SEARCH

SEARCH BY CITATION

Keywords:

  • bipolar disorder;
  • treatment-resistant;
  • bipolar depression;
  • evidence-based guidelines

Abstract

  1. Top of page
  2. Abstract
  3. Clinical recommendations
  4. Additional comments
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interest
  11. References

Objective:  To summarize the conceptual and operational definitions of treatment-resistant bipolar depression and to review the evidence-based therapeutic options.

Method:  Structured searches of PubMed, Index Medicus, Excerpta Medica and Psyclit conducted in December 2008.

Results:  Criteria for treatment resistance in bipolar depression are commonly based on concepts stemming from treatment resistance as defined for unipolar depression, an approach that proved to be inadequate. In fact, the addition of an ad hoc criterion based on lithium and other mood stabilizer unresponsiveness after reaching adequate plasma levels appears to be a patch that attempts to take into account the uniqueness of bipolar depression but fails to become operational. Recent data from randomized clinical trials of new anticonvulsants and second-generation antipsychotics should lead to the development of a modern definition of treatment-resistant bipolar depression, and specific therapeutic algorithms.

Conclusion:  We suggest a redefinition of resistant bipolar I and II depression. We propose different degrees of severity within bipolar depression in a stepwise manner.


Clinical recommendations

  1. Top of page
  2. Abstract
  3. Clinical recommendations
  4. Additional comments
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interest
  11. References
  • • 
    There is large evidence showing that bipolar depression is clinically, hence therapeutically, distinct from unipolar depression. Thus the concept of treatment resistance for bipolar depression should change according to this reality.
  • • 
    Lithium, often seen as the gold standard for the treatment of all phases of bipolar disorder, seems to be more useful against full-blown mania than against depression. To date, no single anticonvulsant has received the approval for the treatment of acute bipolar depression.
  • • 
    The Food and Drug Administration (FDA) has not approved any antidepressant for the treatment of bipolar depression. To date, the only antidepressant drug that showed evidence of efficacy in bipolar depression is fluoxetine but only when given combined with olanzapine, which allowed the FDA, but not the EMEA (European Medicines Agency), to approve a fixed dose olanzapine–fluoxetine combination for the treatment of acute bipolar I depression.
  • • 
    Quetiapine monotherapy is at present the only both FDA- and EMEA-approved treatment for bipolar depression.

Additional comments

  1. Top of page
  2. Abstract
  3. Clinical recommendations
  4. Additional comments
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interest
  11. References
  • • 
    Current criteria for defining resistant bipolar depression are inadequate and not in line with present-day evidence.
  • • 
    When a bipolar depressive episode occurs, in most cases clinicians prescribe any kind of drug from different classes, especially antidepressants, often comprising off-label medications, based not on official guidelines, but rather on the personal experience of each clinician.

Introduction

  1. Top of page
  2. Abstract
  3. Clinical recommendations
  4. Additional comments
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interest
  11. References

Although the most distinctive clinical feature of bipolar disorder is mania/hypomania, depression is clearly much more recurrent, predominant and burdensome (1–3). Bipolar patients spend much more time in depressive episodes, including subsyndromal depressive symptoms, than in hypomania, mixed episodes and mania (4–6).

Bipolar depression accounts for the largest part of morbidity and mortality of bipolar disorder (5). Most of the suicides are known to occur during this phase of illness; Colom et al. (1) found that lifetime history of attempted suicide and higher mean number of suicide attempts were strongly associated with depressive polarity.

Depressive symptoms and number of depressive episodes appear to be primarily responsible for reduced quality of life (7).

The presence of comorbidity contributes to the complex presentation of bipolar depression and may predict a poorer response to treatment. Particularly, anxiety comorbidity within a bipolar depression has been identified as one of several factors associated with a negative impact on treatment outcome (8).

Despite all these facts, the majority of treatment studies in bipolar disorder have highlighted the efficacy of medications for mania and hypomania rather than bipolar depression.

Drug treatment for acute bipolar depression is less than optimal (6, 9), considering, on the one hand, that antidepressants may induce manic switch and rapid cycling and, on the other hand, that most widely used antidepressants for bipolar depression lack any evidence supporting their efficacy in this group of patients, as most of them have only been validated and labeled for unipolar depression.

In summary, despite depression being the longest and most burdensome phase of bipolar disorder, data on its treatment are scanty. Hence, it is not surprising at all that, even if we keep in mind that treatment response to essayed treatment options is usually very poor, the term of resistant bipolar depression is neglected and lacks any kind of operational criteria.

Aims of the study

The main goal of this overview was to reconsider the concept of resistance in bipolar depression, differentiating it from concepts valid in the domain of unipolar depression which may not apply to bipolar depression. Furthermore, we attempted to differentiate as much as possible resistance in bipolar I vs. bipolar II depression and to provide a rationale for their optimal treatment. To this end, we reviewed the most recent data on treatment options for bipolar depression. We suggest new operational criteria for resistant bipolar depression in order to better meet patients’ needs. We focused on bipolar depression and not on other phases of bipolar disorder, although we consider that a switch to the opposite polarity or an unstable, cyclic course of the disorder may, in clinical practice, equal to resistance.

Material and methods

  1. Top of page
  2. Abstract
  3. Clinical recommendations
  4. Additional comments
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interest
  11. References

We critically analyzed published RCTs or open-label trials published between 1 January 1995 and 31 December 2008, focusing on the treatment of bipolar depression. In addition, we included some major meta-analyses, reviews and observational studies. To be considered, papers had to state clearly their methods and results, such as appropriate duration of the trial (at least 8 weeks), and adequate statistical methods for the sample sizes involved (we will comment on some papers that did not include sufficient sample size for definite conclusions to be drawn). Furthermore, the rating of depression should be either with the Hamilton Depression Rating Scale (Ham-D) or with the Montgomery–Åsberg Depression Rating Scale (MADRS).

We carried out a PubMed, Index Medicus, Excerpta Medica and Psyclit search using as keywords ‘resistance’ or ‘resistant’ or ‘refractory’ along with ‘bipolar disorder’ and ‘bipolar depression’ to identify and locate papers dealing with this subject. This yielded over 250 titles as of December 2008, of which 150 were highly pertinent. We used as terms all symptoms, ‘bipolar disorder’ and ‘depression’ or ‘depressive’ (over 8000 titles). Similarly, we used all possible antidepressant drugs and other treatments for bipolar depression and bipolar disorder in general, such as stabilizers and antipsychotics, using both general terms and the names of every individual drug (about 2000 papers).

Results

  1. Top of page
  2. Abstract
  3. Clinical recommendations
  4. Additional comments
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interest
  11. References

Evolution of the concept of treatment resistance: from unipolar to bipolar depression

Treatment resistant depression (TRD) is typically defined as the occurrence of an insufficient remission of depressive symptoms after adequate treatment. Although it is clear that the evidence supporting a definite description of treatment refractoriness is, at least, chequered even in unipolar depression, the concept of resistance in bipolar depression is even more far from a theoretical and operational definition. No systematic attempts have been made to define what constitutes refractoriness in bipolar depression, even if the lack of treatment response is a very common aspect during the depressive phase of bipolar disorder. In a study conducted by Post et al. (9), patients with bipolar depression were double-blind randomized to 10 weeks of therapy with one of three antidepressants (bupropion, sertraline and venlafaxine); responders were offered a double-blind continuation with the same treatment for 1 year. Amongst the patients who entered the continuation trial, 30% dropped out because of depression. Comparing MDD patients with depressed bipolar disorder patients, unresponsiveness to antidepressants was found to be 1.6 times as common in the bipolar group and loss of response during antidepressant treatment occurred 3.4 times as often (10).

Previous attempts to define treatment-resistant bipolar depression

The first operational definitions to conceptualize the meaning of refractory bipolar depression was proposed by Sachs (11), who defined as refractory bipolar depression a depression without remission despite two adequate trials of standard classes of antidepressant agents (at least 6 weeks each and adequately dosed), with or without augmentation.

The second definition was provided by Yatham et al. (12) who state refractory bipolar depression to be a ‘depression that failed to respond to a trial with lithium at serum levels of 0.8 mmol/l and above for 6 weeks’.

In a recent review, Gitlin (13) defined criteria for treatment-resistant bipolar disorder, and suggested to apply to bipolar depression the same criteria used for treatment-resistant unipolar depression, with the condition that failure to respond to mood stabilizers as well as antidepressants, should be added to the definition.

However, there is large evidence showing that bipolar depression is clinically, hence therapeutically, distinct from unipolar depression. Atypical depressive features of hypersomnia, hyperphagia, leaden paralysis and psychotic features are more frequent in bipolar depression. Regarding the course of illness, bipolar depressed patients show an earlier age of onset, shorter duration of episodes, higher number of past depressive episodes, and family history of bipolar disorder (14). Furthermore, patients with bipolar depression show higher rates of short-term non-response, of manic switch, of cycle acceleration and new rapid cycling, of tolerance to therapeutic effect of drugs and of depressive relapse after discontinuation, compared with patients with unipolar depression (10). Thus the concept of treatment resistance for bipolar depression should change according to this reality.

Evidence-based treatment options for bipolar depression

The role of lithium and other mood stabilizers.  Lithium, often seen as the gold standard for the treatment of all phases of bipolar disorder, seems to be more useful against full-blown mania than against depression (15–17), although its anti-suicidal action is out of doubt (18–20).

Lithium has been traditionally associated with rapid antidepressant effects, mostly in resistant depression studies, and is often used as augmentation to overcome resistance to antidepressant agents. Independent meta-analyses show a small effect in accelerating the response to antidepressants, and a solid effect in overcoming antidepressant resistance in augmentation studies (21). By contrast, no statistically significant differences in MADRS scores after 8 weeks between lithium and placebo were found at 0.6 mEq/l mean lithium serum levels in the acute phase of bipolar depression in the EMBOLDEN I study (22), which used a non-specified lithium formulation, still awaiting full publication and currently available only in abstract form.

To date, no single anticonvulsant has received the approval for the treatment of acute bipolar depression (23). Lamotrigine monotherapy showed some efficacy in the treatment of acute bipolar depression at the dose of 50–200 mg only on secondary outcome measures (24). Calabrese et al. (25) recently reported the results of five double-blind, placebo-controlled RCTs of acute bipolar depression I and II, combining the results of their own study (24) and those of four other previous unpublished trials. They concluded that lamotrigine and placebo as monotherapy (50–400 mg/die) did not statistically differ on primary efficacy endpoints.

There are two studies that support the efficacy of lamotrigine as monotherapy for bipolar I depression (26) and for bipolar II depression (27), but both are limited by a lack of placebo arm.

These data of lamotrigine have not been considered strong enough to proceed to its approval as a labeled drug in the treatment of bipolar depression during the acute phase. This, together with the slow titration required to avoid serious rash, seriously hinders the role of lamotrigine monotherapy in the treatment of bipolar depression.

Data on the use of lamotrigine as adjunctive therapy to a mood stabilizer in bipolar depression are more consistent. In the open-label STEP-BD study (National Institute of Mental Health’s Systematic Treatment Enhancement Program for Bipolar Disorder), lamotrigine, inositol or risperidone were used as adjuncts to a mood stabilizer for up to 16 weeks in patients with resistant bipolar I or II depression. Recovery rates were 23.8% with lamotrigine, compared with 17.4% with inositol and 4.6% with risperidone (28).

An 8-week, randomized, double-blind, placebo-controlled study showed lamotrigine to be effective as an add-on to lithium in patients with bipolar I and II depression who were non-responders or partial responders (29).

The efficacy of valproate monotherapy (serum valproic acid levels ranging 70–90 mcg/ml) in acute bipolar depression compared with placebo is supported by two small studies (30, 31). However, because of the limited sample size (25 and 18 patients respectively), there is a need for larger, randomized trials to confirm these preliminary results. Currently, valproic acid monotherapy is recommended as a second-line treatment option (32).

Antidepressants in bipolar depression: long and winding road.  Regarding the use of antidepressants in bipolar depression, there is a long-standing heretofore unresolved debate about their true effectiveness and safety as well as about the risk of manic switch and development of rapid cycling (33–42).

In the largest field study that tested the effectiveness of the addition of paroxetine or bupropion on standard care, the STEP-BD (36), it was suggested that standard antidepressant medications (paroxetine and bupropion) added to traditional mood stabilizers (lithium, valproate, carbamazepine or other FDA-approved antimanic agents), compared with the same mood stabilizers plus placebo, were associated with neither improved recovery from bipolar depression nor increased risk of switch to mania, hypomania or mixed episodes (the authors did not conclude that the lack of switch could extend to other antidepressants as well).

However, it would be dangerous to extend these data, obtained with heterogeneous studies and large patient numbers to individual patients, who may well have different sensitivity to the switching properties of antidepressant medications. Patients given antidepressants for bipolar depression should also receive a mood stabilizer and be closely monitored.

It is to be remarked that most RCTs evaluating the efficacy of antidepressants in acute bipolar depression were not conducted as individual drugs vs. placebo (two of 12 studies) but rather compared the add-on of antidepressants or a comparator or placebo with a pre-existing drug treatment with a stabilizer or antipsychotic, like lithium, olanzapine or valproate (10 of 12 studies); these studies taken together showed increased response and remission rates to antidepressants compared with placebo and no significant differences between the two treatment conditions in the emergence of manic switches (41).

The efficacy of 20 mg/day paroxetine monotherapy for 8 weeks has been compared with placebo in the EMBOLDEN II study of bipolar I and II depression; no significant differences were found as regards efficacy and switch risk (43).

Atypical antipsychotics: the unexpected winner?  The Food and Drug Administration (FDA) has not approved any antidepressant for the treatment of bipolar depression. To date, the only antidepressant drug that showed evidence of efficacy in bipolar depression is fluoxetine but only when given combined with olanzapine (44) which allowed the FDA, but not the European Medicines Agency (EMEA), to approve variable fixed dose olanzapine–fluoxetine combinations (OFCs) under the same brand for the treatment of acute bipolar I depression. Tohen et al. (45) studies showed that the OFC was significantly better than olanzapine or placebo in the treatment of bipolar I depression; olanzapine alone was better than placebo. However, it has been claimed that olanzapine alone was better than placebo not on the core depressive items of the MADRS (23).

In a 7-week acute phase of a randomized double-blind study, the OFC induced a small but statistically significant improvement in the scores of both mania and depression rating scales, and on the CGIs, compared with lamotrigine, but did not yield a higher proportion of responders (26).

Quetiapine monotherapy is at present the only both FDA- and EMEA-approved treatment for bipolar depression. The Bolder study Group I (46) and II (47) evaluated the efficacy, safety and tolerability of 600 and 300 mg/die of quetiapine vs. placebo in the treatment of bipolar I and II depression. Quetiapine at a dose of either 300 or 600 mg/die showed significantly greater improvement, as assessed in terms of the MADRS total scores, compared with placebo from the first evaluation (week 1) across all successive time points in the intent-to-treat group of patients with bipolar I and II depression (week 8). The percentage of patients meeting response (≥50% improvement on the MADRS) and remission (MADRS score ≤12) criteria was significantly higher in the quetiapine groups vs. placebo (58% vs. 36%, and 53% vs. 28% respectively). The incidence of treatment-emergent mania was lower in the quetiapine groups than in the placebo one.

In the EMBOLDEN I and EMBOLDEN II studies (22, 43), quetiapine (300–600 mg/die) showed efficacy in acute bipolar I and II depression compared with placebo, while the differences from placebo for lithium, at a mean serum level of 0.6 mEq/l (EMBOLDEN I), and for paroxetine, at 20 mg/die (EMBOLDEN II), did not reach statistical significance at the 8-week endpoint.

Taken together, these data may further confound, rather than clarify, the ideas of clinicians wishing to adopt some clear-cut guideline to treat their patients while they are in the depressive phase of bipolar disorder.

If everything fails.  To date, relatively few studies have examined the next step treatment strategies for resistant bipolar depression. No clear guidelines exist concerning what to do when the first approved therapies fail. Some evidence has been obtained for adjunctive modafinil (200 mg/die), a non-addictive stimulant agent used for narcolepsy (48); 100–200 mg/day modafinil added to ongoing mood stabilizer to 41 patients for 6 weeks showed about double the efficacy of placebo to 44 patients in inducing response and remission; both response and remission did not exceed half of the patients in the modafinil arm. Adjunctive pramipexole, a dopamine agonist (up to 5 mg/die, mean 1.7 mg/die) (49), added to ongoing stabilizer treatment for 6 weeks to 12 patients with non-psychotic bipolar depression vs. 10 to placebo, induced response in 67% of patients randomized to pramipexole vs. 20% of patients randomized to placebo. A small, open, randomized, non-responder cross-over trial vs. lamotrigine provided additional support for the adjunctive use of the MAOI, tranylcypromine, for the treatment of refractory bipolar depression (50). In this study, patients were taken from the Stanley Foundation Bipolar Network and were openly assigned to receive, in addition to their ongoing mood stabilizer treatment, 20 initial mg/die tranylcypromine, which could be titrated to a maximum of 100 mg/die at weeks 5–10, or 25 mg/die lamotrigine, that could be titrated to a maximum of 400 mg/day at weeks 6–10. Response varied according to criterion used from half patients receiving add-on tranylcypromine to over 60%, the double of figures obtained with lamotrigine; people who did not respond to a drug addition were switched to the other and obtained response in all cases with tranylcypromine and in half cases with lamotrigine. However, the results of these studies were limited by small sample size (only 19 patients) and its open nature.

Hence, treatment options for patients with resistant bipolar depression may include:

  • i)
     optimization of the dosage of the same drug,
  • ii)
     combination, that is adding a new active treatment to the existing one if the latter fails to improve depressive symptoms, and
  • iii)
     substitution, that is, introducing a new drug to replace the old one (12).

This is an empirical guideline, but no unequivocal algorithm is provided (51, 52). In patients with refractory and severe bipolar I and bipolar II depression, the use of electroconvulsive therapy (ECT) should be considered, as this was shown to be one of the most effective treatments (53, 54). ECT sessions are provided every other day, for a maximum of three sessions per week and not more than 9 or 10 sessions for every cycle.

The available data on vagus nerve stimulation (VNS) for the treatment of resistant bipolar depression showed some efficacy in reducing depressive symptoms in the short and long term mainly in open studies. On the contrary, evidences from the only double-blind study are inconclusive, suggesting that further clinical trials are needed to confirm its efficacy (55).

Discussion

  1. Top of page
  2. Abstract
  3. Clinical recommendations
  4. Additional comments
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interest
  11. References

The concept of resistance, even in unipolar depression, has been the subject of intense debate in terms of definitions, including those of response and remission, but also in terms of minimum treatment duration, of what should be considered an adequate treatment dosage, of treatment adherence and of formulating a correct diagnosis (56). It is not astonishing that even stronger a debate might arise when talking about resistance in bipolar depression, which is per se a refractory condition with poor response to treatment (9).

To date, it seems that the available definitions of resistance in bipolar depression do not take into consideration the substantial differences between the two conditions, in terms of dimensional, phenomenological and clinical features. This resulted in transferring data and experiences obtained from research in resistant unipolar depression directly to the treatment of bipolar patients, although it would be more appropriate to pursue the opposite, given that many patients with ‘unipolar depression’ convert to bipolar during the course of their illness, and given that recurrent unipolar major depression appears to be biologically more related to bipolar, rather than unipolar depression (57).

Furthermore, it is not possible to split bipolar depression from bipolar disorder; hence, the focus should be more on the term ‘bipolar’ than on ‘depression’.

While there are no ‘evidence-based’ data to support that the depressive phase of bipolar disorder is a different clinical entity from unipolar depression, it is interesting to remark that the only two drugs officially approved for bipolar depression by the FDA are an atypical antipsychotic and an antipsychotic in combination with an SSRI antidepressant. This gives no room to two ‘adequate trials of standard classes of antidepressant agents’ (11). Currently, there is a discrepancy among FDA recommendations, data emerging from randomized clinical trials (RCTs), and what occurs in clinical practice. In Europe, the only EMEA-approved drug in bipolar depression is quetiapine. The approval is so recent that no European guideline has still incorporated this issue, although the ECNP consensus provided some support to quetiapine as first choice (58).

The result is that, when a bipolar depressive episode occurs, in most cases clinicians prescribe any kind of drug from different classes, often comprising off-label medications, based not on official guidelines, but rather on the personal experience of each clinician (59).

Stating that the use of antidepressants in bipolar depression is always malpractice is unfair; however, at the same time, there are several cases in which their use might worsen the course of the illness right away or improve it in the short run, but negatively affect its overall outcome (60). Another reflection should be reserved to the use of mood stabilizers because if it is true that they are effective for treating acute mania and protect from relapse/recurrence, data supporting their efficacy in the treatment of the acute phase of bipolar depression are less than consistent. Despite this, mood stabilizers are currently prescribed as first-line treatment and one of the suggested criteria of resistant bipolar depression is ‘the failure to respond to a trial of lithium with serum levels of 0.8 mEq/l and above for 8 weeks’ (12). It is obvious that the use of mood stabilizers during the acute phase of bipolar depression is meant not only to counter depression but also to create a protective chemical milieu for an effective antidepressant (lamotrigine, antidepressants and atypical antipsychotics?) to act and protect from future mood oscillations and switches.

Suggested new criteria

Current criteria for defining resistant bipolar depression are inadequate and are not in line with current evidence. Excessively simplistic algorithms, like the ones employed for unipolar depression, should be avoided, given the complexity of the clinical picture of bipolar depression, which may be considered like comorbidity with substance use disorder, at higher risk of drug resistance.

To propose new operational criteria for resistance in bipolar depression, we should first define the remission criteria for this condition. For bipolar depression, remission is traditionally defined as a score of ≤7 on the 17-item version of the Ham-D (59) or ≤12 on the Montgomery–Åsberg Depression Rating Scale (47), i.e. the same criteria applied to unipolar depression. Considering residual mood symptoms in bipolar depression are associated with a higher risk of relapse and poorer functional outcomes, there is a need to better define full remission of an acute depressive episode, as this is a more desirable treatment target.

Berk et al. (61) suggested that a cut-off score of <5 on the MADRS corresponds to a CGI-BP of 1 (normal, not at all ill) and it may approximate a definition of complete remission.

However, in the context of clinical trials, it is difficult to apply this lower cut-off as a measure of full remission at end-point, even if these criteria for remission are more likely to be clinically meaningful. The reason is that this goal would be difficult to reach and would lower the proportion of remitters, making any treatment to appear quite unattractive, despite positive response, and raising serious issues about statistical power. Moreover, if trials had to be powered for full remission, the number of patients who would be exposed to placebo would increase enormously, carrying relevant ethical concerns (62).

The primary goal in treating bipolar depression should be the remission of depressive symptoms with a return to normal levels of psychosocial functioning. On the other hand, as it is not possible to separate bipolar depression from bipolar disorder, the secondary, but not less important targets in starting treatment, should be to prevent subsequent episodes of both illness polarities and avoid switching to an episode of the opposite polarity or to mood cycling. Unfortunately, most currently available treatment options do not entirely meet these criteria.

All the above prompt us to take into account the concept of predominant polarity within bipolar disorder (1) when planning treatment for bipolar depression. This concept might need first to define operational criteria for treatment resistance, that distinguish between bipolar I and bipolar II depression (the latter having a well-recognized depressive predominant polarity). For example, although treatment with antidepressants is recommended in bipolar I depression only for short-term periods and in association with a mood stabilizer, due to the risk of switch, in bipolar II depression the use of antidepressants is more relaxed, as mood elevations in bipolar II disorder were found to be less severe than in bipolar I disorder, not progressing beyond hypomania in most cases (63).

If a depressive episode occurs within treated bipolar disorder it means that the current treatment is no longer sufficient. In our opinion, this is the real first step of resistance. During an acute depressive episode, the recommended approach is to increase the dosage to the maximum tolerated level, in order to maximize benefit from a single medication and to reduce the risk of side effects due to drug combinations. However, optimum serum levels are not established for most available treatments for bipolar disorder; so, it is difficult to target treatment to other than clinical measures. Furthermore, in most cases, multiple drug prescription is the rule in the depressive phase of bipolar disorder. Thus, this recommendation is not easy to apply to routine clinical practice. Moreover, it is not always true that a combination therapy is more unsafe than high dosages of a single drug (64); in some instances, adding a drug may render possible to reduce the dosage of another drug, lowering both its therapeutic threshold (65) and its side effects (66). Furthermore, combination therapies often allow to broaden the scope of treatment, including comorbidities and refractoriness (67).

Provided that diagnosis has been carefully posed, we may suggest a definition for bipolar I and II depression. We also propose different degrees of severity within bipolar depression in a stepwise manner, based on the failure to respond to a defined therapeutic algorithm. We apply an 8-week period as the time needed for resistance to be declared. It should be made clear that a previously effective treatment for bipolar disorder should not be suspended when specifically dealing with the treatment of depression. An ongoing antibipolar (stabilizing) treatment must be confirmed, even if it has obtained partial results; what is seen in Tables 1 and 2 and Fig. 1 should be intended as an additional strategy to address acutely emerging bipolar depression only.

Table 1.   Algorithm to overcome treatment-resistance in bipolar I depression
Quetiapine (≥600 mg) OR Lithium (0.8 mEq/l) (OR other mood stabilizing ongoing regimen) + lamotrigine (50–200 mg)Step 1 8 weeks CGI-BP < 2; HAM-D < 7
Failure in step 1: resistant bipolar I depression 
Olanzapine + fluoxetine (OFC) (6–12 + 25–50 mg) OR Quetiapine (>600 mg) + lamotrigine (50–200 mg)Step 2
Failure in step 1 + step 2: refractory bipolar I depression 
Adjust OFC dose OR Lithium (0.8 mEq/l) + SSRI (20–40 mg)/bupropion (150–300 mg) OR Quetiapine (>600 mg) + SSRI (20–40 mg)/bupropion (150–300 mg) OR Novel agents as adjunts (modafinil: 100–200 mg or pramipexole: mean 1.7 mg/die)Step 3
Failure in step 1 + step 2 + step 3: intractable bipolar I depression 
Electroconvulsive therapy (one full cycle)Step 4
Failure in step 1 + step 2 + step 3 + step 4: involutional bipolar I depression 
Table 2.   Algorithm to overcome treatment-resistance in bipolar II depression
Quetiapine (300–600 mg) OR Lithium (0.8 mEq/l) (OR other mood stabilizing ongoing regimen) + lamotrigine (50–200 mg)Step 1 8 weeks CGI-BP < 2; HAM-D < 7
Failure in step 1: resistant bipolar II depression 
Quetiapine (300–600 mg) + SSRI (20–40 mg)/bupropion(150–300 mg) OR Lithium (0.8 mEq/l) + SSRI (20–40 mg)/bupropion (150–300 mg)Step 2
Failure in step 1 + step 2: refractory bipolar II depression 
Quetiapine (300–600 mg) + MAOIs (tranylcypromine: 20–100 mg/day) OR Lithium (0.8 mEq/l) + MAOIs (tranylcypromine: 20–100 mg/day) OR Novel agents as adjunts (modafinil: 100–200 mg or pramipexole: mean 1.7 mg/die)Step 3
Failure in step 1 + step 2 + step 3: intractable bipolar II depression 
Electroconvulsive therapy (one full cycle)Step 4
Failure in step 1 + step 2 + step 3 + step 4: involutional bipolar II depression 
image

Figure 1.  Algorithms for overcoming treatment-resistant bipolar depression I (left) and II (right).

Download figure to PowerPoint

Bipolar I Depression.  We define as resistant bipolar I depression a depressive episode within bipolar disorder that fails to reach remission with adequately dosed lithium (0.8 mEq/l in the plasma) or to other adequate ongoing mood-stabilizing treatment, plus lamotrigine (50–200 mg/day) or with full dose (≥600 mg/day) of quetiapine as monotherapy.

Step 1. The definition of resistant bipolar I depression also corresponds to the first step of our therapeutic algorithm (Table 1).

Step 2. Failure to reach remission with combinations of fluoxetine plus olanzapine at variable doses or full-dose quetiapine plus lamotrigine.

We propose the definition of refractory bipolar I depression for depression that persists after STEP 1 plus STEP 2 (Table 1).

Step 3. Resistance persists even when dose adjustments take place or antidepressants from other classes (such as SSRIs or bupropion) or novel agents (like modafinil or pramipexole) are added.

We propose the definition of intractable bipolar I depression for depression that persists after STEP 1 plus STEP 2 plus STEP 3 (Table 1).

Step 4. Resistance and unresponsiveness persist after ECT is given a trial.

We propose the definition of involutional bipolar I depression for depression that persists after STEP 1 plus STEP 2 plus STEP 3 plus STEP 4 (Table 1).

Bipolar II depression.  We define as resistant bipolar II depression a depressive episode within bipolar disorder that fails to reach remission with adequately dosed lithium (0.8 mEq/l in plasma) or to other adequate ongoing mood-stabilizing treatment, plus lamotrigine (50–200 mg/day) or with a dose range of 300–600 mg/day quetiapine as monotherapy.

The definition of resistant bipolar II depression also corresponds to the first step of our specific therapeutic algorithm (Table 2).

Step 2. Failure to reach remission with lithium or with quetiapine and an antidepressant (SSRIs or bupropion added).

We propose the definition of refractory bipolar II depression for depression that persists after STEP 1 plus STEP 2 (Table 2).

Step 3. Resistance persists even when the add-on is switched to a MAOI antidepressant or when novel agents (modafinil or pramipexole) are added.

We propose the definition of intractable bipolar II depression for depression that persists after STEP 1 plus STEP 2 plus STEP 3 (Table 2).

Step 4. Resistance and unresponsiveness persist after ECT is given a trial.

We propose the definition of involutional bipolar II depression for depression that persists after STEP 1 plus STEP 2 plus STEP 3 plus STEP 4 (Table 2). For each step, the term ‘or’ means that, provided that the various options showed in similar therapeutic efficacy in various controlled studies, the choice between different strategies will mainly depend on the different clinical and individual characteristics of patients and on how they tolerate the specific drug or treatment involved.

For bipolar II depression, we consider first a treatment with medium doses of quetiapine (300–600 mg/day) or with a stabilizer combination that comprises lithium and a drug with known antidepressant properties, such as lamotrigine. For lamotrigine, the range of dosage is 50–200 mg/day, but the higher dose does not need to be reached. The dose should be maintained once a satisfactory response has been obtained. Anyway, the target dose in bipolar disorder is 200 mg/day (100 mg in conjunction with valproate and 400 mg/day in combination with carbamazepine or other enzyme-inducing drugs). Titration timing (not taking into account enzyme-inducing drugs or valproate) is 25 mg/day at weeks 1 and 2; 50 mg at weeks 3 and 4; 100 mg/day at week 5 and 200 mg/day at week 6.

If remission criteria, set at CGI-BP less than 2 and HAM-D less than 7, are not met within 8 weeks, an add-on standard dose of an SSRI (range 20–40 mg/day) or the noradrenaline/dopamine reuptake blocker, bupropion (150–300 mg/day; titration timing 1 week), that despite its dopaminergic activity showed no abuse potential and no manic switch in large samples (36), should be given. If the above therapeutic goals are still not met, one could try a MAO inhibitor, like tranylcypromine (20–100 mg/day; titration timing 5 weeks) or the stimulant modafinil (100–200 mg/day) and the dopamine D3/D2 enhancer, pramipexole (up to 5 mg/die, mean 1.7 mg/die) as adjuncts. A final step would comprise the use of ECT.

For bipolar I depression, criteria and scores would be similar (Table 1), but quetiapine doses should be higher than for bipolar depression II. This is recommended to obtain more mood stability, which is needed to avoid switches to the mixed state. The mixed state involves more risk of suicide in bipolar I disorder (68). An OFC could be involved in steps 2 and 3, with dosage adjustments made possible by the multiplicity of available formulations. We do not provide differential indications for antidepressant drug dosages between bipolar I and bipolar II depression, but we recommend that lower doses should be used in bipolar I depression, as the risk of manic switch is probably higher in the latter, compared with both bipolar II and unipolar major depression (63). For the same reason, we recommend that the lower fluoxetine doses should be preferred in using the OFC; dose adjustments within the same combination should also take into account this fact. Moreover, always for the same reason, we excluded from our algorithm the administration of MAOIs, due to increased problems with treatment adherence (including dietary restrictions) with these patients, despite no evidence for manic/hypomanic switch for tranylcypromine (50). We decided to hold the third step recommendation for activating drugs like modafinil and pramipexole, as there are no reports of manic switch for the latter, while we recommend some caution for the former because, despite that it was shown not to induce mood swings in a large population (69), a case of manic switch has been recently reported (70).

For the purpose of our current proposal of resistant bipolar depression we accept as adequate an interval of 8 weeks of trial for each step we propose; the outcome to be met in each step is currently accepted definitions of remission [i.e. maximum Ham-D score of 7 (61) or MADRS 12 (47)]. The number of treating episodes needed to consider a patient as ‘involutional’, i.e. highly drug resistant, is according to Tables 1 and 2 one each step. We did not re-propose the term drug-resistant after ‘at least two trials with adequate doses of antidepressants belonging to two or three different classes, etc.’ as this would take 8–12 months to pass the first step alone. In conclusion, we propose a stepwise algorithm to overcome treatment resistance in bipolar depression (Fig. 1).

We will not provide details for treatment with antidepressants, but we recommend that manufacturers’ standard instructions are followed. For patients with psychosis and for those with BP-I, we would recommend to avoid reaching the highest doses with any antidepressant.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Clinical recommendations
  4. Additional comments
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interest
  11. References

The authors thank the support and funding of the Spanish Ministry of Health, Instituto de Salud Carlos III, CIBERSAM. Dr Colom is funded by the Spanish Ministry of Science and Innovation, Instituto Carlos III, through a ‘Miguel Servet’ postdoctoral contract (CP08/00140).

Declaration of interest

  1. Top of page
  2. Abstract
  3. Clinical recommendations
  4. Additional comments
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interest
  11. References

Prof. Eduard Vieta has received grant/research support from Almirall, Astra-Zeneca, Bristol-Myers Squibb, Eli Lilly, the European 7th Framework Program, GlaxoSmithKline, Janssen-Cilag, Novartis, Organon, Otsuka, Pfizer, Sanofi-Aventis, Seny Foundation, Servier, the Spanish Ministry of Health (CIBERSAM), the Spanish Ministry of Science and Education and the Stanley Medical Research Institute; has been a member of the speakers boards for Almirall, Astra-Zeneca, Bristol-Myers Squibb, Eli Lilly, Esteve, GlaxoSmithKline, Janssen, Lundbeck, Novartis, Organon, Otsuka, Pfizer, Sanofi-Aventis and Servier; and has served as a consultant for AstraZeneca, Bristol-Myers Squibb, Eli Lilly, Forest Research Institute, GlaxoSmithKline, Janssen, Jazz, Lundbeck, Novartis, Organon, Otsuka, Pfizer, Sanofi-Aventis, Servier and UBC. Dr Colom has been a member of the speakers’ boards for Bristol-Myers-Squibb, Otsuka, Eli-Lilly, Astra Zeneca, GlaxoSmithKline, Pfizer, Sanofi-Aventis and Tecnifar; and has served as a consultant for Shire Ltd and Astra Zeneca.

References

  1. Top of page
  2. Abstract
  3. Clinical recommendations
  4. Additional comments
  5. Introduction
  6. Material and methods
  7. Results
  8. Discussion
  9. Acknowledgements
  10. Declaration of interest
  11. References