The Primary Study Of Predictive Value Of LBP Gene Polymorphism，APACHEⅡ Score，and Procalcitonin For Acute Respiratory Distress Syndrome

Background：Acute respiratory distress syndrome（ARDS）,a life threatening lunginflammation,is characterised by poor lung compliance,severe hypoxaemia,and non-cardiacpulmonary oedema with poor prognosis due to lung injury from a variety of direct or indirectrisk factors, such as pulmonary infection, trauma and acute pancreatitis. Despite many largerandomized clinical trials for the treatment of ARDS in recent years, effective therapies toimprove patients’outcome remain limited. Thus, early screening patients at high risk forARDS and timely preventing the development of ALI are expected to reduce the incidence ofARDS and may be more effective in improving outcomes. Unfortunately, there were no idealpredictive models due to the complexity of ARDS pathogenic factors and the various factorsinfluencing the disease progression. The previous studies have proved that depending onpatients’clinical data solely or biomarkers related to lung inflammation and cell damage wasunable to obtain satisfactory predictive value. Therefore, we carried out the study to explorewhether the LBP gene polymorphism,or combined APACHEⅡ score and procalcitonin couldpredict the onset of ARDS.Objective: To explore whether the LBP gene polymorphism or combined APACHEⅡscore and procalcitonin could be better predictors for ARDS.Methods:1. Study design:(1) To detect whether certain SNPs of LBP gene related with higher susceptibility toARDS,222endotoxemia patients admitted to ICU were recruited.Patients were divided into2groups with or without ARDS within the following7days.15SNPs in LBP were obtainedusing snapshot method.(2) To investigate whether combined APACHEⅡ and procalcitonin could be valuablepredictor on the development of ARDS in patients with pulmonary infection,78patients with pulmonary infection in ICU were recruited, and the receiver operating characteristic curvewere used to analyse the data.2. Blood and DNA extraction: Peripheral blood dealing with EDTA anticoagulant werecollected in the first24hours from patients admitted to ICU. Plasma were separated fordetection of endotoxin.When extracted, DNA was stored in-80℃refrigerator before using.3. Detection of15SNPs of LBP using snapshot method.4. The basic information including APACHEⅡscore, age, gender, admission source,chronic complications, primary diagnosis, pathogens, and others were collected and analysed.For pulmonary infection patients, blood gas analysis, erythrocyte sedimentation, C-reactiveprotein, lacticacid, procalcitonin, and the results of sputum culture within1week were alsocollected and analysed.5. Statistical methods: Frequencies for allele or genotype of each SNP were determinedby direct gene counting. Genotype distribution of each SNP was tested for Hardy WeinbergEquilibrium (HWE) and P>0.05represented according with HWE. LD figures wasconstructed by Haploview software. The association between rs2232618and ARDS wasanalyzed using Chi-Square Goodness-of-Fit Test. Logistic regression method was used toadjust for confounding factors such as age, gender, and endotoxin concentration. T test,variance analysis or rank sum test were used to compare difference between groups. Thepredictive value was assessed by calculating the area under a receiver operating characteristiccurve. Combination of two factors was obtained by binary logistic regression methods.Results:1. Among the222patients with endotoxemia ranging from61.0±17.3（17～93）years old,154were male,58were female.49patients (49/222,22.1%) met ARDS within7days;66patients (66/222,29.7%) were with gram-negative bacteria infection in our study；46cases(46/222,21.2%) were with mixed infection or other pathogenic bacteria；110cases (110/222,49.2%) were with negative bacteria results.2. The overall MAF of15SNPs of LBP among222endotoxemia patients was similar tothose from HapMap databank. Genotype distribution of13SNPs were according withHWE(P>0.05), indicating that the allele genotype frequency was constant in the geneticprocess. No significant difference was found between APACHEII scores and rs2232618(P>0.05). 3. Only rs2232618was found to be associated with the susceptibility to ARDS bylogistic regression method (adjusted P=0.002). Patients carrying allele C, including genotypeof CC or CT, were more susceptible to ARDS compared with those without allele C(genotypeTT)(P=0.001).4. Among the78patients with pulmonary infection ranging from65.2±17.6(25～93)years old,60were male,18were female.33patients were diagnosed with ARDS with anincidence of42.3%. Compared with non-ARDS group, APACHEⅡ[(19.9±8.3) vs(15.6±6.0),P=0.013] and procalcitonin[(12.6±11.4)ng/L vs (7.5±10.5)ng/L], P=0.046] weresignificantly higher while the oxygenation index [(109.9±82.6) vs (252.1±92.4), P=0.003]was lower in ARDS group.5. Both APACHEⅡ and procalcitonin could solely predicate the occurrence of ARDS(0.651vs0.657, P=0.929), but no improvement was found when they were combined.Conclusions:1. Patients carrying allele C in rs2232618of LBP gene, including genotype of CC or CT,were more susceptible to ARDS than those without allele C(genotype TT). The rs2232618polymorphism T→C substitution may be an ideal predictor for endotoxin-induced ARDS.2. Both APACHEⅡ and procalcitonin in plasma were associated with the occurrence ofARDS in patients with pulmonary infection and could solely predicate the occurrence ofARDS induced by pulmonary infection. But no improvement was found when they werecombined together.