Antipsychotics or Mood Stabilizers in Bipolar Disorder: Towards Evidence-Based Personalised Medicine

Lintunen et al. [1] publish in previous issue an article entitled Dosing Levels of Antipsychotics and Mood Stabilizers in Bipolar Disorder: A Nationwide Cohort Study on Relapse Risk and Treatment Safety. This nationwide study estimates doses of antipsychotics and mood stabilizers associated with the most favourable benefit–risk ratio. Benefit corresponded to a decreased risk of psychiatric hospitalization (prevention of relapse) and risk to an increase in non-psychiatric hospitalization (adverse events). The authors followed individuals with bipolar disorder from diagnosis over an average of 8 years. They compared outcomes over periods with and without antipsychotics or with and without mood stabilizers within individuals, by distinguishing low (< 0.9 DDD), standard (0.9– < 1.1 DDD) and high doses (≥ 1.1 DDD). Only monotherapies and individuals with both treatment changes and outcomes contributed to the findings. This design might have selected individuals with most severe disorders or those who did not receive an effective medication on a first line of treatment, but allowed comparing various treatment patterns.

Considering sensitivity analyses that omitted the 30-day period following treatment changes and selected stable treatments, among antipsychotics, only low and standard doses of aripiprazole (< 16.5 mg/day) were able to prevent relapse. High doses and quetiapine at any dose were associated with an increase in psychiatric hospitalization. While the association between high doses and relapse might be due to confounding by indication (relapse justifying the increase in dose), the absence of preventive effectiveness of antipsychotic monotherapies is alarming and contrasts with their extensive use [2]. Previous publications highlighted the lack of evidence of efficacy of antipsychotics in the maintenance treatment of bipolar disorders, RCTs showing selection bias (enrichment design limiting generalizability, inclusion of bipolar disorder type I only), attrition bias (considerable dropout levels), insufficient duration to demonstrate preventive efficacy, possible adverse effects of abrupt medication discontinuation in the placebo-group with beneficial effects of treatment and possible reporting bias [3, 4]. Parallelly, Lintunen et al. [1] found an increased risk of non-psychiatric hospitalization except for standard doses of olanzapine, risperidone and aripiprazole and low dose of aripiprazole, questioning the benefit–risk ratio of these monotherapies. These safety concerns are added to previous ones concerning mortality or cognitive functioning [2, 5, 6]. A real utility of antipsychotics was shown at short- and mid-term in acute bipolar episodes and in association with mood stabilizers with synergistic effects [7, 8]. Their place in the therapeutic strategy might be re-thought and, for example, re-focused on acute episodes and patients with disorders refractory to conventional treatment, in association with mood stabilizers.

Regarding mood stabilizers, as for antipsychotics, sensitivity analyses showed higher risk of relapse for high dose of valproate (≥ 1100 mg/day) and lamotrigine (≥ 220) but not of lithium (≥ 990) or carbamazepine (≥ 440). Lower doses of the four mood stabilizers prevented psychiatric hospitalizations. Safety concerns were associated with any dose of valproate, high doses of carbamazepine and lamotrigine. Conversely, low and standard doses of lithium were associated with a decreased risk of non-psychiatric hospitalization. Thus, lithium was shown effective on relapse without serious adverse event, in line with most international guidelines putting lithium at the first line of treatment. Despite the increasing evidence of its utility for patients, lithium is still underused in bipolar disorders [2]. This study is a new opportunity to highlight the effectiveness of lithium not only in bipolar disorders but also in the prevention of early mortality, suicidal behaviour, accidents and cognitive decline [9].

This favourable benefit–risk ratio may also reflect the interest of therapeutic drug monitoring that allows personalizing drug treatments, particularly in vulnerable individuals. Therapeutic drug monitoring is essential with lithium treatment and only recommended with other drugs [10]. Its use should be widespread. Globally, the study by Lintunen et al. [1] supports the use of the minimum effective dose for each drug in bipolar disorders to achieve a favourable benefit–risk ratio. This may be connected to studies on receptor occupancy showing few differences between low and high doses of antipsychotics [11]. Increasing the dose may not alter effectiveness, but may compromise safety.

Prevention is central to the treatment of bipolar disorders. Patients face not only a high risk of recurrence but also its consequences, such as early mortality, suicidal behaviour, persistence of residual symptoms and poorer psychosocial functioning. Medication is a key component of prevention. When initiating a treatment for acute episodes, prescribers should already think about the benefit–risk ratio with a view to long-term treatment. Efficacy is not the only parameter, tolerance being a determining factor in quality of life and adherence to medication. Prescribers may be satisfied with the absence of symptoms, while patients' expectations may differ. Drugs' assessment must take account of the user's opinion and consider patient-centred outcomes. For example, the main reasons for discontinuation reported by individuals with bipolar disorder include weight gain, lethargy, sleepiness, sexual dysfunction, shaking and emotional blunting [12]. Regarding chronic disorders, improving quality of life and functioning should be at the forefront of care. Future research will certainly have to continue this work of comparing existing therapeutic alternatives under real-life conditions and as close as possible to the experience of users. Better use of existing resources is a major factor in progress.

The author declares no conflicts of interest.