Detecting Antidepressant Response Trajectory Shapes in the STAR*D Cohort

University Of California Los Angeles, Los Angeles CA

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): Major depressive disorder (MDD) is a common illness, associated with high morbidity, mortality, and a very high economic cost for society. Treatment with antidepressant medications is associated with significant improvements in clinical symptoms of depression, as well as improvements in functional status and quality of life. Still, there is marked heterogeneity of outcomes and there are no reliable means of predicting outcomes for the individual patient. One approach to personalizing treatment for depression involves identifying heterogeneity among clinically-relevant outcomes and identifying clinically-viable means to predict those outcomes for the individual patient. Whereas clinical trials typically assess a primary endpoint outcome, more clinically meaningful outcomes include some assessment of the course of changes in symptom severity over time. Personalized medicine approaches are likely to benefit from exploring the possibility of multiple clinically-meaningful response trajectory outcomes. In this vein, growth mixture modeling (GMM), a multilevel modeling technique that incorporates features of cluster analysis, can be applied to longitudinal data to identify latent "classes" or patterns of change in symptom severity over time. Symptom response trajectory shapes may reveal important clinical subgroups. Preliminary GMM studies, for example, suggest that, in addition to a responder class and non-responder class showing monotonic symptom changes, there may be a subgroup of antidepressant-treated subjects that shows a more volatile course of symptom changes (alternating improvement and worsening) during acute treatment. Other groups may exhibit initial but unsustained improvement. Pretreatment baseline clinical and sociodemographic features may moderate treatment outcomes. These potential predictors are among the most easily accessible measures for clinical care. Baseline subject characteristics may prove to be more closely tied to clinical outcomes when those outcomes follow natural (i.e., data-driven) groupings of symptom response trajectories. The objectives of the proposed study are 1) to examine and identify subtypes of clinical response trajectories among MDD subjects treated with antidepressant medication and 2) to characterize subjects in the various response trajectory classes in terms of their baseline characteristics including sociodemographics and clinical features. We will achieve these objectives through secondary analyses of data from the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial. Because of the large size of STAR*D, broad inclusion criteria, and clinically-rich data, this dataset presents an ideal context for applying GMM to "real world" patients and exploring subject characteristics of those classified according to various response trajectories. The identification of clinically meaningful types of antidepressant response trajectories, and the corresponding characterization of persons classified according to those trajectories, could contribute importantly to 'measurement-based care'and prove to be an especially fruitful approach toward personalizing the treatment of depression. Public Health Relevance: Due to the high prevalence of Major Depressive Disorder (MDD), millions in the U.S. are candidates for treatment with antidepressant medications every year, yet some patients do not appear to show significant symptom improvement during treatment, and others may show unsustained or erratic patterns of symptom changes (alternating improvement and worsening). This project will 1) identify various response trajectory outcomes in antidepressant-treated MDD patients through the use of statistical growth mixture modeling (GMM), and 2) characterize subjects in the various response trajectory classes in terms of sociodemographics and baseline clinical features in secondary analyses of data from the NIMH-funded effectiveness trial of "real world" MDD patients: Sequenced Treatment Alternatives to Relieve Depression (STAR*D). By addressing the need to develop clinically-useful means to determine responses to antidepressant medications for the individual patient, findings of this study could be used to enhance `measurement-based care'approaches to personalize treatment for depression based upon patient characteristics.