Biomarkers and subphenotypes predictive of clinical outcomes in patients with COVID-19 related acute respiratory distress syndrome

PROJECT SUMMARY/ABSTRACT Acute Respiratory Distress Syndrome (ARDS), caused by a variety of triggers including COVID-19, is a devastating critical illness without effective therapies. Despite numerous studies on COVID-19 related ARDS (CARDS), the best available treatments lead to only modest or inconsistent clinical improvements. This paucity of discoveries is likely in part due to heterogeneity of host response to COVID-19. The long-term goal is to bring precision therapies to patients with ARDS and CARDS. Using plasma protein biomarkers and clinical data, Dr. Calfee (sponsor of the applicant) has identified two biologic subphenotypes of non-COVID-19 related or "typical" ARDS. Specifically, patients with the hyperinflammatory subphenotype had higher levels of inflammatory biomarkers, metabolic acidosis, and shock. They also experienced significantly higher mortality and were more likely to benefit from certain ICU therapies. Conversely, most CARDS patients demonstrate a hypoinflammatory subphenotype of typical ARDS, raising questions about the pathophysiology underlying poor outcomes in this population. The overall objective of this project is to (i) identify plasma biomarkers and differential gene expression associated with poor outcomes in CARDS and (ii) search for the presence of CARDS specific subphenotypes of clinical importance. The central hypothesis is that differential gene expression indicative of dysregulated inflammation and high baseline plasma biomarkers of lung injury and disordered coagulation will predict poor outcomes in CARDS, and that distinct CARDS subphenotypes exist associated with differential outcomes and treatment responses. The rationale for this project is that it will offer mechanistic insights into the biology underlying poor outcomes in this population, lead to identifying potential druggable targets, and identify patients most likely to benefit from targeted therapies. The central hypothesis will be tested by pursuing these specific aims: 1) Identify biological predictors of poor clinical outcomes within patients with CARDS; and 2) Identify distinct biologic subphenotypes of CARDS associated with poor clinical outcomes and differential treatment response. Under aim 1, hypothesis-driven plasma protein biomarkers will be measured, and RNA sequencing performed on blood samples of CARDS patients enrolled in ISPY COVID, an ongoing adaptive randomized controlled trial of experimental drugs across multiple US institutions. Under aim 2, latent class analysis will be performed on a combination of protein biomarkers and clinical variables to identify subphenotypes associated with distinct outcomes and responses to the trial's investigational agents. The proposed research is innovative, because extensive biological phenotyping in a large and representative cohort of patients with a uniform trigger for ARDS has not been done and may be the key to identifying successful therapies. Ultimately, this knowledge will pave the way for conducting phenotype-aware trials studying targeted therapies in patients with subphenotypes of CARDS.