Then, as so often happens, some pharmaceutical company does something stupid like rigging clinical trial end points to benefit marketing (see "Pharma R&D Succumbs to the Dark Side").
Pharmaceutical companies may be rigging clinical trials in another way: by carefully recruiting patients who are most likely to benefit from the treatment. New evidence in support of this "dark" side of R&D is presented in the latest issue of JAMA.
In the article, "Antidepressant Drug Effects and Depression Severity: Patient-Level Meta-analysis"; JAMA. 2010;303(1):47-53 (doi:10.1001/jama.2009.1943), the authors find that antidepressants are no better than placebo for most depressed patients (ie, those with mild, moderate, or "even severe" baseline symptoms; see chart below; click on it for an enlarged view).
"What makes our findings surprising," say the authors, "is the high level of depression symptom severity that appears to be required for clinically meaningful drug/placebo differences to emerge, particularly given the evidence that the majority of patients receiving ADM [antidepressant medication] in clinical practice present with scores below these levels."
What about all those clinical trials trials that "prove" effectiveness? It turns out that to be eligible for these trials, patients must have pretty high depression scores (HDRS greater than 18). These are precisely the patients most likely to benefit from treatment. "Such cutoffs," note the authors, "can be expected to exclude nearly half of all patients who meet diagnostic criteria for MDD [major depressive disorder]."
But it is not likely that marketers of antidepressants point this out to physicians, especially not to general practitioners who often prescribe antidepressants.
As the authors note: "This important feature of the evidence base [lack of effectiveness in all but the most severe cases] is not reflected in the implicit messages present in the marketing of these medications to clinicians and the public. There is little mention of the fact that efficacy data often come from studies that exclude precisely those MDD patients who derive little specific pharmacological benefit from taking medications."
I'm wondering about other drugs -- Viagra, for example. How were those clinical trials designed? See "Pfizer's Erection Hardness Meter" for more about that!
Maybe another less "dark side" reason is that placebo effects especially in "diseases of the mind" are so strong as to even produce superior outcomes to drugs. The issue is however what you do with the data and how you spin it.
ReplyDeleteThere are so many subtle, and not so subtle, ways that clinical trial methodology effects outcomes. Additionally, since so many trials are managed by clinical research organizations (CROs), there are financial pressures that can affect data collection and reporting at study sites. Add in how companies than report study findings in CME, the scientific literature and their marketing materials and you have the perfect storm for misleading information and mass confusion. I think 2010 will be the year the professional and business media starts covering this topic. Thanks for starting us off in the right direction.
ReplyDeleteGreat post and discussion. I believe it's well understood that the greater the disease severity, the greater the pharmacological effect that is likely to be observed (though not so if drug has no efficacy or if at the wrong part of the dose-response curve). Clinical trials do not show effectiveness, they show efficacy in a narrow patient population. One needs a wider selection criteria or real world observational studies to measure effectiveness. But these also come with problems and issues. Both data sets are usually beneficial to get a well rounded view of the therapeutic profile of the therapy in question. Notwithstanding the errors and omissions that occur in the execution of and data collection in clinical trials, once the data have been properly analyzed, appropriate conclusions drawn, and published in a peer reviewed journal, then the data need to be positioned in a fair and medically accurate manner. Often pharma will go to KOLs to get their opinion on how the data should be properly positioned and this medical perspective helps keep pharma honest. In Canada, all advertising must be pre-cleared by the PAAB and this increases the likelihood that the claims are medically accurate and consistent with the label. If some inappropriate claims slip through, competitors will point out any misleading claims and report these to the PAAB and the situation gets rectified. The system works fairly well. I am less familiar with the US situation and DDMAC. If marketers ignore appropriate claims and regulatory support, then this has to be called out and corrected. Now I know you probably know all of this, so apologies in advance for pointing out the obvious. I am new to the world of pharmaceutical blogs :-)
ReplyDeleteIt's not just "diseases of the mind." This is exactly what goes on with Biologics, the most expensive drugs in the world. They are tested on a very narrowly defined population of Rheumatoid patients. Then they're shown to be less effective in practice than in clinical trials.
ReplyDeleteAt first I considered other explanations for it (that still could be contributing factors). Then, my own doc explained to me that my own severe case would not fit into parameters for a CT. Shocked, I looked deeper.
This is indeed not a separate issue from the marketing side. The ads for these products are a sore spot for patients since this disease is already so misunderstood by the general public who are being told that we are basically "cured" by these treatments that in reality leave most of us still disabled.