Clinical Study Of Dynamic Stress CT Myocardial Perfusion Imaging In The Diagnosis And Prognosis Of Stable Coronary Artery Disease

BackgroundIschemia with non-obstructive coronary arteries(INOCA)is an important public health problem with an estimated prevalence of 3 to 4 million individuals.Clinical studies have shown the prognosis of INOCA is not benign.Patient with INOCA patients have recurrent angina pectoris,decreased quality of life,and increased risk of major adverse cardiovascular events(MACE).However,clinicians currently have insufficient understanding of the diagnosis,treatment and management of INOCA.At present,the diagnosis of INOCA remains an intractable clinical problem.Invasive diagnostic procedure(IDP)is the gold standard for diagnosing INOCA.However,IDP is difficult to popularize in China in the short term due to its complex operation,high technical requirements for doctors and high price(about 10,500 RMB).Positron emission tomography(PET)myocardial perfusion and magnetic resonance imaging(MRI)have the advantages of absolute quantification of myocardial blood flow(MBF)and myocardial blood flow reserve(MBFR),can be used to diagnose INOCA and identify CMD.However,the widespread clinical use of PET and MRI are limited by its high cost and poor availability.Stress echocardiography(SE)and single photon emission computed tomography myocardial perfusion imaging(SPECT-MPI)is usually assessed by qualitative visual assessment of left ventricular myocardial perfusion and/or)abnormal wall motion.Therefore,diffuse myocardial ischemia may easily be missed by SE and SPECT-MPI.Furthermore,none of the conventional cardiac functional imaging can provide anatomical information of coronary arteries.Therefore,there is an urgent need to establish a non-invasive imaging technique that not only able to qualitatively and quantitatively diagnose INOCA,but also can be widely promoted.In recent years,the combination of coronary computered tomography angiography(CCTA)and computed tomography myocardial perfusion imaging(CT-MPI)enables a"one-stop" non-invasive assessment of coronary artery anatomy and myocardial blood flow,and has the advantages of rapidity,simple operation,high safety profile and the potential for widespread availability.At present,there is only one small-sample,retrospective,cross-sectional study that fails to provide prognostic information.Therefore,more data need to be provided to further verify the feasibility and prognostic value of CCTA combined with dynamic CT-MPI to evaluate INOCA Additional reasons why CT-MPI has not been widely used include concerns about radiation dose and the lack of uniform criteria for diagnosing myocardial ischemia.The performance improvements of the third-generation dual source CT(DSCT)scanner allow acquisition of CT-MPI images at a tube voltage of 70 kV,which can greatly reduce the radiation dose and contrast agent consumption.The performance improvements will promote wide application of CT-MPI technology.There is no consensus among current studies on the optimal cut-off value of MBF for the diagnosis of myocardial ischemia.Therefore,establishing the normal reference value of CT-MPI is an urgent problem to be solved.The variability of MBF is affected by many factors.However,the establishment of a database of normal values of MBF in healthy subjects based on specific scanners and postprocess software remains clinically important for guiding the image interpretation of dynamic CT-MPI and helps to better understand the pattern of myocardial ischemia in coronary artery disease(CAD)and INOCA.Objective1.To describe the distribution of MBF in healthy subjects with dynamic stress CT-MPI scanning protocol based on third-generation DSCT scanner.2.To evaluate the clinical and imaging characteristics of INOCA patients diagnosed by CCTA+CT-MPI,and to explore the related risk factors of INOCA.MethodsThis study prospectively recruited healthy subjects aged 18-60 years from August 2019 to January 2021 to obtain the normal distribution characteristics and normal reference range of MBF derived from dynamic CT-MPI.Patients with suspected or known CAD who underwent CCTA and dynamic stress CT-MPI between January 2017 and June 2021 were consecutively recruited.The patients’ general information,clinical manifestations,complications,and medication history were recorded through the case report form.The patients were followed up by telephone or outpatient clinic every half a year to record the occurrence of MACE.MACE was defined as all-cause death,cardiac death,nonfatal myocardial infarction,readmission for unstable angina,readmission for heart failure,and late revascularization(≥90 days).All healthy subjects and patients underwent a "one-stop" CCTA+dynamic stress CT-MPI examination using third-generation DSCT scanner(SOMATOM Force,Siemens Healthcare,Forchheim,Germany).The degree of coronary artery stenosis,high-risk plaque features were obtained manually.Using a prototype plaque analysis software(Keya Medical,Shenzhen,China)to measure the maximum plaque length,total atheroma volume(TAV),coronary plaque types.The plaques were divided into three types:mixed plaques,calcified plaques,and non-calcified plaques.For quantitative analysis,the region of interest was manually placed to sample the MBF in short-axis according to the American Heart Association’s17 segment model.Obstructive stenosis was defined as≥50%diameter stenosis on CCTA.High-risk plaque features including spotty calcification(SC),low-attenuation plaque(LAP),positive remodeling(PR),and napkin-ring sign(NRS).Abnormal myocardial perfusion was defined as the presence of reduced MBF in at least one myocardial segment.Normal reference range of MBF was defined as mean ± standard deviation × 1.96.It was considered to be reduced MBF if it was less than the lower limit of the normal reference value.Patients were divided into four groups according to whether they had obstructive stenosis and myocardial perfusion abnormalities:Non-ischemic,non-obstructive coronary artery disease group(NINOCA);Ischemia,non-obstructive coronary artery disease group(INOCA);Non-ischemic,obstructive coronary artery disease group(NIOCA),and ischemia with obstructive coronary artery disease group(IOCA).The clinical characteristics,imaging characteristics,and the occurrence of MACE were compared among the groups.Comparisons of continuous variables were performed using independent samples t-test or Mann-Whitney U test.Categorical variable data were compared using the chi-square test.Logistic regression analysis was used to analyze the risk factors associated with INOCA.Survival curves were plotted with the Kaplan-Meier method to show differences in the occurrence of MACE among the groups of patients.All tests were two-sided,andp values<0.05 were considered to be statistically significant.Results1.At the individual level,the vessel level and the segmental level analysis,the normal reference ranges of hyperemic MBF are 117-211ml/100ml/min,116-211ml/100ml/min and 115ml/100ml/min,respectively.According to the normal reference range of hyperemic MBF in healthy subjects and combined with the literature,we finally used MBF<115ml/100ml/min as the criterion for judging the presence of perfusion abnormalities.2.The detection rate of INOCA was 16.0%(54/338)in the entire cohort of 338 patients suspected CAD.Among 211 patients with non-obstructive CAD confirmed by CCTA,the detection rate of INOCA was 25.6%(54/211).The detection rate of INOCA in patients with hypertension was higher than that in patients without hypertension(20.1%vs.10.7%,p<0.05).There was no significant difference in the detection rate of INOCA among patients with different age groups,different genders,and with or without other cardiovascular risk factors(all p>0.05).3.Comparison of clinical characteristics:(1)INOCA group vs.NINOCA group:Age≥60 years old,male and history of hypertension more frequently detected in the INOCA group(all p<0.05).(2)INOCA group vs.NIOCA group:obesity and hypertension were more frequent in the INOCA group.There were no significant differences in age,gender,diabetes,dyslipidemia and smoking history between the two groups(all p>0.05).(3)INOCA group vs.IOCA group:Male,smoking history,dyslipidemia,and at least three cardiovascular risk factors were more frequent in the IOCA group(all p<0.05).The proportion of patients with hypertension and diabetes in the IOCA group tends to higher than that in the INOCA group(p>0.05).4.Comparison of imaging features:(1)INOCA group vs.NINOCA group:the patients in INOCA group had severe stenosis,higher calcification score garde and larger total plaque volume(all p<0.05).PR and mixed plaques were more frequent in the INOCA group(both p<0.05).(2)INOCA group vs.obstructive CAD group(including NIOCA group and IOCA group):calcified plaques,non-calcified plaques,mixed plaques and high-risk plaque(PR and LAP)were more frequent in the obstructive CAD group(all p<0.05).TAV≥250mm3,plaques with length≥30mm were more frequent in obstructive CAD group(NIOCA group and the IOCA group)(all p<0.05).5.Risk factors of INOCA:(1)INOCA group vs.NINOCA group:male,hypertension and the presence of mixed plaques were independent risk factors for INOCA;(2)INOCA group vs.NIOCA group,hypertension was an independent risk factor for INOCA,while calcium score≥100 was risk factors for NIOCA(p<0.05);(3)INOCA group vs.IOCA group:plaque length≥30mm was an independent risk factor for IOCA(p<0.05),while LAP was a potential risk factor for INOCA(p=0.05).6.The median follow-up time was 31 months.MACE occurred in 23 patients.The incidence rate of MACE in NINOCA group,INOCA group,NIOCA group and IOCA group were 1.9%,9.3%,2.2%and 17.3%,respectively.The incidence rate of MACE in the INOCA group was significantly higher than that in the NINOCA group(p=0.019),but the incidence of MACE in the INOCA group was significantly lower than that in the IOCA group(p=0.015),and there was no significant difference in the incidence of MACE between the INOCA and NIOCA groups(p=0.172).Conclusion1.The normal reference range of hyperemic MBF established based on healthy subjects helps to identify INOCA;2.There is no gender difference in the detection rate of INOCA,INOCA was common in both male and female;3.In patients with non-obstructive CAD,male,hypertension and the presence of mixed plaques were independent risk factors for INOCA;4.The INOCA group had a poorer prognosis than that of NINOCA group,but INOCA group had a better prognosis than that of IOCA group.This result strongly supports that INOCA is not a benign disease;5."One-stop" CCTA+dynamic stress CT-MPI examination has the potential to identify INOCA and show promising risk stratification potential for non-obstructive CAD.BBackgroundCoronary CT angiography(CCTA)is a well-established technique in non-invasive coronary artery assessment that yields high sensitivity and negative predictive values in diagnosing obstructive coronary artery disease(CAD).Recently,innovative tools derived from CT,including computed tomography myocardial perfusion imaging(CT-MPI)and fractional flow reserve from computed tomography(CT-FFR),have allowed physiological evaluations of CAD beyond the anatomical information derived from CCTA.Growing evidence suggests that a combination of coronary computed tomography angiography(CCTA)with computed tomography derived flow fractional reserve(CT-FFR)and computed tomography myocardial perfusion imaging(CT-MPI)might have advantages for diagnosing coronary artery disease(CAD).How to optimally implement these noninvasive functional tests in clinical practice is still unclear.Objective(1)To evaluate the consistency between CCTA+CT-FFR+CT-MPI stepwise strategy and invasive ICA/FFR for the diagnosis of functionally significant lesions.(2)To the incremental diagnostic value of CCTA+CT-FFR+CT-MPI stepwise strategy over CCTA+CT-FFR,CCTA+CT-MPI stepwise strategy.MethodsThis retrospective study consecutively included patients who underwent "one-stop"CCTA+CT-MPI examination and completed interventional coronary angiography(ICA)and fractional flow reserve(FFR)measurement for suspected or known CAD between January 2017 and September 2021.All ICA/FFR was performed within 90 days after CCTA+CT-MPI.An intermediate stenosis was defined as an angiographic stenosis diameter of 30%-90%on visual estimation.Invasive fractional flow reserve(FFR)was assessed for vessels with intermediate stenosis(30%to 90%).CT-FFR was computed using a deep learning-based platform.Positive and negative predictive values of CT-FFR and myocardial blood flow(MBF)were calculated using ICA/FFR as the reference standard.The CT-FFR and MBF gray zones were defined by the thresholds for negative and positive predictive values of 95%.A stepwise strategy,in which CT-FFR was restricted to intermediate stenosis on CCTA and CT-MPI was reserved for vessels with gray zone CT-FFR values,was evaluated.For the stepwise diagnostic strategy,the sensitivity,specificity,PPV and NPV were calculated using two approaches:(ⅰ)rule-out approach of considering both the gray zone and ischemia categories as ’positive’,and(ⅱ)rule-in approach of considering only the ischemia category as’positive’.ROC curve analysis was also applied to evaluate the models based on the rule-in and rule-out criteria.In addition,the net reclassification improvement(NRI)was calculated as previously described to quantify the improvements in reclassification among different strategies.A 2-sided p<0.05 was considered statistically significant.ResultsThis study included 87 patients(26 women;mean age,59±10 years)with 211 vessels.Seventy-nine vessels(37%)were diagnosed as functionally significant lesions.This stepwise CCTA+CT-FFR+CT-MPI algorithm theoretically avoided invasive FFR measurement in 78%(93/119)of vessels and avoided CT-MPI examinations in 53%(63/119)of vessels while maintaining 97%agreement with ICA/FFR.This stepwise strategy could rule out functionally significant lesions with a sensitivity and NPV of 92%and 95%and rule in functionally significant lesions with a specificity and PPV of 98%and 97%,respectively.The stepwise CCTA+CT-FFR+MBF strategy showed the highest partial AUC in the range of 90-100%sensitivity for ruling in approach and the highest partial AUC in the range of 90-100%specificity for ruling out approach.The stepwise CCTA+CT-FFR+MBF strategy was superior to CCTA+CT-FFR with an NRI of 0.238(NRI nonischemia+0.136,NRI ischemia+0.101;p<0.001)and superior to CCTA+CT-MPI with an NRI of 0.233(NRI nonischemia+0.106,NRI ischemia+0.127;p<0.001).However,the NRI between the CCTA+CT-FFR and CCTA+CT-MPI strategies was not statistically significant(p>0.05).Conclusions1.Using ICA and FFR as reference standard,deep learning-based CT-FFR and dynamic stress CT-MPI have equivalent diagnostic value in SCAD;2.In diagnosis of SCAD,the stepwise diagnostic algorithm based on CCTA+CT-FFR+CT-MPI has a high consistency(97%)with ICA/FFR in identifying functional ischemic lesions,and can reduce unnecessary radiation exposure and demand for invasive FFR;3.This concept is equally applicable in MBF.The deep learning-based CT-FFR gray zone value is 0.76-0.86.The gray zone of MBF ranged from 86ml/100ml/min to 118ml/100ml/min.BackgroundCoronary computed tomography angiography(CCTA)is an effective tool for risk stratification in patients with suspected coronary artery disease(CAD).However,there is the discordance between severity of coronary stenosis and the myocardial ischemia.Thus,only focusing on the degree of coronary stenosis may expose patients to a significant residual risk of major adverse cardiac events(MACE).Previous studies have shown that high-risk plaque(HRP)features detected by CCTA have incremental prognostic value for MACE in CAD patients independent of traditional risk factors and coronary stenosis.CT-derived fractional flow reserve(CT-FFR)can be used for functional assessment of coronary artery lesions and increase the performance and specificity of CCTA in detection of ischemia.Comprehensive analysis of coronary stenosis degree,high-risk plaque characteristics and CT-FFR information obtained in a single CCTA scan has potential clinical application value for improving risk stratification and predicting MACE in patients with suspected CAD.Dynamic CT myocardial perfusion imaging(CT-MPI)is an effective method to quantitatively measure myocardial blood flow(MBF),and its accuracy in diagnosing myocardial ischemia has been confirmed by many studies.Previous study reported that CT-MPI has an incremental predictive value for MACE based on the prognostic information provided by the severity of coronary stenosis.However,the incremental predictive value of dynamic stress CT-MPI for MACE based on comprehensive coronary stenosis degree,HRP and CT-FFR information is still unclear.In addition,the prognostic value of the stepwise method of CCTA+CT-FFR+CT-MPI proposed in this study for risk stratification in patients with stable CAD has not been evaluated.Objective1.To evaluate whether CT-MPI provides incremental prognostic value for MACE prediction over comprehensive use of coronary stenosis degree,HRP and CT-FFR information for MACE prediction in patients with stable CAD.2.To evaluate the prognostic value of the stepwise CCTA+CT-FFR+ CT-MPI method for risk stratification of patients.MethodsThis study is a prospective,single-center clinical study.A total of 341 patients with suspected or known CAD who underwent CCTA and dynamic stress CT-MPI between January 2017 and June 2021 were consecutively enrolled.The general information,clinical manifestations,comorbidities and medication history of the patients were recorded through the CRF form,and the patients were followed up by telephone or outpatient clinic every six months to record the occurrence of MACE.MACE was defined as all-cause death,cardiac death,nonfatal myocardial infarction,readmission for unstable angina,readmission for heart failure,and late revascularization(≥90 days).All patients underwent a "one-stop" CCTA+dynamic stress CT-MPI examination using third-generation DSCT scanner.The degree of coronary artery stenosis,HRP features and MBF were measured.Obstructive stenosis was defined as>50%diameter stenosis on CCTA.HRP features including spotty calcification,low-attenuation plaque,positive remodeling,and napkin-ring sign.HRP were defined as plaques with at least two HRP features.MBF was calculated using a hybrid deconvolution and maximum slope model model.Hyperemic MBF ≤115ml/100ml/min was defined as a positive.CT-FFR was calculated by deep learning method.CT-FFR ≤0.8 were defined as positive.According to the stepwise CCTA+CT-FFR+CT-MPI approach,the patients were grouped as bellow:(1)The low-risk group was defined as:CCTA showed coronary artery diameter stenosis rate<30%;or CCTA showed coronary artery diameter stenosis rate of 30%-90%and CT-FFR>0.86;or CCTA showed coronary artery diameter stenosis rate of 30%-90%,CT-FFR between 0.76-0.86,MBF>118ml/100ml/min.(2)Intermediate risk group is defined as:CCTA shows coronary artery diameter stenosis rate of 30%-90%,CT-FFR between 0.76-0.86,and MBF is in the range of 86ml/100ml/min-118ml/100ml/min.(3)High-risk group is defined as:CCTA shows coronary artery diameter stenosis rate>90%;or CCTA shows coronary artery diameter stenosis rate of 30%-90%,CT-FFR<0.76;or CCTA showed coronary artery diameter stenosis rate between 30%-90%,CT-FFR between 0.76-0.86,MBF<86ml/100ml/min.Continuous variables were compared using student t test or Mann-Whitney U test.The chi-square test was used for the comparison of categorical variable data.The survival curve was drawn by the Kaplan-Meier method to observe the cumulative incidence of MACE among patients in different groups by each index and stepwise method,the Log-rank test was used to compare the incidence differences of MACE between groups.Cox proportional hazards regression model was used to evaluate the relationship between CCTA,HRP,CT-FFR,MBF and MACE,and the hazard ratio(HR)and its 95%confidence interval(CI)were calculated.Multiple nested multivariate COX regression models were constructed to determine the incremental pronostic value of HRP,CT-FFR,MBF.The C-statistic,net reclassification index(NRI),and integrated discriminative improvement index(IDI)were calculated to evaluate the incremental predictive value of each index.The annualized event rate cut-points used in calculating categorical NRIs were<1%,1%-3%,and>3%.A two-sided P<0.05 was considered statistically significant.Results1.Baseline and follow-up clinical characteristicsA total of 525 patients were enrolled in this study.According to exclusion criteria,184 patients were excluded.Finally,341 cases were included in the analysis.Overall,149 participants(44%)were men,and the median age was 58 years(interquartile range[IQR],51-63).The median follow-up time of this study was 31 months.26 patients(7.62%)developed a MACE during follow-up,including one all-cause death,one non-fatal myocardial infarction,twenty readmission for angina pectoris and four revascularization.2.Comparison of baseline clinical and imaging characteristics of patients in the MACE and MACE-free groups.There were no significant differences in the age,traditional cardiovascular disease risk factors and medication history between subjects with MACE and those without.Male,obstructive CAD,HRP,CT-FFR≤0.8 and MBF ≤115ml/100ml/min were more frequent in the MACE group(all p<0.05).3.Relationship between CCTA、HRP、CT-FFR、CT-MPI and incidence of MACEThe Kaplan-Meier curve showed that incidence rate of MACE was higher in patients with obstructive CAD(14.4%vs 4.0%),HRP(12.3%vs 4.4%),CT-FFR<0.8(17.5%vs 5.4%)and MBF ≤115ml/100ml/min(17.6%vs 1.9%)(all p<0.05).4.Predictive value of CCTA,high-risk plaque,CT-FFR and CT-MPI for MACE.After adjustment for the age,gender and traditional risk factors,the obstructive CAD is an independent predictor of MACE(HR:2.796;95%CI:1.193-6.553;p=0.018).HRP(HR:1.626;95%CI:0.677-3.900;p=0.275)and CT-FFR(HR:1.699;95%CI:0.644-4.478;p=0.284)was no longer significant independent predictors for MACE after adjusting for age,gender,traditional risk factors and obstructive CAD,but MBF ≤115mL/100mL/min(HR:8.954;95%CI:2.812-28.509;p<0.001)was still a significantly independent predictor for MACE.Even after adjusting for age,gender,traditional risk factors,obstructive CAD,high-risk plaques,and CT-FFR,MBF ≤115mL/100mL/min was the only independent predictor of MACE(HR:8.502;95%CI:2.683-26.947;p<0.001).5.Incremental predictive value of high-risk plaques,CT-FFR and CT-MPI.In order to evaluate the incremental predictive value of CT-MPI in predicting MACE,six models were constructed after adjusting for age,gender and traditional risk factors.Model 1:CCTA stenosis;Model 2:CCTA stenosis+HRP;Model 3:CCTA stenosis+CT-FFR;Model 4:CCTA stenosis+HRP+CT-FFR;Model 5:CCTA stenosis+CT-MPI;Model 6:CCTA stenosis+HRP+CT-FFR+CT-MPI.(1)Incremental predictive value of HRP and CT-FFR.There were no significant differences in the C statistics among model 1,2,3 and 4(0.752 vs.0.760 vs.0.755 vs.0.767,p>0.05).Compared with model 1,NRI increases in models 2,3 and 4 were statistically significant(p<0.05).However,IDI of models 2,3 and 4 did not significantly increased compared with model 1.The IDI were 0.003,0.005 and 0.007 for model 2,3 and 4 compared with model 1,respectively(all p>0.05).(2)Incremental value of adding MBF compared with CCTA alone for predicting MACE.Compared to Model 1,the C statistic of model 5 significantly increased(0.819 vs.0.752,p<0.001).Adding MBF to the model 1 significantly improved the prediction of MACE(classification NRI=0.456,95%CI:0.221-0.691,p<0.001;continuous NRI=1.038,95%CI:0.742-1.334,p<0.001;IDI=0.057,95%CI:0.038-0.077,p<0.001).(3)Incremental prognostic value of MBF on top of the CCTA,HRP,and CT-FFR.Compared to model 4,the C statistic of model 6 also increased significantly(0.828 vs.0.767,p<0.001).Adding MBF to the model 4 improved the prediction of MACE(classification NRI=0.496,95%CI:0.263-0.730,p<0.001;continuous NRI=1.038,95%CI:0.742-1.334,p<0.001;IDI=0.059,95%CI:0.040-0.078,p<0.001).There were no significant differences in the C statistics between model 5 and model 6(p>0.05).The change in IDI and NRI were not significant(p>0.05).Model 5 had the lowest AIC among the 6 models.6.Predictive value of stepwise diagnosis for MACEKaplan-meier survival curve showed that the incidence rate of MACE was significantly higher in medium-risk and high-risk groups than in low-risk patients(medium-risk vs.low-risk:17.3%vs.2.53,p<0.001;High-risk vs.low-risk:21.2%vs.2.53%,p<0.001),while there was no significant difference in the incidence of MACE between moderate and high-risk group(p=0.620).After adjusting for age,gender,and traditional risk factors,the risk of MACE was increased in both medium-risk(HR=5.654;95%CI:1.846 17.314;p=0.002)and high-risk groups compared with low-risk group(HR:6.587;95%CI:2.208 19.650;p<0.001).Conclusion1.The degree of coronary artery diameter stenosis and CT-MPI derived MBF can independently predict long-term MACE events in patients with stable CAD,and MBF ≤115mL/100mL/min is a strong predictor of long-term MACE events in patients with stable CAD.2.Hyperemic MBF derived from CT-MPI has incremental value for predicting long-term MACE in patients with stable CAD based on the top of coronary artery diameter stenosis,HRP and CT-FFR.3.Risk stratification for patients with stable CAD by stepwise diagnostic method can predict the incidence of MACE.The incidence of long-term MACE in low-risk group is significantly lower than that in intermediate risk group and high-risk group,and there is no significant difference between the incidence of long-term MACE between the intermediate risk group and high-risk group.