Abstract
Background: This study evaluated the utility of myocardial strain analysis on computed tomography (CT) using state-of-the-art image analysis software to predict the prognosis of patients who underwent transcatheter aortic valve replacement (TAVR).
Methods and Results: We included 126 patients with severe aortic valve stenosis (AS) who underwent preoperative CT. Major adverse cardiovascular events (MACE) were defined as a composite of cardiac death (including unknown death based on medical records), hospitalization due to heart failure, and fatal arrhythmia. Twenty-four (19%) patients experienced MACE. Global longitudinal strain (GLS), circumferential strain (GCS), radial strain (GRS) of the left ventricular (LV) myocardium (LVM), LV ejection fraction on CT, and the percentage of patients administered aspirin or statins was significantly lower among patients with than without MACE (all P<0.05). The percentage of patients with AF, a history of congestive heart failure, and tolvaptan or oral anticoagulants administration was significantly higher among patients with than without MACE (all P<0.05). In multivariate survival analysis using a Cox proportional hazard model, LV-GLS ≥−9.92% on CT (hazard ratio [HR] 4.45; 95% confidence interval [CI] 1.89–10.48; P=0.0007) and aspirin (HR 0.27; 95% CI 0.10–0.70; P=0.0074) or statin (HR 0.33; 95% CI 0.13–0.84; P=0.02) administration were significant predictors of prognosis after TAVR.
Conclusions: Our findings indicate that LV-GLS on CT is a sensitive predictor of prognosis after TAVR.
Aortic valve stenosis (AS) is not merely a valvular disease of the aortic valve (AV); the left ventricular (LV) hypertrophy and myocardial degeneration caused by pressure overload can also be central features of the pathological abnormalities in AS. Hypertrophy of the LV myocardium (LVM) is a compensatory response aimed at restoring wall stress and maintaining cardiac output under the persistent pressure overload caused by AS. However, this can eventually lead to heart failure.1 Transcatheter aortic valve replacement (TAVR) or surgical aortic valve replacement (SAVR) usually induce reverse remodeling of the LVM by weakening wall stress and reducing LV overload. Nevertheless, if the pathological changes in the LVM have advanced, cardiac dysfunction and poor prognosis may persist even after SAVR or TAVR for severe AS.2
Myocardial fibrosis and myocyte cell death in the LVM in AS are key contributors to heart failure due to compensated hypertrophy.3 Cardiac magnetic resonance imaging (MRI) is recognized as the gold-standard non-invasive method for assessing myocardial damage.4 Although late gadolinium enhancement (LGE) is useful for detecting focal myocardial damage, it is difficult to detect diffuse myocardial fibrosis and evaluate myocardial damage quantitatively. T1
mapping for extracellular volume fraction (ECV) analysis provides a valuable quantitative assessment of myocardial damage, even if it is a diffuse type.5,6 Several studies have shown a reduction in diffuse fibrosis following relief of pressure overload with SAVR and have indicated that ECV analysis on MRI can predict prognosis in AS patients.7 In addition, ECV analysis on computed tomography (CT) has been found to be strongly correlated with ECV analysis on MRI.8
Myocardial strain analysis is now conducted through methods such as echocardiography and MRI, and is considered valuable for the early detection of myocardial dysfunction.9–13 The remarkable advances in cardiac CT technology in recent years have enabled cardiac functional analysis, such as LV ejection fraction (LVEF), thanks to improvements in temporal resolution. Using state-of-the-art image analysis software, myocardial strain analysis for structures like the LV can also be performed using CT images. The acquisition of 1 cardiac cycle is necessary for TAVR candidates because of evaluation of the aortic valve in a systolic phase dataset and coronary artery stenosis in a diastolic phase dataset; therefore, they are suitable for cardiac functional analysis on CT. The gold standard of myocardial strain analysis is echocardiography, but factors such as image quality and the accuracy of the slice plane for 2-dimensional analysis are highly dependent on the skill of the examiner.
The purpose of this study was to evaluate the utility of LV strain analysis on CT to predict prognosis in patients with severe AS who underwent TAVR.
Methods
Study Population
This retrospective observational cohort study included 127 consecutive patients with severe AS who underwent preoperative CT prior to TAVR between June 2020 and March 2022. All patients are the same as reported in our previous paper,14 and all underwent transthoracic echocardiography (TTE) before TAVR. Patients who underwent SAVR (n=5), received conservative management (n=7), had severe mitral regurgitation (n=2), previous SAVR (n=1), or poor imaging quality due to mitral annulus calcification artifacts (n=12) were excluded (Figure 1). The reasons we excluded patients after aortic valve replacement (AVR) in this study were because we did not want to risk including patients who underwent AVR because of severe aortic regurgitation, and because patients for whom TAVR is indicated after AVR have a significantly different history leading to severe AS than do other candidates for TAVR. In addition, the CT dataset for cardiac functional analysis was not stored; thus, 126 patients were finally included in the analysis.
Patients’ medical histories, clinical treatments, and laboratory data were obtained from electronic medical records. TAVR was performed via transfemoral or trans-subclavian access using either balloon-expandable devices (SAPIEN [Edwards Lifesciences]; n=96) or self-expanding devices (Evolut or CoreValve [Medtronic]; n=30).
This study adhered to the principles of the Declaration of Helsinki and was approved by the Ethics Committee of Chiba University (Reference no. 3822).
CT Protocol
Participants were positioned supine on the scanner table for the CT procedure. Scanning was conducted using either a 256-row multidetector CT (Revolution CT Apex; GE HealthCare, Waukesha, WI, USA) or a 320-row multidetector CT (Aquilion One/ViSION Edition; Canon Medical Systems, Otawara, Japan). After an initial scout scan, a non-contrast, electrocardiogram (ECG)-gated cardiac scan was performed using a prospective ECG-gated technique before the contrast-enhanced scan. The 320-row CT used a slice thickness of 0.5 mm with a tube voltage of 80–120 kV, whereas the 256-row CT used a slice thickness of 0.625 mm and a tube voltage of 70–120 kV.
To precisely evaluate the aortic valve complex as well as the coronary artery stenosis, a whole cardiac cycle was acquired using a retrospective ECG-gated scan.15 Conventional contrast-enhanced CT scans were performed with a slice thickness of 0.5 mm and a tube voltage of 80–120 kV for the 320-slice CT or a slice thickness of 0.625 mm and a tube voltage of 120 kV for the 256-slice CT.15 Patients with a heart rate exceeding 65 beats/min were administered 12.5 mg landiolol before the scan, unless contraindicated for β-blockers. In addition, 2 doses of isosorbide dinitrate were administered sublingually to assess significant coronary artery disease before scanning. During the first phase, we injected 40–100 mL undiluted iodinate contrast agent (350–370 mgI/mL) at a rate of 3–5 mL/s, followed by 0–50 mL of a 50%/50% saline/contrast medium mixture at a rate of 3–4 mL/s and 20–30 mL pure saline at a rate of 2–4 mL/s. Following the early-phase cardiac scan, a non-ECG-gated, high-pitch scan of the thorax, abdomen, and pelvis was performed to evaluate the access route.
A late-phase scan was conducted 6 min after administration of the iodine contrast medium using a prospective ECG-gating technique.16 Imaging parameters were consistent with those used in the non-contrast scan: slice thickness of 0.5 mm and tube voltage of 80–120 kV for the 320-row CT, and slice thickness of 0.625 mm and tube voltage of 70–120 kV for the 256-slice CT. The tube current was managed by an automatic exposure control system.
CT Analysis
According to the American Heart Association’s guidelines, the coronary arteries were segmented into 15 parts.17 Segments with a diameter of ≥1.5 mm were included in the analysis. Two board-certified cardiologists (M.T. [4 years experience in cardiac CT] and H.T. [14 years experience]) assessed the presence and location of stenotic lesions in each coronary artery using a commercially available workstation (Ziostation 2; Ziosoft, Tokyo, Japan). Any disagreements were resolved by consensus. Obstructive coronary artery disease was defined as a stenosis >70% in ≥1 major epicardial coronary arteries, following the latest guidelines.18 The Agatston calcium score for coronary arteries and the aortic valve was calculated by M.T. using the same software.
The ECV of the LVM was evaluated in non-contrast and late-phase cardiac images using commercially available software (Ziostation 2).19 The ECV of the LVM was calculated using the following formula:
ECV = (∆HUm / ∆HUb) / (1 − Hct)
where ∆HUm represents the change in myocardial CT attenuation in Hounsfield units (HU), ∆HUb represents the change in CT attenuation of the blood, and Hct is the hematocrit level.16
Myocardial strain analysis of the LVM on CT was performed by 2 cardiologists (H.T. [14 years experience in cardiac CT] and M.M. [2 years experience]) using specific software (Medis Suite CT QStrain 4.2; Medis Medical Imaging, Netherlands). Early-phase cardiac CT data were reconstructed at every 5% of the R-R interval, resulting in a 20-phase cardiac CT dataset. Four-dimensional CT images of 2-, 3-, and 4-chamber views were generated, and global (GLS) and segmental longitudinal strain (LS), global circumferential strain (GCS), and global radial strain (GRS) were calculated based on the AHA 17-segment LVM model using the software (Figure 2).9,10 The LVM strain analysis was performed as a whole-layer analysis.
Echocardiographic Measurement
All patients underwent comprehensive echocardiographic assessment prior to TAVR. TTE was performed using standard ultrasound systems (Vivid E9 [GE Healthcare] or EPIQ7 [Philips, Amsterdam, Netherlands]). Cardiac dimensions, function, and the severity of AS were evaluated according to the guidelines of the American Society of Echocardiography and the European Association of Echocardiography.20 The aortic valve area (AVA) was calculated using the continuity equation and adjusted for body surface area.2 LV end-diastolic and end-systolic volumes, along with LVEF, were measured using the modified Simpson’s method. Left atrial volume was assessed with the biplane area-length method from the apical view, then indexed to body surface area.21 In addition, whole-layer GLS analysis was performed using specific software (EchoPac, GE Vingmed). Based on the LS analysis on TTE, the presence or absence of apical sparing was also evaluated according to the previous criteria. However, GLS was not measurable on TTE in 18 patients, primarily because of arrhythmia, including atrial fibrillation.
Outcomes
The primary outcome was a composite of cardiac death, hospitalization for heart failure, and fatal arrhythmia after TAVR. There were cases in which the cause of death was unclear in the medical record, but the possibility of cardiac death could not be ruled out; thus, patients in whom the cause of death could not be clearly confirmed as non-cardiac were included in the cardiac death category. Follow-up information was obtained from patient medical records at the Chiba University Hospital.
Statistical Analysis
Continuous variables are presented as the mean±SD, whereas categorical variables are shown as counts and percentages. Continuous variables were compared using Student’s t-test and categorical variables were compared using Fisher’s exact test. The diagnostic performance of GLS, GCS, and GRS on CT for predicting prognosis was evaluated through receiver operating characteristic (ROC) curve analysis, with the optimal cut-off value determined by the Youden index.
For time-to-event analyses, the Kaplan-Meier method was used, and group differences were assessed using the log-rank test. Hazard ratios (HRs) are reported as the mean with 95% confidence intervals (CIs). A Cox proportional hazards model was also used to explore the relationship between time to major cardiac events and predictor variables.
Statistical significance was set at P<0.05. All statistical analyses were performed using JMP software, version 18.1.1 (SAS Institute Inc., Cary, NC, USA).
Results
Participant Characteristics
During the follow-up period (651±383 days), 12 patients were hospitalized due to heart failure and 1 patient had sustained ventricular tachycardia. Cardiac deaths were recorded for 7 patients, and non-cardiac deaths were recorded for 4 patients, including pneumonia, rupture of an aortic aneurysm, bacteremia, and gastric cancer. There were 4 patients for whom the cause of death was unclear in the medical record, and because the possibility of cardiac origin could be ruled out, these patients were not excluded from the major adverse cardiovascular events (MACE) analysis. Therefore, we considered 24 patients who had MACE during the follow-up period. Table 1 presents baseline characteristics overall, as well as for patients with and without cardiac events. One cycle early-phase cardiac CT dataset was not recorded in 1 patient; therefore, that patient was excluded from the analysis.
Table 1.
Baseline Characteristics in All Patients and in Those With and Without Cardiac Events Separately
| |
All patients
(n=126) |
Cardiac events |
P value |
| Yes (n=24) |
No (n=102) |
| Age (years) |
84±5 |
83±7 |
85±5 |
0.18 |
| Male sex |
50 (40) |
9 (38) |
41 (40) |
1.00 |
| Body mass index (kg/m2) |
23.0±4.1 |
22.4±4.3 |
23.2±4.1 |
0.40 |
| Creatinine (mg/dL) |
1.15±1.00 |
1.10±0.64 |
1.16±1.07 |
0.80 |
| Hematocrit (%) |
36.7±4.9 |
36.7±6.0 |
36.7±4.6 |
0.97 |
| Low-voltage ECG |
44 (35) |
11 (46) |
33 (32) |
0.24 |
| Follow-up period (days) |
651±383 |
259±271 |
743±346 |
<0.0001 |
| Medical history |
| Hypertension |
90 (71) |
18 (75) |
72 (71) |
0.80 |
| Diabetes |
27 (21) |
5 (22) |
22 (22) |
1.00 |
| Dyslipidemia |
67 (53) |
12 (50) |
55 (54) |
0.82 |
| Atrial fibrillation |
30 (24) |
14 (58) |
16 (16) |
<0.0001 |
| Dialysis |
3 (2) |
0 (0) |
3 (3) |
1.00 |
| Chronic kidney disease |
30 (24) |
7 (30) |
23 (23) |
0.59 |
| Previous myocardial infarction |
8 (6) |
0 (0) |
8 (8) |
0.35 |
| Previous PCI |
10 (8) |
0 (0) |
10 (10) |
0.21 |
| Previous CABG |
1 (1) |
0 (0) |
1 (1) |
1.00 |
| Past history of congestive heart failure |
48 (38) |
14 (58) |
34 (33) |
0.035 |
| Echocardiographic measurements |
| LVEF (%) |
59.1±10.9 |
55.7±13.6 |
59.9±10.1 |
0.09 |
| Mean aortic valve gradient (mmHg) |
49.1±16.4 |
50.3±20.2 |
48.8±15.5 |
0.69 |
| Indexed aortic valve area (cm2/m2) |
0.44±0.11 |
0.41±0.14 |
0.45±0.11 |
0.077 |
| LV end-diastolic volume (mL) |
93.0±37.4 |
90.8±37.6 |
93.5±37.5 |
0.75 |
| LV end-systolic volume (mL) |
40.0±26.8 |
42.4±27.0 |
39.5±26.9 |
0.63 |
| LV end-diastolic diameter (mm) |
43.1±7.3 |
44.4±6.2 |
42.7±7.5 |
0.32 |
| LV end-systolic diameter (mm) |
29.7±6.9 |
31.2±6.8 |
29.3±6.9 |
0.24 |
| LA volume index (mL/m2) |
53.4±32.9 |
58.8±21.2 |
52.2±35.1 |
0.38 |
| Mitral E (cm/s) |
84.3±30.8 |
87.3±35.1 |
83.5±29.8 |
0.59 |
| E/A ratio |
0.76±0.42 |
0.69±0.42 |
0.77±0.42 |
0.54 |
| E/e′ ratio |
18.9±9.8 |
18.3±6.7 |
19.0±10.4 |
0.77 |
| Pulmonary artery systolic pressure (mmHg) |
31.1±14.2 |
32.1±16.4 |
30.8±13.8 |
0.71 |
| GLS on TTE (%) |
−14.9±4.9 |
−13.2±4.8 |
−15.2±4.9 |
0.11 |
| Apical sparing |
61 (48) |
11 (61) |
50 (49) |
0.80 |
| Low-flow, low-gradient subtype |
6 (5) |
1 (4) |
5 (5) |
1.00 |
| Bicuspid aortic valve |
2 (2) |
1 (4) |
1 (1) |
0.35 |
| CT measures |
| LV ECV (%) |
31.3±4.3 |
32.4±4.8 |
31.1±4.1 |
0.16 |
| CAD |
49 (39) |
9 (38) |
40 (39) |
1.00 |
| CACS |
1,467±1,594 |
1,415±1,707 |
1,480±1,575 |
0.86 |
| AVCS |
3,085±2,672 |
2,978±1,349 |
3,110±2,901 |
0.83 |
| CT GCS (%) |
−17.7±6.0 |
−14.9±7.5 |
−18.4±5.5 |
0.010 |
| CT GRS (%) |
44.2±15.7 |
36.8±18.9 |
45.9±14.4 |
0.0096 |
| CT GLS (%) |
−12.7±4.4 |
−10.2±5.0 |
−13.3±4.1 |
0.0019 |
| CT LVEF (%) |
52.1±14.1 |
44.1±18.1 |
54.0±12.4 |
0.0016 |
| Medication at discharge |
| Aspirin |
92 (73) |
9 (38) |
83 (81) |
<0.0001 |
| P2Y12 inhibitor |
15 (12) |
4 (17) |
11 (11) |
0.48 |
| OAC |
38 (30) |
15 (63) |
23 (23) |
0.0003 |
| β-blocker |
67 (53) |
16 (67) |
51 (50) |
0.18 |
| ACEi/ARB/ARNI |
55 (44) |
12 (50) |
43 (42) |
0.50 |
| CCB |
50 (40) |
8 (33) |
42 (41) |
0.64 |
| Loop diuretics |
69 (55) |
16 (67) |
53 (52) |
0.26 |
| Tolvaptan |
13 (10) |
8 (33) |
5 (5) |
0.0004 |
| Thiazide |
4 (3) |
0 (0) |
4 (4) |
1.00 |
| MRA |
40 (32) |
10 (42) |
30 (29) |
0.33 |
| SGLT2 inhibitor |
9 (7) |
1 (4) |
8 (8) |
1.00 |
| PPI |
107 (85) |
22 (92) |
85 (83) |
0.53 |
| Histamine H2 receptor blocker |
3 (2) |
0 (0) |
3 (3) |
1.00 |
| Steroid p.o. |
11 (9) |
2 (8) |
9 (9) |
1.00 |
| Statin |
62 (49) |
6 (25) |
56 (55) |
0.012 |
| Ezetimibe |
12 (10) |
4 (17) |
8 (8) |
0.24 |
Unless indicated otherwise, data are given as the mean±SD or n (%). A, late atrial diastolic transmitral flow velocity; ACEi, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor-neprilysin inhibitor; AVCS, aortic valve calcium score; CABG, coronary artery bypass grafting; CACS, coronary artery calcium score; CAD, coronary artery disease; CCB, calcium channel blocker; CT, computed tomography; E, early diastolic transmitral flow velocity; e′, early diastolic mitral annular velocity; ECG, electrocardiogram; ECV, extracellular volume fraction; GCS, global circumferential strain; GLS, global longitudinal strain; GRS, global radial strain; LA, left atrium; LV, left ventricle; LVEF, left ventricular ejection fraction; MRA, mineralocorticoid receptor antagonist; OAC, oral anticoagulant; PCI percutaneous coronary intervention; PPI, proton pump inhibitor; SGLT2, sodium-glucose cotransporter 2; TTE, transthoracic echocardiography.
Predictors of Cardiac Events
GLS of the LVM (LV-GLS) on CT was significantly decreased in patients with than without composite outcomes (−10.2±5.0% vs. −13.3±4.1%; P=0.0019). GCS and GRS of the LVM (LV-GCS and LV-GRS, respectively) on CT were also significantly decreased in patients with than without MACE (LV-GCS: −14.9±7.5 vs. −18.4±5.5%, respectively [P=0.010]; LV-GRS, 36.8±18.9 vs. 46.9±14.4%, respectively [0.0096]). A significantly higher percentage of patients with than without MACE had atrial fibrillation (14 [58%] vs. 16 [16%]; P<0.0001) and a past history of congestive heart failure (14 [58%] vs. 34 [33%]; P=0.035). The LVEF on CT was significantly lower in patients with than without MACE (44.1±18.1% vs. 54.0±12.4%; P=0.0016). A significantly lower percentage of patients with than without MACE were administered aspirin (9 [38%] vs. 83 [81%]; P<0.0001) and statins (6 [25%] vs. 56 [55%]; P=0.012). A significantly higher percentage of patients with than without MACE were administered an oral anticoagulant (OAC; 15 [63%] vs. 23 [23%]; P=0.0003) and tolvaptan (8 [33%] vs. 5 [5%]; P=0.0004; Table 1).
ROC analysis identified –9.92% as the optimal LV-GLS cut-off value on CT (CT-GLS) for predicting cardiac events (Figure 3A). This cut-off value had a sensitivity of 54% and a specificity of 87%, with an area under the curve of 0.69 (P=0.0019). Based on this threshold, patients were divided into 2 groups, and Kaplan-Meier analysis with a log-rank test was used to compare the prognosis between those with higher and lower LV-GLS (Figure 3A). Similarly, ROC analysis determined −18.18% and 36.43% as the best cut-off values for LV-GCS and LV-GRS on CT for predicting cardiac events (Figure 3B,C). Patients with decreased LV-GLS experienced significantly more cardiac events than did patients with higher LV-GLS (P<0.0001; Figure 4).
According to the optimal threshold of CT-GLS determined from the ROC analysis results (−9.92%), patients were divided into 2 groups, one with higher CT-GLS and the other with lower CT-GLS, and the background characteristics compared between the 2 groups. A significantly higher proportion of patients with decreased LV-GLS, compared with those with higher LV-GLS, had had previous PCI (6 [23%] vs. 4 [4%]; P=0.0053), a history of congestive heart failure (18 [69%] vs. 30 [30%]; P=0.0005), CT-GCS ≥−18.18 (26 [100%] vs. 40 [40%]; P<0.0001), and were administered OAC (13 [50%] vs. 25 [25%]; P=0.018), loop diuretics (20 [77%] vs. 49 [49%]; P=0.014), and mineralocorticoid antagonists (MRAs; 15 [58%] vs. 25 [25%]; P=0.0037) at discharge. Patients with decreased LV-GLS, compared with those with higher LV-GLS, had significantly lower LVEF on TTE (50.7±14.4% vs. 61.3±8.6%; P<0.0001), body mass index (21.5±3.1 vs. 23.4±4.2 kg/m2; P=0.036), indexed aortic valve area (0.39±0.13 vs. 0.46±0.1 cm2/m2; P=0.0057), and GLS on TTE (−10.3±3.3 vs. –15.8±4.6; P<0.0001). Patients with decreased LV-GLS, compared with those with higher LV-GLS, had significantly higher LV-ECV (33.6±4.9% vs. 30.8±3.9%; P=0.023), LV endo-systolic diameter (54.9±39.4 vs. 36.2±21.1 mm; P=0.0013) and volume (32.5±9.3 vs. 28.9±6.0 mL; P=0.017), and left atrium volume index (66.7±62.4 vs. 50.0±18.0; P=0.02; Table 2). The number of patients with CT-GRS ≥36.43% was significantly lower in group with lower than higher LV-GLS (5 [19%] vs. 84 [84%]; P<0.0001).
Table 2.
Baseline Characteristics of Patients Stratified According to GLS −9.92% on CT
| |
GLS |
P value |
| <−9.92% (n=100) |
≥−9.92% (n=26) |
| Age (years) |
84±5 |
85±7 |
0.48 |
| Male sex |
41 (41) |
9 (35) |
0.81 |
| Body mass index (kg/m2) |
23.4±4.2 |
21.5±3.1 |
0.036 |
| Creatinine (mg/dL) |
1.10±0.81 |
1.31±1.55 |
0.36 |
| Low-voltage on ECG |
33 (33) |
11 (42) |
0.49 |
| Follow-up period (days) |
711±374 |
419±329 |
0.0004 |
| Medical history |
| Hypertension |
72 (72) |
18 (69) |
0.74 |
| Diabetes |
17 (17) |
10 (38) |
0.0031 |
| Dyslipidemia |
52 (52) |
15 (58) |
0.67 |
| Atrial fibrillation |
21 (21) |
9 (35) |
0.20 |
| Dialysis |
2 (2) |
1 (4) |
0.50 |
| Chronic kidney disease |
21 (21) |
9 (35) |
0.20 |
| Previous myocardial infarction |
4 (4) |
4 (15) |
0.056 |
| Previous PCI |
4 (4) |
6 (23) |
0.0053 |
| Previous CABG |
1 (1) |
0 (0) |
1.00 |
| Past history of congestive heart failure |
30 (30) |
18 (69) |
0.0005 |
| GLS on TTE (%) |
−15.8±4.6 |
−10.3±3.3 |
<0.0001 |
| Apical sparing on TTE |
48 (48) |
13 (50) |
0.46 |
| LVEF (%) |
61.3±8.6 |
50.7±14.4 |
<0.0001 |
| Mean aortic valve gradient (mmHg) |
49.3±16.1 |
48.3±18.1 |
0.80 |
| Indexed aortic valve area (cm2/m2) |
0.46±0.1 |
0.39±0.13 |
0.0057 |
| LV end-diastolic volume (mL) |
90.5±33.4 |
102.6±49.5 |
0.14 |
| LV end-systolic volume (mL) |
36.2±21.1 |
54.9±39.4 |
0.0013 |
| LV end-diastolic diameter (mm) |
42.7±7.2 |
44.3±7.9 |
0.32 |
| LV end-systolic diameter (mm) |
28.9±6.0 |
32.5±9.3 |
0.017 |
| LA volume index (mL/m2) |
50.0±18.0 |
66.7±62.4 |
0.02 |
| Mitral E (cm/s) |
84.5±29.5 |
83.5±35.7 |
0.88 |
| E/A ratio |
0.77±0.39 |
0.69±0.56 |
0.48 |
| E/e′ ratio |
18.5±9.5 |
20.4±10.9 |
0.38 |
| Pulmonary artery systolic pressure (mmHg) |
31.1±14.7 |
30.8±12.7 |
0.93 |
| Low-flow, low-gradient subtype |
4 (4) |
2 (8) |
0.60 |
| Bicuspid aortic valve |
2 (2) |
0 (0) |
1.00 |
| CT measures |
| LV-ECV (%) |
30.8±3.9 |
33.6±4.9 |
0.023 |
| CT GCS ≥−18.18% (lower GCS) |
40 (40) |
26 (100) |
<0.0001 |
| CT GRS ≥36.43% (higher GRS) |
84 (84) |
5 (19) |
<0.0001 |
| CAD |
36 (36) |
13 (50) |
0.26 |
| CACS |
1,377±1,439 |
1,817±2,085 |
0.21 |
| AVCS |
2,849±1,332 |
3,994±5,250 |
0.051 |
| Medication at discharge |
| Aspirin |
77 (77) |
15 (58) |
0.080 |
| P2Y12 inhibitor |
11 (11) |
4 (15) |
0.51 |
| OAC |
25 (25) |
13 (50) |
0.018 |
| β-blocker |
49 (49) |
18 (69) |
0.079 |
| ACEi/ARB/ARNI |
41 (41) |
14 (54) |
0.27 |
| CCB |
40 (40) |
10 (38) |
1.00 |
| Loop diuretics |
49 (49) |
20 (77) |
0.014 |
| Tolvaptan |
8 (8) |
5 (19) |
0.14 |
| Thiazide |
3 (3) |
1 (4) |
1.000 |
| MRA |
25 (25) |
15 (58) |
0.0037 |
| SGLT2 inhibitor |
6 (6) |
3 (12) |
0.39 |
| PPI |
85 (85) |
22 (85) |
1.00 |
| Histamine H2 receptor blocker |
2 (0) |
1 (4) |
0.50 |
| Steroid p.o. |
10 (10) |
1 (4) |
0.46 |
| Statin |
49 (49) |
13 (50) |
1.00 |
| Ezetimibe |
11 (11) |
1 (4) |
0.46 |
Unless indicated otherwise, data are given as the mean±SD or n (%). Abbreviations as in Table 1.
Univariable and multivariable analyses using a Cox proportional hazards model were used to identify significant predictors of cardiac events (Table 3). In univariate analysis, atrial fibrillation (HR 5.79; 95% CI 2.56–13.10; P<0.0001), a previous history of congestive heart failure (HR 2.67; 95% CI 1.18–6.03; P=0.018), LVEF on CT (HR 0.96; 95% CI 0.93–0.98, P=0.0007), LV-GCS on CT ≥−18.18% (HR 3.18; 95% CI 1.26–8.04; P=0.014), LV-GRS on CT ≥36.43% (HR 0.33; 95% CI 0.15–0.74; P=0.0068), and LV-GLS on CT ≥−9.92% (HR 6.35; 95% CI 2.81–14.34; P <0.0001) were significant predictors of cardiac events after TAVR. Aspirin (HR 0.18; 95% CI 0.079–0.42; P<0.0001), OAC (HR 4.6; 95% CI 2.02–10.6; P=0.0003), tolvaptan (HR 8.8; 95% CI 3.59–21.6; P<0.0001), and statin (HR 0.34; 95% CI 0.13–0.85; P=0.021) use were also significant predictors of cardiac events after TAVR.
Table 3.
Univariate and Multivariate Analysis Using the Cox Proportional Hazard Model of Predictive Factors of Cardiac Events
| |
Univariate |
Multivariate |
| HR (95% CI) |
P value |
HR (95% CI) |
P value |
| Age |
0.94 (0.88–1.02) |
0.13 |
|
|
| Male sex |
0.90 (0.39–2.05) |
0.79 |
|
|
| Body mass index |
0.94 (0.85–1.04) |
0.25 |
|
|
| Creatinine |
1.09 (0.57–1.47) |
0.71 |
|
|
| Hematocrit |
0.99 (0.90–1.08) |
0.76 |
|
|
| Low-voltage on ECG |
1.68 (0.75–3.75) |
0.21 |
|
|
| Hypertension |
1.16 (0.46–2.93) |
0.75 |
|
|
| Diabetes |
1.20 (0.45–3.24) |
0.72 |
|
|
| Dyslipidemia |
0.90 (0.40–2.01) |
0.80 |
|
|
| Atrial fibrillation |
5.79 (2.56–13.10) |
<0.0001 |
2.21 (0.85–5.72) |
0.10 |
| Chronic kidney disease |
1.56 (0.65–3.78) |
0.32 |
|
|
| Previous congestive heart failure |
2.67 (1.18–6.03) |
0.018 |
|
|
| GLS on TTE |
1.09 (0.99–1.19) |
0.063 |
|
|
| Apical sparing on TTE |
1.28 (0.50–3.31) |
0.61 |
|
|
| LVEF on TTE |
0.97 (0.94–1.00) |
0.06 |
|
|
| Mean aortic valve gradient |
1.00 (0.98–1.03) |
0.79 |
|
|
| Indexed aortic valve area |
0.07 (0.004–1.95) |
0.10 |
|
|
| LV end-diastolic volume |
1.00 (0.99–1.01) |
0.75 |
|
|
| LV end-systolic volume |
1.00 (0.99–1.02) |
0.51 |
|
|
| LV end-diastolic diameter |
1.03 (0.98–1.07) |
0.18 |
|
|
| LV end-systolic diameter |
1.03 (0.98–1.07) |
0.18 |
|
|
| LA volume index |
1.01 (0.98–1.01) |
0.11 |
|
|
| Mitral E |
1.00 (0.99–1.02) |
0.66 |
|
|
| E/A ratio |
0.62 (0.10–2.14) |
0.53 |
|
|
| E/e′ ratio |
1.00 (0.95–1.03) |
0.84 |
|
|
| Pulmonary artery systolic pressure |
1.00 (0.97–1.03) |
0.92 |
|
|
| Low-flow, low-gradient subtype |
0.94 (0.13–6.99) |
0.95 |
|
|
| Bicuspid aortic valve |
2.48 (0.33–18.41) |
0.14 |
|
|
| LV-ECV on CT |
1.09 (0.99–1.18) |
0.07 |
|
|
| CAD |
0.98 (0.43–2.25) |
0.97 |
|
|
| CT GCS ≥−18.18% |
3.18 (1.26–8.04) |
0.014 |
|
|
| CT GLS ≥−9.98% |
6.35 (2.81–14.34) |
<0.0001 |
4.45 (1.89–10.48) |
0.0007 |
| CT GRS ≥36.43% |
0.33 (0.15–0.74) |
0.0068 |
|
|
| CT LVEF |
0.96 (0.93–0.98) |
0.0007 |
|
|
| AVCS |
1.00 (1.00–1.00) |
0.78 |
|
|
| Medication at discharge |
| Aspirin |
0.18 (0.079–0.42) |
<0.0001 |
0.27 (0.10–0.70) |
0.0074 |
| P2Y12 inhibitor |
1.49 (0.51–4.37) |
0.46 |
|
|
| OAC |
4.6 (2.02–10.6) |
0.0003 |
|
|
| β-blocker |
1.98 (0.84–4.64) |
0.12 |
|
|
| ACEi/ARB/ARNI |
1.33 (0.60–2.96) |
0.49 |
|
|
| CCB |
0.76 (0.32–1.76) |
0.52 |
|
|
| Loop diuretic |
1.79 (0.77–4.19) |
0.18 |
|
|
| Tolvaptan |
8.8 (3.59–21.6) |
<0.0001 |
|
|
| MRA |
1.67 (0.74–3.78) |
0.21 |
|
|
| SGLT2 inhibitor |
0.59 (0.08–4.36) |
0.60 |
|
|
| PPI |
10.20 (0.52–9.36) |
0.29 |
|
|
| Steroid p.o. |
0.99 (0.23–4.23) |
1.00 |
|
|
| Statin |
0.34 (0.13–0.85) |
0.021 |
0.33 (0.13–0.84) |
0.020 |
| Ezetimibe |
2.19 (0.75–6.42) |
0.15 |
|
|
CI, confidence interval; HR, hazard ratio. Other abbreviations as in Table 1.
Past statistical reports have suggested that the number of factors in a multivariate analysis should be approximately one-fifth to one-tenth of the number of interventions (n=24 in the present study).22,23 Therefore, it was necessary to fully discard the covariates in the multivariate analysis. First, the main purpose of this study was to predict prognosis by CT-GLS, so that other factors that were considered highly correlated with CT-GLS (LVEF, LV-GCS, and LV-GRS on CT; a previous history of congestive heart failure) based on the results presented in Table 2, were excluded from the multivariate analysis. Some cardiovascular drugs were administered more frequently in the group with the poor prognosis, which was considered a limitation of this retrospective study. Therefore, loop diuretics or tolvaptan at discharge were also excluded from the multivariate analysis due to the limited number of covariates as described above. Finally, we compared CT-GLS, atrial fibrillation, and the administration of aspirin or statin in the multivariate analysis.24,25 In multivariate survival analysis using the Cox proportional hazard model, only LV-GLS ≥−9.92% on CT (HR 4.45; 95% CI 1.89–10.48; P=0.0007) and the administration of aspirin (HR 0.27; 95% CI 0.10–0.70; P=0.0074) or statin (HR 0.33; 95% CI 0.13–0.84; P=0.020) were independent predictors of patient prognosis after TAVR.
Discussion
This study examined the utility of GLS analysis on preoperative CT for predicting prognosis in TAVR candidates with severe AS. We found that higher LV-GLS on CT was associated with cardiac events after TAVR and was an independent predictor of patient prognosis. The present study also suggests that aspirin and statins may have played a role in improving the prognosis of patients who underwent TAVR, but we believe that larger prospective trials are needed to prove this.
A previous study reported the utility of GLS analysis on CT to predict patient prognosis after TAVR,11 but the utility of GLS to predict MACE was not confirmed. LV-GLS on CT was also reported to be useful for detecting cardiac amyloidosis in TAVR candidates.12
TAVR has emerged as an effective treatment for elderly patients with both severe AS and high risk for SAVR. Nevertheless, a substantial proportion of patients who undergo TAVR are readmitted for heart failure.26,27 Our findings suggest that GLS analysis on CT performed just before TAVR may be useful for assessing the risk of cardiac events and determining the intensiveness of follow-up, as well as drug doses, to reduce the risk of heart failure after TAVR.
Assessment of Myocardial Damage in AS
LV hypertrophy regresses by approximately 20–30% within 1 year after SAVR, a result of reverse myocardial remodeling due to the reduction in afterload.7 Approximately two-thirds of patients with a reduced LVEF experience reverse remodeling after either SAVR or TAVR, with early recovery of LVEF being linked to a better prognosis.28,29 However, complete myocardial normalization is not always achievable, even after surgical intervention for severe AS.
Because myocardial fibrosis contributes to LV dysfunction and the subsequent development of heart failure, both histological and imaging studies have explored the connection between cardiac decompensation and structural myocardial remodeling. In the present study, patients with elevated GLS also exhibited reduced LVEF on TTE and higher LV ECV on CT, which may be indicative of advanced myocardial fibrosis (Table 2).
Myocardial fibrosis generally presents in 2 patterns: diffuse fibrosis or replacement fibrosis. Diffuse fibrosis, characterized by the expansion of collagen fibers and hypertrophy of cardiomyocytes, has been shown to regress following TAVR or SAVR due to its reversible nature.30 Conversely, replacement fibrosis, which involves collagen deposition following myocyte apoptosis or necrosis, represents a more advanced stage of diffuse fibrosis and is considered irreversible. Severe myocardial replacement fibrosis observed during SAVR is linked to an increased risk of cardiac mortality,31 and both types of fibrosis are predictors of ventricular decompensation.3
Clinical Implications of GLS Analysis on CT
Myocardial strain analysis was initially measured using echocardiography and MRI, and has been considered useful for quantifying local wall motion abnormalities and detecting early myocardial damage.32 Recently, advances in the temporal resolution of CT scanners, a reduction in radiation exposure, and the development of new image analysis software have enabled myocardial strain analysis using 4-dimensional images from CT.9–13
The limitation of using CT to analyze cardiac function such as strain is that it requires data collection for 1 heartbeat, which increases the radiation dose to the patient compared with coronary artery diagnosis, which is the original purpose of CT. Preoperative evaluation for TAVR requires assessment of the heart and systemic access routes using CT. In addition, CT is recommended for evaluating the aortic valve complex during systole and the coronary arteries during diastole, with early contrast imaging generally captured in 1 heartbeat. Although radiation exposure from CT-based cardiac function analysis is typically a concern, this is not the case in pre-TAVR patients, where the necessary single heartbeat data for functional analysis is obtained through standard imaging, making this approach reasonable.
Utility of GLS Analysis on CT
This CT-based myocardial strain analysis data has been found to correlate with that from echocardiography, with a particularly good correlation noted for LV-GLS.9 Of the various strain metrics, LV-GLS is considered a sensitive parameter for indicating myocardial damage.3 CT strain has already been shown to be useful in detecting apical sparing in patients with amyloidosis and LV apical ballooning in takotsubo cardiomyopathy, in addition to candidates for TAVR.10,13
The results of ROC analysis inferred that GLS is a useful strain for prognostication after TAVR compared with other CT strains. AS requires the LVM to generate high pressure to pump blood to the aorta. This pressure load causes hypertrophy of myocytes in the LVM to compensate for the contractile force. This results in increased LV wall thickness and decreased LV compliance. The hypertrophied myocardium has an increased demand for oxygen and nutrient supply, but the coronary arterial supply capacity is constant, resulting in relative ischemia. In addition, increased LV pressure decreases coronary perfusion pressure, mainly in the subendocardial layer, resulting in dysfunction and fibrosis from ischemia in the endocardial layer. In the myocardium, fibers in the subendocardial layer mainly contribute to long-axis contraction. Because the long-axis strain strongly reflects the function of the endocardial layer, it is thought to be able to sensitively detect changes in the early stages of myocardial dysfunction, such as ischemia and fibrosis, and thus the results obtained differ from those of strains in other directions.
Prognostic Implications of Myocardial Damage in AS
Current guidelines typically recommend SAVR or TAVR primarily for patients exhibiting symptoms or LV systolic dysfunction due to severe AS.2,33,34 However, in routine clinical practice, it can be challenging to discern symptoms attributable to severe AS in elderly patients with multiple comorbidities. In addition, relying solely on LV systolic dysfunction may mask underlying myocardial histological changes. Research indicates that myocardial fibrosis, as detected through MRI techniques such as ECV and LGE, may develop before symptoms manifest in AS patients.35,36
These findings suggest that some patients with preserved LV systolic function may receive intervention too late and that pre-existing myocardial damage could contribute to heart failure following TAVR. Although not a significant factor in the present study, the fact that more patients in the cardiac event group had a history of heart failure underscores the need for early intervention in severe AS. Our research also highlights the value of GLS analysis using CT, providing crucial insights for the perioperative management of AS patients undergoing TAVR. Further research is necessary to establish the role of imaging-based early myocardial damage detection in patients with severe AS who are asymptomatic.
Study Limitations
Several limitations of our study should be acknowledged. First, the retrospective nature of the study, which was conducted at a single center with a small sample size, limits its generalizability. The number of patients in this study was very small, meaning some analyses, such as propensity score matching, could not be performed. In addition, it is possible that the present study did not adequately examine the details of the TAVR procedure in individual patients. There may have been variations in pharmacotherapy from one physician to another, limiting the verification of its impact on prognosis. In addition, some cardiovascular drugs were administered more frequently in the poor prognosis group. These are considered limitations of the present retrospective study. A larger, prospective study is needed to validate the utility of preoperative GLS analysis on CT for predicting cardiac events after TAVR.
Conclusions
LV-GLS on CT was found to be associated with cardiac events in patients undergoing TAVR. Our analysis suggests that preoperative LV-GLS on CT can help predict prognosis after TAVR in patients with severe AS.
Sources of Funding
This work was supported, in part, by JSPS KAKENHI grants (No. JP24K21117 and JP23K11891).
Disclosures
Y.K. is a member of Circulation Journal’s Editorial Team. The remaining authors have no conflicts of interest related to the contents of this article to declare.
IRB Information
The study was approved by the Ethics Committee of Chiba University (Reference no. 3822).
Data Availability
The deidentified participant data will not be shared.
References
- 1.
Carabello BA, Paulus WJ. Aortic stenosis. Lancet 2009; 373: 956–966.
- 2.
Baumgartner H, Falk V, Bax JJ, De Bonis M, Hamm C, Holm PJ, et al. 2017 ESC/EACTS guidelines for the management of valvular heart disease. Eur Heart J 2017; 38: 2739–2791.
- 3.
Chin CWL, Everett RJ, Kwiecinski J, Vesey AT, Yeung E, Esson G, et al. Myocardial fibrosis and cardiac decompensation in aortic stenosis. JACC Cardiovasc Imaging 2017; 10: 1320–1333.
- 4.
Cummings KW, Bhalla S, Javidan-Nejad C, Bierhals AJ, Gutierrez FR, Woodard PK. A pattern-based approach to assessment of delayed enhancement in nonischemic cardiomyopathy at MR imaging. RadioGraphics 2009; 29: 89–103.
- 5.
Messroghli DR, Moon JC, Ferreira VM, Grosse-Wortmann L, He T, Kellman P, et al. Clinical recommendations for cardiovascular magnetic resonance mapping of T1, T2, T2 and extracellular volume: A consensus statement by the Society for Cardiovascular Magnetic Resonance (SCMR) endorsed by the European Association for Cardiovascular Imaging. J Cardiovasc Magn Reson 2017; 19: 1–24.
- 6.
Nakamori S, Dohi K. Myocardial tissue imaging with cardiovascular magnetic resonance. J Cardiol 2022; 80: 377–385.
- 7.
Treibel TA, Kozor R, Schofield R, Benedetti G, Fontatna M, Bhuva AN, et al. Reverse myocardial remodeling following valve replacement in patients with aortic stenosis. JACC 2018; 71: 860–871.
- 8.
Nacif MS, Kawel N, Lee JJ, Chen X, Yao J, Zavodni A, et al. Interstitial myocardial fibrosis assessed as extracellular volume fraction with low-radiation-dose cardiac CT. Radiology 2012; 264: 876–883.
- 9.
Bernhard B, Grogga H, Zurkirchena J, Demirela C, Hagemeyera D, Okunoa T, et al. Reproducibility of 4D cardiac computed tomography feature tracking myocardial strain and comparison against speckle-tracking echocardiography in patients with severe aortic stenosis. J Cardiovasc Comput Tomogr 2022; 16: 309–318.
- 10.
Kinoshita M, Takaoka H, Ota J, Ikeda JI, Noguchi Y, Nishikawa Y, et al. Four-dimensional computed tomography is useful for detection of apical sparing of cardiac amyloidosis. Int J Cardiovasc Imaging 2024, doi:10.1007/s10554-024-03127-6.
- 11.
Bernhard B, Leib Z, Dobner S, Demirel C, Caobelli F, Rominger A, et al. Myocardial analysis from routine 4D cardiac-CT to predict reverse remodeling and clinical outcomes after transcatheter aortic valve implantation. Eur J Radiol 2024; 175: 111425.
- 12.
Bernhard B, Leib Z, Dobner S, Demirel C, Caobelli F, Rominger A, et al. Routine 4D cardiac CT to identify concomitant transthyretin amyloid cardiomyopathy in older adults with severe aortic stenosis. Radiology 2023; 309: e230425.
- 13.
Miyakoda K, Kato K, Takaoka H, Noguchi Y, Matsumoto M, Nishikawa Y, et al. Utility of myocardial strain analysis on computed tomography in a patient with takotsubo syndrome. Circ J 2025; 89: 517.
- 14.
Takahashi M, Takaoka H, Yashima S, Suzuki-Eguchi N, Ota J, Kitahara H, et al. Extracellular volume fraction by computed tomography predicts prognosis after transcatheter aortic valve replacement. Circ J 2024; 88: 492–500.
- 15.
Takaoka H, Funabashi N, Uehara M, Fujimoto Y, Kobayashi Y. Diagnostic accuracy of coronary 320 slice CT angiography using retrospective electrocardiogram gated acquisition compared with virtual prospective electrocardiogram gated acquisition with and without padding. Int J Cardiol 2013; 168: 2811–2815.
- 16.
Hamdy A, Kitagawa K, Goto Y, Yamada A, Nakamura S, Takafuji M, et al. Comparison of the different imaging time points in delayed phase cardiac CT for myocardial scar assessment and extracellular volume fraction estimation in patients with old myocardial infarction. Int J Cardiovasc Imaging 2019; 35: 917–926.
- 17.
Austen WG, Edwards JE, Frye RL, Gensini GG, Gott VL, Griffith LS, et al. A reporting system on patients evaluated for coronary artery disease: Report of the Ad Hoc Committee for Grading of Coronary Artery Disease, Council on Cardiovascular Surgery, American Heart Association. Circulation 1975; 51(Suppl): 5–40.
- 18.
Narula J, Chandrashekhar Y, Ahmadi A, Abbara S, Berman DS, Blankstein R, et al. SCCT 2021 expert consensus document on coronary computed tomographic angiography: A report of the Society of Cardiovascular Computed Tomography. J Cardiovasc Comput Tomogr 2021; 15: 192–217.
- 19.
Nishikawa Y, Takaoka H, Kato K, Ota J, Noguchi Y, Aoki S, et al. Extracellular volume analysis using computed tomography is useful for follow-up of takotsubo syndrome. Circ J 2024; 88: 1198.
- 20.
Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: An update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2015; 28: 1–39.
- 21.
Nistri S, Galderisi M, Ballo P, Olivotto I, D’Andrea A, Pagliani L, et al. Determinants of echocardiographic left atrial volume: Implications for normalcy. Eur J Echocardiogr 2011; 12: 826–833.
- 22.
Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol 1996; 49: 1373–1379.
- 23.
Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression. Am J Epidemiol 2007; 165: 710–718.
- 24.
Tarantini G, Mojoli M, Windecker S, Wendler O, Lefèvre T, Saia F, et al. Prevalence and impact of atrial fibrillation in patients with severe aortic stenosis undergoing transcatheter aortic valve replacement: An analysis from the SOURCE XT prospective multicenter registry. JACC Cardiovasc Interv 2016; 9: 937–946.
- 25.
Ferro CJ, Chue CD, Belder MAD, Moat N, Wendler O, Trivedi U, et al. Impact of renal function on survival after transcatheter aortic valve implantation (TAVI): An analysis of the UK TAVI registry. Heart 2015; 101: 546–552.
- 26.
Durand E, Doutriaux M, Bettinger N, Tron C, Fauvel C, Bauer F, et al. Incidence, prognostic impact, and predictive factors of readmission for heart failure after transcatheter aortic valve replacement. JACC Cardiovasc Interv 2017; 10: 2426–2436.
- 27.
Auffret V, Bakhti A, Leurent G, Bedossa M, Tomasi J, Belhaj Soulami R, et al. Determinants and impact of heart failure readmission following transcatheter aortic valve replacement. Circ Cardiovasc Interv 2020; 13: e008959.
- 28.
Morris JJ, Schaff HV, Mullany CJ, Rastogi A, McGregor CG, Daly RC, et al. Determinants of survival and recovery of left ventricular function after aortic valve replacement. Ann Thorac Surg 1993; 56: 22–29.
- 29.
Dauerman HL, Reardon MJ, Popma JJ, Little SH, Cavalcante JL, Adams DH, et al. Early recovery of left ventricular systolic function after CoreValve transcatheter aortic valve replacement. Circ Cardiovasc Interv 2016; 9: e003425.
- 30.
Bing R, Cavalcante JL, Everett RJ, Clavel MA, Newby DE, Dweck MR. Imaging and impact of myocardial fibrosis in aortic stenosis. JACC Cardiovasc Imaging 2019; 12: 283–296.
- 31.
Herrmann S, Fries B, Salinger T, Liu D, Hu K, Gensler D, et al. Myocardial fibrosis predicts 10-year survival in patients undergoing aortic valve replacement. Circ Cardiovasc Imaging 2018; 11: e007131.
- 32.
Tröbs SO, Prochaska JH, Schwuchow-Thonke S, Schulz A, Müller F, Heidorn MW, et al. Association of global longitudinal strain with clinical status and mortality in patients with chronic heart failure. JAMA Cardiol 2021; 6: 448–456.
- 33.
Nishimura RA, Otto CM, Bonow RO, Carabello BA, Erwin JP, Guyton RA, et al. 2014 AHA/ACC guideline for the management of patients with valvular heart disease: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2014; 63: e57–e185.
- 34.
Izumi C, Eishi K, Ashihara K, Arita T, Otsuji Y, Kunihara T, et al. JCS/JSCS/JATS/JSVS 2020 guidelines on the management of valvular heart disease. Circ J 2020; 84: 2037–2119.
- 35.
Everett RJ, Tastet L, Clavel MA, Chin CWL, Capoulade R, Vassiliou VS, et al. Progression of hypertrophy and myocardial fibrosis in aortic stenosis: A multicenter cardiac magnetic resonance study. Circ Cardiovasc Imaging 2018; 11: e007451.
- 36.
Kwak S, Everett RJ, Treibel TA, Yang S, Hwang D, Ko T, et al. Markers of myocardial damage predict mortality in patients with aortic stenosis. J Am Coll Cardiol 2021; 78: 545–558.
