A Combined Ratio Change Of Cytokines For Sepsis Severity And Clinical Prognosis

Part I Comprehensive analysis of Hub genes among sepsis through integrated bioinformaticsObjective: Sepsis is often accompanied by an alteration of various cytokines.This project aims to comprehensive analysis of Hub genes among sepsis through integrated bioinformatics.Methods: First of all,microarray data of sepsis(GSE69063)was downloaded from Gene Expression Omnibus(GEO)database,including peripheral blood samples of different times from 68 sepsis(severe sepsis: n = 38,mild sepsis: n = 30).Weighted Gene Co-Expression Network Analysis(WGCNA)combined with GEO2 R analysis were used to identify the different Hub genes between severe and mild cases.Then,we identified the functional enrichment pathways as well as interactive proteins through Gene Ontology(GO)enrichment and Protein-Protein Interaction(PPI)analysis,respectively.Besides,the expression of Hub genes in normal tissues was checked in Genotype-Tissue Expression(GTEx)database.Moreover,the single cell subtype enrichment of Hub genes was analyzed in human lung,as well as peripheral blood cells,through Panglao DB single cell database.Lastly,novel potent drugs targeted for Hub genes were checked through Drug-Gene Interaction Database(DGIdb).Results: There was a total of 1,688 differential expression genes(DEGs)between severe and mild sepsis through GEO2 R analysis,including 644 up-regulated and 1,044down-regulated in severe cases.Under the condition of a scale-free co-expression network,a soft threshold β was set as twelve,we selected different correlated modules,and found the turquoise module had the highest Module Significance(MS),which concluded that eigengenes in this module were significantly assoicated with the severe sepsis.The eigengenes in turquoise module were compared to 1,688 DEGs identified by GEO2 R,and Hub genes(Interlukin-6,Interlukin-8,Interlukin-10,and Tumor Necrosis Factors-a)were common genes between them.These Hub genes were significantly higher expressed in severe sepsis.GO Biological Process(GO-BP)enrichment analysis concluded that Hub genes were enriched in pathways including regulation of macrophage activation,immunoglobulin production,and chemokine production.Besides,proteins expressed by the four Hub genes were closely related to each other by PPI network.Afterwards,through GTEx analysis,IL-6 and TNF-a were highly expressed in human normal lung tissue,on the contrary,lowly expressed in peripheral blood cells.The retrieval of single cell Panglao DB database showed that IL-6 in lung tissue was mainly expressed in alveolar type II cells.IL-8 was widely expressed in dendritic cells,ependymal cells,and alveolar type II cells.TNF-a was mainly expressed in dendritic cells.In the peripheral blood cells,IL-8 was widely distributed in T cells,B cells and dendritic cells,while IL-6 and IL-10 were less distributed in each subcellular groups.DGIdb analysis showed that most of the targeted drugs of hub genes were monoclonal antibody antagonists.Conclusion: Hub genes,including IL-6,IL-8,IL-10,and TNF-a,are highly expressed in severe sepsis.The distribution of Hub genes in human normal tissues and single cells are different,but they interact closely and participate in the regulation of human functions together.Part II A combined ratio change of cytokines for sepsis severity and clinical prognosisObjective: Sepsis is often accompanied by a drastic alteration of different cytokines,while cytokines could indicate the disease severity or prognosis in turn.This project aims to determine the association of a combined ratio change of pro-and anti-inflammatory biomarkers at 72 hours after admission and the sepsis severity as well as prognosis.Methods: Based on the Hub genes associated with sepsis through bioinformatic analysis,adult patients diagnosed for sepsis,and septic shock were retrospectively collected from September 2016 to August 2019 at Tongji Hospital Affiliated to Tongji Medical College of Huazhong University of Science and Technology.Patients were divided into two groups,according to their outcomes of hospitalization(survivors and non-survivors).Blood specimens were obtained on admission(T0)and 72h(T72)after therapies,and Hub genes associated biomarkers were compared between two groups.Serum levels of biomarkers were measured by enzyme linked immunosorbent assay(ELISA)at each time point.Through linear regression analysis,the association between cytokines’ level and Acute Physiology And Chronic Health(APACHE II),as well as Sequential Organ Failure Assessment(SOFA)score on admission was compared to determine the relationship between inflammatory factors and sepsis severity.For survivors discharged from hospital,patients were divided into increased and decreased group,according to the changes of inflammatory factors at T72.For their results,the primary endpoint was 9-month overall survival,and the secondary endpoint included the incidence of persistent bacteraemia(PB),methicillin-resistant staphylococcus aureus(MRSA),multi-drug resistant bacteria(MDR),as well as the relative treatments(Surgery,Glucocorticoid,Continuous Renal Replacement Treatment,Vasopressors)dependence.Lastly,the independent risk factors to predict 9-month overall survival were analyzed by Cox regression model analysis.Results: There were a total of 129 cases diagnosed for sepsis(survivors: 55 cases;non-survivors: 74 cases)included into this clinical research,from the department of Intensive Care Unit of Tongji Hospital Affiliated to Tongji Medical College of Huazhong University of Science and Technology.The result of this study did not show any difference in single cytokines between survivors and non-survivors at T0(all P-values ≥ 0.05),but the ratio of IL-10/IL-6 was significantly higher in survivors group(P = 0.0032).Moreover,cytokines corresponding to the four Hub genes(IL-6,IL-8,IL-10,and TNF-a)indicated a positive correlation with APACHEII score(all P-values < 0.05),besides,IL-6 showed a positive correlation with SOFA score alone(P < 0.0001),which indicated that serum levels of cytokines could predict the sepsis severity on admission.After seventy-two hours effective treatments,serum levels of single cytokines were still comparable between survivors and non-survivors group(all P-values ≥ 0.05),but the ratio of IL-10/TNF-a was significantly increased in the non-survivors group(P = 0.0014).For those survivors discharged from hospital,patients in the IL-6 increased(P = 0.0213),as well as IL-10 decreased group(P = 0.0002)after 72 h effective treatments,revealed a better 9-month overall survival,compared to the other groups.Furthermore,after adjusting for a combined ratio change of anti/pro-inflammatory cytokines at T72,patients with a decreased ratio of IL-10/IL-6,IL-10/IL-8,and IL-10/TNF-a(all P-value < 0.0001)showed a significantly higher 9-month overall survival.In addition,an increased ratio of IL-10/IL-6 was significantly associated with the PB(P = 0.005)and MDR(P = 0.023)incidence,and more surgical interventions(P = 0.037)were needed.While,patients with an increased ratio of IL-10/TNF-a also had a higher PB incidence(P = 0.036),and presented a more need for Continuous Renal Replacement Treatment(P = 0.014).Lastly,Cox regression model analysis indicated that SOFA score at T0(P < 0.0001)and IL-10/IL-6 ratio change at T72(P = 0.019)were two independent risk factors for sepsis long-term prognosis.Conclusion: A combined ratio change of pro-and anti-inflammatory biomarkers was an effective predictor for sepsis severity and prognosis.Monitoring the relative cytokines’ level was a reasonable way to predict the therapeutic effects during the whole treatment.