Each month, three presenters review an interesting journal article in a conversational manner. These articles involve “hot topics” that affect family physicians or “bust” commonly held medical myths. The presenters give their opinions about the clinical value of the studies discussed. The opinions reflect the views of the presenters, not those of AFP or the AAFP.
This Month's Article
Berman RM, Marcus RN, Swanink R, et al. The efficacy and safety of aripiprazole as adjunctive therapy in major depressive disorder: a multicenter, randomized, double-blind, placebo-controlled study. J Clin Psychiatry. 2007;68(6):843–853.
Does augmentation with aripiprazole help patients with major depressive disorder who have not responded to SSRI or SNRI therapy?
Mark: We all have patients who do not respond well to antidepressants. It would be nice to be able to add a rapidly acting augmentation agent to improve the patient's depression. Unfortunately, aripiprazole (Abilify) isn't it.
What does this article say?
Mark: Patients were included in this drug company-funded and -written single-blind study if they had major depressive disorder lasting at least eight weeks and were unresponsive to antidepressant treatment, defined as less than 50 percent reduction in depressive symptoms as measured by the General Hospital Antidepressant Treatment Response Questionnaire (ATRQ-24). This study had three phases:
Phase 1. During this phase, which lasted seven to 28 days, patients were screened and recruited. Patients who previously failed aripiprazole were excluded.
Phase 2. To enter this phase, patients had to have a score of 18 or higher on the 17-item Hamilton Rating Scale for Depression (HAM-D17). They were treated with an SSRI (escitalopram [Lexapro], fluoxetine [Prozac], controlled-release paroxetine [Paxil CR], or sertraline [Zoloft]) or an SNRI (extended-release venlafaxine [Effexor XR]), and were followed for eight weeks. Patients who had a HAM-D17 score indicating a less than 50 percent reduction in symptoms, a total HAM-D17 score of 14 or higher, and a Clinical Global Impressions-Improvement (CGI-I) score of 3 or higher were then included in phase 3.
Phase 3. Patients in this phase were randomized to receive adjunctive aripiprazole (184 patients) or placebo (178 patients), and were followed for six weeks. They received no change in the dose of the SSRI or SNRI.
The primary outcome was the patient's score on the Montgomery-Asberg Depression Rating Scale (MADRS); a lower score means the patient has less severe depression. The authors found a statistically significant difference in the two groups, with the aripiprazole group scoring better than the placebo group (−8.8 versus −5.8 from the patient's baseline, respectively; P < .001). They conclude that aripiprazole is useful as an adjunct to SSRIs and SNRIs.
Should we believe this study?
Mark: No. There are so many design flaws in this study that it would be impossible to cover them all. Let's just look at a couple. First, most (but not all) other psychotropic medications were discontinued. Thus, patients could continue on other drugs besides the study drugs; this was not controlled for. Second, the MADRS is scored from 1 to 60. Is a difference of three points noticeable to the patient (−8.8 versus −5.8)? As it turns out, no.
Andrea: How do we know? The authors had the patients do a self-reported scale called the Inventory of Depressive Symptomatology (IDS-SR). Patients noticed no difference in the way they felt (P = .076), despite a statistical difference of three points in the MADRS. When evaluating a study, you need to know what scale is being used and whether or not a statistical difference (in this case, three points) is clinically meaningful.
Mark: Interestingly, the study authors did a subgroup analysis. Now, we are not fans of subgroup analyses. They should only be used to generate a hypothesis (the derivation set) that should then be tested in a separate study (the validation set). But in this case, it is worth reporting. The difference in the MADRS for men was +0.48 (yes, that is less than one point on a 60-point scale favoring placebo). The difference for women was −5. Intuitively, this doesn't make any sense. Why should the drug work in women, but not in men? Regardless, as noted above, the patients noticed no difference in how they felt.
Bob: Wait, there's more. Physicians wouldn't normally wait eight weeks before upping the dose, adding a second drug, or switching drugs in a patient with major depressive disorder. So, this is a “straw man comparison”; the authors are comparing aripiprazole with no action, a situation that wouldn't happen in the real world.
Andrea: Amazingly, they used five different depression scales in this study (HAM-D17; ATRQ-24; MADRS; IDS-SR; and CGI-I). It would have made more sense to pick one scale and stick with it. They used different scales for entry into the study and for outcomes at the end of the study. The authors do not give a reason for their choice of using mean change on the MADRS as the primary end point, and we have no idea how patients were doing at the end of the study compared with their baseline in phase 1. This does not give me a lot of confidence in the results.
Mark: Importantly, they excluded patients who had previously tried and failed aripiprazole therapy. This biases the experimental group toward patients who might respond, and makes the results look better. Thus, it does not reflect our real-world patients.
Bob: Were there any differences between the two groups? You bet. Akathisia? 4.5 percent in the placebo group versus 23.1 percent in the aripiprazole group. Restlessness? 3.4 versus 14.3 percent. Fatigue? 3.4 versus 6.0 percent. To be fair, there were more headaches in the placebo group (10.8 versus 6.0 percent).
What should the family physician do?
Mark: This study provides no evidence that aripiprazole is useful as an adjunct to SSRIs or SNRIs in patients with major depressive disorder. Based on this study, you should not use aripiprazole for this indication.
Andrea: Other drugs that are useful adjuncts in depression include bupropion (Wellbutrin),1 tricyclic antidepressants, lithium, and triiodothyronine. Other atypical antipsychotics have also been used, but data are limited and not of the best quality.2 MAOIs are contra-indicated with SSRIs and SNRIs because of the risk of serotonin syndrome. When prescribing an augmentation agent, you should have a high clinical suspicion for serotonin syndrome if the patient has muscle rigidity or mental status changes.
Bob: Lithium has the best data and may be particularly effective. It reduces suicide risk in patients with bipolar disorder and it seems to do the same for patients with depression.3 Lithium will be more difficult to manage in patients on NSAIDs, ACE inhibitors, ARBs, or diuretics, and in patients with renal failure, and it may also cause hypothyroidism, which can mimic depression. It has rarely been associated with serotonin syndrome.4
When used as an adjunct to SSRIs or SNRIs, aripiprazole (Abilify) did not significantly improve depression in patients in this study.
Lithium is the best studied augmentation agent and should be considered as an adjunct to SSRIs or SNRIs. However, there are multiple contraindications and drug interactions.
Other antidepressants (but not MAOIs) can be used to augment SSRIs and SNRIs, but you should watch for serotonin syndrome.
When reviewing a study, you must know what the scale measures, whether the scale has been validated, and what change in the scale is actually clinically significant.
Subgroup analyses should only be used to generate a hypothesis (the derivation set). This hypothesis then needs to be tested in a separate randomized study (the validation set).
Excluding patients who have already failed the study drug biases the study in favor of that drug. If these patients had been included (e.g., a random sample of patients), the response rate to the study drug would have been even less. In this study, patients who had already failed aripiprazole were excluded.
Almost anything will look better than placebo. A study should compare the study drug with a real-world scenario (in this case, upping the dose, switching antidepressants, or using another drug for augmentation). That was not done in this study.