2012 John Wiley and Sons A/S Published by Blackwell Publishing Ltd.

Bipolar Disorders 2013: 15: 61–69

BIPOLAR DISORDERS

Review Article

Evidence-based treatment strategies for treatment-resistant bipolar depression: a systematic review Sienaert P, Lambrichts L, Dols A, De Fruyt J. Evidence-based treatment strategies for treatment-resistant bipolar depression: a systematic review. Bipolar Disord 2013: 15: 61–69.  2012 John Wiley & Sons A ⁄ S. Published by Blackwell Publishing Ltd. Objectives: Treatment resistance in bipolar depression is a common clinical problem that constitutes a major challenge for the treating clinician as there is a paucity of treatment options. The objective of this paper was to review the evidence for treatment options in treatmentresistant bipolar depression, as found in randomized controlled trials and with special attention to the definition and assessment of treatment resistance. Methods: A Medline search (from database inception to May 2012) was performed using the search terms treatment resistance or treatment refractory, and bipolar depression or bipolar disorder, supplemented with 43 separate searches using the various pharmacologic agents or technical interventions as search terms. Results: Only seven studies met our inclusion criteria. These studies examined the effects of ketamine (n = 1), (ar)modafinil (n = 2), pramipexole (n = 1), lamotrigine (n = 1), inositol (n = 1), risperidone (n = 1), and electroconvulsive therapy (ECT) (n = 2). Conclusions: The available level I evidence for treatment strategies in resistant bipolar depression is extremely scarce, and although the response rates reported are reassuring, most of the strategies remain experimental. There is an urgent need for further study in homogeneous patient samples using a clear concept of treatment resistance.

Bipolar disorder is a highly prevalent disorder, with poor symptomatic and psychosocial outcome (1). Research interest in bipolar disorder has focused on the treatment of mania, with little consideration given to the treatment of depression (2, 3). However, depressive symptomatology predominates the clinical picture (4, 5), is associated with significantly greater psychosocial impairment, and constitutes a great disease burden to patients and their families. Impairment increases significantly with each increment in symptom severity; depressive symptoms being at least as disabling as manic symptoms at corresponding severity levels (6). In spite of the high prevalence and the important burden of depressive morbid-

Pascal Sienaerta,b, Lore Lambrichtsa, Annemiek Dolsc and Ju¨rgen De Fruytd a

Department of Mood Disorders, bECT Department, University Psychiatric Center, Catholic University Leuven, Campus Kortenberg, Kortenberg, Belgium, cDepartment of Elderly Psychiatry, GGZ inGeest ⁄ VUmc, Amsterdam, The Netherlands, d Department of Psychiatry, General Hospital Sint-Jan Brugge-Oostende AV, Brugge, Belgium doi: 10.1111/bdi.12026 Key words: armodafinil – bipolar disorder – electroconvulsive therapy – inositol – ketamine – lamotrigine – modafinil – pramipexole – risperidone – treatment resistance Received 22 August 2011, revised and accepted for publication 27 August 2012 Corresponding author: Pascal Sienaert, M.D., Ph.D. University Psychiatric Center Catholic University Leuven, Campus Kortenberg Leuvensesteenweg 517 Kortenberg 3070 Belgium Fax: 0032-2-7595380 E-mail: [email protected]

ity, treatment options for bipolar depression are limited (3). Moreover, non-response in bipolar depression is highly prevalent, and occurs in 40% of patients after eight weeks of treatment with quetiapine (7). Other first-line treatments, such as lithium, lamotrigine, olanzapine, or olanzapinefluoxetine combination, have an even less favourable outcome and are less evidence based (7–10). The addition of antidepressants to an ongoing treatment with mood stabilizers will be helpful in only a quarter of patients with bipolar depression (11). Non-response to antidepressants is so commonly encountered that it has become a criterion for the diagnosis of bipolar spectrum disorder (12, 13).

61

Sienaert et al.

In spite of the high prevalence of treatmentresistant bipolar depression, research activity is scarce (14) and hampered by unclear and changing definitions. Sachs (15) defined treatment-resistant bipolar depression as non-remission despite two adequate trials of standard antidepressant agents (six weeks each), with or without augmentation strategies; whereas Yatham and colleagues (9) defined treatment-resistant bipolar depression as non-response to a six-week trial with lithium at serum levels of ‡0.8 mmol ⁄ L. Gitlin (16) proposed using the criteria for treatment-resistant unipolar depression, i.e., non-response to two antidepressants from different classes (six weeks each), with the proviso that failure to respond to mood stabilizers as well as antidepressants should be added to the definition. Recently, Pacchiarotti and colleagues (17) proposed elaborated definitions for treatment-resistant bipolar depression, adding a staging method, based on various degrees of severity. The authors defined resistant, refractory, intractable, and involutional bipolar depression, according to non-response to specific drugs and doses, based on current treatment algorithms. Lipsman and colleagues (18) suggested defining treatment-resistant bipolar depression as Ôa nonresponse to adequate trials of monotherapy with lithium or lamotrigine, as well as lithium or lamotrigine in combination with at least one anticonvulsant or antipsychoticÕ. Finally, Malhi et al. (19) defined treatment-resistant bipolar depression as non-remission despite at least three adequate trials of first-line medications, such as a mood stabilizer. In recent guidelines, various second- and thirdline strategies for the management of bipolar depression are advocated (20). A sound evidence base for most of these strategies is, however, lacking. The objective of this paper was to review the evidence for treatment options in treatment-resistant bipolar depression, as found in randomized controlled trials and with special attention to the definition and assessment of treatment resistance. Materials and methods

We searched Medline databases (January 1966 to May 2012) using the terms treatment resistant or treatment refractory and bipolar disorder or bipolar depression in the search strategy. The strategy was supplemented with 43 separate searches, including: various pharmacological agents or treatment options [electroconvulsive therapy (ECT), vagal nerve stimulation, light therapy, transcranial magnetic stimulation, and deep brain stimulation] and bipolar but limited to Ôrandomized controlled trialsÕ. In addition, we reviewed the citations of retrieved

62

papers and key reviews. Titles and abstracts (or the full copy, if necessary) of the retrieved papers were read and screened to meet the following inclusion criteria: (i) randomized, controlled trial comparing a pharmacological or technical intervention with another pharmacological or technical intervention or placebo; (ii) a study population consisting of patients with bipolar I or bipolar II depression; (iii) a sample size of at least 10 patients; (iv) resistance to at least two treatment steps, of which at least one was a mood stabilizer; (v) the intervention studied was used either as monotherapy or as addon to an ongoing treatment; (vi) a standardized outcome measure was reported; and (vii) if both patients with unipolar and bipolar depression were included, the results for the patients with bipolar depression were analysed and presented separately. Searches and selection were performed independently by two co-authors (PS and LL). Any disagreement was discussed and a final decision was made by consensus (all authors). Results

The data search by key words yielded 404 potentially relevant articles. Four papers met our inclusion criteria: Goldberg et al. (21), Nierenberg et al. (22), Diazgranados et al. (23), and Zarate et al. (24). Additional searches yielded three papers meeting our inclusion criteria: Frye et al. (25), Sienaert et al. (26), and Bailine et al. (27). Characteristics of these studies: (i) design, (ii) sociodemographic and clinical characteristics, and (iii) definition of treatment resistance and concurrent treatment, are summarized in Tables 1, 2, and 3, respectively. Goldberg and colleagues (21) studied the effect of pramipexole or placebo added to an ongoing treatment with mood stabilizers. Response was evident in 67% of patients taking pramipexole versus 20% of patients taking placebo (p = 0.04). One patient in the pramipexole group developed mania, and none in the placebo group. Mean Young Mania Rating Scale (YMRS) scores at the end of the study did not differ significantly between the pramipexole group [mean = 4.4, standard deviation (SD) = 4.6] and the placebo group (mean = 2.0, SD = 2.2) (p = 0.12). In a study by Nierenberg et al. (22), patients were randomly assigned to receive open-label adjunctive lamotrigine, inositol, or risperidone (equipoise randomization: lamotrigine versus risperidone, lamotrigine versus inositol, and risperidone versus inositol). No significant betweengroup differences in recovery rate were seen when any pair of treatments were compared. However, numerical differences were found when compari-

Treatment-resistant bipolar depression Table 1. Design characteristics of randomized controlled trials in treatment-resistant bipolar depression

Study

Design

Goldberg et al. (2004) (21)

RDB

Nierenberg et al. (2006) (22)

Equipoise randomization, open-label

Diazgranados et al. (2010) (23)

RDB, cross-over

Frye et al. (2007) (25)

RDB

Sienaert et al. (2009) (26)

RDB

Bailine et al. (2010) (27)

RDB

Zarate et al. (2012) (24)

RDB, cross-over

Intervention Pramipexole (n = 12); mean dose: 1.7 mg ⁄ day, SD = 1.3 Placebo (n = 10) Lamotrigine (n = 21); target dose: 150–205 mg Inositol (n = 23); target dose 10–25 mg Risperidone (n = 22); up to 6 mg as tolerated Ketamine IV (n = 18); 0.5 mg ⁄ kg Placebo IV (n = 18) Modafinil (n = 41); mean dose: 177 mg ⁄ day Placebo (n = 44) RUL: 6 · ST BF: 1.5 · ST

RUL: 6 · ST (n = 13) BF: 1.5 · ST (n = 22) BT: 1.5 · ST (n = 15) Ketamine IV (n = 15) Placebo IV (n = 15)

Duration (weeks) 6

16

2

6

Unspecifieda

Unspecifieda

2

Outcome measures Response: ‡50% reduction in HDRS for at least two consecutive weeks

Recovery: £ two symptoms meeting DSM-IV threshold criteria for a major depressive, manic, or hypomanic episode, and no significant symptoms present for eight weeks

Response: 50% improvement from baseline on MADRS Remission: MADRS scores