Strategies to Minimize Heterogeneity and Optimize Clinical Trials in Acute Respiratory Distress Syndrome (ARDS): Insights from Mathematical Modeling

Abstract

Background: Mathematical deconvolution of multiple biologic pathways simultaneously active in critical illness may aid in the interpretation of clinical data.Methods: Using a systems biology model of COVID-19, we analyze the effect of immunomodulation on diverse patient phenotypes. We define optimal treatment and optimal treatment initiation time, identify biologic programs which determine heterogeneous clinical responses and predict biomarkers of those programs. Finally, we evaluate the utility of model biomarkers for identifying high-risk patients in a clinical database of COVID-19 patients.Findings: We predict patient phenotypes with disparate clinical responses to the same therapy. Older and hyperinflammed patients respond better to immunomodulation than obese and diabetic patients. Optimal treatment initiation time is driven by neutrophil recruitment dynamics, extra-pulmonary cytokine expression, systemic microthrombosis and the renin-angiotensin system (RAS) in older patients, and by RAS, systemic microthrombosis and trans IL6 signaling for hyperinflamed patients. For older and hyperinflamed patients, IL6 modulating therapy is optimal when initiated very early (th day of infection) and broad immunosuppression therapy (corticosteroids) is better later in the disease (between 7th– 9thday of infection). We show that biologic programs identified by the model correspond to clinically identified markers of disease severity,Interpretation: Modeling of COVID-19 pathobiology suggests combinations of patient phenotypes which may lead to disparate outcome when included in a trial with a single intervention protocol. Systems biology modeling thus has potential for suggesting refinements to clinical trial design by aiding in the identification of biologic mechanisms underlying clinical heterogeneity and predicting markers of those mechanisms.Funding: Jain’s research is supported by NIH grants R35-CA197743, U01-CA224348, R01-CA259253, R01-CA208205 R01-NS118929, National Foundation for Cancer Research, Ludwig Cancer Center at Harvard; Advanced Medical Research Foundation and Jane's Trust Foundation. Lance Munn’s research is supported by R01-CA2044949. No other funding to declare. Declaration of Interest: LLM owns equity in Bayer AG and consultant for SimBiosys. Jain received an Honorarium from Amgen; Consultant fees from Chugai, Elpis, Pfizer, SPARC, SynDevRx; Owns equity in Accurius, Enlight, SynDevRx; Board of Trustees of Tekla Healthcare Investors, Tekla Life Sciences Investors, Tekla Healthcare Opportunities Fund, Tekla World Healthcare Fund and received a research Grant from Boehringer Ingelheim. All others have nothing to declare. Ethical Approval: This analysis was approved by the MGB Institutional Review Board (IRB protocol #2020P000964).