Construction Of CHD Risk Prediction Model In Patients With Hyperlipidemia And Exploration Of The Relationship Between PCSK9 Gene Polymorphism And CH

Background and objective:Cardiovascular diseases are highly prevalent and highly fatal diseases in clinical practice.The incidence rate of these diseases is increasing year by year.The incidence rate of these diseases is becoming younger.Hyperlipidemia is one of the independent risk factors of coronary heart disease(CHD),and early hyperlipidemia is often ignored because of its atypical clinical symptoms.Therefore,the prediction of CHD risk in hyperlipidemia patients is of great significance for reducing the incidence rate and mortality of CHD.This study is committed to constructing the nomogram model of CHD risk prediction in patients with hyperlipidemia,and conducting internal and external validation to evaluate the effectiveness of the model,so as to provide a scientific basis for the screening,prevention and treatment of high-risk groups of hyperlipidemia and coronary heart disease;To clarify the relationship between PCSK9 gene polymorphism and coronary heart disease and blood lipid,and try to provide micro basis for syndrome differentiation of phlegm and blood stasis syndrome of coronary heart disease,and provide clinical basis for the pathogenesis and individualized diagnosis and treatment of CHD.Methods:1 A total of 820 patients with hyperlipidemia admitted to cardiovascular Department 1 and CCU of our hospital from January 2018 to December 2020 were retrospectively analyzed.Collect the following patient information according to the Case Record Form(CRF):Gender,age,BMI,ethnic group,history of smoking,alcohol,history of fatty liver disease,history of diabetes,history of hypertension,cerebrovascular disease history,family history of diabetes,family history of hypertension,family history of hyperlipidemia,family history of coronary heart disease,family history of cerebrovascular disease,carotid ultrasound,intima-media thickness(IMT)on average,an average plaque score,LVEF,HGB,PLT,GLU,BUN,CR,UA,ALB,TC,HDL-C,LDL-C,TG,Clq,HCY,HbAlc.All patients were divided into CHD group and control group according to coronary angiography results or previous clear history.Use binary Logistic regression analysis to screen independent risk factors for CHD,draw a Nomogram,and evaluate the efficacy of the model through internal validation.2 Collect 39 inpatients with hyperlipidemia in our department from January 2021 to February 2022,including 24 patients with CHD and 15 controls.On the basis of the first part of the table to collect clinical data of patients with CRF and external validation Nomogram model for coronary heart disease risk prediction.Use R software to draw clinical decision curves and clinical impact curves to further evaluate the clinical applicability of the model.3 Clinical data and 5ml blood samples were collected from 99 inpatients in our department from January 2021 to February 2022.Genotypes of PCSK9 E670G locus were detected by PCR and MASS spectrometry(ABI3730-XL)in all patients.Population representation was further tested by Hardy-Weinberg genetic balance law to further evaluate the relationship between gene polymorphism and CHD incidence.Results:1 Comparison of general data between the CHD group and the control groupA total of 820 patients were included in this study,according to the intake criteria,including 563 patients in the CHD group and 257 patients in the control group.There were differences in age,sex,ethnicity,smoking history,hypertension,diabetes,family history of hypertension,ultrasound results,mean IMT,plaque score,LVEF,HGB,PLT,ALB,TC,TG,LDL-C,HDL-C,C1q,GLU,BUN,CR,HCY and HbAlc bet-ween the two groups,The difference was statistically significant(P<0.05).2 Results of univariate Logistic regression analysisUnivariate Logistic regression analysis Results of univariate Logistic regression analysis showed that gender,age,ethnicity,smoking history,diabetes,hypertension,family history of hypertension,family history of hyperlipidemia,mean IMT,plaque score,ultrasound results,LVEF,HGB,PLT,GLU,BUN,CR,ALB,TC,TG,HDL-C,LDL-C,C1q,HbAlc were significantly correlated with CHD.(Family history of hyperlipidemia(and PLT P<0.1,all other P<0.05).3 Multi-factor Logistic regression analysis results and model constructionCombined with the univariate Logistic regression results and clinical practice,the multivariate Logistic regression analysis was further carried out.The results showed that gender,age,hypertension,plaque score,LVEF,PLT and HbAlc were independent risk factors for CHD(all P<0.05).A Nomogram model for predicting CHD risk in hyperlipidemia was constructed by converting the continuous variables of these seven factors into categorical variables.H-L test(P=0.575)shows that the goodness of fit of the model is good.The area under the receiver operating characteristic(ROC)curve of the model was 0.881(95%CI 0.858～0.905),P<0.001.When Youden’s index=0.607,the sensitivity and specificity of the model were 79%and 81.7%,indicating that the discrimination of the model was good.The calibration curve of internal verification shows that the curve of the predicted result and the actual result occurrence rate of the nomogram model is very close to the reference line,indicating that the calibration degree of the model is good.4 External validation of the Nomogram model for CHD risk prediction39 clinical patients were selected to externally verify the nomogram model.The verification results showed that the sensitivity of the model was 54.16%,the specificity was 80%,the positive predictive value was 81.25%,the negative predictive value was 52.17%,and the total accuracy was 64.1%.ROC curves were drawn based on the total Nomogram evaluation of each patient for diagnostic variables,with AUC=0.75,95%CI(0.602～0.906),P=0.008<0.01.External validation data suggested that the sensitivity of this model was low and specificity was fair.The calibration curve shows that the curve fluctuation of the predicted result and the actual result incidence is around the reference line,indicating that the calibration degree of the model is acceptable.H-L test results show that P=0.945>0.05,indicating that the model has a good goodness of fit.DCA analysis results showed that the model had good predictive value in the threshold probability range of 40%-60%,and could balance false negative and false positive problems well.The prediction results were basically close to the real situation of patients,and the clinical net benefit was high,indicating that the model had strong clinical applicability.5 The relationship between PCSK9 gene E670G polymorphism and CHD5.1 Gene sequencing showed that there were AA and AG gene types in E670G locus of PCSK9 gene,and there was no significant difference in frequency distribution of AA and AG genotypes between CHD group and control group(P=0.481>0.05).There was no significant difference in the distribution of A and G allele frequencies between the two groups(P=1>0.05).5.2 There were no significant differences in age,BMI and average levels of LVEF,HGB,PLT,GLU,BUN,CR,ALB,HCY and HbAlc in CHD group between AA type and AG type(P>0.05).However,there were statistical differences in UA,TC,TG,HDL-C and LDL-C levels between the two genotypes(P<0.05).The average level of UA and HDL-C in AA type was higher than that in AG type,but the average level of TC,TG and LDL-C in AG type was higher than that in AA type.5.3 Between the two syndromes of phlegm-stasis and non-phlegm-stasis syndrome,there was no statistical significance in age,BMI and average levels of LVEF,HGB,PLT,GLU,BUN,CR,UA,ALB,TC,TG,LDL-C,HCY and HbAlc(P>0.05).However,there was statistical difference in the average HDL-C level between the two syndrome types(P=0.001<0.05),and the average HDL-C level of the syndrome of phlegm-stasis interaction was lower than that of the syndrome of non-phlegm-stasis interaction.Conclusions and Significance:1 Based on carotid ultrasound data and clinical data of hyperlipidemia patients,a nomogram model for CHD risk prediction of hyperlipidemia patients was established to screen out 7 risk factors including gender,age,hypertension,plaque score,LVEF,PLT and HbAlc.This model can be used to identify high-risk patients with CHD and guide the formulation of individualized diagnosis and treatment plans.After internal and external verification and clinical decision analysis(DCA),this model has good predictive ability and high clinical application value.2 In this study,it was found that the polymorphism of E670G locus of PCSK9 gene was not significantly associated with the incidence of CHD,but was correlated with blood lipid level,and HDL-C level may be related to the formation of phlegm-stasis syndrome,providing clinical basis for further research on the upstream and downstream targets and mechanisms of PCSK9.