Identification and development of a five-gene signature to improve the prediction of mechanical ventilator-free days for patients with COVID-19

J.-X. Ni, Y.-B. Qian, Y.-W. Zhang

Emergency Department, The Third People’s Hospital of Hefei, Hefei, China. zhangyunweiahmu@yeah.net

---

• Bioinformatics
• Infectious Diseases

OBJECTIVE: Coronavirus disease 2019 (COVID-19) is a highly contagious infectious disease caused by the newly discovered severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Severe COVID-19 infection causes complications in the respiratory tract, which results in pulmonary failure, thus requiring prolonged mechanical ventilation (MV). An increase in the number of patients with COVID-19 poses numerous challenges to the healthcare system, including the shortage of MV facilities. Despite continued efforts to improve COVID-19 diagnosis and treatment, no study has established a reliable predictive model for the risk assessment of deteriorating COVID-19 cases.

MATERIALS AND METHODS: We extracted the expression profiles and clinical data of the GSE157103, GSE116560 and GSE21802 cohorts from the Gene Expression Omnibus database. Differentially expressed genes (DEGs) were identified as the intersection of the resulting differential genes as analysed via limma, edgeR and DESeq2 R packages. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed using the R package ‘clusterProfiler’. Variables closely related to MV were examined using univariate Cox regression analysis, and significant variables were subjected to least absolute shrinkage and selection operator regression (LASSO) analysis for the construction of a risk model. Kaplan-Meier analysis and receiver operating characteristic (ROC) curves were generated to verify the predictive values of the risk model.

RESULTS: We identified 198 unigenes that were differentially expressed in COVID-19 samples. Moreover, a five-gene signature (BTN3A1, GPR35, HAAO, SLC2A6 and TEX2) was constructed to predict the ventilator-free days of patients with COVID-19. In our study, we used the five-gene signature to calculate the risk score (MV score) for each patient. The results revealed a statistical correlation between the MV score and the scores of the Acute Physiology and Chronic Health Evaluation and Sequential Organ Failure Assessment of patients with COVID-19. Kaplan-Meier analysis revealed that the number of ventilator-free days was significantly reduced in the low-MVscore group compared to the high-MVscore group. The ROC curves revealed that our model had a good performance, and the areas under the ROC curve were 0.93 (3-week ROC) and 0.97 (4-week ROC). The ‘Limma’ package analysis revealed 71 upregulated genes and 59 downregulated genes in the high-MV score group compared to the low-MV score group. These DEGs were mainly enriched in cytokine signalling in immune system and cellular response to cytokine stimulus.

CONCLUSIONS: This study identified a five-gene signature that can predict the length of ventilator-free days for patients with COVID-19.

Free PDF Download