Treatment-resistant schizophrenia: What role for mood stabilizers?

In patients with treatment-resistant schizophrenia, lithium and anticonvulsants have become such common adjuncts to antipsychotics that 50% of inpatients may be receiving them.¹ Evidence supporting this practice is mixed:

• Initial case reports and open-label studies that showed benefit have not always been followed by randomized clinical trials.
• Lack of clear benefit also has been described (Table 1).

This article examines the extent of this prescribing pattern, its evidence base, and mechanisms of action that may help explain why some adjunctive mood stabilizers are more effective than others for treatment-resistant schizophrenia.

EXTENT OF USE

For 5 years, the rate at which inpatients with schizophrenia received adjunctive mood stabilizers has held steady at approximately 50% in New York State Office of Mental Health (NYSOMH) facilities (Figure1). These facilities provide intermediate and long-term care for the seriously, persistently mentally ill. Adjunctive mood stabilizers might not be used as often for outpatients or for inpatients treated in short-stay facilities.

Table 1

Evidence for adjunctive use of lithium or anticonvulsants for treating schizophrenia

Agent	Case reports and open studies	Randomized, double-blind trials	Benefit?
Lithium	Yes (many)	Yes (several, +/-)	Probably not
Carbamazepine	Yes (many)	Yes (several; small total sample)	Limited
Valproate	Yes (many)	Yes	Yes
Gabapentin	Yes (very few, +/-)	None	Probably not
Lamotrigine	Yes (few, +/-)	Yes (2, +)	Yes
Topiramate	Yes (very few, +/-)	Yes (1, +/-)	Probably not
Oxcarbazepine	Yes (very few, +/-)	None	Probably not
+ = positive results
- = negative results
+/- = both positive and negative results

Valproate is the most commonly used anticonvulsant, with one out of three patients with schizophrenia receiving it. Adjunctive gabapentin use is declining, probably because of inadequate efficacy—as will be discussed later. Use of adjunctive lamotrigine is expected to increase as more data become available on its usefulness in treatment-resistant schizophrenia.

Clinicians are using substantial dosages of adjunctive mood stabilizers. During first-quarter 2004, average daily dosages for 4,788 NYSOMH patients (80% with schizophrenia or schizoaffective disorder) receiving antipsychotics were:

• valproate, 1639 mg (n = 1921)
• gabapentin, 1524 mg (n = 303)
• oxcarbazepine, 1226 mg (n = 201)
• carbamazepine, 908 mg (n = 112)
• lithium, 894 mg (n = 715)
• topiramate, 234 mg (n = 269)
• lamotrigine, 204 mg (n = 231).²

Mood-stabilizer combinations were also used. Approximately one-half of patients receiving adjunctive mood stabilizers—with the exception of valproate—were receiving more than one. In patients receiving valproate, the rate of mood-stabilizer co-prescribing was about 25%.²

WHAT IS THE EVIDENCE?

Evidence supporting the use of adjunctive lithium or anticonvulsants to treat schizophrenia varies in quality and quantity (Table 1). Case reports and open studies offer the weakest evidence but can spur double-blind, randomized clinical trials (RCTs). Unfortunately, RCTs are not often done, and the published studies usually suffer from methodologic flaws such as:

• inadequate number of subjects (insufficient statistical power to detect differences)
• lack of control of confounds such as mood symptoms (seen when studies include patients with schizoaffective disorder)
• inadequate duration
• inappropriate target populations (patients with acute exacerbations of schizophrenia instead of persistent symptoms in treatment-resistant schizophrenia).

Because controlled trials of the use of adjunctive mood stabilizers are relatively scarce, clinical practice has transcended clinical research. Clinicians need effective regimens for treatment-resistant schizophrenia, and mood-stabilizer augmentation helps some patients.

Lithium is perhaps the best-known mood stabilizer. Although early studies showed adjunctive lithium useful in treating schizophrenia, later and better-designed trials did not. The authors of a recent meta-analysis of randomized clinical trials (n = 611 in 20 studies) concluded that despite some evidence supporting the efficacy of lithium augmentation, overall results were inconclusive. A large trial would be required to detect a small benefit in patients with schizophrenia who lack affective symptoms.³

Carbamazepine use among patients with schizophrenia is declining, primarily because this drug induces its own metabolism and can require frequent dose adjustments. Adjunctive carbamazepine has been used to manage persistent aggressive behavior in patients with schizophrenia and schizoaffective disorder. Evidence comes primarily from small trials or case reports (Table 2),^4-8 but results of a larger clinical trial (n = 162) by Okuma et al⁶ are also available

In the Okuma report—a double-blind, placebo-controlled trial of carbamazepine in patients with DSM-III schizophrenia or schizoaffective disorder—carbamazepine did not significantly improve patients’ total Brief Psychiatric Rating Scale (BPRS) scores. Compared with placebo, however, some benefit with carbamazepine did emerge in measures of suspiciousness, uncooperativeness, and excitement.

A systematic review and meta-analysis (n = 283 in 10 studies) detected a trend toward reduced psychopathology with carbamazepine augmentation for schizophrenia. BPRS scores declined by 20% and 35% in the six trials (n = 147) for which data were available (P = 0.08 and 0.09, respectively).⁹ Because the double-blind trial by Okuma et al⁶ was not randomized, it was not included in this meta-analysis.⁹

Figure 1 10-year trend in use of adjunctive mood stabilizers for schizophrenia

Valproate. Among the anticonvulsants, the greatest body of evidence supports the use of valproate in patients with schizophrenia,¹⁰ although a recent meta-analysis (n = 378 in 5 studies) indicates inconsistent beneficial effects.¹¹

 

Initial double-blind, RCTs of adjunctive valproate in patients with schizophrenia were limited in size and failed to show benefit with adjunctive valproate^12-15 (Table 2). A more recent study¹⁶ showed that adjunctive valproate affects acute psychotic symptoms rather than mood. This study, however, did not answer whether adjunctive valproate would help patients with persistent symptoms of schizophrenia.

Table 2

Double-blind studies of adjunctive carbamazepine in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Neppe (1983)	11	42	Crossover	Mixed, 8 with schizophrenia	“Overall clinical rating” improved
Dose et al (1987)	22	28	HAL + CBZ vs HAL + placebo	Schizophrenia or schizoaffective disorder	No difference on BPRS
Okuma et al (1989)	162	28	NL + CBZ vs NL + placebo	Schizophrenia or schizoaffective disorder	No difference on BPRS; possible improvement in suspiciousness, uncooperativeness, and excitement
Nachshoni et al (1994)	28	49	NL + CBZ vs NL + placebo	“Residual schizophrenia with negative symptoms”	No difference on BPRS or SANS
Simhandl et al (1996)	42	42	NL + CBZ vs NL + lithium vs NL + placebo	Schizophrenia(treatment- nonresponsive)	No difference on BPRS; CGI improved from baseline in groups receiving CBZ and lithium
n = number of subjects
BPRS = Brief Psychiatric Rating Scale
CGI = Clinical Global Impression scale
CBZ = carbamazepine
HAL = haloperidol
NL = neuroleptic (first-generation antipsychotic)
SANS = Scale for the Assessment of Negative Symptoms

In this large (n = 249), multicenter, randomized, double-blind trial, hospitalized patients with an acute exacerbation of schizophrenia received olanzapine or risperidone plus divalproex or placebo for 28 days. Patients with schizoaffective disorder and treatment-resistant schizophrenia were excluded.

By day 6, dosages reached 6 mg/d for risperidone and 15 mg/d for olanzapine. Divalproex was started at 15 mg/kg and titrated to a maximum of 30 mg/kg by day 14. Mean divalproex dosage was approximately 2300 mg/d (mean plasma level approximately 100 mg/mL).

Positive and Negative Syndrome Scale (PANSS) total scores improved significantly in patients receiving adjunctive divalproex compared with those receiving antipsychotic monotherapy, and significant differences occurred as early as day 3. The major effect was seen on schizophrenia’s positive symptoms. A post-hoc analysis¹⁷ also showed greater reductions in hostility (as measured by the hostility item in the PANSS Positive Subscale) in patients receiving adjunctive divalproex compared with antipsychotic monotherapy. This effect was independent of the effect on positive symptoms or sedation.

A large, multi-site, 84-day acute schizophrenia RCT similar to the 28-day trial—but using extended-release divalproex—is being conducted. An extended-release preparation may be particularly helpful in encouraging medication adherence for patients taking complicated medication regimens.

Table 3

Double-blind studies of adjunctive valproate in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Ko et al (1985)	6	28	Crossover	Neurolepticresistant patients with chronic schizophrenia, not exacerbation	No valproate effect noted
Fisk and York (1987)	62	42	Antipsychotic + valproate vs antipsychotic + placebo	Chronic psychosis and tardive dyskinesia	No differences in mental state and behavior, as measured by“ Krawiecka scale”*
Dose et al (1998)	42	28	HAL + valproate vs HAL + placebo	Acute, nonmanic schizophrenic or schizoaffective psychosis	No difference on BPRS; possible effect on “hostile belligerence”
Wassef et al (2000)	12	21	HAL + valproate vs HAL + placebo	Acute exacerbation of chronic schizophrenia	CGI and SANS scores improved significantly, but BPRS scores did not
Casey et al (2003)	249	28	RIS + valproate vs OLZ + valproate vs RIS + placebo vs OLZ + placebo	Acute exacerbation of schizophrenia	PANSS scores improved
* Krawiecka M, Goldberg D, Vaughan M. A standardized psychiatric assessment scale for rating chronic psychotic patients. Acta Psychiatr Scand 1977;55(4):299-308.
n = number of subjects
BPRS = Brief Psychiatric Rating Scale
CGI = Clinical Global Impression scale
HAL = haloperidol
OLZ = olanzapine
PANSS = Positive and Negative Syndrome Scale
RIS = risperidone
SANS = Scale for the Assessment of Negative Symptoms

In a recent large (n = 10,262), retrospective, pharmacoepidemiologic analysis,¹⁸ valproate augmentation led to longer persistence of treatment than did the strategy of switching antipsychotics. Average valproate dosages were small, however (<425 mg/d), as were antipsychotic dosages (risperidone <1.7 mg/d, quetiapine <120 mg/d, and olanzapine <7.5 mg/d). Patients’ diagnostic categories were not available. One interpretation of this study is that valproate augmentation would be more successful than switching antipsychotics, assuming that treatment persistence can be viewed as a positive outcome.

Table 4

Double-blind studies of adjunctive lamotrigine in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Tiihonen et al (2003)	34	84	Crossover; clozapine with or without lamotrigine	Clozapine-resistant male inpatients with chronic schizophrenia, not exacerbation	BPRS, PANSS positive, and PANSS general psychopathology symptom scores improved Negative symptoms did not improve
Kremer et al (2004)	38	70	Antipsychotic* + lamotrigine vs antipsychotic* + placebo	Treatment- resistant inpatients with schizophrenia	Completers’ PANSS positive, general psychopathology and total symptom scores improved No difference in negative symptoms or total BPRS scores No difference with intent-to-treat analyses
n = number of patients
* First- or second-generation antipsychotic
BPRS = Brief Psychiatric Rating Scale
PANSS = Positive and Negative Syndrome Scale

Lamotrigine is the only other anticonvulsant for which published, double-blind, randomized evidence of use in patients with schizophrenia is available (Table 4).^19,20 Adjunctive lamotrigine may be effective in managing treatment-resistant schizophrenia, as was shown in a small (n = 34), double-blind, placebo-controlled, crossover trial.¹⁹ Hospitalized patients whose symptoms were inadequately controlled with clozapine monotherapy received lamotrigine, 200 mg/d, for up to 12 weeks. Adjunctive lamotrigine improved positive but not negative symptoms.

 

Similar results were seen in treatment-resistant inpatients with schizophrenia (n = 38) in a 10-week, double-blind, parallel group trial by Kremer et al.²⁰ Adjunctive lamotrigine improved PANSS positive, general psychopathology, and total symptom scores in the 31 patients who completed the trial. No differences were seen, however, in negative symptoms, total BPRS scores, or in the intent-to-treat analysis. These results have spurred the launch of a large, multi-site, RCT of adjunctive lamotrigine in patients with schizophrenia who have responded inadequately to antipsychotics alone.

Topiramate, one of the few psychotropics associated with weight loss, has attracted interest as an adjunct to second-generation antipsychotics to address weight gain. Although case reports have shown benefit,²¹ one showed deterioration in both positive and negative symptoms when topiramate was added to second-generation antipsychotics.²²

Table 5

Double-blind study of adjunctive topiramate in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Tiihonen (2004)*	26	84	Crossover; SGA plus topiramate or placebo	Treatment- resistant male inpatients with chronic schizophrenia	PANSS general scores improved No difference in total PANSS, PANSS positive, or PANSS negative scores
* 2004 Collegium Internationale Neuro-Psychopharmacologicum (CINP) presentation, and personal communication (6/22/04)
n = number of patients
PANSS = Positive and Negative Syndrome Scale
SGA = Second-generation antipsychotic (patients were taking clozapine, olanzapine, or quetiapine)

An unpublished, randomized, crossover trial compared second-generation antipsychotics plus topiramate or placebo in 26 male inpatients with chronic schizophrenia. With adjunctive topiramate, the authors found a statistically significant improvement in the PANSS general psychopathology subscale but not in PANSS total, positive subscale, or negative subscale scores (Table 5) (Tiihonen J, personal communication 6/22/04).

Other agents. Very little information—all uncontrolled—supports adjunctive use of gabapentin or oxcarbazepine for patients with schizophrenia.^23-28 Of concern are reports of patients suffering worsening of psychosis with gabapentin²⁵ or of dysphoria and irritability with oxcarbazepine (attributed to a pharmacokinetic interaction).²⁶

Conclusion. More trials are needed to examine the use of adjunctive mood stabilizers in patients with schizophrenia—particularly in those with chronic symptoms. Although mood stabilizers are widely used in this population, important questions remain unanswered, including:

• characteristics of patients likely to require adjunctive treatment
• how long treatment should continue.

MECHANISMS OF ACTION

Unlike antipsychotics, mood stabilizers do not exert their therapeutic effects by acting directly on dopamine (D2) receptors. Differences in mechanism of action among the anticonvulsants may help explain why some—such as valproate and lamotrigine—have been useful for bipolar disorder or schizophrenia and others—such as gabapentin—have not.²⁹

One possibility is that anticonvulsants that affect voltage-gated sodium channels—such as valproate, lamotrigine, carbamazepine and oxcarbazepine—may be most useful for patients with bipolar disorder or schizophrenia. On the other hand, agents that affect voltage-gated calcium channels—such as gabapentin—may be efficacious as anticonvulsants but not as efficacious for bipolar disorder or schizophrenia.

Ketter et al³⁰ proposed an anticonvulsant classification system based on predominant psychotropic profiles:

• the “GABA-ergic” group predominantly potentiates the inhibitory neurotransmitter GABA, resulting in sedation, fatigue, cognitive slowing, and weight gain, as well as possible anxiolytic and antimanic effects
• the “anti-glutamate” group predominantly attenuates glutamate excitatory neurotransmission and is associated with activation, weight loss, and possibly anxiogenic and antidepressant effects.

In the GABA-ergic group are anticonvulsants such as barbiturates, benzodiazepines, valproate, gabapentin, tiagabine, and vigabatrin. The antiglutamate group includes agents such as felbamate and lamotrigine. A “mixed” category includes anticonvulsants with GABA-ergic and anti-glutaminergic actions such as topiramate, which has sedating and weight-loss properties.

Because GABA appears to modulate dopamine neurotransmission,³¹ this may explain valproate’s role as an adjunctive agent for schizophrenia. Similarly, mechanisms related to Nmethyl-D-aspartate (NMDA) and non-NMDA glutamate receptor function may explain lamotrigine’s usefulness in this setting.^19,20

SUMMARY

Clinicians resort to combination therapies when monotherapies fail to adequately control symptoms or maintain response. Co-prescribing of anticonvulsants with antipsychotics for inpatients with schizophrenia is common practice in New York State and most likely elsewhere. In general, antipsychotics’ and mood stabilizers’ different—and perhaps complementary—mechanisms of action explain the synergism between them. Mechanisms of action also may explain why some anticonvulsants help in schizophrenia (or bipolar disorder) whereas others do not.

Evidence for using adjunctive anticonvulsants is variable. The strongest data support using valproate (and perhaps lamotrigine), followed by carbamazepine and then topiramate. Gabapentin and oxcarbazepine have only anecdotal evidence, some of it negative. Well-designed, randomized clinical trials with the appropriate populations are needed.

Related resources

• Stahl SM. Essential psychopharmacology of antipsychotics and mood stabilizers. New York: Cambridge University Press, 2002.
• Harvard Medical School Department of Psychiatry’s psychopharmacology algorithm project. Osser DN, Patterson RD. Consultant for the pharmacotherapy of schizophrenia. Available at http://mhc.com/Algorithms/. Accessed Nov. 5, 2004.

Drug brand names

• Carbamazepine • Tegretol
• Clozapine • Clozaril
• Gabapentin • Neurontin
• Haloperidol • Haldol
• Lamotrigine • Lamictal
• Lithium • Lithobid, Eskalith
• Olanzapine • Zyprexa
• Oxcarbazepine • Trileptal
• Quetiapine • Seroquel
• Risperidone • Risperdal
• Topiramate • Topamax
• Valproate (valproic acid, divalproex sodium) • Depakene, Depakote

 

Disclosure

Dr. Citrome receives research grants/contracts from Abbott Laboratories, AstraZeneca Pharmaceuticals, Bristol-Myers Squibb Co., Eli Lilly & Co., Janssen Pharmaceutica, and Pfizer Inc. He is a consultant to and/or speaker for Bristol-Myers Squibb Co., Eli Lilly & Co., Pfizer Inc., Abbott Laboratories, AstraZeneca Pharmaceuticals, and Novartis Pharmaceuticals Corp.

Acknowledgment

Adapted from Citrome L. “Antipsychotic polypharmacy versus augmentation with anticonvulsants: the U.S. perspective” (presentation). Paris: Collegium Internationale Neuro-Psychopharmacologicum (CINP), June 2004 [abstract in Int J Neuropsychopharmacol. 2004; 7(suppl 1):S69], and from Citrome L. “Mood-stabilizer use in schizophrenia: 1994-2002” (NR350) (poster). New York: American Psychiatric Association annual meeting, May 2004.

In patients with treatment-resistant schizophrenia, lithium and anticonvulsants have become such common adjuncts to antipsychotics that 50% of inpatients may be receiving them.¹ Evidence supporting this practice is mixed:

• Initial case reports and open-label studies that showed benefit have not always been followed by randomized clinical trials.
• Lack of clear benefit also has been described (Table 1).

This article examines the extent of this prescribing pattern, its evidence base, and mechanisms of action that may help explain why some adjunctive mood stabilizers are more effective than others for treatment-resistant schizophrenia.

EXTENT OF USE

For 5 years, the rate at which inpatients with schizophrenia received adjunctive mood stabilizers has held steady at approximately 50% in New York State Office of Mental Health (NYSOMH) facilities (Figure1). These facilities provide intermediate and long-term care for the seriously, persistently mentally ill. Adjunctive mood stabilizers might not be used as often for outpatients or for inpatients treated in short-stay facilities.

Table 1

Evidence for adjunctive use of lithium or anticonvulsants for treating schizophrenia

Agent	Case reports and open studies	Randomized, double-blind trials	Benefit?
Lithium	Yes (many)	Yes (several, +/-)	Probably not
Carbamazepine	Yes (many)	Yes (several; small total sample)	Limited
Valproate	Yes (many)	Yes	Yes
Gabapentin	Yes (very few, +/-)	None	Probably not
Lamotrigine	Yes (few, +/-)	Yes (2, +)	Yes
Topiramate	Yes (very few, +/-)	Yes (1, +/-)	Probably not
Oxcarbazepine	Yes (very few, +/-)	None	Probably not
+ = positive results
- = negative results
+/- = both positive and negative results

Valproate is the most commonly used anticonvulsant, with one out of three patients with schizophrenia receiving it. Adjunctive gabapentin use is declining, probably because of inadequate efficacy—as will be discussed later. Use of adjunctive lamotrigine is expected to increase as more data become available on its usefulness in treatment-resistant schizophrenia.

Clinicians are using substantial dosages of adjunctive mood stabilizers. During first-quarter 2004, average daily dosages for 4,788 NYSOMH patients (80% with schizophrenia or schizoaffective disorder) receiving antipsychotics were:

• valproate, 1639 mg (n = 1921)
• gabapentin, 1524 mg (n = 303)
• oxcarbazepine, 1226 mg (n = 201)
• carbamazepine, 908 mg (n = 112)
• lithium, 894 mg (n = 715)
• topiramate, 234 mg (n = 269)
• lamotrigine, 204 mg (n = 231).²

Mood-stabilizer combinations were also used. Approximately one-half of patients receiving adjunctive mood stabilizers—with the exception of valproate—were receiving more than one. In patients receiving valproate, the rate of mood-stabilizer co-prescribing was about 25%.²

WHAT IS THE EVIDENCE?

Evidence supporting the use of adjunctive lithium or anticonvulsants to treat schizophrenia varies in quality and quantity (Table 1). Case reports and open studies offer the weakest evidence but can spur double-blind, randomized clinical trials (RCTs). Unfortunately, RCTs are not often done, and the published studies usually suffer from methodologic flaws such as:

• inadequate number of subjects (insufficient statistical power to detect differences)
• lack of control of confounds such as mood symptoms (seen when studies include patients with schizoaffective disorder)
• inadequate duration
• inappropriate target populations (patients with acute exacerbations of schizophrenia instead of persistent symptoms in treatment-resistant schizophrenia).

Because controlled trials of the use of adjunctive mood stabilizers are relatively scarce, clinical practice has transcended clinical research. Clinicians need effective regimens for treatment-resistant schizophrenia, and mood-stabilizer augmentation helps some patients.

Lithium is perhaps the best-known mood stabilizer. Although early studies showed adjunctive lithium useful in treating schizophrenia, later and better-designed trials did not. The authors of a recent meta-analysis of randomized clinical trials (n = 611 in 20 studies) concluded that despite some evidence supporting the efficacy of lithium augmentation, overall results were inconclusive. A large trial would be required to detect a small benefit in patients with schizophrenia who lack affective symptoms.³

Carbamazepine use among patients with schizophrenia is declining, primarily because this drug induces its own metabolism and can require frequent dose adjustments. Adjunctive carbamazepine has been used to manage persistent aggressive behavior in patients with schizophrenia and schizoaffective disorder. Evidence comes primarily from small trials or case reports (Table 2),^4-8 but results of a larger clinical trial (n = 162) by Okuma et al⁶ are also available

In the Okuma report—a double-blind, placebo-controlled trial of carbamazepine in patients with DSM-III schizophrenia or schizoaffective disorder—carbamazepine did not significantly improve patients’ total Brief Psychiatric Rating Scale (BPRS) scores. Compared with placebo, however, some benefit with carbamazepine did emerge in measures of suspiciousness, uncooperativeness, and excitement.

A systematic review and meta-analysis (n = 283 in 10 studies) detected a trend toward reduced psychopathology with carbamazepine augmentation for schizophrenia. BPRS scores declined by 20% and 35% in the six trials (n = 147) for which data were available (P = 0.08 and 0.09, respectively).⁹ Because the double-blind trial by Okuma et al⁶ was not randomized, it was not included in this meta-analysis.⁹

Figure 1 10-year trend in use of adjunctive mood stabilizers for schizophrenia

Valproate. Among the anticonvulsants, the greatest body of evidence supports the use of valproate in patients with schizophrenia,¹⁰ although a recent meta-analysis (n = 378 in 5 studies) indicates inconsistent beneficial effects.¹¹

 

Initial double-blind, RCTs of adjunctive valproate in patients with schizophrenia were limited in size and failed to show benefit with adjunctive valproate^12-15 (Table 2). A more recent study¹⁶ showed that adjunctive valproate affects acute psychotic symptoms rather than mood. This study, however, did not answer whether adjunctive valproate would help patients with persistent symptoms of schizophrenia.

Table 2

Double-blind studies of adjunctive carbamazepine in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Neppe (1983)	11	42	Crossover	Mixed, 8 with schizophrenia	“Overall clinical rating” improved
Dose et al (1987)	22	28	HAL + CBZ vs HAL + placebo	Schizophrenia or schizoaffective disorder	No difference on BPRS
Okuma et al (1989)	162	28	NL + CBZ vs NL + placebo	Schizophrenia or schizoaffective disorder	No difference on BPRS; possible improvement in suspiciousness, uncooperativeness, and excitement
Nachshoni et al (1994)	28	49	NL + CBZ vs NL + placebo	“Residual schizophrenia with negative symptoms”	No difference on BPRS or SANS
Simhandl et al (1996)	42	42	NL + CBZ vs NL + lithium vs NL + placebo	Schizophrenia(treatment- nonresponsive)	No difference on BPRS; CGI improved from baseline in groups receiving CBZ and lithium
n = number of subjects
BPRS = Brief Psychiatric Rating Scale
CGI = Clinical Global Impression scale
CBZ = carbamazepine
HAL = haloperidol
NL = neuroleptic (first-generation antipsychotic)
SANS = Scale for the Assessment of Negative Symptoms

In this large (n = 249), multicenter, randomized, double-blind trial, hospitalized patients with an acute exacerbation of schizophrenia received olanzapine or risperidone plus divalproex or placebo for 28 days. Patients with schizoaffective disorder and treatment-resistant schizophrenia were excluded.

By day 6, dosages reached 6 mg/d for risperidone and 15 mg/d for olanzapine. Divalproex was started at 15 mg/kg and titrated to a maximum of 30 mg/kg by day 14. Mean divalproex dosage was approximately 2300 mg/d (mean plasma level approximately 100 mg/mL).

Positive and Negative Syndrome Scale (PANSS) total scores improved significantly in patients receiving adjunctive divalproex compared with those receiving antipsychotic monotherapy, and significant differences occurred as early as day 3. The major effect was seen on schizophrenia’s positive symptoms. A post-hoc analysis¹⁷ also showed greater reductions in hostility (as measured by the hostility item in the PANSS Positive Subscale) in patients receiving adjunctive divalproex compared with antipsychotic monotherapy. This effect was independent of the effect on positive symptoms or sedation.

A large, multi-site, 84-day acute schizophrenia RCT similar to the 28-day trial—but using extended-release divalproex—is being conducted. An extended-release preparation may be particularly helpful in encouraging medication adherence for patients taking complicated medication regimens.

Table 3

Double-blind studies of adjunctive valproate in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Ko et al (1985)	6	28	Crossover	Neurolepticresistant patients with chronic schizophrenia, not exacerbation	No valproate effect noted
Fisk and York (1987)	62	42	Antipsychotic + valproate vs antipsychotic + placebo	Chronic psychosis and tardive dyskinesia	No differences in mental state and behavior, as measured by“ Krawiecka scale”*
Dose et al (1998)	42	28	HAL + valproate vs HAL + placebo	Acute, nonmanic schizophrenic or schizoaffective psychosis	No difference on BPRS; possible effect on “hostile belligerence”
Wassef et al (2000)	12	21	HAL + valproate vs HAL + placebo	Acute exacerbation of chronic schizophrenia	CGI and SANS scores improved significantly, but BPRS scores did not
Casey et al (2003)	249	28	RIS + valproate vs OLZ + valproate vs RIS + placebo vs OLZ + placebo	Acute exacerbation of schizophrenia	PANSS scores improved
* Krawiecka M, Goldberg D, Vaughan M. A standardized psychiatric assessment scale for rating chronic psychotic patients. Acta Psychiatr Scand 1977;55(4):299-308.
n = number of subjects
BPRS = Brief Psychiatric Rating Scale
CGI = Clinical Global Impression scale
HAL = haloperidol
OLZ = olanzapine
PANSS = Positive and Negative Syndrome Scale
RIS = risperidone
SANS = Scale for the Assessment of Negative Symptoms

In a recent large (n = 10,262), retrospective, pharmacoepidemiologic analysis,¹⁸ valproate augmentation led to longer persistence of treatment than did the strategy of switching antipsychotics. Average valproate dosages were small, however (<425 mg/d), as were antipsychotic dosages (risperidone <1.7 mg/d, quetiapine <120 mg/d, and olanzapine <7.5 mg/d). Patients’ diagnostic categories were not available. One interpretation of this study is that valproate augmentation would be more successful than switching antipsychotics, assuming that treatment persistence can be viewed as a positive outcome.

Table 4

Double-blind studies of adjunctive lamotrigine in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Tiihonen et al (2003)	34	84	Crossover; clozapine with or without lamotrigine	Clozapine-resistant male inpatients with chronic schizophrenia, not exacerbation	BPRS, PANSS positive, and PANSS general psychopathology symptom scores improved Negative symptoms did not improve
Kremer et al (2004)	38	70	Antipsychotic* + lamotrigine vs antipsychotic* + placebo	Treatment- resistant inpatients with schizophrenia	Completers’ PANSS positive, general psychopathology and total symptom scores improved No difference in negative symptoms or total BPRS scores No difference with intent-to-treat analyses
n = number of patients
* First- or second-generation antipsychotic
BPRS = Brief Psychiatric Rating Scale
PANSS = Positive and Negative Syndrome Scale

Lamotrigine is the only other anticonvulsant for which published, double-blind, randomized evidence of use in patients with schizophrenia is available (Table 4).^19,20 Adjunctive lamotrigine may be effective in managing treatment-resistant schizophrenia, as was shown in a small (n = 34), double-blind, placebo-controlled, crossover trial.¹⁹ Hospitalized patients whose symptoms were inadequately controlled with clozapine monotherapy received lamotrigine, 200 mg/d, for up to 12 weeks. Adjunctive lamotrigine improved positive but not negative symptoms.

 

Similar results were seen in treatment-resistant inpatients with schizophrenia (n = 38) in a 10-week, double-blind, parallel group trial by Kremer et al.²⁰ Adjunctive lamotrigine improved PANSS positive, general psychopathology, and total symptom scores in the 31 patients who completed the trial. No differences were seen, however, in negative symptoms, total BPRS scores, or in the intent-to-treat analysis. These results have spurred the launch of a large, multi-site, RCT of adjunctive lamotrigine in patients with schizophrenia who have responded inadequately to antipsychotics alone.

Topiramate, one of the few psychotropics associated with weight loss, has attracted interest as an adjunct to second-generation antipsychotics to address weight gain. Although case reports have shown benefit,²¹ one showed deterioration in both positive and negative symptoms when topiramate was added to second-generation antipsychotics.²²

Table 5

Double-blind study of adjunctive topiramate in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Tiihonen (2004)*	26	84	Crossover; SGA plus topiramate or placebo	Treatment- resistant male inpatients with chronic schizophrenia	PANSS general scores improved No difference in total PANSS, PANSS positive, or PANSS negative scores
* 2004 Collegium Internationale Neuro-Psychopharmacologicum (CINP) presentation, and personal communication (6/22/04)
n = number of patients
PANSS = Positive and Negative Syndrome Scale
SGA = Second-generation antipsychotic (patients were taking clozapine, olanzapine, or quetiapine)

An unpublished, randomized, crossover trial compared second-generation antipsychotics plus topiramate or placebo in 26 male inpatients with chronic schizophrenia. With adjunctive topiramate, the authors found a statistically significant improvement in the PANSS general psychopathology subscale but not in PANSS total, positive subscale, or negative subscale scores (Table 5) (Tiihonen J, personal communication 6/22/04).

Other agents. Very little information—all uncontrolled—supports adjunctive use of gabapentin or oxcarbazepine for patients with schizophrenia.^23-28 Of concern are reports of patients suffering worsening of psychosis with gabapentin²⁵ or of dysphoria and irritability with oxcarbazepine (attributed to a pharmacokinetic interaction).²⁶

Conclusion. More trials are needed to examine the use of adjunctive mood stabilizers in patients with schizophrenia—particularly in those with chronic symptoms. Although mood stabilizers are widely used in this population, important questions remain unanswered, including:

• characteristics of patients likely to require adjunctive treatment
• how long treatment should continue.

MECHANISMS OF ACTION

Unlike antipsychotics, mood stabilizers do not exert their therapeutic effects by acting directly on dopamine (D2) receptors. Differences in mechanism of action among the anticonvulsants may help explain why some—such as valproate and lamotrigine—have been useful for bipolar disorder or schizophrenia and others—such as gabapentin—have not.²⁹

One possibility is that anticonvulsants that affect voltage-gated sodium channels—such as valproate, lamotrigine, carbamazepine and oxcarbazepine—may be most useful for patients with bipolar disorder or schizophrenia. On the other hand, agents that affect voltage-gated calcium channels—such as gabapentin—may be efficacious as anticonvulsants but not as efficacious for bipolar disorder or schizophrenia.

Ketter et al³⁰ proposed an anticonvulsant classification system based on predominant psychotropic profiles:

• the “GABA-ergic” group predominantly potentiates the inhibitory neurotransmitter GABA, resulting in sedation, fatigue, cognitive slowing, and weight gain, as well as possible anxiolytic and antimanic effects
• the “anti-glutamate” group predominantly attenuates glutamate excitatory neurotransmission and is associated with activation, weight loss, and possibly anxiogenic and antidepressant effects.

In the GABA-ergic group are anticonvulsants such as barbiturates, benzodiazepines, valproate, gabapentin, tiagabine, and vigabatrin. The antiglutamate group includes agents such as felbamate and lamotrigine. A “mixed” category includes anticonvulsants with GABA-ergic and anti-glutaminergic actions such as topiramate, which has sedating and weight-loss properties.

Because GABA appears to modulate dopamine neurotransmission,³¹ this may explain valproate’s role as an adjunctive agent for schizophrenia. Similarly, mechanisms related to Nmethyl-D-aspartate (NMDA) and non-NMDA glutamate receptor function may explain lamotrigine’s usefulness in this setting.^19,20

SUMMARY

Clinicians resort to combination therapies when monotherapies fail to adequately control symptoms or maintain response. Co-prescribing of anticonvulsants with antipsychotics for inpatients with schizophrenia is common practice in New York State and most likely elsewhere. In general, antipsychotics’ and mood stabilizers’ different—and perhaps complementary—mechanisms of action explain the synergism between them. Mechanisms of action also may explain why some anticonvulsants help in schizophrenia (or bipolar disorder) whereas others do not.

Evidence for using adjunctive anticonvulsants is variable. The strongest data support using valproate (and perhaps lamotrigine), followed by carbamazepine and then topiramate. Gabapentin and oxcarbazepine have only anecdotal evidence, some of it negative. Well-designed, randomized clinical trials with the appropriate populations are needed.

Related resources

• Stahl SM. Essential psychopharmacology of antipsychotics and mood stabilizers. New York: Cambridge University Press, 2002.
• Harvard Medical School Department of Psychiatry’s psychopharmacology algorithm project. Osser DN, Patterson RD. Consultant for the pharmacotherapy of schizophrenia. Available at http://mhc.com/Algorithms/. Accessed Nov. 5, 2004.

Drug brand names

• Carbamazepine • Tegretol
• Clozapine • Clozaril
• Gabapentin • Neurontin
• Haloperidol • Haldol
• Lamotrigine • Lamictal
• Lithium • Lithobid, Eskalith
• Olanzapine • Zyprexa
• Oxcarbazepine • Trileptal
• Quetiapine • Seroquel
• Risperidone • Risperdal
• Topiramate • Topamax
• Valproate (valproic acid, divalproex sodium) • Depakene, Depakote

 

Disclosure

Dr. Citrome receives research grants/contracts from Abbott Laboratories, AstraZeneca Pharmaceuticals, Bristol-Myers Squibb Co., Eli Lilly & Co., Janssen Pharmaceutica, and Pfizer Inc. He is a consultant to and/or speaker for Bristol-Myers Squibb Co., Eli Lilly & Co., Pfizer Inc., Abbott Laboratories, AstraZeneca Pharmaceuticals, and Novartis Pharmaceuticals Corp.

Acknowledgment

Adapted from Citrome L. “Antipsychotic polypharmacy versus augmentation with anticonvulsants: the U.S. perspective” (presentation). Paris: Collegium Internationale Neuro-Psychopharmacologicum (CINP), June 2004 [abstract in Int J Neuropsychopharmacol. 2004; 7(suppl 1):S69], and from Citrome L. “Mood-stabilizer use in schizophrenia: 1994-2002” (NR350) (poster). New York: American Psychiatric Association annual meeting, May 2004.

In patients with treatment-resistant schizophrenia, lithium and anticonvulsants have become such common adjuncts to antipsychotics that 50% of inpatients may be receiving them.¹ Evidence supporting this practice is mixed:

• Initial case reports and open-label studies that showed benefit have not always been followed by randomized clinical trials.
• Lack of clear benefit also has been described (Table 1).

This article examines the extent of this prescribing pattern, its evidence base, and mechanisms of action that may help explain why some adjunctive mood stabilizers are more effective than others for treatment-resistant schizophrenia.

EXTENT OF USE

For 5 years, the rate at which inpatients with schizophrenia received adjunctive mood stabilizers has held steady at approximately 50% in New York State Office of Mental Health (NYSOMH) facilities (Figure1). These facilities provide intermediate and long-term care for the seriously, persistently mentally ill. Adjunctive mood stabilizers might not be used as often for outpatients or for inpatients treated in short-stay facilities.

Table 1

Evidence for adjunctive use of lithium or anticonvulsants for treating schizophrenia

Agent	Case reports and open studies	Randomized, double-blind trials	Benefit?
Lithium	Yes (many)	Yes (several, +/-)	Probably not
Carbamazepine	Yes (many)	Yes (several; small total sample)	Limited
Valproate	Yes (many)	Yes	Yes
Gabapentin	Yes (very few, +/-)	None	Probably not
Lamotrigine	Yes (few, +/-)	Yes (2, +)	Yes
Topiramate	Yes (very few, +/-)	Yes (1, +/-)	Probably not
Oxcarbazepine	Yes (very few, +/-)	None	Probably not
+ = positive results
- = negative results
+/- = both positive and negative results

Valproate is the most commonly used anticonvulsant, with one out of three patients with schizophrenia receiving it. Adjunctive gabapentin use is declining, probably because of inadequate efficacy—as will be discussed later. Use of adjunctive lamotrigine is expected to increase as more data become available on its usefulness in treatment-resistant schizophrenia.

Clinicians are using substantial dosages of adjunctive mood stabilizers. During first-quarter 2004, average daily dosages for 4,788 NYSOMH patients (80% with schizophrenia or schizoaffective disorder) receiving antipsychotics were:

• valproate, 1639 mg (n = 1921)
• gabapentin, 1524 mg (n = 303)
• oxcarbazepine, 1226 mg (n = 201)
• carbamazepine, 908 mg (n = 112)
• lithium, 894 mg (n = 715)
• topiramate, 234 mg (n = 269)
• lamotrigine, 204 mg (n = 231).²

Mood-stabilizer combinations were also used. Approximately one-half of patients receiving adjunctive mood stabilizers—with the exception of valproate—were receiving more than one. In patients receiving valproate, the rate of mood-stabilizer co-prescribing was about 25%.²

WHAT IS THE EVIDENCE?

Evidence supporting the use of adjunctive lithium or anticonvulsants to treat schizophrenia varies in quality and quantity (Table 1). Case reports and open studies offer the weakest evidence but can spur double-blind, randomized clinical trials (RCTs). Unfortunately, RCTs are not often done, and the published studies usually suffer from methodologic flaws such as:

• inadequate number of subjects (insufficient statistical power to detect differences)
• lack of control of confounds such as mood symptoms (seen when studies include patients with schizoaffective disorder)
• inadequate duration
• inappropriate target populations (patients with acute exacerbations of schizophrenia instead of persistent symptoms in treatment-resistant schizophrenia).

Because controlled trials of the use of adjunctive mood stabilizers are relatively scarce, clinical practice has transcended clinical research. Clinicians need effective regimens for treatment-resistant schizophrenia, and mood-stabilizer augmentation helps some patients.

Lithium is perhaps the best-known mood stabilizer. Although early studies showed adjunctive lithium useful in treating schizophrenia, later and better-designed trials did not. The authors of a recent meta-analysis of randomized clinical trials (n = 611 in 20 studies) concluded that despite some evidence supporting the efficacy of lithium augmentation, overall results were inconclusive. A large trial would be required to detect a small benefit in patients with schizophrenia who lack affective symptoms.³

Carbamazepine use among patients with schizophrenia is declining, primarily because this drug induces its own metabolism and can require frequent dose adjustments. Adjunctive carbamazepine has been used to manage persistent aggressive behavior in patients with schizophrenia and schizoaffective disorder. Evidence comes primarily from small trials or case reports (Table 2),^4-8 but results of a larger clinical trial (n = 162) by Okuma et al⁶ are also available

In the Okuma report—a double-blind, placebo-controlled trial of carbamazepine in patients with DSM-III schizophrenia or schizoaffective disorder—carbamazepine did not significantly improve patients’ total Brief Psychiatric Rating Scale (BPRS) scores. Compared with placebo, however, some benefit with carbamazepine did emerge in measures of suspiciousness, uncooperativeness, and excitement.

A systematic review and meta-analysis (n = 283 in 10 studies) detected a trend toward reduced psychopathology with carbamazepine augmentation for schizophrenia. BPRS scores declined by 20% and 35% in the six trials (n = 147) for which data were available (P = 0.08 and 0.09, respectively).⁹ Because the double-blind trial by Okuma et al⁶ was not randomized, it was not included in this meta-analysis.⁹

Figure 1 10-year trend in use of adjunctive mood stabilizers for schizophrenia

Valproate. Among the anticonvulsants, the greatest body of evidence supports the use of valproate in patients with schizophrenia,¹⁰ although a recent meta-analysis (n = 378 in 5 studies) indicates inconsistent beneficial effects.¹¹

 

Initial double-blind, RCTs of adjunctive valproate in patients with schizophrenia were limited in size and failed to show benefit with adjunctive valproate^12-15 (Table 2). A more recent study¹⁶ showed that adjunctive valproate affects acute psychotic symptoms rather than mood. This study, however, did not answer whether adjunctive valproate would help patients with persistent symptoms of schizophrenia.

Table 2

Double-blind studies of adjunctive carbamazepine in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Neppe (1983)	11	42	Crossover	Mixed, 8 with schizophrenia	“Overall clinical rating” improved
Dose et al (1987)	22	28	HAL + CBZ vs HAL + placebo	Schizophrenia or schizoaffective disorder	No difference on BPRS
Okuma et al (1989)	162	28	NL + CBZ vs NL + placebo	Schizophrenia or schizoaffective disorder	No difference on BPRS; possible improvement in suspiciousness, uncooperativeness, and excitement
Nachshoni et al (1994)	28	49	NL + CBZ vs NL + placebo	“Residual schizophrenia with negative symptoms”	No difference on BPRS or SANS
Simhandl et al (1996)	42	42	NL + CBZ vs NL + lithium vs NL + placebo	Schizophrenia(treatment- nonresponsive)	No difference on BPRS; CGI improved from baseline in groups receiving CBZ and lithium
n = number of subjects
BPRS = Brief Psychiatric Rating Scale
CGI = Clinical Global Impression scale
CBZ = carbamazepine
HAL = haloperidol
NL = neuroleptic (first-generation antipsychotic)
SANS = Scale for the Assessment of Negative Symptoms

In this large (n = 249), multicenter, randomized, double-blind trial, hospitalized patients with an acute exacerbation of schizophrenia received olanzapine or risperidone plus divalproex or placebo for 28 days. Patients with schizoaffective disorder and treatment-resistant schizophrenia were excluded.

By day 6, dosages reached 6 mg/d for risperidone and 15 mg/d for olanzapine. Divalproex was started at 15 mg/kg and titrated to a maximum of 30 mg/kg by day 14. Mean divalproex dosage was approximately 2300 mg/d (mean plasma level approximately 100 mg/mL).

Positive and Negative Syndrome Scale (PANSS) total scores improved significantly in patients receiving adjunctive divalproex compared with those receiving antipsychotic monotherapy, and significant differences occurred as early as day 3. The major effect was seen on schizophrenia’s positive symptoms. A post-hoc analysis¹⁷ also showed greater reductions in hostility (as measured by the hostility item in the PANSS Positive Subscale) in patients receiving adjunctive divalproex compared with antipsychotic monotherapy. This effect was independent of the effect on positive symptoms or sedation.

A large, multi-site, 84-day acute schizophrenia RCT similar to the 28-day trial—but using extended-release divalproex—is being conducted. An extended-release preparation may be particularly helpful in encouraging medication adherence for patients taking complicated medication regimens.

Table 3

Double-blind studies of adjunctive valproate in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Ko et al (1985)	6	28	Crossover	Neurolepticresistant patients with chronic schizophrenia, not exacerbation	No valproate effect noted
Fisk and York (1987)	62	42	Antipsychotic + valproate vs antipsychotic + placebo	Chronic psychosis and tardive dyskinesia	No differences in mental state and behavior, as measured by“ Krawiecka scale”*
Dose et al (1998)	42	28	HAL + valproate vs HAL + placebo	Acute, nonmanic schizophrenic or schizoaffective psychosis	No difference on BPRS; possible effect on “hostile belligerence”
Wassef et al (2000)	12	21	HAL + valproate vs HAL + placebo	Acute exacerbation of chronic schizophrenia	CGI and SANS scores improved significantly, but BPRS scores did not
Casey et al (2003)	249	28	RIS + valproate vs OLZ + valproate vs RIS + placebo vs OLZ + placebo	Acute exacerbation of schizophrenia	PANSS scores improved
* Krawiecka M, Goldberg D, Vaughan M. A standardized psychiatric assessment scale for rating chronic psychotic patients. Acta Psychiatr Scand 1977;55(4):299-308.
n = number of subjects
BPRS = Brief Psychiatric Rating Scale
CGI = Clinical Global Impression scale
HAL = haloperidol
OLZ = olanzapine
PANSS = Positive and Negative Syndrome Scale
RIS = risperidone
SANS = Scale for the Assessment of Negative Symptoms

In a recent large (n = 10,262), retrospective, pharmacoepidemiologic analysis,¹⁸ valproate augmentation led to longer persistence of treatment than did the strategy of switching antipsychotics. Average valproate dosages were small, however (<425 mg/d), as were antipsychotic dosages (risperidone <1.7 mg/d, quetiapine <120 mg/d, and olanzapine <7.5 mg/d). Patients’ diagnostic categories were not available. One interpretation of this study is that valproate augmentation would be more successful than switching antipsychotics, assuming that treatment persistence can be viewed as a positive outcome.

Table 4

Double-blind studies of adjunctive lamotrigine in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Tiihonen et al (2003)	34	84	Crossover; clozapine with or without lamotrigine	Clozapine-resistant male inpatients with chronic schizophrenia, not exacerbation	BPRS, PANSS positive, and PANSS general psychopathology symptom scores improved Negative symptoms did not improve
Kremer et al (2004)	38	70	Antipsychotic* + lamotrigine vs antipsychotic* + placebo	Treatment- resistant inpatients with schizophrenia	Completers’ PANSS positive, general psychopathology and total symptom scores improved No difference in negative symptoms or total BPRS scores No difference with intent-to-treat analyses
n = number of patients
* First- or second-generation antipsychotic
BPRS = Brief Psychiatric Rating Scale
PANSS = Positive and Negative Syndrome Scale

Lamotrigine is the only other anticonvulsant for which published, double-blind, randomized evidence of use in patients with schizophrenia is available (Table 4).^19,20 Adjunctive lamotrigine may be effective in managing treatment-resistant schizophrenia, as was shown in a small (n = 34), double-blind, placebo-controlled, crossover trial.¹⁹ Hospitalized patients whose symptoms were inadequately controlled with clozapine monotherapy received lamotrigine, 200 mg/d, for up to 12 weeks. Adjunctive lamotrigine improved positive but not negative symptoms.

 

Similar results were seen in treatment-resistant inpatients with schizophrenia (n = 38) in a 10-week, double-blind, parallel group trial by Kremer et al.²⁰ Adjunctive lamotrigine improved PANSS positive, general psychopathology, and total symptom scores in the 31 patients who completed the trial. No differences were seen, however, in negative symptoms, total BPRS scores, or in the intent-to-treat analysis. These results have spurred the launch of a large, multi-site, RCT of adjunctive lamotrigine in patients with schizophrenia who have responded inadequately to antipsychotics alone.

Topiramate, one of the few psychotropics associated with weight loss, has attracted interest as an adjunct to second-generation antipsychotics to address weight gain. Although case reports have shown benefit,²¹ one showed deterioration in both positive and negative symptoms when topiramate was added to second-generation antipsychotics.²²

Table 5

Double-blind study of adjunctive topiramate in schizophrenia

Author (yr)	n	Duration (days)	Design	Diagnosis	Outcome
Tiihonen (2004)*	26	84	Crossover; SGA plus topiramate or placebo	Treatment- resistant male inpatients with chronic schizophrenia	PANSS general scores improved No difference in total PANSS, PANSS positive, or PANSS negative scores
* 2004 Collegium Internationale Neuro-Psychopharmacologicum (CINP) presentation, and personal communication (6/22/04)
n = number of patients
PANSS = Positive and Negative Syndrome Scale
SGA = Second-generation antipsychotic (patients were taking clozapine, olanzapine, or quetiapine)

An unpublished, randomized, crossover trial compared second-generation antipsychotics plus topiramate or placebo in 26 male inpatients with chronic schizophrenia. With adjunctive topiramate, the authors found a statistically significant improvement in the PANSS general psychopathology subscale but not in PANSS total, positive subscale, or negative subscale scores (Table 5) (Tiihonen J, personal communication 6/22/04).

Other agents. Very little information—all uncontrolled—supports adjunctive use of gabapentin or oxcarbazepine for patients with schizophrenia.^23-28 Of concern are reports of patients suffering worsening of psychosis with gabapentin²⁵ or of dysphoria and irritability with oxcarbazepine (attributed to a pharmacokinetic interaction).²⁶

Conclusion. More trials are needed to examine the use of adjunctive mood stabilizers in patients with schizophrenia—particularly in those with chronic symptoms. Although mood stabilizers are widely used in this population, important questions remain unanswered, including:

• characteristics of patients likely to require adjunctive treatment
• how long treatment should continue.

MECHANISMS OF ACTION

Unlike antipsychotics, mood stabilizers do not exert their therapeutic effects by acting directly on dopamine (D2) receptors. Differences in mechanism of action among the anticonvulsants may help explain why some—such as valproate and lamotrigine—have been useful for bipolar disorder or schizophrenia and others—such as gabapentin—have not.²⁹

One possibility is that anticonvulsants that affect voltage-gated sodium channels—such as valproate, lamotrigine, carbamazepine and oxcarbazepine—may be most useful for patients with bipolar disorder or schizophrenia. On the other hand, agents that affect voltage-gated calcium channels—such as gabapentin—may be efficacious as anticonvulsants but not as efficacious for bipolar disorder or schizophrenia.

Ketter et al³⁰ proposed an anticonvulsant classification system based on predominant psychotropic profiles:

• the “GABA-ergic” group predominantly potentiates the inhibitory neurotransmitter GABA, resulting in sedation, fatigue, cognitive slowing, and weight gain, as well as possible anxiolytic and antimanic effects
• the “anti-glutamate” group predominantly attenuates glutamate excitatory neurotransmission and is associated with activation, weight loss, and possibly anxiogenic and antidepressant effects.

In the GABA-ergic group are anticonvulsants such as barbiturates, benzodiazepines, valproate, gabapentin, tiagabine, and vigabatrin. The antiglutamate group includes agents such as felbamate and lamotrigine. A “mixed” category includes anticonvulsants with GABA-ergic and anti-glutaminergic actions such as topiramate, which has sedating and weight-loss properties.

Because GABA appears to modulate dopamine neurotransmission,³¹ this may explain valproate’s role as an adjunctive agent for schizophrenia. Similarly, mechanisms related to Nmethyl-D-aspartate (NMDA) and non-NMDA glutamate receptor function may explain lamotrigine’s usefulness in this setting.^19,20

SUMMARY

Clinicians resort to combination therapies when monotherapies fail to adequately control symptoms or maintain response. Co-prescribing of anticonvulsants with antipsychotics for inpatients with schizophrenia is common practice in New York State and most likely elsewhere. In general, antipsychotics’ and mood stabilizers’ different—and perhaps complementary—mechanisms of action explain the synergism between them. Mechanisms of action also may explain why some anticonvulsants help in schizophrenia (or bipolar disorder) whereas others do not.

Evidence for using adjunctive anticonvulsants is variable. The strongest data support using valproate (and perhaps lamotrigine), followed by carbamazepine and then topiramate. Gabapentin and oxcarbazepine have only anecdotal evidence, some of it negative. Well-designed, randomized clinical trials with the appropriate populations are needed.

Related resources

• Stahl SM. Essential psychopharmacology of antipsychotics and mood stabilizers. New York: Cambridge University Press, 2002.
• Harvard Medical School Department of Psychiatry’s psychopharmacology algorithm project. Osser DN, Patterson RD. Consultant for the pharmacotherapy of schizophrenia. Available at http://mhc.com/Algorithms/. Accessed Nov. 5, 2004.

Drug brand names

• Carbamazepine • Tegretol
• Clozapine • Clozaril
• Gabapentin • Neurontin
• Haloperidol • Haldol
• Lamotrigine • Lamictal
• Lithium • Lithobid, Eskalith
• Olanzapine • Zyprexa
• Oxcarbazepine • Trileptal
• Quetiapine • Seroquel
• Risperidone • Risperdal
• Topiramate • Topamax
• Valproate (valproic acid, divalproex sodium) • Depakene, Depakote

 

Disclosure

Dr. Citrome receives research grants/contracts from Abbott Laboratories, AstraZeneca Pharmaceuticals, Bristol-Myers Squibb Co., Eli Lilly & Co., Janssen Pharmaceutica, and Pfizer Inc. He is a consultant to and/or speaker for Bristol-Myers Squibb Co., Eli Lilly & Co., Pfizer Inc., Abbott Laboratories, AstraZeneca Pharmaceuticals, and Novartis Pharmaceuticals Corp.

Acknowledgment

Adapted from Citrome L. “Antipsychotic polypharmacy versus augmentation with anticonvulsants: the U.S. perspective” (presentation). Paris: Collegium Internationale Neuro-Psychopharmacologicum (CINP), June 2004 [abstract in Int J Neuropsychopharmacol. 2004; 7(suppl 1):S69], and from Citrome L. “Mood-stabilizer use in schizophrenia: 1994-2002” (NR350) (poster). New York: American Psychiatric Association annual meeting, May 2004.