Improved Trauma Survivor Phenotypes Are Critical for Better Biomarkers, Prediction Tools, and Treatments: Initial Results From the AURORA study
McLean SA, An X, House S, Beaudoin FL, Musey P, Hendry P, Jones CW, Lewandowski C, and Storrow A. 338 Improved Trauma Survivor Phenotypes Are Critical for Better Biomarkers, Prediction Tools, and Treatments: Initial Results From the AURORA study. Ann Emerg Med 2019; 74(4):S133.
Ann Emerg Med
Study Objectives: Adverse posttraumatic neuropsychiatric sequelae (APNS) are common among civilian trauma survivors and military veterans. APNS, as traditionally classified, include posttraumatic stress, post-concussion syndrome, depression, and regional or widespread pain. These traditional classifications artificially fragment APNS into siloed, syndromic diagnoses unmoored to discrete components of brain functioning. These traditional classifications are typically studied in isolation, and do not accurately reflect actual posttraumatic neuropsychiatric phenotypes. Most trauma survivors experience complex patterns of overlapping/co-occurring symptoms across multiple traditional classifications, and increasing evidence indicates that co-occurring symptoms can share an interwoven neurobiological substrate. Determining more discrete, homogenous APNS outcomes over time may improve the ability to index APNS to brain function, and categorizing individuals across these outcomes may provide more holistic, accurate phenotyping. Developing such phenotypes is a goal of the ongoing AURORA study, a ∼40 million dollar effort funded by NIH, the DoD, and foundation and industry partners. Methods: 5,000 trauma survivors presenting to 29 EDs for care are enrolled. Core self-report posttraumatic phenotype trajectories are developed using data obtained via serial administration of brief smartphone-based self-report surveys. Biomarkers for these outcomes are identified via serial assessments of neurocognitive, physiologic, digital phenotyping, psychophysical, neuroimaging, and genomic domains. After derivation and validation, these biomarkers will be integrated into outcome phenotype definitions. In the present abstract, the development of core self-report outcomes in the immediate aftermath of trauma is described. The Mindstrong™ App is used to assess symptom indicators for homogenous outcomes across the APNS spectrum (sleep, pain, loss, nightmares, avoidance, re-experiencing, anxiety, hyperarousal, cognition, and somatic). Each indicator group is assessed 6 times during the initial 8 weeks after trauma. Data from an initial study subsample experiencing a common trauma exposure (MVC, n=837) were used to develop measurement models and latent growth curves across timepoints for each outcome. Initial multidimensional outcome groups were developed using latent profile analyses. Results: Construct measurement models provided a good fit to the data (eg, pain CFI 0.99, Loss CFI 0.97). Latent growth curves were developed and mixture model classes created for display purposes (Figure 1). Multidimensional outcomes were identified/selected based on relative model fit and clinical utility; non-recovered groups had markedly different inter-construct profiles. Conclusion: New phenotyping categorizations such as those resulting from the AURORA study have the potential to advance the evaluation and treatment of trauma survivors.