Steven Dubovsky, MD

This Special Report covers some important aspects of the treatment of mood disorders. All of us who treat patients with these conditions keep hoping that the last word is finally in on mood disorders. On the other hand, if treatment of complex illnesses could be readily condensed to simple algorithms, perhaps anyone could do it. Let’s see if examining some of the more popular elements of conventional wisdom helps us keep our perspective.

Thanks to straightforward diagnoses, treatment options are clear. The categorical mood disorder diagnoses on which we base our treatment plans have a number of subtypes, but these do not begin to describe the true clinical range of affective syndromes. Within each of the broad categories (eg, bipolar I, bipolar II, rapid cycling, psychotic depression) are numerous dimensions of illness that influence course and response. Within any category, some illnesses have an early onset, and some start later in life. Some are highly recurrent, some are chronic, and others are sporadic; some are familial, and some are associated with cognitive changes, trauma, or substance abuse.

For some patients, the change in mood is the main problem, while in others personality traits coalesce around a pervasive negative affective orientation. Some patients with mood disorders have a robust response to lithium, an antidepressant, or psychotherapy; others require combinations of somatic therapies along with individual and conjoint psychotherapy. Yet there are very few gold standard studies in which specific features of illness are correlated with treatment response.

A diagnosis of cancer does not begin to convey the exact nature of the illness or the therapy that is needed. Oncologists need to know not only which organ the malignancy started in, its cellular and molecular pathology, its degree of extension, and its genetic profile, but also the stage of disease. As cancer progresses, it changes fundamentally, and as its pathophysiology becomes more complex, different treatments with different mechanisms of action become necessary, often in combination.

Mood disorders can be equally progressive, with the additional complication that psychology and the social environment change as much as physiology with continued illness. In both cases, a diagnosis made at one time is not the same as the diagnosis later in the course of illness. As in other medical specialties, the notion of a static diagnosis and an unchanging treatment plan is a deception.

We finally have evidence-based treatment approaches. Recent years have seen a proliferation of evidence-based practice guidelines and expert opinions, some of which even agree with each other. But what is the nature of the evidence?

With the exception of a few studies (eg, STEP-BD, STAR*D), the vast majority of clinical trials in mood disorders have been sponsored by industry, with the primary goal of getting regulatory approval or permission to advertise for a new indication for an approved product.

Such studies are not designed to investigate the best treatment for patients with a particular disorder, and certainly not for patients with specific features, but to demonstrate that the sponsor’s product is statistically superior to placebo in producing a response (usually defined as 50% reduction in a rating scale score). In these studies, remission is a secondary outcome defined as a low depression or mania rating scale score (but not 0), without regard to functioning. A positive study means that patients are half as depressed or manic as when they started, and in the case of remission, they are about two-thirds better. Since any residual symptoms, including social or occupational impairment, predict later major relapse, few patients who do well in clinical trials can be said to be truly recovered.

To maximize the chance of a positive finding from an extremely expensive multicenter clinical trial, sponsors want to enroll patients with a high likelihood of response to their medication and a lower likelihood of getting better with placebo. This means that it is necessary to exclude patients with mood disorders that manifest with psychosis, or that are too acute, too chronic, too severe, treatment-resistant, or associated with suicide risk or psychiatric or medical comorbidity, including substance abuse.

A fairly homogeneous sample precludes assessment of the impact of features such as stage of illness, family history, course, and dimensions of illness. Because patients in clinical trials resemble primary care patients rather than the average psychiatric patient, it is not possible to generalize results of these studies to most patients in the specialist’s practice. Moreover, because the outcome of interest is the number of patients who respond or the average rating scale score, nothing is known about which specific features predict a better response.

Because monotherapy or at most the addition of one agent to an established treatment is studied in patients with relatively uncomplicated illnesses, little is known about the combinations of treatments that are necessary for more complex and advanced mood disorders. The unending search for another blockbuster ($1 billion in sales) to offset the development costs of the many drugs that do not make a profit leads to attempts to prove that a single medication can treat everyone with a particular diagnosis, as well as patients with unrelated diagnoses. As a result, we have no information about how to combine a drug with other treatments in specific subtypes or for patients with truly refractory disorders.

Primary care physicians (PCPs) and nurse practitioners can make up for the shortage and cost of psychiatrists. The notions of stable diagnoses and straightforward treatments, along with reductions in health care financing, have led policymakers and insurers to claim that even if a psychiatrist is needed to make the initial diagnosis, nonpsychiatric practitioners can continue to monitor the treatment. Indeed, an academic leader in family medicine told me that psychiatry will soon be entirely replaced by family medicine.

It is true that 60% of depression treatment is provided by PCPs. However, these practitioners are best at treating depression that can be diagnosed with a brief questionnaire and is not severe or complex. Even for these cases, ongoing psychiatric oversight is necessary to ensure adherence to practice guidelines. Although PCPs are writing an increasing number of prescriptions for antipsychotic drugs and anticonvulsants, the ability of PCPs and nonphysician professionals to diagnose, let alone treat, bipolar disorder has repeatedly been questioned by empirical research. Until metastatic cancer can be treated in these settings, psychiatry will remain essential to the accurate diagnosis and effective treatment of all but the simplest cases of depression and the majority of cases of bipolar disorder.

What is the future of psychiatric assessment and treatment of mood disorders? Research is beginning to identify dimensions of illness and endophenotypes (objective, reproducible, discrete features that are more relevant to genetic markers and response to specific treatments than broad categorical diagnoses) that will enhance specific treatment planning. With research into mood disorders that have not responded to multiple aggressive treatment modalities, rather than the 1 or 2 treatment attempts for patients in studies of antipsychotic augmentation of antidepressants or transcranial magnetic stimulation, we will know more about how to cure truly refractory mood disorders.

We will ultimately know who requires which specific combinations of treatments at which stage of illness, and we will have as our goal complete functional as well as symptomatic recovery. This will enable us to prevent further affective episodes, but it will require more time, expertise, and teamwork than a 15-minute “med check” and referral to a nonpsychiatric professional can provide. In the meantime, we can amalgamate the advice in this issue with our own experience and that of those we trust and the emerging literature, and we can advocate vigorously for more rational funding of mental health care.

Dr Dubovsky is Professor and Chair of the department of psychiatry at the State University of New Yzork at Buffalo. He reports that he has received research support from Hoffmann-LaRoche, Lilly, Otsuka, Pfizer, and Takeda. 

Articles

What is the future of psychiatric assessment and treatment of mood disorders? The articles in this Special Report explore some important aspects and issues.

Steven Dubovsky, MD

Articles

Introduction: The Conventional Wisdom About Mood Disorders