Phenytoin: an anti-bipolar anticonvulsant?

Transcription

1 International Journal of Neuropsychopharmacology (2006), 9, Copyright f 2005 CINP doi: /s Phenytoin: an anti-bipolar anticonvulsant? Yuly Bersudsky Ministry of Health Mental Health Center, Faculty of Health Sciences, Ben Gurion University of the Negev, Beersheva, Israel TRENDS AND PERSPECTIVES Abstract Phenytoin, a classical anticonvulsant has been little studied in bipolar disorder. We completed a trial of phenytoin in mania and schizoaffective disorder, manic type. Thirty-nine patients entered a 5-wk doubleblind controlled trial of haloperidol+phenytoin vs. haloperidol+placebo; 30 patients completed at least 3 wk; 25 completed 5 wk. Significantly more improvement was observed in those patients receiving phenytoin. Phenytoin has not previously been studied prophylactically in bipolar patients. Bipolar patients were studied who had at least one episode per year in the previous 2 yr despite ongoing prophylaxis. Patients were stable for a mean of 4 months (range 1 13) before entering the study. Phenytoin or placebo was added to their current therapy in a double-blind cross-over design for 6 months in each phase. Thirty observation periods of 6 months each were studied for 23 patients. Three patients had relapse on phenytoin and nine had relapse on placebo. There was a significant prophylactic effect of phenytoin in bipolar disorder [Cox s F test for comparing survival in two groups: F(6, 18)=3.44, p=0.02]. This study suggests prophylactic effects of add-on phenytoin in bipolar illness. However, the number of patients was small and confirmation is necessary. Lamotrigine has recently been reported to have antidepressant effects. In the past, small studies showed antidepressant effects for carbamazepine and valproate. To determine if such effects could be a class property of other voltage-activated sodium channel blockers such as phenytoin, we performed a double-blind controlled trial of phenytoin vs. fluoxetine in unipolar depression. Thirty-three depressed patients entered the study, and 28 completed at least 3 weeks and were included in data analyses. Weekly Hamilton Depression Scales for 6 wk showed no difference between fluoxetine and phenytoin. Clearly pharmaceutical company funding for clinical trials or advertising for phenytoin is minimal and this must be taken into account in evaluating literature on phenytoin vs. other drugs. The present data suggests that effects on affective disorder may be common to many anticonvulsants. Received 17 March 2005; Reviewed 7 May 2005; Revised 28 July 2005; Accepted 29 July 2005; First published online 5 October 2005 Key words: Antidepressant, controlled trial, mania, phenytoin, prophylaxis. Introduction The mechanism of action of anticonvulsants in epilepsy is not completely known (Macdonald and Kelly, 1995). Clinical effects are also not well defined and often a trial-and-error method of adding various anticonvulsants to clinically non-responsive patients in a polypharmacy approach yields unpredictably positive results (Post, 1989). As in bipolar disorder, many patients are unresponsive to monotherapy and even rational polypharmacy is often unsuccessful in attaining complete seizure control. Address for correspondence : Y. Bersudsky, M.D, Ph.D., Beersheva Mental Health Center, PO Box 4600, Beersheva, Israel. Tel. : Fax : yuly@bgumail.bgu.ac.il One startling point of contrast between anti-bipolar and anticonvulsant treatment has been phenytoin. Phenytoin was the first non-sedative anticonvulsant and is still the most widely used anticonvulsant worldwide. Phenytoin was reported by Kalinowsky and Putnam (1943) to be effective in eight manic patients, but Post (1989) reported no effect in five bipolar patients unresponsive to carbamazepine. In animal models of bipolar treatment, phenytoin was often used as a negative control compared with antibipolar anticonvulsants such as carbamazepine (Post, 1989). New anticonvulsants such as lamotrigine and topiramate have come under investigation as possible antimanic compounds. While numerous biochemical properties have been found for carbamazepine, valproate, lamotrigine, and topiramate, all seem to

2 480 Y. Bersudsky share with phenytoin, the classical anticonvulsant, powerful inhibition of voltage-activated sodium channels (Dichter, 1998). It thus seems of considerable theoretical importance to determine whether phenytoin is antimanic. Early uncontrolled reports of improvement of mania with phenytoin exist (Kalinowsky and Putnam, 1943). Recently, its cognitive side-effects have been re-assessed favourably (Devinsky, 1995). Because individual patients sometimes respond to one agent of a class and not to another, it could be of clinical as well as theoretical value to add phenytoin to our therapeutic armamentation in bipolar illness. We, therefore, decided to conduct a trial of phenytoin in acute mania, in bipolar prophylaxis and in depression. Phenytoin in acute mania The study (Mishory et al., 2000) was approved by our Helsinki Committee and all patients gave informed written consent. Newly admitted patients to the Beersheva Mental Health Center could participate if they met DSM-IV criteria for mania or schizoaffective disorder, manic type and had no serious physical illness. Generally patients who had side-effects or inadequate response to mood-stabilizing treatment of previous episodes were referred to the study. Patients admitted to the study were started on haloperidol at doses of physicians discretion. Trihexyphenidyl was available as necessary for extrapyramidal symptoms and benzodiazepines for sleep. Phenytoin or placebo was begun at doses of 300 mg/d after completion of screening physical and laboratory examination. Dose was increased to 400 mg after 4 d, with the first blood level being drawn on the third day after that. Patients received phenytoin or identical capsules of placebo. Phenytoin was begun in all patients within 5 d of onset of haloperidol treatment. Weekly ratings by a psychiatrist blind to the study drug were conducted using the Brief Psychiatric Rating Scale (BPRS; Overall and Gorham, 1962), the Young Mania Rating Scale (YMS; Young et al., 1978), and the Clinical Global Impression (CGI; Guy, 1976). Weekly blood was drawn for phenytoin levels and the results reported to the control psychiatrist who created dummy levels for placebo patients and reported results to the treating physician, who adjusted the dose of phenytoin accordingly. Thirty-nine patients entered the study over 1 yr; 30 completed at least 3 wk; and 25 completed 5 wk. Patients completing at least 3 wk were included in the data analysis on a last value carried forward basis. BPRS total score Baseline * * * Time (wk) Figure 1. The effect of phenytoin ( &, n=6) or placebo ( 2, n=6) addition to haloperidol in mania. Figure 1 illustrates the results for the bipolar subgroup. Results for the schizoaffective patients were less marked (Mishory et al., 2000). Analyses (Greenhouse Geisser corrected) were three-way MANCOVA, covariance for baseline, for diagnosis, treatment and time. Diagnosis, treatment and time showed a significant three-way interaction (F=5.09, d.f.=2.3, 60.6, p=0.006), with a significant two-way interaction between treatment and time (F=3.83, d.f.=2.3, 60.6, p=0.02) and significant two-way interaction between treatment and diagnosis (F=4.42, d.f.=1, 25, p=0.046). Bipolar manic patients showed significant two-way interaction for treatment and time (F=5.6, d.f.=1.7, 17.4, p=0.016). Differences between phenytoin and placebo were significant only for bipolar manic patients from week 3 to week 5 (Tukey HSD post-hoc, p 3 =0.03, p 4 =0.02 and p 5 =0.01). Two-way ANOVA for schizoaffective manic patients was n.s. (F=0.31). The present study is limited by small size and diagnostic variability. The fact that all patients received haloperidol makes it more difficult to determine the nature of the phenytoin effect. Phenytoin has been reported to reduce haloperidol blood levels (Young et al., 1978) and it is possible that the effect of phenytoin would be more marked in a placebo-controlled monotherapy design. The mean daily haloperidol dose did not significantly differ between patients taking phenytoin and placebo at any point in the study and rose from 14.4 mg (S.D.=7.6 mg) in week 1 to 16.8 mg (S.D.=7.5 mg) in week 5.

3 Phenytoin: an anti-bipolar anticonvulsant? 481 Phenytoin in bipolar prophylaxis Valproate and carbamazepine treatment of bipolar illness in both the acute and prophylactic phase has been extensively studied, although some have questioned the quality and conclusiveness of the data on carbamazepine since only small trials exist (Bowden, 2001). The pivotal study of valproate prophylaxis of bipolar disorder was inconclusive (Bowden et al., 2000). The study of phenytoin (Bersudsky et al., 2003) was approved by our Helsinki Committee (IRB) and patients gave written informed consent. Patients participated if they met DSM-IV criteria for bipolar disorder I or schizoaffective disorder, and had no unstable physical illness. Patients could enter the study if they had been stably out of hospital for at least 1 month and had inadequate prophylaxis in the past on lithium, carbamazepine or valproate, and had at least one episode per year for the previous 2 yr despite compliance with their mood stabilizer. The last episode could be either manic or depressive. All patients were evaluated by the clinical treating physician at baseline, weekly during the first month and monthly thereafter. Ongoing prophylactic treatment was not changed (lithium, carbamazepine, valproate or neuroleptic). Phenytoin or placebo (as randomized) was added slowly (100 mg/wk) to ongoing anti-bipolar treatment. Blood levels of all anti-bipolar treatments were monitored weekly for the first month and adjusted as necessary. BPRS, YMS, Hamilton Depression Rating Scale (HAMD; Hamilton, 1960), GCI were done by the clinical treating psychiatrist at baseline and once monthly thereafter. Patients who completed 6 months without relapse were crossed over during a month of weekly visits with one drug (phenytoin or placebo) being reduced by 100 mg/wk and the other increased by 100 mg/wk. This was done with individualized packets of capsules, such that the double-blind was maintained. Blood levels of all antibipolar drugs were monitored weekly during this adjustment phase, and doses adjusted as necessary. Monitoring of blood levels was monthly after the adjustment phase. The treating psychiatrist who also performed the rating scales was blind to whether the patient was on phenytoin or placebo. He received from the control psychiatrist a bottle of tablets of phenytoin or placebo, according to a pre-arranged random order. Blood levels of phenytoin were reported by the laboratory to the treating psychiatrist after dummy levels were assigned by the control psychiatrist to patients on Table 1. Clinical events (mania or depression) in phenytoin or placebo-treated bipolar patients Phenytoin treatment group (n=14) Placebo treatment group (n=16) Clinical events 3 (21 %) 9 (56%) Clinically stable patients 11 (79 %) 7 (44%) placebo. The second phase also lasted for up to 6 months of monthly clinical ratings. Four patients dropped out within first 3 wk of the study during the gradual rise of phenytoin dosage. The primary outcome measure (end-point) was the time to event, an affective relapse. For statistical analysis we included all patients randomized to treatment and providing at least one post baseline outcome assessment. Patients who discontinued the study for reasons other than an event were censored from analysis at the point of their discontinuation. A cumulative proportion Kaplan Meier survival curve was generated for the time-to-event data. Differences among treatment groups were compared using Cox s F test for comparing survival in two groups, which may be most powerful for small sample sizes (Cox, 1964; Lee, 1980). Thirty observation periods (6-month phases) were studied for 23 patients (seven patients that were stable in the first phase of study crossed over after 6 months and are counted on both phases of the study) with a total 140 months (survival time) of patient treatment, including 12 clinical events (Table 1). Fourteen observation periods were in the phenytoin treatment group and 16 in the placebo group. Three patients (21.4%) had relapse ( event ) on phenytoin and nine (56.3%) had a relapse ( event ) on placebo. Of the events on placebo, six were manic and three depressive. Of the events on phenytoin, two were manic and one depressive. Seven patients ended one 6-month phase and did not continue because the study terminated and three patients completed the full year with no episodes. There was a significant prophylactic effect of phenytoin in bipolar disorder [Cox s F test for comparing survival in two groups: F(6, 18)=3.44, p=0.02] (Table 1). One female patient complained of slight weakness and sleepiness during both periods of treatments, but her affective state was stable. Three patients on phenytoin reported temporary dizziness that resolved without any changes in the treatment. One patient that

4 482 Y. Bersudsky received phenytoin with lithium developed psoriasislike symptoms and stopped the study as a result. There were two additional cases of dropout without clinical relapse: both of them were stable for 5 months on phenytoin treatment, but requested to stop the study without explanation. This study suggests prophylactic effects of add-on phenytoin in bipolar illness. The patient group was relatively resistant and concomitantly treated with numerous effective mood stabilizers, making the significant added benefit of phenytoin more impressive. However, the number of patients studied was small and the study duration was short. The implications of phenytoin s possible effectiveness in bipolar disorder prophylaxis for the mechanism of action of anticonvulsant anti-bipolar drugs is unclear. Some have claimed that anti-bipolar compounds can be divided into those with predominantly antimanic prophylactic action and those with additional antidepressant anti-bipolar action (Mitchell and Malhi, 2002). Others have suggested that most anticonvulsants have a common mechanism of action both in epilepsy and in bipolar disorder (Levine et al., 2003). However, this makes it hard to explain the trial-and-error, multiple polypharmacy nature of actual clinical practice in both epilepsy and bipolar disorder. Phenytoin in unipolar depression Recently, lamotrigine has attracted considerable interest as an anticonvulsant treatment of bipolar disorder with possible specific activity in the treatment and prophylaxis of the depressive phase. Calabrese et al. (1999) reported acute efficacy of lamotrigine in depressed bipolar patients. Bowden et al. (2003) reported prophylactic efficacy of lamotrigine of depressive episodes of bipolar disorder over 18 months in over 200 patients vs. placebo. Lamotrigine s efficacy in the depressed phase of bipolar disorder raises the critical theoretical question as to whether it is different in this matter from other anti-bipolar anticonvulsants or indeed from the classic mood stabilizer lithium. Carbamazepine, the first anticonvulsant mood stabilizer, was first tried in bipolar disorder because of reports by neurologists of mood improvement in depressed epileptic patients. Only small studies exist of carbamazepine in bipolar or unipolar depression but these are generally positive (Neumann et al., 1984; Post et al., 1986; Small, 1990). Supporting the concept that carbamazepine too has antidepressant properties are the reports that carbamazepine, like lithium, can augment the effects of monoamine blockers in patients who have failed to fully respond to those monoamine blockers (Steinacher et al., 2002). Valproate, a widely used mood stabilizer in the USA today, for which large controlled studies have been conducted with the patented divalproex sodium, has also been studied in depression (Davis et al., 1996; Petty et al., 1999; Sachs, 1996; Schaff et al., 1993; Winsberg et al., 2001). These studies were preceded by early reports of its beneficial effects on mood in epileptic children and adults with depression or irritability. Interestingly, in the 1940s and 1950s numerous reports were published on improvement of depressive symptoms and irritable symptoms in adults and children with epilepsy who were treated with phenytoin. Some controlled studies of phenytoin in depressionrelated syndromes were published (Boelhouwer et al., 1968; Case et al., 1969; Jonas, 1967; Turner, 1967). However, no modern study has been done of the efficacy of phenytoin in DSM-IV-diagnosed major depression. Given the apparent efficacy of other mood stabilizers in depression the possible efficacy of phenytoin would be of significant theoretical importance. Moreover, the treatment of epilepsy is characterized by a trial-and-error polypharmacy approach. Despite the fact that many anti-epileptic compounds have similar modes of action on voltage-activated sodium channels and GABA, it is clinically well accepted that efficacy can be improved by finding the right drug for the right patient and/or adding a second drug when the first is not entirely effective even if they have similar mechanisms of action. Therefore, the possibility that phenytoin may be useful in depression may be of value to some patients who have not responded to other antidepressants. We, therefore, performed a controlled trial of phenytoin vs. fluoxetine in unipolar depression (Nemets, 2005). Ethical considerations made a trial of placebo vs. phenytoin problematic in a disorder such as depression with a clear standard effective therapy. The proposal was approved by the Helsinki Committee. Patients entered the study after baseline physical examination, blood chemistry and hematology, EKG, and written informed consent. Patients with significant risk of suicide, need for hospitalization, or unstable medical illnesses were excluded. All patients met DSM-IV criteria for major depression. After a 3-d washout of any previous medications, patients were randomized to fluoxetine or phenytoin in identical capsules. Each capsule contained 100 mg phenytoin or 7 mg fluoxetine+corn starch. Patients started on one tablet daily and increased every other day until one tablet three times daily with meals.

5 Phenytoin: an anti-bipolar anticonvulsant? 483 HAMD (average±s.d.) Baseline Time (wk) Figure 2. Phenytoin ( & ) vs. fluoxetine ( 2 ) treatment of unipolar major depression. Clearly pharmaceutical company funding for clinical trials or advertising for phenytoin is minimal and this must be taken into account in evaluating literature on phenytoin vs. other drugs. It is difficult to create a level playing field between drugs with and without pharmaceutical company backing in the race for scientific attention. Phenytoin might be a reasonable option, third or fourth, in depressed patients who have not responded to monoamine reuptake blockers or mood stabilizers. Most important research screening for new anti-bipolar drugs should not use phenytoin as a negative control, and a search for a common mechanism of the anti-bipolar anticonvulsants including phenytoin might facilitate progress in this area. Blood levels of phenytoin were taken after 1, 3 and 6 wk and dosage adjusted to achieve blood levels between mg/ml, to a maximum dose of four capsules per day or a minimum dose of two capsules per day. Fluoxetine patients were assigned a dummy blood level by the control psychiatrist to guarantee that fluoxetine patients received a minimum of 20 mg/d. The treating psychiatrist who performed the rating scales was blind to whether the patient was on phenytoin or fluoxetine. Upon entry of a patient into the study the treating psychiatrist received from the control psychiatrist a bottle of tablets of phenytoin or fluoxetine, according to a pre-arranged random order. Patients were evaluated at baseline and weekly thereafter with the 24-item HAMD by an experienced psychiatrist. The trial lasted 6 wk. Thirty-three patients entered the study, and 28 completed at least 3 wk and were included in data analysis (see Figure 2). Twenty-eight patients completed the entire 6 wk. Patients who dropped out after week 3 (three patients) were included in the study as last value carried forward. Figure 2 illustrates the results. There was no difference in overall rate of response or speed of response of phenytoin vs. fluoxetine. Blood levels of phenytoin were mg/ml at week 1, mg/ml at week 3, and mg/ ml at week 6. Dummy levels were similar. The ethical and methodological advantages and disadvantages of a placebo arm in a study of disorders with known effective treatment has been extensively discussed (Garattini et al., 2003). The absence of a placebo arm in our study allows the possibility that neither was more effective than placebo. However, the minimum HAMD score at baseline of 18 and the lack of even any numerical trend in favour of fluoxetine make the latter possibility unlikely. Acknowledgements I thank all my clinical collaborators in these studies: Dr Alex Mishory, Dr Boris Nemets, Dr Yuri Yaroslavsky and Professor R. H. Belmaker. This paper received the Clinical Prize 2004 from the Israel Society for Biological Psychiatry. Statement of Interest None. References Bersudsky Y, Winokur M, Mishory A (2003). Prophylactic effects of phenytoin in bipolar illness. Bipolar Disorders 5, Boelhouwer C, Henry CE, Glueck Jr. BC (1968). Positive spiking: a double-blind control study on its significance in behavior disorders, both diagnostically and therapeutically. American Journal of Psychiatry 125, Bowden C, Calabrese J, McElroy S, Gyulai L, Wassef A, Petty F, Pope Jr. HG, Chou JC, Keck Jr. PE, Rhodes L, Swann A, Hirschfeld R, Wozniak P (2000). A randomized, placebo-controlled 12-month trial of divalproex and lithium in treatment of outpatients with bipolar I disorder. Archives of General Psychiatry 57, Bowden C, Calabrese J, Sachs G, Yatham L, Asghar S, Hompland M, Montgomery P, Earl N, Smoot T, DeVeaugh-Geiss J, and Lamictal 606 Study Group (2003). A placebo-controlled 18-month trial of lamotrigine and lithium maintenance treatment in recently manic or hypomanic patients with bipolar I disorder. Archives of General Psychiatry 60, Bowden CL (2001). Novel treatments for bipolar disorder. Expert Opinion on Investigative Drugs 10, Calabrese JR, Bowden CL, Sachs GS, Ascher JA, Monaghan E, Rudd GD (1999). A double-blind placebo-controlled study of lamotrigine monotherapy in

6 484 Y. Bersudsky outpatients with bipolar I depression. Journal of Clinical Psychiatry 60, Case WG, Rickels K, Bazilian S (1969). Diphenylhydantoin in neurotic anxiety. American Journal of Psychiatry 126, Cox DR (1964). Some applications of exponential ordered scores. Journal of the Royal Statistical Society 26, Davis LL, Kabel D, Patel D, Choate AD, Foslien-Nash C, Gurguis GN, Kramer GL, Petty F (1996). Valproate as an antidepressant in major depressive disorder. Psychopharmacological Bulletin 32, Devinsky O (1995). Cognitive and behavioral effects of antiepileptic drugs. Epilepsia 32 (Suppl. 2), S Dichter MA (1998). Mechanisms of action of new antiepileptic drugs. Advances in Neurology 76, 1 9. Garattini S, Bertele V, Li Bassi L (2003). How can research committees protect patients better? British Medical Journal 326, Guy W (1976). Assessment Manual for Psychopharmacology. US Dept of Health, Education and Welfare publication, Hamilton M (1960). A rating scale for depression. Journal of Neurology, Neurosurgery and Psychiatry 23, Jonas AD (1967). The diagnostic and therapeutic use of diphenylhydantoin in the subictal state and non-epileptic dysphoria. International Journal of Neuropsychiatry 3 (Suppl. 2), Kalinowsky LB, Putnam TJ (1943). Attempts at treatment of schizophrenia and other non-epileptic psychoses with Dilantin. Archives of Neurological Psychiatry 49, Lee ET (1980). Statistical Methods for Survival Data Analysis. Belmont, CA: Lifetime Learning. Levine J, Bersudsky Y, Nadri C, Yaroslavsky Y, Abed N, Mishory A, Agam G, Belmaker R (2003). Mechanism of action of new mood stabilizing drugs. In: Soares J, Gershon S (Eds.), Handbook of Medical Psychiatry (pp ). New York: Marcel Dekker. Macdonald RL, Kelly KM (1995). Antiepileptic drug mechanisms of action. Epilepsia 36 (Suppl. 2), S2 12. Mishory A, Yaroslavsky Y, Bersudsky Y, Belmaker RH (2000). Phenytoin as an antimanic anticonvulsant : a controlled study. American Journal of Psychiatry 157, Mitchell PB, Malhi GS (2002). The expanding pharmacopoeia for bipolar disorder. Annual Review of Medicine 53, Nemets B, Bersudsky Y, Belmaker RH (2005). Controlled double-blind trial of phenytoin vs. fluoxetine in major depressive disorder. Journal of Clinical Psychiatry 66, Neumann J, Seidel K, Wunderlich HP (1984). Comparative studies of the effect of carbamazepine and trimipramine in depression. In: Einrich HM, Okuma T, Muller AA (Eds.), Anticonvulsants in Affective Disorders (pp ). Amsterdam: Elsevier Science Publishing. Overall J, Gorham D (1962). The brief psychiatric rating scale. Psychological Reports 10, Petty F, Davis AL, Nugent M, Bollig M, Hendricks KR (1999). Valproate treatment of bipolar depression. 54th Annual Convention Society of Biological Psychiatry 45, p. 71S, Washington. Post RM (1989). Use of anticonvulsants in the treatment of manic-depressive illness. In: Post RM, Trimble MR, Pippenger CE (Eds.), Clinical Use of Anticonvulsants in Psychiatric Disorders (pp ). New York: Demos Publications. Post RM, Uhde TW, Roy-Byrne PP, Joffe RT (1986). Antidepressant effects of carbamazepine. American Journal of Psychiatry 143, Sachs GS (1996). Treatment-resistant bipolar depression. Psychiatric Clinics of North America 19, Schaff MR, Fawcett J, Zajecka JM (1993). Divalproex sodium in the treatment of refractory affective disorders. Journal of Clinical Psychiatry 54, Small JG (1990). Anticonvulsants in affective disorders. Psychopharmacology Bulletin 26, Steinacher L, Vandel P, Zullino DF, Eap CB, Brawand-Amey M, Baumann P (2002). Carbamazepine augmentation in depressive patients non-responding to citalopram: a pharmacokinetic and clinical pilot study. European Neuropsychopharmacology 12, Turner WJ (1967). The usefulness of diphenylhydantoin in treatment of non-epileptic emotional disorders. International Journal of Neuropsychiatry 3 (Suppl. 2), Winsberg ME, DeGolia SG, Strong CM, Ketter TA (2001). Divalproex therapy in medication-naive and moodstabilizer-naive bipolar II depression. Journal of Affective Disorders 67, Young RC, Biggs JT, Ziegler VE, Meyer DA (1978). A rating scale for mania: reliability, validity and sensitivity. British Journal of Psychiatry 133,