Having previous experience with antidepressants can change a person’s future response to both medications and to placebo, according to a new study. The findings could have implications for clinical trials designed to test new drugs.
For the new study, researchers led by Aimee Hunter, assistant professor of psychiatry at University of California, Los Angeles, studied the brain activity of 89 depressed adults who were enrolled in clinical trials of the antidepressant drug Prozac (fluoxetine) or Effexor (venlafaxine).
During the first week of each trial, all participants took placebo and then after that, they were given either a study drug or placebo. Using a technique called quantitative electroencephalography (QEEG), which involves a cap of electrodes placed painlessly on the scalp, researchers measured and mapped the participants’ brain activity.
They found differences based on the participants’ previous history of antidepressant use — a factor that wasn’t taken into account in the trials. “We found that prior history affects the response to placebo,” Hunter says. “If they had a prior history of taking antidepressants and were given placebo, the brain changes were indistinguishable from those of the people given medication.”
The changes were seen on a measure called prefrontal theta-band cordance (PFC). People who had previously taken antidepressants showed a larger decline in PFC than those who had not, regardless of whether they were given a real medication or placebo. “That’s the overall finding,” Hunter says. “If [someone has] taken medication before, then the brain changes, whether on medications or placebo, showed decreases.”
Hunter and her colleagues had previously shown that a decline in this measure is associated with recovery from depression for those taking medication, although curiously, those who got better on placebo showed a slight increase in PFC.
In the new study, however, even after the researchers controlled for whether depression improved or not, the change in brain activity associated with prior medication history remained measurable and significant. “Treatment history still has an effect on brain function over and above that accounted for by clinical improvement or worsening,” says Hunter.
The findings, if replicated, could have implications for drug development and testing, since the Food and Drug Administration relies on placebo-controlled trials to determine which drugs to approve. Because many people who participate in clinical trials have a prior history of medication use, their response to the experimental drug may be altered, making effective medications appear not to work and vice versa, depending on the make-up of the groups in the trial. Recently, drug companies have struggled with strong placebo effects — which make drugs look less effective — and a growing population of patients with prior medication history may be a factor.
“I hate to think that a medication either will or won’t remain in the development pipeline, or be approved or not approved, based on how it compares in physiological effects to placebo, when placebo response is driven by this thing we’re not controlling for at all,” says Hunter.
The reasons for the difference in response are probably related to learning and expectations. Although Hunter and her colleagues didn’t have data about whether their participants’ response to prior antidepressant treatment had been positive or negative, that may well be a key determinant of future results.
Earlier research has suggested that the brain’s response to drugs involves conditioning. Drug tolerance isn’t just a matter of pharmacology, but also involves an unconscious learning process. For example, one study found that heroin overdoses were more likely in people who injected the drug in unfamiliar places. Without the familiar cues associated with their normal injecting routine, users’ typical tolerance response didn’t kick in and their regular dose became an overdose.
With regard to antidepressants, people who have previously taken the drugs and felt better might have a similar physiological response to the cue of being given a pill that is said to be an antidepressant. In contrast, those who took a pill and got worse or did not improve might come to have a negatively conditioned response. This could explain at least some of why depression becomes resistant to treatment.
“We’ve always known this to some degree,” says Dr. David Healy, professor of psychological medicine at Cardiff University in Wales and a leading expert on antidepressant research. He explains that rats given drugs in different environments also have variations in drug tolerance related to cues. But he doesn’t think it’s necessary to control for it, considering that clinical trials have larger issues, such as failure to disclose data.
Hunter sees a serious problem, however. “It’s entirely possible that you learn that this type of stimulus is not going to make you better,” she says. “Then let’s say a brand new powerful antidepressant [becomes available] — you might be psychologically conditioned not to be able to benefit even from something with a novel mechanism.”
Hunter believes that if her results stand up to further study, participants’ prior history of medication use may become another factor to consider when studying antidepressants. But since the response is learned, it’s also possible that helping patients differentiate the new drug, as well as the environment in which it’s taken, from past medication experiences could eliminate the problem and possibly even enhance drug effects.
The research, published in European Neuropsychopharmacology, was funded by both pharmaceutical companies and the National Institute of Mental Health and the authors report prior industry funding.