Clinical Practice Guidelines for Bipolar Disorder

MANAGEMENT OF PERSONS WITH PSYCHOSES
Current Bipolar Depressive Episode Annotations

MODULE F

View Flow Chart for Module F.

Go to Module E.

Go to Module G.

A. Determine Appropriate Setting for Care

See Screening Algorithm for Psychosis and Bipolar Disorder, Module A.

B. Begin Psychoeducational Tasks and Evaluate for Psychotherapy and/or Psychosocial Rehabilitation

See Current Manic/Hypomanic/Mixed Episode Module E, annotation C.

C. Ensure Normal Thyroid Function

See Current Manic/Hypomanic/Mixed Episode Module E, annotation C.

D. Is Person Experiencing Insomnia, Agitation, or Anxiety?

Insomnia is difficulty falling asleep and staying asleep or waking too early, accompanied by a subjective feeling of not being rested after sleep.

Agitation is defined as disruptive behavior over which the person has poor control. This may take the form of verbal or physical aggression, increased activity, excessive talking, or intrusiveness. The agitated client may respond briefly to admonition, but cannot control impulsive behavior for any extended period of time. While agitation may accompany many other psychiatric disorders, it is a hallmark of the manic state.

Anxiety is a state of abnormal and inappropriate hyperarousal. Signs and symptoms of anxiety may include:

1. Tremor
2. Palpitations
3. Nausea
4. Diarrhea
5. Hyperventilation
6. Chest pain
7. Feelings of impending doom
8. Tingling
9. Fear of losing control

E. Consider Benzodiazepine Treatment

Benzodiazepines should be prescribed on a standing dose basis, rather than as needed (PRN), since the goal of anxiolytic therapy is to prevent anxiety. Duration of anxiolytic therapy should be individualized with a "drug holiday" scheduled every 3 to 6 months to determine whether the anxiety, agitation, and/or insomnia is still a clinical problem. Upon discontinuation, anxiolytics should always be tapered over the course of one to four weeks.

In persons with a history of multiple substance abuse, benzodiazepines should be used with great caution and frequent monitoring, since most (85 percent) benzodiazepine abuse occurs in persons with the diagnosis of multiple substance abuse. For persons with a recent (within 12 months) diagnosis of substance dependence, benzodiazepines should be used with care and in the context of an addiction recovery treatment program. Benzodiazepines may be coadministered with disulfiram if this is helpful for sobriety maintenance. However, the clinician should be assured that there is no clinical evidence that use of benzodiazepines will cause an "addiction" de novo in a person with no history of substance abuse. Persons may need reassurance that the addiction potential of benzodiazepines is low when they are prescribed appropriately and used as directed.

F. Does Person Have Psychotic Features Present?

See Current Manic/Hypomanic/ Mixed Episode Module E, annotation I.

G. Initiate/Optimize a Mood Stabilizer, Reassess at 2 to 4 Weeks

A mood stabilizer (lithium, carbamazepine, valproate) should be initiated prior to antidepressant treatment. For the treatment of acute bipolar depression, lithium is the preferred mood stabilizer. Eight controlled studies of lithium for the treatment of acute bipolar depression have been conducted. Lithium was shown to be superior to placebo in several studies, with an overall response rate of over 75 percent (Srisurapanont et al. 1995). Although the data are not strong, there are some reports that either carbamazepine or valproate also may be efficacious in the treatment of bipolar depression. Three small controlled studies on the use of carbamazepine showed a response in 58 percent of persons (Ballenger & Post 1980; Post et al. 1983; Kramlinger & Post 1989), and two open studies showed a response rate of 55 percent (Okuma et al. 1975; Dilsaver et al. 1996). There are no controlled studies on the use of valproate for bipolar depression. An open study of rapid cycling bipolar persons with acute depression showed limited efficacy with an overall response rate of 19 percent (Calabrese et al. 1992). Thus, for persons in whom there is a medical contraindication or history of lithium intolerance or inefficacy, carbamazepine or valproate should be initiated. The time interval for response to mood stabilizers in the treatment of bipolar depression is not well established but appears to be longer than when treating mania. Although some persons will experience improvement in symptoms in two to three weeks, others may not experience full effects for up to six weeks after initiation of pharmacotherapy (Goodwin & Jamison 1990). For this reason, it is recommended that symptoms be monitored for at least four to six weeks prior to reassessment if clinically feasible.

Some persons with bipolar depression will experience a resolution of symptoms in response to an increase in the dose of their current mood stabilizer. Thus, given the risks associated with the use of antidepressants in this population (i.e., induction of mania or rapid cycling), optimization of the current mood stabilizer is recommended as the initial step of treatment. For persons currently on lithium or valproate, the dosage should be adjusted to achieve a target 12-hour trough serum concentration of 0.8 to 1.2 mEq/L or 45 to 120 mcg/mL, respectively. A therapeutic range of 4 to 12 mcg/mL (which has been established for the treatment of seizure disorders) is generally used for carbamazepine; however, the blood level does not correlate well with response to treatment and therefore the dose should be titrated according to clinical response. Symptoms should be reassessed two to four weeks after a therapeutic steady-state serum concentration of mood stabilizer has been achieved. If depressive symptoms are still present, the clinician should proceed to the next algorithm point.

H. Initiate Prophylaxis and Consider Psychosocial Rehabilitation

See Bipolar/ Schizoaffective Disorder Core Module D, annotation F.

I. Initiate Lithium or Antidepressant Treatment; Reassess Response at 2 to 4 Weeks

Lithium is an effective yet underused treatment for bipolar depression. Studies have shown that it is effective as a monotherapeutic agent in up to 60 percent of clients with bipolar depression versus only 30 percent of clients with unipolar depression. Lithium may have synergistic antidepressant effects in combination with anticonvulsant mood stabilizers (Kramlinger & Post 1989; Delucchi and Calabrese 1989). Therefore, clients whose depressive symptoms have not responded to adequate doses of carbamazepine or valproate should be considered for a trial of lithium augmentation prior to the initiation of an antidepressant. As there is uncertainty about optimum serum concentration and time of response when using lithium as an adjunct to anticonvulsants, the literature suggests that lithium be initiated and the dose adjusted to obtain a serum trough concentration of 0.8 to 1.2 mEq/L. However, since lower concentrations may be effective and may reduce the likelihood of side effects, in some cases the clinician may elect to use lower doses. Positive response to adjunctive lithium therapy in the treatment of bipolar depression may not be fully evident for up to six weeks after initiation of pharmacotherapy; however, some persons may show clinical improvement within a shorter time period. It is therefore recommended that symptoms be reassessed two to four weeks after a steady-state lithium level is achieved. Depending upon a person’s clinical status, antidepressant treatment may be indicated if no improvement is seen within this time interval.

Antidepressants should be used conservatively in the treatment of persons with bipolar disorder. Two key issues face the clinician when choosing an antidepressant for persons with bipolar disorder are the tendency of antidepressants to induce mania and rapid cycling in this group and the limited literature regarding the efficacy of available antidepressant agents for treatment of bipolar depression. It is likely that all standard antidepressants are capable of inducing a "switch" from depression to mania in persons with bipolar I disorder (Srisurapanont et al. 1995). The degree to which this risk is present in persons with bipolar II disorder is unclear. It also appears that antidepressants can induce rapid cycling or dysphonia in susceptible persons (Wehr & Goodwin 1987; Bauer et al. 1994; Altshuler et al. 1995; Coryell et al. 1991). Unfortunately, these phenomena have not been systematically studied and therefore data concerning the relative propensity of antidepressants to induce mania or rapid cycling are limited.

With regard to efficacy, ECT has been well studied and is the most effective treatment for the depressed phase of bipolar disorder (Zornberg & Pope 1993). The efficacy of antidepressant drugs is much less clear. Although there is an extensive literature showing efficacy for all standard antidepressants in the treatment of unipolar depression, most of these studies excluded persons with bipolar disorder. There is thus a paucity of controlled studies of antidepressant drugs in the treatment of bipolar depression. Although clinical experience suggests that all standard antidepressants are probably efficacious, the clinician is faced with a dilemma when choosing among the various antidepressant agents since data on overall and relative efficacy are minimal.

All persons with bipolar I disorder (and most with bipolar II disorder) should be treated with an adequate dose of a mood stabilizer prior to the initiation of an antidepressant. It is generally agreed that the starting dose, rate of dose titration, and target therapeutic dose should be similar for bipolar depression and unipolar depression (Frances et al. 1996). However, a more conservative dosage may be necessary, particularly if a person has a history of rapid switches into mania. It is crucial that persons be monitored closely for the emergencies of hypomanic or manic symptoms when antidepressant treatment is used. If signs or symptoms of hypomanic or mania occur, the antidepressant should be discontinued or dosage decreased if symptoms are mild.

Antidepressant Treatments

1. Electroconvulsive therapy (ECT)-- ECT is the most efficacious treatment for bipolar depression (Zornberg & Pope 1993). Several studies have evaluated the efficacy of ECT for bipolar depression; response rates varied from 43 percent to 100 percent. More than half of these studies compared ECT with antidepressants; the remainder found ECT clearly more efficacious, whereas one found ECT to be equally efficacious (Srisurapanont et al. 1995). ECT is often reserved for persons who have depression with psychotic features, contraindications to pharmacotherapy (i.e., unstable medical illness, pregnancy), or medication resistance. However, it should always be considered a first-line treatment option for bipolar depression and does not have to be reserved only for these cases.
2. Monoamine oxidase inhibitors (MAOIs)-- There are data that MAOIs may be the most efficacious antidepressant agents for the treatment of bipolar depression. Two controlled studies and one open study showed the MAOI tranylcypromine to be clearly more efficacious than imipramine in bipolar I and bipolar II persons with anergic, hypersomnic depression, with response rates of 53 percent to 100 percent (Himmelhoch et al. 1982, 1991; Thase et al. 1992). In addition, the MAOIs may be less likely to induce mania than other antidepressants (Stoll et al. 1994). MAOIs appear to be particularly efficacious for bipolar depression with atypical features. However, given the safety issues associated with these agents as well as the availability of newer, better-tolerated antidepressants, the MAOIs are actually considered to be second-line treatment.
3. Tricyclic antidepressants (TCAs)-- Several controlled studies have shown TCAs to be efficacious in the treatment of bipolar depression, with an overall response rate of approximately 55 percent (Zornberg & Pope 1993). However, there are data showing that these agents are more likely than other antidepressants to induce mania (Peet 1994) or accelerate the natural course of bipolar disorder ("cycle acceleration") (Altshuler et al. 1995). For this reason, TCAs should be used with caution in persons with bipolar disorder.
4. Bupropion-- Many clinicians believe bupropion is a first-line agent for the treatment of bipolar depression (Dilsaver 1996; Frances et al. 1996). A small open study of treatment-resistant bipolar persons showed a response rate of 100 percent; all were maintained on bupropion for one year without induction of mania (Wright et al. 1985). A second controlled study of buproprion and desipramine showed both agents to be similarly efficacious; however, the rate of mania cycling was significantly lower in persons taking bupropion (Sachs et al. 1994). The studies suggest that bupropion also may be particularly useful for rapidly cycling bipolar II depression, with less propensity to induce cycling (Haykal & Akiskal 1990); this finding was also observed in a retrospective study in which bupropion was found to induce mania less frequently than TCAs and SSRIs (Stoll et al. 1994).
5. Selective serotonin reuptake inhibitors (SSRIs)-- Despite lack of conclusive data regarding the efficacy of SSRIs in bipolar depression, clinical experience suggests that these agents (fluoxetine, fluvoxamine, paroxetine, sertraline) are probably efficacious (Frances et al. 1996). A large retrospective study including unipolar depressives and persons receiving lithium (Peet 1994) and using data from multiple trials involving SSRIs suggested that their propensity to induce mania may be less accentuated than with TCAs. More data are needed, as several reports indicate that a substantial risk for induction of mania may be associated with these agents, particularly when used at higher doses (Stoll et al. 1994; Howland 1996; Vesely et al. 1997). In addition, the SSRI fluoxetine should be used with caution; with its extremely long half-life (10 to 14 days) more than four weeks are required for washout should mania cycling occur or the person needs to be switched to an MAOI.
6. Venlafaxine, nefazodone, mirtazapine, trazodone--There are no systematic data on these newer agents for the treatment of bipolar depression. However, anecdotal reports indicate that these agents may be efficacious. Clinical experience indicates that venlafaxine may be useful (Frances et al. 1996), although more studies are needed.
7. Therapeutic antidepressant response may not be fully evident for six weeks or longer; however, some persons will respond sooner. Data from a trial in which antidepressants were used on persons with unipolar depression indicate that if no response is evident after four weeks of treatment, the likelihood of eventual response is significantly diminished (Nierenberg 1995). Therefore, if no response is observed after four weeks of therapy at an adequate dose, the agent should be discontinued and a second started; if a partial response is present at four weeks, the trial should be continued for two to four additional weeks.

J. Response

See Current Manic/Hypomanic/Mixed Episode Module E, annotation E.

K. Assess for Need for Continued Antidepressant Treatment--For persons who experience a full response, the issue of when to discontinue the antidepressant agent arises. Maintenance therapy with antidepressant drugs is generally not advised in persons with bipolar disorder. It is recommended that antidepressants be tapered and discontinued within 6 to 12 weeks after remission is achieved (Sachs 1996). On the other hand, there appears to be a subgroup of chronically depressed bipolar persons who will require antidepressant treatment for extended periods. The decision to use long-term antidepressants in these persons should be individualized and dictated by a well-established history of recurrent emergence of depression despite optimal mood stabilizer therapy, and/or repeated relapses when antidepressants are discontinued.

L. Partial Response?

Clinically, "partial response" is defined as some reduction in symptom severity or frequency, but less than 50 percent and usually not of sufficient magnitude to be considered adequate or significant. Partial response may be seen early in treatment or when doses of medication are not optimal.

The person should be monitored closely for the emergence of manic or hypomanic symptoms, which, although more likely to occur during the first week of treatment (Bunney et al. 1972), can also occur at any time during the course of antidepressant therapy. If symptoms of mania occur, the antidepressant should be discontinued immediately. If hypomanic symptoms occur, the antidepressant should be discontinued; however, the clinician may choose to lower the antidepressant dose or increase the mood stabilizer dose and continue to monitor symptoms closely.

M. Augment or Combine Drug Therapies

There are minimal data concerning augmentation or combination antidepressant therapy for bipolar depression; therefore, strategies to treat unipolar depression are generally used. The most effective treatment augmentation strategies include the addition of lithium, triiodothyronine (T₃), or levothyroxine (T₄). While fewer data are available, stimulants (dextroamphetamine, methylphenidate, pemoline), dopaminergic agents (bromocriptine, pergolide), anticonvulsant mood stabilizers (carbamazepine, valproate), and bright light therapy also may be effective adjuncts (Joffe et al. 1996; Frances et al. 1996).

With regard to combination treatments, the use of an MAOI with a TCA has long been thought to be more effective than either drug given alone (Goodwin & Jamison 1990). However, there are few data to support this claim and, due to the potential risks associated with this combination, it is generally not recommended. The combination of a TCA with fluoxetine was found to be quite efficacious in treatment-resistant depressed persons (Weilburg et al. 1989). However, as previously mentioned, TCAs should be used cautiously in bipolar persons due to possible increased risk of mania induction. Anecdotal and case reports also imply that the combination of an SSRI with buspirone or bupropion may increase efficacy in partially responsive clients. Trials of augmentation pharmacotherapy should generally be continued for at least four weeks. If partial efficacy is observed, a second agent can be added, whereas if no response is observed, the augmentation agent should be discontinued and a different one started.

N. Switch Antidepressants; Assess Response in 4 to 6 Weeks

For persons who have not responded to an initial antidepressant, trials of other agents are indicated. There are limited data to suggest that bupropion is efficacious and less likely to induce mania cycling. While fewer data are available, there is general agreement that SSRIs also are efficacious. In addition, both are relatively safe and well tolerated and therefore are generally considered first-line choices for the treatment of bipolar depression. MAOIs, although quite effective due to their lower safety and side effect profile, are generally reserved for persons who have not responded to antidepressants or SSRIs. Thus, for persons who have not responded to bupropion or an SSRI, a trial of MAOI is probably indicated. If this is not feasible, venlafaxine is a reasonable alternative. For persons who have failed all of these agents, a TCA or one of the newer agents should be considered. The number of antidepressant trials that should be initiated depends upon the person’s clinical status, although in general at least three agents (including a MAOI) is recommended.

O. Consider ECT or Alternative Therapies

ECT is highly effective and should always be considered a first-line treatment, particularly in psychotic depression or in medication-resistant persons. In addition, it has been shown that as many as 50 percent of persons who fail treatment with at least one antidepressant will respond to ECT (Prudic et al. 1990). Several alternative therapies exist for persons who have not responded to standard agents. These include combination mood stabilizers, newer anticonvulsants, atypical antipsychotics, sleep deprivation and bright light therapy (phototherapy).

Combination anticonvulsant therapy has become increasingly accepted and today is a popular strategy in the treatment of bipolar disorder. The literature indicates that the combination of carbamazepine and valproate may have synergistic effects in treatment-resistant persons. (Ketter et al. 1992; Tohen et al. 1994). Reports indicate that the newer anticonvulsants such as lamotrigine may have mood-stabilizing properties (Calabrese et al. 1996).

It has also been demonstrated that the atypical antipsychotics clozapine, risperidone, and olanzapine may have antidepressant effects in bipolar depression; a controlled study of olanzapine versus haloperidol in schizoaffective, bipolar persons with depression revealed superior efficacy of olanzapine (Beasly et al. 1997).

Finally, nonpharmacologic therapies may also be effective for bipolar depression. Sleep deprivation may be effective in some persons; a review of 60 studies found a response rate of 74 percent in bipolar depressives (Goodwin & Jamison 1990). Unfortunately, although dramatic antidepressant responses are sometimes observed after one night of total or partial sleep deprivation, this effect is usually lost after the next full night of sleep. In addition, hypomania and mania can be induced by this treatment. Bright light also may be effective in bipolar depression. Although more study is needed, a small trial found it to be effective in persons with bipolar II depression (Deltito et al. 1991). However, light therapy also should be used with caution, as induction of mania has been reported.

View Flow Chart for Module F.

Go to Module E.

Go to Module G.