A proportion of manic patients show only partial improvement or initial improvement followed by partial relapse with antipsychotic drugs. There is little evidence that increasing the dose will produce further improvement. Clozapine may prove to be useful in resistant mania as in resistant schizophrenia. However, at the present time, the main alternatives or adjuncts to the antipsychotic drugs are lithium and the anticonvulsants, valproate, and carbamazepine.
Current American guidelines recommend a combination of lithium or valproate with an antipsychotic as the first-line treatment for “severe” mania, but a choice between lithium, valproate, and an atypical antipsychotic for “mild” mania.
It is noteworthy that no definition of severe mania is offered. Other guidelines, including the British, recommend a choice of either an antipsychotic or valproate for severe mania, an antipsychotic for psychotic mania, and for the “less ill” either lithium, valproate, or carbamazepine.
Clinical Trial Evidence for Combination Treatment
Evidence concerning the efficacy of combination treatment is complex. Although several studies have demonstrated an advantage of combining an antipsychotic with lithium or valproate over lithium or valproate alone, only one study has investigated the advantage of the combination over antipsychotic alone.
The majority of these studies involved a design in which patients, who had not responded to one drug administered for at least three weeks, had additional treatment with the combination of drugs using a placebo control for the added drug. In some studies, a proportion of patients commenced on the combination without previous treatment, and the control group received only the lithium, valproate, or carbamazepine with placebo instead of antipsychotic. Thus, there have been no studies of combination treatment in which a control group receives only placebo. It is therefore not possible to determine directly the size of effect of giving combination treatment to drug-free patients with mania over giving placebo alone, or to determine the size of advantage of combination therapy over monotherapy initiated in drug-free patients with mania.
The studies of combination treatment shown in Table: Studies of Combination of Antipsychotic with Lithium, Valproate, or Carbamazepine have proved that several antipsychotics provided additional efficacy when added to lithium or valproate compared with lithium or valproate alone. These are: haloperidol, risperidone, olanzapine, and quetiapine. A similar study with ziprasidone was negative.
Risperidone in combination with lithium or valproate was shown to be more effective than either lithium or valproate alone and as efficacious as haloperidol in combination with either lithium or valproate, in patients with manic or mixed episodes. In combination with carbamazepine, risperidone did not provide significant additional efficacy perhaps because risperidone blood levels were reduced by about 50% in patients on carbamazepine. This was presumably because of its induction of liver enzymes that metabolize drugs. The advantage of risperidone over placebo was far more evident in those who had already been on a mood stabilizer for at least 2 weeks before randomization, than in those who started mood stabilizer shortly before starting risperidone or placebo. Response to risperidone was independent of the presence or absence of psychotic symptoms.
In the study of Sachs et al., 63% of patients had already received lithium or valproate before randomization to risperidone or placebo, and might therefore be considered non-responders to the mood stabilizer; the rest commenced on lithium or valproate at the same time as being randomized to start on risperidone or placebo. The benefit of additional antipsychotic medication was much less apparent in the latter group. Significant improvement in response to either risperidone or haloperidol was limited to patients with pure mania and was not evident in those with mixed manic states, who comprised 21% of patients.
In combination studies with olanzapine, this drug was shown to provide additional efficacy when added in patients who had already received lithium or valproate for at least three weeks. The effect was statistically significant in the subgroup on valproate but only a trend in the smaller subgroup on lithium.
Conversely, the addition of valproate to classical antipsychotics (mainly haloperidol) has been shown to produce greater improvement than addition of placebo.