The Association Between Plasma Homocysteine And In-stent Restenosis In Patients With Coronary Heart Disease Treated With Drug-eluting Stents

Background and ObjectiveCoronary heart disease is the second leading cause of death in China,and it is increasing rapidly.Percutaneous coronary intervention(PCI)is the main method for the treatment of coronary heart disease.China Intervention Network report shows that the amount of PCI is increasing year by year.In 2018,the number of interventions reached 910,000.The clinical application of PCI has greatly reduced the mortality of patients with acute myocardial infarction and improved the clinical prognosis of most patients with coronary heart disease.However,the occurrence of in-stent restenosis(ISR)is one of the main reasons leading to the failure of advanced PCI treatment,and it is also a difficult problem in the interventional field.With the development of devices and drugs,the incidence of ISR has shown a significant downward trend.However,even in the current era of extensive clinical application of drug-eluting stents(DES),the incidence of ISR is still above 5-10%.Previous studies have suggested that biomarkers may provide prognostic information in addition to risk factors such as traditional clinical and angiography.Beside of traditional clinical ISR risk factors,there are few studies on biomarkers of ISR.Exploring the controllable impact biomarkers of ISR and combining them with traditional ISR risk factors may result in a more accurate risk assessment system.It is very important for early identification of ISR,choice of treatment options,and postoperative follow-up arrangements.Dyslipidemia is an important risk factor for the occurrence and development of coronary heart disease,and it is also an important cause of in-stent restenosis(ISR)in patients with coronary heart disease after PCI.Previous studies have confirmed that patients still have a higher risk of ISR after active lipid-lowering treatment,suggesting that beside of dyslipidemia,there may be other potential risk factors that play an important role in the occurrence and development of ISR.Recently,plasma homocysteine(HCY)has been considered as a new intervenable risk factor for cardiovascular disease,which is closely related to the occurrence and development of cardiovascular disease.Experimental studies have found that high homocysteine may cause vascular inflammation,lipid oxidation,promote atherosclerosis,destroy the function of platelets and coagulation system,directly damage vascular endothelial cells,and cause smooth muscle cells and collagen fibers to proliferate.Which may lead to increase the incidence of ISR.However,the relationship between HCY and ISR in clinical evidence-based medicine research is still controversial,and the research evidence of the association between the HCY and DES-ISR is still limited.Additionally,the association between HCY and DES-ISR in different blood lipid levels is still unclear.Folic acid supplementation has a good economic benefit ratio in reducing HCY,but previous studies have controversial conclusions about the effect of folic acid supplementation on ISR.Based on the above research background,this research aims to explore the following four parts:(1)Analyze the independent association between HCY and DES-ISR in patients treated with DES;(2)Explore the relationship between HCY and DES-ISR in different baseline blood lipid levels;(3)To study whether supplementation of folic acid or folic acid combination can reduce the risk of DES-ISR and ISR-related target vessel revascularization;(4)Combine traditional ISR risk factors and related biomarkers to construct a DES-ISR nomogram prediction model and conduct internal verification and clinical application value evaluation.MethodsThis study is a single-center,retrospective longitudinal study.According to the standard a total of 5823 patients with coronary heart disease underwent DES in the Second Affiliated Hospital of Nanchang University from January 1,2015 to October1,2019.The baseline data(demographic information,clinical characteristics,anthropometric indicators,blood biochemical tests and inspection indicators,etc.)were extracted from the big data center platform of the Second Affiliated Hospital of Nanchang University.The Epidata database was established to supplement the receipt of the patient's medication status,angiographic characteristics,and surgery-related information.The data analysis using R 3.4.3 software,Empower software,and SPSS 20.0.The specific content is divided into 4 parts.Part 1: Univariable and multivariable logistic regression were performed to evaluate the independent relation between HCY and DES-ISR.Using generalized additive model(Generalized additive model,GAM)and smooth curve fitting(penalty spline method)to evaluate the true linear correlation between HCY and DES-ISR.Markov-chain Monte Carlo method for multiple interpolation to reduce the influence of the bias of missing variables on the results.Part 2: The baseline blood lipid level was divided into: normal blood lipid group(LDL-C <1.8mmol/L and non-HDL-C <2.6mmol/L)and dyslipidemia group(LDL-C ?1.8mmol/L or non-HDL-C ?2.6mmol/L).A logistic regression analysis was performed to evaluate the relationship between HCY and DES-ISR in different blood lipid levels.In addition,a generalized additive model and smooth curve fitting were performed to explore the dose-response relationship between HCY and DES-ISR in different blood lipid levels.Part 3: Propensity score matching analysis was performed 1:5 matching of baseline data between groups for patients taking folic acid or folic acid combined with themselves,and the matching caliper value was set to 0.05.The main endpoint event were DES-ISR and target vessel revascularization(TLR).The logistic regression was used to calculate odd ratio(OR)and 95% confidence interval(CI),and to evaluate the interaction between different subgroup levels and treatment strategies.Part 4: LASSO regression analysis and literature review were used to screen predictive covariates.Next,we further conduct a multivariate logistics regression analysis on the selected covariates.Receiver operating curve(ROC)evaluation model distinguishing ability,calibration curve calibration chart to evaluate the calibration level.The data is randomly divided into 70% training set and 30%validation set,using Bootstrap to repeat sampling 1000 times for internal model verification.Using C-statistics to compare with traditional predictive models(PRESTO-1,PRESTO-2,EVENT),and perform decision curve(DCA)analysis to evaluate the clinical application value of the nomogram model.ResultsPart 1: A total of 716 patients with coronary heart disease treated with DES were included in the analysis of this study,with an average age of 65.7 years and76.4% of men.The average time for rechecking coronary angiography was 18.5months.Multivariate logistic regression analysis found that for every 1 ?mol/L increase in HCY,the risk of DES-ISR increased by 3%(95% CI: 1.00,1.05)(P =0.042).Smooth curve fitting suggests that HCY and DES-ISR are non-linearly positively correlated.The two-sided logistic regression models showed that the ORs of DES-ISR on both sides of the inflection point HCY 15 ?mol/L were 1.00(95% CI:0.90,1.11),P = 0.956 and 1.03(95% CI: 1.00,1.06),P = 0.042.In the stratified analysis,there was a significant interaction between age grouping status(P = 0.019).Part 2: In this study,682 patients with coronary heart disease who were treated with DES were included(excluding taking folic acid or folic acid combination preparations).In the fully adjusted model,the risk of DES-ISR increased by 17% for every 1 ?mol/L increase in HCY in the normal lipid group(OR = 1.17,95% CI:1.02-1.35;P = 0.030).In the dyslipidemia group,there was no significant increase in the risk of DES-ISR for every 1 ?mol/L increase in HCY(OR = 1.02,95% CI:0.99-1.05;P-interaction = 0.039).Additionally,we converting HCY from a continuous variable to a binary variable,in the normal lipid group,compared with the HCY ? 15 ?mol/L group,the risk of DES-ISR in the hyperhomocysteinemia group increased 9.75 times(OR = 9.75,95% CI: 1.31-72.62;P = 0.026).However,in the dyslipidemia group,compared with the HCY ? 15?mol/L group,there was no significant increase in the risk of DES-ISR in the hyperhomocysteinemia group(OR= 1.16,95% CI: 0.74-1.82;P = 0.525).Smooth curve fitting showed that there was a significant positive linear correlation between HCY and DES-ISR in the normal lipid group,but there was no significant correlation in the dyslipidemia group.Part 3: After matching the propensity score,the number of people in the test group(treated with folic acid or combination of folic acid)and control group(not treated with folic acid or combination of folic acid)were 77 and 385,respectively.Compared with the control group,the test group was not significantly related to DES-ISR and TLR,and the hazard ratios were 8%(OR: 1.08,95% CI: 0.57-2.06;P= 0.804)and 10%(OR: 1.10,95 % CI: 0.51-2.39;P = 0.803).In different subgroups,folic acid supplementation treatment group also did not find significant interaction effects on DES-ISR,including: age,gender,body mass index(BMI),hypertension,diabetes,minimum stent diameter,baseline HCY level,and e GFR.Part 4: A total of 845 participants in this study were followed up for an average of 18.44 ± 11.26 months,and the incidence of DES-ISR was 16.21%.The modeling variables included in the multivariate logistics regression analysis were the number of lesions,ostial lesions,ACC\AHA lesion classification,acute coronary syndrome,and baseline HCY levels.Use the above variables to conduct a predictive model and build a visual nomogram.The calibration curve and Hosmer-Lemeshow test have no statistically significant difference(P >0.05).Use all the data to build the model.The AUC of the four modeling methods are Full=0.735,BS Full=0.719,Stepwise=0.701,BS Stepwise=0.701,the best threshold is 13.33,the sensitivity is 0.63,and the specificity is 0.70.In the verification of random data segmentation,the AUC of the training set is 0.639,and the AUC of the verification set is 0.659.The best threshold in the verification set is 14.1,the sensitivity is 0.61,and the specificity is 0.77.The C statistics of the traditional ISR prediction models PRESTO-1,PRESTO-2 and EVENT in this study population are 0.58,0.59 and 0.58,respectively.The prediction model AUC of this study is 0.68.Compared with the traditional prediction models,our model significantly improves the predict ability of ISR.DCA analysis shows that its net benefit is better than the other three traditional models.ConclusionsThere was a positively relationship between plasma HCY levels and the risk of DES-ISR in patients with coronary heart disease.In patients with normal blood lipids at baseline,HCY ?15umol/L is still an important residual risk factor for DES-ISR.Supplementation of folic acid or a combination of folic acid(0.8 mg)failed to reduce the risk of DES-ISR and TLR,the conclusion need for further large-scale randomized controlled trial research verification.The nomogram model built on the basis of traditional ISR risk factors in combination with HCY has a better predictive effect on the risk of DES-ISR,which is conducive to the assessment and prevention of high-risk patients with ISR in contemporary PCI treatment.