Posttraumatic stress disorder is not a unitary disorder, but rather a heterogeneous syndrome, both in its symptomatology and response to treatment. Although there have been important breakthroughs in our understanding of neurobiological pathways associated with PTSD, its neurobiological heterogeneity has impeded the identification of consistent biomarkers, which remain elusive. A major limitation of previous work is that clinical symptoms and behavioral subtypes of PTSD do not adequately capture variation in the underlying neurobiology, as the same symptoms can stem from different underlying neurobiological mechanisms. To address this critical gap, the proposed study will pioneer a new analytic methodology to identify neurophysiological subtypes, or biotypes of PTSD, based on shared patterns of brain dysfunction in resting fMRI connectivity. The following aims to discover and validate MRI-based biotypes of PTSD in our Veterans have wide-ranging future applications such as improving objective diagnostic tools, providing targets for interventions, and predicting who will response to a particular treatment. DESIGN AND METHODS: The current proposal will use multi-site neural, clinical, and cognitive data to discover resting fMRI-based biotypes of PTSD, determine their reliability and replicability, and relationship to symptoms, comorbidities, and cognition. Veterans' imaging and clinical data will come primarily from the VA Boston (n>500) as well as a replication site at the Houston VA (n>200). An additional 300 participants will be recruited for a state-of-the-art cognitive battery, to determine if these biotypes have cognitive signatures that can be inferred from sensitive cognitive tests. Analytically the research will use a general set of techniques at the forefront of an exciting new era for brain imaging- MRI-based “fingerprinting”, or measuring and modeling the reproducible and yet substantial individual variation in the fMRI-based connectome (functional connectivity). OBJECTIVES. Aim 1. Using existing data (n>500), we will A) Define distinct biotypes of PTSD using patterns of fMRI connectivity. We will further test the reliability of these biotypes across two consecutive resting scans, and the ability to predict PTSD biotype in individual Veterans. We will also explore if certain symptoms and comorbidities differentially cluster with PTSD biotypes. Aim 2. We will determine the across- time reliability of these biotypes by examining repeated assessments of the same participants 1-2 years later (n>300). Aim 3. We will validate these biotypes in an independent replication data set, from the Houston VA (n>200 Veterans). Aim 4. We will collect a comprehensive state-of-the-art cognitive battery (n=300) measuring PTSD-specific mechanisms (e.g., fear learning, inhibitory control, attentional bias) to determine if these biotypes can be inferred from cognitive performance profiles.

Posttraumatic stress disorder (PTSD) is a highly pervasive and debilitating condition in Veterans returning from OEF/OIF/OND. Though PTSD is clearly a distinct disorder, it is heterogeneous both in its clinical presentation and biological underpinnings. The overarching goal of this proposal is to use brain imaging to discover subtypes (biotypes) of PTSD that can be identified in individual Veterans' brains. Discovery of these subtypes could lead to treatment development and better prediction of treatment outcomes. These methods are not only applicable to PTSD, but could be used to better understand other mental health problems.