Treatment-Resistant Bipolar Disorder

Definition
Treatment of bipolar illness involves four primary aims: treatment of mania, treatment of depression, prevention of manic recurrences, and prevention of depressive recurrences. Lithium has come to be regarded as the treatment of choice for bipolar illness, since double-blind controlled trials have demonstrated its efficacy for all four treatment goals (Goodwin et al. 1969; Maggs 1963; Stallone et al. 1973; Stokes et al. 1971). The concept of treatment resistance is often applied to bipolar illness when one or more of these goals are not achieved despite adequate treatment with lithium or another appropriate medication.

Kraepelin’s (1921) naturalistic data suggested that, if untreated, many individuals with bipolar disorder are likely to have one or more acute episodes yearly. Kraepelin believed that resolution of mood symptoms with complete restitution of functioning was the defining feature of the illness. As applied to bipolar illness, treatment resistance can be defined as failure to effect recovery from an episode within a reasonable time frame despite receiving adequate treatment and/or as failure to prevent recurrences despite institution of appropriate prophylactic regimens.

Estimating the prevalence of treatment-resistant bipolar illness is hampered by the clinical complexity of the illness. Research on unipolar illness has been aided by relatively straightforward definitions of recovery, relapse, and recurrence. However, the nature of bipolar illness, with its various irregular episodic courses and the potential for mislabeling a switch in polarity as recovery from the preceding phase, makes it inappropriate to apply the same definitions to bipolar illness. Thus, the absence of consensus regarding appropriate definitions of recovery and treatment resistance continues to present a significant obstacle to research in this important public health problem. Even interpreting the available literature can be frustrating.

Typical of the conundrums presented by patients with bipolar disorder is the treatment of rapid cycling. Studies of treatment outcome may define recovery or marked improvement in rapid cycling as the cessation (or near cessation) of the occurrence of new episodes. This definition appears reasonable but fails to distinguish treatment success (euthymia without functional impairment) from a clear treatment failure (the induction of a prolonged depression). Furthermore, interpreting the results of interventions for any particular phase of rapid cycling is always clouded by the possibility that the improvement is merely part of the natural transition of the illness to the next episode.

Prevalence estimates for treatment-resistant bipolar illness are highly dependent on the definition chosen. The duration of acute episodes and the frequency of recurrence provide two reasonable and distinct measures of treatment response. First, the median duration of acute episodes provides a context for defining treatment resistance. Keller et al. (1986) followed the course of illness in 155 patients during open naturalistic treatment. The reported median duration of acute episodes was 10 weeks for pure mania, 19 weeks for pure depression, and 36 weeks for mixed episodes.

Since the time of Kraepelin, longitudinal studies have informed us that recurrence rather than sustained remission is the rule. Many patients diagnosed in adolescence could expect to have 10 or more acute episodes during their lifetimes (Carlson et al. 1974; Fukuda et al. 1983; Kraepelin 1921; Perris 1968; Zis et al. 1980). The enormous variation between individuals limits the usefulness of predicted averages of duration of episodes and cycle frequency in determining whether a given patient has treatment-resistant bipolar illness. For more consistency in diagnosis, we offer the following definitions for treatment-refractory bipolar illness:

•  Treatment-refractory mania: Mania without remission despite 6 weeks of adequate therapy with at least two antimanic agents used together (lithium, antipsychotic, anticonvulsant, etc.) in the absence of antidepressants or other mood-elevating agents.

•  Treatment-refractory bipolar depression: Depression without remission despite two adequate antidepressant treatment trials including at least one augmentation strategy.

•  Treatment-refractory mood cycling: Continued cycling despite maximal tolerated lithium combined with valproate and/or carbamazepine for a period of three times the cycle length and not less than 6 months in the absence of antidepressants or other cycle-promoting agents.

•  Recovery from acute episode: A period of at least 8 weeks with sustained remission of mood symptoms. During the first 8 weeks of recovery, remission is characterized by no more than 1 day in any week with significant mood abnormality (depression, lack of interest, irritability, expansiveness, euphoria), no more than two neurovegetative symptoms, and an Axis V (Global Assessment of Functioning) rating above 60.

Fortunately, the vast majority of acute episodes appear to be treatment-responsive. Sachs et al. (1994b) reported the course of illness during a year of open treatment for 100 unselected outpatients with bipolar disorder. During the follow-up year, 58% experienced one or more new acute episodes and 32% remained euthymic. As Kraepelin might have expected, however, all but 10% met the criteria for recovery at some time during the follow-up year.

Chronicity, defined as no recovery during 1 year of follow-up, provides a measure of treatment resistance among patients who experience new acute episodes. Several studies provide estimates of the incidence of treatment resistance based on chronicity of acute episodes during open naturalistic treatment. Keller et al. (1986) reported a 20% chronicity rate among patients after an acute episode. At 18-month follow-up, pure mania had the lowest rate of chronicity (8%), followed by pure bipolar depressive episodes (22%). The highest rate of chronicity (38%) was found among patients with a mixed index episode. Coryell et al. (1989) studied patients with type II bipolar disorder and found that as many as 10% remained ill for 5 years or longer. Failure of prophylaxis is the most common problem seen in clinical practice, but treatment-resistant acute episodes may be a significant problem as well, because one out of every four or five acute episodes appears to follow a chronic course.