Diagnostic and Prognostic Biomarkers in SARS-CoV-2 Infections

Overview

Literature data document that SARS-CoV-2 RNA is present not only in the respiratory tract but also in the feces of infected patients, suggesting a potential additional route of transmission: the oro-fecal route. In this context, it becomes essential to have data on the use of serological tests in suspected SARS-CoV-2 patients and the presence of viral RNA in biological samples from affected patients, to quickly and reliably identify infected individuals and provide recommendations on the duration of patient isolation. In particular, such data could support the indication for contact isolation similar to that used for all highly contagious gastrointestinal infections, such as Clostridium difficile, with a longer duration than respiratory isolation. The objective of this study is to verify the presence of diagnostic and prognostic biomarkers in patients with SARS-CoV-2 infection.

\- Sample Size Calculation Given the exploratory nature of the study, no formal sample size calculation was performed. A total of 370 patients is expected to be enrolled. \- Data Analysis (as outlined in the approved protocol) Data will be analyzed using t-tests (or Mann-Whitney tests, depending on the data type). The correlation between obtained results and clinical outcomes will be tested using Spearman's rank correlation coefficient to identify potential biomarkers. For microbiota analysis, intra-sample diversity (alpha diversity) will be assessed using Faith's phylogenetic diversity metrics, observed OTUs, and the Shannon index. Inter-sample diversity (beta diversity) will be evaluated using weighted and unweighted UniFrac distances, which will serve as input for principal coordinates analysis (PCoA). PCoA plots, heatmaps, and bar plots will be created using the Made4 and Vegan packages in R. Statistical analysis will be conducted using the Vegan and Stats packages. The separation of data in PCoA will be tested using a permutation test with pseudo-F ratios (Adonis function in Vegan). Fisher's exact test will be used to assess the significance of differences between clusters obtained through hierarchical clustering analysis. The Wilcoxon test (for paired or unpaired data) will be employed to compare alpha and beta diversity, as well as the relative abundance of microbial groups (or functional groups) between subject groups and over time. Discriminatory features (taxa or genes) will be identified using Random Forests (Breiman, 2001). Microbiota sequences from healthy subjects, matched for age, sex, and BMI, will be retrieved from publicly accessible databases for comparative purposes. p-values will be adjusted for multiple comparisons using the Benjamini-Hochberg method. A false discovery rate \<0.05 will be considered statistically significant. Correlations between variables will be assessed using Kendall's correlation test with the cor.test function from the Stats package in R. For microbiota analysis, intra-sample diversity (alpha diversity) will be assessed using Faith's phylogenetic diversity metrics, observed OTUs, and the Shannon index. Inter-sample diversity (beta diversity) will be evaluated using weighted and unweighted UniFrac distances, which will serve as input for principal coordinates analysis (PCoA). PCoA plots, heatmaps, and bar plots will be created using the Made4 and Vegan packages in R. Statistical analysis will be conducted using the Vegan and Stats packages. The separation of data in PCoA will be tested using a permutation test with pseudo-F ratios (Adonis function in Vegan). Fisher's exact test will be used to assess the significance of differences between clusters obtained through hierarchical clustering analysis. The Wilcoxon test (for paired or unpaired data) will be employed to compare alpha and beta diversity, as well as the relative abundance of microbial groups (or functional groups) between subject groups and over time. Discriminatory features (taxa or genes) will be identified using Random Forests (Breiman, 2001). Microbiota sequences from healthy subjects, matched for age, sex, and BMI, will be retrieved from publicly accessible databases for comparative purposes. p-values will be adjusted for multiple comparisons using the Benjamini-Hochberg method. A false discovery rate \<0.05 will be considered statistically significant. Correlations between variables will be assessed using Kendall's correlation test with the cor.test function from the Stats package in R.