This interview took place on October 25, 2007, and was conducted by Norman Sussman, MD.
This interview is also available as an audio PsychCastTM at http://psychcast.mblcommunications.com.
Disclosure: Dr. Thase is consultant to AstraZeneca, Bristol-Myers Squibb, Cephalon, Cyberonics, Eli Lilly, GlaxoSmithKline, Janssen, MedAvante, Neuronetics, Novartis, Organon, Sepracor, Shire US, Supernus, and Wyeth; is on the speaker’s bureaus of AstraZeneca, Bristol-Myers Squibb, Cyberonics, Eli Lilly, GlaxoSmithKline, Organon, sanofi-aventis, and Wyeth; has equity in MedAvant; and receives book royalties from American Psychiatric Publishing, Guilford Publications, and Herald House.
Dr. Thase is professor of psychiatry at the University of Pennsylvania School of Medicine in Philadelphia. He is a distinguished fellow of the American Psychiatric Association and a founding fellow of the Academy of Cognitive Therapy, a non-profit mental health organization dedicated to serving people with mental health disorders. He is director of the American Society of Clinical Psychopharmocology and vice chairman on the Scientific Advisory Board of the National Depression and Bipolar Support Alliance. Dr. Thase’s research centers on the evaluation and treatment of mood disorders.
What factors most commonly influence clinician perceptions of different antidepressants?
A clinician’s firsthand experiences shape how he or she views a medication’s efficacy and tolerability. The first 10 experiences involving the prescription of a new medication is usually what determines whether one values it as a well-tolerated or highly effective medication.
What was the methodology used in the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trials?
Approximately 4,000 patients relatively new to treatment were given citalopram, a standard first-line selective serotonin reuptake inhibitor (SSRI). Citalopram was chosen for several reasons. Because it is the least used of the common SSRIs, the probability of a patient having had experience with this treatment is low. In addition, changes in the marketplace emphasize generic SSRIs over patent-protected medication, and citalopram was to be the first generic SSRI. Furthermore, citalopram has relatively little effect on the cytochrome P450 system, is midline among the SSRIs in terms of discontinuation symptoms, and is easy for elderly patients to use. Patients who did not respond to or could not tolerate citalopram accepted randomization to at least two out of seven treatment options. Only 3% to 4% consented to all seven options.
Four switch strategies were available. The first involved a choice of three antidepressants, bupropion, venlafaxine, and sertraline. The fourth option was cognitive therapy with neither the original nor the alternative medications. The medication switch experiments elucidated whether or not switching out of the popular SSRI class yielded a better result than staying within it.
The augmentation options included adding bupropion, buspirone, or cognitive therapy to ongoing citalopram therapy. Because few patients consented to switching or adding medications, the efficacy of both multi-action antidepressants and different SSRI classes remains unknown.
More than two-thirds of the patients declined psychotherapy even though the cost of treatment within the STAR*D study was partially supported. Medications were provided free of charge. However, waiving the co-payment cost for psychotherapy was not possible for legal reasons.
The study was designed to reliably conclude that 15% differences between the various pair-wise comparisons of the treatments were real. We found many differences of 7% to 10% in the STAR*D study but no 15% differences. In retrospect, because so many sites were involved with the study, we should have conducted fewer comparisons with a larger number of patients. If that were the case, a few more statistically significant differences may have been found.
Based on clinical trials and an intent-to-treat analysis, a true effect is approximately a 35% remission rate with any one of the antidepressants. Could one argue that clinical practice yields a better response?
Of course. Drug administration in clinical practice is conducted more flexibly, and the main focus is on the patient getting better as opposed to clinical trials, in which the goal is to answer a particular study question. In a clinical practice, one can readily augment or manage a side effect outside of a protocol. A clinician is also more free to try psychotherapeutic interventions.
Could patients in the STAR*D trials receive augmentation with lithium?
Lithium did not do well in the STAR*D study. The dropout rate on lithium was higher than the proportion of patients who received it and use of blood-level monitoring was lower than expected. With lithium augmentation and, subsequently, with tranylcypromine, the STAR*D study exemplified how many doctors in practice today are not confident or comfortable with prescribing some of the older interventions.
Tricyclic antidepressants (TCAs) also were surprisingly softly prescribed in the STAR*D study. Fewer patients reached the maximum dosage than we expected, especially in a difficult-to-treat patient population.
Should findings in the STAR*D study change the way people approach the treatment of depression?
The findings from the STAR*D study illustrate that depression is a serious illness for which current treatments, if provided in a rigorous, iterative, step-by-step way, can get a majority of people better. The findings also illustrate that our first-line treatments are imperfect and that in current practice, standard remedies typically used as the third-, fourth-, and fifth-line treatments require a level of training and expertise that may be beyond many practitioners’ means.
Our house staff needs to know how to prescribe a TCA, lithium, and a monoamine oxidase inhibitor before they leave training because these three old-guard treatments underperformed in the STAR*D study. They were not used as well as they should have been.
What common misconceptions can be corrected about antidepressant selection and use?
There is a continuous tension between how clinicians match medications to patient symptoms. For example, most of us believe that some antidepressants are more effective for patients who oversleep whereas others are better for patients with significant insomnia. However, we cannot prove that this kind of matching actually exists or works in practice. Acknowledging that we do not always know how to pick the right medication for the right person is important.
Will any mechanistically unique treatments be available in the near future?
We may get a dual-reuptake inhibitor with a stronger norepinephrine effect or a triple-reuptake inhibitor people can tolerate that does not have abuse potential. However, for the most part, we are close to having the best reuptake inhibitors possible. That obligates us to look toward other mechanisms, which may yield exquisitely good effects for small percentages of patients who are not well treated with the current standards. For example, a medication for sleep that works through a novel melatonin mechanism may be better for sleep than any other antidepressant already available, even though it is not as effective for depression as other antidepressants.
An antidepressant that works through the stress-response system, such as a corticotropin-releasing hormone or a corticotropin-releasing factor antagonist, may treat a subset of people with stress-related depression in a safer, simpler way. However, such an antidepressant would be useless for people whose exaggerated stress response is not a destructive health issue.
The role of glutamate and medications that dampens the role of excitatory amino acids in longer-term negative consequences of depression may be an example of a treatment that is used in combination with other treatments, not to treat the mood symptoms but to lessen the consequences of protracted periods of depression.
What is most interesting about ketamine?
Ketamine has rapid mood-elevating properties. Related mechanisms do not simply offset the negative consequences of glutamate, but they may also have some interesting mood-elevating properties. If a drug mimicked ketamine, it would have major street value; it has great abuse potential. However, if we could get the rapidity of the effect without the intoxication and the hallucinogenic or illusionary effects, then the field would move forward substantially. Such a medication would be cautiously used lower in the treatment algorithm. A fair number of mood disorders experts already use psychostimulants, some more liberally than others. They have made peace with the fact that psychostimulants have potential abuse and street use. They are aware that difficult-to-treat patients get a palliative benefit from the psychostimulants’ energy, concentration, and motivation-enhancing effects.
What adverse effects are under-recognized or exaggerated to the point where people do not use drugs associated with them?
The original underrecognized side effects of SSRIs were inhibition of orgasm and potential for significant weight gain. Anorgasmia was partly overlooked because we were asking the questions neither correctly enough nor often enough. Weight gain was overlooked because it happened so insidiously. Weighing patients during each visit was not a standard part of our practice at that point. As for potential adverse reactions that have been exaggerated, the heightened awareness and concern about treatment-emergent suicidality was blown way beyond the evidence. The fact is that a small portion of people treated with antidepressants initially get worse instead of better. Eliciting or provoking unrecognized bipolar disorder may be the most important mechanism explaining why someone’s condition worsens rather than improves. If you think about a depressed person’s often negative cognitive set and add an activated, uncomfortable, driven agitation to it, then a 1% to 2% incidence of worsened suicidal ideology is not surprising.
How has reduced use of antidepressants affected suicide rates?
In mid-2007, the suicide rate among youths increased for the first time in more than a decade. This was during the time antidepressant-use in youths decreased. This is not necessarily a cause-and-effect relationship, but it is certainly suggestive that antidepressant treatment of depressed youths was saving lives.
Is the recent increase in bipolar depression diagnosis driven by drugs currently used for bipolar patients?
The availability of novel therapeutic mechanisms always leads to increased awareness of the possibility that a patient is bipolar. This happened in Europe during the 1950s and 1960s when lithium first became available. The very same thing happened in the United States during the early 1970s. Likewise, the diagnosis of schizophrenia increased when antipsychotics first became available in the US. When safer SSRIs came around, antidepressant-use increased, and family practitioners were trained to recognize depression because they were better able to treat it. Having said that, one should consider bipolar disorder when an antidepressant is not working, especially when it seems to be making the circumstances worse. Each patient with unipolar depression has a measurable risk for bipolarity on a stepwise factor scale. People can have either a low or a high chance of becoming bipolar. Screening for potential heritable risk factors like early age of onset, hypersomnolence, and hyperphagia both characterizes the depression and clarifies particular inherited variabilities. We may be able to recognize who is at a 50-50 risk of converting from unipolar disorder to bipolar disorder earlier, allowing more bipolar-specific treatments for such patients to be used earlier as well.
Does lamotrigine have acute antidepressant activity?
It does, but it has not been approved by the FDA for this purpose. Two unequivocal trials are necessary in order to receive an FDA indication as an antidepressant. Of the numerous properly done large bipolar trials, only one of those was clearly, unequivocally positive.1 However, the whole program of studies definitely shows a signal of antidepressant benefit.
In a sense, lamotrigine suffers in its efficacy because it lacks nonspecific benefits such as being strongly anxiolytic or improving sleep. The magnitude of the long-acting effect in lamotrigine bipolar disorder trials is only slightly below the magnitude of that of SSRIs in regular depression.
Do any of the atypical antipsychotics have more clearly demonstrated antidepressant effects than others?
Quetiapine has shown the best demonstrated antidepressant effect, as noted by two studies in which both 300 mg and 600 mg worked.2,3 In the large, major, pivotal trial, olanzapine also showed effect in bipolar I depression.4 However, the magnitude of that effect was much smaller than that of olanzapine and fluoxetine combination. If there had been a quetiapine-plus-antidepressant arm in the quetiapine studies, the same result may have been observed.
In all fairness, quetiapine’s magnitude of effect was somewhat larger than olanzapine’s magnitude of effect. However, there has never been, nor will there ever be, a head-to-head study of those two atypical antipsychotics in bipolar I depression.
Although aripiprazole did not show a significant antidepressant effect in the bipolar depression studies,5 the dosing characteristics were not properly determined in advance. As doses were increased in an effort to have partial remitted patients achieve remission, a fair number of people dropped out or developed akathisia. What looked like a clear therapeutic effect during weeks 3 and 4 eroded enough during the study endpoint that neither of the aripiprazole studies were positive.
Risperidone has shown add-on antidepressant effects in a couple of studies, and the same is true for ziprasidone.
Atypical antipsychotics do have antidepressant effects and the added advantage of protecting people with bipolar disorder against treatment-emergent mania.
Now that atypical antipsychotics are being used, is tardive dyskinesia going to be considered less of a risk?
I think we will see more of it over time but neither as quickly as nor to the degree of that observed with haloperidol. In this era, less tardive dyskinesia has been seen with thioridazine than with haloperidol. Differences even within the family of conventional antipyschotics were present. Atypical antipsychotics may be either at or below the level of thioridazine. We can expect significantly less trouble with atypical than with conventional antipsychotics. PP
1. Calabrese JR, Bowden CL, Sachs GS, Ascher JA, Monaghan E, Rudd GD. A double-blind placebo-controlled study of lamotrigine monotherapy in outpatients with bipolar I depression. Lamictal 602 Study Group. J Clin Psychiatry. 1999;60(2):79-88.
2. Calabrese JR, Keck PE Jr, Macfadden W, et al. A randomized, double-blind, placebo-controlled trial of quetiapine in the treatment of bipolar I or II depression. Am J Psychiatry. 2005;162(7):1351-1360.
3. Thase ME, Macfadden W, Weisler RH, et al. Efficacy of quetiapine monotherapy in bipolar I and II depression: a double-blind, placebo-controlled study (the BOLDER II study). J Clin Psychopharmacol. 2006;26(6):600-609. Erratum in: J Clin Psychopharmacol. 2007;27(1):51.
4. Tohen M, Vieta E, Calabrese J, et al. Efficacy of olanzapine and olanzapine-fluoxetine combination in the treatment of bipolar I depression. Arch Gen Psychiatry. 2003;60(11):1079-1088. Erratumin: Arch Gen Psychiatry. 2004;61(2):176.
5. Thase ME, Jonas A, Khan A, et al. Aripiprazole monotherapy in non-psychotic bipolar I depression: results of two randomized, placebo-controlled studies. J Clin Psychopharmacol. In press.