Much attention has been drawn to a recent PLoS Medicine article criticizing the evidence base for the use of Abilify as maintenance treatment for bipolar disorder. The major points emphasized by most critics are, first, that the FDA approved Abilify for this purpose in 2005 on the basis of flawed and scanty evidence and, secondly, that the literature since that time has failed to point out the deficiencies in the original study.
While the above may be true, I believe these criticisms miss a more important point. Instead of lambasting the FDA or lamenting the poor quality of clinical research, we psychiatrists need to use this as an opportunity to take a closer look at what we treat, why we treat, and how we treat.
Before elaborating, let me summarize the main points of the PLoS article. The authors point out that FDA approval of Abilify was based on only one “maintenance” trial by Keck et al published in 2007. The trial included only 161 patients (only 7 of whom, or 1.3% of the total 567 who started the study, were followed throughout 26 weeks of stabilization and 74 follow-up weeks of maintenance). It also consisted of patients who had already been stabilized on Abilify; thus, it was “enriched” for patients who had already shown a good response to this drug. Furthermore, the “placebo failures” consisted of patients who were abruptly withdrawn from Abilify and placed on placebo; their relapses might thus be attributed to the researchers’ “randomized discontinuation” design rather than the failure of placebo. (For more commentary, including follow-up from Bristol-Myers Squibb, Abilify’s manufacturer, please see this excellent post on Pharmalot.)
These are all valid arguments. But as I read the PLoS paper and the ongoing discussion ever since, I can’t help but think, so what?? First of all, most psychiatrists probably don’t know about the PLoS paper. And even if they did, the major questions for me would be: would the criticism of the Keck et al. study change the way psychiatrists practice? Should it?
Let’s think about psychiatric illness for a moment. Most disorders are characterized by an initial, abrupt onset or “episode.” These acute episodes are usually treated with medications (plus or minus psychotherapy or other psychosocial interventions), often resulting in rapid symptomatic improvement—or, at the very least, stabilization of those symptoms.
One big, unanswered (and, unfortunately, under-asked) question in psychiatry is, then what? Once a person is stabilized (which in some cases means nothing more than “he’s no longer a danger to himself or others”), what do we do? We don’t know how long to treat patients, and there are no guidelines for when to discontinue medications. Instead we hear the common refrain: depression, schizophrenia, and bipolar disorder, are lifelong illnesses—”just like hypertension or diabetes”—and should be treated as such.
But is that true? At the risk of sounding like a heretic (and, indeed, I’d be laughed out of residency if I had ever asked this question), are there some cases of bipolar disorder—or schizophrenia, or depression, for that matter—which only require brief periods of psychopharmacological treatment, or none at all?
The conventional wisdom is that, once a person is stabilized, we should just continue treatment. And why not? What doctor is going to take his patient off Abilify—or any other mood stabilizer or antipsychotic which has been effective in the acute phase—and risk a repeat mood episode? None. And if he does, would he attribute the relapse to the disease, or to withdrawal of the drug? Probably to the disease.
For another example of what I’m talking about, consider Depakote. Depakote has been used for decades and is regarded as a “prototypical” mood stabilizer. Indeed, some of my patients have taken Depakote for years and have remained stable, highly functional, and without evidence of mood episodes. But Depakote was never approved for the maintenance treatment of bipolar disorder (for a brilliant review of this, which raises some of the same issues as the current Abilify brouhaha, read this article by The Last Psychiatrist). In fact, the one placebo-controlled study of Depakote for maintenance treatment of bipolar disorder showed that it’s no better than placebo. So why do doctors use it? Because it works (in the acute phase.) Why do patients take it? Again, because it works—oh, and their doctors tell them to continue taking it. As the old saying goes, “if it ain’t broke, don’t fix it.”
However, what if it is broke[n]? Some patients indeed fail Depakote monotherapy and require additional “adjunctive” medication (which, BTW, has provided another lucrative market for the atypical antipsychotics). In such cases, most psychiatrists conclude that the patient’s disease is worsening and they add the second agent. Might it be, however, that after the patient’s initial “response” to Depakote, the medication wasn’t doing anything at all?
To be sure, the Abilify study may have been more convincing if it was larger, followed patients for a longer time, and had a dedicated placebo arm consisting of patients who had not been on Abilify in the initial stage. But I maintain that, regardless of the outcome of such an “improved” trial, most doctors would still use Abilify for maintenance treatment anyway, and convince themselves that it works—even if the medication is doing absolutely nothing to the underlying biology of the disease.
The bottom line is that it’s easy to criticize the FDA for approving a drug on the basis of a single, flawed study. It’s also easy to criticize a pharmaceutical company for cutting corners and providing “flawed” data for FDA review. But when it comes down to it, the real criticism should be directed at a field of medicine which endorses the “biological” treatment of a disorder (or group of disorders) whose biochemical basis and natural history are not fully understood, which creates post hoc explanations of its successes and failures based on that lack of understanding, and which is unwilling to look itself in the mirror and ask if it can do better.