Inflammatory profiles revealed the dysregulation of cytokines in adult patients of HFMD

Background: Adult patients of HFMD might act as potential enterovirus reservoirs. As enterovirus infection will cause acute inflammatory response, identifying the association between the dysregulation of cytokines and the development and prognosis of HFMD in adult patients has vital clinical significance. Methods: 60 patients from 266 laboratory-confirmed adult HFMD cases were included in this study, with 40 healthy adult subjects serving as the controls. Social-demographic data were collected through follow-up phone calls. Serum samples were collected from the participants. Enterovirus genotype was tested by RT-PCR, and the expression of cytokines were examined according to the manufacturer’s instructions. Cases were classified using the cytokine profiles with machine learning algorithm. Results: Adult patients of HFMD presented with dysregulation of cytokines. 15 cytokines of adult patients were significantly elevated and 11 cytokines were decreased compared with those of controls. Correlation analysis showed some cytokines have positive correlation with the clinical characteristics and others have negative correlation. All of the enteroviral genotype presented cytokine dysregulation, and five cytokines were significantly different between genotypes. Using a random forest algorithm, we could classify the cytokine profiles into HFMD class and control class with a very high accuracy. Conclusion: These findings suggested that cytokine expression was correlated with the enteroviral infection, genotype and clinical presentation. The inflammatory profiles could be developed as markers to identify HFMD cases with machine learning algorithm. Keywords: Adult HFMD, Cytokines, Inflammation, Random forest