Predictive Role Of Atrial Cardiomyopathy Markers In Prognosis Of Patients With Acute Myocardial Infarction

Background and objective: Acute myocardial infarction(AMI),one of the leading causes of death worldwide,has a poor prognosis despite advances in reperfusion treatment.Atrial cardiomyopathy(ACM),a complex of structural,architectural,contractile or electrophysiological changes affecting the atria,has been considered an entity preceding the presence of AF and stroke in general population.Although the diagnostic criteria for ACM is not clear at present,several markers including P-wave terminal force in ECG lead V1(PTFV1),left atrial enlargement(LAE)and B-type natriuretic peptide(BNP)have been used to identify ACM.Large-sample prospective studies have shown that markers of ACM can predict the onset of atrial fibrillation as well as ischemic stroke independent of atrial fibrillation in community population.To the best of our knowledge,the relationship between ACM markers and prognosis of patients with AMI has not been adequately addressed.Therefore,we investigated the relationship between ACM markers and the prognosis of patients with AMI from three aspects: ACM markers and new onset atrial fibrillation(NOAF),ACM markers and Ischemic cerebrovascular events(ICVE)as well as ACM markers and major adverse cardiovascular events(MACE)after AMI.Method: Part 1We retrospectively analyzed 4713 consecutive patients with AMI without a documented history of atrial fibrillation(AF)who underwent coronary angiography between January 2016 to June 2021.We measured markers of ACM including PTFV1,Left atrial dimension(LAD),and BNP.Patients were stratified into tertiles of PTFV1,LAD,and BNP levels.Associations between markers and NOAF were evaluated using logistic regression analysis.By using risk factors for incident AF defined in the 2020 ESC guideline and risk factors reported from previous studies,we further constructed a multivariable logistic regression model(clinical model)to predict NOAF.Receiver operating characteristic(ROC)curves,area under the ROC curve(AUC),reclassification improvement(NRI)and integrated discrimination improvement(IDI)were used to estimate the predictive value of the combined model.The calibration curve analysis and decision curve analysis(DCA)were applied to evaluate the agreement between the predicted probability and the actual probability as well as clinical usefulness of the combined model.Part 2We retrospectively analyzed 4206 consecutive patients with AMI who underwent coronary angiography between January 2016 to June 2021.Patients were assessed for ACM markers including BNP,PTFV1 and LAD.Left atrial enlargement(LAE)and PTFV1 enlargement was defined by previously published cut-off points.The primary outcome was incident ICVE composed of ischemic stroke(IS)and transient ischemic attack(TIA).Receiver operating curve(ROC)analyses,NRI and IDI were used to compare the predictive performance of the CHA2DS2-VASc score combined with ACM markers to the CHA2DS2-VASc score alone.Part 3We retrospectively analyzed 3902 consecutive patients with AMI who underwent coronary angiography between January 2016 to June 2021.We measured markers of ACM including PTFV1,LAD and BNP.Patients were stratified into quartiles based on PTFV1,LAD,and BNP levels.We compared prognosis between patients in the highest quartile and those in other quartiles of the ACM markers.The primary composite outcome was major adverse cardiovascular events(MACE)including: all-cause death,heart failure-induced readmission,recurrent myocardial infarction,target vessel revascularization(TVR)and ischemic stroke.Multivariate Cox regression was used to assess the relationship between ACM markers and MACE.The predictive value of the ACM markers combined with the GRACE(Global Registry of Acute Coronary Events)risk score was estimated by C-statistics,NRI and IDI.The calibration curve analysis and decision curve analysis were applied to evaluate the agreement between the predicted probability and the actual probability as well as clinical usefulness of the combined model.Results:1、Overall,222(4.71%)patients had NOAF out of 4713 patients.The prevalence of NOAF increased gradually with PTFV1,LAD,and BNP tertiles.On multivariable regression analysis with potential confounders,elevated PTFV1(OR per 1000μV*ms,1.184;95% CI,1.126-1.246,P<0.001),LAD(OR per millimeter,1.090;95% CI,1.050-1.131,P<0.001)and Log BNP(OR=1.269;95% CI,1.146-1.404,P<0.001)were significantly associated with an increased risk of NOAF.The addition of PTFV1,LAD,and BNP to the AF risk factors recommended by the 2020 ESC Guidelines significantly improved risk discrimination for NOAF with AUC increasing from 0.762 to 0.810(p<0.001).The calibration curve showed good prediction agreement with actual observation.Decision curve analysis demonstrated that the combined model would be clinically useful.2、During a median follow-up of 44.0 months,229(5.44%)ICVE occurred.Of these,156 individuals developed IS and the remaining 73 cases were diagnosed with TIAs.The ICVE group showed larger PTFV1 and increased LAD as well as elevated BNP levels at baseline.In the multivariate analysis,we found significant associations with ICVE for PTFV1(HR per 1,000 μV*ms,1.140;95% CI,1.090-1.193),LAD(HR per millimeter,1.147;95% CI,1.106-1.189),but not Log BNP(HR=1.053;95% CI,0.957-1.159,P=0.288)after adjusting for known ICVE risk factors and interim AF.The addition of PTFV1 enlargement and LAE improved the predictive accuracy of the CHA2DS2-VASc score with C-statistic increasing from 0.708 to 0.761(p<0.001).3、MACE occurred in 766(19.6%)of the 3902 patients at a median of 45 months.On multivariable regression analysis with potential confounders,elevated PTFV1(HR per 1000μV*ms,1.032;95% CI,1.004-1.062,P=0.026)、LAD(HR per millimeter,1.067;95% CI,1.045-1.089,P<0.001)和 Log BNP(HR=1.119;95% CI,1.070-1.171,P<0.001)were significantly associated with an increased risk of MACE.The addition of the ACM markers to the GRACE score had an incremental effect on the predictive value for MACE[1-year,3-year and 5-year C-statistic increasing from0.631(0.605-0.657),0.564(0.550-0.578),0.549(0.538-0.560)to 0.676(0.652-0.700),0.608(0.594-0.621),0.566(0.555-0.577),respectively(p<0.001)].The calibration curve showed good prediction agreement with actual observation.Decision curve analysis demonstrated that the combined model would be clinically useful.Conclusion:1、ACM markers were strongly associated with NOAF after AMI.The prediction performance of the clinical model for NOAF was increased by the addition of these markers.2、ACM markers were associated with incident ICVE independent of well-established risk factors and AF occurrence.The addition of ACM markers with CHA2DS2-VASc score may well discriminate individuals at high risk of ICVE in AMI patients.3、ACM markers were strongly associated with MACE.The prediction performance of the GRACE score was increased by the addition of these markers.4、Assessment of ACM markers may add further information in the prediction of outcome in patients with AMI.ACM markers can be used as a reliable tool for prognostic stratification in AMI patients.