Bipolar depression continues to be one of the most challenging, least studied, and yet relatively common problems in clinical psychopharmacology.1 In this article, we summarize the flowcharts and recommendations for the treatment of bipolar depression from the most recent version of the Algorithm for the Pharmacotherapy of Depression, which is an Internet-based computerized expert system developed as part of the Harvard Psychopharmacology Algorithm Project. The initial Internet version was completed in 1996, but it evolved from written precursors published in 19882 and 1993.3 The Internet version is continually updated and revised based on new data. In 1998, a summary of the Internet algorithms for the treatment of non-bipolar forms of depression was published.4 Readers are encouraged to experience "virtual psychopharmacology consultation" for patients with depression by trying the full program. This may be found at the Internet address www.mhc.com/expert.html.
To prepare the current revision of the algorithms for the treatment of bipolar depression, the authors resurveyed the literature and evaluated other sources of expert opinion such as published algorithms,5"10 expert consensus practice guidelines,71112 and a variety of Internet-based resources.9,10 The goal was to create an up-to-date, evidence-based treatment sequence that could serve as a guide for practitioners. However, because of the limitations of current data and the wide variability among individual patients, this is at best an educated guess as to what constitutes optimal practice. The key to improving this algorithm and mamtaining it as a high-quality tool for practitioners will be alterations in it based in part on user feedback. The introduction to this issue of Psychiatric Annals describes the potential for collecting patient outcome data from algorithm users over the Internet or an intranet.
As with the other algorithms in this issue, the flow of logic as one follows the steps in the algorithm may be compared with the series of questions that a psychopharmacology consultant might ask a consultee about a patient under discussion. It attempts to reproduce the thinking process that a psychopharmacologist uses to make a decision based on many complex considerations. It tries to address some, but by no means all, of the nuances that might affect the decision of the prescriber in each individual case.
INITIAL EVALUATION STEPS
As shown in Figure 1, the initial step in evaluating any depressed patient is to screen for medical causes of depression, and, if present, to institute appropriate medical management of treatable conditions. The patient's current medications and known comorbid medical conditions are also reviewed, both for any possible causative relationship to the patient's current symptoms and for any potential for interactions with proposed treatments.4 Female patients of childbearing age should be screened for pregnancy because that could alter the sequence of choices in ways not addressed in the current algorithms.1314 Actively substance-abusing patients should have completed withdrawal from their substances (and from any medications required for that withdrawal) for at least a week. This appears to be the minimal time by which drug-placebo differences may become apparent in the treatment of depression persisting after abstinence from alcohol.15 Although there may be occasions when it may be reasonably safe and appropriate to use medications for patients who are actively abusing substances, algorithms for that situation are not being proposed here.
The psychiatrist now diagnoses the depressive symptoms presented by the patient using the DSM-IV criteria,16 noting any Axis I comorbidity. With each sequential trial of treatment in this algorithm that fails to produce the desired result, these diagnoses, including the medical differential, should be reviewed and reestablished.
Figure 1 shows the flowchart for the initial identification of depressive subtype. If the diagnosis falls into one of the non-bipolar categories, appropriate questions and recommendations may be found in our other publications4 and in the Internet version of the Algorithm for the Pharmacotherapy of Depression.
If the patient has depression associated with bipolar disorder, this algorithm applies best to the treatment of depressive episodes meeting DSMIV criteria for major depression. If the patient has a subsyndromal or partially remitted depression, the evidence basis for selecting treatment is much less clear. The algorithm's standard recommendations are generally applicable, but psychotherapeutic interventions may be useful as alternative or augmenting treatments.11,17
Urgen! Indication for Electroconvulsive Therapy?
If the patient has any of the variants of bipolar depression, the algorithm next asks whether the use of electroconvulsive therapy (ECT) as the initial treatment is appropriate. ECT is the most effective treatment known for depressive episodes associated with bipolar disorder.1819 Therefore, the initial decision to be made is whether there is an urgent indication for ECT. Examples of urgent indications include (but are not limited to) severe states of suicidality, inanition, catatonia, psychosis, or incapacitating depression in the context of certain medical conditions, such as pregnancy.20
Are Psychotic Symptoms Present?
If psychotic symptoms are present, and ECT was refused, unsuccessful, or not found to be urgently indicated as the initial treatment option, then an antipsychotic should be given in addition to the other medications (mood stabilizers and antidepressants) that are proposed below. The evidence is strong that psychotic depression in unipolar patients responds best to treatment with a combination of an antipsychotic and an antidepressant.4,21 These studies included a small number of bipolar patients, but too few to enable a separate outcome analysis for them. Some older studies have suggested that bipolar patients whose conditions were maintained with typical neuroleptics have an increased risk of cycling into depression.22 However, a more recent study using depot neuroleptics found that bipolar psychotic depression relapses were prevented by addition of the neuroleptic.23
A major problem in using typical neuroleptics is the increased vulnerability of bipolar patients to extrapyramidal side effects and tardive dyskinesia.22 One study (N = 131) showed a 35% prevalence of tardive dyskinesia in bipolar patients treated with typical neuroleptics, compared with a total absence of tardive dyskinesia (0% prevalence) in the group not treated with them.24 The rate of tardive dyskinesia in vulnerable populations associated with the use of the newer antipsychotics appears to be significantly lower than that with typical neuroleptics.25
Therefore, our first-line recommendation in choosing an antipsychotic to treat psychotic bipolar depression is to employ, when possible, one of the newer generation of antipsychotics: olanzapine, quetiapine, or risperidone.26 There is also consistent evidence that the newer agents have antidepressant activity that is superior to that of typical neuroleptics in patients with schizophrenia and schizoaffective disorder.27'29 The data specific to patients with bipolar psychotic depression are sparse, but two reports on olanzapine involving a. total of 18 patients found that 14 had responded well.30,31 All but 2 were taking an antidepressant also. However, the new antipsychotics may occasionally precipitate agitation or mania (perhaps due to these same antidepressant properties).32"34 This may necessitate substitution of a different antipsychotic or use of a lower dose. If a parenteral or depot antipsychotic is needed due to patient agitation or noncompliance, a typical neuroleptic must be used, as currently none of the newer generation of antipsychotics is available for parenteral or depot use.
Clozapine is sometimes effective for the treatment-refractory case of bipolar psychotic depression.35
BIPOLAR I; BIPOLAR II; RAPID CYCLING?
After evaluating whether concurrent treatment with an antipsychotic agent is necessary, the next step is to determine the patient's bipolar disorder subtype, as subsequent treatment steps will vary depending on this subtype. The patient may have a depression associated with non-rapid-cyáing bipolar I disorder (Fig. 2) or non-rapid-cycling bipolar ? disorder (Fig. 3), or the patient's depression may occur in the context of bipolar I or ? disorder with rapid cycling (four or more episodes of mood disturbance per year; Fig. 4). An algorithm is not provided for patients who meet criteria for bipolar I disorder, mixed episode.
Bipolar I Disorder, Currently Depressed. No History of Rapid Cycling
In the algorithm for this subgroup, initial attempts are made to use the potential antidepressant efficacy of lithium in bipolar depression.12 Lithium has demonstrated antidepressant efficacy in bipolar depression/8 but other older mood stabilizers do not seem to have an equivalent antidepressant effect. Carbamazepine may possess some weak antidepressant activity, but less than lithium.36 To date, valproic acid has not demonstrated a clear-cut acute antidepressant effect,37,38 although a controlled study in progress should provide more information. A large placebo-controlled study comparing valproate sodium with lithium for 1 year of maintenance treatment was recently completed and showed a very low rate of depressive relapse requiring hospitalization with either of the active treatments.39 However, the lithium group had the highest rate of suicidal ideation, which raises questions as to the validity of this study. Previous studies have consistently shown that lithium treatment reduces suicidal activity.40
Lamotrigine is an anticonvulsant that may have particular efficacy in bipolar depression. A recent double-blind, placebo-controlled, prospective study of lamotrigine in the treatment of major depressive episodes associated with bipolar I disorder, which included 195 patients, found that 51% of the patients treated with 200 mg/d of lamotrigine showed significant clinical improvement, 41% improved with 50 mg, and 26% improved with placebo.41 Lamotrigine may become a first-line option in the treatment of bipolar depression if further studies continue to demonstrate its utility and safety. However, reports of switches to mania indicate it may sometimes behave moré like an antidepressant.41,42 Its efficacy in acute mania and for prevention of recurrences of bipolar mood episodes requires more study. A disadvantage of lamotrigine is the occurrence of potentially serious dermatologie reactions, and there have been rashassociated deaths. According to the package insert, rashes occur in 10% of patients, and the appearance of rash necessitates discontinuation of the medication. The dosage must be increased slowly to minimize the occurrence of these dangerous allergic reactions. Another disadvantage is that improvement in depression may occur rather slowly: 4 to 6 weeks were required in the study cited above. However, lithium may require a similar period to achieve its acute antidepressant efficacy.11
Other anticonvulsants may have some utility in treating bipolar depression. Gabapentin appears to have some weak antidepressant efficacy in bipolar depression.43,44 The roles of other newer mood-stabilizing agents such as topiramate45 and tiagabine46 remain to be defined.
Current Mood Stabilizer Medication? The clinical approach to the patient with bipolar I depression depends on his or her medication regimen at the time of presentation.
If the patient is taking lithium, whether as the sole mood-stabilizing agent or in combination with others, such as valproic acid, carbamazepine, or both, the first step is to check the lithium level. If the lithium level is less than 0.8 mEq/L, it is recommended that the lithium dosage be raised until the level exceeds 0.8 mEq/L, with a suggested range of 0.8 to 1.2, to optimize the potential for an acute antidepressant response.47 If the lithium level is already more than 0.8 or the patient remains depressed after the lithium level is maintained at more than 0.8 for several weeks, or if the patient is unable to tolerate an increase in lithium dosage, an antidepressant should be initiated. This recommendation is based in part on a recent study that showed that in these clinical circumstances, initiation of antidepressant treatment is better tolerated than adding a second mood stabilizer, although efficacy was comparable.48
If the patient is not taking lithium, but is currently taking valproic acid, carbamazepine, or another mood stabilizer, the dosage levels should be optimized to minimize the possibility of cycling in preparation for the addition of an antidepressant to the treatment regimen. The valproic acid level should be adjusted to the therapeutic range of 50 to 125 ng/mL,49 and optimally, if tolerated, to the upper half of that range; the carbamazepine level should be in the therapeutic range of 4 to 12 ng/mL. Concurrently, an antidepressant should be added.
If the patient is not taking a mood stabilizer and his or her depression is currently mild, a trial of lithium monotherapy is the first-line recommendation,19 if the patient agrees to defer the potential for a more rapid or robust effect from starting an antidepressant in exchange for the advantages of monotherapy with a mood stabilizer. If the patient's depression is more severe, the first-line recommendation is to begin both a mood stabilizer (eg, lithium or valproic acid) and an antidepressant.12
Antidepressant Selection and Sequence. All patients with bipolar I disorder should be taking a mood stabilizer prior to or concurrently with the initiation of antidepressant therapy to minimize the risk of antidepressant-induced mania or rapid cycling.50
The first-line antidepressants for bipolar depression are the selective serotonin reuptake inhibitors (SSRIs) and bupropion hydrochloride, due to their lower risk of inducing antidepressant-associated mania12'51 and their relatively favorable side effect profiles. All of the SSRIs, including fluoxetine hydrochloride, sertraline hydrochloride, paroxetine hydrochloride, Citalopram, and fluvoxamine maleate, are effective treatments for bipolar depression, although the only placebo-controlled study was with fluoxetine hydrochloride.52 For patients with histories of antidepressant-induced mania or in whom a later trial of a monoamine oxidase inhibitor (MAOI) can be anticipated, fluoxetine hydrochloride's long half-Life (7 to 10 days for its metabolite norfluoxetine) is a relative disadvantage.
Bupropion hydrochloride, in its immediaterelease form, has a higher risk for inducing seizures (as high as 0.44% at doses up to 450 mg/d53) than do the SSRIs (seizure risk is roughly 0.1% to 0.2% according to the package inserts), but the seizure risk of the sustained-release form is lower (0.1% at doses of 300 mg/d or below, rising to 0.4% at a dose of 400 mg/d according to the package insert), making it the preferred form for general use.54 Patients with bipolar disorder who are taking anticonvulsants are possibly at lower risk, but the use of bupropion hydrochloride is relatively contraindicated in patients with comorbid eating disorders such as bulimia or anorexia nervosa, or with a history of seizures.
If the initial trial of the SSRI or bupropion hydrochloride added to the mood stabilizer is unsuccessful, then the second antidepressant trial would be with whichever class of drug was not used first.
The third antidepressant recommended in the algorithm sequence in Figure 2 is venlafaxine hydrochloride. Recent uncontrolled published evidence55 and expert opinion37 are converging to suggest that venlafaxine hydrochloride may be particularly efficacious in bipolar depression with a low incidence of inducing mania.
For a fourth trial, an MAOI should be considered because MAOIs have demonstrated efficacy in bipolar depression, especially tranylcypromine sulfate for anergic bipolar depression.56 There are some practical difficulties that argue against this, however. If the patient was taking bupropion hydrochloride, paroxetine hydrochloride, sertraline hydrochloride, fluvoxamine maleate, Citalopram, or venlaiaxine hydrochloride, a washout period of 2 weeks should occur before an MAOI can be started, and if the SSRI used was fluoxetine hydrochloride, the washout period will need to be at least 5 weeks. The other difficulty is that a depressed bipolar patient might not maintain the level of dietary compliance required for the safe administration of an irreversible, nonselective MAOI. For this reason, some experts still favor the use of tricyclic antidepressants57 in combination with a mood stabilizer at this point, due to their proven efficacy in the treatment of acute bipolar depression. However, the use of tricyclic antidepressants (TCAs) is associated with a high rate of manic switch, even when mood stabilizers are given,58 so TCAs are avoided in this algorithm and other options are favored. These are described below in the section on treatmentrefractory bipolar depression (Table).
Finally, it should be noted that ECT can be reconsidered as an alternative to any of the four antidepressant pharmacotherapy trials. This highly effective treatment for bipolar depression appears to be underused by clinicians.1
Bipolar II Disorder, Currently Depressed. No History of Rapid Cycling
There are relatively few studies specifically addressing the treatment of bipolar ? patients, but an important difference in the treatment response of bipolar I and II depression is that antidepressant-associated mood elevation appears to be less problematic in bipolar II.59 When it does occur, the severity of the mood elevation in bipolar II disorder tends to be milder.60,61 Therefore, some experts feel that with cautious monitoring, some patients with bipolar ? depression can be treated with antidepressants alone.62 Fluoxetine hydrochloride,63,64 venlaiaxine hydrochloride,55 and tranylcypromine sulfate63 have been used successfully as monotherapy for bipolar II depression.
The bipolar ? algorithm attempts to identify individuals for whom antidepressant monotherapy might be a reasonably low-risk option. Bipolar II patients with a history of rapid cycling are not included: they are addressed in a separate algorithm (Fig. 4). Rapid cycling is found more commonly in patients with bipolar II disorder than in those with bipolar I disorder.60 Akiskal and Mallya have also suggested that bipolar ? patients with a hypomanic or "hyperthymic" temperament may develop very treatment-resistant mixed states if given antidepressants alone.66 They, too, should probably be treated with the rapid-cycling algorithm (Fig. 4).3 The phenomenology and treatment of bipolar ? disorder has been reviewed at greater length elsewhere.67
Figure 3 details the algorithm for the treatment of bipolar II depression in patients who are not rapid cyclers. The first question asks whether the patient is currently taking a mood stabilizer or is appropriate for one because of a history of antidepressant-induced hypomania or mania or significant impairment during prior hypomanic episodes, or if there is any other reason why the prescriber would prefer that the patient be treated with a mood stabilizer. If any of these exceptions apply, the patient would be treated in the same way as non-rapid-cycling bipolar I depression. The small remaining group of patients may be candidates for monotherapy with an antidepressant.
The potential risks of antidepressant-associated hypomania or mania or rapid mood cycling should be explained to the patient.68,69 The risk of hypomanic or manic switches may be 2% to 4% during the first year of treatment.55,63,64 The potential benefits of mood stabilizers in decreasing this risk should also be explained, as well as their disadvantages, which include the possible loss of periods of mild hypomania associated with high productivity, creativity, and quality of life and other potential side effects (eg, weight gain).
If the patient prefers to take a mood stabilizer, the treatment sequence would be the same as in the algorithm for non-rapid-cycling bipolar I depression (Fig. 2). Although the recommended mood stabilizer would be either lithium or valproic acid, some clinicians might consider a trial of gabapentin at this point in view of its benign side effect profile, although its efficacy is less well established.44
If the patient elects not to begin a mood stabilizer, and the prescriber agrees that antidepressant monotherapy is a reasonable option, the subsequent sequence of antidepressant recommendations is similar to the bipolar I algorithm. SSRIs and bupropion hydrochloride are selected first, then venlafaxine hydrochloride. Some experts might consider lithium augmentation next. However, the response to lithium augmentation in bipolar depression does not seem particularly robust.70 Therefore, again, we prefer considering a trial of franylcypromine sulfate, which has been used safely and effectively as monotherapy in bipolar ? patients with anergic depressions.65 Because of the greater mania-inducing effects of MAOIs when compared with newer agents,50 the addition of a mood stabilizer concurrently with or prior to beginning therapy with MAOIs might be considered; however, there is some evidence to suggest that when MAOIinduced mood elevations occur, they tend to be relatively mild.71
If the patient's depressive symptoms persist, further options for refractory bipolar depression are detailed in the table and are discussed later.
Bipolar I or II With History of Rapid Cycling
The algorithm pathway for bipolar patients, currently depressed, who have a history of rapid cycling is shown in Figure 4. These patients are at particularly high risk for mood destabilization from treatment with antidepressants, even if currently taking mood stabilizers. The algorithm therefore emphasizes use of more than one mood stabilizer treatment before cautiously introducing antidepressants,72 which can be an effective treatment for many patients.48
As noted, mixed states are not addressed in this algorithm for bipolar depression. Such patients meet full DSM-IV criteria for mania, as well as major depression, and are conceptualized as having a subtype of mania. They are covered in algorithms for the treatment of mania. Most experts consider monotherapy with divalproex sodium the first-line pharmacotherapy, with secondary evidence for carbamazepine. Lithium appears to be much less effective.73
The initial step in patients with rapid cycling is to consider possible predisposing factors. These should be treated, if appropriate. Hypothyroidism should be ruled out, including subclinical forms.74 Undiagnosed complex partial seizures,75 multiple sclerosis,76 and other neurologic conditions77 have been associated with rapid cycling. Covert substance abuse could be a factor, and concomitant antidepressants and other medications (eg, stimulants) should be assessed for their possible role in the cycling78: tricylics are clearly the greatest offenders, but SSRIs and others could also be contributing. Noncompliance should also be considered as a factor: patients who abruptly discontinue lithium therapy often bring on rapid relapses,79 and rapid cycling may develop if they do this repeatedly.80
Once predisposing factors have been dealt with, the initial treatment depends on whether the patient is currently taking a mood stabilizer. If this depressed patient is not taking a mood stabilizer, our first-line recommendation is to begin valproic acid.73 There is evidence from the pre-valproate literature that lithium may be effective,81 but most experts now consider valproic acid to be the first-line recommendation here.12 Side effect considerations and patient preference may determine which drug is selected. Carbamazepine does not seem particularly effective in rapidcycling bipolar depression.82
If the mood stabilizer chosen is ineffective after a several-week trial, or if the patient was already taking a mood stabilizer, the first-line recommendation is to add a second mood stabilizer. The choice will usually be valproate sodium or lithium, depending on what was chosen for first-line treatment. Some clinicians are considering gabapentin at this point, again because of its more benign side effect profile, although it is much less established as an effective treatment. When it does work, the benefits appear to be modest.44 However, in mild to moderately ill patients, this alternative may have a risk-benefit ratio favorable to its consideration, and it is being increasingly used by clinicians. If it is ineffective, the standard second-step recommendation should be employed. Lamotrigine has been reported to be useful in rapid-cycling patients with depression83; however, given its greater side effect risks (rashes), consideration would usually be postponed to a later point.
Figure 3. Flowchart of the algorithm for treatment of patients with non-rapid-cycling bipolar Il depression. SSRI = selective serotonin reuptake inhibitor; MAOI = monoamine oxidase inhibitor.
If the patient remains depressed after two mood stabilizers in combination, the treatments to consider next depend on whether this rapidcycling patient has bipolar I or bipolar ? disorder. The addition of a third mood stabilizer is an option common to both subtypes,8 but patients with bipolar I disorder, especially those with a clear history of antidepressant-induced cycling, could be candidates for ECT or the addition of one of the newer antipsychotic agents (Fig. ?).84 If none of these considerations are preferred, a cautious trial of an antidepressant is the next recommendation. The first-line preference is for an SSRI or bupropion hydrochloride. Due to its long halflife, ñuoxetine hydrochloride might be best avoided in patients with rapid cycling.
In treating bipolar depression, some have advocated that the antidepressant selected should be started at a low dose, and raised more slowly than usual.12,85 This would seem an especially reasonable approach in the patient with rapid-cycling bipolar disorder.
If the initial antidepressant selected fails to ameliorate the depression, the algorithm proposes that the physician continue to follow the same succession of antidepressants proposed for bipolar I and ?, non-rapid-cycling depressions, but with the precautions indicated above (ie, low dose, slow titration). Venlaiaxine hydrochloride and perhaps tranylcypromine sulfate might be tried as third- or fourth-line options. ECT may be reconsidered at any point.
If the patient does respond to one of these antidepressants without cycling, it is advised that the antidepressant be discontinued as soon as possible,12 unless the history indicates that the patient is able to tolerate maintenance antidepressant therapy without cycling.
If all these efforts are unsuccessful, the patient is considered to have a refractory bipolar depression.
REFRACTORY BIPOLAR DEPRESSION
The table lists some treatment considerations in patients with refractory bipolar depression. Possible covert compliance problems and substance abuse need to be explored and addressed if present. The diagnosis should be reviewed and possible comorbid medical conditions should again be excluded, such as thyroid dysfunction, underlying neurologic problems, and drug interactions. Comorbid Axis I and Axis ? disorders that may be contributing to the refractoriness of the depression should be diagnosed and appropriate treatment initiated and optimized. The computer version of this algorithm links to pathways to address these comorbidities, and text versions have been published recently elsewhere.4
If the patient's depression remains refractory, there are other potential treatment options to be considered. The examples given below are not listed in any order of preference. The supporting evidence is limited for many of these treatment modalities when applied to patients with bipolar depression. Individual patient considerations will determine their appropriateness.
A trial of ECT should certainly be a prime consideration if it has not previously been tried. It may be followed by maintenance ECT, which has also been shown to be a safe and cost-effective treatment alternative for some highly refractory cases of bipolar depression.86 Antidepressant augmentation strategies include addition of thyroid hormone, using either levothyroxine sodium (T4) or liothyronine sodium (T3),87'88 and addition of stimulants.1,89 Supraphysiologic thyroid treatment (ie, to produce T4 levels 150% of the upper limit of normal) has been successful in treating some cases of highly refractory bipolar depression.90"92 Other monotherapy antidepressant options include nefazodone hydrochloride and mirtazapine. This may be the time to consider the newer mood stabilizers, including lamotrigine, gabapentin, topiramate, or tiagabine. Different combinations of mood stabilizers72,93 could also be considered. The calcium channel blockers, in particular nimodipine, have occasionally been effective for the depressed phase of bipolar patients.94,95 Some patients, especially those with a seasonal component to their depression, have benefited from bright light therapy, and others have reportedly responded to repeated cycles of total sleep deprivation.96
Antipsychotic agents may be useful treatments for some refractory patients, even if they are not psychotic.84'97'98 Patients with psychotic bipolar depression may benefit from a switch to a different antipsychotic agent. As noted earlier, the new-generation antipsychotics are probably better for treating bipolar depression than the older neuroleptics. The atypical antipsychotic clozapine was reportedly effective in approximately 50% of psychotic bipolar depressed patients with unsatisfactory response to other antipsychotics.35
Treatment of Refractory Bipolar Depression
ALGORITHMS FOR THE PHARMACOTHERAPY OF ACUTE DEPRESSION IN PATIENTS WITH RIPOLAR DISORDER: WHAT'S NEXT?
Treatment of bipolar depression remains a challenging problem. Although a minority of patients can be treated successfully using monotherapy with lithium or an antidepressant, rational copharmacy is the rule rather than the exception.59'100 This article outlines one approach to the serial consideration of treatment options. However, this is a work in progress. We welcome suggestions about how this algorithm could be improved. Feedback can be given electronically through our web program at the address www.mhc.com/expert.html by clicking on "Send us a message."
1. Zarate CA Jr, Tohen M, Baraibar G, Kando JC, Mirin J. Prescribing trends of antidepressants in bipolar depression. / Clin Psychiatry. 1995;56:260-264.
2. Osser DN. Treatment resistant problems. In: Tupin JP, Shader PJ, Harnett DS, eds. Clinical Handbook of Psychopharmacology, 2nd ed. New York: Jason Aronson; 1988:269-328.
3. Osser DN. A systematic approach to the classification and pharmacotherapy of nonpsychotic major depression and dysthymia. / Clin Psychopharmacol. 1993;13:133-144.
4. Osser DN, Patterson RD. Algorithms for the pharmacotherapy of depression: part one and part two. Directions in Psychiatry. 1998;18:303-336.
5. Gilbert DA, Altshuler KZ, Rago WV, et al. Texas Medication Algorithm Project: definitions, rationale, and methods to develop medication algorithms. / Clin Psychiatry. 1998;59:345-351.
6. Suppes X Calabrese JR, Mitchell PB, Pazzaglia PJ, Potter WZ, Zarin DA. Algorithms for the treatment of bipolar manic-depressive illness. Psychopharmacol Bull. 1995; 31:469-474.
7. Bauer MS, Callahan AM, Jampala C, et al. Clinical Practice Guidelines for Bipolar Disorder from the Department of Veterans Affairs. J Clin Psychiatry. 1999;60:9-21.
8. Calabrese JR, Woyshvüle MJ. A medication algorithm for treatment of bipolar rapid cycling? j Clin Psychiatry. 1995;56:11-18.
9. Suppes T, Brown S, Habermacher E. Bipolar Disorders Module Guideline Procedures Manual. TMAP Procedural Manual. Austin, TX: Texas Medication Algorithm Project; 1998.
10. Kusumaker V, Yatham L, Parikh S, Matte R. Bipolar Disorder: A Summary of Clinical Issues and Treatment Options. Vancouver, Canada: Bipolar Disorder Sub-Committee, Canadian Network for Mood and Anxiety Treatments (CANMAT); 1997. Available Internet: www. mentalhealth.com / rx2 / bp-canl .html.
11. American Psychiatric Association. Practice guideline for the treatment of patients with bipolar disorder. Am J Psychiatry. 1994;151:1-36.
12. Frances AJ, Kahn DA, Carpenter D, Docherty JP, Donovan SL. The Expert Consensus Guidelines for treating depression in bipolar disorder. / Clin Psychiatry. 1998; 59(suppl 4):73-79.
13. Leibenluft E. Issues in the treatment of women with bipolar illness. / Clin Psychiatry. 1997;58(suppl 15):5-11.
14. Viguera AC, Cohen LS. The course and management of bipolar disorder during pregnancy. Psychopharmacol Bull. 1998;34:339-346.
15. Mason BJ, Kocsis JH, Ritvo EC, Cutler RB. A double-blind, placebo-controlled trial of desipramine for primary alcohol dependence, stratified on the presence or absence of major depression. JAMA. 1996;275:761-767.
16. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC: American Psychiatric Association; 1994.
17. Coursey RD, Keller AB, Farrell EW. Individual psychotherapy and persons with serious mental illness: the clients' perspective. Schizophr Bull. 1995;21:283-301.
18. Zornberg GL, Pope HG Jr. Treatment of depression in bipolar disorder: new directions for research. J Clin Psychopharmacol. 1993;13:397-408.
19. Goodwin FK, Jamison KR. Manir Depressive Illness. New York: Oxford University Press; 1990.
20. Coffey CE. The pre-ECT evaluation. Psychiatric Annals. 1998;29:506-508.
21. Schatzberg AF, Rothschild AJ. Psychotic (delusional) major depression: should it be included as a distinct syndrome in DSM-IV? Am J Psychiatry. 1992;149:733-745.
22. Tohen M Zarate CA Jr. Antipsychotic agents and bipolar disorder. J Clin Psychiatry. 1998;59:38-48.
23. Litrlejohn R, Leslie F, Cookson J. Depot antipsychotics in the prophylaxis of bipolar affective disorder. Br J Psychiatry. 1994;165:827-829.
24. Mukherjee S, Rosen AM, Caracd G, Shukla S. Persistent tardive dyskinesia in bipolar patients. Arch Gen Psychiatry. 1986;43:342-346.
25. Brecher M, Jeste D, Okamoto A, Napolitano J, Kane JM. Low frequency of tardive dyskinesia in elderly patients with dementia. American College of Neuropsychopharmacology 37th Annual Meeting; Las Croabas, Puerto Rico; 1998.
26. Frye MA, Ketter TA, Altshuler LL, et al. Clozapine in bipolar disorder: treatment implications for other atypical antipsychotics. / Affect Disord. 1998;48:91-104.
27. Marder SR, Davis JM, Chouinard G. The effects of risperidone on the five dimensions of schizophrenia derived by factor analysis: combined results of the North American trials. J Clin Psychiatry. 1997;58:538-546.
28. Cantillon M, Goldstein JM. Efficacy of quetiapine in affective symptoms of schizophrenia (NR 445). American Psychiatric Association Annual Meeting; Toronto, Ontario, Canada; 1998.
29. Tollefson GD, Sanger TM, Thieme ME. Depressive signs and symptoms in schizophrenia: a prospective blinded trial of olanzapine and haloperidol. Arch Gen Psychiatry. 1998;55:250-258.
30. Rothschild AJ, Bates KS, Boehringer KL, Syed A. Olanzapine response in psychotic depression. J Clin Psychiatry. 1999;60:116-118.
31. Zarate CA Jr, Narendran R, Tohen M, et al. Clinical predictors of acute response with olanzapine in psychotic mood disorders. J Clin Psychiatry. 1998;59:24-28.
32. Lindenmayer JP, Klebanov R. Olanzapine-induced maniclike syndrome. J Clin Psychiatry. 1998;59:318-319. Letter.
33. London JA. Mania associated with olanzapine. J Am Acad Child Adolesc Psychiatry. 1998;37:135-136. Letter.
34. Barkin JS, Pais VM, Gaffney MF. Induction of mania by risperidone in patients resistant to mood stabilizers. / CHn Psychopharmacol. 1997;17:57-58.
35. Zarate CA Jr, Tohen M, Baldessarini RJ. Clozapine in severe mood disorders. J Clin Psychiatry. 1995;56:411-417.
36. Post RM, Uhde TW, Roy-Byrne PP. Antidepressant effects of carbamazepine. Am J Psychiatry. 1986;143:29-34.
37. Potter WZ. Bipolar depression: specific treatments. J Clin Psychiatry. 1998;59(suppl 18):30-36.
38. Sachs G. Is divalproex sodium an appropriate initial treatment for bipolar depression? Psychiatric Annals. 1996;26:S454-S459.
39. Bowden CL, Calabrese JR, Nemeroff CB, Lydiard RB, McElroy SL. Effects on suicidality of divalproex, lithium, and placebo in randomized one-year maintenance treatment of bipolar disorder (Poster 24). New Clinical Drug Evaluation Unit Conference; Boca Raton, Florida; 1998.
40. Baldessarini RJ, Tondo L, Hennen J. Effecte of lithium treatment and its discontinuation on suicidal behavior in bipolar manic-depressive disorders. J Clin Psychiatry. 1999;60(suppl 2):77-84.
41. Calabrese JR, Bowden CL, Sachs GS, Ascher JA, Monaghan E, Rudd GD. A double-blind placebo-controlled study of lamotrigine monotherapy in outpatients with bipolar I depression. J Clin Psychiatry. 1999;60:79-88.
42. Sporn J, Sachs G. The anticonvulsant lamotrigine in treatment-resistant manic-depressive illness. J Clin Psychopharmacol. 1997;17:185-189.
43. Ferrier IN. Lamotrigine and gabapentin. Neuropsychobiology. 1998;38:192-197.
44. Ghaemi SN, Katzow JJ, Desai SP, Goodwin FK. Gabapentin treatment of mood disorders: a preliminary study. J Clin Psychiatry. 1998;59:426-429.
45. Marcotte D. Use of topiramate, a new anti-epileptic as a mood stabilizer. J Affect Disord. 1998;50:245-251.
46. Kaufman KR. Adjunctive tiagabine treatment of psychiatric disorders: three cases. Ann Clin Psychiatry. 1998;10:181-184.
47. Gelenberg A, Kane J, Keller M, et al. Comparison of standard and low serum levels of lithium for maintenance treatment of bipolar disorder. N Engl J Med. 1989; 321:1489-1493.
48. Joffe RT, Young LT, Robb JC, MacQueen GM, Marriott M. A randomized controlled trial of the acute treatment of bipolar depression (Abstract 54). American College of Neuropsychopharmacology 37th Annual Meeting; Las Croabas, Puerto Rico; 1998.
49. Schaff MR, Fawcett J, Zajecka JM. Divalproex sodium in the treatment of refractory affective disorders. J Clin Psychiatry. 1993;54:380-384.
50. Boerlin HL, Gitlin MJ, Zoellner LA, Hammen CL. Bipolar depression and antidepressant-induced mania: a naturalistic study. J Clin Psychiatry. 1998;59:374-379.
51. Peet M. Induction of mania with selective serotonin reuptake inhibitors and tricyclic antidepressants. Br J Psychiatry. 1994;164:549-550.
52. Cohn JB, Collins G, Ashbrook E, Wernicke JF. A comparison of fluoxetine, Imipramine, and placebo in patients with bipolar depressive disorder, int Clin Psychopharmacol. 1989;4:313-322.
53. Rosenstein DL, Nelson JC, Jacobs SC. Seizures associated with antidepressants: a review. J Clin Psychiatry. 1994;54:289-299.
54. Dunner DL, Zisook S, Billow AA, Batey SR, Johnston JA, Ascher JA. A prospective safety surveillance study for bupropion sustained-release in the treatment of depression. J Clin Psychiatry. 1998;59:366-373.
55. Amsterdam J. Efficacy and safety of venlaiaxine in the treatment of bipolar ? major depressive episode. / Clin Psychopharmacol. 1998;18:414-417.
56. Thase ME, Mallinger AG, McKnight D, Himmelhoch JM. Treatment of irriprarrune-resistant recurrent depression: IV. A double-blind crossover study of tranylcypromine for anergic bipolar depression. Am J Psychiatry. 1992;149: 195-198.
57. Yatham LN, Kusumakar V, Parikh SV, et al. Bipolar depression: treatment options. Can J Psychiatry. 1997; 42(suppl 2):87S-91S.
58. Bottlender R, Rudolf D, Strauss A, Moller HJ. Antidepressant-associated maniform states in acute treatment of patients with bipolar-I depression. Eur Arch Psychiatry Clin Neurosa. 1998;248:296-300.
59. Kupfer DJ, Carpenter LL, Frank E. Possible role of antidepressants in precipitating mania and hypomania in recurrent depression. Am J Psychiatry. 1988;145:804-808.
60. Vieta E, Gasto C, Otero A, Nieto E, Vallejo J. Differential features between bipolar I and bipolar ? disorder. Compr Psychiatry. 1997;38:98-101.
61. Coryell W, Andreasen NC, Endicott J, Keller M. The significance of past mania or hypomania in die course and outcome of major depression. Am J Psychiatry. 1987;144:309-315.
62. Frances A, Docherty JP, Kahn DA. The expert consensus guidelines for the treatment of bipolar disorder. J Clin Psychiatry. 1996;57(suppl 12A):20.
63. Amsterdam JD, Garcia-Espana F> Fawcett J, et al. Efficacy and safety of fluoxetine in treating bipolar II major depressive episode. J Clin Psychopharmacol. 1998;18: 435-440.
64. Simpson SG, DePaulo JR. Fluoxetine treatment of bipolar H depression. J Clin Psychopharmacol. 1991;11:52-54.
65. Himmelhoch JM, Thase ME, Mallinger AG, Houck P. Tranylcypromine versus Imipramine in anergic bipolar depression. Am J Psychiatry. 1991;148:910-916.
66. Akiskal HS, Mallya G. Criteria for the "soft" bipolar spectrum: treatment implications. Psychopharmacol Bull. 1987;23:68-73.
67. Fawcett J. Bipolar II disorder. Psychiatric Annals. 1996;26:S440-S443.
68. Altshuler LL, Post RM, Leverich GS, Mikaiauskas K, Rosoff A, Ackerman L. Antidepressant-induced mania and cycle acceleration: a controversy revisited. Am J Psychiatry. 1995;152:1130-1138.
69. Howland RH. Induction of mania with serotonin reuptake inhibitors. J Clin Psychopharmacol. 1996;16:425-427.
70. Price LH, Charney DS, Henninger GR. Variability of response to lithium augmentation in refractory depression. Am J Psychiatry. 1986;143:1387-1392.
71. Stall AL, Mayer PV, Kolbrener M, et al. Antidepressantassodated mania: a controlled comparison with spontaneous mania. Am J Psychiatry. 1994;151:1642-1645.
72. Freeman MP, Stoll AL. Mood stabilizer combinations: a review of safety and efficacy. Am J Psychiatry. 1998;155: 12-21.
73. Bowden CL, Brugger AM, Swann AC, et al. Efficacy of divalproex vs lithium, placebo in the treatment of mania. JAMA. 1994;271:918-924.
74. Chang KD, Keck PE Jr, Stanton SP, McElroy SL, Strakowski SM, Geracioti TD Jr. Differences in thyroid function between bipolar manic and mixed states. Biol Psychiatry. 1998;43:730-733.
75. Brewerton TD. The phenomenology of psychosis associated with complex partial seizure disorder. Ann Clin Psychiatry. 1997;9:31-51.
76. Casanova MF, Kruesi M, Mannheim G. Multiple sclerosis and bipolar disorder: a case report with autopsy findings. J Neuropsychiatry Clin Neurosa. 1996;8:206-208.
77. Jan JE, Abroms IF, Freeman RD, Brown GM, Espezel H, Connolly MB. Rapid cycling in severely multidisabled children: a form of bipolar affective disorder? Pediatric Neurol. 1994;10:34-39.
78. Simpson HB, Hurowitz GI, Liebowitz MR. General principles in the pharmacotherapy of antidepressant-induced rapid cycling: a case series. J Clin Psychopharmacol. 1997;17:460-466.
79. Baldessarini RJ, Tondo L, Floris G, Rudas N. Reduced morbidity after gradual discontinuation of lithium treatment for bipolar I and ? disorders: a replication study. Am J Psychiatry. 1997;154:551-553.
80. Goodwin GM. Recurrence of mania after lithium withdrawal: implications for the use of lithium in the treatment of bipolar affective disorder. Br J Psychiatry. 1994;164:149-152.
81. Wehr TA, Sack DA, Rosenthal NE, Cowdry RW. Rapidcycling affective disorder: contributing factors and treatment responses in 51 patients. Am J Psychiatry. 1988;145:179-184.
82. Okuma T. Effects of carbamazepine and lithium on affective disorders. Neuropsychobiology. 1993;27:138-145.
83. Fatemi SH, Rapport DJ, Calabrese JR, Thuxas P. Lamotrigine in rapid-cycling bipolar disorder. J Clin Psychiatry. 1997;58:522-527.
84. Vieta E, Gasto C, Colom F, Martinez A, Otero A, Vallejo J. Treatment of refractory rapid cycling bipolar disorder with risperidone. J Clin Psychopharmacol. 1998;18:172-174. Letter.
85. Dantzler A, Salzman C. Treatment of bipolar depression. Psychiatr Sew. 1995;46:229-230.
86. Bonds C, Frye MA, Coudreaut MF, et al. Cost reduction with maintenance ECT in refractory bipolar disorder. J ECT. 1998;14:36-41.
87. Prange AJ Jr. Novel uses of thyroid hormones in patients with affective disorders. Thyroid. 1996;6:537-543.
88. Joffe RT. The use of thyroid supplements to augment antidepressant medication. J Clin Psychiatry. 1998;59:26-29.
89. Fawcett J, Kravitz HM, Zajecka JM, Schaff MR. CNS stimulant potentiation of monoamine oxidase inhibitors in treatment-refractory depression. J Clin Psychopharmacol. 1991;11:127-132.
90. Baumgartner A, Bauer M, Hellweg R. Treatment of intractable non-rapid cycling bipolar affective disorder with high-dose thyroxine: an open clinical trial. Neuropsychopharmacology. 1994;10:183-189.
91. Bauer M, Hellweg R, Graf KJ, Baumgartner A. Treatment of refractory depression with high-dose thyroxine. Neu ropsychopharmacology. 1 998; 18:444-455.
92. Bauer MS, Whybrow PC. Rapid cycling bipolar affective disorder: 11. Treatment of refractory rapid cycling with high-dose levothyroxine: a preliminary study. Arch Gen Psychiatry. 1990;47:435-440.
93. Post RM, Leverich GS, Denicoff KD, Frye MA, Kimbrell TA, Dunn R. Alternative approaches to refractory depression in bipolar illness. Depress Anxiety. 1997;5:175-189.
94. Pazzaglia PJ, Post RM, Ketter TA, George MS, Marangell LB. Preliminary controlled trial of nimodipine in ultrarapid cycling affective dysregulation. Psychiatry Res. 1993;49:257-272.
95. Pazzaglia PJ, Post RM, Ketter TA, et al, Nimodipine monotherapy and carbamazepine augmentation in patients with refractory recurrent affective illness. J Clin Psychopharmacol. 1998;18:404-413.
96. Barbini B, Colombo C, Benedetti F, Campori E, Bellodi L, Smeraldi E. The unipolar-bipolar dichotomy and the response to sleep deprivation. Psychiatry Res. 1998;79: 43-50.
97. Shelton R, Tollefson G, Tohen M, et al. The study of olanzapine plus fluoxetine in treatment-resistant major depressive disorder without psychotic features (Poster 90). New Clinical Drug Evaluation Unit Conference; Boca Raton, Florida; 1998.
98. Hendrick V, Altshuler LL, Szuba MP. Is there a role for neuroleptics in bipolar depression? J Clin Psychiatry. 1994;55:533-535.
99. Post RM, Ketter TA, Pazzaglia PJ, et al. Rational polypharmacy in the bipolar affective disorders. Epilepsy Res Suppl. 1996;11:153-180.
100. Janicak P. Copharmacy strategies for bipolar disorder. Psychiatric Annals. 1998;28:357-363.
Treatment of Refractory Bipolar Depression