Our seventh installment of The Placebo Problem continues our look at the increasingly high placebo response that is plaguing clinical trials in analgesia and psychiatry. Check out the other posts here.
The placebo response is broad. It goes far beyond the effects of merely consuming a sugar pill; it is the patient’s response to the entire therapeutic context in which that treatment is administered. In a placebo-controlled trial, because the drug group also receives treatment in this same therapeutic context, the drug group’s overall response includes the placebo response. The pharmacological effect of the drug—what we really want to know—is then calculated as the overall drug response minus the placebo response.
Additive or interactive?
This method of calculating the true effect of a drug is predicated on one fundamental assumption: that the drug effect and placebo response are additive. That is, that the response in the drug group equals the sum of the placebo response and the drug’s pharmacological effect. Said another way, it assumes that the placebo response is the same in the placebo group as in the drug group. It also predicts that the pharmacological effect of the drug isn’t affected by whether or not the patient thinks he or she has received the drug.
This so-called additivity hypothesis has been guiding clinical trial interpretation for decades, yet it’s never been directly tested. In fact, over the last 10 years a growing body of evidence, though far from conclusive, suggest that perhaps drug-placebo interactions aren’t so simple.
Meta-analyses of placebo-controlled trials have shown that some factors that are associated with a higher response in the placebo group, such as lower symptom severity at the start of the trial, are not associated with the response of the drug group. This suggests that different mechanisms underlie the placebo response in the placebo and drug arms.
Neurobiological evidence also calls the additive model into question. Made possible using computer-controlled drug infusion pumps, studies using an open-hidden design, where one group of participants is aware they are receiving a drug (open group) and another group has no idea (hidden group), have demonstrated that hidden administration of a drug substantially reduces or even completely negates its pharmacological effect. Thus, patient expectancy likely plays a role in a drug’s pharmacological effect. Neuroimaging data also suggest that placebo effects during open drug treatment are different from those in the placebo group.
Taken together, this body of research suggests that the additive hypothesis may not apply, at least not in all trials. An interactive model of the drug-placebo relationship that produces unequal placebo responses in the drug and placebo treatment groups has also been proposed, but not yet directly tested. According to this hypothesis, drug-specific and nonspecific (placebo) effects in the drug treatment group interact to yield a total treatment response that is different than the simple sum of the placebo group response plus the pharmacological effect of the drug.
With the additive model of the drug-placebo relationship called into question, where do we go from here? One conclusion is that current clinical trial designs are misleading with respect to the true pharmacological effects of drugs and are not the best way to assess the efficacy of novel compounds. To say this realization would have a dramatic effect on the pharmaceutical industry is an understatement; fortunately, this is a premature conclusion at this point. However, these findings do suggest that it would be prudent to devote more research to the topic, as well as explore alternative trial designs that could better incorporate non-additive drug-placebo interactions.