Francesco Monaco MD*
Marco Mula MD**
* Chief and Professor of Neurology, Department of Neurology, University of Piemonte Orientale “Amedeo Avogadro,” Novara, Italy
** Research Fellow, Department of Neurology, University of Piemonte Orientale “Amedeo Avogadro,” Novara, Italy
The use of antiepileptic drugs (AEDs) in the treatment of psychiatric disorders has reached a new phase of clinical and theoretical interest (1). Initially, as various became available, anecdotal observations and clinical studies suggested that they could be associated with either improvement or worsening of mood and psychosis in patients with epilepsy (2). Later assessments of the behavioral effects of AEDs yielded a series of clinical trials on psychiatric patients who had no evidence of seizure disorders (3). Phenytoin (PHT) (Chapter 58) was one of the first AEDs to be used in psychiatric disorders (4). Conversely, to our knowledge, the other hydantoins (mephenytoin and ethotoin) were very seldom employed for this use (5), and no significant data are available in the literature.
SPECTRUM OF EFFICACY IN PSYCHIATRIC DISORDERS
In 1943, Kalinowsky and Putnam (4) treated 41 patients with schizophrenia with an oral dose of 700 mg/day of diphenylhydantoin (the older term for PHT) sodium, and these investigators reported substantial improvement in excitement, agitation, hyperactivity, and aggression in these patients. In 1945, Freyhan (6) treated 25 patients with schizophrenia with 300 to 600 mg of PHT daily and observed encouraging results, mainly in the alleviation of catatonic excitement. In 1946, Kubanek and Rowell (7) administered PHT (300 to 700 mg/day) to 35 chronically psychotic hospitalized patients, eight of whom showed marked improvement in their behavior. Taking into account all the patients studied in the three (uncontrolled) studies, there was an overall improvement rate of 45%, a rather impressive figure considering the severe and long-term nature of the disturbances shown by most of these patients. In 1969, Haward (8), in the first double-blind, placebo-controlled study performed for the use of PHT in the treatment of chronic schizophrenia, reported that both aggressive behavior and assaultive ward behavior were reduced in the 20 patients studied when PHT (≤200 mg/day) was administered.
In 1974, Simopoulos et al. (9) conducted a double-blind, placebo-controlled trial to clarify the effect of PHT in patients with chronic schizophrenia, with particular interest on the effect of the drug on hostility. In this study, 26 men and 50 women (age range, 20 to 59 years) with a primary diagnosis of schizophrenia, who had been hospitalized and had been taking continuous phenothiazine medication for at least 6 months, were selected. PHT was administered according to the following dosage: 375 mg/day for the first 2 weeks, 500 mg for the third and fourth weeks, and 625 for the remaining 4 weeks. Measures used were the Brief Psychiatric Rating Scale, the Nurses Observation Scale for Inpatient Evaluation, and the Crownsville Psychiatric Scale (parts I and II). Patients receiving both PHT and placebo worsened on both measures (in this study, all phenothiazines were discontinued on the first day of PHT administration); differences in favor of PHT were found for the Brief Psychiatric Rating Scale variables of hostility and thought content. The optimal dosage appeared to be <500 mg/day (because retardation and blunted affect were more predominant at the highest dose levels), and the indication seemed to be a high initial disorder score, especially for hostility. The therapeutic effects of PHT were of short duration and occurred in the early phases of treatment. These effects did not continue into the later period of treatment, when doses were >525 mg/day. The same authors (10) conducted a double-blind study to evaluate the response of patients with chronic schizophrenia to a combination of PHT and neuroleptics, and results were the same as those of the previous study.
The early studies, particularly in the 1940s, were largely based on uncontrolled clinical observations. In 1943, Kalinowsky and Putnam (4) reported that eight of nine patients with mania improved during treatment with PHT, as did three of five patients in the depressed phase. In 1945, Freyhan (6) reported that five of six patients were doubtful responders to treatment with PHT. However, he reported one patient who appeared to be an unequivocal responder. This patient's severe manic illness improved on four occasions in response to PHT and showed repeated exacerbations when the drug was discontinued. In 1967, both Jonas (11) and Turner (12) studied the effects of PHT in less severely ill outpatients with mood disturbances who could be classified as having neurotic or nonendogenous depressive illness. In 1983, Post and Uhde (2), in a double-blind, placebo-controlled trial with three different AEDs (carbamazepine, PHT, and valproic acid) in patients with recurrent manic illness, showed a notable response to carbamazepine but showed no evidence of response to treatment with the PHT or valproic acid. From this finding, these investigators theorized a different biochemical mechanism of action and neural substrate for carbamazepine and PHT, in spite of the similar clinical and experimental spectrum of clinical efficacy. Compared with PHT, carbamazepine is actually more effective in inhibiting amygdala-kindled afterdischarges and seizures and in raising the threshold for amygdala seizures (13, 14, 15, 16, 17).
As a new phase of interest in the use of PHT for mania has begun (1,18,19), and the cognitive side effects of PHT have been reassessed favorably (20,21), it may be of clinical value to add PHT to the therapeutic regimen in some patients with bipolar illness. Mishory et al. (19) conducted a controlled study on the use of PHT for mania and used an add-on design with ongoing neuroleptic treatment; 39 patients with either bipolar I disorder, manic type, or schizoaffective disorder, manic type, entered a 5-week, double-blind, placebo-controlled trial of haloperidol plus PHT versus haloperidol plus placebo. PHT or placebo was begun at a dose of 300 mg/day, and the dose was increased to 400 mg after 4 days. Of 39 patients, 30 completed at least 3 weeks, and 25 completed 5 weeks of treatment. Improvement with PHT was observed on the Brief Psychiatric Rating Scale and Clinical Global Impression scores in patients with bipolar mania but not with schizoaffective mania. This conclusion should be interpreted cautiously, because the Young Mania Rating Scale scoring did not show significant improvement. The study, as underlined by the authors themselves, is limited by the small sample size, by the finding that the schizoaffective group included more patients nonresponsive to previous treatment, and by the situation that all patients received haloperidol, the plasma levels of which have been demonstrated to be reduced by PHT (22).
Aggression and Behavior Disorders
Many different drugs have been advocated to be useful in treating aggressive patients (23). Aggression is a symptom, and it is important for the clinician to clarify underlying causal agents or psychological factors before treating the patient. For example, violence can exist in association with a psychotic condition such as catatonic excitement or an acute toxic psychotic state. Organic factors such as latent epilepsy or, in rare instances, overt partial temporal lobe epilepsy, with complex symptoms, may or may not underlie such conditions. Episodic rage, fear attacks, memory disturbances, fugues, and impulsive actions are typical of partial epilepsy and are strongly reminiscent of various aspects of other illnesses such as character disorders of hysterical and emotionally unstable type (2,24,25). Because of the similarity in clinical features, investigators have suggested that AEDs be used in treating such patients (26).
PHT has long been advocated to reduce aggressiveness in psychiatric patients (27,28). In 1945, Freyhan (29) reported good results using PHT in adult patients with agitated depression. In 1967, Klein and Greenberg (26) studied the efficacy of PHT in hospitalized older adolescents and adult psychiatric patients with a main diagnosis of agitated depression and behavior characterized by impulsive actions and rage attacks unrelated to appropriate stimuli; 13 patients were treated with PHT as the sole psychotropic agent, and 15 others were treated with PHT added as an adjunctive medication to another psychotropic agent. All patients received 100 mg of PHT three times a day for at least 2 weeks; control measures were not taken, nor there was any attempt to mask the identification of the drug. This study showed no beneficial effect of PHT, whether administered alone or in association with phenothiazines or imipramine. The only patient who had a maximum clinical response to PHT had a previous diagnosis of “psychomotor” epilepsy. Several controlled studies have shown that the drug is no more effective than placebo or other medications (30, 31, 32).
In 1968, Boelhouwer et al. (33), in a double-blind study, showed that a combination of benzodiazepines and PHT was useful in aggressive patients, and another controlled study showed that PHT alone was useful for treating anxiety, irritability, and anger in neurotic patients (34). The complexities of clinical diagnosis and evaluation are further emphasized by isolated reports of worsening of episodic disorders by PHT and chlordiazepoxide (35). Many of these studies were carried out on criminals, in whom the degree of aggressiveness was not specified and was not associated with a well-defined psychiatric disorder (31).
To date, there is more negative than positive evidence in the literature on this topic. For example, in a controlled trial, Yudofsky et al. (36) demonstrated no efficacy for PHT in aggression. More recently, Barratt (37) theorized that the equivocal findings in the early studies resulted from poor criterion measures, inappropriate inclusion and exclusion criteria
for patients, and a different nosologic basis for classifying various types of aggression. He therefore proposed that aggression could be divided into three broad categories: (a) medically related (aggression is a symptom related to a psychiatric, neurologic, or other medical disorder); (b) premeditated or planned (the aggressive act is an instrumental response); and (c) impulsive. Barratt hypothesized that selected anticonvulsants (e.g., PHT) would have a therapeutic effect on impulsive aggression. For these reason, Barratt et al. conducted a double-blind, placebo-controlled, crossover study (38). Sixty inmates were divided into two groups on the basis of committing impulsive aggressive acts or premeditated aggressive acts committed while the inmates were in prison; medical aggression was ruled out by subject selection. As hypothesized, PHT (300 mg/day) significantly reduced impulsive aggressive acts but not premeditated aggressive acts. However because aggression is a complex phenomenon that is undoubtedly multifactorially determined, it appears very difficult to evaluate a specific antiaggressive pharmacotherapeutic agent.
The pharmacotherapy of anorexia nervosa and bulimia is mainly based on antidepressant drugs (39,40); alternative drugs therapies include anticonvulsants (PHT and carbamazepine) and lithium (41). Evidence for the use of AEDs in eating disorders is based on the early observations of Lundberg and Walinder (42) who, in 1967, reported on five patients with anorexia nervosa who showed clear clinical signs of underlying neurologic disease. In 1974, Green and Rau (5) speculated that some patients with disturbances in eating behavior could suffer from hypothalamic neurologic dysfunction. Whether such dysfunction could be directly caused by lesions in the hypothalamus or from the feedback effects of cortical lesions was unknown. Furthermore, persons with severe eating disorders resulting from cerebral lesions could be expected to have abnormal electroencephalographic (EEG) patterns, and these patients could be responsive to AEDs. In the study by Green and Rau (5), 10 patients with symptoms of compulsive eating were treated pharmacologically with PHT at doses ranging from 200 to 400 mg/day. Nine of these patients were treated successfully, with normalization of eating behavior and weight loss in the overweight group. Two patients developed allergic reactions; one dropped out of the study, and the second patient was less successfully treated with mephenytoin (300 mg/day). For these two patients, interruption of medication was shortly followed by a recurrence of symptoms. In the patient who could resume PHT treatment, the compulsive eating behavior decreased significantly. All these patients had abnormal EEGs (six of them showed 14- and six-per-second positive spiking in the temporal and occipital areas, and four of them showed spiking in temporal and occipital areas), indicating that a neurologic dysregulation may have been an etiologic factor.
The foregoing study stimulated Wermuth et al. (43) to test the efficacy of PHT treatment in a controlled clinical trial. Nineteen subjects completed a 12-week, double-blind, crossover study comparing PHT (300 mg/day) with placebo. The data were first analyzed by comparing the mean number of binges per week estimated by subjects (historical control) and those during both experimental periods (placebo-PHT and PHT-placebo sequence). In the placebo-PHT sequence, the number of binges per week during the placebo period was not significantly different from the historical control. However, there was a significant decrease during PHT treatment. Nevertheless, the efficacy of PHT was considerably less in this study than in the previous one, in which 90% of the subjects returned to normal eating habits. As suggested by the authors, this discrepancy could have resulted from the inclusion of more seriously affected subjects (in this study, patients who had frequent binges were intentionally selected; those with milder forms were therefore excluded) and the shorter duration of treatment. Results were further confounded by the effect of treatment sequence, in which there was an unexpected persistence of treatment effect in the placebo period for subjects in the PHT-placebo sequence. The action of PHT ended soon after the drug was discontinued, and during placebo period, virtually all subjects had no detectable plasma drug concentrations. Thus, a prolonged drug effects was most unlikely, so it could be speculated that a learning effect was created once the pattern of binge eating was interrupted.
MODE OF USE
The previously described data provide the tempting suggestion that PHT, in carefully adjusted doses, may be of therapeutic value at least in some psychiatric disorders, particularly for its lack of sedative action. An initial dose of 300 mg/day seems to be useful in mood disorders and aggressive behavior. In some patients, such as those with schizophrenia, particularly those with high levels of hostility, the initial dose of 300 mg/day of PHT must be raised 100 mg/wk until a good clinical response is obtained; the optimal dose appears to be 500 mg/day. However, a good clinical response has been described mainly for patients with EEG abnormalities (34).
Long-term prophylactic studies of PHT in patients with bipolar disorder and with schizophrenia will need to take into account the danger of gingival hyperplasia, leukopenia, or anemia and the risks of toxicity resulting from nonlinear kinetic and pharmacokinetic and pharmacodynamic interactions with other drugs in such combined therapies. Moreover, patients with organic mania induced by PHT are described in the literature (44, 45, 46). For clinicians managing psychiatric disorders, PHT pharmacokinetic interactions with psychotropic drugs are quite relevant (47). PHT is 90% metabolized by the cytochrome P450 enzyme system, mainly by the subfamily CYP2C9 and to a lesser extent also
by CYP2C19 (Chapter 52), and it is a potent inducer of the CYP3A4, epoxide hydrolase, and the uridine diphosphate glucuronosyltransferase enzyme system (48). Chapter 59 deals with the pharmacologic interactions of PHT, so here we simply mention some aspects of the problem.
As far as antipsychotic drugs are concerned, an increase in PHT plasma concentrations and neurotoxicity was reported after the addition of typical phenothiazines (thioridazine, chlorpromazine, and prochlorperazine) (49,50), whereas plasma concentrations of butyrophenone (as haloperidol) have been demonstrated to be reduced by PHT (21). For atypical antipsychotic drugs, anecdotal reports have described a reduction of clozapine plasma levels in patients treated with PHT (51), as well as the appearance of extrapyramidal symptoms during the coadministration of PHT and risperidone (52). Among antidepressant drugs, fluoxetine (53), viloxazine (54), trazodone, and tricyclic antidepressants (55) demonstrated an increase in PHT plasma levels, whereas PHT was shown to induce paroxetine metabolism and to decrease its plasma concentrations by about 50% (56). In conclusion, PHT, a powerful AED of the first generation, shows many properties, other than its anticonvulsant action, that could have relative efficacy in some psychiatric disorders.