Introduction

Scientific background

Bipolar disorder has a lifetime prevalence of 2.5%1 and is associated with an 8- to 12-year reduction in life expectancy.2 Current UK NHS care pathways are defined by National Institute for Health and Care Excellence (NICE) clinical guidelines for the management of bipolar disorder.3 Patients with bipolar are symptomatic around 50% of the time, the vast majority of which is depression,4,5 for which NICE guidelines list three treatments: lamotrigine, quetiapine and olanzapine (with or without fluoxetine),3 with the latter two poorly tolerated due to weight gain and sedation.6,7 Around 70% of currently depressed bipolar patients in the UK are on at least one antidepressant,8 despite evidence that they lack efficacy,3 demonstrating the clinical challenge posed and that many patients have treatment-resistant bipolar depression (TRBD). The prevalence of TRBD is unknown due to a lack of a consensus definition. However, around 50% of patients remain depressed at 6 months, and 30% at a year because of non-response, intolerance, or non-acceptance of treatment,9 and current evidence-based treatment options for bipolar depression (BD) are extremely limited.3,10 As a result, TRBD is the major contributor to the enormous burden of disease associated with bipolar disorder.

The potential role of pramipexole as a treatment for depressive episodes in bipolar disorder is supported by a number of lines of investigation, including preclinical investigations in animal models. Antidepressant-like effects of pramipexole have been observed in animal models of depression and models known to be responsive to drugs with antidepressant efficacy, such as stress-induced suppression of sucrose intake in rats,11 the forced swim test,12,13 social interaction test13 and olfactory bulbectomised rats.14 It has also been shown to increase hippocampal neurogenesis,13 an effect believed to be common to antidepressants.15,16 Pramipexole has extensive evidence for efficacy in Parkinson’s disease,17 for which it has a marketing licence. A meta-analysis of pramipexole in Parkinson’s disease reported improvement in depressive symptoms on the Unified Parkinson’s Disease Rating Scale (UPDRS).18 This led to a 12-week randomised double-blind placebo-controlled trial of pramipexole in patients with Parkinson’s disease and significant depressive symptoms which reported a significant benefit that was independent of any motor improvements seen.19 Such findings, together with hypothesised roles for a hypo-dopaminergic state underlying BD20 and naturalistic and open trial data,21–27 led to two randomised controlled trials (RCTs) in BD.28,29 Goldberg et al. studied 22 patients (mainly type I).28 All had failed to respond to at least two adequate trials of standard antidepressants during the current episode. They were treated with a mean of 1.7 mg/day [standard deviation (SD = 1.3)] pramipexole in combination with lithium or anticonvulsant mood stabilisers. (NB: Pramipexole doses as published in the literature refer to the salt form.) Mean improvement in 17-item Hamilton Depression Rating Scale (HDRS) scores at 6 weeks were 48% for pramipexole and 21% for placebo (p = 0.05). Zarate and colleague’s RCT included 21 patients with bipolar type II,29 who had failed at least one trial of a standard antidepressant. A mean dose of 1.7 mg/day (range 0.375–4.5 mg/day) pramipexole was given in combination with lithium or valproate. Sixty per cent of patients treated with pramipexole achieved a response [50% decrease in Montgomery–Åsberg Depression Rating Scale (MADRS)] at 6 weeks, compared with 9% taking placebo (p = 0.02).

Pramipexole pharmacology and issues regarding the coadministration with an antipsychotic

There is a potential pharmacodynamic interaction between pramipexole and antipsychotics. As a class, antipsychotics are dopamine receptor antagonists, while pramipexole is a dopamine agonist. There is a risk that concomitant treatment with an antipsychotic will block the effects of pramipexole.

The original trial brief from the National Institute for Health and Care Research (NIHR) Human Tissue Authority panel was that patients on antipsychotics should be excluded from the trial. However, antipsychotics are used extensively in BD. Data provided by the UK Prescribing Observatory for Mental Health (POMH) suggest that between 65% and 80% of patients with BD are prescribed at least one antipsychotic. The reason for such high rates is that they are used for many reasons. They are first-line treatment options for mania and some have an evidence base for efficacy in BD (quetiapine, olanzapine and lurasidone).10 Some antipsychotics also have an evidence base for long-term prophylaxis.10 However, in addition, antipsychotics are also used clinically to treat psychosis and for night-time sedation and anxiolysis. This explains why a patient with ongoing depression, despite being on an antipsychotic, might be reluctant to withdraw it for fear of insomnia and worsening anxiety, a concern raised by our patient and public involvement (PPI) group. This concern and the impact of excluding patients taking antipsychotics led to a protocol amendment to allow for the coadministration of trial medication with an antipsychotic based on a detailed understanding of the pharmacology of pramipexole and antipsychotics. This amendment was approved by the NIHR HTA panel, ethics and the Medicines and Healthcare products Regulatory Agency (MHRA). All regulatory approvals were in place by 1 April 2021.

Antipsychotics are a broad and diverse group of medications, including both oral and depot formulations. They vary with regard to their pharmacological activity but have a common feature of blocking dopamine D2 receptors and there is evidence for a close correlation between their D2 affinity and their clinical potency as antipsychotics.30 However, most bind to a range of additional receptors and uptake sites, including other dopamine receptors.

Pramipexole is a relatively ‘clean’ dopamine agonist, but it has activity at three dopamine receptors: D2, D3 and D4. The binding affinity of pramipexole is highest at the D3 receptor (Ki = 0.5 nM) – around an order of magnitude greater than at D2 (Ki = 3.9 nM) and D4 (Ki = 5.1 nM) receptors.31–33 D3 receptors have a distinct distribution in the brain with high levels in the limbic system, nucleus accumbens and the olfactory tubercles,34 brain areas critically involved in the regulation of motivation and reward.35,36 An impairment of the mesolimbic dopaminergic pathway may be responsible for anhedonia, one of the major symptoms of depression. D3 receptor knockout mice have been reported to exhibit depressive and anxious features.36 Positron emission tomography (PET) imaging data in humans suggest D3 receptor expression may be related to motivation for rewards.37 Such findings support a hypothesis that the mechanism of action of pramipexole in BD may be mediated via D3 receptors. This is supported by additional observations. For example, pramipexole induces a dose-dependent increase of dendritic arborisation and soma size, effects antagonised by selective D3 antagonists and mediated by brain-derived neurotrophic factor.38 This is of potential relevance given hypothesised associations between depression and impairments of neuroplasticity,39 known to be reversed by antidepressants, exercise and electroconvulsive therapy (ECT).40–42 Similarly, pramipexole has antidepressant effects on depression-like behaviour in mice induced by an inflammatory challenge while a D3 selective antagonist makes mice susceptible to depression-like effects.43 Depression and BD in men are associated with raised pro-inflammatory cytokines with this hypothesised to be of aetiological significance.44,45 In summary, the exact mechanism of pramipexole is not known. However, pramipexole binds preferentially to D3 over D2 receptors and there is evidence suggesting a key role of D3 receptors in treating depression.

Based on the receptor pharmacology of pramipexole and antipsychotics, and their hypothesised mechanism of actions, it is plausible that it may clinically be possible to capitalise on the D3 effects of pramipexole alongside the D2 antagonist effects of antipsychotics (along with their other pharmacological actions, e.g. the sedative effects of histamine receptor blockade) by combining pramipexole with an antipsychotic. Distinctly different effects of these drugs are suggested by the differential effect of the antipsychotic quetiapine and pramipexole on cerebral blood flow.46 There is no published prospective or randomised controlled data on the combination of pramipexole with antipsychotics in patients with affective disorders. Clinical evidence can be drawn from specialist centres (such as those run by the co-applicants: Young and Stokes in London; Geddes in Oxford; Morriss in Nottingham; McAllister-Williams and Watson in Newcastle and Smith in Edinburgh) where pramipexole is prescribed for patients with difficult-to-treat BD and depression, sometimes in combination with antipsychotics. In general, expert opinion suggests that the combination is well tolerated, perhaps because some of the potential adverse effects of pramipexole are blocked by the antipsychotic. Collating clinical experience, El-Mallakh et al.24 reported a retrospective naturalistic study of the effects of pramipexole in BD, finding a favourable effect on those who continued to take pramipexole for several months. Of the 16 patients in the report, 9 were prescribed antipsychotics (quetiapine n = 3, aripiprazole n = 3, risperidone n = 2, clozapine n = 1). The report does not detail whether those on antipsychotics were more or less likely to respond or experience side effects of the pramipexole. No cases of emergent mania were observed. When using pramipexole alongside an antipsychotic, clinicians tend to favour antipsychotics that have higher affinity for D2 versus D3 receptors, based on the hypothesis that pramipexole’s antidepressant effects are mediated via D3 agonism. Generally, doses of antipsychotic are kept to a minimum and titrated against pramipexole to maximise antidepressant effects while retaining the sedative or anxiolytic benefits from the antipsychotic. The aim is to maximise D3 occupancy by pramipexole while allowing for D2 occupancy of the antipsychotic.

Table 23 (see Appendix 1) describes the affinities of the more commonly used antipsychotics in the UK. As can be seen, antipsychotics vary in the ratio of their D2 : D3 affinity. While pramipexole has an affinity for D3 receptors nearly eight times that of D2 receptors, most antipsychotics have higher affinity for D2 rather than D3 receptors. Based on these data, a protocol amendment allowed for antipsychotics to be used in addition to study mediation in PAX-BD as long as the dose of the former was within specified limits (see Table 24 in the Appendix 1).

Rationale

Based on epidemiological data,47 over a 12-month period in excess of 1 million people in the UK will suffer from an episode of BD and for 95% of these individuals this will herald a lifelong relapsing and recurring illness. For the vast majority, this episode will be a depressive one,4,5 and at the end of a year 30% will remain ill because of non-response, intolerance or non-acceptance of treatment,9 at least in part due to the limited treatment options currently available for BD.3,10 This high prevalence and rate of treatment resistance accounts for the estimated annual UK bipolar disorder costs of £5.2B, with direct NHS costs of £342M, at 2010 prices.48,49 Patients are eligible for this trial if their current episode of BD has not responded to two adequate trials from the three NICE-recommended medications,3 or the additional evidence-based, though expensive, treatment with lurasidone.50,51 A ‘failed’ trial, in this study, is defined as a clinically determined inadequate response to an ‘adequate trial’ (at least 8 weeks for lamotrigine) at an ‘adequate dose’ (based on current guidance), or an inability to tolerate or accept treatment. This reflects the current point of clinical equipoise, supported by our PPI group due to the limited number of treatment options available for TRBD, and defines a point at which there is potential for significant change to clinical and cost trajectories.

The trial was intended to be an appropriately powered, pragmatic, RCT with economic evaluation. The primary outcome was set at 12 weeks post randomisation with a total of 52 weeks of follow-up. The trial was intended to determine efficacy and longer-term cost-effectiveness in a real-world design. The use of a validated,52 self-reported primary outcome measure, Quick Inventory of Depressive Symptoms, Self-Rated (QIDS-SR) has successful precedents,53–56 and allows a patient-centric assessment and experience, as well as critically facilitating the inclusion of a large and broad representative sample. The secondary outcome measures examine the broader potential for positive and negative impacts of treatment. Economic evaluation was intended to further allow consideration of the utility of this treatment regime, using the NHS/Personal Social Services (PSS) perspective preferred by NICE with secondary analyses incorporating wider societal costs.

The PAX-BD trial therefore had the potential to have a major impact on the burden of disease experienced by patients and UK health services. A positive finding of cost-effectiveness coupled with a favourable tolerability profile has the potential to impact on UK provision of care for patients with TRBD due to the current dearth of options available to clinicians. Additionally, any evidence of efficacy of pramipexole supports the potential role of pro-dopaminergic treatments for the management of BD.

Aims and objectives

Aim

The PAX-BD trial aimed to examine if pramipexole could be an effective treatment for patients with TRBD. To achieve this, the following objectives were set.

Primary objective

The PAX-BD trial aimed to evaluate the clinical effectiveness of pramipexole versus placebo alongside standard mood-stabilising medication, over 12 weeks, in the management of patients with TRBD. The objective was assessed with the primary outcome measure being the QIDS-SR, measured at 12 weeks and with scores in the two treatment arms compared using an analysis of covariance (ANCOVA) covarying for baseline scores.

Secondary objectives

Secondary efficacy objectives

• To examine the impact of pramipexole treatment on mood and anxiety symptoms for the duration of trial involvement, and pleasure symptoms over 12 weeks.
• To examine the impact of pramipexole on psychosocial function for the duration of trial involvement.

Secondary safety and acceptability objectives

• The trial examined known possible side effects of pramipexole: the risk of switching to mania and the occurrence of psychosis or impulse control disorders.
• Rates of impulsivity during treatment with pramipexole were examined.
• Side effects and overall acceptability of pramipexole treatment were examined.
• Tolerability of pramipexole was reviewed by reporting rates of adverse events (AEs), serious adverse events (SAEs) and suspected unexpected serious adverse reactions (SUSARs).
• Adherence to medication was also reviewed.

Health economic objectives

• To examine the quality of life, well-being, health and social care and broader societal costs of participants randomised to either pramipexole or placebo.

Comparison analysis objectives

• Content validity was increased by using gold-standard observer-rated scales to measure participant scores on mania and depression to compare participant scores with previous studies in the field.