(CCSP001) MACHINE LEARNING ANALYSIS OF LONG COVID MULTIOMICS REVEALS CLINICAL PHENOTYPES AND PROGNOSTIC BIOMARKERS

Background: An estimated 10-30% of individuals convalescing from SARS-CoV-2 infection continue to experience long COVID, characterized by fatigue, sleep disturbance, confusion, and dyspnea, alongside many other debilitating symptoms resulting in significant impairments in their quality of life. The risk of developing new diagnoses of pulmonary, cardiovascular, gastrointestinal, metabolic, psychiatric, and nervous system disorders was greatly elevated, associated with higher hospitalization rates and worse prognoses at six months post-infection. Although female sex, pre-existing comorbidities, and severity of the acute infection have been proposed as risk factors for PASC, the underlying cause behind such heterogeneity in disease sequelae is not yet understood.

METHODS AND RESULTS: To better understand the pathogenesis and elucidate therapeutic targets of long COVID, we obtained plasma samples from 117 individuals during and six months following their initial COVID-19 hospitalization to comprehensively profile and assess changes in cytokines, proteome, and metabolome. Repeat blood sampling was performed over a median duration of 6.3 (IQR: 6.0 - 7.1) months. Our results revealed that as patients transition from acute to long COVID, there remains a persistently high burden of symptoms and adverse clinical outcomes. Network analysis revealed sustained inflammatory response, platelet degranulation, and cellular activation in long COVID accompanied by dysregulation in arginine biosynthesis, methionine metabolism, taurine metabolism, and tricarboxylic acid (TCA) cycle processes. Unsupervised clustering analysis unveiled three distinct clinical phenotypes, with cluster B characterized by a predominance of triglyceride and organic acid signature. In comparison, cluster C had a higher proportion of women and more frequently reported symptoms such as insomnia, palpitation, dyspnea, weakness, and fatigue. Furthermore, we developed a prognostic model composed of 20 molecules involved in regulating T-cell exhaustion and energy metabolism that can reliably predict adverse clinical outcomes following discharge from COVID-19 hospitalization with 83% accuracy and an area under the curve (AUC) of 0.96.

Conclusion: Given the lack of proven effective therapies for long COVID, our results point toward several potential therapeutic targets that may be explored in future clinical trials. Firstly, persistent immune activation can impair wound healing and contribute to neuroinflammation. Secondly, the higher risk of thromboembolic disorders observed in long COVID, together with abnormal platelet degranulation and blood coagulation processes, supports the use of anticoagulant therapy. Lastly, global metabolomic analyses revealed specific alterations in methionine metabolism and the TCA cycle, suggesting a potential role of antioxidants and treatment strategies to support mitochondrial function and energy production.