotonergic mechanisms, as deletion of noradrenergic and serotonergic neurons significantly prevents or reduces the anticonvulsant effect of VNS against electroshock- or pentylenetetrazol-induced seizures. It could be speculated that the effect of VNS on the locus coeruleus and raphe may be responsible for its antidepressant effects identified in humans (18,19,20).

The clinician needs to be mindful of the fact that neither the therapeutic nor the untoward psychiatric effects of AEDs are necessarily related to their anticonvulsant properties. For example, in placebo-controlled trials of vigabatrin (VGB), the rate of depressive episodes in patients treated with the drug was 12.1%, versus 3.5% in the placebo group (21,22). However, analysis of the data (23) showed that none of the patients with a depressive episode secondary to the introduction of VGB had become seizure-free.

AEDs can have both positive and negative effects on psychiatric symptoms. In some cases, they can even cause seizure. Sedation, weight gain, and cognitive impairment are the most common side effects. These are some common side effects that can have a major impact on compliance. Our knowledge of the negative psychiatric effects of conventional AEDs is largely empirical, based on uncontrolled data.

There is also a paucity of controlled trials in patients with epilepsy. Even when studies are available in an epilepsy population, it is often difficult to ascertain whether improvements in psychiatric symptoms in response to administration of an AED represent a direct psychotropic effect or favorable effect in response to improvement in seizures.

It has to be assumed, on the basis of case reports and uncontrolled trials, that any AED can be beneficial to some patients. However, the percentage of individuals who may benefit may be too small for the efficacy to be proved in large controlled trials; hence the paradox that some drugs are extensively used although they are not FDA approved. The list of disorders includes anxiety, depression, impulse control, self-mutilation, and substance abuse. Some of the more common psychiatric applications of and psychiatric side effects of AEDs are listed drug-wise in the subsequent text.

Benzodiazepines

Benzodiazepines are sometimes used in psychiatric disorders for their anxiolytic properties. They are commonly used in the acute treatment of seizure exacerbations or status epilepticus but are considered to be less desirable for long-term daily therapy because of their adverse effects, such as tolerance, and the availability of many alternatives. Although benzodiazepines are often used for their sedating or anxiolytic properties, they may cause a paradoxical disinhibition syndrome characterized by agitation, aggressiveness, irritability, and hyperactivity (24). Special care should be taken to avoid abrupt
withdrawal, because severe seizure exacerbation and mental status changes, including even psychosis, may occur (25).

Valproate

Valproate is FDA approved for the treatment of acute mania and mixed states. Up to 75% of patients with acute mania respond to valproate. It is also clearly effective as a maintenance treatment, but the pivotal clinical trial for this indication failed, mainly because of misjudgments in its design. The combination of divalproex with the high potency DA receptor antagonist haloperidol has been linked to the development of encephalopathy. It is not clear if the same risk exists when the drug is used in combination with the newer antipsychotic agents.

Bowden et al. (26), in a 12-month randomized controlled trial, compared divalproex (n = 187), lithium (n = 90), and placebo (n = 92) for maintenance treatment of bipolar disorder. The time for development of any mood episode was similar in all three groups, with a trend favoring divalproex over lithium (p = 0.06). The median time to survival without a mood episode was 40 weeks for divalproex, 24 weeks for lithium, and 28 weeks for placebo. Uncontrolled studies also suggest that valproate is effective for the maintenance treatment of pediatric bipolar disorder (27), for the treatment of bipolar II depression (28), for the treatment of agitation (29) but not aggression (30) associated with dementia, and for the treatment of impulsive aggression (31). In an 18-month open-label study comparing divalproex and lithium for maintenance treatment of pediatric bipolar I and II disorders, median (± SE) time for the occurrence of mood symptoms was 114 ± 57.4 days with divalproex therapy and 112 ± 56 days with lithium (27). This was a negative outcome with respect to the study’s aim, which was to determine whether divalproex would be superior to lithium for this purpose, and not to establish the efficacy of either drug in comparison with placebo.

Winsberg et al. (28) reported a 57% decrease (from 22.2 to 9.6) in the mean Hamilton Rating Scale for Depression (HAM-D) score of 19 patients with bipolar II depression who were given divalproex for 12 weeks. However, there was no control or comparator group in this open-label study. Porsteinsson et al. (29) evaluated divalproex in the treatment of agitation associated with dementia in 56 patients in a 6-week controlled trial. Change in the Brief Psychiatric Rating Scale (BPRS) agitation score did not differ significantly in the placebo and divalproex groups (p = 0.08), although post hoc analysis of covariance showed a significant treatment effect (p = 0.05).

Valproate had no significant effect, compared with placebo, on the aggressive behavior of 42 patients with senile dementia in a randomized crossover trial (30). However, in another controlled study, divalproex had a significant antiaggressive effect in patients with Cluster B personality disorders (n = 96) but not in patients with intermittent explosive disorder (n = 116) or post-traumatic stress disorder (PTSD) (n = 34)] (31). Valproate was comparable to phenytoin and superior to placebo or CBZ in a controlled trial evaluating treatment for impulsive aggression (discussed in the “Phenytoin” section) (32).

Open trials have suggested efficacy of VPA in the treatment of panic disorder. In a study of 13 patients with panic disorder who had failed to respond to standard antipanic pharmacologic therapy and behavior therapy, Baetz and Bowen found a significant improvement in panic symptoms and associated symptoms of depression and anxiety in 10 patients (34). In a separate open trial of 12 patients with panic disorder, Woodman and Noyes also found a moderate-to-marked improvement in all patients, 11 of whom elected to remain on this drug for 6 months (34). Clearly, these data will need to be replicated under double-blind placebo controlled conditions.

Although the psychiatric literature is replete with studies of the psychotropic properties of valproate (divalproex), controlled
trials of its psychotropic effects in patients with epilepsy are lacking.

Phenytoin

The use of phenytoin to treat bipolar disorder has only recently been evaluated. Phenytoin has been studied in controlled trials of acute therapy for bipolar mania (35), maintenance treatment of bipolar disorder (36), and even treatment of major depressive disorder (37) or impulsive aggression (32).

Patients with bipolar I disorder who were admitted to a hospital for acute mania and had been unable to tolerate or had not responded adequately to previous treatment with conventional mood stabilizers (lithium, CBZ, and valproate) were randomized to treatment with phenytoin (n = 6) or placebo (n = 6) for 5 weeks, in addition to individualized haloperidol regimens (35). At week 5, the mean daily haloperidol dose was 16.8 mg, and the mean phenytoin level was 21.4 μg per mL. Patients receiving phenytoin showed statistically significant improvement in the BPRS scores compared with placebo recipients at week 3 (p = 0.03), week 4 (p = 0.02), and week 5 (p = 0.01), but their scores on the Young Mania Rating Scale (YMRS) did not differ significantly from those of the placebo recipients. In another study (36), the prophylactic effect of phenytoin was examined in 23 patients with bipolar I disorder who had had inadequate prophylaxis with conventional mood stabilizers. Patients were given either phenytoin or placebo in addition to their ongoing prophylactic treatment and those who completed 6 months without a relapse crossed over to the other treatment. Seven patients crossed over after the first 6 months, and thirty 6-month observation periods (14 for phenytoin and 16 for placebo) were evaluated. A significant prophylactic effect of phenytoin was demonstrated by the Cox F-test for comparison of survival in the two groups (F[6,18] = 3.44, p = 0.02). In a 6-week trial, phenytoin (dose adjusted to blood levels of 10 to 20 μg per mL, n = 14) and fluoxetine (20 mg per day, n = 14) demonstrated similar efficacy in treatment of major depressive disorder (37). Twelve patients in each group improved more than 50% on the HAM-D.

Phenytoin (n = 7), CBZ (n = 7), valproate (n = 7), and placebo (n = 8) were compared in a parallel-group study of patients with impulsive aggression (32). The aggression score, a global severity index from the Overt Aggression Scale, was significantly lower with phenytoin or valproate treatment compared with placebo (p = 0.001 for each comparison), but the difference between the aggression scores of carbamazepine-treated patients and placebo recipients did not reach statistical significance.

An older, review article–based literature (38) promotes the notion of a correlation of phenytoin with depressive symptoms. Although a direct negative psychotropic effect may be responsible, distress over developing adverse cosmetic effects may also play a role. Dose-related sedation and delirium, chronic encephalopathy (39), and agitated psychosis (40) have been described. As with many AEDs, adverse behavioral effects have been described in children (41).

Carbamazepine

CBZ was approved as a treatment for acute mania in 2004, decades after it was recognized as an effective alternative to lithium in the management of bipolar illness. en-label trials suggest that CBZ is effective in the prophylaxis of bipolar disorder or acute mania but may be less effective than valproate or lithium (24,25,42). This together with CBZ’s side effect profile may be responsible for the much more frequent use of valproate for bipolar disorder. In one study (43), 171 patients with bipolar disorder were randomized to prophylactic treatment with CBZ (n = 85) or lithium (n = 86) and followed up for 30 months. A subgroup analysis showed that lithium was superior in preventing episodes requiring hospitalization in patients described as
having classical bipolar disorder (bipolar I disorder without mood-incongruent delusions and without comorbidity), but there was a trend favoring CBZ for patients with nonclassical bipolar disorder (all patients other than those with classical bipolar disorder). A 2-year controlled study compared CBZ (n = 50) and lithium (n = 44) for prophylaxis in patients with bipolar disorder (44). Lithium appeared to be more effective than CBZ in patients with a hypomanic or manic index episode (p < 0.01). An analysis of outcome based on the type of bipolar disorder was not possible because patients with bipolar II disorder were not equally distributed between the treatment groups, and only 10 patients with rapid cycling were included in the study. In two multicenter studies (45,46), 443 patients with bipolar disorder whose current episodes were manic or mixed were randomized to treatment with extended-release CBZ (n = 225) or placebo (n = 220) for a 21-day double-blind evaluation following a 7-day single-blind placebo lead-in period. The intention-to-treat population for the primary efficacy end point (change from baseline to last observation on the YMRS) included 192 patients in the first study and 235 patients in the second. In both studies, the decrease in YMRS total score was significantly greater with CBZ treatment than with placebo (first study, p = 0.033; second study, p < 0.0001). In the second study, CBZ had a large effect size (0.85).

The side effect and safety profile of CBZ are well known to neurologists, who are comfortable prescribing the drug. The multiple warnings associated with the drug and the necessity for regular laboratory testing, coupled with its status as an off-label drug until recently, has limited the acceptance of CBZ by psychiatrists. The initial concerns over the hematologic adverse events of this drug have diminished and the old practice of checking complete blood counts (CBCs) on a regular basis is discouraged today.

In rare cases, CBZ has been used to treat aggressive behavior and facilitate sedative withdrawal, but no controlled studies are available to date to establish its efficacy in this domain.

There have been no systematic evaluations of this AED in patients with epilepsy and comorbid mood disorders.

Tiagabine

Tiagabine has also been touted as a treatment for generalized anxiety disorder, PTSD, insomnia, cocaine addiction, and impulse control disorder (without much evidence to support widespread efficacy in those conditions). However, reports that use of tiagabine is associated with new-onset seizures and status epilepticus in patients without epilepsy has shifted the perceived risk–benefit ratio of such off-label use. Dose appears to be a predisposing factor in the development of seizures, although they have been reported in patients taking doses as low as 4 mg per day. Some seizures occurred near the time of a dose increase, even after periods of prior stable dosing.

One notable series in epilepsy suggested that tiagabine promotes improved mood and adjustment independent of seizure control (47). Mental status changes due to tiagabine-induced generalized nonconvulsive status epilepticus may occur even in patients with underlying partial seizure disorders (48).

Only gold members can continue reading. Log In or Register to continue