While the majority of coronavirus disease 2019 patients develop a mild disease, up to 20% become severely ill, with a severe interstitial pneumonia with high levels of acute phase mediators (cytokine storm) and other complications. There is a lack of knowledge on the role of individual genetic variability in conferring differential viral susceptibility, response to treatments, and severity of disease. This study aims at addressing this question, to identify factors predictive for the different evolution of the disease.
After performing Whole Genome Sequencing in 200 Covid19 patients, stratified on disease severity (100 each with mild and severe symptoms), response to therapies and presence of co-morbidities, genetic, clinical, and laboratory datasets underwent biased and unbiased burden tests for rare pathogenic variants, in addition to machine learning (ML) and genome-wide association (GWAS) analyses for common predisposing or protecting variants
ML confirmed CRP as the major factor discriminating patient status and showed that IL18 variants accumulation correlates with mild patients, thus reflecting a protective role of this cytokine with respect to severe Covid19. GWAS confirmed presence of more variants than expected by chance in the IL18 gene, in addition to replicate the already known involvement of SLC6A20/LZTFL1 SNP rs35081325 in severe versus mild Covid19
Finally, functional tests revealed a role of the complement activation through increased levels of C5a and C5b9 levels, found to be predictive for adverse outcomes
To clarify the pathogenic mechanisms inducing either a severe outcome or mild disease in patients affected with Covid-19, in the past 2 years we have collected, in the Reggio Emilia Hospital Unit (UO2), a large set of whole blood and serum/plasma samples from 100 Covid-19patients who did not require hospitalization (mild symptoms) and 100 Covid-19 patients who were hospitalized (severe symptoms). DNA samples thus extracted were transferred to the IIT Unit (Genoa, Italy – UO4) where they have been subjected to Whole Genome Sequencing (WGS). Primary data analysis, variants calling and further genomic, statistical and AI analyses have been carried out in the IIT and IGG Units (Genoa, Italy – UO1).
Clinical, laboratory and genetic datasets underwent i) search and analysis of rare variants, possibly responsible for congenital conditions able to modulate the response to the SARS-CoV-2 infection, ii) machine learning, an artificial intelligence approach, to identify relevant biological markers and molecular signatures, iii) a genome wide association study (GWAS) to identify common variant possibly responsible for increased or decreased susceptibility to severe Covid-19. To this end, patients were grouped based on disease severity, response to the therapies, presence of pre-existing morbidities, such as rheumatologic chronic diseases, and familial clustering.
Functional tests focusing on the complement activation were carried out in 97 patients (54 mild and 43 severe disease) (Istituto Auxologico, Milan, Italy – UO3). The disease severity was associated with therapy independent increased levels of C5a and C5b9 levels suggesting that complement activation products may be predictive for the negative outcome.