Commentary on N. Ghaemi's “Hippocratic Psychopharmacology of Bipolar Disorder” Treating Bipolar Disorder

Abstract

Objective: This is a commentary on Hippocratic Psychopharmacology in Bipolar Disorder, an article by S. Nasir Ghaemi in this issue of Psychiatry 2006. Design: Based on Dr. Ghaemi's article and relevant literature, I discuss implications of Hippocratic treatment, i.e., the principle of treating an underlying illness using methods that enhance the patient's adaptive responses, rather than using symptomatic treatments. Results and Discussion: Key points include 1) the course of illness, as heterogeneous as it is, seems to have episodic-stable and inherently unstable forms; 2) course of illness interacts with episode characteristics, with mixed or polyphasic episodes associated with an unstable and complicated course of illness; 3) consequences in terms of treatment response, with lithium being more effective in treating the episodic-stable than the unstable form of the illness; and 4) the fact that, in determining treatment response, course of illness trumps episode characteristics. The goal of Hippocratic/Oslerian medicine, curative treatment aimed at underlying mechanisms of disease, is the aim of treatment, but is still elusive.

Most of us physicians may generally get through the day as well as we can, without thinking critically about the assumptions and models that underlie our treatments or even our research. Dr. Ghaemi's article on Hippocratic psychopharmacology is a refreshing and critical view of the conceptual bases of modern psychiatric treatment. It should be required reading for those who treat psychiatric patients or who are learning to do so.

The basic principles of Hippocratic psychopharmacology, as formulated by Dr. Ghaemi, are that treatment should be aimed at the underlying illness rather than its symptoms and that treatment should work in harmony with the patient's endogenous adaptive mechanisms.¹ Dr. Ghaemi proposes that the basis of bipolar disorder is recurrence rather than episodes, so Hippocratic treatment should focus on recurrence.

Based on these principles, Dr. Ghaemi presents a critical and useful view of potentially useful treatments. The critical evaluation of the existing data, distinguishing true prophylaxis from relapse prevention and from withdrawal syndromes, is much needed. The succinct review and critical discussion of underlying evidence have great practical value, complementing the theoretical interest of the paper. I will discuss the following three consequences of Dr. Ghaemi's analysis: Interactions between the course of illness and characteristics of episodes, heterogeneity in the course of illness, and relationships between course of illness and effectiveness of treatments.

Course of Illness and Episode Characteristics

I think it is vital to distinguish the mechanisms of episodes, which appear to be medically nonspecific syndromes, from the underlying illness that generates them. This underlying illness, however, may influence characteristics of episodes. Patients with episodes that are mixed, combining depressive and manic features,² or polyphasic, passing between predominately depressive and manic states without resolution,³ appear to have a worse course of illness than do those with simple depressive or manic episodes. This is manifested by earlier onset of illness, more frequent episodes, and higher incidence of severe complications like suicidal behavior and substance abuse, in patients with mixed or polyphasic episodes.^2–4

Heterogeneous Course of Illness

There are at least two dimensions of variability in the course of bipolar disorder. First, while the course of illness is heterogeneous, it can be formulated as falling into two basic types: An episodic-stable course, with pure manic or depressive episodes that are relatively infrequent, and an inherently unstable course, with more frequent episodes that can be mixed or polyphasic.³ Accordingly, severity of recurrence, rather than of depressive or manic symptoms in an index episode, predicts poor functional outcome.^5,6

Second, some patients may experience episodes that are predominately depressive, while others may have mostly manic episodes.⁷ This may account for the finding that, in maintenance treatment studies, the course of illness differs between index depressive and manic episodes, with recurrences more likely to be the same as the index episode.⁸

Treatment Consequences

Both episode and illness course characteristics may affect acute or chronic treatment response, but course of illness may trump episode characteristics. Lithium appears less effective in treating mixed than pure manic episodes.^9,10 Lithium also appears less effective in patients whose course is unstable, with acute¹¹ and chronic¹² responses reduced in patients with many previous episodes or with histories of complications such as substance abuse.¹³ Even patients in pure manic episodes have reduced response to lithium if they have many previous episodes¹¹ or histories of substance abuse.¹³

Preventive treatment responses may depend on characteristics of the index episode, which may itself be related to whether the patient has a predominately depressive or manic course.⁷ For example, lithium has been reported to protect against depressive episodes after an index depressive episode,^14,15 but not after an index manic episode.^16,17 Divalproex protected against depression after a manic episode¹⁷ but has never been studied after a depressive episode. Lamotrigine, on the other hand, appeared to protect against depression regardless of index episode characteristics but was less effective against relapse into mania.¹⁸

There are at least two competing models for effectiveness in prevention of episodes. First, as discussed by Ghaemi, it is vital to distinguish effects of withdrawal from true relapse prevention in order to understand when treatment really works. This point also brings home the importance of continuing effective treatment. Further, true prophylaxis, determined by randomization, after stabilization from an episode to treatments independent of the index episode is different from prevention as measured in continuation studies, determined by randomization to treatment including that of the index episode. As Ghaemi points out, prophylaxis is more generalizeable because it extends to patients regardless of initial treatment, while continuation applies only to those initially treated with the same drug. In real practice, however, continuation is the more relevant model, because it is most practical to be able to continue treatment with the same agent that made the patient better, when possible.

Effectiveness of treatment depends on the interaction between pharmacological and experiential components of treatment to increase the patient's adaptive mastery of the illness.¹ Both long-term impairment and treatment adherence are more strongly related to caregiver stress¹⁹ and attitudes of the patient and caregivers about the illness²⁰ than to severity of index episode symptoms.

What Remains

The Hippocratic/Oslerian model calls for curative treatment aimed at the mechanism underlying the illness. In the case of bipolar disorder, we have no objective markers of illness. Despite tantalizing possibilities, we also do not know what the underlying mechanisms are. We are lucky in having found potentially effective treatments. We know many effects of these treatments, like neuroprotective effects, blockade or stabilization of second messenger systems, modulation of ion fluxes, and prevention of behavioral sensitization or kindling. Knowing effects, however, is a long way from knowing mechanisms of action.

Finally, we have no treatments that are curative. Even after years of successful lithium maintenance, patients relapse when lithium treatment is discontinued.²¹

In summary, the Hippocratic model provides valuable insights into evaluating treatments for bipolar disorder. Ghaemi's article provides a much-needed reminder that our effectiveness as clinicians demands that we examine the historical, theoretical, and scientific bases of our treatments critically. With this firm basis, we will be able to help our patients with current knowledge and with advances that are sure to come.