COVID-19 and cholesterol biosynthesis: Towards innovative decision support systems

Area of Science:

• Biochemistry
• Infectious Diseases
• Medical Diagnostics

Background:

• COVID-19 (Coronavirus Disease 2019) requires ongoing research for effective patient stratification and biomarker discovery as it becomes endemic.
• Understanding host-pathogen interactions, including metabolic pathways like cholesterol biosynthesis, is crucial for managing COVID-19 severity.

Purpose of the Study:

• To investigate the association between cholesterol biosynthesis intermediates and COVID-19.
• To develop and evaluate predictive models for COVID-19 disease severity using clinical parameters and sterol biomarkers.

Main Methods:

• Compared concentrations of 10 cholesterol biosynthesis intermediates in 164 hospitalized COVID-19 patients at different disease stages.
• Developed machine learning models using 8 clinical parameters and a subset of sterols to predict disease severity.
• Validated model performance using Area Under the Curve (AUC) and compared it with existing clinical risk scores like COVID-GRAM.

Main Results:

• Significant alterations in zymosterol, 24-dehydrolathosterol, desmosterol, and zymostenol levels were observed in COVID-19 patients.
• Machine learning models based on clinical parameters alone achieved excellent accuracy (AUC = 0.96) in predicting disease severity.
• Incorporating sterol levels into the models further enhanced predictive performance compared to existing scores.

Conclusions:

• This study is the first to demonstrate a link between cholesterol biosynthesis pathways and COVID-19 severity.
• Specific sterols are significantly associated with COVID-19 disease severity and can serve as potential biomarkers.
• Combining clinical data with sterol biomarkers offers a promising approach for improved COVID-19 patient stratification and outcome prediction.