Atypical Antipsychotics For Depression: Now With "Considerable Evidence"

I've been wanting to write about this for months. Here goes. We know that antipsychotics are the new panacea for all things mental health-related, including depression (1, 2, 3). But critics kept pointing to a pesky lack of evidence that such treatments actually worked. Bristol-Myers Squibb, manufacturer of Abilify, has been running a disinformation campaign in medical journals to tout its drug as an antidepressant. Their attempts to paint a positive picture of Abilify's antidepressant properties and its allegedly fantastic safety/tolerability profile have been simultaneously tragic and amusing (1, 2, 3).

We're now moving on to something bigger... It ain't just Abilify, folks. It's all the atypicals. They are all antidepressants. According to the authors of a recent meta-analysis, for atypical antipsychotics: "At present, this body of evidence is considerably larger than that for any other augmentation strategy in the treatment of major depressive disorder." In other words, if you are not prescribing atypicals for your patients who don't show adequate response to antidepressants, you are not practicing evidence-based medicine. You are a [bleeping] cowboy who is willfully disregarding science. You are denying your patients the best possible treatment. The authors don't actually say any of those things, but those are the implications. If the evidence for using antipsychotics is "considerably larger" than the evidence for anything else, then the implications are clear-cut. And this is exactly how this study will be cited. Salespeople, from drug reps to academic psychiatrists, to practitioners looking to earn a few thousand extra bucks on the side through pharma speaking gigs, will discuss this study as if it were a landmark finding.

Response and Remission: But the "evidence" is not all that convincing. Here's why... The authors pooled together the results of 16 randomized controlled trials. In these studies, patients had failed to respond adequately (using various definitions) to an antidepressant. Patients were then assigned to receive either an atypical antipsychotic or a placebo in addition to their antidepressant. Outcomes were then tabulated somewhere between 4 and 12 weeks later. The results seem clear cut -- if your brain is turned to "off" -- the response rates for atypicals was 44% compared to 30% for placebo. The remission rates were 31% for atypicals and 17% for placebo. The advantage for atypicals is statistically significant. Well, there you have it. Done deal. Ask your doctor about Abilify/Zyprexa/Seroquel today...

But the most important thing in a treatment outcome study is... the outcomes. The authors of the meta-analysis did not bother to actually measure change in scores on rating scales. Instead, they only used response and remission rates. There is absolutely no good reason for doing this. It's potentially quite misleading. Doctors like remission and response rates because they provide the illusion that we are measuring depression exactly. A "responder" got a lot better and is functioning reasonably well whereas a "non-responder" is in bed 12 hours a day while spending the rest of her time watching the E! Network, eating Bon-Bons, and sobbing constantly. But it's not nearly that scientific. A "responder" is usually defined as someone who got 50% better on his or her depression rating score during the study period. So Bob's depression rating score improved by 52% (he's a responder), but Amy's score only improved by 48%, so she's a nonresponder. Is this 4% difference really meaningful?

Let's look at the following dataset for 20 participants in a fictional study...

Improvements in depression over course of 10 week study

Drug	Placebo
40%	30%
55%	60%
50%	45%
55%	48%
52%	48%
60%	55%
60%	55%
10%	25%
20%	10%
25%	30%

Using a 50% improvement to determine if a patient is a "responder", we get a 60% response rate on drug and a 30% response rate on placebo. Lazy logic says: Oooh -- the drug is twice as effective as placebo. But is we take the average for each group, we get an average improvement of 42.7% on the drug compared to 40.6% on placebo. See the problem with response and remission rates? Similar arguments have been made by smarter people than myself.

Putting outcomes into convenient little categories makes good sense when the categories themselves make sense - events like having a heart attack, getting pregnant, or dying. If the death rate on a drug is 4% compared to 2% on a placebo, then the drug really reduced death by 50%. But if the "remission rate" or "response rate" for depression is 40% on drug compared to 20% on placebo, that does not mean the drug is twice as effective as placebo in treating depression. If you need to score a 7 or below on a depression rating scale to be "in remission", but you score an 8, are you really much worse off than the person who scored a 7?

Am I saying that the drugs really just squeaked by placebo in these studies? Well, I've read the Abilify studies and posted on them previously - in those studies, Abilify barely beat the placebo. And in the opinion of the patients themselves, Abilify didn't beat placebo at all. And the studies were designed to benefit Abilify, not to actually see if the drug worked. As I noted previously...

Patients were initially assigned to receive an antidepressant plus a placebo for eight weeks. Those who failed to respond to treatment were assigned to Abilify + antidepressant or placebo + antidepressant. Those who responded during the initial 8 weeks were then eliminated from the study. So we've already established that antidepressant + placebo didn't work for these people -- yet they were then assigned to treatment for 6 weeks with the same treatment (!) and compared to those who were assigned antidepressant + Abilify. So the antidepressant + placebo group started at a huge disadvantage because it was already established that they did not respond well to such a treatment regimen. No wonder Abilify came out on top (albeit by a modest margin).

Here's an analogy. A group of 100 students is assigned to be tutored by Tutor A regarding math. The students are all tutored for 8 weeks. The 50 students whose math skills improve are sent on their merry way. That leaves 50 students who did not improve under Tutor A's tutelage. So Tutor B comes along to tutor 25 of these students, while Tutor A sticks with 25 of them. Tutor B's students do somewhat better than Tutor A's students on a math test 6 weeks later. Is Tutor B better than tutor A? Not really a fair comparison between Tutor A and Tutor B, is it?

I've not read the other antipsychotics for depression studies. I'll even give them the benefit of the doubt and assume they were not designed in the same biased manner as the Abilify trials. It is, however, worth noting that the "benefit" of Abilify, in terms of response and remission rates compared to placebo, was about the same as for the other atypicals. Which leads me to think that the other atypicals probably show similar marginal benefits for depression.

But now, based solely on potentially quite misleading response and remission rates, an article appears in the American Journal of Psychiatry - a piece that has the potential to ramp up the prescribing of antipsychotics for depression to an even more ridiculous level. Let the good times roll.

Source of ironclad evidence that atypical antipsychotics are antidepressants (until you actually read the paper):

Nelson, J., & Papakostas, G. (2009). Atypical Antipsychotic Augmentation in Major Depressive Disorder: A Meta-Analysis of Placebo-Controlled Randomized Trials American Journal of Psychiatry, 166 (9), 980-991 DOI: 10.1176/appi.ajp.2009.09030312

Treatment Guidelines and GSK's Open Disclosure

Last week, I noted that a recently published article had found that studies favoring GSK's "mood stabilizer" Lamictal tended to get published in medical journals while articles reaching less favorable conclusions tended to remained unpublished. I wrote that "GSK worked the system expertly and it paid off." A reader commented that he thought my characterization of GSK as hiding negative data on Lamictal was inaccurate. I appreciate his well-written critical comments, which are linked here and are partially reproduced below:

Acute Depression - All of the acute depression studies (there were 5 not 3 as you reported) were presented at scientific meetings over the years and were recently published in Bipolar Disorders (Calabrese et al. 2008). Why so long to publish? The paper was rejected twice and took 3 years to get accepted because journal reviewers did not find the data of interest.

I responded via comment that, if his history is accurate, then the reviewers should be flogged. He added that GSK had provided negative Lamictal data to numerous authors who wrote review articles on Lamictal. In some cases, this appears to be true. However, in at least one notable case, either GSK failed to provide the data or the authors completely ignored the negative data. The data here appeared in a 2004 "academic highlight" (i.e., lowlight) in the Journal of Clinical Psychiatry. Of relevance, the article was funded by an "unrestricted educational grant" from GSK. The article bashes antidepressant treatment in bipolar as unsupported by evidence. Then the expert panel of authors/key opinion leaders put together their guidelines for treating bipolar disorder.

The article begins by discussing bipolar depression. Lithium is discussed first and receives a positive review. Then comes Lamictal, GSK's mood stabilizer. They discuss, in detail, the positive results from Calabrese et al. The authors then discuss some positive long-term findings for lamotrigine before moving on to olanzapine and olanzapine/fluoxetine. They conclude that lithium and Lamictal have the best evidence for treating bipolar depression as can be seen here:

Category 1 evidence is the best evidence, so hooray for lamotrigine/Lamictal! But what don't they discuss in their "expert" review of the data? How about two negative studies -- SCA40910 (completed in 2002) and SCAB2001 (completed in 1997) -- GSK titles of studies that both showed negative results for Lamictal in treating depression in bipolar disorder. A reader tracked these down and sent them -- you can find them if you head to GSK's clinical trial registry. Given that these "International Consensus Guidelines" were published in February of 2004, you'd think the authors would have included data from both of GSK's unpublished studies unless:
A. They didn't know about their existence (and why would they unless GSK told them)
B. They knew about them but opted to not include them in this "expert review"

Given that a GSK employee has told me how open and honest GSK has been with their data, I'd be interested in seeing his response as to which of the above he believes took place. Keep in mind that the Journal of Clinical Psychiatry, in which this so-called "academic highlight" appeared is a very widely read journal. According to Google Scholar, this piece has been cited 46 times, many of which have doubtlessly recycled the inaccurate claim that Lamictal is an effective treatment for acute bipolar depression.

The same pattern as usual: Company conducts research, selectively publishes positive results, funds "educational" pieces such as "academic highlights" to paint an overly rosy picture of treatment effectiveness and/or safety, and physicians, based upon the "evidence base" delude themselves into thinking that they are writing prescriptions based on the best scientific data.

Antidepressants: Hiding and Spinning Negative Data

As I alluded to yesterday, a whopper of a study has just appeared in the New England Journal of Medicine. It tracked each study antidepressant submitted to the FDA, comparing the results as seen by the FDA in comparison with the data published in the medical literature. The FDA uses raw data from the submitting drug companies for each study. This makes great sense, as the FDA statisticians can then compare their analyses to the analyses from drug companies, in order to make sure that the drug companies were analyzing their data accurately.

After studies are submitted to the FDA, drug companies then have the option of submitting data from their trials for publication in medical journals. Unlike the FDA, journals are not checking raw data. Thus, it is possible that drug companies could selectively report their data. An example of selective data reporting would be to assess depression using four measures. Suppose that two of the four measures yield statistically significant results in favor of the drug. In such a case, it is possible that the two measures that did not show an advantage for the drug would simply not be reported when the paper was submitted for publication. This is called "burying data," "data suppression," "selective reporting," or other less euphemistic terms. In this example, the reader of the final report in the journal would assume that the drug was highly effective because it was superior to placebo on two of two depression measures, left completely unaware that on two other measures the drug had no advantage over a sugar pill. Sadly, we know from prior research that data are often suppressed in such a manner. In less severe cases, one might just switch the emphasis placed on various outcome measures. If a measure shows a positive result, allocate a lot of text to discussing that result and barely mention the negative results.

But wait, there's an even better way to suppress data. Suppose that a negative study is submitted to the FDA. There is no commercial value in presenting negative results on a product. Indeed, it makes no sense from a commercial vantage point to submit a clinical trial that shows no advantage for one's drug for publication in a medical journal. While it earns a bit of good PR for being honest, it would of course hurt sales for the drug, which would not please shareholders. From an amoral, purely financial view, there is no reason to publish negative trial results.

On the other hand, there is science. One of the first things that any medical student hopefully learns is that scientists should report all of their results so that other scientists, physicians, the media, and the general public have an up-to-date and comprehensive understanding of all scientific findings. Yes, this may sound naive, but this is how science is supposed to work in an ideal world.

Back to the NEJM study. Were manufacturers of antidepressants playing by the rules of science or the rules of the almighty dollar? Take a look at this table excerpted from the study...

The FDA concluded that 38 studies yielded positive results. 37 of these 38 studies were published. The FDA found mixed or "questionable" results in 12 studies. Of these 12 studies, six were not published, and six others were published as if they were positive findings. Of the 24 studies that the FDA concluded were negative, three were published accurately, five were published as if they were positive findings, and 16 were not published. To summarize, positive studies were nearly always reported while mixed and negative studies were nearly always either not published or published in a manner that spun the results unreasonably. How does one turn a questionable or negative finding into a positive one? As mentioned above, report the results that are favorable to your product and sweep the remaining results under the rug.

Overall, how do the statistics for this group as prepared by the FDA compare to the statistics in medical journal publications? Remember, physicians are trained to highly value medical journals, as they are the storehouse for "evidence-based medicine." I'll borrow a quote from the study authors:

For each drug, the effect-size value based on published literature was higher than the effect-size value based on FDA data, with increases ranging from 11 to 69%

Well, that's not very reassuring. Effect size refers to the magnitude of the difference between the drug and placebo. Note that for every single drug, the effect size as reported in the medical literature (the foundation for "evidence based medicine) was greater than the effect size calculated from the FDA's data. Remember, the FDA's data is based on raw data submitted by drug companies, and is thus much less subject to bias than data that the drug companies manipulate prior to submitting for publication in a medical journal. Other highlights from the authors:

Not only were positive results more likely to be published, but studies that were not positive, in our opinion, were often published in a way that conveyed a positive outcome... we found that the efficacy of this drug class is less than would be gleaned from an examination of the published literature alone. According to the published literature, the results of nearly all of the trials of antidepressants were positive. In contrast, FDA analysis of the trial data showed that roughly half of the trials had positive results. The statistical significance of a study’s results was strongly associated with whether and how they were reported, and the association was independent of sample size.

I'll say it one more time: Every single drug had an inflated effect size in the medical literature in comparison with the data held by the FDA. To move into layman's terms for a moment, manufacturers of every single drug appear to have cheated. This is not some pie in the sky statistics review -- this is the medical literature (the foundation of "evidence-based medicine") being much more optimistic about the effects of antidepressants than is accurate. This is marketing trumping science.

The drugs that were found to have increased their effects as a result of selective publication and/or data manipulation:

• Bupropion (Wellbutrin)
• Citalopram (Celexa)
• Duloxetine (Cymbalta)
• Escitalopram (Lexapro)
• Fluoxetine (Prozac)
• Mirtazapine (Remeron)
• Nefazodone (Serzone)
• Paroxetine (Paxil)
• Sertraline (Zoloft)
• Venlafaxine (Effexor)

That is every single drug approved by the FDA for depression between 1987 and 2004. Just a few of many tales of data suppression and/or spinning can be found below:

• Data reported on only 1 of 15 participants in an Abilify study
• Data hidden for about 10 years on a negative Zoloft for PTSD study
• Suicide attempts vanishing from a Prozac study
• Long delay in reporting negative results from an Effexor for youth depression study
• Data from Abilify study spun in dizzying fashion. Proverbial lipstick on a pig.
• A trove of questionable practices involving a key opinion leader
• Corcept heavily spins its negative antidepressant trial results

Props to the Wall Street Journal (David Armstrong and Keith Winstein in particular) and the New York Times (Benedict Carey) for quickly getting on this important story.

There are some people who seem unmoved by this story. Indeed, some people are crying that this is an unfair portrayal of the drug industry. More on their curious take on the situation coming later.

I'll close with a question: What does this say about the key opinion leaders whose names appear as authors on most of these published clinical trials in which the data is reported inaccurately?

TMAP and Bipolar: Where's the Beef?

Much has been made of the Texas Medication Algorithm Project. I have written about it earlier (1, 2), as have others (3). A lawsuit has been filed alleging that TMAP was a sneaky way to convince state mental health programs to switch their patients to newer, much more expensive medications. TMAP defenders, on the other hand, say that TMAP was simply allowing patients access to the state of the art, most effective medications.

What is TMAP? Essentially, TMAP is a program that used “expert consensus” to develop treatment guidelines for patients (depression, bipolar, and schizophrenia) in the public mental health system. On one hand, I can see why care should be improved – not many people seriously argue that patient care is very good in most public mental health systems.

According to the TMAP model, medication treatment is provided in stages according to these guidelines. If you are not responding to treatment #1, then you move to treatment #2, and if that does not work, then to treatment #3, and so on. Naturally, the “objective experts” who developed said guidelines stuck the newer, more expensive medications on the top of the list for treatments, especially as the guidelines have been revised during the past couple of years.

TMAP was unfurled in the mid 1990’s and similar programs have since been sweeping across many states.

Was this a way to backdoor newer medications onto patients? Well, I think that is probably the case, but the issue I am going to address here is one that I think is even more important…

Does TMAP Work? Do TMAP patients show more improvement than patients who were not on the TMAP treatment regimen? The TMAP team has produced some evidence in which they claim that TMAP treatment works better than “treatment as usual,” which was standard state mental health care. In this case, we’ll discuss TMAP for bipolar patients.

The Study: Some patients received TMAP, which included both a standardized medication algorithm as follows:

Manic:
Stage 1 – Depakote or Lithium or Tegretol
Stage 2 – Depakote + Lithium OR Tegretol + Lithium
Stage 3 – Depakote + Lithium OR Tegretol + Lithium
Stage 4 – Depakote + Tegretol
Stage 5 – Add atypical antipsychotic to mood stabilizer
Stage 6 – ECT
Stage 7 – Other (e.g., Lamictal, Neurontin)

Depressed:
Stage 1 – Wellbutrin or SSRI + mood stabilizer
Stage 2 – Wellbutrin or SSRI or Effexor or Serzone + mood stabilizer
Stage 3 – Mood Stabilizer + two antidpressants
Stage 4 – Mood Stabilizer and MAOI antidepressant
Stage 5 – ECT
Stage 6 – Other (e.g., Lamictal)

Note that in the 2005 revision of this standard, atypical antipsychotics are featured much more prominently. But when the study on bipolar patients was conducted, this was not the case.

Those who received “treatment as usual” (TAU) received whatever care they would normally receive.

The Results: On some measures, TMAP patients did modestly better than TAU patients. This could be interpreted as evidence that these strict treatment algorithms that involve a high frequency of prescribing Depakote and Tegretol, and to a lesser extent, newer antipsychotics, are a good idea for patients. However, one would be fooling oneself to buy this conclusion. Why?

The (Huge) Caveat: The TMAP patients all received: Group education, consumer to consumer discussion groups, individual patient education from the physician, referrals to therapy groups and more. These interventions were rolled out exclusively for the TMAP group. Why does this matter? Well, patients in the TMAP group were likely getting more time with their physician, which is likely going to boost their relationship with the physician, which will likely lead to better outcomes regardless of the medication taken. In addition, the patient education groups provide additional support for patients, which has been shown to improve outcomes.

Even the study authors, to their credit, admit this is a gigantic potential issue:

At this time, the relative contributions of different elements [i.e., medication versus the extra patient care] of the “disease management package” to the obtained results has not been evaluated.

A Better Idea for TMAP: If you wanted a study that would have compared the effects of a) extra patient education and support, b) use of medication algorithms that favored use of newer medications and c) “treatment as usual” – regular care in the state mental health system, then why not have a study that looks like this:

A) Treatment as usual (No extra patient support)
B) Extra Patient Support + TMAP Algorithms
C) TMAP Algorithms (No extra patient support)
D) Extra Patient Support + Treatment as usual

If C’s outcomes are better than A’s outcomes, then you can shout about the evidence base of your algorithms. If B is greater than D, then you can also do your evidence-based practice speech. However, the TMAP bipolar study as was actually conducted was A versus B, – that is a pretty lame comparison. Was it the patient support (which is my guess) or was it the algorithms?

Why Was TMAP Investigated This Way? This is where it gets interesting but murky. Could such a study have been designed because it was biased to find favorable results for the TMAP intervention (due to TMAP patients receiving extra support not received by treatment as usual)? Then, finding positive results, it becomes a lot easier to sell the program to other states because TMAP is now “evidence based”. Maybe I’m off in Conspiracyville, but one has to admit this is pretty weird stuff.