Uses of Carbamazepine for Psych iatricDisorders: A Review

Edward Kim , M.D.

INTROD UCTION

Carbamazepine, first synthesized in 1953, was in it ially mark eted in Europeas an anticonvulsant. In 197 1 Dalby reported th e drug's psych otrop ic effec ts,most no tably mood stabi lization, in patients with temporal lobe epilepsy (TLE)(I). Other psychiatric applications such as th e treatment of affectively an dbehaviorally labil e patients are being explored. These stud ies are yie ldi ng dat arelevant to the treatment and understanding of th e neu robiology of mentalilln ess. For this reason , fami liarity wit h car bamazep ine is becoming increasinglyimportant to psychiatrists .

CHEMISTRY AND PHARMACOLOGY

Ca rbamazepine is structurally sim ilar to th e tricyclic antidepressant imi ­p ramine (Fig. 1) (2,3). Plasma leve ls peak four to e ight hours afte r ing est ion ,with 75 % of the drug bound to pla sma proteins. CSF levels a re dependent on th elevel of unbound drug in plasma. Metabolism occurs primaril y in the liver via th ecytochrome P-450 oxi dase system , producing carbamazepi ne- l 0 , ll-epoxidewhich is as active an d may reach levels up to half that of carba mazepine. This isalmost entire ly converted to carbamazepine-trans-l 0, l l -dihydrodiol by epoxidehyd rolase before excretion in the urine (Fig 2) (4). T he epo xide is 50 %protein-bound in p lasma (5).

Carbamazep ine induces hepatic en zymes, thus enhanc ing its ow n metabo­lism . Therefore, the 30 -hour half-life of a single oral dose drops to 20 hours atthree weeks at 12 hours afte r several months. The half-life is reduced in patient swho simultaneously ta ke other inducers of hepatic en zymes suc h as ph en oba rbi­tal, p henytoin, or alcohol (4). Ca rbamazepine also enhances the metabo lism o fphenytoin an d warfarin for th e same reason (6) .

CLINICAL GU IDELINES

T he most co m mon side effect is an a llergic skin rash whi ch occurs in 5% to15% of treated patients (7) . T his ma y be accompanied by sore th roat , mucosalu lce ration and low grade fever. Antihistamines can be effect ive in e lim inating

63

64 JEFFERSO N JO U RN A L O F PSYCH I A T RY

Chlorpromazine Imipramine Carbamazepine

FIGURE 1. The tricyclic structure of carbamazepine resembles that o f imipramineand, to a lesser extent, chlorpromazine. Carbamazepine blocks norepinephrine reup­take but does not block dopamine receptors

a llergic symptoms, th ough th e rash may require di sco ntinuatio n of ca rba maze­pine. Carbamazepine-induced agranulocytosis has recei ved much attentionbecause of its mortality rate of up to 50%. Howeve r , less th an thirty cases havebeen reported (2). T h is id iosyncrat ic reacti on has a preva lence o f one in 20,000to 40 ,000 and is unrelated to a harmless 25 % decrease in whi te ce ll co unt wh ichis fairly co mmon (7). About 10% of pati ents receiving the drug develop a milddose-r elated leukopenia whic h resolves within the first four months. O f these,app roxima tely 2% co nt inue with a persistent leukopenia which necessitatesdi scontinu ation o f th erapy (2) . Neurotoxic side effec ts of ca r ba mazepine ma yincl ude drowsin ess, ver t io , atax ia, diplopia , and blurred visio n . Patien ts ac utelyintoxicated on th e drug ca n exh ib it seda tio n , respiratory depression , hyper ir r i­ta bi lity, and even se izures. Other side effects incl ude na usea and vomiting , andhepato to xicit y ranging fr om mild e levation of serum transaminases to acutehepatitis. In addition , ch ron ic ad m inistra t ion ca n also cause a condition resem­bl ing the synd rome of inappropriate secretion of antidiuretic hormone(SIADH) , with wat er retention , hyponatremia , an d possib le water int oxicatio n(2).

T he in itial dose o f ca r ba mazepine in healthy adu lts is 200 mg o nce or twicea day. This is increased by no more th an 20 0 mg a day to a tota l of200 mg t.i.d , Aser um level should be chec ked afte r five days and dosage adj usted accordi ng ly(7). The actui ty of th e patient di ctates how aggressive ly the dosage should beincreased, bu t th e appearance of neu rotoxi cit y is a clear indication fo r sma ller,more gradual increments. Mainten an ce doses average about 1000 mg/day,thoug h the ra nge may vary from 20 0 to 160 0 mg/ day (8). In pati ents who aree lder ly or have live r di sease , dosage and increments should be appropriatelydecreased. T herape ut ic se rum levels ra nge from 6- 12 JLg/ ml; neurotox ic sideeffects are mo re frequent abo ve 10 JLg/ ml (2,4). Leve ls shou ld be checkedfreque nt ly dur ing initial th erapy since 90% of enzy me induct ion occu rs in the

FIGURE 2. Metabolic pathway of carbamazepine. Both the epoxide and dihydrodiol are excreted in the urine.

c:::(Jl

t""I

o'Tl("J

>:0Ct'>:::::>NtTl:::Zt""I'Tlo:0-e(Jl

-e("J

:t

>-J:0

ne(Jl

o:0ot""I:0~

>:0t""I<

~

OHH

N

OANH2CBZ-trans -lO, I !-dihydrodiol

EpoxideHydrolase ..

H 0 H

N

OA NH2CBZ -! 0, I I-epoxide

•Cytochrome

P-450

N

oANH2Carbamazepine

(CBZ)

Ol

""

66 J EFF ERSON JO UR NAL OF PSYCHIAT RY

first month (7). Moreover, a complete blood cou nt and live r panel should bech ecked prior to and periodicall y during treatment (2).

NE UROLOGIC AND ENDOCRI NE APPLICATIO NS

Carbamazepine, a potent anticonvulsant is th e d rug of choice in temporallobe epilepsy (9), and a first-line drug in controlling grand mal se izures. It is alsoeffective in the treatment of trigeminal neuralgia and other chronic painsyndromes (10- 13). Its antidiuretic properties have led to its use in managingdiabetes insipidus (10).

APPLICATIONS IN MOOD DISORDERS

Most psychiatric trials ofcarbamazepine have been co nd ucted with patientswith bipolar disorder, although some studies have included sch izoaffect ivepatients in their test groups. Beginning in the early 1970's Okuma et a l (15,16)in Japan demonstrated that carbamazep ine is ben eficial in th e treatment ofacutely manic patients and in longterm prophylaxis against ma nia. In a double­blind controlled study , Post et a l (17) at th e National In sti tu tes of Menta l Healthreported carbamazep ine 's clear efficacy over placebo in symptom re lief in sevenout of nine (88%) manic patients. Of th ese , four (57%) relapsed d uring p laceboad m inistrat ion.

The drug's acute anti-ma nic efficacy is approx ima tely th at of chlorproma­zine (18). In over half of treat ed patients, peak antiman ic effects were reachedwithin ten da ys of administrati on o f either ch lorp romazi ne or ca rbamazepine.Carbamazepine also lowered scores on th e Clin ica l Global Im pressio n (CGI )sca le on average as much as lithium (19). Lithium, howeve r , had a moreco nsiste nt anti-manic effect than ca rba mazepine: th e changes in CGI sco restended to cluster around the mean in patients t reated with lith ium, while themean CGI change in car bamazep ine -treated patients was largely d ue to a fewwho responded extremely well. A double-blind 3-year study compa ring lith ium(n = 27) and carbamazepine (n = 29) found both drugs effec tive in about twothirds of treated patients (26). Peak symptom relief was reached with in threemonths of treatment onset in both groups. The on ly striking difference betweenthe drugs was their relative dropout rates . Lithium-treated patien ts with moreschizophrenic or schizoaffective symptoms had a sign ificant ly higher d ropoutrate than carbarnazepine-treated patients (50 % vs. 20 %, respecti vely). In con­trast, patients who demonstrated " classical" bipolar psych otic fea tures had alower dropout rate when treat ed with lithium (20 % vs. 35%, respect ive ly).

Investigators are attempting to define a population of pat ients who areparticularly responsive to ca rbamazepine and who, th erefore , would be consid­ered prime candidates for this treatment early in th eir illn ess. Some patientswith a history of lithium resistance responded better when ca rbamazepine wasadded to or substituted for lithium (17, 2 1). In two open clinical tria ls with

USE OF CARBAMAZEPINE FOR PSYCHIATRIC DISORDERS: A REVI EW 67

bipolar and schizoaffective patients resistant to lithium and/or neurolepticmaintenance (22,23), a total of 14 out of 21 (67 %) of those patients improvedwhen carbamazepine was added or substituted into their treatment regimen.Observations that rapid-cycling bipolar patients respond well to car ba mazepine(17,21,22,24) have yet to be demonstrated in controlled clinical trials .

The drug's therapeutic effect in depressed patients is somewhat less clear ,with five of thirteen (38 %) bipolar patients showing marked improvement inPost et aI's original report (17), and twelve of thirty-five (34 %) showing markedimprovement in a subsequent study which involved bipolar (n = 24) and unipo­lar (n = 11) depressed patients (25). The antidepressant effect is related toplasma levels of the epoxide metabolite rather than carbamazepine itsel f (26) .The drug's time course of action in depression is similar to th at of tri cyclicantidepressants (TCA's), with substantial improvements occurring after twoweeks and maximum effect after three weeks (27). Because TCA's are known toinduce more rapid cycling in some bipolar patients treated for depression (28) ,carbamazepine may be considered as an alternative treatment in patients with ahistory of rapid cycling or mania during TCA therapy. One depressed bipolarpatient with a history of TCA-induced mania despite lithium prophylaxisresponded well to carbamazepine without cycling into mania (29). A recent casereport described the successful treatment of a bipolar depressed patient usingcarbamazepine and electroconvulsive therapy (30) .

OTHER PSYCHIATRIC APPLICATIONS

Carbamazepine was found to markedly decrease the severity and fr equen cyof behavioral dyscontrol in some women with borderline personality d isorder(30) . The patients often reported an enhanced ability to think before acting, aphenomenon referred to as "reflective delay. " This resulted in fewer ang ryoutbursts, episodes of violence, and suicide threats or attempts during th e tria lperiod. In contrast, the placebo group showed no such improvements, and sevenof eleven placebo patients (64 %) were removed from the trial early becau se o fbehavioral deterioration, compared with one of fourteen of th e car bamazepinetrials (7%). In addition, the drug has also proven helpful in th e treatment ofpatients with mixed frontal lobe disease and schizophrenia or schizoaffectivedisorder with affective or behavioral liability (31).

. Carbamazepine is more effective than placebo when used in conj unc tionwith haloperidol in the treatment of excited psychoses of schi zophrenic andschizoaffective, as well as manic, patients (32). Because of this appa rent efficacyin patients exhibiting more schizophrenic features (22) , the effect s of carbama­zepine on the dopaminergic pathway have been studied by Po st et a l (33) . T heyfound that carbamazepine decreased probenecid-induced CSF accumulati on ofhomovanillic acid (HVA), a dopamine metabolite, and had no effec t on base lineCSF concentrations of HV A. This is in contrast to neuroleptics which increaseboth baseline and probenecid-treated HVA levels in CSF. This, in addition to

68 J EFF ERSON JO URNAL OF PSYCHIATRY

the absence of reported cases of ta rdive dyskinesia, sugge sts th at carbamazepinedoes not exert its effect by dopamine receptor blockade. T h is is consistent withprevious animal studies (34). Rather, th e drug ma y work by decreasing dopa­mine turnover in the brain. Thus far , carbamazepine ha s not proven useful intreating tardive d yskinesia (35).

DRUG INTERACTIONS

Two recent reports have demonstrated that ca rba mazepine can reduceplasma levels of haloperidol when the two are administered simultaneously(36,37). This ma y be due to hepatic enzyme induction resulting in acceleratedmetabolism of haloperidol. Therefore, larger neuroleptic doses may be neces­sary for the management of patients treated concurrently with ca r ba mazepine.

Another potential complicat ion of carbamazepine treatment is neurotoxic­ity during coadministration with lithium (38). Patients may present with confu­sion, drowsin ess , weakness, lethargy, coarse tremor, hyper reflexia , nystagmus,and cerebellar ataxia. These symptoms can be seen in both lithium (39) andcarbamazepine toxicity (40). In this case , however, neurotoxicity may develop at" nontoxic" plasma levels of each drug. Therefore , monitoring o f drug levelsma y not protect patients agains t developing ad verse effec ts . Fortunately theneurotoxicity is rapidly reversible following dis continuation of treatm ent. Ahistory of lithium-induced neurotoxicity is a prominent risk factor for th iscomplication (38).

On the other hand, coadministration of carbamazepine and lith ium offers apotentially beneficial interaction. Carbamazepine can induce wate r re tentionleading to hyponatremia and wat er intoxication (41 ,42). Lithium, on the otherhand, has been used to treat such water retention in SI ADH by inducing anephrogenic diabetes insipidus (43). These opposite effec ts on plasma sodiumand fluid balance can offset one another so that patients receiving lith ium inaddition to carbamazepine may be less likely to develop hyponatremia thanpatients treated with carbamazepine alo ne (44).

The use of carbamazepine in conjunction with TCA's and monoamineoxidase inhibitors (MAOI's) has not been systemat ically stu d ied . However , weca n make some inferences on the basis of carbamazepine' s structure andpharmacology. The drug inhibits presynaptic reuptake of norepinephrine byapproximately 25 %, compared to 85 % in the case of imipramine (3). T hereforecar bamazepine should be used like other TCA's and started prior to or at thesame time as MAOI's when combination therapy is desired (45). CombinedTCA-MAOI treatment, while raising concerns about hyperten sive crisis, hasbeen extensively reviewed and found to be safe and efficac ious when adminis­tered properly (46, 47) .

A number ofdrugs inhibit the metabolism of ca r ba mazepine and may causeincreased serum levels. These include ery th romycin , verapamil , diltiazem,isoniazid, nicotine, and th e opiate propoxyphene (4) . T hroug h enzyme indue-

USE OF CARBAMAZEPINE FOR PSYCHIATRIC DISORDERS;' A REVI EW 69

tion car bamazepine decreases serum levels of phenytoin and warfarin as well asdoxycycline, sodium valproate, e thosux im ide, and clonazeparn (3,4).

THEORETICAL CONSlDERATIO S

The efficacy of ca rbamazep ine in the treatment of exci ted psych oses raisesquestions about the neurobiology of psychiatric illness. In the late 1960 's it wasobserved that dominant hemisphere TLE predisposes patients to psych osis,whil e nondominant hemisphere TLE is associa ted with bipolar illness (48). Thisstudy compared fifty TLE patients with psychotic features with fifty nonpsy­chotic TLE patients. The presence of psychosis was inversel y co r related withseizure activity . A more recent study found that II of 17 (65%) of T LE pat ientsand zero of seven generalized epileptics had at least one of th e e igh t Schneider­ian first-rank symptoms for schizophrenia (49) . Dalby's study (I ) goes on todemonstrate that TLE patients treated with carbamazepine had mar kedl yreduced affective symptoms such as anxiety , paranoia, and aggressive beh aviorin 24 of 58 (44 %) patients. Moreover, several patients improved psychiatricallydespite a lack of seizure control. While nearly half of Dalby's epileptic patientshad structural brain abnormalities, most subsequent stud ies of th e psych otrop iceffects of carbamazepine have screened out patients with str uc tura l or electro­en cephalographic abnormalities.

Post et al (50) proposed a th eory of agitated psych osis wh ich involved aphenomenon called " kind ling. " Repetitive subth reshold stimulatio n of thelimbic system produces a hyperexcitable " kind led " state in whic h mot or seizuresma y be induced by stimulation that was previously non-seizurogenic . Reviewingthe literature on the effects of chronic administration of cocaine and other CNSst im ulants, they noted that subjects developed hypersensivity ra ther thantolerance to the effects of th ese substances on pathological behavior andse izu res.

Carbamazepine, in addition to phenobarbital and diazepam, is very effec­tive in blocking experimental electrically kindled amygda loid se izu res in avariety of animal species (51). Ph enytoin , another potent anticonvul sant, is muchless effective in blocking kindled am ygdaloid seizures. Moreover, phenytoinexh ibits no significant therapeutic effect on manic patients (52). T h is suggeststhe possibility of a limbic focus for the generation of mania and other excitedpsychoses. The mechanism may involve subthreshold kindling of limbic st ruc­tures. The absence of structural and physiologic abnormalities in such patientscould be more reflective of instrumental limitations than th e absen ce of organicpathology.

CO NCLUSIO NS

Carbamazepine, once used only in the treatment of epilepsy and neuropat­hic pain, is becoming increasingly important in th e treatment of some psychiat-

70 JEFFERSON JOURNAL OF PSYCHIATRY

ric patients, but that group is not yet clearly defined. Because of potentialadverse effects and drug interactions, it should be administered j udicious ly inselected patients who can be carefully monitored. Nevertheless, it is becomingevident that for a certain subset of psychiatric patients, carbamazepine a lone orin combination with .o ther drugs may prove to be the only effec tive treatmentthat is currently available.

REFERENCES

1. Dalby MA : Antiepileptic and psychotropic effect of carbamazepine (Tegretol[rJ) inthe treatment of psychomotor epilepsy. Epilepsia 12:325-334, 1971

2. Gilman AG , Goodman LS, Rail TW, et al (eds): The Pharmacologi cal Basis ofTherapeutics. New York, MacMillan Publishing Company, pp. 457-459 , 1985

3. Post RM, Ballenger JC, Uhde TW et al: Efficacy of carbam azep ine in manic­depressive illness, in Post RM and Ballenger JC: Neurobiology of Mood Disorders.Baltimore, Williams and Wilkins, pp 777-816, 1984

4. Levy RH , Kerr BM: Clinical pharmacokinetics of carbamaze pine. J Clin Psychiatry49(suppl.): 58-61 , 1988

5. Johannsen SI: Antiepileptic drugs: pharmacokinetics and clin ical asp ects. T her Dru gMonit3: 17-37, 1974

6. Hansen JM, Siersbaek-Nielsen K, Skovsted L: Carbamazepine-induced accelerationof diphenylhydantoin and warfarin metabolism in man. Clin Pharm T he r 12:539­543, 1971

7. Neppe VM, Tucker GJ , Wilensky AJ: Introduction: fundamentals of carbamazepineuse in neuropsychiatry.J Clin Psychiatry 49(suppl.): 4-6, 1988

8. Hyman SE, Arana GW: Handbook of Psychiatric Drug Therap y. Baltimore, Little ,Brown, and Company, pp. 98-114,1987

9. Bird CAK, Griffin BP, Miklaszewska JM , et al: Tegretol (carbam azepine): acontrolled trial of a new anticonvulsant. Br J Psych 112:737-742, 1966

10. Ekbom K: Carbamazepine in the treatment of tabetic lightning pains. Arch Neurol26 :374-378, 1972

11. Elliot F, Little A, Millbrandt W: Carbamazepine for ph antom-limb phenomena(letter). N Engl J Med 295:678, 1976

12. Taylor JC, Brauer S, Espir MLE: Long-term treatment of trigeminal neuralgia withcarbamazepine. Postgrad MedJ 57 :16-18,1981

13. Sweet WH: The treatment of trigeminal neuralgia (tic doloroureux). N Engl J Med315:174-177,1986

14. Wales JK: Treatment of diabetes insipidus with carbamazepine . Lan cet 1:948-951,1975

15. Okuma T, Kishimoto A, Inoue K, et al: Anti-manic and prophylacti c effects ofcarbamazepine on manic-depressive psychosis. Folia Psych Neurol Jpn 27:283-297,1973

16. Okuma T, Inanaga K, Otsuki S, et al: A preliminary double-blind stu dy on theefficacy of carbamazepine in prophylaxis of manic-depressive illness. Psych opharm73:95-96, 1981

17. BallengerJC, Post RM: Carbamazepine in manic-depressive illness: a new treatment.AmJ Psych 137:782-790, 1980

USE OF CARBAMAZEPINE FOR PSYCHIATRIC DISORDERS: A REVI EW 71

18. Okuma T, Inanaga K, Otsuki S, et al: Comparison of th e an timanic efficacy ofcarbamazepine and chlorpromazine: a double-blind controlled study. Psychoph arm66 :211-217,1979

19. Lerer B, Moore N, Meyendorff E, et al : Carbamazepine versus lithium in mania: adouble-blind study.J Clin Psych 48:89-93, 1987

20 . Placidi GF, Lenzi A, Lazzerini F, et al : The comparative efficacy and safety ofcarbamazepine versus lithium: a randomized, double-blind 3-year tri al in 83 patients.J Clin Psych 47 :490-494,1986

21. Fawcett J, Kravitz HM: The long-term management of bipolar di sorders withlithium, carbamazepine, and antidepressants. J Clin Psych 46 :58-60 , 1985

22. Folks DG , King LD, Dowdy SB, et al: Carbamazepine treatment of selectedaffectively disordered inpatients. Am J Psych 139: 115-11 7, 1982

23 . Elphick M: An open clinical trial of carbamazepine in treatment-r esistant bipo lar an dschizo-affective psychotics. Br J Psych 147: 198-200, 1958

24. Roy-B yrne PP, Joffe RT, Uhde TW, et al: Approaches to th e eva luation andtreatment of rapid-cycling affective illn ess . Br J Psych 145 :54 3- 550 , 1984

25 . Post RM, Uhde TW, Roy-Byrne PP, et al: Antidepressant effects of carbamazepine .AmJ Psych 143:29-34, 1986

26. Post RM, Uhde TW, Ballenger JC: Carbamazepine and its -10, ll -epoxide metabo­lite in plasma and CSF-relationship to antidepressant re sponse. Arch Gen Psych40:673-676, 1983

27. Post RM: Time course of clinical effects of carbamazepine implicati ons for mecha­nism of action .J Clin Psychiatry 49 (suppl) : 35-40, 1988

28 . Wehr T A, Goodwin FK: Rapid cycling in manic-depressives induced by tricyclicantidepressants. Arch Gen Psych 36:555-559, 1979

29. Number HG, FinkelJA: Carbamazepine in bipolar-depressed di sorder complicatedby tricyclic switching: case report

30 . Cantor C: Carbamazepine and ECT: a paradoxical combination (le tter) . J Clin Psych47:276-277 ,1986

3 1. Gardner DL , Cowdry RW : Positive effects of carbamazepine on behavioral dyscon­trol in borderline personality. Am J Psych 143:519-522, 1986

32. McAllister TW: Carbamazepine in mixed frontal lobe and psychi atr ic d iso rde rs. JClin Psychiatry 46:393-394, 1985

33 . Klein E, Bental E, Lerer B, et al : Carbamazepine and haloperidol v p lacebo andhaloperidol in excited psychoses. Arch Gen Psych 41 :165-1 70, 1984

34. Post RM , Rubinow DR, Uhde TW, et al : Dopaminergic effec ts of carbam azepine .Arch Gen Psych 43 :392-396, 1986

35. Marangos PJ, Post RM , Patel ], et al : Specific and potent interac tions of carbamaze­pine with brain adenosine receptors . Eur J Pharmacol 93 :175-182 , 1983

36. Yassa R: Carbamazepine in the treatment of tardive dyskinesia (letter). Am J Psych143:1191 ,1986

37 . Fast DK,Jones BD, Kusalic M, et al: Effect of carbamazepine on neurolepti c plasmalevels and efficacy (letter). Am J Psych 143 :117-118, 1986

38 . Arana GW, Goff DC , Friedman H: Do es carbam azepine-induced reduction ofplasma haloperidol levels worsen psychotic symptoms? Am J Psych 143:650- 65 1,1986

39 . Shukla S, Godwin CD, Long LEB , et al: Lithium-carbamazepine neu ro toxicity andrisk factors. Am J Psych 14 1:1604-1606, 1984

72 J EFFERSO N J OURNAL OF PSYCHI AT RY

40. Sh opsin D, Johnson G, Gershon S: Neuroto xicity with lithium : d ifferential drugre sponsiveness. Int Pharmacopsych 5: 170-182, 1970

41 . Reynolds EM: Neurotoxicity of carbamazepine. Adv Ne uro l 2:345-354, 197 542. Flegel KM, Col CH: Inapprop riate antidiuresis during carbamazepine treatment.

Ann Intern Med 87:722-723 , 197743. Ashton MG , Ball SG, T homas TH, et al : Water intoxication associat ed with

carbamazepine treatment. Br Med J 1:1134-11 35, 197744 . White MG, Fetner CD: T reatment of the syndrome of inappropr iate secret ion of

antidiuretic hormone with lithium carbonate . N EnglJ Med 292:390- 392, 197 545. Vieweg V, Glick JL, Herring S, et al: Absence of carbamazepine-induced hypona­

tremia among patients also given lithium. Am J Psych 144 :943- 94 7, 198746. Schatzberg AF : Evaluation and treatment of the refr actory depressed patient, in

BernsteinJG: Clinical Psychopharmacology, 2nd ed ., Boston,John Wright-PSG Inc.,pp.77-92,1984

47 . Stern SL, Mendels J: Drug combinations in the treatment of refractory depression: areview.J Clin Psych 42:368-373,1981

48 . FeighnerJP, HerbsteinJ, Damlouji N: Combined MAOI , TCA, and di rect stimulanttherap y of treatment-resistent depression. J Clin Psych 46 :206-209, 1985

49 . Flor-Henry P: Psychosis and temporal lob e epilepsy-a co ntrolled investigation.Epilepsia 10:363-395, 1969

50. Perez MM, Trimble MR , Murray NMF et al : Epileptic psychosis: an evaluati on ofPSE profiles. Br J Psych 141: I 55-163 , 1985

5 1. Post RM, Kopanda RT: Cocaine , kindling, and psych osis. Am J Psych 133 :627-633,1976

52. Wada JA: Pharmacological prophylaxis in the kind ling model of ep ileps y. ArchNeurol 34:389-395, 1977

53 . Post RM , Berretini W , Uhde TW, et al : Selective response to the anticonvulsantcarba mazepine in manic-depressive illness: a case study . J Clin Psychoph ar m 4:178­185,1984