Should bipolar disorder treatment be modified depending on staging?

KEYWORDS:

• Staging
• bipolar disorder
• functional impairment
• treatment refractoriness
• cognitive impairment

1. Introduction

Current treatment guidelines in bipolar disorder focus mainly on the prevention of recurrence and stabilization of acute mood episodes while neglecting outcomes related to the longitudinal course of illness [1]. These unfavorable outcomes include refractoriness, functional, and neurocognitive impairment [2]. It is known that 42% of euthymic patients show functional impairment [3]. Recent studies also have shown that the percentage of patients with bipolar disorder and clinically significant neurocognitive impairments may vary from 30% to 62% [2]. In addition, nonresponse in bipolar depression, for instance, occurs in 40% of patients [4]. In view of that, bipolar disorder has been conceived in terms of differential stages by some researchers in order to tailor treatments according to patients’ needs [5,6]. Staging models imply that natural history of bipolar disorder moves through a predictable temporal progression, and provision of stage-appropriate treatment can modify this course. Some medical specialties such as oncology and cardiology have advanced in this field, developing strategies to prevent unfavorable outcomes.

In 2007, Berk and colleagues reported a staging model for bipolar disorder [6]. Subjects at stage 0 are those at increased risk of bipolar disorder (e.g. family history). Subjects with prodromal symptoms of bipolar disorder were classified at stage 1. Those who had the first mood episode were classified in stage 2. Subjects with recurrent mood episodes were classified at stage 3, whereas patients with unremitting symptoms were classified at stage 4. The rationale for this model is that early intervention is likely to be associated with a better response to treatment [6]. In this same vein, Kapczinski and colleagues reported a model that emphasizes the assessment of patients in the inter-episodic period: latent phase – subjects who present family history and anxiety and/or mood symptoms; stage I – bipolar disorder patients who present well-established periods of euthymia and absence of overt psychiatric morbidity between episodes; stage II – patients who present symptoms in between episodes, mostly due to axis I or II comorbidities; stage III – patients who present a clinically relevant pattern of neurocognitive and functioning impairment; and stage IV – patients who are unable to live autonomously due to functional impairment related to bipolar disorder [5]. Authors also suggested that patients at late stages are more likely to present altered brain scans and biomarkers [5]. The progression across these presentations could be determined by the cumulative exposure to mood episodes, trauma, drug abuse, and inherited vulnerability [5]. As a consequence of these early descriptions, the notion of functional staging emerged in the field of bipolar disorder. A study showed a strong linear association between The Functioning Assessment Short Test (FAST) scores and the clinical stages described by Kapczinski [5], suggesting a progressive functional decline from early stages through late-stage of bipolar disorder [7]. Emerging evidence suggests that correlates of staging such as number of mood episodes, functioning or neurocognition, could be potential moderators of outcomes. Therefore, the present editorial aims to review the literature related to these clinical correlates of staging.

2. Treatment refractoriness

One interesting hypothesis of staging models is that response to treatment in the early stages is more favorable than in later stages. It was suggested that recurrent mood episodes could induce treatment refractoriness [8]. Therefore, number of episodes was the clinical correlate assessed in studies of treatment refractoriness in patients with bipolar disorder. For instance, patients with an increased number of episodes have a worse response to lithium during manic episodes [9]. Similarly, a clinical trial of cognitive behavioral therapy showed that only patients who had fewer than 12 episodes of illness improved [10]. Moreover, a study showed that response rates for olanzapine in acute mood episodes or during the maintenance phase were significantly higher among individuals with less than five previous episodes [11]. In addition, a study with thirty-five treatment-resistant patients with bipolar disorder reported that response to lamotrigine monotherapy was associated with fewer hospitalizations [12]. Furthermore, a 5-year follow-up post hoc analysis study showed that patients with more than seven episodes at study entry did not show any significant improvement with psychoeducation according to time to recurrence [13]. Likewise, a post hoc analysis from a 15-month randomized controlled trial reported that psychoeducation for caregiver in the prophylaxis of mood episodes appeared beneficial only in patients at early stages [14]. Last, a cross-sectional study reported that monotherapy was more frequent in stage I, and combination of two drugs in stage II, while patients at stages III and IV needed either clozapine or a combination of more than three medications [15].

However, a number of studies in patients with bipolar disorder did not find such effect. For instance, studies showed no association between frequency of mood episodes and poorer responses to lithium in patients with bipolar disorder [16–18].

3. Functioning and neurocognitive impairment

Although available treatments may treat and prevent mood episodes, they are less effective in recovering functioning or neurocognition. For instance, a study showed that while almost all patients with first-episode mania had recovery from the syndrome within 2 years, only one-third of them achieved functional recovery [19]. A functional remediation therapy was recently developed in order to address this functional impairment [20]. A multicenter randomized controlled trial with 239 outpatients with bipolar disorder showed that functional remediation presented significant efficacy in improving functional impairment as compared with treatment as usual for euthymic patients over 21 weeks of treatment [20]. Later on, the same authors showed that improvement in psychosocial functioning was maintained after 1-year follow-up [21]. In addition, a secondary analysis of data from the functional remediation trial showed that this intervention may also improve verbal memory and functioning in a sample of neurocognitively impaired bipolar patients at 6-month follow-up [22]. Another promising treatment for late-stage bipolar patients is clozapine [23]. A systematic review that included 15 clinical trials showed that clozapine was associated with improvement in social functioning [23]. It is worth mentioning that clozapine may be also effective for treatment-resistant patients with bipolar disorder [23–25]. Despite proven effectiveness in bipolar disorder, there are concerns regarding some significant side effects, including agranulocytosis, myocarditis, cognitive disturbances, and weight gain. These side effects continue to limit the use of clozapine in bipolar disorder [26].

The research efforts into new therapeutic interventions for neurocognitive impairment in patients with bipolar disorder have produced only preliminary results so far. A cross-over study showed that mifepristone improved spatial working memory in patients with bipolar disorder [27]. In addition, another clinical trial reported that erythropoietin was effective in sustained attention, facial expression recognition, and measures of executive function [28]. Furthermore, a small preliminary clinical trial showed that cognitive remediation improved memory encoding and retention [29], although a recent randomized controlled trial revealed no cognitive benefits of short-term cognitive remediation [30]. It is noteworthy that research into novel neurocognitive treatments for patients with bipolar disorder faces some methodological challenges. For instance, there is no consensus regarding which cognitive measures should be assessed to track treatment efficacy or whether included patients can have current mood symptoms or should be euthymic [31].

4. Conclusions

It appears that the use of staging systems in psychiatry is not at the point where it could be applied clinically. The predictive validity of staging models is still in its infancy and needs longitudinal validation. It is recognized that heterogeneity is a major issue among patients with bipolar disorder and the assumption that patients move through stages in linear fashion is not supported by data [2]. However, there is a very productive debate in the area of staging in bipolar disorder that will surely yield benefits for patients and the field as a whole [32–34]. In addition, retrospective studies present important limitations to assess number of episodes. It is also unclear whether length or severity of mood episodes can moderate the impact of the number of mood episodes in treatment refractoriness.

In summary, we are of the opinion that treatments such as psychoeducation, lithium monotherapy, and cognitive behavior therapy seem to be better suited to prevent recurrences within patients with fewer episodes. Late-stage patients with higher functional impairment may benefit from complex pharmacological strategies, clozapine, and functional remediation. However, current staging models are not yet ready for clinical use. Future studies using new analytic approaches and multimodal data may help to determine the usefulness of staging models in bipolar disorder.

Declaration of interest

F. Kapczinski has received grants or research support from AstraZeneca, Eli Lilly, Janssen-Cilag, Servier, NARSAD, and the Stanley Medical Research Institute; has been a member of speakers’ boards for AstraZeneca, Eli Lilly, Janssen, and Servier; and has served as a consultant for Servier. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Acknowledgments

The authors would like to thank the Graduation Program in Psychiatry at Universidade Federal do Rio Grande do Sul (UFRGS).

Additional information

Funding

This paper was not funded.