EEG Biomarkers of Response in Depression

Our previous work indicates that a combination of neurophysiologic, symptom, and cognitive measures may predict response more accurately than brain functional measures alone. The purpose of this study is to replicate results from our earlier work using a similar study design (96-06-291-11), and also to prospectively gather additional information to substantiate that a more comprehensive approach to subject assessment will yield more accurate and reproducible prediction of treatment response.

One of the major challenges involved in clinical trials for major depressive disorder (MDD) is that of placebo response. The placebo response rate has been estimated at 20 - 50% of those subjects who enter a standard clinical trial for MDD. This high rate of response to placebo, which may not differ substantially from the response rate to medication, can make it difficult to demonstrate the efficacy of new antidepressant compounds. Identification of MDD subjects with a placebo responder (PR) phenotype, either at the beginning or end of a clinical trial, could have two major potential benefits. First, identification of placebo responders prior to enrollment in a clinical trial might make it possible to have restrictive entry criteria, excluding such subjects from the trial. Segregation of PR phenotype subjects a priori could reduce variance in the outcome data and increase the drug-placebo difference. This exclusion could reduce the number of subjects required for clinical trials and render the trials more efficient. Second, identification of a PR phenotype during a clinical trial could make it possible to distinguish "true medication" from placebo response. This distinction could make it possible to identify subgroups in the trial, enhancing precision in the study of medication effects.

Our research group has performed a series of placebo-controlled treatment trials in MDD and has used a combination of behavioral ratings, self-report, and neurophysiologic measurements with quantitative electroencephalography (QEEG) to identify predictors of both placebo response and medication response. The preliminary results from previous studies suggest that a combination of neurophysiologic, symptom, and cognitive measures may be useful for pretreatment prediction and/or early treatment detection of different types of treatment response.

In this study our primary goal is to assess the neurophysiologic, behavioral, and cognitive assessments of subjects with MDD in the setting of a clinical trial to replicate prospectively these initial results and more completely identify the characteristics of different types of treatment response in a clinical trial.