SEARCH

SEARCH BY CITATION

Keywords:

  • affective disorders;
  • antipsychotics;
  • bipolar disorder;
  • psychopharmacology;
  • review of the literature

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Declaration of interest
  8. References

Objective

The integration of new treatments into the market and routine clinical practice should be dependent on robustness of evidence from randomised controlled trials (RCTs). We assessed study designs of long-term studies for bipolar disorder of all second-generation antipsychotics (SGAs) submitted to the Food and Drug Administration (FDA) and the completeness of evidence submitted to the regulatory agency.

Method

Systematic review of double-blind RCTs comparing SGAs with placebo or active drugs in adults. FDA website and electronic databases were searched until July 2013.

Results

Six placebo-controlled trials comparing aripiprazole, olanzapine, quetiapine and ziprasidone were found in the FDA website. Electronic searches found four additional RCTs about aripiprazole, olanzapine or quetiapine. All RCTs (either submitted to FDA or not) selected patients who responded to acute treatment to increase the treatment effect observed in the long-term phase (enrichment design). By contrast, in the prescribing information sheets for all SGAs, the reported indication was ‘maintenance treatment of bipolar disorder’.

Conclusion

Extrapolation of results from enrichment studies to the more general population of patients should be carried out cautiously because average treatment benefits are likely to be less in unselected patients. Clear guidance for regulatory submission of RCTs is needed.

Summations
  • All randomised controlled trials (RCTs) submitted to Food and Drug Administration (FDA) investigating efficacy and acceptability of second-generation antipsychotics (SGAs) for long-term treatment in bipolar disorder used the enrichment design, which uses a prerandomisation active run-in to select patients who had already responded to an acute treatment phase with the same drug.
  • The treatment effect observed in the long-term phase of RCTs with an enrichment design can only be expected in patients who have responded to acute-phase therapy. Extrapolation of the results of enrichment studies to the more general population of patients should therefore be carried out cautiously because these average treatment benefits are likely to be less in unselected patients.
  • A clear guidance for regulatory submission of RCTs with clinically appropriate designs examining the relative efficacy and acceptability of psychotropic drugs in the long-term treatment for bipolar disorder is needed.
Considerations
  • An adequate duration of the maintenance phase is of crucial importance to determine the efficacy and safety of relapse prevention treatments. Longer studies are expensive and difficult to conduct; however, this information is needed before widespread deployment of costly and untested new drugs.
  • All the studies submitted to FDA have been now published, although there was a significant delay between regulatory submission and the time they were made publicly available.
  • The quality of data available in the FDA website was sometimes obscure.

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Declaration of interest
  8. References

Bipolar disorder is a chronic relapsing and remitting disorder associated with significant morbidity, mortality and a considerable economic burden on society [1]. Long-term treatment is often required to minimise the risk of recurrence [2]. Even though both psychological and psychosocial interventions have an important part to play, the primary treatments are pharmacological [3]. The term ‘continuation therapy’ has been applied to the continuation of a treatment which has shown to be effective in the acute phase, with the purpose of preventing relapse. The term ‘maintenance treatment’ is applied to longer treatment aimed at preventing recurrence in patients at high risk of new mood episodes. The need for maintenance treatment is supported by the desire to prevent the costs of future episodes, and it has been suggested that the optimal duration of maintenance therapy for patients at risk of relapse is 5 years [2]. One of the key issues in the design of long-term maintenance studies is the use of active run-in periods, which are used to select patients to be entered in a clinical trial before random allocation [4]. By selecting for specific patient populations (non-compliant subjects, placebo responders, subjects not tolerating or not responding to active drug), active run-in phases may have huge implications in the interpretation of study results and in their application into everyday clinical practice [5, 6].

Several second-generation antipsychotics (SGAs) have been recently approved by the US Food and Drug Administration (FDA) for long-term treatment for bipolar disorder (www.fda.gov) and are likely to be used clinically because they are effective in the acute manic phase [7]. The integration of new treatments into routine clinical practice, as well as their relevance to particular illness phases, should be dependent upon the strength and robustness of randomised controlled trials (RCTs). To assess methodological rigour and consequent robustness of evidence is essential in considering the place of SGAs in therapy.

Aims of the study

The aim of this article is to provide a systematic overview of study designs of long-term studies for bipolar disorder of all second-generation antipsychotics submitted to Food and Drug Administration, as an example of long-term treatment studies in psychiatry in general. To have supplemental information on FDA studies and to check the completeness of evidence submitted to the regulatory agency, we also carried out a systematic review of the available scientific literature.

Material and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Declaration of interest
  8. References

Included studies

Double-blind RCTs comparing SGAs with placebo or other active drugs for long-term treatment for bipolar disorder, either as monotherapy or add-on treatment, were included. Long-term treatment was defined as a treatment which had been instituted specifically or mainly to prevent further episodes of illness. Continuation trials, in which patients were randomised only before the first phase of acute treatment to see whether the effect of treatment seen in the short term is maintained during the whole episode of illness, were therefore excluded.

Participants

Patients aged 18 years or older with a primary diagnosis of bipolar disorder were included. Studies adopting any standardised diagnostic criteria to define patients suffering from bipolar disorder (all subtypes: Type I and II, rapid cycling and other) were included. RCTs of patients with cyclothymia were excluded.

Search strategy

Appropriate terms for bipolar disorder (bipolar disorder or manic-depressive psychosis or bipolar depression or manic depression) and individual SGAs (aripiprazole, Abilify, asenapine, Saphris, iloperidone, Fanapt, olanzapine, Zyprexa, paliperidone, Invega, quetiapine, Seroquel, risperidone, Risperdal, sertindole, Serdolect, ziprasidone, Geodon, Zeldox) were searched in the FDA website (www.fda.gov) between 1 November 2010 and 31 July 2013 (Fig. 1). The additional systematic search was carried out using the following electronic databases: Cochrane Collaboration Depression, Anxiety and Neurosis Controlled Trials Register and Cochrane Central Register of Controlled Trials (July 2008), MEDLINE (1966-July 2013), EMBASE (1980-July 2013), CINAHL (1982-July 2013) and PsycINFO (1872-July 2013). The WHO registry for randomised controlled studies (http://www.who.int/ictrp/en/) and clinicaltrials.gov were also searched (July 2013). No language restrictions were applied. Websites of pharmaceutical companies marketing SGAs were checked, and possibly authors of included studies were contacted to provide supplemental or unpublished data.

image

Figure 1. Included and excluded studies (PRISMA flow diagram). FDA, Food and Drug Administration; RCT, randomised controlled trial. *, studies included in both FDA website and electronic search. §, studies retrieved by the electronic search but not included in the FDA website.

Download figure to PowerPoint

Study selection and data extraction

All identified studies (either published or unpublished) were screened and reviewed independently by two reviewers. Data were extracted by one reviewer and double-checked by a second reviewer. Any disagreement was resolved by consensus discussion with a third member of the research team.

Criteria of study description

To the best of our knowledge, the FDA website does not provide any tool defining which criteria should be employed when designing a long-term study for bipolar disorder. By contrast, in 2001, the Committee for Proprietary Medicinal Products of the European Medicines Agency (EMA), the European regulatory agency, had issued some notes intended to provide a guidance on clinical investigation of medical products in the treatment for bipolar disorder, to assist applicants in the interpretation with respect to specific problems presented by compounds in the prevention of recurrent episodes of bipolar disorder [8]. We used this guidance to identify the key methodological features addressed in long-term studies for bipolar disorder: i) study enrolment criteria (study population and selection of patients), ii) study design (active run-in phase, length of follow-up, study comparator) (Table 1). As we were interested in deviations from available evidence, information about prescribing labelling and the approval history of each drug was also collected from the FDA website.

Table 1. Summary of main methodological aspects reported in the ‘Guidance on clinical investigation of medicinal products for the prevention of bipolar disorder’ [8] and used in our review
Indication
Recurrence preventionEfficacy has to be demonstrated in long-term prophylactic treatment studies with the purpose of preventing recurrence of manic/depressive symptoms (new episodes)
Bipolar disorderA compound should show efficacy in the treatment for manic episodes, episodes of major depression and prevention of manic/depressive episodes
Method to assess efficacyBoth the proportion of patients developing mania or depression over time and time to either event are of importance to describe the effect of a product in the prevention of recurrence (both should be available in statistical analysis). The number of patients should be sufficient to address both recurrence of depression and of mania. Researches should use the same scales to document recurrence as they have used in the acute studies. Usually a recurrence includes the reappearance of manic or depressive symptoms, scored on validated scale during one or more visits
Study population and selection of patientsPatients with bipolar I disorder who are in full remission should be included in the study. The disorder should be classified according to an internationally acknowledge classification system, preferably DSM-IV, using the diagnostic criteria given herein. Detailed clinical history should be documented. To demonstrate recurrence prevention, it is recommended to include patients with a reasonably high recurrence rate only. However, patients included should be free of manic/depressive symptoms for a sustained period of time at the start of the study so that if manic or depressive symptoms occur, recurrence can be distinguished from relapse
Study designLong-term trials are needed vs. placebo and/or suitable comparator (e.g. lithium) at adequate dosage. The duration of the study should be long enough to demonstrate recurrence prevention. When concomitant medication is allowed that could affect outcome, it should be documented in detail and its impact on efficacy should be evaluated. Prior medication and ‘other’ concomitant medication have to be documented in detail
Initial studiesThe usefulness of including more than one dose of new compound to investigate the optimal dose for long-term treatment should be considered
Main studiesControlled studies against placebo and reference therapy (3-way studies) are first choice. In this way, both the effect of the new compound can be shown and the relative benefit/risk can be assessed. If only an active comparator is used, assay sensitivity should be addressed adequately. Due to the changing course of bipolar disorder and the chronic character with recurrences of depressive and (hypo)manic symptoms, studies with a long duration (at least 1 year) are necessary

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Declaration of interest
  8. References

On the FDA website, six RCTs evaluating four SGAs in the long-term treatment for bipolar disorder were found (two studies for aripiprazole and quetiapine, and one for olanzapine and ziprasidone respectively) (Fig. 1). All six RCTs were placebo-controlled, and no studies had any active comparator [9-14]. Aripiprazole was the only SGA to be investigated both as monotherapy and as adjunctive treatment to mood stabiliser (lithium or valproate). Olanzapine was investigated as monotherapy, while ziprasidone and quetiapine as add-on treatment to lithium or valproate only (Table 2). The marketing authorisation for maintenance treatment for bipolar disorder was given by FDA between 2004 and 2011 as follows: olanzapine and aripiprazole as monotherapy in 2004 and 2005; quetiapine, ziprasidone and aripiprazole as combination treatment with lithium or valproate in 2008, 2009 and 2011 respectively. Study data posted in the FDA website were usually reported in a narrative way, with few figures (often percentages without actual numbers); few data about adverse events occurring during the trials were reported and were seldom suitable for checking or reanalysis. The duration of the follow-up period ranged between 6 and 24 months, and the studies were published between 2006 and 2011: aripiprazole [9, 14], olanzapine [10], quetiapine [11, 12] and ziprasidone [13].

Table 2. Key methodological aspects of all studies included in the present systematic review (either included or not included in the FDA website)
Study IDComparisons (N)Enrichment design?Marketing authorisation for long-term treatment for bipolar disorder?
MonoAdd-onSummary of marketing authorisation
  1. N, number of patients randomised; ARI, aripiprazole; LTI, lithium; OLZ, olanzapine; PBO, placebo; QTP, quetiapine; ZIP, ziprasidone; MS, mood stabilisers (i.e. lithium or valproate); Mono, monotherapy; Add-on, add-on (or augmentation) treatment; Lam, Lamotrigine.

  2. a
  3. b
  4. c
  5. d
Included in FDA website
Kecka [9]ARI (78)PBO (83)YesYesYesMaintenance treatment for bipolar I disorder, both as monotherapy and as an adjunct to lithium or valproate
Marcusa [14]ARI + MS (168)PBO + MS (169)Yes
Tohenb [10]OLZ (225)PBO (136)YesYesNo…and maintenance treatment for bipolar I disorder
Vietac [11]QTP + MS (336)PBO + MS (367)YesNoYesMaintenance treatment for bipolar disorder as an adjunct to lithium or divalproex
Suppesc [12]QTP + MS (310)PBO + MS (313)Yes
Bowdend [13]ZIP + MS (127)PBO + MS (113)YesNoYesMaintenance treatment for bipolar I disorder as an adjunct to lithium or valproate
Not included in FDA website
Tohen [15]OLZ + MS (51)PBO + MS (48)Yes
Tohen [16]OLZ (217)LIT (214)Yes
Weisler [17]QTP (404)LIT (418)Yes
PBO (404)
Carlson [18]ARI + Lam (178)PBO + Lam (173)Yes

The electronic search found four additional RCTs not included in the FDA website: two 2-arm studies using olanzapine [15, 16] one 3-arm study using quetiapine [17] and one 2-arm study using aripiprazole as investigational drug [18]. Three of these studies had a placebo-controlled arm [15, 17, 18] and two compared SGA with lithium [16, 17] (Fig. 1 and Table 2). Full description of study characteristics and references to unpublished material are reported in Table 3.

Table 3. Full description of studies included in the systematic review (either included or not included in the FDA website)
StudyMethodsParticipantsInterventionPrimary outcome, randomisation and blinding
Kecka [9]Randomised, double-blind, parallel-group, placebo-controlled, multicentre study, enrolling patients from 76 centres in three countries (Argentina, Mexico, US) from March 2000 to June 2003. During a stabilisation phase, patients received open-label treatment with aripiprazole for 6–18 weeks. Patients remained in this phase until their symptoms were stable (YMRS score ≤10 and MADRS score ≤13 over a minimum of 6 weeks). A total of 633 patients were enrolled in the study; 567 entered the stabilisation phase, including 333 who had participated in earlier studies of aripiprazole for acute mania. Of the 567 participants, 206 (36%) completed the stabilisation phase and 361 (64%) discontinued, mainly because adverse events (22%), lack of efficacy (12%), or withdrawal of consent (12%). A total of 161 patients (28%) who completed the stabilisation phase were randomly assigned to receive either aripiprazole or placebo. Study duration: 6 monthsAll study participants met DSM-IV criteria for bipolar I disorder. Patients (men and women aged 18 years and older) could enter the stabilisation phase if they recently completed a 3-week, placebo-controlled, acute mania study of aripiprazole; if they met eligibility criteria for an acute mania study but declined participation; or they experienced a manic or mixed episode requiring hospitalisation and treatment within the previous 3 months. Exclusion criteria: pregnancy or lactation; history or symptoms of a cognitive disorder, schizophrenia or schizoaffective disorder, psychotic symptoms better explained by another medical condition or attributed to substance abuse, allergy or hypersensitivity to aripiprazole or other quinolinones, history of neuroleptic malignant syndrome, seizure disorder, participation in a clinical trial with another investigational agent within the past month or electroconvulsive therapy within the past 2 monthsStudy medications were administered orally, once a day, at approximately the same time each day. During the stabilisation phase, patients received open-label treatment with aripiprazole at a starting dose of 30 mg/day. The dose could be decreased o 15 mg/day at any time for tolerability. Subsequently, patients were randomised to the aripiprazole dose they were taking at the end of stabilisation phase or to PBO. At any time during the double-blind phase, the dose of aripiprazole or placebo could be increased or decreased to 30 or 15 mg, respectively, based on the investigator's assessment of therapeutic effect and tolerability. During the whole study, all psychotropic medications except lorazepam and anticholinergic agents were excludedTime to relapse for a mood episode (whether manic, depressive or mixed) during the double-blind phase. Relapse was defined by a discontinuation of the study attributed to lack of efficacy (i.e. hospital admission, new psychotropic drug for manic and/or depressive symptoms). Randomisation in a 1 : 1 ratio (no other details about randomisation or blinding were reported). Aripiprazole was not tapered in patients randomly assigned to placebo. The double-blind period was designed a priory to consist of an initial 26-week phase, followed by a prospective, 74-week extension phase. Patients who completed the initial 26-week period without a relapse were offered to continue a further 74 weeks of double-blind treatment under the same treatment regimen to allow them a total of 100 weeks of double-blind treatment [29]
Tohenb [10]

Randomised, double-blind, placebo-controlled, parallel study. In the open-label phase, patients who met the enrolment criteria after a 2- to 7-day screening period received open-label olanzapine in a flexible dose (5–20 mg/day) for 6–12 weeks. Patients unable to tolerate the minimum dose were discontinued. During the first 3 weeks of open-label treatment, additional antimanic medications were tapered. After the third week, concomitant psychotropic medications were discontinued, excluded the benzodiazepines (not to exceed 6 mg/day of lorazepam equivalents). Beginning with week 6 (and no later than week 12), patients in symptomatic remission for at least 2 consecutive weekly visits became eligible for randomisation to the maintenance phase.

Study duration: 12 months

Bipolar disorder patients with symptomatic remission after open-label acute treatment with olanzapine for an index manic or mixed episode. In- and out-patients (at least 18 years old) were recruited at 44 sites in the US and five sites in Romania. 361 participants entered the double-blind phase (225 with olanzapine; 136 with placebo). The patients met the DSM-IV criteria for an index manic or mixed episode of bipolar I disorder on the basis of the SCID-IV and were required to have a YMRS total score ≥20 at screening and at the start of open-label acute treatment. In addition, patients had to have experienced at least two prior manic or mixed episodes within 6 years of study entryOlanzapine: from 5 to 20 mg/day. During double-blind treatment, study drugs were initially dispensed at a dose that was equivalent to that received at the end of open-label treatment. After 1 day, the daily dose could be adjusted within the allowed dose range (5–20 mg/day). Dose decreases in any number of tablets were permitted in response to adverse events. Benzodiazepines were permitted during the double-blind phase but could not exceed 4 mg/day of lorazepam equivalents during the first week and could not exceed 2 mg/day of lorazepam equivalents during the remainder of the study period (total duration of benzodiazepines use could not exceed 60 cumulative days). Anticholinergic use was permittedTime to and incidence of symptomatic relapse into any mood episode. Symptomatic relapse (any mood episode): YMRS total score ≥15, HAMD score ≥15, or hospitalisation for manic, mixed or depressive episode. Patients were randomly assigned in a 2 : 1 ratio to double-blind maintenance treatment with either 5–20 mg/day of olanzapine or placebo. This assignment ratio minimised the number of patients receiving placebo
Tohenc [16]

Randomised, double-blind, controlled trial. Four study periods:

Screening (two visits over 2–7 days)

Open-label cotreatment (6–12 weeks; twice weekly visits for the first 2 weeks, weekly thereafter)

Double-blind taper (4 weeks; weekly visits)

Double-blind monotherapy (48 weeks; biweekly during the first 4 weeks, monthly thereafter).

From 87 in-patient and out-patient settings across Western Europe, Canada, South Africa, Israel, Australia, and New Zealand between August 1999 and June 2002. Four hundred and thirty-one participants entered the double-blind phase (217 with olanzapine; 214 with lithium). Study duration: 12 months

Bipolar disorder patients (aged more than 18 years), meeting DSM-IV criteria for bipolar disorder (current episode manic or mixed) as determined with the SCID-IV. Patients were required to have a YMRS total score ≥20 at baseline and a history of at least two manic or mixed episodes in the preceding 6 years. Patients were excluded from the study if they had a serious, unstable medical illness; met DSM-IV substance dependence criteria (nicotine or caffeine excepted) within the past 30 days; had been treated with a depot neuroleptic in the previous 6 weeks; were at serious suicide risk; had a history of intolerance, or lack of response, to an adequate trial of lithium or olanzapine as determined by the investigatorOpen-label cotreatment: allowed dosages of olanzapine were 5–20 mg/day (starting dose: 15 mg); required to LI target blood level of 0.6–1.2 mEq/l by week 4 (starting dose: 600 mg). Only oral or i.m. haloperidol and zuclopenthixol were permitted for extreme agitation during the open-label period. Benzodiazepines were allowed: maximum 8 mg/day in lorazepam equivalents up to the first 6 weeks of the open-label period, maximum 6 mg/day for the remainder of the open-label period and during the first 2 weeks of the taper period. Up to 4 mg/day for the remaining 2 weeks of the taper period and then to 2 mg/day (for not more than 60 cumulative days) for the double-blind monotherapy period. Anticholinergics were permitted for treatment-emergent extrapyramidal symptomsIncidence of symptomatic relapse into any mood episode. Symptomatic relapse (any mood episode): YMRS total score ≥15 and/or HAMD score ≥ 15. Randomisation (1 : 1 ratio) by means of a unique drug kit number via a call-in interactive voice response system. Patients, site personnel and sponsor investigators were blind to randomisation codes. During the double-blind taper period, patients remained on their current dose of randomly assigned treatment, and the dose of the discontinued drug was tapered in a blinded, a priori-determined, stepwise manner over 4 weeks. To maintain the blind associated with blood draws, all patients randomly assigned to olanzapine also had blood drawn. For every outlier report generated for a lithium patient, a sham lithium outlier report was sent to an olanzapine patient
Tohend [15]

Double-blind, randomised, controlled trial. After a screening and washout period (2–7 days), a 6-week double-blind period follows (olanzapine – from 5 to 20 mg/day – plus lithium or valproate vs. placebo and lithium or valproate). After a rerandomisation, there is a 78-week double-blind therapy period for responders (olanzapine – from 5 to 20 mg/day – plus lithium or valproate vs. placebo and lithium or valproate).

Study duration: 18 months

99 participants, aged 18–70 years (51 with olanzapine plus lithium or valproate; 48 with placebo plus lithium or valproate) with diagnosis of bipolar I disorder, manic or mixed episode, with or without psychotic features, using the SCID–IV, in syndromic remission. In addition, participants had to have experienced at least two previous mood episodes (manic, depressed or mixed) and a documented trial at a therapeutic blood level of lithium or valproate for at least 2 weeks immediately prior to enrolment, demonstrating persistent manic symptoms (YMRS score ≥16). Exclusion criteria: serious and unstable medical illness; DSM–IV substance dependence within the past 30 days (except nicotine or caffeine); pregnancy; documented history of intolerance to olanzapine; and serious suicidal riskOlanzapine (range 5-20 mg/day; starting dose: 10 mg) in combination with lithium or valproate (combination therapy), or placebo added to lithium or valproate (monotherapy). Patients continued taking the same mood stabiliser that they had received during the acute phase, the choice of which was determined by the site investigator. During this relapse prevention phase of the study, the blood levels of lithium (0.6–1.2 mm) and valproate (50– 125 mg/ml) remained within the therapeutic range. Only benzodiazepines (max. 2 mg lorazepam equivalent per day for max. five consecutive days, or 60 days cumulatively) and anticholinergics (for treatment for extrapyramidal side-effects but not for prophylaxis) were permittedTime to syndromic relapse of bipolar disorder (either mania or depression) as defined by DSM-IV criteria. Randomisation (unique drug kit number via a call-in interactive voice response system) in a 1 : 1 ratio. Patients, study site personnel and sponsor investigators were masked to the randomisation codes. Only the addition of olanzapine or placebo was conducted under double-masked conditions. To maintain masking, treatment took the form of two 5-mg capsules (either olanzapine or placebo), titrated up in increments of one capsule, or down by any number of decrements at the investigator's discretion, as indicated by each patient's symptom improvement and treatment tolerance
Vietae [11]Multicentre, randomised, parallel-group, double-blind study, comparing quetiapine with placebo, as adjunct to lithium or valproate in adult patients with bipolar I disorder (both mania and depression). The study was conducted in 37 in the United States and 140 in other countries, including Australia, Europe and South Africa between April 2004 and October 2006. The study consisted of enrolment and two phases, the initial open-label treatment phase (12–36 weeks) and the subsequent randomised treatment phase (up to 104 weeks). To be randomised, patients must have been treated with quetiapine within the range of 400–800 mg/day and lithium or valproate for at least 12 weeks during the open-label treatment phase and also YMRS and MADrS total scores ≤12. Study duration: up to 2 yearsMale or female patients, aged 18 years or older, with bipolar I disorder, who had experienced an acute manic, depressed or mixed episode at enrolment; or a past manic, depressed or mixed episode within 26 weeks, as documented by medical records, treated with quetiapine and mood stabiliser (lithium or valproate). The patients should have had at least one manic, depressed or mixed episode in the 2 years prior to the index episodeOral dose of open-label quetiapine, 400–800 mg/day in divided doses, (recommended target dose: 600 mg/day – doses could be adjusted). After randomisation, open-label 100-mg quetiapine tablets were titrated down and replaced with 100-mg tablets of blinded investigational product at a rate of one tablet per every 2 days. The dose of blinded investigational product could be adjusted as clinically indicated. Patients began or continued on an oral dose of open-label lithium or valproate chosen by the investigator according to clinical judgment. Lithium or valproate dose was adjusted at the discretion of the investigator (target serum concentrations: 0.5–1.2 mEq/l for lithium and 50–125 μg/ml for valproateTime to recurrence of a mood event. Recurrence was defined as i) initiation of an antipsychotic, antidepressant, mood stabiliser other than the assigned mood stabiliser, anxiolytic other than lorazepam or any other medication to treat a manic, depressed or mixed event, ii) hospitalisation for a manic, depressed or mixed event, iii) YMRS score ≥20 at two consecutive assessments or at the final assessment if the patient discontinues, or MADRS score ≥20 at two consecutive assessments or at the final assessment if the patient discontinues, or iv) discontinuation from the study by the patient if, in the opinion of the investigator, the discontinuation was due to a manic, depressed or mixed event
Suppesf [12]

Multicentre (127 centres in US and Canada, between March 2004 and September 2006), randomised, double-blind study comparing quetiapine with placebo, as adjunct to lithium or valproate in the maintenance treatment of adult patients with bipolar I disorder (both mania and depression). The design of this study is identical to the study by Vieta et al. [11].

Study duration: up to 2 years

The design of this study is identical to the study by Vieta [11] (see above)The design of this study is identical to the study by Vieta [11] (see above)The design of this study is identical to the study by Vieta [11] (see above)
Weislerg [17]Multicentre, randomised, parallel-group, double-blind, placebo-controlled study to evaluate the efficacy and safety of quetiapine and lithium for up to 104 weeks in adult patients with bipolar I disorder. The study was conducted between March 2005 and July 2007. Countries of recruitment: Argentina, Bulgaria, Colombia, India, Lithuania, Malaysia, Mexico, Peru, Philippines, Romania, Russian Federation, Taiwan, Thailand, Ukraine, United States. The study was terminated early after planned interim analysis provided positive results

Male or female patients, aged 18 years or older. Inclusion Criteria:

Diagnosis of bipolar I disorder (current manic, mixed or depressed)

Any mood episode during the last 26 weeks that was treated with quetiapine

Need of reliable method of contraception for female patients.

Exclusion Criteria:

Pregnancy

Substance or alcohol dependence

Medical illness

Use of an experimental drug within 30 days of enrolment

Patients aged ≥18 years with DSM-IV-diagnosed bipolar I disorder and a current or recent manic, depressive, or mixed episode received open-label quetiapine (300–800 mg/day) for 4–24 weeks. Patients achieving stabilisation were randomised to continue quetiapine or to switch to placebo or lithium (0.6–1.2 mEq/l) for up to 104 weeks in a double-blind trialPrimary outcome: time from randomisation to recurrence of a mood event. Secondary outcomes: time from randomisation to recurrence of a manic or depressed event. Safety assessments included adverse events and laboratory values
Bowdenh [13]

Subjects with DSM-IV bipolar I disorder with a Mania Rating Scale score ≥14 were enrolled. Subjects achieving ≥8 consecutive weeks of stability with open-label ziprasidone (80–160 mg/day) and lithium or valproate over a period of at least 10 weeks and up to 16 weeks. (period 1) were randomly assigned in the double-blind maintenance period (period 2) to ziprasidone plus mood stabiliser or placebo plus mood stabiliser. The study was conducted from December 2005 to May 2008.

Study duration: 6 months

Inclusion criteria: adults meeting DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition) criteria for bipolar I disorder (currently with manic or mixed symptoms)

Exclusion Criteria: Ultra rapid cyclers and subjects with significant cardiovascular disease including history of QT prolongation and/or congenital long QT syndrome

Patients randomised to ziprasidone continued to receive the same stable treatment regimen achieved during the open-label treatment, that is, either 80, 120 or 16 mg/day for up to 24 weeks of double-blind treatment. Patients randomised to placebo were tapered off the open-label ziprasidone by 40 mg/day every 2 days (in a double-blinded manner) until they were completely off ziprasidone and were on matching placebo capsules for up to 24 weeks of double-blind treatment

Primary outcome: time to intervention for a mood episode. Secondary outcomes:

time to discontinuation for any reason

Change from baseline in Mania Rating Scale

Change from baseline in Clinical Global Impression Severity score

Clinical Global Impression - Improvement

Change from baseline in Montgomery-Asberg Rating Scale Score

Change from baseline in Positive and Negative Syndrome Scale score

Marcusi [14]

Multicentre, randomised, double-blind, placebo-controlled out-patient study.

In Phase 1, patients with acute manic or mixed symptomatology were assigned lithium (0.6–1.0 mm) or valproate (50–125 μg/ml). Patients with a partial non-response (Y-MRS ≥16) were eligible to enter Phase 2, where aripiprazole (10–30 mg/day; starting dose: 15 mg/day) was assigned in single-blind fashion. Three hundred and thirty-seven patients responding to the combination treatment, and maintaining response for 12 weeks, were randomised (1 : 1) double-blind to continuation of aripiprazole plus mood stabiliser or placebo plus mood. Study duration: 12 months

The population included in this study were male and female out-patients, ≥18 years of age, with a diagnosis of bipolar I disorder as defined by DSM-IV-TR who displayed an acute manic or mixed episode as confirmed by the SCID (Y-MRS total score ≥16), with or without psychotic features. Patients who are using lithium or valproate immediately prior to study entry were included as long as they met the criteria

outlined above.

Phase 1 (2–8 weeks): 250- and 300-mg oral lithium tablets or 250-mg valproic acid oral tablets

Phase 2 (13–24 weeks): aripiprazole 10-and 15-mg oral tablets and 250- and 300-mg oral lithium tablets or 250-mg valproic acid oral tablets

Phase 3 (52 weeks): aripiprazole 10- and 15-mg oral tablets and 250- and 300-mg oral lithium tablets or 250-mg valproic acid oral tablets

Primary outcome: time to relapse to any mood episode.

Secondary outcomes: mean change from baseline in the following scales: Y-MRS total score; MADRS total score; CGI-BP severity of illness score; percentage of patients maintaining remission during Phase 3; time to manic relapse; time to depressive relapse.

Safety outcome measures included the frequency and severity of adverse events, serious adverse events (clinical and laboratory) and discontinuations from study due to adverse events, vital signs (supine and standing positions), body weight, electrocardiogram, routine and special laboratory tests

Carlsonj [18]

Subjects with DSM-IV-TR bipolar I disorder with a manic or mixed episode were enrolled. Subjects achieving at least 8 consecutive weeks of stability (YMRS and MADRS total score ≤12) with single-blind aripiprazole (10-30 mg/day) and open-label lamorigine (100-200 mg/day) over a period of 9–24 weeks (stabilisation phase) were randomly assigned to aripiprazole (double-blind) plus lamotrigine (open label) or placebo (double-blind) plus lamotrigine (open label). The study was conducted from December 2005 to July 2009.

Study duration: 52 months

Inclusion criteria: adults meeting DSM-IV-TR criteria for bipolar I disorder with a manic or mixed episode in the previous 3 months [YMRS ≥ 16], with or without psychotic features and with or without a history of rapid cycling (≥4 but ≤7 mood episodes per year). Exclusion criteria: DSM-IV-TR diagnosis of delirium, dementia, amnestic or other than bipolar I disorder; pregnancy; first manic or mixed episode; treatment refractory manic or mixed episodes; a history of aripiprazole treatment within the previous 3 months; a prestudy lamotrigine dose >200 mg/day that could not be titrated down to 100 or 200 mg/day; risk of committing suicide

Starting dose for single-blind aripiprazole: 10 mg/day, with a target dose of 15 mg/day (then, dose could be adjusted by increments or decrements of 5 mg, between 10–30 mg/day). Open-label lamotrigine (as add-on to aripiprazole) was started at 25 mg/day (target dose of 200 mg/day). Patients stabilised were randomised (1 : 1) to continue treatment with either double-blind aripiprazole or double-blind placebo in combination with open-label lamotrigine.

Doses of aripiprazole and lamotrigine could be adjusted. Clinical assessments: every 2 weeks for the first 8 weeks and then monthly up to Week 52

The criteria for relapse included one or more of the following events: hospitalisation for a manic or mixed episode; a serious adverse event or worsening disease during the study; or discontinuation due to a lack of efficacy (as determined by the investigator).

Secondary endpoints were time to any relapse (manic, mixed, depressive), time to depressive relapse, and time to discontinuation for any reason. Other outcome measures were the mean change from baseline in YMRS and MADRS total scores, and Clinical Global Impression–Bipolar Disorder (CGI-BP) severity of illness score (mania, depression, overall), and mean CGI-BP change from preceding phase score (mania, depression, overall)

All RCTs included in the present systematic review (either submitted to FDA or not) used a so-called enrichment design. This design consists of two stages: the first stage serves as a non-randomised active run-in, which serves as a screening process for selecting a certain subgroup of participant (usually responders) [19]; the succeeding stage is typically a randomised comparison of continued treatment with discontinuation and placebo substitution, within the selected (enriched) subpopulation [20]. Details on active run-in phases and other methodological issues are reported by compound and summarised in the corresponding figures (Figs 2 and 3).

image

Figure 2. Study design for second generation antipsychotics vs. placebo for long term treatment in bipolar disorder, using an enriched design and one randomisation. SGA, second generation antipsychotic; PBO, placebo; R, randomization procedure to allocate remitters to maintenance phase.

Download figure to PowerPoint

image

Figure 3. Study design for olanzapine (a) or quetiapine (b) vs. lithium (or placebo), using an enriched design and one randomisation. OLZ, olanzapine; LIT, lithium; R, randomization procedure to allocate remitters to maintenance phase; N, number; YMRS, Young Mania Rating Scale; MADRS, Montgomery-Asberg Depression Rating Scale; HDRS, Hamilton Depression Rating Scale.

Download figure to PowerPoint

Aripiprazole

In the two aripiprazole studies [9, 14] found in the FDA website, the duration of follow-up of was 6 and 12 months respectively (Fig. 2). In the monotherapy study, 567 patients entered the open-label phase (6–18 weeks), but only 67 (11.8%) completed the overall study and 28 of them (41.8%) had been randomised to placebo after responding to aripiprazole in the first phase of the trial [9]. In the add-on study [14], the enrichment design selected manic patients partially resistant to lithium or valproate, because during the run-in phase all manic patients received first lithium or valproate as open-label treatment and only patients who were still manic after 2–8 weeks were given aripiprazole in a single-blind fashion as add-on treatment (Fig. 2). The completion rate was higher than in the monotherapy study probably because of the combination therapy (27.9% vs 11.1%), but it is noteworthy that 46.3% (89 out of 192) of the patients who completed the study had been randomised to placebo but continued to take lithium or valproate, even though there would be a high likelihood of continued non-response to these mood stabilisers.

The third aripiprazole study [18] was not included in the FDA website. During the 9- to 24-week stabilisation phase, eligible patients were stabilised with open-label lamotrigine and aripiprazole administered in single-blind fashion and then randomised (1 : 1) to continue treatment with either double-blind aripiprazole in combination with open-label lamotrigine or double-blind placebo in combination with open-label lamotrigine. Seven hundred and eighty-seven patients entered and 435 (55.3%) discontinued during the stabilisation phase. Among the 351 patients who were randomised into the double-blind relapse assessment phase, the discontinuation rates were 63.5% and 69.4% for the aripiprazole combination and placebo combination treatment arms respectively.

Olanzapine

In the FDA website, only one monotherapy RCT was found [10]. The enrichment design used was similar to that deployed for aripiprazole monotherapy, but with a somewhat shorter active run-in open-label phase (up to 12 instead of 18 weeks) (Fig. 2). At study end, only 7.8% of the 731 patients who had entered the open-label phase completed the whole 11-month follow-up. The two studies retrieved by the electronic search and not included in the FDA website enrolled a similar patient population in the active run-in phase. One RCT compared olanzapine as monotherapy vs. lithium and lasted for 12 months [16]; the second study investigated olanzapine vs. placebo in combination with lithium or valproate, with a 18-month follow-up [15]. In the monotherapy study, eligible patients entered an open-label cotreatment phase with olanzapine and lithium (Fig. 3a). At randomisation, patients remained on their current dose of assigned treatment, and the dose of the discontinued drug was tapered off in a blinded manner over 4 weeks. At study end, 31.4% of the 543 patients who had entered the open-label phase completed the whole 12-month follow-up.

The add-on treatment study [15] was conceptually similar in design to other SGAs' placebo-controlled RCTs, but uniquely used an enrichment strategy with two sequential randomised phases (Fig. 4). The first randomisation allocated to add-on treatment with olanzapine or placebo patients who were partially non-responsive to valproate or lithium monotherapy (open label). The second randomisation allocated patients who recovered to continue or discontinue olanzapine (with placebo substitution). Only 9.1% of the 229 patients who had originally been randomised to cotreatment with olanzapine and lithium or valproate completed the whole 18-month follow-up.

image

Figure 4. Study design for olanzapine as add-on treatment to lithium or valproate vs. placebo plus lithium or valproate, using an enriched design and two randomisations. OLZ, olanzapine; PBO, placebo; LIT, lithium; DVP, valproate; R, randomization procedure; N, number of patients enrolled; YMRS, Young Mania Rating Scale; HDRS, Hamilton Depression Rating Scale. *, Two definitions of syndromic remission were used by study authors: (a) DSM–IV ‘A’ criteria for current manic episode no worse than mild (score ≤3), ‘B’ criteria no worse than mild (score ≤3), and no more than two ‘B’ criteria that were mild (score 3); or (b) all DSM–IV ‘A’ criteria for current major depressive episode no worse than mild (score ≤3), and no more than three ‘A’ criteria mild (score 3).

Download figure to PowerPoint

Quetiapine

Two identical RCTs comparing quetiapine with placebo as add-on treatment were found in the FDA website [11, 12]. The enrichment design was similar to those used for olanzapine and ariprazole, but treatment regimens and patient inclusion criteria were slightly different (Table 3). After randomisation, the dose of quetiapine could be adjusted as clinically indicated, while lithium or divalproex doses were maintained the same unless tolerability-related issues emerged. At study end, only 15.2% of the 3424 patients who had entered the open-label phase completed the whole 24-month follow–up, and of these 523 patients, 33.6% had been allocated to placebo.

The third quetiapine study was not included in the FDA website [17]. After up to 24 weeks of open-label treatment with quetiapine, patients in remission for at least 4 weeks were randomised to continue quetiapine or switch to placebo or lithium for a 2-year follow-up (Fig. 3b). For patients allocated to lithium or placebo, the replacement of quetiapine was completed by 2 weeks following randomisation. Quoting the evidence from the two FDA quetiapine studies [11, 12] as backing evidence for a better efficacy profile of quetiapine than originally supposed, an interim analysis was added as an amendment to the study protocol to account for the possibility of early closure of the study in case of observed treatment difference between quetiapine and placebo. Based on positive results from the interim analysis (366 vs 364 patients respectively), the study was terminated early; however, only 547 patients of 2438 enrolled in the open-label phase (22.4%) completed the treatment, and most of them were taking lithium or placebo (54.5%).

Ziprasidone

In the ziprasidone add-on study [14], the enrichment design selected manic patients partially resistant to lithium or valproate (Fig. 2). At study entry, treatment resistance or intolerance to ziprasidone, but not to lithium or valproate was an exclusion criterion. The study was designed with a 6-month follow-up only (Fig. 2). Despite this short study duration, only 23.5% of the 586 patients who had entered the combination treatment open-label phase reached the end of the study; of these, 138 patients (39.1%) were allocated to placebo.

Information about prescribing labelling and the approval history

In the FDA website treatment, indications were usually detailed, highlighting the efficacy of SGAs in preventing mainly manic relapses in patients who responded during an acute treatment (for information about prescribing labelling and FDA approval history, see Box 1). It is worth noting, however, that in the front page of the prescribing information sheets for all SGAs, it was reported that the indication was ‘maintenance treatment of bipolar disorder’ (Table 2). Moreover, some discrepancies and inconsistencies between supporting evidence and treatment indications were found, most of all because they changed over time but no clear explanations were provided. For instance, in March 2005 it was reported that one SGA was ‘effective in maintaining efficacy in patients with bipolar I disorder with a recent manic or mixed episode who had been stabilised and then maintained for at least 6 weeks’, but in June 2005 there was a significant change: the effectiveness of the same SGA ‘for more than 3 weeks of treatment for an acute episode, and for prophylactic use in mania, has not been established in controlled clinical trials’. No reason for the change is given on the FDA website, and the accompanying letter did not report the revised indication. In February 2006, the same indication of March 2005 was printed again, using the same study as supporting evidence but without discussing the re-revised indication. In the FDA website, other similar examples are available (www.fda.gov).

Box 1. Information about prescribing labelling and the approval history as reported in the FDA website (with links to web pages)

Treatment indications were usually detailed, highlighting the efficacy of second-generation antipsychotics (SGAs) in preventing mainly manic relapses in patients who responded during an acute treatment, also reminding the physician who elects to use SGAs for extended periods to periodically re-evaluate the long-term risks and benefits of the drug for the individual patient (www.accessdata.fda.gov/drugsatfda_docs/label/2004/20592se1-019_zyprexa_lbl.pdf; www.accessdata.fda.gov/drugsatfda_docs/label/2006/021436s010,021713s006lbl.pdf; www.accessdata.fda.gov/drugsatfda_docs/summary_review/2008/020639se1-037_SUMR.pdf). However, the quality of data available was sometimes obscure. For instance, in March 2005 it was reported as follows: ‘the efficacy of ABILIFY in maintaining efficacy in patients with Bipolar I Disorder with a recent manic or mixed episode who had been stabilized and then maintained for at least 6 weeks, was demonstrated in a double-blind, placebo-controlled trial. Prior to entering the double-blind, randomization phase of this trial, patients were clinically stabilized and maintained their stability for 6 consecutive weeks on ABILIFY. Following this 6-week maintenance phase, patients were randomized to either placebo or ABILIFY and monitored for relapse …’ (www.accessdata.fda.gov/drugsatfda_docs/label/2005/21436s005,008,21713s003lbl.pdf).

The accompanying letter quotes: ‘Supplemental new drug application NDA 21-436/S-005 provides for the use of Abilify® Tablets as maintenance therapy in Bipolar I Disorder; supplemental new drug application NDA 21-713/S-003 is a labeling supplement to provide for similar use of Abilify® Oral Solution’ (www.accessdata.fda.gov/drugsatfda_docs/appletter/2005/21436s005,008,21713s003ltr.pdf).

In June 2005, there was a significant change as follows: ‘The efficacy of ABILIFY was established in two placebo-controlled trials (3 week) of inpatients with DSM-IV criteria for Bipolar I Disorder who were experiencing an acute manic or mixed episode with or without psychotic features … However, the effectiveness of ABILIFY for longer-term use, that is, for more than 3 weeks of treatment of an acute episode, and for prophylactic use in mania, has not been established in controlled clinical trials. Therefore, physicians who elect to use ABILIFY for extended periods should periodically re-evaluate the long-term risks and benefits of the drug for the individual patient…’ (www.accessdata.fda.gov/drugsatfda_docs/label/2005/021713s004,021436s007lbl.pdf). No reason for the change is acknowledged anywhere in the FDA website and the accompanying letter did not even report the revised indication (www.accessdata.fda.gov/drugsatfda_docs/appletter/2005/021713s004,021436s007ltr.pdf).

Interestingly, in February 2006, the same indication of March 2005 was printed again, using the same study as supporting evidence but without discussing the re-revised indication: ‘The efficacy of ABILIFY in maintaining efficacy in patients with Bipolar I Disorder with a recent manic or mixed episode who had been stabilized and then maintained for at least 6 weeks, was demonstrated in a double-blind, placebo-controlled trial. Prior to entering the double-blind, randomization phase of this trial, patients were clinically stabilized and maintained their stability for 6 consecutive weeks on ABILIFY. Following this 6-week maintenance phase, patients were randomized to either placebo or ABILIFY and monitored for relapse ··· Physicians who elect to use ABILIFY for extended periods, that is, longer than 6 weeks, should periodically re-evaluate the long-term usefulness of the drug for the individual patient…’ (www.accessdata.fda.gov/drugsatfda_docs/label/2006/021436s010,021713s006lbl.pdf).

As before, in the accompanying letter, nothing is reported on the matter (www.accessdata.fda.gov/drugsatfda_docs/appletter/2006/021436s010,021713s006ltr.pdf).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Declaration of interest
  8. References

This study shows that all RCTs submitted to FDA investigating efficacy and acceptability of SGAs for long-term treatment in bipolar disorder used a very similar design, the so-called enrichment design. The enrichment design uses a prerandomisation active run-in to select patients who had already responded to an acute treatment phase with the same drug. This has a major impact on the clinical interpretation of results and prescribing indications, because the enrichment will increase the treatment effect observed in the long-term phase, and this level of average benefit could only be expected in patients who have responded to acute-phase therapy. Extrapolation of the results of enrichment studies to the more general population of patients without such careful acute-phase screening should therefore be carried out cautiously with the recognition that average treatment benefits are likely to be less in unselected patients [2].

All RCTs submitted to FDA were placebo-controlled - which is required for regulatory approval of the efficacy of drug treatments. However, active comparisons are also needed when more than one therapeutic option is available. Our search found two active comparator studies not submitted to FDA for regulatory purposes, which both used lithium as control agent [16, 17]. The run-in phase in the quetiapine study [17] clearly favoured this antipsychotic drug because patients after responding for at least 4 weeks to open-label quetiapine monotherapy were switched to lithium or placebo quite abruptly, within 2 weeks only. Interestingly, despite this short tapering period, quetiapine failed to outperform lithium. In the olanzapine study, the run-in phase was balanced, because it used both olanzapine and lithium [16]. However, some concerns were raised [21] about the methodology of this study because it was likely that patients with a higher probability to be olanzapine responders with a history of intolerance or lack of response to an adequate trial of lithium were recruited.

An adequate duration of the maintenance phase is of crucial importance to determine the efficacy and safety of relapse prevention treatments [2]. In the RCTs submitted to FDA, follow-ups ranged between 6 and 24 months. Even though the majority of single studies comparing SGAs with placebo were reasonably large, we consider that a 6-month duration is too short for a true maintenance study in a disorder where treatment is typically continued for many years. Longer studies allow for more accurate assessment of the impact of the relapse prevention treatment on the disease course by including also important secondary outcome measures as adverse effects, direct and indirect medical costs and quality of life [22]. Longer studies are expensive and difficult to conduct, but this information has to be warranted before widespread deployment of costly and untested new drugs.

All the studies submitted to FDA have been now published, although there was a significant delay between regulatory submission and the time they were made publicly available. For instance, olanzapine was approved for long-term treatment for bipolar I disorder in January 2004 (www.accessdata.fda.gov/drugsatfda_docs/appletter/2004/20592se1-019ltr.pdf), but the study supporting this indication was published in February 2006 [10]. In the meanwhile, two other studies investigating the efficacy of olanzapine in long-term treatment for bipolar disorder were published, but they were not used for regulatory purposes [15, 16]. The study with olanzapine as add-on treatment to lithium or valproate [15] is quite similar in design to other two add-on studies [13, 14] submitted to FDA, selecting patients resistant to lithium or valproate who responded to the antipsychotic add-on treatment. From a clinical point of view, it is not clear why aripiprazole and ziprasidone are approved as adjunctive treatment to mood stabilisers, while olanzapine is licensed only as monotherapy.

We did not find any guidance on clinical investigation of drugs for the treatment and prevention of bipolar disorder in the FDA website, despite a comprehensive search. There is a long-standing scientific debate on the matter [23], and the EMA issued a guidance on this topic in 2001, reporting that a product may be developed for the indication bipolar disorder if demonstrated to be effective in both the treatment for manic episodes/depressive and the prevention of further manic/depressive episodes [8]. Clinically, it is worth distinguishing between acute treatment continuation design and recurrence prevention designs. The two different designs answer different clinical questions. Continuation studies address the issue of how long to continue a treatment that has proven to be effective in the acute phase of illness. Maintenance studies should aim at preventing new episodes of illness, regardless of the acute treatment phase, and should include patients with bipolar disorder who are in full remission and free of manic/depressive symptoms for a sustained period of time at the start of the study. If affective symptoms occur, recurrence can be distinguished from relapse, and this type of studies would be clearly more informative for clinicians because of higher external validity. The design of the run-in phase may be crucial for such studies and a recent trial with an active, balanced run-in phase reported efficiency benefits [24].

Even though unlikely, we cannot rule out that one or two studies submitted to FDA are missing in our analysis, as information in the regulatory website was not easy to retrieve. However, after a comprehensive search, we found only one RCT for some SGAs, and it is still not clear how one single placebo-controlled study can be considered sufficient for licencing new drugs. There is the need for more transparent approval policies and greater consistency within regulatory agencies [25]. The best way to do so is to give the public free and easy access to the submitted data to let interested people understand and utilise scientific information, and check robustness of findings.

The quality of data available in the FDA website was sometimes obscure. It is hard to understand why the same study may be used with different conclusions, if no clear reasons or explanations are reported (see Box 1). Considering the available evidence, to report the indication of ‘maintenance treatment for bipolar disorder’ for SGAs is an unjustifiably extended and broadened indication which neglects the enriched design and limited duration. As FDA provides marketing authorisation for these agents, it is likely to lead to wide use of these drugs beyond those situations which have a sound evidence base [26, 27].

In conclusion, there is a need for drafting a clear guidance for regulatory submission of longer term and larger sample size RCTs, with clinically appropriate designs examining the relative efficacy and acceptability of SGAs (and other antipsychotic agents or mood stabilisers) compared with placebo and other maintenance treatment drugs in the long-term treatment for bipolar disorder [28].

Declaration of interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Declaration of interest
  8. References

Drs. Cipriani, Barbui and Rendell report no financial relationships with commercial interests. Dr. Geddes currently receives research funding from the UK Medical Research Council, UK Economic and Social Research Council, the National Institute for Health Research and the Stanley Medical Research Institute. He was expert witness for Doctor Reddy's Laboratories and is Chief Investigator on the CEQUEL trial to which GlaxoSmithKline has contributed and supplied investigational drugs.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Declaration of interest
  8. References
  • 1
    Soares-Weiser K, Bravo Vergel Y, Beynon S et al. A systematic review and economic model of the clinical effectiveness and cost-effectiveness of interventions for preventing relapse in people with bipolar disorder. Health Technol Assess 2007;11:iiiiv, ix-206.
  • 2
    National Institute for Health and Clinical Excellence (NICE). Bipolar disorder: the management of bipolar disorder in adults, children and adolescents, in primary and secondary care (Issue date: July 2006). Available at: www.nice.org.uk/CG038. Last accessed 31st July 2013.
  • 3
    Dias VV, Balanzá-Martinez V, Soeiro-de-Souza MG et al. Pharmacological approaches in bipolar disorders and the impact on cognition: a critical overview. Acta Psychiatr Scand 2012;126:315331.
  • 4
    Cipriani A, Geddes JR. What is a run-in phase? Epidemiol Psichiatr Soc 2010;19:2122.
  • 5
    Pablos-Méndez A, Barr RG, Shea S. Run-in periods in randomized trials: implications for the application of results in clinical practice. JAMA 1998;279:222225.
  • 6
    Berger VW, Rezvani A, Makarewicz VA. Direct effect on validity of response run-in selection in clinical trials. Control Clin Trials 2003;24:156166.
  • 7
    Cipriani A, Barbui C, Salanti G et al. Comparative efficacy and acceptability of antimanic drugs in acute mania: a multiple-treatments meta-analysis. Lancet 2011;378:13061315.
  • 8
    European Medicines Agency. Committee for proprietary medicinal products: note for Guidance on clinical investigation of medicinal products for the treatment and prevention of bipolar disorder, April 2001. Available at: http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2009/09/WC500003528.pdf.
  • 9
    Keck PE Jr, Calabrese JR, McQuade RD et al. A randomized, double-blind, placebo-controlled 26-week trial of aripiprazole in recently manic patients with bipolar I disorder. J Clin Psychiatry 2006;67:626637.
  • 10
    Tohen M, Calabrese JR, Sachs GS et al. Randomized, placebo-controlled trial of olanzapine as maintenance therapy in patients with bipolar I disorder responding to acute treatment with olanzapine. Am J Psychiatry 2006;163:247256.
  • 11
    Vieta E, Suppes T, Eggens I et al. Efficacy and safety of quetiapine in combination with lithium or divalproex for maintenance of patients with bipolar I disorder (international trial 126). J Affect Disord 2008;109:251263.
  • 12
    Suppes T, Vieta E, Liu S et al. Maintenance treatment for patients with bipolar I disorder: results from a north American study of quetiapine in combination with lithium or divalproex (trial 127). Am J Psychiatry 2009;166:476488.
  • 13
    Bowden CL, Vieta E, Ice KS, Schwartz JH, Wang PP, Versavel M. Ziprasidone plus a mood stabilizer in subjects with bipolar I disorder: a 6-month, randomized, placebo-controlled, double-blind trial. J Clin Psychiatry 2010;71:130137.
  • 14
    Marcus R, Khan A, Rollin L et al. Efficacy of aripiprazole adjunctive to lithium or valproate in the long-term treatment of patients with bipolar I disorder with an inadequate response to lithium or valproate monotherapy: a multicenter, double-blind, randomized study. Bipolar Disord 2011;13:133144.
  • 15
    Tohen M, Chengappa KN, Suppes T et al. Relapse prevention in bipolar I disorder: 18-month comparison of olanzapine plus mood stabiliser v. mood stabiliser alone. Br J Psychiatry 2004;184:337345.
  • 16
    Tohen M, Greil W, Calabrese JR et al. Olanzapine versus lithium in the maintenance treatment of bipolar disorder: a 12-month, randomized, double-blind, controlled clinical trial. Am J Psychiatry 2005;162:12811290.
  • 17
    Weisler RH, Nolen WA, Neijber A et al. Continuation of quetiapine versus switching to placebo or lithium for maintenance treatment of bipolar I disorder (trial 144: a randomized controlled study). J Clin Psychiatry 2011;72:14521464.
  • 18
    Carlson BX, Ketter TA, Sun W et al. Aripiprazole in combination with lamotrigine for the long-term treatment of patients with bipolar I disorder (manic or mixed): a randomized, multicenter, double-blind study (CN138-392). Bipolar Disord 2012;14:4153.
  • 19
    Ivanova A, Tamura RN. A two-way enriched clinical trial design: combining advantages of placebo lead-in and randomized withdrawal. Stat Methods Med Res 2011. doi: 10.1177/0962280211431023 [Epub ahead of print].
  • 20
    Fedorov VV, Liu T. Enrichment design. In: D'Agostino RB, Sullivan LM, Massoro JM, eds. Wiley encyclopedia of clinical trials. Hoboken, NJ: John Wiley & Sons, Inc, 2007. doi: 10.1002/9780471462422.eoct342.
  • 21
    Cipriani A, Rendell J, Geddes JR. Olanzapine in the long-term treatment of bipolar disorder: a systematic review and meta-analysis. J Psychopharmacol 2010;24:17291738.
  • 22
    Calabrese JR, Rapport DJ, Shelton MD, Kimmel SE. Evolving methodologies in bipolar disorder maintenance research. Br J Psychiatry 2001;178:S157S163.
  • 23
    Bowden CL, Swann AC, Calabrese JR et al. Maintenance clinical trials in bipolar disorder: design implications of the divalproex-lithium-placebo study. Psychopharmacol Bull 1997;33:693699.
  • 24
    Geddes JR, Goodwin GM, Rendell J et al. Lithium plus valproate combination therapy versus monotherapy for relapse prevention in bipolar I disorder (BALANCE): a randomised open-label trial. Lancet 2010;375:385395.
  • 25
    Barbui C, Baschirotto C, Cipriani A. EMA must improve the quality of its clinical trial reports. BMJ 2011;342:d2291.
  • 26
    Bulloch AG, Bresee LC, Beck CA, Patten SB. Substantial changes in prescription recommendations for bipolar disorder in Canada: 2002–2010. Can J Psychiatry 2012;57:263268.
  • 27
    Alexander GC, Gallagher SA, Mascola A, Moloney RM, Stafford RS. Increasing off-label use of antipsychotic medications in the United States, 1995–2008. Pharmacoepidemiol Drug Saf 2011;20:177184.
  • 28
    Ehret MJ, Levin GM. Long-term use of atypical antipsychotics in bipolar disorder. Pharmacotherapy 2006;26:11341147.
  • 29
    Keck PE Jr, Calabrese JR, McIntyre RS et al. Aripiprazole monotherapy for maintenance therapy in bipolar I disorder: a 100-week, double-blind study versus placebo. J Clin Psychiatry 2007;68:14801491.