Rational polypharmacy in the bipolar affective disorders.

Bipolar affective illness represents a syndrome not readily treated by single agents. Approximately 50% of patients are inadequately responsive to lithium and the majority of patients require supplemental antidepressants, antimanic, antipsychotic or hypnotic medications. These traditional adjunctive medications are associated with potential problems. Antidepressants may precipitate mania (at a rate about double that of placebo) or cause cycle acceleration. Neuroleptics may be associated with either more profound or longer depressive phases, and clearly increase the risk of tardive dyskinesia, to which bipolar patients appear particularly predisposed. Moreover, there are subgroups of patients who are known to be poorly responsive to lithium. These include patients with rapid cycling, dysphoric mania, co-morbid drug or alcohol abuse, a pattern of depression-mania-well interval (D-M-I as opposed to the M-D-I pattern), and patients without a family history of bipolar illness in first-degree relatives. There is increasing recognition that the anticonvulsants carbamazepine and valproate are effective alternatives or adjuncts to lithium in the acute and long-term treatment of bipolar illness. Ideally, one would want to assess whether patients who were unresponsive to lithium were responsive to an anticonvulsant alone prior to utilizing lithium in addition to anticonvulsant combination therapy. However, from the clinical perspective, it is often more expedient to use an anticonvulsant adjunctively to lithium to assess the efficacy of this combination and establish mood stabilization. When lithium is not discontinued, the increased morbidity during lithium withdrawal also would not occur and would not confound the evaluation of the new agent. We suggest the initial use of acute adjuncts to lithium with the anticonvulsants carbamazepine or valproate (instead of neuroleptics) so that their efficacy can be assessed in the individual's acute episode, with the likelihood of a positive response in longer-term prophylaxis. Hypnotic benzodiazepines with anticonvulsant properties, such as clonazepam or lorazepam, are often used to help to induce sleep in escalating bipolar patients, and may be useful adjuncts as well. Patients who were inadequately responsive to either carbamazepine or valproate alone may be responsive to the anticonvulsant combination. In a similar fashion, one can also utilize several mood-stabilizing drugs (lithium and an anticonvulsant such as carbamazepine or valproate) in the treatment of depressive breakthroughs, and then augment this combination (if necessary) with a catecholamine-active antidepressant such as bupropion or a serotonin-selective reuptake inhibitor (SSRI) such as fluoxetine, paroxetine, sertraline or if necessary a monoamine oxidase inhibitor (MAOI). Once the patient has responded to a combination of drugs, it becomes problematic to decide whether the last agent added was the crucial ingredient in helping the patient achieve remission or that remission might have occurred with this agent alone. A conservative approach would have merit in patients who are finally stabilized on complex polypharmacy regimens only after many years of sequential trials; in this instance, the potential risk of re-exacerbating the illness with a taper of one of the drugs in the regimen. Rational polypharmacy should thus be implemented with careful delineation of the prior course of illness (typically using life chart methodology) and targeted treatment outcomes titrated against side effects, using sequential clinical trials in individual patients who have not adequately responded to monotherapy. In this fashion, it is hoped that pharmacodynamic differences among agents can be maximized and pharmacokinetic and side effects minimized.