Implications of a Biosignature Study of the Placebo Response in Major Depressive Disorder

The article by Pecina and colleagues1 in this issue of JAMA Psychiatry reports the results of a study evaluatingwith a neuroimaging paradigm theneurochemicalmechanismsunderlying the formation of placebo effects in patients withmajor depressive disorder (MDD). Theirstudyinvolvedperforminga single-blinded2-weekcrossover randomizedclinical trial of 2 identical oral placebos (describedashavingeither activeor inactive fast-acting antidepressantlike effects), followed by a 10-weekopen-label treatmentwithanantidepressant.Theparticipants with MDD were studied with positron emission tomography and the μ-opioid receptor–selective radiotracer [11C]carfentanil after each 1-week inactive and active oral placebo treatment. In addition, 1mLof isotonic salinewasadministered intravenouslywithin sightof thevolunteerduringpositron emission tomographic scanning every 4minutes over 20 minutesonlyafter the 1-weekactiveplacebo treatment.Reductions in depressive symptoms after 1 week of active placebo treatment, comparedwith the inactiveplacebotreatment,were associatedwithincreasedplacebo-inducedμ-opioidneurotransmission in a network of regions implicated in emotion, stress regulation, and thepathophysiology ofMDD, namely, the subgenual anterior cingulate cortex, nucleus accumbens,midline thalamus, andamygdala.Their findingshavenumerous implications. The identificationofabiological response,namely,amodulation of μ-opioid neurotransmission, to placebo exposure amongpatientswithMDD is consistentwith theview thatplacebo treatment can, in fact, induce robust neurobiological changes and that the brain responds to this type of intervention in away that is quite robust and leads to changes in brain functioning.These findingsappear toprovideanadditional rationale for the use of open-label placebo as a possible firstline treatment for depression.2 Therefore, in addition to the traditional rationales—such as double-blind placebos being roughly 80% as effective as antidepressants in randomized clinical trials, classic conditioning being a possible mechanism for automatic self-healing, andpositive expectations increasing placebo effects3—the evidence of an actual biosignatureof the effect of placebo inMDDmayprovidepatientswith MDDwith thedemonstration that placebos canbeactually active treatments in their condition. To our knowledge, our group3 conducted the first-ever pilot randomized clinical trial of open-label placebo in 20 adult outpatients with a primary diagnosis of nonpsychotic MDD. For the first 2 weeks, patients were randomized to either open-label placebo or waitlist control. After 2 weeks, participants originally randomized toopen-label placebocontinued for anadditional 2weeks taking open-label placebo, and participants originally assigned to thewaitlist controlwere switched to open-label placebo for an additional 4 weeks, as long as they continued to meet the eligibility requirements after 2 weeks on the waitlist. At the 2-week end point, although there were no statistically significant differences between open-label placebo and waitlistprimarilyowing to thesmall sample size, theeffect size for themainoutcomemeasure (17-itemHamiltonScale forDepressionscore)wasmedium(d = 0.54).3 Inaddition, therewere statistically significant pre-post improvements on all 3 outcome measures after 4 weeks of open-label placebo treatment, with moderate to large effect sizes.3 Therefore, the results of the study by Pecina and colleagues1 confirm the impression that the administration of placebo in MDD leads to brain changes that go beyond simple regression toward the meanandspontaneous improvements andare consistentwith the findings of our study of open-label placebo in MDD.3 Another oneof thekey findingsof the studybyPecina and colleagues1 is that placebo-induced endogenous opioid release inanetworkof regions implicated inemotion, stress regulation, and thepathophysiologyofMDD is associatedwithbetter open-label antidepressant treatment response, predicting 43% of the variance in symptom improvement at the end of the antidepressant trial. A possible interpretation of this finding is that such correlation is accounted for by the fact that, nested within the response to active antidepressant therapy, there is a substantial placebo response. As described by our group,4 onemayclassify treatedpatients in adepression clinical trial based on each participant’s propensity to respond to a given type of treatment. The “D−P−” population comprises patients who are not responsive to both active treatment (D) and inactive,placebo treatment (P). Indepression trials innonresistantpopulations, theD−P−group typically represents50% of the populations, based on the nonresponse rates frequently observed with active treatments.5 The “D+P+” population comprises patients who are responsive to either active (D) or placebo (P) treatments. In a very largemeta-analysis of 182 antidepressant clinical trials pooled (N = 36385) fromour group,6 the average placebo response rate in trials with 2 active treatments and placebowas about 40%,which is a proxy for the response rate of the D+P+ group.When the placebo response (and, therefore, theD+P+group) is40%inagivenpopulation and theD−P− group is 50%, thismeans that only 10%of the treated patients belong to the D+P− group, which comprisespatientswho respond toactive treatmentbutnot toplaRelated article page 1087 Opinion