The PAX-BD trial does not provide a definitive answer, but it does provide some support for the hypothesis that pramipexole may treat TRBD. However, it is not a simple treatment to use. Dose titration was complex and there were errors made by participants and clinicians. There also appears to be a risk of hypomanic/mania early on in treatment, though this may be partially ameliorated by coadministration of an antipsychotic. The treatment does appear to improve not only mood but also psychosocial functioning and quality of life and is probably a cost-effective treatment. All of these conclusions must be considered extremely tentative given the small sample size and variable length of follow-up of participants due to the study being closed prior to achieving its target sample size.
The nature of the commissioned call (examining a drug that would always be extremely challenging to implement into routine clinical practice), requirements laid down in the call (that participants must not be on an antipsychotic when randomised) and then the emergence of the COVID-19 pandemic produced a perfect storm of issue hampering the prosecution of the PAX-BD study. Nevertheless, it was demonstrated that it was possible to successfully put into place systems to allow a study of a drug associated with several complex issues around dosing and monitoring to run largely remotely. In addition, a cogent scientific argument that allowing antipsychotic prescription alongside pramipexole was accepted by funder and ethics committee, with very tentative observations that this did not fully prevent the beneficial effects of pramipexole but may have reduced the occurrence of adverse effects. Unfortunately, the timing of approval of this amendment was not till during the pandemic which impacted on the ability to recruit.
Extensive PPI input to the application to proposed extension and completion of the trial. When the application was rejected, there was a great deal of unhappiness and feeling that the PPI voice was not being listened to.
The continuation of the PAX-BD study was considered by the funder due to the initial contract coming to an end, with the study not being as progressed as would be expected at that point, not least because of the period when the study was suspended along with all non-COVID-related research during the pandemic. Unfortunately, the request for a funded extension to complete the study with a revised target sample size of 100 was rejected, with no right of appeal. The results of the study demonstrate that with a short extension, the study could have been successfully completed. Data obtained suggest that in fact a sample size of just 90 would have been sufficient, with the study being 40% of the way to this target and recruiting 4–5 participants per month. As a result, the opportunity to obtain more definitive data regarding the efficacy and safety of pramipexole in TRBD was lost. The findings in the study confirm the concerns raised by the applicants when they first applied for funding for a study of lurasidone rather than pramipexole, in that it would be difficult to implement the treatment in routine care. However, this does not negate the fact that a definitive result regarding pramipexole would have had important implications for our understanding of BD and its treatment.
Interpretation
As things stand, the findings of the PAX-BD trial are insufficient to recommend changes in routine clinical practice. Where pramipexole is being used off-licence for the management of BD, for example in highly specialised, tertiary level, services, observations from the PAX-BD study tentatively suggest that clinicians, in addition to all of the usual considerations in using a drug with a limited evidence base off-licence, should consider the following:
- It may take time for the beneficial effects of the drug to be seen – many of the beneficial effects seen in the trial were not evident until 36 weeks post randomisation.
- Increases in the ability to experience pleasure may proceed improvements in mood.
- Pramipexole does not seem to exacerbate comorbid anxiety, but this may improve with improvement in mood.
- There appears to be a significant risk of increased hypomanic and manic symptoms with treatment, particularly in the first 6–12 weeks of treatment and great care is recommended in terms of patient monitoring at this time.
- Coadministration of an antidepressant along with pramipexole may be possible. Consideration of the precise pharmacology and dose of the antipsychotic used seems parsimonious.
- It is important to be vigilant for the emergence, or exacerbation, of possible impulse control problems, identification of which can be problematic given trait impulsivity associated with bipolar disorder and overlap with hypomanic symptomatology.
- While complex, the dosing schedule used in PAX-BD may be an option to consider with patients. In the trial, if the dose was reduced due to tolerability issues in the titration phase, then it was not re-increased. In clinical practice, it may be possible to attempt this, as previously described.27
The inability to persuade the NIHR HTA panel to fund a study of the more easily implementable drug lurasidone in the first place, and not to challenge the requirement regarding antipsychotics in the commissioned call for a study of pramipexole, must be laid squarely at the door of the CI and applicants. However, the circumstances around the commissioned call that led to the PAX-BD do raise questions regarding the nature of how decisions are reached as to what the subject of commissioned calls should be, and the detail specified in them.
It is possible to run a trial, including of a drug that is complex to use, largely remotely. One significant challenge that relates to such trials in mental health is that there are only a relatively small number of NHS Trusts and health boards that have expertise in the conduct of CTIMPs. As a result, significant input is required to support less experienced sites to build up their expertise. This takes significant resources but has the potential to increase the UK’s capacity to undertake CTIMPs in mental health.
Recommendations for research
Aside from repeating the trial and achieving an adequate sample size, there are a number of avenues for further research. These include:
- Undertaking a meta-analysis of the now three small RCTs undertaken of pramipexole in patients with BD.
- Investigating the impact of coadministration of an antipsychotic alongside pramipexole both in terms of efficacy and safety.
- Further exploration of the mechanisms of effect of pramipexole in improving mood, including the possible role of an early impact on the ability to experience pleasure.