Association between layer-specific global longitudinal strain and adverse outcomes following acute coronary syndrome

Association between layer-specific global longitudinal strain and adverse outcomes following... Abstract Aims To investigate the prognostic value of layer-specific global longitudinal strain (GLS) in predicting heart failure (HF) and cardiovascular death (CD) following acute coronary syndrome (ACS). Methods and results In this retrospective study, 465 ACS patients underwent transthoracic echocardiography following percutaneous coronary intervention (PCI). The primary endpoint was the composite of HF and/or CD with a median follow-up time of 4.6 (0.2–6.3) years. During follow-up 199 patients (42.7%) suffered HF and/or CD (176 developed HF and 38 suffered CD). Absolute endomyocardial global longitudinal strain (GLSendo) (12% vs. 17%, P < 0.001), GLS (11% vs. 14%, P < 0.001), and epimyocardial global longitudinal strain (GLSepi) (9% vs. 13%, P < 0.001) were all reduced in patients with an adverse outcome. In multivariable Cox regressions, which included clinical baseline characteristics and conventional echocardiographic measurements, GLS obtained from all layers remained independently associated with the composite outcome; GLSendo [hazard ratio: 1.19 (1.10–1.28), P < 0.001, per 1% decrease], GLS [hazard ratio 1.24 (1.14–1.35), P < 0.001, per 1% decrease], and GLSepi [hazard ratio 1.26 (1.15–1.39), P < 0.001, per 1% decrease]. No other echocardiographic measures remained independently associated with the composite outcome in these models. Finally, GLS and GLSepi provided incremental prognostic information on the risk of developing the composite endpoint, when added to all other clinical and echocardiographic measures [adding GLS (c-statistics: 0.76 vs. 0.74, P = 0.048) or adding GLSepi (c-statistics: 0.76 vs. 0.74, P = 0.039)]. Conclusion In ACS patients, layer-specific strain provides independent prognostic information regarding risk of developing HF and/or CD. Furthermore, only GLS and GLSepi provided incremental prognostic information when added to all other significant predictors. acute coronary syndrome , 2D-speckle tracking echocardiography , heart failure , cardiovascular death , layer-specific 2D-speckle tracking Introduction A common and adverse complication following a myocardial infarction (MI) is heart failure (HF). To prevent HF, and improve the prognosis it is necessary to identify high risk patients to initiate more intensive monitoring or treatment. Prior to the clinical symptoms of HF become clear, patients can develop asymptomatic systolic or diastolic left ventricular (LV) dysfunction caused by structural or functional cardiac abnormalities which are precursors of HF.1–6 Recent studies have shown global longitudinal strain (GLS) obtained from 2D-speckle tracking echocardiography to be a strong predictor of adverse outcomes in several cardiac diseases.5,7–11 GLS has already been suggested as a more sensitive marker of cardiac function than conventional echocardiography.4,4–6,12,13 It has been demonstrated to be a more powerful predictor of outcome than left ventricular ejection fraction (LVEF) as it may reflect subclinical LV systolic dysfunction.6,14 Novel echocardiographic software can now sectionalize the myocardial wall in layers. This has made it possible to detect which layers of the myocardial wall that suffer from reduced deformation. Layer-specific GLS might be important since the LV wall is heterogeneous and consists of three different layers of muscle fibres.15 Especially in ischaemic heart disease, layer-specific strain may be particularly useful since longitudinally orientated myocardial fibres located in the sub-endocardium is known to be most susceptible to ischaemia.16 There are no studies that have investigated the prognostic value of layer-specific GLS in an acute coronary syndrome (ACS) population. This is what we aim to examine, especially the association between layer-specific GLS and the risk of developing HF and cardiovascular death (CD). Methods Study population The Cardiology Department of Copenhagen University Hospital Gentofte is an invasive hub for 10 non-invasive cardiology departments. In the period January 2003 to November 2008, 5003 patients were admitted to the department for PCI, these patients were all included in a clinical registry. All patients in this registry were eligible for our study if they had an echocardiogram available. We hereby identified 580 ACS patients [ST-segment elevation myocardial infarction (STEMI), non-ST-segment elevation myocardial infarction (NSTEMI), or unstable angina pectoris (UAP)] who were admitted to Copenhagen University Hospital Gentofte, Denmark for percutaneous coronary intervention (PCI) in the period and had an echocardiogram performed at Gentofte Hospital. A considerable number of the 5003 patients were admitted to other non-invasive cardiology departments and were rapidly transported back to their local hospital post-PCI. The echocardiographic examinations were performed [median 2 days (1–3 days)] after the PCI procedure. In addition, 115 patients were excluded due to a non-sinus rhythm, missing images, or inadequate image quality. Outcomes and endpoints Information regarding endpoints—development of HF and CD—were retrieved from the Danish National Board of Health’s National Patient Registry and the Danish Register of Causes of Death using International Classifications of Diseases codes. Echocardiography The echocardiographic examinations were performed by experienced clinicians and sonographers on GE Vivid machines (GE Healthcare, Little Chalfont, UK). The images were stored in a GE Health care image vault and analysed offline with EchoPac version 113 (GE Healthcare, Horten, Norway) by an experienced investigator blinded to clinical baseline data and endpoints. Conventional 2D echocardiography The LV dimensions, interventricular septal thickness (IVSd), LV internal diameter (LVIDd), and LV posterior wall (LVPWd) thickness were measured in the parasternal long-axis view at the tip of the mitral valve leaflet in end-diastole. LV mass index (LVMI) was calculated as the anatomical mass17 divided by body surface area. Mitral valve inflow patterns: peak early filling (E-wave), atrial filling (A-wave) and deceleration time of the early filling (DT) were measured with pulsed-wave Doppler imaging at the tips of the mitral valve leaflets in the apical 4-chamber view. Additionally, early (e′) peak myocardial diastolic velocity was measured with pulsed-wave tissue Doppler imaging in the apical 4-chamber view at the lateral and septal walls at the mitral annulus. The e′ was indexed to the E-wave velocity to obtain the E/e′. LVEF was measured with the modified Simpson’s biplane method obtained from the apical 4- and 2-chamber views. Left atrial volume (LAV) was obtained through Simpson’s biplane method as well. Using m-mode tricuspid annular plane systolic excursion (TAPSE) was measured in the apical 4-chamber view. 2D speckle tracking echocardiography Speckle tracking analysis was performed in the three apical views—4-chamber, 2-chamber, and the longitudinal long-axis (Figure 1). The endocardium of the LV was traced with a semiautomatic function and adjusted manually with a point-and-click method if the investigator thought the tracing to be inaccurate. Width of the regions of interest (ROI) covered the endo-, myo- and epicardium and was adjusted regionally when required. Each view covered six segments—consequently a total of 18 segments were included. Global values were calculated manually as the mean value of the peak systolic longitudinal strain of each segment. Segments if deemed untraceable could be excluded at the discretion of the analyst. Two layers of longitudinal layer-specific strain: the endomyocardial (GLSendo) and epimyocardial (GLSepi) were calculated by the software. Whole wall GLS which encompasses the whole ROI was calculated by the software as well. Figure 1 View largeDownload slide Examples of 2D-speckle tracking echocardiography. Screenshots of 2D-speckle tracking echocardiographic analysis with tracking of the LV with regional longitudinal strain traces. The views shown are the apical 4-chamber (top), 2-chamber (middle), and longitudinal long-axis (bottom). Figure 1 View largeDownload slide Examples of 2D-speckle tracking echocardiography. Screenshots of 2D-speckle tracking echocardiographic analysis with tracking of the LV with regional longitudinal strain traces. The views shown are the apical 4-chamber (top), 2-chamber (middle), and longitudinal long-axis (bottom). Statistical analysis Student’s t-test was used in comparing continuous variables exhibiting Gaussian distribution, which were displayed as mean values ± standard deviation. Wilcoxon rank-sum test was utilized for comparing continuous non-Gaussian distributed variables (LVMI and E/e′) with interquartile ranges displayed. χ2 test was used in comparing categorical variables and expressed as frequencies (percentages). Uni- and multivariable Cox regression models were created to correlate clinical and echocardiographic findings to the endpoints. TAPSE was excluded from the multivariable Cox regression due to missing values in 58 (12%) patients. For multivariable Cox regressions, three models with incremental number of confounders were constructed. Model 1 included either GLSendo, GLS, or GLSepi and the following confounders: age, gender, body mass index (BMI), prior cardiovascular disease (CVD), family history of CVD, diabetes mellitus (DM), current smoker, hypercholesterolaemia, systolic blood pressure, diastolic blood pressure, and heart rate (HR). In Model 2, we also included variables obtained from the coronary angioplasty and ACS category—STEMI, multivessel disease, culprit lesions, left anterior descending (LAD) occlusion, and left main stem coronary artery (LMS) occlusion. Model 3 additionally included all significant echocardiographic parameters—LVMI, LVEF, DT, e′, and E/e′. To compare prognostic potential of our variables, Harrell c-statistics was calculated from the univariable Cox regressions for the variables included in the multivariable Cox regressions. The Kaplan–Meier curves were created for our cohort stratified into tertiles of layer-specific GLS. Comparisons of the Harrell’s c-statistics of the diagnostic models (Models 1, 2, and 3 with or without a GLS variable) were made to assess the incremental prognostic information gained from layered GLS. A comparison with a P-value ≤0.05 in two-tailed tested was deemed statistically significant. STATA Statistics/Data analysis, MP 13.0 (StataCorp, TX, USA) was used for the statistical analysis. Results Of the 465 patients included, 176 (37.8%) developed HF, 38 suffered CD (8.2%), with 15 (3.2%) developing both. Overall there were 199 (42.7%) events during a median follow-up time of 4.6 (0.2–6.3) years. Follow-up was 100%. The mean age of our study sample was 66 ± 12 years and 74.0% were males. The HF/CD group were significantly older (67 years vs. 64 years, P = 0.019), had higher HR during the echocardiogram [79 beats per minute (bpm) vs. 71 bpm, P < 0.001], had a higher frequency of prior CVD (14.1% vs. 6.4%, P = 0.006), had a lower frequency of history of CVD in the family (23.6% vs. 34.6%, P = 0.011), a higher rate of LAD lesion (59.3% vs. 43.6%, P < 0.001), and LMS stenosis (1.5% vs. 0.0%, P = 0.045) and a lower rate of right coronary artery (RCA) (29.6% vs. 40.2%, P = 0.018) and circumflex artery (Cx) occlusion (9.5% vs. 16, 2%, P = 0.045) (Table 1). Table 1 Baseline and echocardiographic characteristics for patients stratified according to composite outcome of HF and/or CD or not Variable All (n = 465) No heart failure or cardiovascular death (n = 266) Heart failure or cardiovascular death (n = 199) P-value Baseline characteristics  Age (years) 66 ± 12 64 ± 12 67 ± 12 0.019  Male gender 74% 74% 74% 0.96  Prior CVD 9.7% 6.4% 14.1% 0.006  Hypertension 43.0% 43.6% 42.2% 0.46  Systolic blood pressure (mmHg) 137 ± 26 138 ± 25 134 ± 28 0.11  Diastolic blood pressure (mmHg) 81 ± 16 82 ± 16 80 ± 16 0.19  Heart rate (bpm) 74 ± 15 71 ± 13 79 ± 17 <0.001  Hypercholesterolaemia 24% 23.7% 25.1% 0.65  Diabetes mellitus 9.7% 7.9% 12.1% 0.13  Current smoker 46.0% 45.1% 47.2% 0.65  Body mass index (kg/m2) 26.5 ± 4.3 26.4 ± 4.2 26.5 ± 4.5 0.75  Family history of CVD 29.9% 34.6% 23.6% 0.011 Acute coronary syndrome information  STEMI 75.7% 74.4% 77.7% 0.46  NSTEMI and/or UAP 24.3% 25.6% 22.6%  Multivessel lesion 6.2% 7.5% 4.5% 0.19  Culprit lesion <0.001   LAD 50.3% 43.6% 59.3%   Cx 13.3% 16.2% 9.5%   RCA 35.7% 40.2% 29.6%   LMS 0.6% 0.0% 1.5%  Type of lesion 0.17   Type A 12.0% 13.9% 9.5%   Type B 39.1% 40.6% 37.2%   Type C 48.8% 45.5% 53.3% Echocardiographic measures  IVSd (cm) 1.1 ± 0.6 1.1 ± 0.8 1.1 ± 0.2 0.46  LVIDd (cm) 5.0 ± 2.2 5.0 ± 2.8 5.0 ± 0.6 0.75  LVPWd (cm) 1.0 ± 0.5 1.0 ± 0.7 1.0 ± 0.2 0.95  LVMI (g/m2) 90.1 [76.6, 106.8] 89.6 [76.1, 100.6] 92.5 [78.4, 111.4] 0.026  LVEF (%) 40.8 ± 11.7 44.9 ± 9.5 35.2 ± 12.1 <0.001  E/A-ratio 1.1 ± 0.41 1.0 ± 0.32 1.1 ± 0.51 0.15  DT (ms) 170.7 ± 46.2 176.3 ± 44.5 162.7 ± 47.5 0.002  e′ (cm/s) 7.5 ± 2.3 7.8 ± 2.3 7.0 ± 2.3 <0.001  E/e′ 9.7 [7.8, 12.3] 9.3 [7.5, 11.4] 10.4 [8.1, 14.1] <0.001  TAPSE (cm) 1.88 ± 0.43 2.0 ± 0.39 1.8 ± 0.46 <0.001  LAVI (mL/m2) 26.0 ± 9.2 26.0 ± 9.0 26.0 ± 9.5 0.98 Absolute layer-specific strain  GLSendo (%) 14.8 ± 4.3 16.6 ± 3.7 12.5 ± 3.8 <0.001  GLS (%) 12.8 ± 3.7 14.4 ± 3.2 10.7 ± 3.3 <0.001  GLSepi (%) 11.1 ± 3.3 12.6 ± 2.8 9.3 ± 2.9 <0.001 Variable All (n = 465) No heart failure or cardiovascular death (n = 266) Heart failure or cardiovascular death (n = 199) P-value Baseline characteristics  Age (years) 66 ± 12 64 ± 12 67 ± 12 0.019  Male gender 74% 74% 74% 0.96  Prior CVD 9.7% 6.4% 14.1% 0.006  Hypertension 43.0% 43.6% 42.2% 0.46  Systolic blood pressure (mmHg) 137 ± 26 138 ± 25 134 ± 28 0.11  Diastolic blood pressure (mmHg) 81 ± 16 82 ± 16 80 ± 16 0.19  Heart rate (bpm) 74 ± 15 71 ± 13 79 ± 17 <0.001  Hypercholesterolaemia 24% 23.7% 25.1% 0.65  Diabetes mellitus 9.7% 7.9% 12.1% 0.13  Current smoker 46.0% 45.1% 47.2% 0.65  Body mass index (kg/m2) 26.5 ± 4.3 26.4 ± 4.2 26.5 ± 4.5 0.75  Family history of CVD 29.9% 34.6% 23.6% 0.011 Acute coronary syndrome information  STEMI 75.7% 74.4% 77.7% 0.46  NSTEMI and/or UAP 24.3% 25.6% 22.6%  Multivessel lesion 6.2% 7.5% 4.5% 0.19  Culprit lesion <0.001   LAD 50.3% 43.6% 59.3%   Cx 13.3% 16.2% 9.5%   RCA 35.7% 40.2% 29.6%   LMS 0.6% 0.0% 1.5%  Type of lesion 0.17   Type A 12.0% 13.9% 9.5%   Type B 39.1% 40.6% 37.2%   Type C 48.8% 45.5% 53.3% Echocardiographic measures  IVSd (cm) 1.1 ± 0.6 1.1 ± 0.8 1.1 ± 0.2 0.46  LVIDd (cm) 5.0 ± 2.2 5.0 ± 2.8 5.0 ± 0.6 0.75  LVPWd (cm) 1.0 ± 0.5 1.0 ± 0.7 1.0 ± 0.2 0.95  LVMI (g/m2) 90.1 [76.6, 106.8] 89.6 [76.1, 100.6] 92.5 [78.4, 111.4] 0.026  LVEF (%) 40.8 ± 11.7 44.9 ± 9.5 35.2 ± 12.1 <0.001  E/A-ratio 1.1 ± 0.41 1.0 ± 0.32 1.1 ± 0.51 0.15  DT (ms) 170.7 ± 46.2 176.3 ± 44.5 162.7 ± 47.5 0.002  e′ (cm/s) 7.5 ± 2.3 7.8 ± 2.3 7.0 ± 2.3 <0.001  E/e′ 9.7 [7.8, 12.3] 9.3 [7.5, 11.4] 10.4 [8.1, 14.1] <0.001  TAPSE (cm) 1.88 ± 0.43 2.0 ± 0.39 1.8 ± 0.46 <0.001  LAVI (mL/m2) 26.0 ± 9.2 26.0 ± 9.0 26.0 ± 9.5 0.98 Absolute layer-specific strain  GLSendo (%) 14.8 ± 4.3 16.6 ± 3.7 12.5 ± 3.8 <0.001  GLS (%) 12.8 ± 3.7 14.4 ± 3.2 10.7 ± 3.3 <0.001  GLSepi (%) 11.1 ± 3.3 12.6 ± 2.8 9.3 ± 2.9 <0.001 Cx, circumflex artery; DT, deceleration time; GLS, whole wall global longitudinal strain; GLSendo, endomyocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; IVSd, interventricular septal thickness; LAD, left anterior descending; LAVI, left atrial volume indexed; LMS, left main stem coronary artery; LVEF, left ventricular ejection fraction; LVIDd, left ventricular internal diameter; LVMI, left ventricular mass indexed; LVPWd, left ventricular posterior wall; NSTEMI, non-ST-segment elevation myocardial infarction; RCA, right coronary artery; STEMI, ST-segment elevation myocardial infarction; TAPSE, tricuspidal annular plane systolic excursion; UAP, unstable angina pectoris. Table 1 Baseline and echocardiographic characteristics for patients stratified according to composite outcome of HF and/or CD or not Variable All (n = 465) No heart failure or cardiovascular death (n = 266) Heart failure or cardiovascular death (n = 199) P-value Baseline characteristics  Age (years) 66 ± 12 64 ± 12 67 ± 12 0.019  Male gender 74% 74% 74% 0.96  Prior CVD 9.7% 6.4% 14.1% 0.006  Hypertension 43.0% 43.6% 42.2% 0.46  Systolic blood pressure (mmHg) 137 ± 26 138 ± 25 134 ± 28 0.11  Diastolic blood pressure (mmHg) 81 ± 16 82 ± 16 80 ± 16 0.19  Heart rate (bpm) 74 ± 15 71 ± 13 79 ± 17 <0.001  Hypercholesterolaemia 24% 23.7% 25.1% 0.65  Diabetes mellitus 9.7% 7.9% 12.1% 0.13  Current smoker 46.0% 45.1% 47.2% 0.65  Body mass index (kg/m2) 26.5 ± 4.3 26.4 ± 4.2 26.5 ± 4.5 0.75  Family history of CVD 29.9% 34.6% 23.6% 0.011 Acute coronary syndrome information  STEMI 75.7% 74.4% 77.7% 0.46  NSTEMI and/or UAP 24.3% 25.6% 22.6%  Multivessel lesion 6.2% 7.5% 4.5% 0.19  Culprit lesion <0.001   LAD 50.3% 43.6% 59.3%   Cx 13.3% 16.2% 9.5%   RCA 35.7% 40.2% 29.6%   LMS 0.6% 0.0% 1.5%  Type of lesion 0.17   Type A 12.0% 13.9% 9.5%   Type B 39.1% 40.6% 37.2%   Type C 48.8% 45.5% 53.3% Echocardiographic measures  IVSd (cm) 1.1 ± 0.6 1.1 ± 0.8 1.1 ± 0.2 0.46  LVIDd (cm) 5.0 ± 2.2 5.0 ± 2.8 5.0 ± 0.6 0.75  LVPWd (cm) 1.0 ± 0.5 1.0 ± 0.7 1.0 ± 0.2 0.95  LVMI (g/m2) 90.1 [76.6, 106.8] 89.6 [76.1, 100.6] 92.5 [78.4, 111.4] 0.026  LVEF (%) 40.8 ± 11.7 44.9 ± 9.5 35.2 ± 12.1 <0.001  E/A-ratio 1.1 ± 0.41 1.0 ± 0.32 1.1 ± 0.51 0.15  DT (ms) 170.7 ± 46.2 176.3 ± 44.5 162.7 ± 47.5 0.002  e′ (cm/s) 7.5 ± 2.3 7.8 ± 2.3 7.0 ± 2.3 <0.001  E/e′ 9.7 [7.8, 12.3] 9.3 [7.5, 11.4] 10.4 [8.1, 14.1] <0.001  TAPSE (cm) 1.88 ± 0.43 2.0 ± 0.39 1.8 ± 0.46 <0.001  LAVI (mL/m2) 26.0 ± 9.2 26.0 ± 9.0 26.0 ± 9.5 0.98 Absolute layer-specific strain  GLSendo (%) 14.8 ± 4.3 16.6 ± 3.7 12.5 ± 3.8 <0.001  GLS (%) 12.8 ± 3.7 14.4 ± 3.2 10.7 ± 3.3 <0.001  GLSepi (%) 11.1 ± 3.3 12.6 ± 2.8 9.3 ± 2.9 <0.001 Variable All (n = 465) No heart failure or cardiovascular death (n = 266) Heart failure or cardiovascular death (n = 199) P-value Baseline characteristics  Age (years) 66 ± 12 64 ± 12 67 ± 12 0.019  Male gender 74% 74% 74% 0.96  Prior CVD 9.7% 6.4% 14.1% 0.006  Hypertension 43.0% 43.6% 42.2% 0.46  Systolic blood pressure (mmHg) 137 ± 26 138 ± 25 134 ± 28 0.11  Diastolic blood pressure (mmHg) 81 ± 16 82 ± 16 80 ± 16 0.19  Heart rate (bpm) 74 ± 15 71 ± 13 79 ± 17 <0.001  Hypercholesterolaemia 24% 23.7% 25.1% 0.65  Diabetes mellitus 9.7% 7.9% 12.1% 0.13  Current smoker 46.0% 45.1% 47.2% 0.65  Body mass index (kg/m2) 26.5 ± 4.3 26.4 ± 4.2 26.5 ± 4.5 0.75  Family history of CVD 29.9% 34.6% 23.6% 0.011 Acute coronary syndrome information  STEMI 75.7% 74.4% 77.7% 0.46  NSTEMI and/or UAP 24.3% 25.6% 22.6%  Multivessel lesion 6.2% 7.5% 4.5% 0.19  Culprit lesion <0.001   LAD 50.3% 43.6% 59.3%   Cx 13.3% 16.2% 9.5%   RCA 35.7% 40.2% 29.6%   LMS 0.6% 0.0% 1.5%  Type of lesion 0.17   Type A 12.0% 13.9% 9.5%   Type B 39.1% 40.6% 37.2%   Type C 48.8% 45.5% 53.3% Echocardiographic measures  IVSd (cm) 1.1 ± 0.6 1.1 ± 0.8 1.1 ± 0.2 0.46  LVIDd (cm) 5.0 ± 2.2 5.0 ± 2.8 5.0 ± 0.6 0.75  LVPWd (cm) 1.0 ± 0.5 1.0 ± 0.7 1.0 ± 0.2 0.95  LVMI (g/m2) 90.1 [76.6, 106.8] 89.6 [76.1, 100.6] 92.5 [78.4, 111.4] 0.026  LVEF (%) 40.8 ± 11.7 44.9 ± 9.5 35.2 ± 12.1 <0.001  E/A-ratio 1.1 ± 0.41 1.0 ± 0.32 1.1 ± 0.51 0.15  DT (ms) 170.7 ± 46.2 176.3 ± 44.5 162.7 ± 47.5 0.002  e′ (cm/s) 7.5 ± 2.3 7.8 ± 2.3 7.0 ± 2.3 <0.001  E/e′ 9.7 [7.8, 12.3] 9.3 [7.5, 11.4] 10.4 [8.1, 14.1] <0.001  TAPSE (cm) 1.88 ± 0.43 2.0 ± 0.39 1.8 ± 0.46 <0.001  LAVI (mL/m2) 26.0 ± 9.2 26.0 ± 9.0 26.0 ± 9.5 0.98 Absolute layer-specific strain  GLSendo (%) 14.8 ± 4.3 16.6 ± 3.7 12.5 ± 3.8 <0.001  GLS (%) 12.8 ± 3.7 14.4 ± 3.2 10.7 ± 3.3 <0.001  GLSepi (%) 11.1 ± 3.3 12.6 ± 2.8 9.3 ± 2.9 <0.001 Cx, circumflex artery; DT, deceleration time; GLS, whole wall global longitudinal strain; GLSendo, endomyocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; IVSd, interventricular septal thickness; LAD, left anterior descending; LAVI, left atrial volume indexed; LMS, left main stem coronary artery; LVEF, left ventricular ejection fraction; LVIDd, left ventricular internal diameter; LVMI, left ventricular mass indexed; LVPWd, left ventricular posterior wall; NSTEMI, non-ST-segment elevation myocardial infarction; RCA, right coronary artery; STEMI, ST-segment elevation myocardial infarction; TAPSE, tricuspidal annular plane systolic excursion; UAP, unstable angina pectoris. Supplementary material online, Tables1 and 2 display baseline characteristics for the cohort stratified according to the individual endpoints separately. The HF/CD group had a significantly larger LVMI (93 g/m2 vs. 90 g/m2, P = 0.026), lower LVEF (35% vs. 45%, P < 0.001), a shorter DT (163 ms vs. 176 ms, P = 0.002), a lower e′ (7.0 cm/s vs. 7.8 cm/s, P < 0.001), a higher E/e′ (10.4 vs. 9.3, P < 0.001), and reduced TAPSE (1.8 cm vs. 2.0 cm, P < 0.001). The longitudinal strain measurements—GLSendo (12% vs. 17%, P < 0.001), GLS (11% vs. 14%, P < 0.001), and GLSepi (9% vs. 13%, P < 0.001) were all significantly lower in the HF/CD group as well (Table 1). Outcome Both GLSendo, GLS, and GLSepi remained independently associated measures after adjustments in Models 1 and 2 with onset of the composite outcome (HF/CD) and each outcome when evaluated separately (Table 2). After adjusting for all the echocardiographic parameters (Model 3) GLSendo, GLS and GLSepi remained independently associated with the composite outcome and HF alone; however, only GLSepi remained independently associated with CD (Table 2). No other echocardiographic measures remained independently associated with the composite outcome (Table 2). Table 2 Multivariable Cox regression and c-statistics (unadjusted) for developing HF and/or CD Composite endpoint (HF and/or CD) (220 events) hazard ratio (95% CI) P-value HF (190 events) hazard ratio (95% CI) P-value CD (46 events) hazard ratio (95% CI) P-value Unadjusted  GLSendo per 1% decrease 1.22 (1.18–1.26) <0.001 1.21 (1.17–1.26) <0.001 1.29 (1.19–1.41) <0.001 C-stat 0.73 C-stat 0.73 C-stat 0.77  GLS per 1% decrease 1.26 (1.21–1.31) <0.001 1.25 (1.20–1.30) <0.001 1.34 (1.22–1.48) <0.001 C-stat 0.74 C-stat 0.73 C-stat 0.77  GLSepi per 1% decrease 1.29 (1.23–1.34) <0.001 1.27 (1.22–1.33) <0.001 1.38 (1.25–1.53) <0.001 C-stat. 0.74 C-stat 0.73 C-stat 0.77 Model 1  GLSendo per 1% decrease 1.20 (1.15–1.25) <0.001 1.20 (1.15–1.26) <0.001 1.17 (1.06–1.29) 0.002  GLS per 1% decrease 1.24 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.20 (1.07–1.34) 0.002  GLSepi per 1% decrease 1.26 (1.20–133) <0.001 1.26 (1.20–1.34) <0.001 1.22 (1.08–1.37) 0.001 Model 2  GLSendo per 1% decrease 1.20 (1.15–1.26) <0.001 1.20 (1.15–1.26) <0.001 1.25 (1.12–1.39) <0.001  GLS per 1% decrease 1.25 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.29 (1.14–1.46) <0.001  GLSepi per 1% decrease 1.26 (1.20–1.35) <0.001 1.27 (1.20–1.35) <0.001 1.32 (1.16–1.51) <0.001 Model 3  GLSendo per 1% decrease 1.19 (1.10–1.28) <0.001 1.18 (1.09–1.28)| <0.001 1.20 (0.97–1.48) 0.10  GLS per 1% decrease 1.24 (1.14–1.35) <0.001 1.23 (1.13–1.35) <0.001 1.26 (0.98–1.63) 0.071  GLSepi per 1% decrease 1.26 (1.15–1.39) <0.001 1.26 (1.14–1.39) <0.001 1.34 (1.00–1.80) 0.048 Composite endpoint (HF and/or CD) (220 events) hazard ratio (95% CI) P-value HF (190 events) hazard ratio (95% CI) P-value CD (46 events) hazard ratio (95% CI) P-value Unadjusted  GLSendo per 1% decrease 1.22 (1.18–1.26) <0.001 1.21 (1.17–1.26) <0.001 1.29 (1.19–1.41) <0.001 C-stat 0.73 C-stat 0.73 C-stat 0.77  GLS per 1% decrease 1.26 (1.21–1.31) <0.001 1.25 (1.20–1.30) <0.001 1.34 (1.22–1.48) <0.001 C-stat 0.74 C-stat 0.73 C-stat 0.77  GLSepi per 1% decrease 1.29 (1.23–1.34) <0.001 1.27 (1.22–1.33) <0.001 1.38 (1.25–1.53) <0.001 C-stat. 0.74 C-stat 0.73 C-stat 0.77 Model 1  GLSendo per 1% decrease 1.20 (1.15–1.25) <0.001 1.20 (1.15–1.26) <0.001 1.17 (1.06–1.29) 0.002  GLS per 1% decrease 1.24 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.20 (1.07–1.34) 0.002  GLSepi per 1% decrease 1.26 (1.20–133) <0.001 1.26 (1.20–1.34) <0.001 1.22 (1.08–1.37) 0.001 Model 2  GLSendo per 1% decrease 1.20 (1.15–1.26) <0.001 1.20 (1.15–1.26) <0.001 1.25 (1.12–1.39) <0.001  GLS per 1% decrease 1.25 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.29 (1.14–1.46) <0.001  GLSepi per 1% decrease 1.26 (1.20–1.35) <0.001 1.27 (1.20–1.35) <0.001 1.32 (1.16–1.51) <0.001 Model 3  GLSendo per 1% decrease 1.19 (1.10–1.28) <0.001 1.18 (1.09–1.28)| <0.001 1.20 (0.97–1.48) 0.10  GLS per 1% decrease 1.24 (1.14–1.35) <0.001 1.23 (1.13–1.35) <0.001 1.26 (0.98–1.63) 0.071  GLSepi per 1% decrease 1.26 (1.15–1.39) <0.001 1.26 (1.14–1.39) <0.001 1.34 (1.00–1.80) 0.048 Model 1 is adjusted for age, gender BMI, prior CVD, family history of CVD, diabetes mellitus, current smoker, hypercholesterolaemia, systolic blood pressure, diastolic blood pressure, HR. Model 2 is adjusted for previous mentioned variables in Model 1 as well as STEMI, multivessel culprit lesions, LAD occlusion, and LMS occlusion. Model 3 is adjusted for all previous mentioned variables in Models 1 and 2, additionally LVMI, LVEF, DT, e′, and E/e′. Table 2 Multivariable Cox regression and c-statistics (unadjusted) for developing HF and/or CD Composite endpoint (HF and/or CD) (220 events) hazard ratio (95% CI) P-value HF (190 events) hazard ratio (95% CI) P-value CD (46 events) hazard ratio (95% CI) P-value Unadjusted  GLSendo per 1% decrease 1.22 (1.18–1.26) <0.001 1.21 (1.17–1.26) <0.001 1.29 (1.19–1.41) <0.001 C-stat 0.73 C-stat 0.73 C-stat 0.77  GLS per 1% decrease 1.26 (1.21–1.31) <0.001 1.25 (1.20–1.30) <0.001 1.34 (1.22–1.48) <0.001 C-stat 0.74 C-stat 0.73 C-stat 0.77  GLSepi per 1% decrease 1.29 (1.23–1.34) <0.001 1.27 (1.22–1.33) <0.001 1.38 (1.25–1.53) <0.001 C-stat. 0.74 C-stat 0.73 C-stat 0.77 Model 1  GLSendo per 1% decrease 1.20 (1.15–1.25) <0.001 1.20 (1.15–1.26) <0.001 1.17 (1.06–1.29) 0.002  GLS per 1% decrease 1.24 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.20 (1.07–1.34) 0.002  GLSepi per 1% decrease 1.26 (1.20–133) <0.001 1.26 (1.20–1.34) <0.001 1.22 (1.08–1.37) 0.001 Model 2  GLSendo per 1% decrease 1.20 (1.15–1.26) <0.001 1.20 (1.15–1.26) <0.001 1.25 (1.12–1.39) <0.001  GLS per 1% decrease 1.25 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.29 (1.14–1.46) <0.001  GLSepi per 1% decrease 1.26 (1.20–1.35) <0.001 1.27 (1.20–1.35) <0.001 1.32 (1.16–1.51) <0.001 Model 3  GLSendo per 1% decrease 1.19 (1.10–1.28) <0.001 1.18 (1.09–1.28)| <0.001 1.20 (0.97–1.48) 0.10  GLS per 1% decrease 1.24 (1.14–1.35) <0.001 1.23 (1.13–1.35) <0.001 1.26 (0.98–1.63) 0.071  GLSepi per 1% decrease 1.26 (1.15–1.39) <0.001 1.26 (1.14–1.39) <0.001 1.34 (1.00–1.80) 0.048 Composite endpoint (HF and/or CD) (220 events) hazard ratio (95% CI) P-value HF (190 events) hazard ratio (95% CI) P-value CD (46 events) hazard ratio (95% CI) P-value Unadjusted  GLSendo per 1% decrease 1.22 (1.18–1.26) <0.001 1.21 (1.17–1.26) <0.001 1.29 (1.19–1.41) <0.001 C-stat 0.73 C-stat 0.73 C-stat 0.77  GLS per 1% decrease 1.26 (1.21–1.31) <0.001 1.25 (1.20–1.30) <0.001 1.34 (1.22–1.48) <0.001 C-stat 0.74 C-stat 0.73 C-stat 0.77  GLSepi per 1% decrease 1.29 (1.23–1.34) <0.001 1.27 (1.22–1.33) <0.001 1.38 (1.25–1.53) <0.001 C-stat. 0.74 C-stat 0.73 C-stat 0.77 Model 1  GLSendo per 1% decrease 1.20 (1.15–1.25) <0.001 1.20 (1.15–1.26) <0.001 1.17 (1.06–1.29) 0.002  GLS per 1% decrease 1.24 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.20 (1.07–1.34) 0.002  GLSepi per 1% decrease 1.26 (1.20–133) <0.001 1.26 (1.20–1.34) <0.001 1.22 (1.08–1.37) 0.001 Model 2  GLSendo per 1% decrease 1.20 (1.15–1.26) <0.001 1.20 (1.15–1.26) <0.001 1.25 (1.12–1.39) <0.001  GLS per 1% decrease 1.25 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.29 (1.14–1.46) <0.001  GLSepi per 1% decrease 1.26 (1.20–1.35) <0.001 1.27 (1.20–1.35) <0.001 1.32 (1.16–1.51) <0.001 Model 3  GLSendo per 1% decrease 1.19 (1.10–1.28) <0.001 1.18 (1.09–1.28)| <0.001 1.20 (0.97–1.48) 0.10  GLS per 1% decrease 1.24 (1.14–1.35) <0.001 1.23 (1.13–1.35) <0.001 1.26 (0.98–1.63) 0.071  GLSepi per 1% decrease 1.26 (1.15–1.39) <0.001 1.26 (1.14–1.39) <0.001 1.34 (1.00–1.80) 0.048 Model 1 is adjusted for age, gender BMI, prior CVD, family history of CVD, diabetes mellitus, current smoker, hypercholesterolaemia, systolic blood pressure, diastolic blood pressure, HR. Model 2 is adjusted for previous mentioned variables in Model 1 as well as STEMI, multivessel culprit lesions, LAD occlusion, and LMS occlusion. Model 3 is adjusted for all previous mentioned variables in Models 1 and 2, additionally LVMI, LVEF, DT, e′, and E/e′. The Kaplan–Meier curves (Figure 2) displaying the patient cohort stratified into tertiles of GLSendo, GLS, and GLSepi, respectively, demonstrates the incrementally increased risk of developing HF and/or CD with lower tertile of GLS. The lowest tertile displayed a eight- to nine-fold increased risk of developing HF and/or CD compared to the highest tertile [GLSendo: 1 tertile vs. 3 tertile: hazard ratio 8.0 (5.1–12.5), P < 0.001; GLS: 1 tertile vs. 3 tertile: hazard ratio 9.2 (5.8–14.7), P < 0.001; GLSepi: 1 tertile vs. 3 tertile: hazard ratio 8.5 (5.4–13.2), P < 0.001]. The intermediate tertile had a three- to four-fold risk compared to the highest tertile [GLSendo: hazard ratio 3.2 (2.0–5.2), P < 0.001; GLS: hazard ratio 4.0 (2.5–6.6), P < 0.001; GLSepi: hazard ratio 3.5 (2.2–5.7), P < 0.001]. Figure 2 View largeDownload slide Layer-specific strain and outcome. The Kaplan–Meier curves displaying the probability of staying event free. The horizontal axis displays the time from the ACS expressed in days. The vertical axis represents the cumulative probability of staying event free of HF and/or CD. The study population is stratified into three groups based on the tertiles of GLSendo, GLS, or GLSepi. 95% CI and hazard ratios included. Number of patient at risk at every 1000 days is displayed below each curve. CD, cardiovascular death; CI, confidence interval; GLS, whole wall global longitudinal strain; GLSendo, endocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; HF, heart failure. Figure 2 View largeDownload slide Layer-specific strain and outcome. The Kaplan–Meier curves displaying the probability of staying event free. The horizontal axis displays the time from the ACS expressed in days. The vertical axis represents the cumulative probability of staying event free of HF and/or CD. The study population is stratified into three groups based on the tertiles of GLSendo, GLS, or GLSepi. 95% CI and hazard ratios included. Number of patient at risk at every 1000 days is displayed below each curve. CD, cardiovascular death; CI, confidence interval; GLS, whole wall global longitudinal strain; GLSendo, endocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; HF, heart failure. Incremental prognostic yield of GLS obtained from the different layers and whole wall When comparing the Harrell’s c-statistics obtained from unadjusted Cox regressions for GLSendo with the c-statistics obtained for GLS (0.733 vs. 0.739, P = 0.017) and c-statistics obtained for GLS compared to the c-statistics obtained for GLSepi (0.739 vs. 0.741, P = 0.38), we found the prognostic power of GLS to be significantly higher than that of GLSendo, and the c-statistics for GLS was lower than for GLSepi; however, this difference was not statistically significant. Adding either GLS or GLSepi to all the variables included in Model 3 resulted in a significant increase in the c-statistics than Model 3 alone [Model 3 with GLS: 0.76 (0.72–0.80) vs. 0.74 (0.70–0.77), P = 0.048; Model 3 with GLSepi: 0.76 (0.72–0.80) vs. 0.74 (0.70–0.77), P = 0.039]. Adding GLSendo to the variables included in Model 3, only displayed a non-significant increase in the c-statistics when compared to Model 3 alone [Model 3 with GLSendo: 0.75 (0.72–0.79) vs. 0.74 (0.70–0.77), P = 0.075]. Models 1 and 2 significantly improved when adding GLSendo, GLS, or GLSepi (Figure 3). Figure 3 View largeDownload slide Incremental prognostic information when adding layer-specific strain. (A) Column diagram displaying Harrell’s C-statistics comparison in various models for predicting HF and/or CD with or without GLSendo GLS or GLSepi. Model 1 includes age, gender BMI, prior CVD, family history of CVD, DM, current smoker, hypercholesterolaemia, systolic blood pressure, diastolic blood pressure, and heart rate. Model 2 includes the same variables as Model 1 along with the PCI variables—STEMI, multivessel culprit lesions, LAD occlusion and LMS occlusion. Model 3 includes the same variable as Models 1 and 2 in addition to LVMI, LVEF, DT, e′, and E/e′. 95% CI included. BMI, body mass index; CD, cardiovascular death; CI, confidence interval; CVD, cardiovascular disease; DM, diabetes mellitus; DT, deceleration time; e′, early peak myocardial diastolic velocity; E/e′, peak early filling velocity divided by early peak myocardial diastolic velocity; GLS, whole wall global longitudinal strain; GLSendo, endocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; HF, heart failure; LAD, left anterior descending, LMS, left main stem; LVEF, left ventricular ejection fraction; LVMI, left ventricle mass index; PCI, percutaneous coronary intervention; STEMI, ST-segment elevation myocardial infarction. Figure 3 View largeDownload slide Incremental prognostic information when adding layer-specific strain. (A) Column diagram displaying Harrell’s C-statistics comparison in various models for predicting HF and/or CD with or without GLSendo GLS or GLSepi. Model 1 includes age, gender BMI, prior CVD, family history of CVD, DM, current smoker, hypercholesterolaemia, systolic blood pressure, diastolic blood pressure, and heart rate. Model 2 includes the same variables as Model 1 along with the PCI variables—STEMI, multivessel culprit lesions, LAD occlusion and LMS occlusion. Model 3 includes the same variable as Models 1 and 2 in addition to LVMI, LVEF, DT, e′, and E/e′. 95% CI included. BMI, body mass index; CD, cardiovascular death; CI, confidence interval; CVD, cardiovascular disease; DM, diabetes mellitus; DT, deceleration time; e′, early peak myocardial diastolic velocity; E/e′, peak early filling velocity divided by early peak myocardial diastolic velocity; GLS, whole wall global longitudinal strain; GLSendo, endocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; HF, heart failure; LAD, left anterior descending, LMS, left main stem; LVEF, left ventricular ejection fraction; LVMI, left ventricle mass index; PCI, percutaneous coronary intervention; STEMI, ST-segment elevation myocardial infarction. Finally, we found no significant changes in Harrell’s c-statistics when adding all the conventional echocardiographic parameters to the models when either one of the GLS measures were already included (Figure 3). Hence, neither of the models significantly improved when adding all other echocardiographic measures to the models where the GLS measures were already included. Discussion This is the first study evaluating the association between layer-specific GLS and HF and/or CD following ACS. In the present study, we found that all three layer-specific GLS measurements provided superior prognostic information on the risk of developing HF and/or CD compared to all other echocardiographic measure. Both GLS obtained from the epicardial layer and the whole wall provided incremental prognostic information when added to all other predictors in the current population. GLSepi was, however, the only independent echocardiographic predictor of CD alone. In addition, the risk assessment did not significantly improve when adding all other significantly associated echocardiographic parameters to the models when either GLSendo, GLS, or GLSepi were already included (Figure 3). In our study, patients developing HF and/or CD were older, had a higher tendency of prior CVD, and increased HR, which is consistent with the literature.3 Cardiac imaging especially transthoracic echocardiography plays a central role in the assessment of LV function following MI18 and in diagnosis of HF.3 One of the strongest predictors of survival in ACS patients is LV function evaluated with an echocardiogram before hospital discharge.18 Echocardiography is the most common imaging method to evaluate regional and global LV function after STEMI. In the presence of severe LV systolic dysfunction, it is recommended to re-evaluate with a subsequent echocardiogram around 6–12 weeks after initial hospitalization.18 Currently only LVEF and not GLS is recommended when assessing systolic function.17 This is in spite of the fact that LVEF can be preserved in the presence of HF. This is hypothesized to be due to an increase in the circumferential and radial myocardial deformation occurring concomitant with an impaired longitudinal deformation which can be identified using speckle tracking imaging.4,6 Hence, LVEF remains preserved despite of a reduced longitudinal deformation. Layer-specific GLS has been suggested as a potential method of assessing LV function.19,20 Several studies have demonstrated layer-specific GLS to be independently associated with the presence of coronary artery disease. Yet these studies display a discrepancy in which specific layer they found to have the strongest diagnostic power.21–24 GLSendo and circumferential endocardial strain have also been displayed to be associated with subsequent cardiac events in patients with chronic ischaemic cardiomyopathy.25 However, there are no published studies evaluating the prognostic value of layer-specific GLS for HF and/or CD following ACS. In the present report, the magnitude of layer-specific GLS decreased from the endocardium to the epicardium (Tables 1 and 2), which is in concordance with previous studies.19,21–23 We found all layers of GLS to be associated with HF and/or CD, after adjusting for clinical and echocardiographic parameters. Yet, GLSepi and GLS displayed stronger prognostic power (Table 2, Figure 3). In our adjusted models, the hazard ratios of layer-specific GLS increased in size from the endocardium to the epicardium per 1% decrease (Table 2). GLSepi remained as the only independent predictor when predicting only CD (Table 2). Our group has recently demonstrated similar results when using layer-specific GLS to diagnose ischaemia in a cohort of 80 patients with stable angina pectoris. In this study, we found that GLSepi was the only independent echocardiographic measure associated with the presence of significance of coronary artery stenosis after multivariable adjustment.23,24 These findings, like the results of the present report, which demonstrates that GLSepi might be the best measure both to diagnose significant coronary artery stenosis, but more importantly, subsequent outcome, are however still in contrary to other studies using layer-specific GLS. Zhang et al.21 (cohort of 139 patients with 79 having complex coronary artery disease) and Sarvari et al.22 (cohort of 77 patients with 49 having significant coronary artery disease) found GLSendo to be the best measurement in identifying complex coronary artery disease and predicting the severity of coronary lesions in NSTEMI patients. We have tested the intra- and inter-observer variability of layered 2D-speckle tracking echocardiography in 20 patients with ischaemic heart disease. Intra-observer Bland–Altman analysis displayed a bias of 0.473 ± 1.45 with coefficients of variation (CV) of 6.6% for GLSendo, 0.447 ± 1.21 with CV of 6.4% for GLS, and 0.420 ± 1.07 with CV of 6.5% for GLSepi. Whereas, the inter-observer analysis displayed a bias of −0.965 ± 2.36 with CV of 11.1% for GLSendo, −0.887 ± 1.92 with a CV of 10.5% for GLS, and −0.815 ± 1.64 with a CV of 10.3% for GLSepi.23,24 The better reproducibility of 2D-speckle tracking of the epimyocardium than that of the endomyocardium might be one of the reasons several new studies found GLSepi to be the best measure for diagnostic and prognostic purposes. This might be due to the better traceability of the epimyocardium as opposed to that of the endocardial border. The endomyocardial region might generally be affected to a higher degree than the epimyocardial region from several factors. Such factors could be the more varied contour of the endocardial border, which can make it more difficult to line up the ROI. Apical foreshortening might also influence the traceability of apical endocardial border, whereas the epicardial border would be less evasive. Apical foreshortening has been shown to affect longitudinal strain values significantly.26 Additionally, regional epimyocardial longitudinal strain has been shown to be more homogenous than the inner two layers, which have significant variance from base to apex.19 As mentioned previously it is currently not recommended to evaluate GLS after ACS or when diagnosing HF.3,17 Applying speckle tracking echocardiography to the standard echocardiographic work-up following ACS would require little additional time and effort. Furthermore, due to the semi-automated nature of the method, the intra- and inter-observer variability is low. The superiority of layer-specific longitudinal strain as well as whole wall longitudinal strain when compared to LVEF indicates that it might be time to reassess the recommendations on LV function measurements. We demonstrated both layer-specific GLSendo and GLSepi and GLS to provide superior prognostic information regarding the risk of developing HF and/or CD compared to all other echocardiographic measures. We find that it might be clinically relevant to add GLSepi, GLSendo, and GLS to the risk stratification scheme of ACS patients. This could assist in identifying patients at greatest risk of adverse morbidity and mortality. Limitations There may be residential confounders as this is a retrospectively planned study. However, the analysis was based on consecutive patients from a well-defined registry. Retrieving endpoints—development of HF and CD—from the Danish National Board of Health’s National Patient Registry and the Danish Register of Causes of Death using International Classifications of Diseases codes may be questioned. However, diagnostic codes obtained through the Danish National Patient Registry have proven highly accurate. In 2011, Thygesen et al. investigated the positive predictive value of International Classification of Diseases-10 diagnostic codes. The authors found diagnostic codes of HF obtained through the Danish National Patient Registry to have a positive predictive value of 100% (92.9–100%), when reviewing discharge summaries and medical records.27,28 Unfortunately, the degree of mitral regurgitation was not reassessed in the present study. Information about the medication of the patients such as the use of betablockers was not available either. Data on medication as well as mitral regurgitation would have resulted in a more thorough description of the cohort. Furthermore, we regret we were not able to examine changes in medication following discharge from the hospital. There might be some concerns regarding overfitting, when utilizing a multivariable Cox regression with as many variables included as in Model 3. However, interpretation of an overfitted multivariable Cox regression can still be acceptable, when it is done to control for potential confounders and not for building a prediction model.29 One of the challenges of speckle tracking is defining the epicardial border, which in some cases can appear unclear and hence may bias GLSepi. However, we have found that intra- and inter-observer reproducibility is superior for GLSepi when compared to GLSendo and whole wall GLS, hence, this issue does not seem to make GLSepi a less valid measure. Regretfully, follow-up echocardiograms were not performed for the present study. Lastly, we cannot exclude the possibility of patients having an undiagnosed pre-existing condition prior to their initial ACS hospitalization influencing our results. Conclusion Layer-specific GLSendo, GLS, and GLSepi were all independently associated with HF and/or CD. GLS and GLSepi provided incremental prognostic information when added to all other significantly associated measures. Supplementary data Supplementary data are available at European Heart Journal - Cardiovascular Imaging online. Conflict of interest: None declared. References 1 Wang TJ , Evans JC , Benjamin EJ , Levy D , LeRoy EC , Vasan RS. Natural history of asymptomatic left ventricular systolic dysfunction in the community . Circulation 2003 ; 108 : 977 – 82 . Google Scholar CrossRef Search ADS PubMed 2 Effect of enalapril on mortality and the development of heart failure in asymptomatic patients with reduced left ventricular ejection fractions. The SOLVD Investigattors . N Engl J Med 1992 ; 327 : 685 – 91 . CrossRef Search ADS PubMed 3 Ponikowski P , Voors AA , Anker SD , Bueno H , Cleland JGF , Coats AJS et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: the Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC . Eur J Heart Fail 2016 ; 18 : 891 – 975 . Google Scholar CrossRef Search ADS PubMed 4 Biering-Sørensen T , Santos M , Rivero J , McCullough SD , West E , Opotowsky AR et al. Left ventricular deformation at rest predicts exercise-induced elevation in pulmonary artery wedge pressure in patients with unexplained dyspnoea . Eur J Heart Fail 2016 ; 19 : 101 – 110 . Google Scholar CrossRef Search ADS PubMed 5 Biering-Sørensen T , Biering-Sørensen SR , Olsen FJ , Sengeløv M , Jørgensen PG , Mogelvang R et al. Global longitudinal strain by echocardiography predicts long-term risk of cardiovascular morbidity and mortality in a low-risk general population: the Copenhagen City Heart Study . Circ Cardiovasc Imaging 2017 ; 10 : e005521 . Google Scholar CrossRef Search ADS PubMed 6 Biering-Sørensen T , Solomon SD. Assessing contractile function when ejection fraction is normal: a case for strain . Circ Cardiovasc Imaging 2015 ; 8 : e004181. Google Scholar CrossRef Search ADS PubMed 7 Kalam K , Otahal P , Marwick TH. Prognostic implications of global LV dysfunction: a systematic review and meta-analysis of global longitudinal strain and ejection fraction . Heart Br Card Soc 2014 ; 100 : 1673 – 80 . 8 Sengeløv M , Jørgensen PG , Jensen JS , Bruun NE , Olsen FJ , Fritz-Hansen T et al. Global longitudinal strain is a superior predictor of all-cause mortality in heart failure with reduced ejection fraction . JACC Cardiovasc Imaging 2015 ; 8 : 1351 – 9 . Google Scholar CrossRef Search ADS PubMed 9 Biering-Sørensen T , Knappe D , Pouleur A-C , Claggett B , Wang PJ , Moss AJ et al. Regional longitudinal deformation improves prediction of ventricular tachyarrhythmias in patients with heart failure with reduced ejection fraction: a MADIT-CRT Substudy (Multicenter Automatic Defibrillator Implantation Trial-Cardiac Resynchronization Therapy) . Circ Cardiovasc Imaging 2017 ; 10 : e005096 . Google Scholar CrossRef Search ADS PubMed 10 Olsen FJ , Pedersen S , Jensen JS , Biering-Sørensen T. Global longitudinal strain predicts incident atrial fibrillation and stroke occurrence after acute myocardial infarction . Medicine (Baltimore) 2016 ; 95 : e5338. Google Scholar CrossRef Search ADS PubMed 11 Shah AM , Claggett B , Sweitzer NK , Shah SJ , Anand IS , Liu L et al. Prognostic importance of impaired systolic function in heart failure with preserved ejection fraction and the impact of spironolactone . Circulation 2015 ; 132 : 402 – 14 . Google Scholar CrossRef Search ADS PubMed 12 Mor-Avi V , Lang RM , Badano LP , Belohlavek M , Cardim NM , Derumeaux G et al. Current and evolving echocardiographic techniques for the quantitative evaluation of cardiac mechanics: ASE/EAE consensus statement on methodology and indications endorsed by the Japanese Society of Echocardiography . Eur J Echocardiogr J Work Group Echocardiogr Eur Soc Cardiol 2011 ; 12 : 167 – 205 . Google Scholar CrossRef Search ADS 13 Joyce E. LVEF: long-standing monarch of systolic dysfunction, buckling under the strain? Eur J Heart Fail 2014 ; 16 : 1270 – 2 . Google Scholar CrossRef Search ADS PubMed 14 Russo C , Jin Z , Elkind MSV , Rundek T , Homma S , Sacco RL et al. Prevalence and prognostic value of subclinical left ventricular systolic dysfunction by global longitudinal strain in a community-based cohort . Eur J Heart Fail 2014 ; 16 : 1301 – 9 . Google Scholar CrossRef Search ADS PubMed 15 Vendelin M , Bovendeerd PHM , Engelbrecht J , Arts T. Optimizing ventricular fibers: uniform strain or stress, but not ATP consumption, leads to high efficiency . Am J Physiol Heart Circ Physiol 2002 ; 283 : H1072 – 81 . Google Scholar CrossRef Search ADS PubMed 16 Reimer KA , Lowe JE , Rasmussen MM , Jennings RB. The wavefront phenomenon of ischemic cell death. 1. Myocardial infarct size vs duration of coronary occlusion in dogs . Circulation 1977 ; 56 : 786 – 94 . Google Scholar CrossRef Search ADS PubMed 17 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 . Eur Heart J Cardiovasc Imaging 2015 ; 16 : 233 – 70 . Google Scholar CrossRef Search ADS PubMed 18 Ibanez B , James S , Agewall S , Antunes MJ , Bucciarelli-Ducci C , Bueno H et al. 2017 ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation: the Task Force for the management of acute myocardial infarction in patients presenting with ST-segment elevation of the European Society of Cardiology (ESC) . Eur Heart J 2017 ; 39 : 119 – 177 . Google Scholar CrossRef Search ADS 19 Leitman M , Lysiansky M , Lysyansky P , Friedman Z , Tyomkin V , Fuchs T et al. Circumferential and longitudinal strain in 3 myocardial layers in normal subjects and in patients with regional left ventricular dysfunction . J Am Soc Echocardiogr 2010 ; 23 : 64 – 70 . Google Scholar CrossRef Search ADS PubMed 20 Shi J , Pan C , Kong D , Cheng L , Shu X. Left ventricular longitudinal and circumferential layer-specific myocardial strains and their determinants in healthy subjects . Echocardiography (Mt Kisco NY) 2016 ; 33 : 510 – 8 . Google Scholar CrossRef Search ADS 21 Zhang L , Wu W-C , Ma H , Wang H. Usefulness of layer-specific strain for identifying complex CAD and predicting the severity of coronary lesions in patients with non-ST-segment elevation acute coronary syndrome: compared with Syntax score . Int J Cardiol 2016 ; 223 : 1045 – 52 . Google Scholar CrossRef Search ADS PubMed 22 Sarvari SI , Haugaa KH , Zahid W , Bendz B , Aakhus S , Aaberge L et al. Layer-specific quantification of myocardial deformation by strain echocardiography may reveal significant CAD in patients with non-ST-segment elevation acute coronary syndrome . JACC Cardiovasc Imaging 2013 ; 6 : 535 – 44 . Google Scholar CrossRef Search ADS PubMed 23 Hagemann C , Hoffmann S , Olsen FJ , Jørgensen P , Jensen J , Biering-Sorensen T. Layer-specific global longitudinal strain reveals impaired cardiac function in patients with reversible ischemia [abstract] . J Am Coll Cardiol 2016 ; 67 : 1592. Google Scholar CrossRef Search ADS 24 Hagemann C , Hoffmann S , Andersen R , Hansen T , Olsen FJ , Jørgensen P et al. Layer-specific strain analysis by two-dimensional speckle tracking echocardiography: improvement of coronary artery disease diagnostics in patients with stable angina pectoris [abstract] . J Am Coll Cardiol 2017 ; 69 : 1634. Google Scholar CrossRef Search ADS 25 Hamada S , Schroeder J , Hoffmann R , Altiok E , Keszei A , Almalla M et al. Prediction of outcomes in patients with chronic ischemic cardiomyopathy by layer-specific strain echocardiography: a proof of concept . J Am Soc Echocardiogr 2016 ; 29 : 412 – 20 . Google Scholar CrossRef Search ADS PubMed 26 Tarr A , Stöbe S , Trache T , Kluge J-G , Varga A , Pfeiffer D et al. The impact of foreshortening on regional strain–a comparison of regional strain evaluation between speckle tracking and tissue velocity imaging . Ultraschall Med 2013 ; 34 : 446 – 53 . Google Scholar CrossRef Search ADS PubMed 27 Thygesen SK , Christiansen CF , Christensen S , Lash TL , Sørensen HT. The predictive value of ICD-10 diagnostic coding used to assess Charlson comorbidity index conditions in the population-based Danish National Registry of Patients . BMC Med Res Methodol 2011 ; 11 : 83. Google Scholar CrossRef Search ADS PubMed 28 Christiansen MN , Køber L , Weeke P , Vasan RS , Jeppesen JL , Smith JG et al. Age-specific trends in incidence, mortality, and comorbidities of heart failure in Denmark, 1995 to 2012 . Circulation 2017 ; 135 : 1214 – 23 . Google Scholar CrossRef Search ADS PubMed 29 Vittinghoff E , McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression . Am J Epidemiol 2007 ; 165 : 710 – 8 . Google Scholar CrossRef Search ADS PubMed Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Heart Journal – Cardiovascular Imaging Oxford University Press

Association between layer-specific global longitudinal strain and adverse outcomes following acute coronary syndrome

Loading next page...
 
/lp/ou_press/association-between-layer-specific-global-longitudinal-strain-and-fZUy6Zr8oA
Publisher
Oxford University Press
Copyright
Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com.
ISSN
2047-2404
D.O.I.
10.1093/ehjci/jey004
Publisher site
See Article on Publisher Site

Abstract

Abstract Aims To investigate the prognostic value of layer-specific global longitudinal strain (GLS) in predicting heart failure (HF) and cardiovascular death (CD) following acute coronary syndrome (ACS). Methods and results In this retrospective study, 465 ACS patients underwent transthoracic echocardiography following percutaneous coronary intervention (PCI). The primary endpoint was the composite of HF and/or CD with a median follow-up time of 4.6 (0.2–6.3) years. During follow-up 199 patients (42.7%) suffered HF and/or CD (176 developed HF and 38 suffered CD). Absolute endomyocardial global longitudinal strain (GLSendo) (12% vs. 17%, P < 0.001), GLS (11% vs. 14%, P < 0.001), and epimyocardial global longitudinal strain (GLSepi) (9% vs. 13%, P < 0.001) were all reduced in patients with an adverse outcome. In multivariable Cox regressions, which included clinical baseline characteristics and conventional echocardiographic measurements, GLS obtained from all layers remained independently associated with the composite outcome; GLSendo [hazard ratio: 1.19 (1.10–1.28), P < 0.001, per 1% decrease], GLS [hazard ratio 1.24 (1.14–1.35), P < 0.001, per 1% decrease], and GLSepi [hazard ratio 1.26 (1.15–1.39), P < 0.001, per 1% decrease]. No other echocardiographic measures remained independently associated with the composite outcome in these models. Finally, GLS and GLSepi provided incremental prognostic information on the risk of developing the composite endpoint, when added to all other clinical and echocardiographic measures [adding GLS (c-statistics: 0.76 vs. 0.74, P = 0.048) or adding GLSepi (c-statistics: 0.76 vs. 0.74, P = 0.039)]. Conclusion In ACS patients, layer-specific strain provides independent prognostic information regarding risk of developing HF and/or CD. Furthermore, only GLS and GLSepi provided incremental prognostic information when added to all other significant predictors. acute coronary syndrome , 2D-speckle tracking echocardiography , heart failure , cardiovascular death , layer-specific 2D-speckle tracking Introduction A common and adverse complication following a myocardial infarction (MI) is heart failure (HF). To prevent HF, and improve the prognosis it is necessary to identify high risk patients to initiate more intensive monitoring or treatment. Prior to the clinical symptoms of HF become clear, patients can develop asymptomatic systolic or diastolic left ventricular (LV) dysfunction caused by structural or functional cardiac abnormalities which are precursors of HF.1–6 Recent studies have shown global longitudinal strain (GLS) obtained from 2D-speckle tracking echocardiography to be a strong predictor of adverse outcomes in several cardiac diseases.5,7–11 GLS has already been suggested as a more sensitive marker of cardiac function than conventional echocardiography.4,4–6,12,13 It has been demonstrated to be a more powerful predictor of outcome than left ventricular ejection fraction (LVEF) as it may reflect subclinical LV systolic dysfunction.6,14 Novel echocardiographic software can now sectionalize the myocardial wall in layers. This has made it possible to detect which layers of the myocardial wall that suffer from reduced deformation. Layer-specific GLS might be important since the LV wall is heterogeneous and consists of three different layers of muscle fibres.15 Especially in ischaemic heart disease, layer-specific strain may be particularly useful since longitudinally orientated myocardial fibres located in the sub-endocardium is known to be most susceptible to ischaemia.16 There are no studies that have investigated the prognostic value of layer-specific GLS in an acute coronary syndrome (ACS) population. This is what we aim to examine, especially the association between layer-specific GLS and the risk of developing HF and cardiovascular death (CD). Methods Study population The Cardiology Department of Copenhagen University Hospital Gentofte is an invasive hub for 10 non-invasive cardiology departments. In the period January 2003 to November 2008, 5003 patients were admitted to the department for PCI, these patients were all included in a clinical registry. All patients in this registry were eligible for our study if they had an echocardiogram available. We hereby identified 580 ACS patients [ST-segment elevation myocardial infarction (STEMI), non-ST-segment elevation myocardial infarction (NSTEMI), or unstable angina pectoris (UAP)] who were admitted to Copenhagen University Hospital Gentofte, Denmark for percutaneous coronary intervention (PCI) in the period and had an echocardiogram performed at Gentofte Hospital. A considerable number of the 5003 patients were admitted to other non-invasive cardiology departments and were rapidly transported back to their local hospital post-PCI. The echocardiographic examinations were performed [median 2 days (1–3 days)] after the PCI procedure. In addition, 115 patients were excluded due to a non-sinus rhythm, missing images, or inadequate image quality. Outcomes and endpoints Information regarding endpoints—development of HF and CD—were retrieved from the Danish National Board of Health’s National Patient Registry and the Danish Register of Causes of Death using International Classifications of Diseases codes. Echocardiography The echocardiographic examinations were performed by experienced clinicians and sonographers on GE Vivid machines (GE Healthcare, Little Chalfont, UK). The images were stored in a GE Health care image vault and analysed offline with EchoPac version 113 (GE Healthcare, Horten, Norway) by an experienced investigator blinded to clinical baseline data and endpoints. Conventional 2D echocardiography The LV dimensions, interventricular septal thickness (IVSd), LV internal diameter (LVIDd), and LV posterior wall (LVPWd) thickness were measured in the parasternal long-axis view at the tip of the mitral valve leaflet in end-diastole. LV mass index (LVMI) was calculated as the anatomical mass17 divided by body surface area. Mitral valve inflow patterns: peak early filling (E-wave), atrial filling (A-wave) and deceleration time of the early filling (DT) were measured with pulsed-wave Doppler imaging at the tips of the mitral valve leaflets in the apical 4-chamber view. Additionally, early (e′) peak myocardial diastolic velocity was measured with pulsed-wave tissue Doppler imaging in the apical 4-chamber view at the lateral and septal walls at the mitral annulus. The e′ was indexed to the E-wave velocity to obtain the E/e′. LVEF was measured with the modified Simpson’s biplane method obtained from the apical 4- and 2-chamber views. Left atrial volume (LAV) was obtained through Simpson’s biplane method as well. Using m-mode tricuspid annular plane systolic excursion (TAPSE) was measured in the apical 4-chamber view. 2D speckle tracking echocardiography Speckle tracking analysis was performed in the three apical views—4-chamber, 2-chamber, and the longitudinal long-axis (Figure 1). The endocardium of the LV was traced with a semiautomatic function and adjusted manually with a point-and-click method if the investigator thought the tracing to be inaccurate. Width of the regions of interest (ROI) covered the endo-, myo- and epicardium and was adjusted regionally when required. Each view covered six segments—consequently a total of 18 segments were included. Global values were calculated manually as the mean value of the peak systolic longitudinal strain of each segment. Segments if deemed untraceable could be excluded at the discretion of the analyst. Two layers of longitudinal layer-specific strain: the endomyocardial (GLSendo) and epimyocardial (GLSepi) were calculated by the software. Whole wall GLS which encompasses the whole ROI was calculated by the software as well. Figure 1 View largeDownload slide Examples of 2D-speckle tracking echocardiography. Screenshots of 2D-speckle tracking echocardiographic analysis with tracking of the LV with regional longitudinal strain traces. The views shown are the apical 4-chamber (top), 2-chamber (middle), and longitudinal long-axis (bottom). Figure 1 View largeDownload slide Examples of 2D-speckle tracking echocardiography. Screenshots of 2D-speckle tracking echocardiographic analysis with tracking of the LV with regional longitudinal strain traces. The views shown are the apical 4-chamber (top), 2-chamber (middle), and longitudinal long-axis (bottom). Statistical analysis Student’s t-test was used in comparing continuous variables exhibiting Gaussian distribution, which were displayed as mean values ± standard deviation. Wilcoxon rank-sum test was utilized for comparing continuous non-Gaussian distributed variables (LVMI and E/e′) with interquartile ranges displayed. χ2 test was used in comparing categorical variables and expressed as frequencies (percentages). Uni- and multivariable Cox regression models were created to correlate clinical and echocardiographic findings to the endpoints. TAPSE was excluded from the multivariable Cox regression due to missing values in 58 (12%) patients. For multivariable Cox regressions, three models with incremental number of confounders were constructed. Model 1 included either GLSendo, GLS, or GLSepi and the following confounders: age, gender, body mass index (BMI), prior cardiovascular disease (CVD), family history of CVD, diabetes mellitus (DM), current smoker, hypercholesterolaemia, systolic blood pressure, diastolic blood pressure, and heart rate (HR). In Model 2, we also included variables obtained from the coronary angioplasty and ACS category—STEMI, multivessel disease, culprit lesions, left anterior descending (LAD) occlusion, and left main stem coronary artery (LMS) occlusion. Model 3 additionally included all significant echocardiographic parameters—LVMI, LVEF, DT, e′, and E/e′. To compare prognostic potential of our variables, Harrell c-statistics was calculated from the univariable Cox regressions for the variables included in the multivariable Cox regressions. The Kaplan–Meier curves were created for our cohort stratified into tertiles of layer-specific GLS. Comparisons of the Harrell’s c-statistics of the diagnostic models (Models 1, 2, and 3 with or without a GLS variable) were made to assess the incremental prognostic information gained from layered GLS. A comparison with a P-value ≤0.05 in two-tailed tested was deemed statistically significant. STATA Statistics/Data analysis, MP 13.0 (StataCorp, TX, USA) was used for the statistical analysis. Results Of the 465 patients included, 176 (37.8%) developed HF, 38 suffered CD (8.2%), with 15 (3.2%) developing both. Overall there were 199 (42.7%) events during a median follow-up time of 4.6 (0.2–6.3) years. Follow-up was 100%. The mean age of our study sample was 66 ± 12 years and 74.0% were males. The HF/CD group were significantly older (67 years vs. 64 years, P = 0.019), had higher HR during the echocardiogram [79 beats per minute (bpm) vs. 71 bpm, P < 0.001], had a higher frequency of prior CVD (14.1% vs. 6.4%, P = 0.006), had a lower frequency of history of CVD in the family (23.6% vs. 34.6%, P = 0.011), a higher rate of LAD lesion (59.3% vs. 43.6%, P < 0.001), and LMS stenosis (1.5% vs. 0.0%, P = 0.045) and a lower rate of right coronary artery (RCA) (29.6% vs. 40.2%, P = 0.018) and circumflex artery (Cx) occlusion (9.5% vs. 16, 2%, P = 0.045) (Table 1). Table 1 Baseline and echocardiographic characteristics for patients stratified according to composite outcome of HF and/or CD or not Variable All (n = 465) No heart failure or cardiovascular death (n = 266) Heart failure or cardiovascular death (n = 199) P-value Baseline characteristics  Age (years) 66 ± 12 64 ± 12 67 ± 12 0.019  Male gender 74% 74% 74% 0.96  Prior CVD 9.7% 6.4% 14.1% 0.006  Hypertension 43.0% 43.6% 42.2% 0.46  Systolic blood pressure (mmHg) 137 ± 26 138 ± 25 134 ± 28 0.11  Diastolic blood pressure (mmHg) 81 ± 16 82 ± 16 80 ± 16 0.19  Heart rate (bpm) 74 ± 15 71 ± 13 79 ± 17 <0.001  Hypercholesterolaemia 24% 23.7% 25.1% 0.65  Diabetes mellitus 9.7% 7.9% 12.1% 0.13  Current smoker 46.0% 45.1% 47.2% 0.65  Body mass index (kg/m2) 26.5 ± 4.3 26.4 ± 4.2 26.5 ± 4.5 0.75  Family history of CVD 29.9% 34.6% 23.6% 0.011 Acute coronary syndrome information  STEMI 75.7% 74.4% 77.7% 0.46  NSTEMI and/or UAP 24.3% 25.6% 22.6%  Multivessel lesion 6.2% 7.5% 4.5% 0.19  Culprit lesion <0.001   LAD 50.3% 43.6% 59.3%   Cx 13.3% 16.2% 9.5%   RCA 35.7% 40.2% 29.6%   LMS 0.6% 0.0% 1.5%  Type of lesion 0.17   Type A 12.0% 13.9% 9.5%   Type B 39.1% 40.6% 37.2%   Type C 48.8% 45.5% 53.3% Echocardiographic measures  IVSd (cm) 1.1 ± 0.6 1.1 ± 0.8 1.1 ± 0.2 0.46  LVIDd (cm) 5.0 ± 2.2 5.0 ± 2.8 5.0 ± 0.6 0.75  LVPWd (cm) 1.0 ± 0.5 1.0 ± 0.7 1.0 ± 0.2 0.95  LVMI (g/m2) 90.1 [76.6, 106.8] 89.6 [76.1, 100.6] 92.5 [78.4, 111.4] 0.026  LVEF (%) 40.8 ± 11.7 44.9 ± 9.5 35.2 ± 12.1 <0.001  E/A-ratio 1.1 ± 0.41 1.0 ± 0.32 1.1 ± 0.51 0.15  DT (ms) 170.7 ± 46.2 176.3 ± 44.5 162.7 ± 47.5 0.002  e′ (cm/s) 7.5 ± 2.3 7.8 ± 2.3 7.0 ± 2.3 <0.001  E/e′ 9.7 [7.8, 12.3] 9.3 [7.5, 11.4] 10.4 [8.1, 14.1] <0.001  TAPSE (cm) 1.88 ± 0.43 2.0 ± 0.39 1.8 ± 0.46 <0.001  LAVI (mL/m2) 26.0 ± 9.2 26.0 ± 9.0 26.0 ± 9.5 0.98 Absolute layer-specific strain  GLSendo (%) 14.8 ± 4.3 16.6 ± 3.7 12.5 ± 3.8 <0.001  GLS (%) 12.8 ± 3.7 14.4 ± 3.2 10.7 ± 3.3 <0.001  GLSepi (%) 11.1 ± 3.3 12.6 ± 2.8 9.3 ± 2.9 <0.001 Variable All (n = 465) No heart failure or cardiovascular death (n = 266) Heart failure or cardiovascular death (n = 199) P-value Baseline characteristics  Age (years) 66 ± 12 64 ± 12 67 ± 12 0.019  Male gender 74% 74% 74% 0.96  Prior CVD 9.7% 6.4% 14.1% 0.006  Hypertension 43.0% 43.6% 42.2% 0.46  Systolic blood pressure (mmHg) 137 ± 26 138 ± 25 134 ± 28 0.11  Diastolic blood pressure (mmHg) 81 ± 16 82 ± 16 80 ± 16 0.19  Heart rate (bpm) 74 ± 15 71 ± 13 79 ± 17 <0.001  Hypercholesterolaemia 24% 23.7% 25.1% 0.65  Diabetes mellitus 9.7% 7.9% 12.1% 0.13  Current smoker 46.0% 45.1% 47.2% 0.65  Body mass index (kg/m2) 26.5 ± 4.3 26.4 ± 4.2 26.5 ± 4.5 0.75  Family history of CVD 29.9% 34.6% 23.6% 0.011 Acute coronary syndrome information  STEMI 75.7% 74.4% 77.7% 0.46  NSTEMI and/or UAP 24.3% 25.6% 22.6%  Multivessel lesion 6.2% 7.5% 4.5% 0.19  Culprit lesion <0.001   LAD 50.3% 43.6% 59.3%   Cx 13.3% 16.2% 9.5%   RCA 35.7% 40.2% 29.6%   LMS 0.6% 0.0% 1.5%  Type of lesion 0.17   Type A 12.0% 13.9% 9.5%   Type B 39.1% 40.6% 37.2%   Type C 48.8% 45.5% 53.3% Echocardiographic measures  IVSd (cm) 1.1 ± 0.6 1.1 ± 0.8 1.1 ± 0.2 0.46  LVIDd (cm) 5.0 ± 2.2 5.0 ± 2.8 5.0 ± 0.6 0.75  LVPWd (cm) 1.0 ± 0.5 1.0 ± 0.7 1.0 ± 0.2 0.95  LVMI (g/m2) 90.1 [76.6, 106.8] 89.6 [76.1, 100.6] 92.5 [78.4, 111.4] 0.026  LVEF (%) 40.8 ± 11.7 44.9 ± 9.5 35.2 ± 12.1 <0.001  E/A-ratio 1.1 ± 0.41 1.0 ± 0.32 1.1 ± 0.51 0.15  DT (ms) 170.7 ± 46.2 176.3 ± 44.5 162.7 ± 47.5 0.002  e′ (cm/s) 7.5 ± 2.3 7.8 ± 2.3 7.0 ± 2.3 <0.001  E/e′ 9.7 [7.8, 12.3] 9.3 [7.5, 11.4] 10.4 [8.1, 14.1] <0.001  TAPSE (cm) 1.88 ± 0.43 2.0 ± 0.39 1.8 ± 0.46 <0.001  LAVI (mL/m2) 26.0 ± 9.2 26.0 ± 9.0 26.0 ± 9.5 0.98 Absolute layer-specific strain  GLSendo (%) 14.8 ± 4.3 16.6 ± 3.7 12.5 ± 3.8 <0.001  GLS (%) 12.8 ± 3.7 14.4 ± 3.2 10.7 ± 3.3 <0.001  GLSepi (%) 11.1 ± 3.3 12.6 ± 2.8 9.3 ± 2.9 <0.001 Cx, circumflex artery; DT, deceleration time; GLS, whole wall global longitudinal strain; GLSendo, endomyocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; IVSd, interventricular septal thickness; LAD, left anterior descending; LAVI, left atrial volume indexed; LMS, left main stem coronary artery; LVEF, left ventricular ejection fraction; LVIDd, left ventricular internal diameter; LVMI, left ventricular mass indexed; LVPWd, left ventricular posterior wall; NSTEMI, non-ST-segment elevation myocardial infarction; RCA, right coronary artery; STEMI, ST-segment elevation myocardial infarction; TAPSE, tricuspidal annular plane systolic excursion; UAP, unstable angina pectoris. Table 1 Baseline and echocardiographic characteristics for patients stratified according to composite outcome of HF and/or CD or not Variable All (n = 465) No heart failure or cardiovascular death (n = 266) Heart failure or cardiovascular death (n = 199) P-value Baseline characteristics  Age (years) 66 ± 12 64 ± 12 67 ± 12 0.019  Male gender 74% 74% 74% 0.96  Prior CVD 9.7% 6.4% 14.1% 0.006  Hypertension 43.0% 43.6% 42.2% 0.46  Systolic blood pressure (mmHg) 137 ± 26 138 ± 25 134 ± 28 0.11  Diastolic blood pressure (mmHg) 81 ± 16 82 ± 16 80 ± 16 0.19  Heart rate (bpm) 74 ± 15 71 ± 13 79 ± 17 <0.001  Hypercholesterolaemia 24% 23.7% 25.1% 0.65  Diabetes mellitus 9.7% 7.9% 12.1% 0.13  Current smoker 46.0% 45.1% 47.2% 0.65  Body mass index (kg/m2) 26.5 ± 4.3 26.4 ± 4.2 26.5 ± 4.5 0.75  Family history of CVD 29.9% 34.6% 23.6% 0.011 Acute coronary syndrome information  STEMI 75.7% 74.4% 77.7% 0.46  NSTEMI and/or UAP 24.3% 25.6% 22.6%  Multivessel lesion 6.2% 7.5% 4.5% 0.19  Culprit lesion <0.001   LAD 50.3% 43.6% 59.3%   Cx 13.3% 16.2% 9.5%   RCA 35.7% 40.2% 29.6%   LMS 0.6% 0.0% 1.5%  Type of lesion 0.17   Type A 12.0% 13.9% 9.5%   Type B 39.1% 40.6% 37.2%   Type C 48.8% 45.5% 53.3% Echocardiographic measures  IVSd (cm) 1.1 ± 0.6 1.1 ± 0.8 1.1 ± 0.2 0.46  LVIDd (cm) 5.0 ± 2.2 5.0 ± 2.8 5.0 ± 0.6 0.75  LVPWd (cm) 1.0 ± 0.5 1.0 ± 0.7 1.0 ± 0.2 0.95  LVMI (g/m2) 90.1 [76.6, 106.8] 89.6 [76.1, 100.6] 92.5 [78.4, 111.4] 0.026  LVEF (%) 40.8 ± 11.7 44.9 ± 9.5 35.2 ± 12.1 <0.001  E/A-ratio 1.1 ± 0.41 1.0 ± 0.32 1.1 ± 0.51 0.15  DT (ms) 170.7 ± 46.2 176.3 ± 44.5 162.7 ± 47.5 0.002  e′ (cm/s) 7.5 ± 2.3 7.8 ± 2.3 7.0 ± 2.3 <0.001  E/e′ 9.7 [7.8, 12.3] 9.3 [7.5, 11.4] 10.4 [8.1, 14.1] <0.001  TAPSE (cm) 1.88 ± 0.43 2.0 ± 0.39 1.8 ± 0.46 <0.001  LAVI (mL/m2) 26.0 ± 9.2 26.0 ± 9.0 26.0 ± 9.5 0.98 Absolute layer-specific strain  GLSendo (%) 14.8 ± 4.3 16.6 ± 3.7 12.5 ± 3.8 <0.001  GLS (%) 12.8 ± 3.7 14.4 ± 3.2 10.7 ± 3.3 <0.001  GLSepi (%) 11.1 ± 3.3 12.6 ± 2.8 9.3 ± 2.9 <0.001 Variable All (n = 465) No heart failure or cardiovascular death (n = 266) Heart failure or cardiovascular death (n = 199) P-value Baseline characteristics  Age (years) 66 ± 12 64 ± 12 67 ± 12 0.019  Male gender 74% 74% 74% 0.96  Prior CVD 9.7% 6.4% 14.1% 0.006  Hypertension 43.0% 43.6% 42.2% 0.46  Systolic blood pressure (mmHg) 137 ± 26 138 ± 25 134 ± 28 0.11  Diastolic blood pressure (mmHg) 81 ± 16 82 ± 16 80 ± 16 0.19  Heart rate (bpm) 74 ± 15 71 ± 13 79 ± 17 <0.001  Hypercholesterolaemia 24% 23.7% 25.1% 0.65  Diabetes mellitus 9.7% 7.9% 12.1% 0.13  Current smoker 46.0% 45.1% 47.2% 0.65  Body mass index (kg/m2) 26.5 ± 4.3 26.4 ± 4.2 26.5 ± 4.5 0.75  Family history of CVD 29.9% 34.6% 23.6% 0.011 Acute coronary syndrome information  STEMI 75.7% 74.4% 77.7% 0.46  NSTEMI and/or UAP 24.3% 25.6% 22.6%  Multivessel lesion 6.2% 7.5% 4.5% 0.19  Culprit lesion <0.001   LAD 50.3% 43.6% 59.3%   Cx 13.3% 16.2% 9.5%   RCA 35.7% 40.2% 29.6%   LMS 0.6% 0.0% 1.5%  Type of lesion 0.17   Type A 12.0% 13.9% 9.5%   Type B 39.1% 40.6% 37.2%   Type C 48.8% 45.5% 53.3% Echocardiographic measures  IVSd (cm) 1.1 ± 0.6 1.1 ± 0.8 1.1 ± 0.2 0.46  LVIDd (cm) 5.0 ± 2.2 5.0 ± 2.8 5.0 ± 0.6 0.75  LVPWd (cm) 1.0 ± 0.5 1.0 ± 0.7 1.0 ± 0.2 0.95  LVMI (g/m2) 90.1 [76.6, 106.8] 89.6 [76.1, 100.6] 92.5 [78.4, 111.4] 0.026  LVEF (%) 40.8 ± 11.7 44.9 ± 9.5 35.2 ± 12.1 <0.001  E/A-ratio 1.1 ± 0.41 1.0 ± 0.32 1.1 ± 0.51 0.15  DT (ms) 170.7 ± 46.2 176.3 ± 44.5 162.7 ± 47.5 0.002  e′ (cm/s) 7.5 ± 2.3 7.8 ± 2.3 7.0 ± 2.3 <0.001  E/e′ 9.7 [7.8, 12.3] 9.3 [7.5, 11.4] 10.4 [8.1, 14.1] <0.001  TAPSE (cm) 1.88 ± 0.43 2.0 ± 0.39 1.8 ± 0.46 <0.001  LAVI (mL/m2) 26.0 ± 9.2 26.0 ± 9.0 26.0 ± 9.5 0.98 Absolute layer-specific strain  GLSendo (%) 14.8 ± 4.3 16.6 ± 3.7 12.5 ± 3.8 <0.001  GLS (%) 12.8 ± 3.7 14.4 ± 3.2 10.7 ± 3.3 <0.001  GLSepi (%) 11.1 ± 3.3 12.6 ± 2.8 9.3 ± 2.9 <0.001 Cx, circumflex artery; DT, deceleration time; GLS, whole wall global longitudinal strain; GLSendo, endomyocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; IVSd, interventricular septal thickness; LAD, left anterior descending; LAVI, left atrial volume indexed; LMS, left main stem coronary artery; LVEF, left ventricular ejection fraction; LVIDd, left ventricular internal diameter; LVMI, left ventricular mass indexed; LVPWd, left ventricular posterior wall; NSTEMI, non-ST-segment elevation myocardial infarction; RCA, right coronary artery; STEMI, ST-segment elevation myocardial infarction; TAPSE, tricuspidal annular plane systolic excursion; UAP, unstable angina pectoris. Supplementary material online, Tables1 and 2 display baseline characteristics for the cohort stratified according to the individual endpoints separately. The HF/CD group had a significantly larger LVMI (93 g/m2 vs. 90 g/m2, P = 0.026), lower LVEF (35% vs. 45%, P < 0.001), a shorter DT (163 ms vs. 176 ms, P = 0.002), a lower e′ (7.0 cm/s vs. 7.8 cm/s, P < 0.001), a higher E/e′ (10.4 vs. 9.3, P < 0.001), and reduced TAPSE (1.8 cm vs. 2.0 cm, P < 0.001). The longitudinal strain measurements—GLSendo (12% vs. 17%, P < 0.001), GLS (11% vs. 14%, P < 0.001), and GLSepi (9% vs. 13%, P < 0.001) were all significantly lower in the HF/CD group as well (Table 1). Outcome Both GLSendo, GLS, and GLSepi remained independently associated measures after adjustments in Models 1 and 2 with onset of the composite outcome (HF/CD) and each outcome when evaluated separately (Table 2). After adjusting for all the echocardiographic parameters (Model 3) GLSendo, GLS and GLSepi remained independently associated with the composite outcome and HF alone; however, only GLSepi remained independently associated with CD (Table 2). No other echocardiographic measures remained independently associated with the composite outcome (Table 2). Table 2 Multivariable Cox regression and c-statistics (unadjusted) for developing HF and/or CD Composite endpoint (HF and/or CD) (220 events) hazard ratio (95% CI) P-value HF (190 events) hazard ratio (95% CI) P-value CD (46 events) hazard ratio (95% CI) P-value Unadjusted  GLSendo per 1% decrease 1.22 (1.18–1.26) <0.001 1.21 (1.17–1.26) <0.001 1.29 (1.19–1.41) <0.001 C-stat 0.73 C-stat 0.73 C-stat 0.77  GLS per 1% decrease 1.26 (1.21–1.31) <0.001 1.25 (1.20–1.30) <0.001 1.34 (1.22–1.48) <0.001 C-stat 0.74 C-stat 0.73 C-stat 0.77  GLSepi per 1% decrease 1.29 (1.23–1.34) <0.001 1.27 (1.22–1.33) <0.001 1.38 (1.25–1.53) <0.001 C-stat. 0.74 C-stat 0.73 C-stat 0.77 Model 1  GLSendo per 1% decrease 1.20 (1.15–1.25) <0.001 1.20 (1.15–1.26) <0.001 1.17 (1.06–1.29) 0.002  GLS per 1% decrease 1.24 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.20 (1.07–1.34) 0.002  GLSepi per 1% decrease 1.26 (1.20–133) <0.001 1.26 (1.20–1.34) <0.001 1.22 (1.08–1.37) 0.001 Model 2  GLSendo per 1% decrease 1.20 (1.15–1.26) <0.001 1.20 (1.15–1.26) <0.001 1.25 (1.12–1.39) <0.001  GLS per 1% decrease 1.25 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.29 (1.14–1.46) <0.001  GLSepi per 1% decrease 1.26 (1.20–1.35) <0.001 1.27 (1.20–1.35) <0.001 1.32 (1.16–1.51) <0.001 Model 3  GLSendo per 1% decrease 1.19 (1.10–1.28) <0.001 1.18 (1.09–1.28)| <0.001 1.20 (0.97–1.48) 0.10  GLS per 1% decrease 1.24 (1.14–1.35) <0.001 1.23 (1.13–1.35) <0.001 1.26 (0.98–1.63) 0.071  GLSepi per 1% decrease 1.26 (1.15–1.39) <0.001 1.26 (1.14–1.39) <0.001 1.34 (1.00–1.80) 0.048 Composite endpoint (HF and/or CD) (220 events) hazard ratio (95% CI) P-value HF (190 events) hazard ratio (95% CI) P-value CD (46 events) hazard ratio (95% CI) P-value Unadjusted  GLSendo per 1% decrease 1.22 (1.18–1.26) <0.001 1.21 (1.17–1.26) <0.001 1.29 (1.19–1.41) <0.001 C-stat 0.73 C-stat 0.73 C-stat 0.77  GLS per 1% decrease 1.26 (1.21–1.31) <0.001 1.25 (1.20–1.30) <0.001 1.34 (1.22–1.48) <0.001 C-stat 0.74 C-stat 0.73 C-stat 0.77  GLSepi per 1% decrease 1.29 (1.23–1.34) <0.001 1.27 (1.22–1.33) <0.001 1.38 (1.25–1.53) <0.001 C-stat. 0.74 C-stat 0.73 C-stat 0.77 Model 1  GLSendo per 1% decrease 1.20 (1.15–1.25) <0.001 1.20 (1.15–1.26) <0.001 1.17 (1.06–1.29) 0.002  GLS per 1% decrease 1.24 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.20 (1.07–1.34) 0.002  GLSepi per 1% decrease 1.26 (1.20–133) <0.001 1.26 (1.20–1.34) <0.001 1.22 (1.08–1.37) 0.001 Model 2  GLSendo per 1% decrease 1.20 (1.15–1.26) <0.001 1.20 (1.15–1.26) <0.001 1.25 (1.12–1.39) <0.001  GLS per 1% decrease 1.25 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.29 (1.14–1.46) <0.001  GLSepi per 1% decrease 1.26 (1.20–1.35) <0.001 1.27 (1.20–1.35) <0.001 1.32 (1.16–1.51) <0.001 Model 3  GLSendo per 1% decrease 1.19 (1.10–1.28) <0.001 1.18 (1.09–1.28)| <0.001 1.20 (0.97–1.48) 0.10  GLS per 1% decrease 1.24 (1.14–1.35) <0.001 1.23 (1.13–1.35) <0.001 1.26 (0.98–1.63) 0.071  GLSepi per 1% decrease 1.26 (1.15–1.39) <0.001 1.26 (1.14–1.39) <0.001 1.34 (1.00–1.80) 0.048 Model 1 is adjusted for age, gender BMI, prior CVD, family history of CVD, diabetes mellitus, current smoker, hypercholesterolaemia, systolic blood pressure, diastolic blood pressure, HR. Model 2 is adjusted for previous mentioned variables in Model 1 as well as STEMI, multivessel culprit lesions, LAD occlusion, and LMS occlusion. Model 3 is adjusted for all previous mentioned variables in Models 1 and 2, additionally LVMI, LVEF, DT, e′, and E/e′. Table 2 Multivariable Cox regression and c-statistics (unadjusted) for developing HF and/or CD Composite endpoint (HF and/or CD) (220 events) hazard ratio (95% CI) P-value HF (190 events) hazard ratio (95% CI) P-value CD (46 events) hazard ratio (95% CI) P-value Unadjusted  GLSendo per 1% decrease 1.22 (1.18–1.26) <0.001 1.21 (1.17–1.26) <0.001 1.29 (1.19–1.41) <0.001 C-stat 0.73 C-stat 0.73 C-stat 0.77  GLS per 1% decrease 1.26 (1.21–1.31) <0.001 1.25 (1.20–1.30) <0.001 1.34 (1.22–1.48) <0.001 C-stat 0.74 C-stat 0.73 C-stat 0.77  GLSepi per 1% decrease 1.29 (1.23–1.34) <0.001 1.27 (1.22–1.33) <0.001 1.38 (1.25–1.53) <0.001 C-stat. 0.74 C-stat 0.73 C-stat 0.77 Model 1  GLSendo per 1% decrease 1.20 (1.15–1.25) <0.001 1.20 (1.15–1.26) <0.001 1.17 (1.06–1.29) 0.002  GLS per 1% decrease 1.24 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.20 (1.07–1.34) 0.002  GLSepi per 1% decrease 1.26 (1.20–133) <0.001 1.26 (1.20–1.34) <0.001 1.22 (1.08–1.37) 0.001 Model 2  GLSendo per 1% decrease 1.20 (1.15–1.26) <0.001 1.20 (1.15–1.26) <0.001 1.25 (1.12–1.39) <0.001  GLS per 1% decrease 1.25 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.29 (1.14–1.46) <0.001  GLSepi per 1% decrease 1.26 (1.20–1.35) <0.001 1.27 (1.20–1.35) <0.001 1.32 (1.16–1.51) <0.001 Model 3  GLSendo per 1% decrease 1.19 (1.10–1.28) <0.001 1.18 (1.09–1.28)| <0.001 1.20 (0.97–1.48) 0.10  GLS per 1% decrease 1.24 (1.14–1.35) <0.001 1.23 (1.13–1.35) <0.001 1.26 (0.98–1.63) 0.071  GLSepi per 1% decrease 1.26 (1.15–1.39) <0.001 1.26 (1.14–1.39) <0.001 1.34 (1.00–1.80) 0.048 Composite endpoint (HF and/or CD) (220 events) hazard ratio (95% CI) P-value HF (190 events) hazard ratio (95% CI) P-value CD (46 events) hazard ratio (95% CI) P-value Unadjusted  GLSendo per 1% decrease 1.22 (1.18–1.26) <0.001 1.21 (1.17–1.26) <0.001 1.29 (1.19–1.41) <0.001 C-stat 0.73 C-stat 0.73 C-stat 0.77  GLS per 1% decrease 1.26 (1.21–1.31) <0.001 1.25 (1.20–1.30) <0.001 1.34 (1.22–1.48) <0.001 C-stat 0.74 C-stat 0.73 C-stat 0.77  GLSepi per 1% decrease 1.29 (1.23–1.34) <0.001 1.27 (1.22–1.33) <0.001 1.38 (1.25–1.53) <0.001 C-stat. 0.74 C-stat 0.73 C-stat 0.77 Model 1  GLSendo per 1% decrease 1.20 (1.15–1.25) <0.001 1.20 (1.15–1.26) <0.001 1.17 (1.06–1.29) 0.002  GLS per 1% decrease 1.24 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.20 (1.07–1.34) 0.002  GLSepi per 1% decrease 1.26 (1.20–133) <0.001 1.26 (1.20–1.34) <0.001 1.22 (1.08–1.37) 0.001 Model 2  GLSendo per 1% decrease 1.20 (1.15–1.26) <0.001 1.20 (1.15–1.26) <0.001 1.25 (1.12–1.39) <0.001  GLS per 1% decrease 1.25 (1.18–1.31) <0.001 1.24 (1.18–1.31) <0.001 1.29 (1.14–1.46) <0.001  GLSepi per 1% decrease 1.26 (1.20–1.35) <0.001 1.27 (1.20–1.35) <0.001 1.32 (1.16–1.51) <0.001 Model 3  GLSendo per 1% decrease 1.19 (1.10–1.28) <0.001 1.18 (1.09–1.28)| <0.001 1.20 (0.97–1.48) 0.10  GLS per 1% decrease 1.24 (1.14–1.35) <0.001 1.23 (1.13–1.35) <0.001 1.26 (0.98–1.63) 0.071  GLSepi per 1% decrease 1.26 (1.15–1.39) <0.001 1.26 (1.14–1.39) <0.001 1.34 (1.00–1.80) 0.048 Model 1 is adjusted for age, gender BMI, prior CVD, family history of CVD, diabetes mellitus, current smoker, hypercholesterolaemia, systolic blood pressure, diastolic blood pressure, HR. Model 2 is adjusted for previous mentioned variables in Model 1 as well as STEMI, multivessel culprit lesions, LAD occlusion, and LMS occlusion. Model 3 is adjusted for all previous mentioned variables in Models 1 and 2, additionally LVMI, LVEF, DT, e′, and E/e′. The Kaplan–Meier curves (Figure 2) displaying the patient cohort stratified into tertiles of GLSendo, GLS, and GLSepi, respectively, demonstrates the incrementally increased risk of developing HF and/or CD with lower tertile of GLS. The lowest tertile displayed a eight- to nine-fold increased risk of developing HF and/or CD compared to the highest tertile [GLSendo: 1 tertile vs. 3 tertile: hazard ratio 8.0 (5.1–12.5), P < 0.001; GLS: 1 tertile vs. 3 tertile: hazard ratio 9.2 (5.8–14.7), P < 0.001; GLSepi: 1 tertile vs. 3 tertile: hazard ratio 8.5 (5.4–13.2), P < 0.001]. The intermediate tertile had a three- to four-fold risk compared to the highest tertile [GLSendo: hazard ratio 3.2 (2.0–5.2), P < 0.001; GLS: hazard ratio 4.0 (2.5–6.6), P < 0.001; GLSepi: hazard ratio 3.5 (2.2–5.7), P < 0.001]. Figure 2 View largeDownload slide Layer-specific strain and outcome. The Kaplan–Meier curves displaying the probability of staying event free. The horizontal axis displays the time from the ACS expressed in days. The vertical axis represents the cumulative probability of staying event free of HF and/or CD. The study population is stratified into three groups based on the tertiles of GLSendo, GLS, or GLSepi. 95% CI and hazard ratios included. Number of patient at risk at every 1000 days is displayed below each curve. CD, cardiovascular death; CI, confidence interval; GLS, whole wall global longitudinal strain; GLSendo, endocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; HF, heart failure. Figure 2 View largeDownload slide Layer-specific strain and outcome. The Kaplan–Meier curves displaying the probability of staying event free. The horizontal axis displays the time from the ACS expressed in days. The vertical axis represents the cumulative probability of staying event free of HF and/or CD. The study population is stratified into three groups based on the tertiles of GLSendo, GLS, or GLSepi. 95% CI and hazard ratios included. Number of patient at risk at every 1000 days is displayed below each curve. CD, cardiovascular death; CI, confidence interval; GLS, whole wall global longitudinal strain; GLSendo, endocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; HF, heart failure. Incremental prognostic yield of GLS obtained from the different layers and whole wall When comparing the Harrell’s c-statistics obtained from unadjusted Cox regressions for GLSendo with the c-statistics obtained for GLS (0.733 vs. 0.739, P = 0.017) and c-statistics obtained for GLS compared to the c-statistics obtained for GLSepi (0.739 vs. 0.741, P = 0.38), we found the prognostic power of GLS to be significantly higher than that of GLSendo, and the c-statistics for GLS was lower than for GLSepi; however, this difference was not statistically significant. Adding either GLS or GLSepi to all the variables included in Model 3 resulted in a significant increase in the c-statistics than Model 3 alone [Model 3 with GLS: 0.76 (0.72–0.80) vs. 0.74 (0.70–0.77), P = 0.048; Model 3 with GLSepi: 0.76 (0.72–0.80) vs. 0.74 (0.70–0.77), P = 0.039]. Adding GLSendo to the variables included in Model 3, only displayed a non-significant increase in the c-statistics when compared to Model 3 alone [Model 3 with GLSendo: 0.75 (0.72–0.79) vs. 0.74 (0.70–0.77), P = 0.075]. Models 1 and 2 significantly improved when adding GLSendo, GLS, or GLSepi (Figure 3). Figure 3 View largeDownload slide Incremental prognostic information when adding layer-specific strain. (A) Column diagram displaying Harrell’s C-statistics comparison in various models for predicting HF and/or CD with or without GLSendo GLS or GLSepi. Model 1 includes age, gender BMI, prior CVD, family history of CVD, DM, current smoker, hypercholesterolaemia, systolic blood pressure, diastolic blood pressure, and heart rate. Model 2 includes the same variables as Model 1 along with the PCI variables—STEMI, multivessel culprit lesions, LAD occlusion and LMS occlusion. Model 3 includes the same variable as Models 1 and 2 in addition to LVMI, LVEF, DT, e′, and E/e′. 95% CI included. BMI, body mass index; CD, cardiovascular death; CI, confidence interval; CVD, cardiovascular disease; DM, diabetes mellitus; DT, deceleration time; e′, early peak myocardial diastolic velocity; E/e′, peak early filling velocity divided by early peak myocardial diastolic velocity; GLS, whole wall global longitudinal strain; GLSendo, endocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; HF, heart failure; LAD, left anterior descending, LMS, left main stem; LVEF, left ventricular ejection fraction; LVMI, left ventricle mass index; PCI, percutaneous coronary intervention; STEMI, ST-segment elevation myocardial infarction. Figure 3 View largeDownload slide Incremental prognostic information when adding layer-specific strain. (A) Column diagram displaying Harrell’s C-statistics comparison in various models for predicting HF and/or CD with or without GLSendo GLS or GLSepi. Model 1 includes age, gender BMI, prior CVD, family history of CVD, DM, current smoker, hypercholesterolaemia, systolic blood pressure, diastolic blood pressure, and heart rate. Model 2 includes the same variables as Model 1 along with the PCI variables—STEMI, multivessel culprit lesions, LAD occlusion and LMS occlusion. Model 3 includes the same variable as Models 1 and 2 in addition to LVMI, LVEF, DT, e′, and E/e′. 95% CI included. BMI, body mass index; CD, cardiovascular death; CI, confidence interval; CVD, cardiovascular disease; DM, diabetes mellitus; DT, deceleration time; e′, early peak myocardial diastolic velocity; E/e′, peak early filling velocity divided by early peak myocardial diastolic velocity; GLS, whole wall global longitudinal strain; GLSendo, endocardial global longitudinal strain; GLSepi, epimyocardial global longitudinal strain; HF, heart failure; LAD, left anterior descending, LMS, left main stem; LVEF, left ventricular ejection fraction; LVMI, left ventricle mass index; PCI, percutaneous coronary intervention; STEMI, ST-segment elevation myocardial infarction. Finally, we found no significant changes in Harrell’s c-statistics when adding all the conventional echocardiographic parameters to the models when either one of the GLS measures were already included (Figure 3). Hence, neither of the models significantly improved when adding all other echocardiographic measures to the models where the GLS measures were already included. Discussion This is the first study evaluating the association between layer-specific GLS and HF and/or CD following ACS. In the present study, we found that all three layer-specific GLS measurements provided superior prognostic information on the risk of developing HF and/or CD compared to all other echocardiographic measure. Both GLS obtained from the epicardial layer and the whole wall provided incremental prognostic information when added to all other predictors in the current population. GLSepi was, however, the only independent echocardiographic predictor of CD alone. In addition, the risk assessment did not significantly improve when adding all other significantly associated echocardiographic parameters to the models when either GLSendo, GLS, or GLSepi were already included (Figure 3). In our study, patients developing HF and/or CD were older, had a higher tendency of prior CVD, and increased HR, which is consistent with the literature.3 Cardiac imaging especially transthoracic echocardiography plays a central role in the assessment of LV function following MI18 and in diagnosis of HF.3 One of the strongest predictors of survival in ACS patients is LV function evaluated with an echocardiogram before hospital discharge.18 Echocardiography is the most common imaging method to evaluate regional and global LV function after STEMI. In the presence of severe LV systolic dysfunction, it is recommended to re-evaluate with a subsequent echocardiogram around 6–12 weeks after initial hospitalization.18 Currently only LVEF and not GLS is recommended when assessing systolic function.17 This is in spite of the fact that LVEF can be preserved in the presence of HF. This is hypothesized to be due to an increase in the circumferential and radial myocardial deformation occurring concomitant with an impaired longitudinal deformation which can be identified using speckle tracking imaging.4,6 Hence, LVEF remains preserved despite of a reduced longitudinal deformation. Layer-specific GLS has been suggested as a potential method of assessing LV function.19,20 Several studies have demonstrated layer-specific GLS to be independently associated with the presence of coronary artery disease. Yet these studies display a discrepancy in which specific layer they found to have the strongest diagnostic power.21–24 GLSendo and circumferential endocardial strain have also been displayed to be associated with subsequent cardiac events in patients with chronic ischaemic cardiomyopathy.25 However, there are no published studies evaluating the prognostic value of layer-specific GLS for HF and/or CD following ACS. In the present report, the magnitude of layer-specific GLS decreased from the endocardium to the epicardium (Tables 1 and 2), which is in concordance with previous studies.19,21–23 We found all layers of GLS to be associated with HF and/or CD, after adjusting for clinical and echocardiographic parameters. Yet, GLSepi and GLS displayed stronger prognostic power (Table 2, Figure 3). In our adjusted models, the hazard ratios of layer-specific GLS increased in size from the endocardium to the epicardium per 1% decrease (Table 2). GLSepi remained as the only independent predictor when predicting only CD (Table 2). Our group has recently demonstrated similar results when using layer-specific GLS to diagnose ischaemia in a cohort of 80 patients with stable angina pectoris. In this study, we found that GLSepi was the only independent echocardiographic measure associated with the presence of significance of coronary artery stenosis after multivariable adjustment.23,24 These findings, like the results of the present report, which demonstrates that GLSepi might be the best measure both to diagnose significant coronary artery stenosis, but more importantly, subsequent outcome, are however still in contrary to other studies using layer-specific GLS. Zhang et al.21 (cohort of 139 patients with 79 having complex coronary artery disease) and Sarvari et al.22 (cohort of 77 patients with 49 having significant coronary artery disease) found GLSendo to be the best measurement in identifying complex coronary artery disease and predicting the severity of coronary lesions in NSTEMI patients. We have tested the intra- and inter-observer variability of layered 2D-speckle tracking echocardiography in 20 patients with ischaemic heart disease. Intra-observer Bland–Altman analysis displayed a bias of 0.473 ± 1.45 with coefficients of variation (CV) of 6.6% for GLSendo, 0.447 ± 1.21 with CV of 6.4% for GLS, and 0.420 ± 1.07 with CV of 6.5% for GLSepi. Whereas, the inter-observer analysis displayed a bias of −0.965 ± 2.36 with CV of 11.1% for GLSendo, −0.887 ± 1.92 with a CV of 10.5% for GLS, and −0.815 ± 1.64 with a CV of 10.3% for GLSepi.23,24 The better reproducibility of 2D-speckle tracking of the epimyocardium than that of the endomyocardium might be one of the reasons several new studies found GLSepi to be the best measure for diagnostic and prognostic purposes. This might be due to the better traceability of the epimyocardium as opposed to that of the endocardial border. The endomyocardial region might generally be affected to a higher degree than the epimyocardial region from several factors. Such factors could be the more varied contour of the endocardial border, which can make it more difficult to line up the ROI. Apical foreshortening might also influence the traceability of apical endocardial border, whereas the epicardial border would be less evasive. Apical foreshortening has been shown to affect longitudinal strain values significantly.26 Additionally, regional epimyocardial longitudinal strain has been shown to be more homogenous than the inner two layers, which have significant variance from base to apex.19 As mentioned previously it is currently not recommended to evaluate GLS after ACS or when diagnosing HF.3,17 Applying speckle tracking echocardiography to the standard echocardiographic work-up following ACS would require little additional time and effort. Furthermore, due to the semi-automated nature of the method, the intra- and inter-observer variability is low. The superiority of layer-specific longitudinal strain as well as whole wall longitudinal strain when compared to LVEF indicates that it might be time to reassess the recommendations on LV function measurements. We demonstrated both layer-specific GLSendo and GLSepi and GLS to provide superior prognostic information regarding the risk of developing HF and/or CD compared to all other echocardiographic measures. We find that it might be clinically relevant to add GLSepi, GLSendo, and GLS to the risk stratification scheme of ACS patients. This could assist in identifying patients at greatest risk of adverse morbidity and mortality. Limitations There may be residential confounders as this is a retrospectively planned study. However, the analysis was based on consecutive patients from a well-defined registry. Retrieving endpoints—development of HF and CD—from the Danish National Board of Health’s National Patient Registry and the Danish Register of Causes of Death using International Classifications of Diseases codes may be questioned. However, diagnostic codes obtained through the Danish National Patient Registry have proven highly accurate. In 2011, Thygesen et al. investigated the positive predictive value of International Classification of Diseases-10 diagnostic codes. The authors found diagnostic codes of HF obtained through the Danish National Patient Registry to have a positive predictive value of 100% (92.9–100%), when reviewing discharge summaries and medical records.27,28 Unfortunately, the degree of mitral regurgitation was not reassessed in the present study. Information about the medication of the patients such as the use of betablockers was not available either. Data on medication as well as mitral regurgitation would have resulted in a more thorough description of the cohort. Furthermore, we regret we were not able to examine changes in medication following discharge from the hospital. There might be some concerns regarding overfitting, when utilizing a multivariable Cox regression with as many variables included as in Model 3. However, interpretation of an overfitted multivariable Cox regression can still be acceptable, when it is done to control for potential confounders and not for building a prediction model.29 One of the challenges of speckle tracking is defining the epicardial border, which in some cases can appear unclear and hence may bias GLSepi. However, we have found that intra- and inter-observer reproducibility is superior for GLSepi when compared to GLSendo and whole wall GLS, hence, this issue does not seem to make GLSepi a less valid measure. Regretfully, follow-up echocardiograms were not performed for the present study. Lastly, we cannot exclude the possibility of patients having an undiagnosed pre-existing condition prior to their initial ACS hospitalization influencing our results. Conclusion Layer-specific GLSendo, GLS, and GLSepi were all independently associated with HF and/or CD. GLS and GLSepi provided incremental prognostic information when added to all other significantly associated measures. Supplementary data Supplementary data are available at European Heart Journal - Cardiovascular Imaging online. Conflict of interest: None declared. References 1 Wang TJ , Evans JC , Benjamin EJ , Levy D , LeRoy EC , Vasan RS. Natural history of asymptomatic left ventricular systolic dysfunction in the community . Circulation 2003 ; 108 : 977 – 82 . Google Scholar CrossRef Search ADS PubMed 2 Effect of enalapril on mortality and the development of heart failure in asymptomatic patients with reduced left ventricular ejection fractions. The SOLVD Investigattors . N Engl J Med 1992 ; 327 : 685 – 91 . CrossRef Search ADS PubMed 3 Ponikowski P , Voors AA , Anker SD , Bueno H , Cleland JGF , Coats AJS et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: the Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC . Eur J Heart Fail 2016 ; 18 : 891 – 975 . Google Scholar CrossRef Search ADS PubMed 4 Biering-Sørensen T , Santos M , Rivero J , McCullough SD , West E , Opotowsky AR et al. Left ventricular deformation at rest predicts exercise-induced elevation in pulmonary artery wedge pressure in patients with unexplained dyspnoea . Eur J Heart Fail 2016 ; 19 : 101 – 110 . Google Scholar CrossRef Search ADS PubMed 5 Biering-Sørensen T , Biering-Sørensen SR , Olsen FJ , Sengeløv M , Jørgensen PG , Mogelvang R et al. Global longitudinal strain by echocardiography predicts long-term risk of cardiovascular morbidity and mortality in a low-risk general population: the Copenhagen City Heart Study . Circ Cardiovasc Imaging 2017 ; 10 : e005521 . Google Scholar CrossRef Search ADS PubMed 6 Biering-Sørensen T , Solomon SD. Assessing contractile function when ejection fraction is normal: a case for strain . Circ Cardiovasc Imaging 2015 ; 8 : e004181. Google Scholar CrossRef Search ADS PubMed 7 Kalam K , Otahal P , Marwick TH. Prognostic implications of global LV dysfunction: a systematic review and meta-analysis of global longitudinal strain and ejection fraction . Heart Br Card Soc 2014 ; 100 : 1673 – 80 . 8 Sengeløv M , Jørgensen PG , Jensen JS , Bruun NE , Olsen FJ , Fritz-Hansen T et al. Global longitudinal strain is a superior predictor of all-cause mortality in heart failure with reduced ejection fraction . JACC Cardiovasc Imaging 2015 ; 8 : 1351 – 9 . Google Scholar CrossRef Search ADS PubMed 9 Biering-Sørensen T , Knappe D , Pouleur A-C , Claggett B , Wang PJ , Moss AJ et al. Regional longitudinal deformation improves prediction of ventricular tachyarrhythmias in patients with heart failure with reduced ejection fraction: a MADIT-CRT Substudy (Multicenter Automatic Defibrillator Implantation Trial-Cardiac Resynchronization Therapy) . Circ Cardiovasc Imaging 2017 ; 10 : e005096 . Google Scholar CrossRef Search ADS PubMed 10 Olsen FJ , Pedersen S , Jensen JS , Biering-Sørensen T. Global longitudinal strain predicts incident atrial fibrillation and stroke occurrence after acute myocardial infarction . Medicine (Baltimore) 2016 ; 95 : e5338. Google Scholar CrossRef Search ADS PubMed 11 Shah AM , Claggett B , Sweitzer NK , Shah SJ , Anand IS , Liu L et al. Prognostic importance of impaired systolic function in heart failure with preserved ejection fraction and the impact of spironolactone . Circulation 2015 ; 132 : 402 – 14 . Google Scholar CrossRef Search ADS PubMed 12 Mor-Avi V , Lang RM , Badano LP , Belohlavek M , Cardim NM , Derumeaux G et al. Current and evolving echocardiographic techniques for the quantitative evaluation of cardiac mechanics: ASE/EAE consensus statement on methodology and indications endorsed by the Japanese Society of Echocardiography . Eur J Echocardiogr J Work Group Echocardiogr Eur Soc Cardiol 2011 ; 12 : 167 – 205 . Google Scholar CrossRef Search ADS 13 Joyce E. LVEF: long-standing monarch of systolic dysfunction, buckling under the strain? Eur J Heart Fail 2014 ; 16 : 1270 – 2 . Google Scholar CrossRef Search ADS PubMed 14 Russo C , Jin Z , Elkind MSV , Rundek T , Homma S , Sacco RL et al. Prevalence and prognostic value of subclinical left ventricular systolic dysfunction by global longitudinal strain in a community-based cohort . Eur J Heart Fail 2014 ; 16 : 1301 – 9 . Google Scholar CrossRef Search ADS PubMed 15 Vendelin M , Bovendeerd PHM , Engelbrecht J , Arts T. Optimizing ventricular fibers: uniform strain or stress, but not ATP consumption, leads to high efficiency . Am J Physiol Heart Circ Physiol 2002 ; 283 : H1072 – 81 . Google Scholar CrossRef Search ADS PubMed 16 Reimer KA , Lowe JE , Rasmussen MM , Jennings RB. The wavefront phenomenon of ischemic cell death. 1. Myocardial infarct size vs duration of coronary occlusion in dogs . Circulation 1977 ; 56 : 786 – 94 . Google Scholar CrossRef Search ADS PubMed 17 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 . Eur Heart J Cardiovasc Imaging 2015 ; 16 : 233 – 70 . Google Scholar CrossRef Search ADS PubMed 18 Ibanez B , James S , Agewall S , Antunes MJ , Bucciarelli-Ducci C , Bueno H et al. 2017 ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation: the Task Force for the management of acute myocardial infarction in patients presenting with ST-segment elevation of the European Society of Cardiology (ESC) . Eur Heart J 2017 ; 39 : 119 – 177 . Google Scholar CrossRef Search ADS 19 Leitman M , Lysiansky M , Lysyansky P , Friedman Z , Tyomkin V , Fuchs T et al. Circumferential and longitudinal strain in 3 myocardial layers in normal subjects and in patients with regional left ventricular dysfunction . J Am Soc Echocardiogr 2010 ; 23 : 64 – 70 . Google Scholar CrossRef Search ADS PubMed 20 Shi J , Pan C , Kong D , Cheng L , Shu X. Left ventricular longitudinal and circumferential layer-specific myocardial strains and their determinants in healthy subjects . Echocardiography (Mt Kisco NY) 2016 ; 33 : 510 – 8 . Google Scholar CrossRef Search ADS 21 Zhang L , Wu W-C , Ma H , Wang H. Usefulness of layer-specific strain for identifying complex CAD and predicting the severity of coronary lesions in patients with non-ST-segment elevation acute coronary syndrome: compared with Syntax score . Int J Cardiol 2016 ; 223 : 1045 – 52 . Google Scholar CrossRef Search ADS PubMed 22 Sarvari SI , Haugaa KH , Zahid W , Bendz B , Aakhus S , Aaberge L et al. Layer-specific quantification of myocardial deformation by strain echocardiography may reveal significant CAD in patients with non-ST-segment elevation acute coronary syndrome . JACC Cardiovasc Imaging 2013 ; 6 : 535 – 44 . Google Scholar CrossRef Search ADS PubMed 23 Hagemann C , Hoffmann S , Olsen FJ , Jørgensen P , Jensen J , Biering-Sorensen T. Layer-specific global longitudinal strain reveals impaired cardiac function in patients with reversible ischemia [abstract] . J Am Coll Cardiol 2016 ; 67 : 1592. Google Scholar CrossRef Search ADS 24 Hagemann C , Hoffmann S , Andersen R , Hansen T , Olsen FJ , Jørgensen P et al. Layer-specific strain analysis by two-dimensional speckle tracking echocardiography: improvement of coronary artery disease diagnostics in patients with stable angina pectoris [abstract] . J Am Coll Cardiol 2017 ; 69 : 1634. Google Scholar CrossRef Search ADS 25 Hamada S , Schroeder J , Hoffmann R , Altiok E , Keszei A , Almalla M et al. Prediction of outcomes in patients with chronic ischemic cardiomyopathy by layer-specific strain echocardiography: a proof of concept . J Am Soc Echocardiogr 2016 ; 29 : 412 – 20 . Google Scholar CrossRef Search ADS PubMed 26 Tarr A , Stöbe S , Trache T , Kluge J-G , Varga A , Pfeiffer D et al. The impact of foreshortening on regional strain–a comparison of regional strain evaluation between speckle tracking and tissue velocity imaging . Ultraschall Med 2013 ; 34 : 446 – 53 . Google Scholar CrossRef Search ADS PubMed 27 Thygesen SK , Christiansen CF , Christensen S , Lash TL , Sørensen HT. The predictive value of ICD-10 diagnostic coding used to assess Charlson comorbidity index conditions in the population-based Danish National Registry of Patients . BMC Med Res Methodol 2011 ; 11 : 83. Google Scholar CrossRef Search ADS PubMed 28 Christiansen MN , Køber L , Weeke P , Vasan RS , Jeppesen JL , Smith JG et al. Age-specific trends in incidence, mortality, and comorbidities of heart failure in Denmark, 1995 to 2012 . Circulation 2017 ; 135 : 1214 – 23 . Google Scholar CrossRef Search ADS PubMed 29 Vittinghoff E , McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression . Am J Epidemiol 2007 ; 165 : 710 – 8 . Google Scholar CrossRef Search ADS PubMed Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

European Heart Journal – Cardiovascular ImagingOxford University Press

Published: Mar 30, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off