Strain Analysis Using Feature Tracking Cardiac Magnetic Resonance Predict Prognosis of Patients With Ventricular Aneurysm After Myocardial Infarction

Objective We aim to assess the left ventricular strain in patients with ventricular aneurysm(VA) after myocardial infarction(MI) using cardiac magnetic resonance-feature tracking (CMR-FT) and to evaluate its value for long term prognosis of patients. Methods Sixty-ve patients who underwent CMR with VA after MI from January 2018 to December 2019 in Drum Tower Hospital Aliated Hospital of Nanjing University School of Medicine were selected for the study. They were divided into two groups based on New York Heart Association (NYHA): 25 cases of NYHA I as group A and 40 cases of NYHA II-IV as group B. CMR was performed in both groups to quantify the parameters of overall and segmental left ventricular myocardial strain in patients with aneurysm. 37 of whom underwent a second CMR 3-12 months after cardiac infarction to investigate the effects of aneurysm on patients' left ventricular strain and left ventricular cardiac function. Results Patients from group B have larger VA basilar transverse diameter and signicant more impaired LV Global longitudinal strain(GLS) (cid:0) Global circumferential strain(GCS) (cid:0) Global radial strain(GRS) (-12.34±7.31 vs. -7.68±6.11;p=0.0072, -21.31±13.49 vs. -14.93±10.44;p=0.0361, 37.13±27.87 vs. 22.00±20.05;p=0.0135) without change in infarct size. GLS, GCS, GRS were signicant indicators of NYHA classication after AMI by multivariate regression analysis. Conclusions Myocardial strain assessed by CMR-FT may be an independent predictor of NYHA of patients with aneurysm after MI and could be used for identifying high-risk patients with VA.


Abstract
Objective We aim to assess the left ventricular strain in patients with ventricular aneurysm(VA) after myocardial infarction(MI) using cardiac magnetic resonance-feature tracking (CMR-FT) and to evaluate its value for long term prognosis of patients. Background Ventricular aneurysm(VA) is a common mechanical complication of myocardial infarction(MI) and can be accompanied by ventricular appendage thrombosis, valvular regurgitation, ventricular wall rupture, ventricular tachycardia or sudden cardiac death, with a high mortality rate and a poor clinical prognosis for patients 1 . Therefore, it is important to investigate how to e ciently and accurately detect VA and to investigate their relationship with left ventricular function, in order to provide better guidance for further treatment by surgical treatment or medical drugs.
Cardiac magnetic resonance imaging(CMI) is more accurate than echocardiographic in detecting cardiac anatomy and function, and was recently suggested also to have an important role in pre-and postsurgical evaluation of patients with left VA 2 . CMR cine images is proposed as the most accurate and reproducible method of measuring left ventricular ejection fraction (LVEF) and volumes. LVEF can be used to evaluate the function of heart valve, especially to explore the relationship between the atrioventricular valves to the aneurysmal structure 3 . However, LVEF can not early identify heart failure patients with preserved ejection fraction, nor can it re ect the changes of regional myocardial systolic function. Myocardial strain, de ned as the percentage change of myocardial dimension in a speci c direction, is an important parameter to assess myocardial performance as it adds independent value to Page 4/15 other well-recognized prognosticators 4,5 . CMR myocardial strain technique can quantitatively evaluate GLS and GCS beyond the LVEF, and has higher sensitivity in the detection of myocardial dysfunction. CMR-FT technique has been validated against the gold standard myocardial tagging and is now considered a preferred CMR solution for strain assessment 6 . However, the role of CMR-FT technique in strain evaluation of left VA is still unknow.
The purpose of this study is to assess strain changes in patients with left VA after MI from January 2018 to December 2019 at Drum Tower Hospital and evaluate clinical value for patient prognosis compared with conventional LVEF.

study population
For the current study, the population includes patients with CMR suggestive of MI with VA from January 2018 to December 2019 in Drum Tower Hospital, A liated Hospital of Nanjing University medical school. All patients were con rmed by coronary angiography, electrocardiogram (ECG) and echocardiography. There were 52 men and 13 women with a mean age of 61.8±11.2 (46-84) years; The extent of the patient's VA includes the apical, lateral, anterior and inferior walls of the left ventricle.

Clinical data
Clinical and demographic characteristics of patients were recorded as previous reported 7 . Diabetes mellitus was de ned as having a history of diabetes mellitus and currently medical therapy with insulin, oral hypoglycemic drugs. Hypertension was de ned as previous use of antihypertensive medications or diagnosed with hypertension. Smoking was de ned as currently or previously smoking. During invasive coronary angiography, the culprit vessel was identi ed and multivessel disease was de ned as more than one vessel with >50% luminal stenosis.

Cardiac function analysis
The CMR images were analyzed on commercially available workstation (IntelliSpace Poral (ISP), Philips Healthcare). The LV epicardial and endocardial contour were delineated semi-automatically based on SA cine images. The left ventricular end-diastolic volume (LVEDV), left ventricular end-systolic volume (LVESV), myocardial mass, LVEF, cardiac out-put (CO), peak ejection rate, rst peak lling rate and second peak lling rate were measured using CMR function package within ISP workstation. Spatial enhancement analysis with percentage of the entire LV myocardial mass were performed to visualize and quantify the transmural extent of infraction based on LGE images. The parameters of ventricular aneurysm are obtained based on the largest plane of VA in the cine images.
The infarct size (IS) was calculated as previous decribed 8 . Brie y, the infarcted regions were de ned as hyper-enhanced regions with + 5SD signal intensity above the normal remote myocardium. The IS was then expressed as the percentage of left ventricular volume mass. A representative gure shows IS measurement were shown in Figure 1.
The measurement of the parameters of VA is mainly based on cine images. The largest plane showing the largest VA during systole is selected as the measurement slice. The diameter of VA is measured three times: 1) the transverse diameter of the base of AV body (Aneurismal-W, de ned as the short diameter of the base section of ventricular wall tumor during the selected plane contraction). The speci c measurement method is to measure the basal diameter of VA during contraction of four-chamber, twochamber or short axial position, and the shortest diameter is taken as its transverse diameter. 2) the height of VA (Aneurismal-H), which is measured by taking the maximum vertical distance from the edge of the tumor to the transverse diameter of the base. 3) the end-diastolic volume (Aneurismal-ED) and endsystolic volume (Aneurismal-SD) of the ventricular wall tumor were automatically recorded by the cardiac function analysis software, combined with the method of manual adjustment, the lm sequence of the selected plane was measured, and the endocardial and adventitia boundaries of each systolic and diastolic phase were determined. The report was automatically generated and the end-diastolic volume (Aneurismal-ED) and end-systolic volume (Aneurismal-SD) were recorded.

Strain analysis
Strain measurements were performed as previous described. 9 Brie y, we use the FT-CMR software method of Medis QStrain Software (Medis Medical Imaging Systems, version 2.0.12.2.) (example of the analysis is in the Supplementary File). All two longitudinal-axis views were used to determine peak GLS.
Endocardial contours were manually drawn during end-diastole and end-systole with subsequent automatic tracking during the cardiac cycle. For the assessment of GCS and segmental circumferential strain, the corelab contours for the short-axis images were used. Peak GCS was calculated from 3 shortaxis views (basal, mid, and apical). For peak segmental strain, short-axis images were used to de ne the segments according to the 16-segment model after manual insertion of a reference point (delineated at the anterior insertion of the right ventricle). All studies were loaded into the software and analyzed in a random order by one investigator blinded for randomization outcome under supervision of a CMR cardiologist with > 15 year-experience. The reproducibility of GLS measurements was assessed in 30 CMR scans (15 patients with baseline and follow-up CMR). The intraclass correlation coe cient for interobserver agreement was 0.97 (95% CI 0.89 to 0.99; p<0.005). A representative gure shows strain measurement were shown in Figure 2.

echocardiography
The heart function of 65 patients was detected by two-dimensional echocardiography, and the LVEF of the patients was measured. The wall movement was detected by tissue Doppler method.

Statistical analyze method
Data were processed as previous decribed. 10 Continuous variables were expressed as means ± standard deviation (normal distribution) or median with interquartile range (nonnormal distribution). Categorical data were expressed as numbers (n) with percentages (%). Differences between continuous variables were analyzed using Student's t-test (normal distribution) or Mann-Whiney's U test (nonnormal distribution). Categorical variables were compared by the chi-square test. However, univariate regression analysis was used to determine the available variables predicting NYHA in patients with VA after cardiac infarction. Any variable with unadjusted p < 0.1 was included in the multivariate logistic regression analysis. Spearman correlation coe cient was used to test the relationships between continuous variables. Statistical analyses were performed with SaS software.  The baseline clinical characteristics of patients are shown in Table 1. Patients with coronary artery diameter stenosis greater than 70% and diameter greater than 2mm were included in the screening criteria. For patients with NHYA≥II after MI with aneurysm, they were more female, more diabetic, more diuretics used (62.5% vs. 32%; p=0.0167), and have higher BNP (5.76±1.28 vs. 5.03±1.33; p=0.0305), when compared with patients with NYHA I.

Discussion
Left ventricular aneurysm is a common complication of myocardial infarction at present, which would be accompanied by arrhythmia, thrombosis, rupture, accelerated ventricular dilatation and so on and may further impair patient's cardiac function after cardiac infarction 11 . In this study, We nd person with poorer cardiac function tends to have larger VA basilar transverse diameter, which is consistent with previous reports. However, there still lack indicators for risk strati cation and prognosis prediction for left ventricular aneurysm after cardiac infarction 12 .
Segment strain analysis with CMR-FT in cardiac infarction predicts future cardiovascular events mortality over and above LV ejection fraction and infarct size 13,14 . A previous study have reported that GLS is an independent predictor of medium-term prognosis post STEMI 15 . However, for person with VA after AMI, there still lack evidences. Our study is the rst study for studying cardiac strain in patients with VA after cardiac infarction. In our study of 65 person with left VA, we nd person with poorer cardiac function have smaller GLS, GCS and GRS. GLS, GCS, GRS could predict NYHA class after AMI. This study rst demonstrates the value of myocardial strain in VA and substantiated its prediction value of cardiac function in patients with VA after AMI.
One of the limitations of our study was that our population is low. So, our study may less convincing and larger population need to be enrolled. The second limitation is we do the follow-up 3-6 months after AMI, which may neglect the long-term prognosis.
In conclusion, by assessing baseline myocardial strain after AMI using CMR-FT, we observed GLS, GCS, GRS were correlated to indices of VA and could predict progress of people with VA after AMI. The analysis of VA after MI by CMR can early warn the NYHA of patients, and provide important guidance for clinicians to judge the prognosis of patients, early drugs, surgery or instrument treatment. Figure 1 Representative images show measurement of infarct size (SI)