Abstract
Background
The prognostic value of follow-up cardiovascular magnetic resonance (CMR) in dilated cardiomyopathy (DCM) patients is unclear. We aimed to investigate the prognostic value of cardiac function, structure, and tissue characteristics at mid-term CMR follow-up.
Methods
The study population was a prospectively enrolled cohort of DCM patients who underwent guideline-directed medical therapy with baseline and follow-up CMR, which included measurement of biventricular volume and ejection fraction, late gadolinium enhancement, native T1, native T2, and extracellular volume. During follow-up, major adverse cardiac events (MACE) were defined as a composite endpoint of cardiovascular death, heart transplantation, and heart-failure readmission.
Results
Among 235 DCM patients (median CMR interval: 15.3 months; interquartile range: 12.5–19.2 months), 54 (23.0%) experienced MACE during follow-up (median: 31.2 months; interquartile range: 20.8–50.0 months). In multivariable Cox regression, follow-up CMR models showed significantly superior predictive value than baseline CMR models. Stepwise multivariate Cox regression showed that follow-up left ventricular ejection fraction (LVEF; hazard ratio [HR], 0.93; 95% confidence interval [CI], 0.91–0.96; p < 0.001) and native T1 (HR, 1.01; 95% CI, 1.00–1.01; p = 0.030) were independent predictors of MACE. Follow-up LVEF ≥ 40% or stable LVEF < 40% with T1 ≤ 1273 ms indicated low risk (annual event rate < 4%), while stable LVEF < 40% and T1 > 1273 ms or LVEF < 40% with deterioration indicated high risk (annual event rate > 15%).
Conclusions
Follow-up CMR provided better risk stratification than baseline CMR. Improvements in the LVEF and T1 mapping are associated with a lower risk of MACE.
Keywords: Cardiac magnetic resonance, Dilated cardiomyopathy, Reverse remodeling, Fibrosis, Prognosis
Graphical abstract
Background
Dilated cardiomyopathy (DCM) refers to a heterogeneous group of diseases characterized by left ventricular or biventricular systolic dysfunction and dilatation without abnormal loading conditions or coronary artery disease [1]. Guideline-directed medical therapy (GDMT) is the first-line treatment for non-ischemic DCM patients and has been shown to improve survival, reduce hospital admissions [2], and facilitate reverse remodeling [3], [4], [5]. While GDMT could lead to significantly improved clinical symptoms and induced reverse remodeling of biventricular volume and systolic function [6], [7], which indicates a better long-term prognosis [5]. However, the follow-up protocols for patients with chronic heart failure are currently inconclusive, and serial echocardiography or cardiac magnetic resonance (CMR) scans are not necessarily recommended [8]. The CMR can yield quantitative assessments with high spatial resolution and good reproducibility in quantitative evaluation of the cardiac structure, function, and myocardial tissue characteristics during follow-up [9]. However, limited evidence is available regarding the prognostic value of follow-up CMR in patients with DCM.
Furthermore, while the underlying mechanism of reverse remolding is unclear, myocardial remodeling may also be an important indicator to determine the therapeutic effect in DCM patients. Myocardial fibrosis is a key pathological change in DCM patients and is significantly associated with poor prognosis in DCM [10], [11], [12]. Growing evidence shows that myocardial fibrosis may reverse after GDMT. A previous observational study showed a significant decrease in myocardial T1 values in DCM patients achieving left ventricular reverse remodeling [9]. Moreover, Mandawat et al. demonstrated that the progression of myocardial fibrosis evaluated by the extent of late gadolinium enhancement (LGE) is associated with minimal left ventricular ejection fraction (LVEF) improvement and a poorer long-term prognosis [13]. The pathophysiology underlying the changes in myocardial fibrosis after GDMT in DCM patients is not fully understood and the potential long-term prognostic value needs further validation.
Thus, the principal aims of the study were to explore the potential prognostic advantage of cardiovascular magnetic resonance (CMR) assessment during mid-term follow-up. We hypothesized that mid-term CMR follow-up has greater prognostic value than the baseline CMR examination, and that dynamic changes in myocardial tissue characteristics could supplement the prognostic value of dynamic changes in cardiac function.
Methods
Study design and informed consent
The study is based on a prospective and consecutive registry study of DCM patients (URL: http://www.chictr.org.cn; Unique identifier: ChiCTR1800017058). Patients diagnosed with DCM who underwent CMR evaluation at baseline and after 1–2 years of treatment were enrolled between February 2016 and December 2020. The study protocol was approved by the Institutional Ethics Committee. Written informed consent was obtained from all participants. All procedures were conducted following the Declaration of Helsinki (2000).
The diagnosis of DCM was based on the World Health Organization/International Society and Federation of Cardiology definition [1]. The inclusion criteria were as follows: (1) reduced LVEF (<45%) on CMR; (2) left ventricle (LV) end-diastolic dimension >55 mm. Exclusion criteria were as follows: (1) significant coronary artery disease assessed by coronary angiography or coronary artery computed tomography (>50% luminal stenosis) or evidence of an ischemic LGE pattern (i.e., perfusion defects or segmental endocardial LGE); (2) primary mitral or aortic valve disease; (3) evidence of myocarditis or hypertensive heart disease, end-stage hypertrophic cardiomyopathy, tachycardia-induced cardiomyopathy, and infiltrative diseases; (4) patient unwillingness to provide informed consent or undergo follow-up CMR examinations; (5) contraindications to CMR examination or poor image quality. The flowchart of patient selection is presented in Fig. 1. All patients underwent GDMT according to the guidelines for the management of patients with heart failure [14].
Fig. 1.
Flowchart of patient selection. CMR: cardiovascular magnetic resonance, GDMT: guideline-directed medical therapy, MACE: major adverse cardiac events.
Follow-up
Patients were followed up for assessment of major adverse cardiac events (MACE), which were defined as a composite of cardiovascular death, heart transplantation, and heart-failure readmission. Follow-up CMR examinations to monitor the therapeutic effect were performed 1–2 years from the baseline with the same scan protocol as the baseline CMR examination. To better evaluate the prognostic value of follow-up clinical and CMR evaluations, MACE that occurred after the mid-term follow-up were taken into consideration for the prognostic analyses. The time to event was calculated from the baseline CMR study. Patients who did not experience events were censored at the last follow-up. MACE evaluation was performed by routine clinic follow-up visits, telephone interviews, and hospital medical records.
Cardiovascular magnetic resonance
All participants underwent CMR imaging on a 3.0T CMR scanner (MAGNETOM Trio or Skyra, Siemens Healthineers, Erlangen, Germany) with a standardized imaging protocol, including cine imaging, LGE assessment, T1 mapping, and T2 mapping [15]. Typical scan parameters of sequences are presented in the Supplementary Materials.
The images were post-processed by Medis Suite 3.2 (Medis, Leiden, the Netherlands) following the guidelines of the Society of Cardiovascular Magnetic Resonance [16]. Quantitative measurements of biventricular end-diastolic and end-systolic volumes and ejection fraction were based on the manually delineated endocardial and epicardial myocardial contours on consecutive short-axis slices with careful avoidance of papillary muscles, blood pool, and epicardial fat [15]. Native T1, T2, and extracellular volume fraction (ECV) values were measured on the same mid-ventricular short-axis slice by delineating the endocardial and epicardial myocardial contours. LGE was independently identified by two experienced CMR observers blinded to patients' information. The LGE extent was assessed using five standard deviation (SDs) threshold above the signal intensity of the remote normal myocardial region on consecutive short-axis slices and calculated as a percentage of LV mass [15]. To reduce the possible differences caused by different scanners and sites, z-scores of T1, T2, and ECV values are provided. The z-score was calculated as (patient value – mean of reference range)/( standard deviation of the reference range), the mean and SD were based on the normative reference data of our previously published cohort [17], [18].
Statistical analysis
The data were presented as mean ± SD or median (interquartile range) for continuous variables and frequency (percentage) for categorical variables. The normality of the distribution was assessed by the Kolmogorov-Smirnov test. Comparisons were performed using the independent t-test or Mann-Whitney U test for continuous variables and the chi-square test or Fisher’s exact test for categorical variables. Univariable Cox regression analysis was performed to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) of characteristics at baseline and follow-up for outcomes of MACE. To identify independent predictors at baseline, follow-up, and the delta change, respectively, variables with P-values <0.10 in univariate models were considered for entry into the stepwise multivariate Cox analysis. In addition, to compare the prognostic value of different clinical and imaging models at baseline, follow-up, and the delta change, multivariate Cox regression models were created as follows: (1) variables showing significant associations with MACE in the univariate Cox analysis were selected, of which variables exhibiting significantly independent value in multivariate Cox models were preferred; (2) age and sex were added as basic clinical considerations in the multivariate models; (3) the best multi-variable models based on the Akaike information criterion (AIC) were determined, with the variance inflation factor calculated to check collinearity. To evaluate the ability of different models to reclassify patients, we calculated net reclassification improvement and integrated discrimination improvement [19]. The discrimination, calibration, and global goodness-of-fit of the multivariable models were evaluated by chi-square and AIC [20]. Kaplan-Meier survival curves were estimated with LVEF and T1 subgroups by using the log-rank test. The highest Youden index of the receiver-operating characteristic curve for MACE was used as the optimal cutoff for LVEF and T1 values during follow-up. The annual event rate of MACE was calculated by dividing the event rate by each group. Annual risk of <4% was considered “low-risk,” 4%−8% was considered “intermediate-risk,” and ≥8% was considered “high-risk” [21]. Two-sided p < 0.05 was statistically significant.
Results
Characteristics of the participants
The study cohort consisted of 235 non-ischemic DCM patients in the Chinese Han population (mean age, 46.1 ± 14.1 years), of whom 158 (67.2%) were male. The baseline LVEF and right ventricular ejection fraction (RVEF) were 25.9% ± 11.2% and 37.5% ± 13.7%, respectively. The cohort was divided based on the occurrence of MACE after CMR follow-up. At baseline, in comparison with the MACE-free survival group, those with MACE were older and had lower body mass index (BMI), a lower frequency of diabetes, lower LVEF, and a larger biventricular volume.
During the mid-term follow-up after a median interval of 15.3 months (interquartile range: 12.5–19.2 months), the MACE-free patients showed significantly greater BMI, lower heart rate, and N-terminal proB natriuretic peptide (NT-proBNP) and troponin T levels, improved New York Heart Association (NYHA) class, significantly improved biventricular ejection fraction, and reduced biventricular volume and LV mass and native T1 values. Patients with MACE also showed significant improvements in NYHA class, NT-proBNP levels, right ventricle (RV) volume, and RVEF, but not LV volume, LV mass, LVEF, and T1. The two patient groups showed no significant difference in mid-term follow-up intervals. Baseline and follow-up characteristics stratified by MACE occurrence are displayed in Table 1. The z-scores of native T1, T2, and ECV values are given in Supplementary Table 1.
Table 1.
Clinical and imaging characteristics at baseline and follow-up
| Patients with event-free survival (n = 181) |
Patients with MACE (n = 54) |
|||||||
|---|---|---|---|---|---|---|---|---|
| Baseline | Follow-up | p value | Baseline | Follow-up | p value | |||
| Clinical parameters | ||||||||
| Age, y | 45.5 ± 14.1 | 47.3 ± 14.4 | <0.001 | 48.2 ± 14.2 | 51.0 ± 17.0 | 0.030 | ||
| Males, n (%) | 122(67.4) | 36 (66.7) | 0.128 | |||||
| BMI, kg/m2 | 24.8 ± 4.4 | 25.5 ± 4.2 | 0.002 | 23.4 ± 3.9* | 23.7 ± 3.8 | 0.091 | ||
| SBP, mmHg | 117.5 ± 16.2 | 120.5 ± 16.2 | 0.082 | 110.2 ± 14.2* | 115.2 ± 19.8 | 0.094 | ||
| DBP, mmHg | 77.6 ± 12.4 | 79.1 ± 12.4 | 0.458 | 74.3 ± 10.6 | 73.7 ± 14.1 | 0.787 | ||
| Heart rate, beats/min | 83.4 ± 16.6 | 75.5 ± 14.6 | <0.001 | 81.7 ± 23.0 | 80.9 ± 14.4 | 0.830 | ||
| HF duration (mo) | 6.0 (1.8−24.0) | 10.5 (2.0-24.0) | 0.860 | |||||
| New onset HF, n (%) | 100 (55.2) | 30 (55.6) | 0.368 | |||||
| NYHA functional class | <0.001 | <0.001 | ||||||
| I, n (%) | 24 (13.3) | 107 (59.1) | 2 (3.7) | 14 (25.9) | ||||
| II, n (%) | 80 (44.2) | 66 (36.5) | 15 (27.8) | 31 (57.4) | ||||
| III, n (%) | 60 (33.1) | 8 (4.4) | 28 (51.9) | 7 (13.0) | ||||
| IV, n (%) | 17 (9.4) | 0 (0) | 9 (16.7) | 2 (3.7) | ||||
| Comorbidity | ||||||||
| Hypertension, n (%) | 34 (18.9) | 29 (16.0) | 5 (9.3) | 4 (7.4) | ||||
| Diabetes, n (%) | 22 (12.2) | 23 (12.7) | 8 (14.8)* | 8 (14.8) | ||||
| Laboratory examination | ||||||||
| Hct | 0.45 ± 0.06 | 0.44 ± 0.05 | 0.124 | 0.46 ± 0.05 | 0.44 ± 0.05 | 0.037 | ||
| NT-proBNP, pg/mL | 813 (306−2246)* | 145 (37−402) | <0.001 | 1621 (825−4026) | 845 (232−1776) | 0.011 | ||
| Troponin T, ng/L | 13 (9−21)* | 9 (6−12) | 0.017 | 17 (13−30) | 22 (11−47) | 0.465 | ||
| Meditations, n (%) | ||||||||
| ARNI, n (%) | 64 (35.4) | 12 (22.2) | ns | |||||
| ACEI/ARB, n (%) | 108 (59.7) | 42 (77.8) | ns | |||||
| β-blocker, n (%) | 166 (91.7) | 46 (85.2) | 0.022 | |||||
| MRA, n (%) | 133 (73.5) | 47 (87.0) | ns | |||||
| Diuretics, n (%) | 117 (64.6) | 47 (87.0) | 0.008 | |||||
| Digoxin, n (%) | 34 (18.8) | 14 (25.9) | ns | |||||
| Warfarin, n (%) | 14 (7.7) | 7 (13.0) | ns | |||||
| CMR parameters | ||||||||
| interval, mo | 15.5 (12.6-19.3) | 14.8 (12.5-19.2) | ns | |||||
| LVEDVi, mL/m2 | 160.4 ± 47.5 | 128.3 ± 46.0 | <0.001 | 189.3 ± 53.7* | 183.5 ± 58.3 | ns | ||
| LVESVi, mL/m2 | 120.3 ± 48.3 | 83.0 ± 43.9 | <0.001 | 150.4 ± 52.9* | 142.5 ± 58.7 | ns | ||
| LVmassi, g/m2 | 76.3 ± 24.5 | 55.4 ± 16.8 | <0.001 | 74.8 ± 26.0 | 70.6 ± 22.5 | ns | ||
| LVEF, % | 27.0 ± 11.2 | 37.9 ± 11.0 | <0.001 | 22.2 ± 10.3* | 24.4 ± 10.9 | ns | ||
| RVEDVi, mL/m2 | 95.9 ± 35.4 | 82.2 ± 21.0 | <0.001 | 106.8 ± 34.2 | 91.6 ± 27.3 | 0.007 | ||
| RVESVi, mL/m2 | 61.9 ± 35.7 | 41.8 ± 17.2 | <0.001 | 74.4 ± 35.8* | 55.3 ± 28.3 | 0.001 | ||
| RVEF, % | 38.6 ± 13.0 | 50.1 ± 9.0 | <0.001 | 33.6 ± 15.0* | 42.2 ± 15.8 | 0.003 | ||
| LGE, n (%) | 71 (38.8) | 73 (39.9) | ns | 33 (63.4) | 35 (67.3) | ns | ||
| LGE extent, % | 10.1 (6.2−15.3) | 6.1(4.8−12.2) | ns | 15.0 (7.6−17.1) | 10.2(6.4−14.2) | ns | ||
| Native T1 mapping, ms | 1311.8 ± 72.7 | 1266.2 ± 68.7 | <0.001 | 1325.2 ± 71.1 | 1311.6 ± 79.8 | ns | ||
| T2 mapping, ms | 42.1 ± 3.3 | 41.9 ± 3.8 | ns | 42.6 ± 3.4 | 43.0 ± 4.0 | ns | ||
| ECV, % | 30.2 ± 5.4 | 29.6 ± 6.0 | ns | 31.5 ± 6.4 | 31.7 ± 9.2 | ns | ||
P value indicates a significant difference between baseline and follow-up in patients with and without MACE.
Abbreviations: AF atrial fibrillation, BMI body mass index, CMR cardiovascular magnetic resonance, DBP diastolic blood pressure, ECV extracellular volume fraction, HR heart rate, LBBB complete left bundle branch block, LGE late gadolinium enhancement, LVEDVi left ventricular end-diastolic volume index, LVEF left ventricular ejection fraction, LVESVi left ventricular end-systolic volume index, LVmassi left ventricular mass index, NYHA New York Heart Association functional classification, NT-proBNP N-terminal prohormone of brain natriuretic peptide, RVEDVi right ventricular end-diastolic volume index, RVEF right ventricular ejection fraction, RVESVi right ventricular end-systolic volume index, SBP systolic blood pressure, MACE major adverse cardiovascular events.
p value indicates a significant difference between patients with DCM with and without MACE at baseline.
Association of baseline and follow-up characteristics
During a median follow-up period of 31.2 months (IQR: 20.8–50.0 months), 54 (23.0%) patients reached the MACE endpoints, which included heart failure (9, 3.8%), sudden cardiac death (6, 2.6%), and heart-failure readmission (48, 20.4%).
At baseline, BMI, systolic blood pressure (SBP), NYHA functional class, NT-proBNP level, LV end-diastolic volume index (LVEDVi), LVEF, RVEF, and the LGE presence and extent showed significant univariable associations with MACE (all p < 0.05). At the follow-up assessment, age, BMI, SBP, diastolic blood pressure (DBP), NYHA functional class, NT-proBNP level, LVEDVi, LV mass index (LVmassi), LVEF, RV end-diastolic volume index (RVEDVi), RVEF, LGE presence, LGE extent, T1, T2, and ECV showed significant associations with MACE (all p < 0.05). In addition, the delta changes in SBP, DBP, NT-proBNP level, LVEDVi, LVmassi, LVEF, and T1 showed significant correlations with MACE. The results are presented in Table 2.
Table 2.
Univariate Cox regression analysis for MACE
| Parameters | Univariate analysis HR (95% CI) |
p value | Multivariate analysis HR (95% CI) |
p value |
|---|---|---|---|---|
| Baseline | ||||
| Clinical parameters | ||||
| Age, y | 1.01 (0.99−1.03) | 0.167 | ||
| Sex, males | 0.97 (0.55−1.71) | 0.919 | ||
| BMI, kg/m2 | 0.92 (0.86−0.99) | 0.019 | ||
| SBP, mmHg | 0.97 (0.95−0.99) | 0.003 | ||
| DBP, mmHg | 0.98 (0.96−1.00) | 0.083 | ||
| HR, beat/min | 0.99 (0.98−1.01) | 0.354 | ||
| NYHA functional class | 1.64 (1.18−2.28) | 0.003 | ||
| ln (NT-proBNP) | 1.34 (1.08−1.68) | 0.009 | ||
| CMR parameters | ||||
| LVEDVi, mL/m2 | 1.01 (1.01−1.02) | 0.001 | 1.01 (1.00−1.02) | 0.004 |
| LVmassi, mL/m2 | 1.00 (0.99−1.01) | 0.953 | ||
| LVEF, % | 0.96 (0.94−0.99) | 0.013 | ||
| RVEDVi, mL/m2 | 1.01 (1.00−1.01) | 0.209 | ||
| RVEF, % | 0.98 (0.96−1.00) | 0.039 | ||
| LGE, n (%) | 3.21 (1.80−5.73) | <0.001 | 2.73 (1.37−5.45) | 0.005 |
| LGE, extent (%) | 1.02 (1.01−1.04) | 0.001 | ||
| T1 mapping, per 10 ms increase | 1.03 (0.99−1.06) | 0.131 | ||
| T2 mapping, ms | 1.05 (0.97−1.14) | 0.227 | ||
| ECV, % | 1.04 (1.00−1.09) | 0.066 | ||
| At follow-up | ||||
| Clinical parameters | ||||
| Age, y | 1.02 (1.00−1.04) | 0.073 | ||
| BMI, kg/m2 | 0.90 (0.84−0.97) | 0.004 | ||
| SBP, mmHg | 0.98 (0.96−1.00) | 0.054 | ||
| DBP, mmHg | 0.97 (0.94−0.99) | 0.018 | ||
| HR, beat/min | 1.02 (1.00−1.04) | 0.139 | ||
| NYHA functional class | 2.65 (1.74−4.02) | <0.001 | ||
| ln (NT-proBNP) | 1.84 (1.51−2.24) | <0.001 | ||
| CMR parameters | ||||
| LVEDVi, mL/m2 | 1.01 (1.01−1.02) | <0.001 | ||
| LVmassi, mL/m2 | 1.03 (1.02−1.04) | <0.001 | ||
| LVEF, % | 0.92 (0.90−0.94) | <0.001 | 0.93 (0.91−0.96) | <0.001 |
| RVEDVi, mL/m2 | 1.01 (1.00−1.02) | 0.011 | ||
| RVEF, % | 0.94 (0.92−0.97) | <0.001 | ||
| LGE, n (%) | 3.79 (1.93−7.44) | <0.001 | ||
| LGE, extent (%) | 1.03 (1.01−1.05) | <0.001 | ||
| T1 mapping, per 10 ms increase | 1.08 (1.04−1.12) | <0.001 | 1.01 (1.00−1.01) | 0.030 |
| T2 mapping, ms | 1.08 (1.01−1.15) | 0.027 | ||
| ECV, % | 1.04 (1.01−1.08) | 0.013 | ||
| Changes (follow-up – baseline) | ||||
| Clinical parameters | ||||
| △BMI, kg/m2 | 0.96 (0.87−1.06) | 0.439 | ||
| △SBP, mmHg | 0.99 (0.99−1.00) | 0.041 | ||
| △DBP, mmHg | 0.99 (0.98−1.00) | 0.016 | ||
| △HR, beat/min | 1.00 (0.99−1.00) | 0.189 | ||
| △NYHA functional class | 1.07 (0.79−1.45) | 0.681 | ||
| △ln (NT-proBNP) | 1.47 (1.19−1.81) | <0.001 | 1.32 (1.04−1.69) | 0.024 |
| CMR parameters | ||||
| △LVEDVi, mL/m2 | 1.01 (1.01−1.02) | <0.001 | ||
| △LVmassi, mL/m2 | 1.03 (1.02−1.04) | <0.001 | 1.02 (1.01−1.04) | 0.005 |
| △LVEF, % | 0.96 (0.94−0.98) | 0.001 | ||
| △RVEDVi, mL/m2 | 1.00 (0.99−1.01) | 0.689 | ||
| △RVEF, % | 0.98 (0.96−1.00) | 0.073 | ||
| △LGE, extent (%) | 1.02 (1.00−1.04) | 0.192 | ||
| △T1 mapping, per 10 ms increase | 1.07 (1.02−1.12) | 0.006 | ||
| △T2 mapping, ms | 1.04 (0.97−1.12) | 0.264 | ||
| △ECV, % | 0.98 (0.98−1.09) | 0.196 | ||
BMI body mass index, CMR cardiovascular magnetic resonance, DBP diastolic blood pressure, ECV extracellular volume fraction, HR heart rate, LGE late gadolinium enhancement, LVEDVi left ventricular end-diastolic volume index, LVEF left ventricular ejection fraction, LVmassi left ventricular mass index, NT-proBNP N-terminal prohormone of brain natriuretic peptide, NHYA New York Heart Association functional classification, RVEDVi right ventricular end-diastolic volume index, RVEF right ventricular ejection fraction, SBP systolic blood pressure.
In multivariable Cox analyses, the LVEDVi (HR, 1.01; 95% CI: 1.00–1.02; p = 0.004) and LGE presence (HR, 2.73; 95% CI: 1.37–5.45; p = 0.005) showed independent significant associations with MACE in the multivariable-adjusted model at baseline. During the follow-up, the LVEF (HR: 0.93, 95% CI: 0.91–0.96, p < 0.001) and native T1 (per 10-ms increase, HR: 1.01, 95% CI: 1.00–1.01, p = 0.030) remained significant independent predictors of MACE. Among the delta changes, the changes in ln (NT-proBNP) (HR: 1.32, 95% CI: 1.04–1.69, p = 0.024) and LVmassi (HR: 1.02, 95% CI: 1.01–1.04, p = 0.005) showed independent significant associations with MACE.
In the multivariable enter models (without stepwise selection), variables with the greatest Wald statistic were included avoiding collinearity. At baseline, the clinical parameters including age, sex, NYHA functional class, and ln (NT-proBNP). The addition of LVEF to the clinical model yielded a significantly improved model, while the addition of LGE presence and LVEF yielded a significant increase in predictive power. The clinical and imaging models based on follow-up evaluations showed significant incremental prognostic power compared to the corresponding models at baseline. The results are presented in Supplementary Figure 1 and Supplementary Table 2.
Outcomes stratified by LVEF and native T1 values
Using Kaplan–Meier analysis, we compared the cumulative incidence of MACE based on the follow-up status and delta changes in LVEF and native T1. The best cutoff LVEF at the follow-up to predict MACE was 40.7% (sensitivity, 0.88; specificity, 0.54; AUC = 0.68), and the best cutoff native T1 at follow-up was 1273.5 ms (sensitivity, 0.76; specificity, 0.61; AUC = 0.76). Based on the LVEF value, four groups were identified: Group 1: stable LVEF ≥ 40%, with LVEF ≥ 40% at baseline and follow-up; Group 2: improvement to LVEF ≥ 40%, with baseline LVEF < 40% and follow-up LVEF ≥ 40%; Group 3: stable LVEF < 40%, with follow-up LVEF < 40% and delta LVEF ≥ −5%; Group 4: deteriorated LVEF < 40%, with follow-up LVEF < 40% and delta LVEF < −5%. Similarly, based on the native T1 value, four groups were identified: Group 1: stable native T1 ≤ 1273 ms, with native T1 ≤ 1273 ms at both baseline and follow-up; Group 2: improvement to native T1 ≤ 1273 ms, with baseline native T1 > 1273 ms and follow-up native T1 ≤ 1273 ms; Group 3: stable native T1 > 1273 ms, with follow-up native T1 > 1273 ms and delta native T1 < 50 ms; and Group 4: deteriorated native T1, with native T1 > 1273 ms and delta native T1 ≥ 50 ms.
In the LVEF subgroup analysis, patients in Groups 3 and 4 showed a significantly higher risk of MACE than those in Groups 1 and 2 (both p < 0.001), while patients in Group 4 showed a significantly higher risk than those in Group 3 (p = 0.023). Thus, in comparison with patients showing follow-up LVEF ≥ 40%, those in Group 3 (HR, 11.2; 95% CI, 3.5–36.5, p < 0.001) and Group 4 (HR, 20.1; 95% CI, 5.8–70.3, p < 0.001) showed a significantly higher risk of reaching MACE. In the T1 mapping subgroup analysis, patients in Groups 3 and 4 showed a significantly higher risk of MACE than those in Groups 1 and 2 (both p < 0.001). Thus, in comparison with patients showing follow-up LVEF ≤ 1273 ms, those in Groups 3 (HR, 5.8; 95% CI, 3.0–11.6; p < 0.001) and 4 (HR, 6.5; 95% CI, 2.4–17.6; p < 0.001) showed a significantly higher risk of reaching MACE. The Kaplan–Meier curves stratified by the LVEF and native T1 subgroups are shown in Fig. 2.
Fig. 2.
Kaplan–Meier survival curve showing event-free survival stratified by LVEF and native T1 subgroups. Subgroups stratified by LVEF: Group 1: stable LVEF ≥ 40%; Group 2: improved to LVEF ≥ 40%; Group 3: stable LVEF < 40%; Group 4: LVEF < 40% and deterioration. Subgroups stratified by native T1: Group 1: stable native T1 ≤ 1273 ms; Group 2: improved to native T1 ≤ 1273 ms; Group 3: stable native T1 > 1273 ms; Group 4: native T1 > 1273 ms and deterioration. CI: confidence interval, HR: hazard ratio, LVEF: left ventricular ejection fraction.
To further analyze the prognostic value of CMR follow-up, we performed a subgroup analysis based on LVEF plus native T1 status, and six subgroups were identified as shown in Fig. 3. The subgroups showed a gradient increase in the risk of MACE. Among patients with stable LVEF < 40%, those with follow-up native T1 < 1273 ms showed significantly lower risk than those with native T1 ≥ 1273 ms (log-rank test, p < 0.001). In comparison, among patients with stable LVEF ≥ 40% and deterioration, survival was relatively unaffected by the status of native T1 during follow-up.
Fig. 3.
Risk stratification analyses based on the combination of LVEF measurements and T1 mapping. Kaplan–Meier analysis (A) based on subgroups stratified by the combination of LVEF and native T1: Group 1: follow-up LVEF ≥ 40% and native T1 ≤ 1237 ms; Group 2: follow-up LVEF ≥ 40% and native T1 > 1237 ms; Group 3: stable follow-up LVEF < 40% and native T1 ≤ 1237 ms; Group 4: stable follow-up LVEF < 40% and native T1 > 1237 ms; Group 5: follow-up LVEF < 40% with deterioration and native T1 ≤ 1237 ms; Group 4: stable follow-up LVEF < 40% and native T1 > 1237 ms. Bar graph showing the event rate (B) and annual event rate (C) stratified by the six subgroups. LVEF: left ventricular ejection fraction.
The Kaplan–Meier curves and bar graphs of event rates stratified by the combination of LVEF and native T1 status are presented in Fig. 3. The event rates were 3.0%, 4.3%, 13.5%, 43.9%, 44.4%, and 57.9% for Group 1 to Group 6, respectively. Patients with LVEF ≥ 40% and native T1 ≤ 1273 ms at follow-up experienced the lowest rate of MACE (0.8% per year). In contrast, patients with LVEF < 40% and deterioration and native T1 > 1273 ms had the highest event rate of 24.6%. Notably, among patients with stable LVEF < 40%, those with native T1 > 1273 ms showed a significantly higher risk of MACE than those with native T1 ≤ 1273 ms (annual event rates of 3.8% and 17.8%).
In the graphical abstract, we show a simplified algorithm to stratify patients into two categories based on the annual event rates. The low-risk group includes patients with follow-up LVEF ≥ 40% and those with stable follow-up LVEF < 40% and follow-up native T1 ≤ 1273 ms. Meanwhile, the high-risk group includes patients with stable follow-up LVEF < 40% and follow-up native T1 > 1273 ms as well as those with follow-up LVEF < 40% and deterioration.
Discussion
This is the first study to comprehensively investigate the prognostic value of mid-term CMR follow-up in DCM patients. The main findings are as follows: (1) follow-up CMR showed better risk stratification than baseline CMR, and the LVEF and native T1 values at follow-up had independent and significant prognostic value. (2) After 1–2 years of GDMT, DCM patients who achieved LVEF ≥ 40% or maintained LVEF at ≥40% showed a significantly lower risk of MACE than patients who had LVEF < 40% at follow-up. During follow-up CMR, patients with stable LVEF < 40% and those with LVEF < 40% and deterioration showed 11.2-fold and 20.1-fold higher risk of MACE than patients with LVEF ≥ 40%, respectively. (3) T1 mapping at follow-up CMR allowed further risk stratification of patients with stable LVEF < 40%, since patients with T1 > 1273 ms showed a 3.4-fold greater risk of MACE than patients with T1 ≤ 1273 ms.
These results are of important clinical significance. While most studies to date have focused on the prognostic value of baseline parameters in patients with DCM and chronic heart failure patients [10], [22], [23], we confirmed that the model including follow-up characteristics showed additional prognostic power than the baseline model [5], [6], and the combination of follow-up LVEF and native T1 status showed a good risk-stratifying effect.
We found that after 1–2 years of GDMT, patients with LVEF ≥ 40% and native T1 < 1273 ms showed significantly better prognosis compared with other groups. Thus, the achievement or maintenance of LVEF ≥ 40% and native T1 < 1273 ms may be considered a new, simple, and efficient clinically feasible standard for “reverse remodeling” in DCM patients and serve as a treatment goal. Till now, one of the major hurdles for the clinical application of imaging follow-up in DCM patients is the absence of standardization in the definition of recovery or remission [24], [25], [26], [27]. At present, the definition of left ventricular reverse remodeling is mainly based on an improvement in LVEF and a reduction in LV volume during follow-up. However, the standards are not uniform. While some studies are based on the LVEF value at follow-up, which usually ranges from 35%−45%, others evaluate the absolute increase in the LVEF value [5], [28], [29].
Although echocardiography was more widely available during follow-up compared with CMR, we evaluate the dynamic changes of cardiac structure, function, and myocardial tissue characteristics using CMR due to the advantage of better reproducibility and the ability of tissue characteristics evaluation. The results further confirmed that the myocardial fibrosis levels provided by follow-up CMR could provide additional prognostic values in DCM patients. In a previous study, Mandawat et al. found that the progression of myocardial fibrosis evaluated with LGE extent was associated with a poor prognosis in DCM patients [13]. In this study, we quantitatively evaluated the progression of diffuse myocardial fibrosis based on native T1 values, which offered the following advantages: (1) native T1 measurement was less time-consuming than LGE extent analysis (1 min per patient vs. 5–8 min per patient). Moreover, the measurements of LGE extent depend on the normal myocardium as a reference, potentially introducing subjectivity in the judgment process. (2) The LGE extent showed a downward trend but no significant changes while the native T1 increased significantly, suggesting that T1 mapping may be more sensitive to changes in myocardial tissue characteristics in DCM patients. (3) Changes in native T1 and LGE presence are significantly independent variables. Although LGE-negative patients could not undergo dynamic assessment of changes in myocardial fibrosis by LGE, T1 mapping could allow such assessments in this patient population [12]. Moreover, the good prognostic value of follow-up T1 mapping for evaluation of myocardial fibrosis and the limited scope of changes in LGE presence and patterns highlight the discussion on necessity of enhanced CMR during follow-up in patients with DCM.
Importantly, this study presents a new algorithm for risk stratification for DCM patients based on the baseline and follow-up CMR examinations. While exploring the most reasonable risk stratification analysis, we divided patients into multiple small subgroups, which may not have sufficient power due to the relatively small cohort. The Kaplan–Meier analysis of LVEF showed that patients with LVEF > 40% during the follow-up period generally had a better prognosis, irrespective of whether LVEF was maintained at ≥40% or improved further. On the other hand, for patients with LVEF < 40% at follow-up, the dynamic change in LVEF showed prognostic value. If the LVEF during the follow-up showed deterioration of >5%, the MACE risk increased in comparison with relatively stable patients. Unlike LVEF evaluations, T1 mapping at follow-up with a cutoff of 1273 ms can be well used for risk stratification of patients. Therefore, we finally combined the groups based on the annual event rates and established a simplified algorithm to stratify them into high- and low-risk groups to facilitate clinical practice.
There are still some important clinical issues that remained unresolved in this study. In patients with MACE, we found that there was no significant decrease in heart rate, which may indicate that the treatment effect of these patients is not optimal, or may explain the occurrence of MACE in those patients [30]. Thus, how to evaluate the treatment efficacy of GDMT in patients with poor response and whether the optimal treatment should be adjusted in time is worth further discussion. In addition, as NT-proBNP level decreased significantly at follow-up, this may indicate that in the baseline model, some patients are in the stage of acute decompensated heart failure, while in the follow-up model, most patients are in a chronic state. Future studies may need to discuss whether different disease states affect myocardial tissue characteristic parameters, such as T1 and T2 values. Furthermore, Manca et al. observed that approximately 40% DCM patients with improved ejection fraction (impEF) after treatment would undergo a recurrent decline in LVEF [28]. Further studies are needed to investigate whether myocardial fibrosis would affect the subsequent functional decline in impEF patients.
Limitations
This study had several limitations: (1) the follow-up time after CMR revaluation of the cohort was limited. Although the current MACE rate after CMR follow-up has sufficient statistical power, a longer follow-up period is needed for more detailed observations in the future; (2) although sodium glucose co-transporter 2 inhibitor (SGLT2i) is a new therapeutic agent with Class of Recommendation 2a for HF patients, it was not widely available at the time of our observational study, thus the analyses of the usage of SGLT2i were not included in the present study; (3) the dose-dependent patterns of medicine usage were not monitored. The medication usage recorded by us was the main medication during the treatment period. The dose of the medication and the short-term adjustment were not recorded in detail, which may not be conducive to the analysis of the dose-dependent manner; (4) the inclusion of patients who had completed follow-up CMR examinations could be a source of bias.
Conclusion
CMR follow-up provides better risk stratification than baseline CMR, with LVEF and native T1 values showing independent and significant prognostic value. DCM patients with LVEF ≥ 40% at follow-up had the best prognosis regardless of the baseline level, and the patients with LVEF < 40% and deterioration showed the worst prognosis. Among patients with stable LVEF < 40%, those showing maintenance or improvement of myocardial native T1 to ≤1237 ms had significantly less risk than those with follow-up T1 > 1237 ms.
Funding
This work was supported by grants from the National Natural Science Foundation of China (Grant Nos. 81571638, 82202248, and 82302290), China Postdoctoral Science Foundation [Grant Nos. 2023M732461x and 2023TQ0227], Natural Science Foundation of Sichuan Province [Grant No. 23NSFSC4589], and 1.3.5 Project for Disciplines of Excellence, West China Hospital, Sichuan University (Grant No. ZYGD22013).
Author contributions
Wang Jie: Data curation, Formal analysis. Li Weihao: Data curation, Formal analysis, Investigation, Methodology. Tan Yinxi: Data curation, Formal analysis. Ghaithan Saeed: Writing – review and editing. Wan Ke: Conceptualization, Methodology, Project administration, Resources, Writing – review and editing. Wang Shiqian: Data curation, Methodology. Li Yangjie: Conceptualization, Data curation, Resources, Software, Writing – original draft. Xu Yuanwei: Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Writing – original draft. Chen Yucheng: Conceptualization, Funding acquisition, Writing – review and editing. Zhang Qing: Conceptualization. Sun Jiayu: Conceptualization, Visualization. Cheng Wei: Data curation, Validation, Visualization. Guo Jiajun: Data curation, Visualization. Han Yuchi: Conceptualization.
Ethics approval and consent
The study was approved by the local Institutional Ethics Committee of each center, and all procedures were following the Declaration of Helsinki. All participants provided written informed consent.
Consent for publication
Not applicable.
Declaration of competing interest
None declared.
Acknowledgments
Acknowledgements
None declared.
Footnotes
Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jocmr.2024.101002.
Appendix A. Supplementary material
Supplementary material
.
References
- 1.Pinto Y.M., Elliott P.M., Arbustini E., Adler Y., Anastasakis A., Böhm M., et al. Proposal for a revised definition of dilated cardiomyopathy, hypokinetic non-dilated cardiomyopathy, and its implications for clinical practice: a position statement of the ESC working group on myocardial and pericardial diseases. Eur Heart J. 2016;37(23):1850–1858. doi: 10.1093/eurheartj/ehv727. [DOI] [PubMed] [Google Scholar]
- 2.Merlo M., Pivetta A., Pinamonti B., Stolfo D., Zecchin D., Barbati G., et al. Long-term prognostic impact of therapeutic strategies in patients with idiopathic dilated cardiomyopathy: changing mortality over the last 30 years. Eur J Heart Fail. 2014;16(3):317–324. doi: 10.1002/ejhf.16. [DOI] [PubMed] [Google Scholar]
- 3.Halliday B.P., et al. Myocardial remodelling after withdrawing therapy for heart failure in patients with recovered dilated cardiomyopathy: insights from TRED-HF. Eur J Heart Fail. 2021;23(2):293–301. doi: 10.1002/ejhf.2063. [DOI] [PubMed] [Google Scholar]
- 4.Halliday B.P., Owen R., Gregson J., Vassiliou V.S., Chen X., Wage R., et al. Beta-blocker use is associated with prevention of left ventricular remodeling in recovered dilated cardiomyopathy. J Am Heart Assoc. 2021;10(12) doi: 10.1161/JAHA.120.019240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Merlo M., Pyxaras S.A., Pinamonti B., Barbati G., Lenarda A.D., Sinagra G. Prevalence and prognostic significance of left ventricular reverse remodeling in dilated cardiomyopathy receiving tailored medical treatment. J Am Coll Cardiol. 2011;57(13):1468–1476. doi: 10.1016/j.jacc.2010.11.030. [DOI] [PubMed] [Google Scholar]
- 6.Merlo M., Gobbo M., Stolfo D., Losurdo P., Ramani F., Barbati G., et al. The prognostic impact of the evolution of RV function in idiopathic DCM. JACC Cardiovasc Imaging. 2016;9(9):1034–1042. doi: 10.1016/j.jcmg.2016.01.027. [DOI] [PubMed] [Google Scholar]
- 7.Cojan-Minzat B.O., Zlibut A., Agoston-Coldea L. Non-ischemic dilated cardiomyopathy and cardiac fibrosis. Heart Fail Rev. 2021;26(5):1081–1101. doi: 10.1007/s10741-020-09940-0. [DOI] [PubMed] [Google Scholar]
- 8.McDonagh T.A., Metra M., Adamo M., Gardner R.S., Baumbach A., Böhm M., et al. ESC guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021;42(36):3599–3726. doi: 10.1093/eurheartj/ehab368. [DOI] [PubMed] [Google Scholar]
- 9.Xu Y., Li W., Wan K., Liang Y., Jiang X., Wang J., et al. Myocardial tissue reverse remodeling after guideline-directed medical therapy in idiopathic dilated cardiomyopathy. Circ Heart Fail. 2021;14(1) doi: 10.1161/CIRCHEARTFAILURE.120.007944. [DOI] [PubMed] [Google Scholar]
- 10.Alba A.C., Gaztañaga J., Foroutan F., Thavendiranathan P., Merlo M., Alonso-Rodriguez D., et al. Prognostic value of late gadolinium enhancement for the prediction of cardiovascular outcomes in dilated cardiomyopathy: an international, multi-institutional study of the MINICOR group. Circ Cardiovasc Imaging. 2020;13(4) doi: 10.1161/CIRCIMAGING.119.010105. [DOI] [PubMed] [Google Scholar]
- 11.Puntmann V.O., Carr-White G., Jabbour A., Yu C.Y., Gebker R., Kelle S., et al. T1-mapping and outcome in nonischemic cardiomyopathy all-cause mortality and heart failure. JACC Cardiovasc Imaging. 2016;9(1):40–50. doi: 10.1016/j.jcmg.2015.12.001. [DOI] [PubMed] [Google Scholar]
- 12.Li S., Zhou D., Sirajuddin A., He J., Xu J., Zhuang B., et al. T1 mapping and extracellular volume fraction in dilated cardiomyopathy: a prognosis study. JACC Cardiovasc Imaging. 2022;15(4):578–590. doi: 10.1016/j.jcmg.2021.07.023. [DOI] [PubMed] [Google Scholar]
- 13.Mandawat A., Chattranukulchai P., Mandawat A., Blood A.J., Ambati S., Hayes B., et al. Progression of myocardial fibrosis in nonischemic DCM and association with mortality and heart failure outcomes. JACC Cardiovasc Imaging. 2021;14(7):1338–1350. doi: 10.1016/j.jcmg.2020.11.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Heidenreich P.A., Bozkurt B., Aguilar D., Allen L.A., Byun J.J., Colvin M.M., et al. AHA/ACC/HFSA guideline for the management of heart failure: executive summary: a report of the American College of Cardiology/American Heart Association Joint Committee on clinical practice guidelines. Circulation. 2022;145(18):e876–e894. doi: 10.1161/CIR.0000000000001062. [DOI] [PubMed] [Google Scholar]
- 15.Schulz-Menger J., Bluemke D.A., Bremerich J., Flamm S.D., Fogel M.A., Friedrich M.G., et al. Standardized image interpretation and post-processing in cardiovascular magnetic resonance - 2020 update: Society for Cardiovascular Magnetic Resonance (SCMR): Board of Trustees Task Force on Standardized Post-Processing. J Cardiovasc Magn Reson. 2020;22(1):19. doi: 10.1186/s12968-020-00610-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Schulz-Menger J., Bluemke D.A., Bremerich J., Flamm S.D., Fogel M.A., Friedrich M.G. Standardized image interpretation and post processing in cardiovascular magnetic resonance: Society for Cardiovascular Magnetic Resonance ( SCMR) Board of Trustees Task Force on Standardized Post Processing. J Cardiovasc Magn Reson. 2013;15(1):35. doi: 10.1186/1532-429X-15-35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Liang Y., Xu Y., Li W., Wan K., Sun J., Lin J., et al. Left ventricular function recovery in peripartum cardiomyopathy: a cardiovascular magnetic resonance study by myocardial T1 and T2 mapping. J Cardiovasc Magn Reson. 2020;22(1):2. doi: 10.1186/s12968-019-0590-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Dong Y., Yang D., Han Y., Cheng W., Sun J., Wan K., et al. Age and gender impact the measurement of myocardial interstitial fibrosis in a healthy adult Chinese population: a cardiac magnetic resonance study. Front Physiol. 2018 6;9:140. doi: 10.3389/fphys.2018.00140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Pencina M.J., D'Agostino R.B., D'Agostino R.B., Vasan R.S. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med. 2008;27(2):157–172. doi: 10.1002/sim.2929. [DOI] [PubMed] [Google Scholar]
- 20.Verdonschot J.A.J., Hazebroek M.R., Wang P., Wijk S.S., Merken J.J., Adriaansen Y.A., et al. Clinical phenotype and genotype associations with improvement in left ventricular function in dilated cardiomyopathy. Circ Heart Fail. 2018;11(11) doi: 10.1161/CIRCHEARTFAILURE.118.005220. [DOI] [PubMed] [Google Scholar]
- 21.Melichova D., et al. Strain echocardiography improves prediction of arrhythmic events in ischemic and non-ischemic dilated cardiomyopathy. Int J Cardiol. 2021;342:56–62. doi: 10.1016/j.ijcard.2021.07.044. [DOI] [PubMed] [Google Scholar]
- 22.Melichova D., Nguyen T.M., Salte I.M., Klaeboe L.G., Sjøli B., Karlsen S., et al. The combination of carboxy-terminal propeptide of procollagen type I blood levels and late gadolinium enhancement at cardiac magnetic resonance provides additional prognostic information in idiopathic dilated cardiomyopathy – a multilevel assessment of my. Eur J Heart Fail. 2021;23(6):933–944. doi: 10.1002/ejhf.2201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Muser D., Liang J.J., Castro S.A., Lanera C., Enriquez A., Kuo L., et al. Performance of prognostic heart failure models in patients with nonischemic cardiomyopathy undergoing ventricular tachycardia ablation. JACC Clin Electrophysiol. 2019;5(7):801–813. doi: 10.1016/j.jacep.2019.04.001. [DOI] [PubMed] [Google Scholar]
- 24.Aimo A., Vergaro G., González A., Barison A., Lupón J., Delgado V., et al. Cardiac remodelling – part 2: clinical, imaging and laboratory findings. a review from the Study Group on Biomarkers of the Heart Failure Association of the European Society of Cardiology. Eur J Heart Fail. 2022;24(6):944–958. doi: 10.1002/ejhf.2522. [DOI] [PubMed] [Google Scholar]
- 25.Kao D.P., Lowes D.B., Gilbert E.M., Minobe W., Epperson L.E., Meyer L.K., et al. Therapeutic molecular phenotype of β-blocker-associated reverse-remodeling in nonischemic dilated cardiomyopathy. Circ Cardiovasc Genet. 2015;8(2):270–283. doi: 10.1161/CIRCGENETICS.114.000767. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Hazebroek M.R., Moors S., Dennert R., Wijngaard A..V.D., Krapels I., Hoos M., et al. Prognostic relevance of gene-environment interactions in patients with dilated cardiomyopathy applying the MOGE(S) classification. J Am Coll Cardiol. 2015;66(12):1313–1323. doi: 10.1016/j.jacc.2015.07.023. [DOI] [PubMed] [Google Scholar]
- 27.Wilcox J.E., Fang J.C., Margulies K.B., Mann D.L. Heart failure with recovered left ventricular ejection fraction: JACC Scientific Expert Panel. J Am Coll Cardiol. 2020;76(6):719–734. doi: 10.1016/j.jacc.2020.05.075. [DOI] [PubMed] [Google Scholar]
- 28.Manca P., Stolfo D., Merlo M., Gregorio C., Cannatà A., Ramani F., et al. Transient versus persistent improved ejection fraction in non-ischaemic dilated cardiomyopathy. Eur J Heart Fail. 2022;24(7):1171–1179. doi: 10.1002/ejhf.2512. [DOI] [PubMed] [Google Scholar]
- 29.Henkens M.T.H.M., Stroeks S.L.V.M., Raafs A.G., Sikking M.A., Tromp J., Ouwerkerk W., et al. Dynamic ejection fraction trajectory in patients with dilated cardiomyopathy with a truncating titin variant. Circ Heart Fail. 2022;15(8) doi: 10.1161/CIRCHEARTFAILURE.121.009352. [DOI] [PubMed] [Google Scholar]
- 30.Halliday B.P., Vazir A., Owen R., Gregson J., Wassall R., Lota A.S., et al. Heart rate as a marker of relapse during withdrawal of therapy in recovered dilated cardiomyopathy. JACC Heart Fail. 2021;9(7):509–517. doi: 10.1016/j.jchf.2021.03.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary material