ABSTRACT
Background: Accurate prognosis is important either after acute infection or during long-term follow-up of patients infected by severe acute respiratory syndrome coronavirus 2. This study aims to predict coronavirus disease 2019 (COVID-19) severity based on clinical and biological indicators, and to identify biomarkers for prognostic assessment. Methods: We included 261 Vietnamese COVID-19 patients, who were classified into moderate and severe groups. Disease severity prediction based on biomarkers and clinical parameters was performed by applying machine learning and statistical methods using the combination of clinical and biological data. Results: The random forest model could predict with 97% accuracy the likelihood of COVID-19 patients who subsequently worsened to the severe condition. The most important indicators were interleukin (IL)-6, ferritin and D-dimer. The model could still predict with 92% accuracy after removing IL-6 from the analysis to generalise the applicability of the model to hospitals with limited capacity for IL-6 testing. The five most effective indicators were C-reactive protein (CRP), D-dimer, IL-6, ferritin and dyspnoea. Two different sets of biomarkers (D-dimer, IL-6 and ferritin, and CRP, D-dimer and IL-6) are applicable for the assessment of disease severity and prognosis. The two biomarker sets were further tested through machine learning algorithms and relatively validated on two Danish COVID-19 patient groups (n=32 and n=100). The results indicated that various biomarker sets combined with clinical data can be used for detection of the potential to develop the severe condition. Conclusion: This study provided a simple and reliable model using two different sets of biomarkers to assess disease severity and predict clinical outcomes in COVID-19 patients in Vietnam.
ABSTRACT
Highly methylated Long Interspersed Nucleotide Elements 1 (LINE-1) constitute approximately 20% of the human genome, thus serving as a surrogate marker of global genomic DNA methylation. To date, there is still lacking a consensus about the precise location in LINE-1 promoter and its methylation threshold value, making challenging the use of LINE-1 methylation as a diagnostic, prognostic markers in cancer. This study reports on a technical standardization of bisulfite-based DNA methylation analysis, which ensures the complete bisulfite conversion of repeated LINE-1 sequences, thus allowing accurate LINE-1 methylation value. In addition, the study also indicated the precise location in LINE-1 promoter of which significant variance in methylation level makes LINE-1 methylation as a potential diagnostic biomarker for lung cancer. A serial concentration of 5-50-500 ng of DNA from 275 formalin-fixed paraffin-embedded lung tissues were converted by bisulfite; methylation level of two local regions (at nucleotide position 300-368 as LINE-1.1 and 368-460 as LINE-1.2) in LINE-1 promoter was measured by real time PCR. The use of 5 ng of genomic DNA but no more allowed to detect LINE-1 hypomethylation in lung cancer tissue (14.34% versus 16.69% in non-cancerous lung diseases for LINE-1.1, p < 0.0001, and 30.28% versus 32.35% for LINE-1.2, p < 0.05). Our study thus highlighted the optimal and primordial concentration less than 5 ng of genomic DNA guarantees the complete LINE-1 bisulfite conversion, and significant variance in methylation level of the LINE-1 sequence position from 300 to 368 allowed to discriminate lung cancer from non-cancer samples.