The Role Of Soluble Form Of Suppression Of Tumorigenicity-2 For In-hospital Prognosis And Inflammatory Metabolic Pathway In Pneumonia Patients

Section I The joint action of soluble form of suppression of tumorigenicity-2 and random survival forest model for in-hospital prognosis in pneumonia patients Objective:Pneumonia is a common,potentially fatal disease.The soluble form of the suppression of tumorigenicity-2(sST2)is a biomarker for risk classification and prognosis of heart failure,and its production and secretion in the alveolar epithelium is significantly correlated with the inflammation-inducing in pulmonary diseases.This study investigated the potential value in prognosis and risk classification of sST2 in pneumonia patients.Methods:We measured the serum levels of sST2 in admission and conducted a 30-day cohort study for newly admitted adult pneumonia patients.Prognosis models to predict the 30-day in-hospital prognosis were developed using random survival forest(RSF)method.Results:A total of 247 adult pneumonia patients were studied and 208(84.21%)of them reached clinical stability within 30 days.Patients who did not reach clinical stability had significantly higher levels of sST2 at admission than Patients who did(P < 0.05).The sST2 was an independent predictor of clinical stability and the addition of sST2 to the prognosis model could improve the performance of the prognosis model.The Cindex of the RSF model to predict clinical stability was 0.83(95%CI,0.81-0.86),which is higher than 0.72(95%CI,0.69-0.74)of confusion,urea level,respiratory rate,blood pressure,and age>65 years(CURB-65 score),0.80(95%CI,0.78-0.83)of pneumonia severity index(PSI)score and 0.82(95% CI,0.81-0.83)of proportional hazards(Cox)regression.In addition,the RSF model was associated with adverse clinical events during hospitalization,intensive care unit(ICU)admissions and short-term in-hospital mortality.Conclusions:The serum sST2 level at admission can predict the disease severity and the possibility of patients’ restoring clinical stability of pneumonia patients.We combined sST2 and other clinical indices to construct a well-calibrated RSF model that can evaluate the prognostic risk of adult pneumonia patients and predict whether the patients can reach clinical stability.The performance of the model in prognostic evaluation is better than pre-existing CURB-65 and PSI score.In addition,the introduction of sST2 could improve the prediction performance of RSF model.Section II Combining soluble form of suppression of tumorigenicity-2 and metabolomics to explore the change of illness and inflammatory pathway in pneumonia patients Objective:In the first part of the study,we have demonstrated that sST2 was a potential biomarker to predict the short-term prognosis of in-hospital pneumonia patients.However,the detail mechanism of how the sST2 is involved in the pathological process of pneumonia remains incompletely illuminated.Metabolomics can be used to detect downstream metabolite of biological system.It can quantitatively describe the overall status of endogenous metabolites and the dynamic responses to the changes of both endogenous factors and exogenous factors,reflecting the pathophysiological changes happening in the body.The aim of our study is to screen metabolites that influence the emerging biomarker sST2 and the condition of pneumonia patients by untargeted metabolomic analysis,which will further classify risk and predict prognosis of pneumonia patients.Methods:A total of 71 individuals including 58 pneumonia patients and 13 healthy controls were enrolled.Serum samples were determined by untargeted ultra-high performance liquid chromatography-quadrupole time-of-flight mass spectrometry(UHPLC-Q-TOFMS)-based metabolomics.The levels of sST2 of patients in admission were measured.Pearson’s correlation analysis was performed to assess relationships between the identified metabolites and clinical indices such as sST2.We used the least absolute shrinkage and selection operator(Lasso)algorithm and selected variables to construct proportional hazards(Cox)regression model.Results:Pneumonia patients had higher sST2 levels than healthy controls,and their sST2 levels increased with disease severity.Nineteen metabolites were found to be significantly dysregulated in pneumonia patients,in which 5 metabolites could distinguish severe pneumonia patients from non-severe pneumonia,including Octanoic Acid(OA),5-Hydroxyeicosatetraenoic Acid(5-HETE),12-Hydroxy-8,10-Octadecadienoic Acid(12-OH-8,10-ODDEA),13-Hydroxyoctadecadienoic Acid(13-HODE)and Isodeoxycholic Acid(IDCA).5-HETE and 13-HODE were positively correlated with clinical indices [including sST2,procalcitonin(PCT),neutrophils(NEUT)and D-dimer] and clinical scoring systems [including PSI,CURB-65 and Acute Physiology and Chronic Health Evaluation II(APACHE II)].We constructed Cox regression model with age,D-dimer,sST2,oxygenation index(OI),5-HETE and13-HODE.The model for predicting disease severity,ICU admission and in-hospital death exhibited better Area Under Curve(AUC)(0.94,0.83 and 0.94,respectively,all P<0.05)than PSI,CURB-65 and APACHE II.Conclusions:The study demonstrated that UHPLC-Q-TOF-MS-based metabolomics can be successfully used to identify specific metabolic changes of pneumonia,and established a metabolite signature correlated with disease severity.Moreover,we found that identified metabolites were significantly correlated with sST2.Combining sST2 with fatty acid metabolites can assess disease severity and predict short-term in-hospital outcomes in patients with pneumonia,and might provide a new insight and evidence for the pathophysiology of pneumonia.General conclusions of the study:1.The serum sST2 level at admission can predict the disease severity and the possibility of patients’ restoring clinical stability of pneumonia patients.We combined sST2 and other clinical indices to construct a well-calibrated RSF model that can evaluate the prognostic risk of adult pneumonia patients and predict whether the patients can reach clinical stability.The performance of the model in prognostic evaluation is better than pre-existing CURB-65 and PSI score.In addition,the introduction of sST2 could improve the prediction performance of RSF model.2.The study demonstrated that UHPLC-Q-TOF-MS-based metabolomics can be successfully used to identify specific metabolic changes of pneumonia,and established a metabolite signature correlated with disease severity.Moreover,we found that identified metabolites were significantly correlated with sST2.Combining sST2 with fatty acid metabolites can assess disease severity and predict short-term in-hospital outcomes in patients with pneumonia,and might provide a new insight and evidence for the pathophysiology of pneumonia.