Are traditional approaches diminishing the bipolar drug pipeline?

KEYWORDS:

• Bipolar disorder
• bipolar inventory symptom scale (BISS)
• symptom domains
• multi-state outcome analysis of treatment (MOAT)

1. Introduction

Bipolar disorder is a chronic illness with profound and potentially devastating consequences. Social isolation, disrupted relationships, and financial instability are common sequelae. Effective treatment of the condition reduces suffering for those afflicted as well as their support system. Despite the existence of multiple medications that have been shown to benefit the condition, many struggle with residual symptoms that interfere with quality of life and increase the propensity for mood episode recurrence [1,2].

Although innovative compounds continue to be investigated, such as those that target synthetic glucose kinase-3 and inositol monophosphatase inhibition pathways, research is still in the early phases and promising results are limited [3]. Studies of new molecules to address bipolar symptoms have been constrained by numerous factors, including the substantial cost of drug research and development as well as the lack of specific etiological targets. Pharmaceutical executives understandably choose to pursue treatments for illnesses with identified genetic or biological targets as these are less risky investments [4].

In addition to these factors, psychiatric nosology and the continued use of traditional rating scales to assess treatment effectiveness interfere with the systematic assessment of novel treatments for bipolar disorder. A treatment that might benefit a specific component of bipolar symptomatology is unlikely to be tested with adequate sensitivity with traditional scales. Historically, bipolar disorder has been conceptualized as a cyclical illness consisting of manic, depressive and mixed episodes. Currently, the American guide defining diagnosis of psychiatric illness, the Diagnostic and Statistical Manual for Mental Disorders, DSM-5, defines major depressive episodes identically for bipolar and unipolar illnesses [5]. The consequent absence of distinguishing criteria other than the presence of a hypomanic or manic episode has led to frequent misdiagnosis. Not surprisingly, a prospective, naturalistic study found that 40% of patients with mood disorders were misdiagnosed as having unipolar major depressive disorder rather than bipolar disorder, leading to an overuse of antidepressants in this population [6]. Treatment with antidepressants in bipolar patients has been identified as an independent predictor of polarity changes (p < 0.001) and mixed symptoms (p = 0.01), whereas exposure to mood stabilizers is an independent predictor of time spent without symptoms (p = 0.019) [7]. While the role of antidepressants in the treatment of bipolar disorder remains a controversial and gray area, the effectiveness of mood stabilizers for decreasing illness symptomatology has been well established. Misdiagnosis of bipolar disorder as unipolar depression delays the initiation of first-line treatments for the condition. While using DSM-5 criteria can help differentiate between the conditions, incorporating illness course and family history into the diagnostic criteria may help providers distinguish the two conditions more reliably [8].

Traditional conceptualizations of the illness impacted which rating scales were chosen to assess treatment response and effectiveness during clinical trials. The Montgomery Asberg Depression Rating Scale (MADRS) and the Young Mania Rating Scale (YMRS) have been selected frequently as appropriate measures for assessing symptom severity, with the MADRS assessing depressive symptoms and the YMRS quantifying manic symptoms [9,10]. However, continued use of these scales restricts new drug development, consequent to lacking items that support principal component analyses, or identifying treatments that might benefit a particular illness element such as anxiety or irritability.

Medications can be effective for some symptom domains, while leaving others untouched. Evidence indicates that multiple mood stabilizers have a relatively selective efficacy on components of bipolar conditions. For example, lithium and divalproex are both superior to placebo in reducing hyperactivity, but only divalproex reduces irritability [11]. Lamotrigine benefits sadness, depressive cognitions, and psychomotor slowing, but lacks specific antimanic properties [12]. In the aggregate, these studies indicate that best treatment outcomes usually involve multiple drugs to address the multiple symptom domains present.

In a recent review of 26 published reports, multiple precursors were identified for bipolar disorder, including mood lability, psychosis, and subsyndromal mood symptoms [13], suggesting that an accurate scale for assessing bipolar symptoms would include assessment of multiple illness components in addition to manic and depressive symptoms. The development of the Bipolar Inventory of Symptoms Scale (BISS) was intended to provide one scale that addressed the full spectrum of symptoms experienced by persons with mood disorders [14,15]. The dimensions of disturbed behavior are divided into the following domains: mania, depression, anxiety, irritability, and psychosis. By including all these domains, the BISS identifies symptoms that may go unrecognized by clinicians and unquantified by other scales. As reported in Acta Scandinavica, symptom subsets measured by the BISS help distinguish mixed episodes from manic/hypomanic episodes and depressive episodes [16].

Approximately 40% of patients with bipolar disorder struggle with a comorbid anxiety condition [17]. In addition, many have an anxiety component to their bipolar illness that remits when mood symptoms are controlled without the use of antidepressants or anxiolytics. Most anxiety items measured by the BISS were rated as significantly more severe in patients with mixed symptoms compared to either mania/hypomania or depression [16]. Bipolar patients with high anxiety levels tend to experience more time in depressive episodes than in manic or hypomanic episodes [18]. No medications are currently approved for treatment of anxiety associated with bipolar disorder. Consequently, an opportunity exists to reduce illness burden, improve quality of life, and decrease treatment resistance by addressing this domain.

Mood lability, also called affective lability, is another symptom recognized and incorporated into the BISS which has yet to be identified as a specific medication target. Many bipolar patients continue with labile mood despite depressive or manic symptom remission. Such lability often interferes with functioning as well as with interpersonal relationships. Developing a medication to target this symptom could substantially benefit many bipolar patients who have this impairing adverse effect.

At the present time, several bipolar symptoms are best approached by overall patient management of habits and routine. The prime example is inadequacy of sleep. The bipolar patient tends to stay up late, and manic states are particularly associated with sleep disruption. This, coupled with the consequences of sleeping later in the morning or napping, promotes a cycle of disrupted sleep. Associated symptoms include daytime fatigue and decreased productivity. Treatment that focuses on re-regulation of circadian rhythms rather than solely on increasing sleep promotes compliance, health, and improved function in bipolar patients. Family and friend support are frequently helpful for reestablishing a consistent and structured sleep schedule. Lithium and valproate influence circadian rhythms in ways that may promote mood stability [19–21]. Exploring the use of other medications that impact circadian rhythm could benefit patients with bipolar disorder, although this is not a traditional perspective for illness management.

Survival analysis techniques dominate the analysis of treatment outcomes for studies of bipolar disorder, as they do in other areas of medicine. Multi-state outcome analysis of treatment (MOAT) is an additional tool that has been developed which is particularly useful for evaluating treatment outcomes in chronic illnesses, such as bipolar disorder, where wellness is rare [22]. MOAT facilitates evaluating responses to treatment during discrete periods of time, allowing for the assessment of treatment effectiveness during subsyndromal periods. This technique also integrates tolerability into assessments, which adds a valuable dimension to data analysis. While statistical methods exist for the study of multiple events with survival analyses, these yield a series of group averages, or snapshots. In contrast, MOAT is able to look at multiple variables at the level of the individual.

2. Conclusion

In conclusion, at present, a substantial number of bipolar symptoms remain unaddressed despite the availability of multiple approved medications for the condition. While assessing depressive and manic symptoms is important for assessing treatment response, these are not the only relevant symptom domains to discuss with patients. Evaluating for the presence and severity of anxiety, mood lability, and sleep disruption is also essential for a complete understanding of whether remission has been achieved. The BISS provides a comprehensive assessment of bipolar disorder which quantifies symptoms in five domains: mania, depression, psychosis, irritability, and anxiety. The scale identifies clinically common symptoms that are frequently unrecognized, especially during subsyndromal periods. Applying MOAT to data analysis can identify helpful and tolerable treatments for subsyndromal and syndromal periods. Utilizing the BISS and MOAT to systematically explore the effectiveness of medications may yield novel treatments for patients with bipolar disorder by detecting tolerable and effective treatments for previously unaddressed symptoms.

3. Expert opinion

Psychopharmacological approaches for treatment of bipolar disorder remain largely embedded in paradigms developed over half a century ago. These procedures enrich protocols to favor the drug of interest over a comparator or placebo. Inclusion and exclusion criteria frequently utilized for trials limited study generalizability as comorbid conditions and depressive symptoms of insufficient severity were often excluded [23]. In addition, high dropout rates often handicap studies’ validity and generalizability.

Traditional assessment tools for bipolar disorder fail to adequately evaluate all symptom domains pertinent to the illness. While systematizing the assessment of affective illnesses has been helpful for research purposes, patients spend a substantial amount of time suffering from mood symptoms that are not captured with historical manic or depressive rating scales. Adapting a categorical approach has consequently limited study generalizability and the identification of potential helpful treatments.

In contrast, a more comprehensive assessment of bipolar symptoms should consider all common domains, also called dimensional approaches, which focus on symptom myriads. Considering multiple domains as well as subsyndromal symptoms rather than just the presence or absence of specific mood episodes provides a more complete understanding of illness status as pathognomonic mood symptoms can decrease in severity but persist, or remit while associated symptoms such as anxiety, sleep disruptions, and mood lability remain. For example, patients can present primarily with anxiety and mood lability rather than depression or euphoria. Excluding them from research studies because the full criteria for a mood episode are not met reduces the chances of identifying helpful treatments for a large portion of the patient population. Discerning between mood-related anxiety and a comorbid anxiety illness can take time; however, the distinction can be important as it can alter treatment approach. Future approaches for evaluating bipolar disorder may thoughtfully incorporate assessments that include evaluating for the multiple domains that are part of the bipolar illness, especially anxiety.

Additional effective treatments for symptoms of bipolar disorder are sorely needed. Investigating treatments that target symptom domains rather than mood episodes would be a novel approach to drug development and a definite departure from previous models. Conceptualizing bipolar disorder in this way creates opportunities for drug development, as more targets for treatment are identified. Data analysis techniques that allow clinical states to be entered multiple times, such as MOAT, should provide scientific advantages to previously used analytic approaches as they would help identify whether treatments are useful during subsyndromal as well as syndromal states. Additional variables such as influence of gender and comorbidity on treatment response might also be important to consider [24]. Through the investigative process, more will be learned about the illness, and the information gained will further the development of effective treatments.

Declaration of Interest

C Bowden is a Scientific Director for Biomedical Development Corp. and has acted as a consultant for Myriad Genetics Inc through the Biomedical Development Corp. 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. One referee declares having received grants/research support, consulting fees and honoraria within the last three years from Gedeon Richter, Genericon, Janssen-Cilag, Lundbeck, Otsuka, Pfizer Inc and Servier.

Additional information

Funding

C Bowden is a Scientific Director for Biomedical Development Corp. and C Bowden has received funding from the National Institute of Mental Health via grants 5P30MH086045, 2/R24MH091997 and STTR/R24MH091997 (subcontracted with the Biomedical Development Corp). He has also received funding from Myriad Genetics Inc. M Martinez has received funding from Myriad Genetics Inc as well as well from the National Institute of Mental Health via grants STTR/R24MH091997 (subcontracted with the Biomedical Development Corp) and 5P30MH086045. She has also received funding from Methodist Healthcare Ministries via the South Texas Access to Recovery grant.