Abstract
Aims
Deformation imaging has demonstrated incremental prognostic value compared to left ventricular ejection fraction (LVEF) in specific cardiovascular diseases; however, it remains widely underused. Consequently, we sought to investigate the significance of deformation imaging in an all-comers population to assess its value in a routine in- and outpatient setting.
Methods and results
Patients were prospectively recruited to the single-centre cardiac magnetic resonance registry. Volumetric right and left ventricular (RV/LV) analyses and global circumferential strain were assessed on short-axis stacks, RV and LV global longitudinal strain (GLS) on long-axis views respectively. Follow-up was conducted for primary (all-cause mortality and heart failure hospitalization) and secondary (all-cause mortality, hospitalized angina, infarction, and stroke) endpoints. In total, n = 1655 patients met n = 68 primary and n = 107 secondary endpoints during a median follow-up of 399 days. GLS (HR 1.12, 95%CI 1.04–1.21 P = 0.002; HR 1.10, 95%CI 1.03–1.16 P = 0.002) but not LVEF (P = 0.406; 0.209) was an independent predictor for the primary and secondary endpoint. GLS was an independent predictor for the primary endpoint after correction for commonly considered risk factors including, age, NYHA class, tissue characterization native T1, biomarker NT-proBNP, and glomerular filtration rate (HR 1.09, 95%CI 1.03–1.15, P = 0.003). After dichotomization at the median of −16.4%, GLS added incremental value to risk stratification for the primary endpoint on Kaplan-Meier plots in patients with LVEF above (P = 0.045) and below (P = 0.017) the median of 55%.
Conclusion
In an all-comers, low-risk patient population, GLS emerged as an independent risk predictor with incremental prognostic value relative to LVEF. This finding may support clinical routine implementation of GLS in cardiac in- and outpatient clinics.
Keywords: strain, GLS, outcome, incremental value, all-comers
Graphical Abstract
Graphical Abstract.
Introduction
The burden of cardiovascular diseases (CVD) remains high1 despite all progress made in the field of ischaemic2 and non-ischaemic heart disease.3 Optimized risk prediction in CVD is a highly debated topic. Despite being continuously challenged as a prognostic indicator, left ventricular ejection fraction (LVEF) remains the clinical go-to standard and is commonly used for heart failure (HF) classification.4 Fibrosis has been recognized as an additional risk factor in considering implantable cardioverter defibrillator (ICD) implantation in dilated cardiomyopathy with LVEF >35%. In line with these findings, the ongoing cardiovascular magnetic resonance (CMR)-ICD trial (NCT04558723) was designed to investigate the role of fibrosis in dilated cardiomyopathy with LVEF <35%. Furthermore, fibrosis has long been implicated as a substrate for ventricular arrhythmias in ischaemic heart disease.5
Both simple echocardiographic and CMR-derived deformation imaging techniques have already been shown to have incremental prognostic value in the full range of specific CVD,6–9 yet clinical routine implementation remains rare. Vendor-dependent reference ranges10 or potential differing prognostic impacts across the range of CVD may hinder or delay routine clinical application. Although CMR has proven to have incremental value in high-risk patients following acute myocardial infarction,11 advanced heart failure,12 and even advanced renal disease,6 the majority of patients, especially in outpatient clinics, are at lower risk. The additional value of deformation imaging in an unselected, low-risk setting remains unknown and may therefore be underestimated, which in turn may prompt the use of simple LVEF approximation for overall risk evaluation. Consequently, we sought to investigate the role of deformation imaging in an all-comers, low-risk patient population to investigate the prognostic benefit of using this technique in a real-world, routine imaging setting.
Methods
Study population
Patients referred to the imaging unit of the Kerckhoff Heart and Thorax Centre in Bad Nauheim, Germany, were prospectively recruited to the all-comers CMR/biobank (BioCVI) registry between April 2017 and May 2023 (Figure 1). Clinical indications for CMR comprised assessment of myocardial function and inflammation, ischaemia, and viability testing as well as evaluation of the cardiomyopathy. A follow-up assessment was conducted using a standardized questionnaire one year after recruitment. The registry was approved by the local ethics committee of the University of Giessen and complied with the principles of the Helsinki Declaration. All patients gave written informed consent before participation.
Figure 1.
Study flow chart.
Cardiovascular magnetic resonance
CMR was conducted on a clinical 3.0 Tesla Skyra (Siemens Healthineers, Erlangen, Germany) using a 18-array coil in head-first supine position. Post-processing analyses were performed using commercially available software solutions (CVI42, Calgary, Canada). The clinical protocol followed Society of Cardiovascular Magnetic Resonance guidelines13 and included balanced steady-state free precession (bSSFP) cine sequences in 2/3- and 4-chamber views (Ch), long-axis (LAX) orientations, and a short-axis stack (SAX) covering both ventricles. Right and left ventricular (RV/LV) volumetric analyses were conducted on the SAX stack, including end-diastolic/systolic and stroke volumes (EDV/ESV/SV) and EF as well as LV mass excluding trabecularization. LV global longitudinal strain (GLS) was assessed on all LAX orientations and RV GLS on 4 Ch only. LV global circumferential strain was assessed on SAX orientations14,15 (Figure 2). Tissue characterization was accomplished by native T1 (MOdified Look Locker Inversion recovery sequences 3(2)3(2)5)) (Goethe CVI approaches®) and T2 mapping based on 5-into-3 planning as well as late gadolinium enhancement (LGE) covering the entire SAX. Native T1 and T2 values were taken in the centre of the basal or midventricular septum by means of a region of interest covering at least two voxels that avoided partial volume effects. Inversion recovery-segmented gradient echo sequences were used for LGE assessment 10–15 min after contrast media application (Dotarem®, Guerbet, Villepinte, France; 0.15 mmol/kg bodyweight).
Figure 2.
Cardiovascular magnetic resonance feature-tracking. Assessment of left ventricular global longitudinal strain (GLS) on 4/3/2 chamber view (Ch) long axis orientations as well as right ventricular GLS on 4 Ch respectively. Short axis views (SAX) were used for global circumferential strain assessment as well as all volumetric analyses.
Clinical endpoints
The primary clinical endpoint was defined as all-cause mortality and heart failure hospitalization according to the AHA key data elements and definitions for cardiovascular endpoints in clinical trials.16 The secondary endpoint comprised all-cause mortality and hospitalized but non-fatal myocardial infarction, stroke, or unstable angina pectoris.
Statistics
Categorical parameters are presented as frequencies with corresponding percentage and were compared using the chi-squared test. Continuous parameters are reported as medians with interquartile ranges (IQR). Independent continuous parameters were compared using the Mann-Whitney U test. Correlations were evaluated using the Spearman’s rank correlation coefficient. Outcome analyses were based on 1) univariate and multivariable Cox proportional hazards models using continuous variables reporting hazard ratios (HR) with corresponding 95% confidence intervals (95%CI) and 2) Kaplan-Meier plots with associated log-rank testing. A two-tailed P-value <0.05 was considered to be statistically significant. Analyses were performed with SPSS version 29.0.2.0 (IBM, Armonk, New York, USA) and MedCalc version 23.1.3 (MedCalc Software bvba, Ostend, Belgium).
Results
Study population
The final population consisted of n = 1655 patients classified as normal (n = 454), chronic coronary syndrome (n = 314), ischaemic heart disease (n = 279), dilated (n = 236), hypertrophic (n = 37) or other (n = 22) cardiomyopathies, hypertensive heart disease (n = 28), storage disorders (n = 8), inflammation (n = 103), acute myocardial injury (n = 11), right heart failure (n = 19), and others (n = 144). A total of n = 68 primary and n = 107 secondary endpoints were met during a median follow-up of 395 days (394, 460).
Table 1 reports the baseline characteristics of the study population. Compared to the cohort overall, patients who met the primary or secondary endpoint were, respectively, 9 or 10 years older (P < 0.001), suffered from a higher degree of dyspnoea according to the NYHA classification (P < 0.001), and showed higher N-terminal fragment of pro-brain natriuretic peptide (NT-proBNP) as well as lower estimated glomerular filtration rate (eGFR) on laboratory testing (P = 0.001 to <0.001). Whilst the presence of coronary artery disease did not differ between patients without and with event during follow-up (P = 0.223 and 0.106, respectively), previous myocardial infarction and interventional or coronary artery bypass graft therapy were more common among those reaching the secondary endpoint (P < 0.001).
Table 1.
Baseline characteristics
| Variable | All patients n = 1655 |
No endpoint n = 1518 |
Primary endpoint n = 68 |
Secondary endpoint n = 107 |
P-value |
|---|---|---|---|---|---|
| Age | 61 (51, 71) | 60 (50, 70) | 69 (59, 77) | 70 (60, 77) | <0.001/<0.001 |
| Sex (m) | 589/1655 (35.6) | 547/1518 | 22/68 (32.4) | 33/107 (30.8) | 0.536/0.278 |
| NYHA class | I 133/1655 (8.0) | I 124 (8.2) | I 7/68 (10.3) | I 5 (4.6) | <0.001/<0.001 |
| II 325/1655 (19.6) | II 301 (19.8) | II 13/68 (19.1) | II 19 (17.6) | ||
| III 287/1655 (17.3) | III 246 (16.2) | III 24/68 (35.3) | III 31 (29.0) | ||
| IV 139/1655 (8.4) | IV 114 (7.5) | IV 14/68 (20.6) | IV 19 (17.6) | ||
| Cardiovascular risk factors | |||||
| Active smoking | 356/1649 (21.5) | 324/1512 (21.3) | 13/68 (19.1) | 24/107 (22.4) | 0.199/0.530 |
| Hypertension | 1020/1654 (61.6) | 916/1517 (60.3) | 52/68 (76.5) | 80/107 (74.8) | 0.028/0.012 |
| Hyperlipoproteinemia | 789/1646 (47.7) | 706/1509 (46.5) | 40/68 (58.8) | 72/107 (67.3) | 0.122/<0.001 |
| Diabetes | 308/1650 (18.6) | 270/1513 (17.8) | 25/68 (36.8) | 28/107 (26.2) | <0.001/0.083 |
| Family history | 571/1621 (34.5) | 527/1489 (34.7) | 16/67 (23.5) | 37/104 (34.6) | 0.148/0.429 |
| Body mass index | 26.8 (24.1, 30.4) | 26.8 (24.1, 30.4) | 26.6 (24.1, 31.0) | 26.2 (24.2, 29.4) | 0.970/0.373 |
| History of coronary heart disease | |||||
| Coronary heart disease | 613/1646 (37.0) | 531/1511 (35.1) | 41/67 (60.3) | 70/105 (65.4) | 0.223/0.106 |
| sclerosis | 87 | 78 | 8 | 7 | |
| 1-vessel | 135 | 122 | 7 | 11 | |
| 2-vessel | 128 | 115 | 3 | 11 | |
| 3-vessel | 202 | 166 | 15 | 31 | |
| unspecified | 61 | 50 | 7 | 10 | |
| Previous infarction | 326/1647 (19.7) | 283/1512 (18.6) | 20/68 (29.4) | 33/105 (30.8) | 0.078/<0.001 |
| Previous PTCA | 392/1654 (23.7) | 335/1517 (22.1) | 25/68 (36.8) | 49/107 (45.8) | 0.018/<0.001 |
| Previous CABG | 99/1655 (6.0) | 79/1518 (5.2) | 7/68 (10.3) | 18/107 (16.8) | 0.070/<0.001 |
| Sudden cardiac death | 77/1642 (4.7) | 67/1507 (4.4) | 5/67 (7.4) | 8/106 (7.5) | 0.224/0.036 |
| Atrial fibrillation | 361/1647 (21.8) | 310/1511 (20.4) | 26/68 (38.2) | 42/106 (39.3) | 0.002/<0.001 |
| Systolic blood pressure | 125 (115, 137) | 125 (116, 138) | 120 (110, 136) | 127 (117, 138) | 0.407/0.854 |
| Diastolic blood pressure | 79 (70, 83) | 80 (70, 84) | 70 (63, 80) | 72 (64, 80) | 0.016/0.057 |
| Laboratory Testing | |||||
| NTproBNP | 247 (82, 887) | 233 (80, 794) | 1736 (588, 4506) | 555 (101, 1652) | <0.001/0.001 |
| eGFR | 91 (72, 109) | 92 (74, 110) | 67 (46, 95) | 80 (57, 99) | <0.001/<0.001 |
| Medication | |||||
| Beta-blocker | 848/1652 (51.2) | 758/1515 (49.9) | 45/68 (66.2) | 69/107 (64.5) | 0.031/0.014 |
| ACE inhibitor | 459/1650 (27.7) | 414/1514 (22.9) | 23/68 (33.8) | 36/107 (33.6) | 0.452/0.313 |
| AT1 antagonist | 385/1651 (23.3) | 348/1514 (22.9) | 13/68 (19.1) | 30/107 (28.0) | 0.693/0.425 |
| Sacubitril/valsartan | 63/1245 (3.8) | 54/1146 (4.7) | 5/49 (10.2) | 6/85 (5.6) | <0.001/0.490 |
| Aldosterone antagonist | 234/1652 (14.1) | 198/1515 (13.0) | 23/68 (33.8) | 22/107 (20.6) | <0.001/0.082 |
| Ca channel blocker | 255/1652 (15.4) | 226/1515 (14.9) | 14/68 (20.6) | 22/107 (20.6) | 0.414/0.263 |
| Loop diuretic | 358/1652 (21.6) | 294/1515 (19.4) | 41/68 (60.3) | 45/107 (42.1) | <0.001/<0.001 |
| Statin | 636/1652 (38.4) | 567/1515 (37.4) | 33/68 (48.5) | 60/107 (56.1) | 0.169/<0.001 |
| Aspirin | 559/1651 (33.8) | 501/1515 (33.0) | 25/67 (36.8) | 49/106 (45.8) | 0.095/0.007 |
| Malignancy <5 years | 100/1649 (6.0) | 88/1514 (5.8) | 9/68 (13.2) | 9/105 (8.4) | 0.042/<0.001 |
Data presented as n/N (%) or median (IQR). P-values were calculated for primary/secondary endpoint compared to no endpoint met. NT-proBNP in pg/mL.
NYHA, New York Heart Association; PTCA, Percutaneous transluminal coronary angioplasty; CABG, coronary artery bypass graft; eGFR, estimated glomerular filtration rate (mL/min/1.73 m2); ACE, angiotensin converting enzyme; AT1, angiotensin II receptor type 1; Ca, Calcium.
Cardiac functional evaluation
Table 2 shows echocardiographic and CMR-derived functional characteristics. Cardiac valvular functional assessment by echocardiography was performed in 945 to 949 cases, depending on the valve. The severity of aortic valve stenosis was greater in patients who met either the primary or secondary endpoint (P < 0.001), although there was no difference in the severity of regurgitation. Mitral and tricuspid function was impaired in the primary endpoint group only (P = 0.004 to <0.001).
Table 2.
Functional assessments
| Variable | All patients n = 1655 |
No endpoint n = 1518 |
Primary endpoint n = 68 |
Secondary endpoint n = 107 |
P-value |
|---|---|---|---|---|---|
| Echocardiography | |||||
| Aortic valve stenosis | 0 822 | 0 748 | 0 39 | 0 54 | <0.001/<0.001 |
| I 25 | I 23 | I 2 | I 1 | ||
| II 18 | II 11 | II 4 | II 6 | ||
| III 14 | III 12 | III 0 | III 2 | ||
| Aortic valve regurgitation | 0 703 | 0 646 | 0 30 | 0 40 | 0.051/0.066 |
| I 158 | I 135 | I 11 | I 20 | ||
| II 17 | II 14 | II 3 | II 2 | ||
| III 2 | III 2 | III 0 | III 0 | ||
| Mitral valve stenosis | 0 857 | 0 778 | 0 39 | 0 59 | 0.004/0.121 |
| I 10 | I 9 | I 1 | I 1 | ||
| II 5 | II 3 | II 2 | II 2 | ||
| III 3 | III 3 | III 0 | III 0 | ||
| Mitral valve regurgitation | 0 485 | 0 444 | 0 20 | 0 32 | <0.001/0.339 |
| I 357 | I 323 | I 16 | I 25 | ||
| II 36 | II 28 | II 8 | II 4 | ||
| III 9 | III 6 | III 2 | III 2 | ||
| Tricuspid valve regurgitation | 0 517 | 0 474 | 0 23 | 0 31 | <0.001/0.078 |
| I 322 | I 291 | I 13 | I 27 | ||
| II 25 | II 21 | II 4 | II 2 | ||
| III 9 | III 4 | III 5 | III 2 | ||
| Cardiovascular magnetic resonance imaging | |||||
| Left ventricle | |||||
| LV mass | 51 (41, 64) | 51 (41, 63) | 60 (46, 73) | 53 (42, 65) | 0.013/0.201 |
| LV EDV | 82 (69, 101) | 81 (68, 100) | 97 (75, 124) | 83 (69, 106) | <0.001/0.743 |
| LV ESV | 36 (27, 52) | 35 (26, 51) | 49 (33, 93) | 38 (27, 54) | <0.001/0.253 |
| LV SV | 43 (36, 50) | 43 (36, 50) | 39 (28, 47) | 41 (32, 49) | <0.0010.068 |
| LVEF | 55 (43, 63) | 56 (44, 63) | 41 (24, 55) | 52 (39, 62) | <0.001/0.060 |
| FT LV GLS | −16.4 (−12.4, −19.3) | −16.6 (−12.6, −19.5) | −11.8 (−7.0, −15.5) | −14.8 (−11.6, −18.3) | <0.001/0.002 |
| FT LV GCS | −17.6 (−13.1, −20.9) | −17.7 (−13.4, −21.0) | −12.5 (−7.8, −17.0) | −16.1 (−11.8, −20.1) | <0.001/0.026 |
| Right ventricle | |||||
| RV EDV | 78 (65, 93) | 78 (65, 93) | 81 (65, 106) | 78 (65, 95) | 0.159/0.827 |
| RV ESV | 38 (29, 50) | 38 (29, 50) | 45 (35, 60) | 41 (30, 52) | 0.002/0.234 |
| RV SV | 39 (31, 46) | 39 (32, 46) | 34 (25, 45) | 36 (30, 44) | 0.021/0.072 |
| RVEF | 50 (47´4, 57) | 51 (44, 57) | 45 (38, 53) | 49 (42, 56) | <0.001/0.086 |
| FT RV GLSa | −23.9 (−19.1, −27.6) | −24.0 (−19.4, −27.6) | −17.6 (−13.4, −23.6) | −23.2 (−16.9, −27.8) | <0.001/0.363 |
| Tissue characterization | |||||
| Native T1 | 1126 (1090, 1169) | 1125 (1089, 1167) | 1168 (1119, 1238) | 1131 (1096, 1188) | <0.001/0.053 |
Data presented as n/N (%) or median (IQR). P-values were calculated for primary/secondary endpoint compared to no endpoint met.
LV, left ventricular; RV, right ventricular; EDV, end-diastolic volume; ESV, end-systolic volume; SV, stroke volume; FT, Feature-Tracking; LVEF, left ventricular ejection fraction; GLS/GCS, global longitudinal/circumferential strain.
a n = 1074/983/71/38. Volumes are given in mL/m2 body surface area, EF and strain in % and T1 in ms.
In patients who met the primary endpoint, LV EDV and ESV were higher with consequently lower SV and EF (P < 0.001), whilst no volumetric changes were observed in patients who met the secondary endpoint. In contrast, both GLS and global circumferential strain were impaired in the primary (P < 0.001) and secondary (P < 0.001/0.032) endpoint groups. Volumetric and deformation-based RV function was impaired in patients who reached the primary (P < 0.001) but not secondary endpoint.
Outcome
Table 3 reports uni- and multivariable Cox regression models for the primary and secondary endpoint. Whilst GLS was associated with both the primary (HR 1.15, 95%CI 1.10–1.21, P < 0.001) and secondary endpoints (HR 1.06, 95%CI 1.03–1.11, P = 0.001), LVEF was only significantly associated with the primary endpoint (HR 0.96, 95%CI 0.95–0.97, P < 0.001). GLS but not LVEF was an independent predictor for the primary endpoint (GLS HR 1.12, 95%CI 1.04–1.21, P = 0.002 vs. LVEF HR 0.99, 95%CI 0.97–1.01, P = 0.406). GLS emerged as an independent risk predictor for the primary endpoint (HR 1.09, 95%CI 1.03–1.15, P = 0.003) in a multivariable model based on clinical factors (age, NYHA classification), laboratory testing (eGFR and NTproBNP), and CMR tissue characterization (native T1). LV GLS dichotomized at the median (16.4%) added incremental value for risk prediction of the primary endpoint in patients with an LVEF below (P = 0.017) and above (P = 0.045) 55% dichotomized at the median (Figure 3). In a model accounting for RV function for primary endpoint prediction which included the same clinical (age, NYHA) and laboratory (eGFR and NTproBNP) parameters in addition to LV GLS and RV EF, NYHA (HR 1.44, P < 0.001), eGFR (HR 0.99, P = 0.014) and GLS (HR 1.09, P = 0.012) but not RV EF (HR 0.99, P = 0.404) emerged statistically significant. Considering the same model, neither LV GLS (P = 0.623) nor RV EF (P = 0.287) predicted the secondary endpoint.
Table 3.
Risk assessment
| Variable | Univariate primary endpoint | Multivariable primary endpoint | Univariate secondary endpoint | Multivariable secondary endpoint |
|---|---|---|---|---|
| Clinical factors | ||||
| Age | 1.04 (1.02–1.06) P < 0.001 | 1.01 (0.98–1.03) P = 0.656 | 1.05 (1.03–1.07) P < 0.001 | 1.04 (1.02–1.06) P < 0.001 |
| NYHA class | 1.63 (1.38–1.94) P < 0.001 | 1.36 (1.11–1.67) P = 0.003 | 1.39 (1.22–1.60) P < 0.001 | 1.31 (1.12–1.54) P < 0.001 |
| Laboratory testing | ||||
| eGFR | 0.98 (0.97–0.99) P < 0.001 | 0.99 (0.98–1.00) P = 0.046 | 0.99 (0.98–0.99) P < 0.001 | 1.00 (0.99–1.01) P = 0.477 |
| NT-proBNP | 1.00 (1.00, 1.00) P < 0.001 | 1.00 (1.00–1.00) P = 0.455 | 1.00 (1.00, 1.00) P = 0.002 | 1.00 (1.00–1.00) P = 0.659 |
| Echocardiography | ||||
| Aortic valve stenosis | 1.00 (0.99–1.02) P = 0.553 | 1.00 (0.99, 1.01) P = 0.890 | ||
| Mitral valve regurgitation | 1.00 (0.99–1.02) P = 0.445 | 1.00 (0.99, 1.01) P = 0.931 | ||
| Cardiovascular magnetic resonance | ||||
| Native T1 | 1.01 (1.01–1.01) P < 0.001 | 1.01 (1.00–1.01) P = 0.001 | 1.01 (1.00–1.01) P = 0.001 | 1.00 (1.00–1.01) P = 0.100 |
| LVEF | 0.96 (0.95–0.97) P < 0.001 | 0.99 (0.98–1.00) P = 0.089 | ||
| LV GLS | 1.15 (1.10–1.21) P < 0.001 | 1.09 (1.03–1.15) P = 0.003 | 1.06 (1.03–1.11) P = 0.001 | 0.99 (0.95–1.04) P = 0.747 |
| LV GCS | 1.12 (1.07–1.17) P < 0.001 | 1.03 (1.00–1.07) P = 0.064 | ||
| RVEF | 0.96 (0.94–0.98) P < 0.001 | 0.98 (0.97–1.00) P = 0.045 | ||
| RV GLSa | 1.05 (1.02–1.08) P < 0.001 | 1.02 (0.99–1.05) P = 0.137 | ||
The data is presented as hazard ratios with associated 95% confidence intervals in parentheses. Variables with univariate significance (P < 0.05) were included in multivariable Cox regression models and are presented if they emerged as statistically significant (P < 0.05). Due to collinearity of functional parameters, only GLS was considered as a CMR-derived functional parameter in multivariable models.
NYHA, New York Heart Association; eGFR, estimated glomerular filtration rate; LV EF/GLS/GCS, left ventricular ejection fraction/global longitudinal/circumferential strain.
aRV GLS assessment only available in n = 1074 (n = 37/60 primary/secondary endpoints). RV function was not considered in this specific multivariate analysis.
Figure 3.
Kaplan-Meier plot—primary endpoint. (A) The graph shows the percentage of patients with LVEF above the median of 55% who met the primary endpoint dichotomized according to the median of GLS. (B) The graph shows the percentage of patients with LVEF below the median of 55% who met the primary endpoint dichotomized according to the median of GLS.
Discussion
The results of this monocentric longitudinal registry demonstrate the high prognostic yield of GLS in an unselected, real-world in- and outpatient clinical setting. GLS but not LVEF was found to be a predictor for both the primary and secondary endpoints. GLS was a predictor for the primary and secondary endpoints independent of LVEF and provided incremental prognostic value in patients with LVEF either below or above the median of 55% for primary endpoint prediction. Finally, GLS remained independent in a multivariable model including clinical, laboratory, and CMR-derived information for primary endpoint prediction.
LVEF is commonly used for HF classification4 and is the most commonly established parameter for risk assessment.17 Park et al.18 demonstrated the incremental value of GLS in patients classified as having HF with reduced, mildly reduced, or preserved EF. Nevertheless, multiple limitations for general clinical practice warrant further consideration. Whilst an acute HF population with a 5-year mortality rate of 40%18 is in line with data from the ESC guidelines,4 this may not be a representative of the day-to-day population of in- and outpatient cardiology clinics. Thus, application of the results presented by Park et al.18 especially to general outpatient clinics with lower risk profiles and fewer events, may be limited. The present population consisted of 1655 patients across the full spectrum of cardiovascular diseases, both ischaemic and non-ischaemic in nature, including cardiomyopathies, rare storage diseases, myocarditis, and chronic coronary syndrome. Furthermore, there was a distinct percentage of patients with no abnormal findings, as may be expected from a specialized outpatient clinic based on referrals from external cardiologists and internal medicine specialists. As a consequence, this population may better reflect the general cardiological population: only about 4.1% primary and 6.4% secondary endpoints were met by patients during the 1-year timeframe. Despite this low frequency, GLS emerged as a highly significant outcome predictor, which underlines its clinical significance across the full spectrum of CVD.
As all patients presented with acute heart failure,18 the vast range of EF values across these patients intrinsically demonstrates the limited value of EF for HF hospitalization prediction. Indeed, the presence of HF with preserved EF (HFpEF) itself limits EF as the sole risk assessment tool.19,20 In addition, different subtypes of HFpEF have been described based on the range of LVEF that differentiate between HFpEF patients with LVEF from 50% to 60% and >60%21: The former group demonstrated impaired contractility and ventriculo-arterial coupling on invasive catheterization in conjunction with ventricular fibrotic remodelling, whereas the latter was associated with increased contractility but higher diastolic and systolic stiffness despite less fibrotic remodelling.21 Indeed, CVD may not be adequately described by volumetric assessments alone, neither by overall cardiac volume nor by EF. As a result, the recent ESC guideline on cardiomyopathies has introduced the non-dilated LV cardiomyopathy subtype.3 Whilst in the present population both GLS and EF predict the primary endpoint of all-cause mortality and HF hospitalization, only GLS emerged as an independent predictor. Furthermore, when dichotomized at the median, GLS provided incremental prognostic value in patients both below and above the LVEF median. These findings underline the significance of GLS for detailed, prognosis-relevant functional characterization. Beyond GLS, LA strain has demonstrated incremental diagnostic19 and prognostic value22–24 across the spectrum of HFrEF and HFpEF, even in addition to GLS.25 On the one hand, LA function may reflect LV functional deterioration and congestion, on the other hand, it may add specific value stemming from intrinsic atrial dysfunction. This may underline the future role of LA functional analyses even compared to GLS which warrants further clinical trials. Notwithstanding, GLS alone has demonstrated prognostic value both in HFpEF and HFrEF.18
The broad spectrum of CVD calls for more than one parameter for detailed patient classification and risk assessment. In that regard, patient characteristics such as age and symptom severity (NYHA class), laboratory findings such as eGFR and NT-proBNP, and the presence of additional valvular dysfunction and tissue remodelling such as T1 should all be taken into account for the evaluation of CVD. However, GLS emerged as a predictor for the primary endpoint independent of the aforementioned characteristics. Most hospitalizations of HF patients (63%) do not qualify as HF hospitalizations as reported by the ESC heart failure guideline.4 Indeed, GLS was the only functional parameter with prognostic significance in univariate Cox regression analysis for the secondary endpoint of all-cause mortality and hospitalized angina, myocardial infarction, and stroke, although significance was not retained on the multivariable clinical model. Consequently, GLS emerges as the best functional parameter, but further clinical risk assessment is warranted for overall patient classification. Although RV function, both volumetric and deformation imaging derived, held prognostic impact on univariate analyses for primary endpoint prediction, multivariate analyses considering the same clinical and laboratory baseline parameters in addition to LV GLS and RV EF continuously supported the role of LV GLS.
Cardiac deformation assessment based either on echocardiography or CMR imaging remains widely underused. Although the availability of CMR imaging may be limited to tertiary centres, this is not the case for echocardiography, and thus limited availability is not the main reason for limited application. Vendor-specific differences in reference ranges10,26 do impact overall applicability27; a reference population may be established for about 30 patients, as is the case for mapping in CMR according to general recommendations. Data from speckle-tracking echocardiography28 as well as CMR-FT using CVI 4229,30 show that roughly −16% GLS is the lower limit of normal, which would be in line with the present population showing a median value of −16.4% for myocardial strain. Thus, this value may be used as a landmark for overall interpretation of GLS as it provided incremental prognostic value. Finally, whilst a body of evidence exists for specific CVD such as ischaemic heart disease following acute myocardial infarction,8,11 non-ischaemic dilated cardiomyopathy,31 and even chronic kidney disease,6 the present findings underline the applicability of GLS in patients presenting for the first time with yet unknown specific underlying cardiovascular disease. GLS may serve as a rather universal risk assessment tool upon initial patient contact.
Limitations
The underlying data and conclusions are derived from an observational registry with its associated general limitations. Whilst both in- and outpatients were prospectively recruited to the registry, this poses an inherent limitation in terms of selection bias, as inpatients unfit for CMR were not included. The low number of endpoints met impacts statistical power; however, multivariable models were not overfitted, and the low number of events may be due to a population that is more homogeneous than that seen in an unselected outpatient setting on a daily basis. RV GLS was not available for all patients which may have resulted in underestimated statistical power and clinical significance.
Conclusion
In an all-comers, low-risk patient population, GLS emerged as an independent risk predictor compared to LVEF as well as established clinical parameters associated with poor outcome. In addition, GLS adds prognostic value to that of LVEF evaluation. As a result, these findings promote clinical routine implementation of GLS for all patients irrespective of their cardiovascular background.
Acknowledgements
We thank Elizabeth Martinson, PhD, of the KHFI Editorial Office for editing the manuscript.
Contributor Information
Sören J Backhaus, Department of Cardiology, Campus Kerckhoff of the Justus-Liebig-University Giessen, Kerckhoff-Heart-Centre, Benekestr. 2-8, 61231 Bad Nauheim, Germany; German Center for Cardiovascular Research (DZHK), Partner Site Rhine-Main, Bad Nauheim, Germany; Department of Cardiology and Angiology, Medical Clinic I, University Hospital Giessen, Justus-Liebig-University Giessen, Klinikstr. 33, 35392 Giessen, Germany.
Julia M Treiber, Department of Cardiology, Campus Kerckhoff of the Justus-Liebig-University Giessen, Kerckhoff-Heart-Centre, Benekestr. 2-8, 61231 Bad Nauheim, Germany; German Center for Cardiovascular Research (DZHK), Partner Site Rhine-Main, Bad Nauheim, Germany.
Jan Sebastian Wolter, Department of Cardiology, Campus Kerckhoff of the Justus-Liebig-University Giessen, Kerckhoff-Heart-Centre, Benekestr. 2-8, 61231 Bad Nauheim, Germany; German Center for Cardiovascular Research (DZHK), Partner Site Rhine-Main, Bad Nauheim, Germany.
Steffen D Kriechbaum, Department of Cardiology, Campus Kerckhoff of the Justus-Liebig-University Giessen, Kerckhoff-Heart-Centre, Benekestr. 2-8, 61231 Bad Nauheim, Germany; German Center for Cardiovascular Research (DZHK), Partner Site Rhine-Main, Bad Nauheim, Germany.
David Saamann, Department of Cardiology, Campus Kerckhoff of the Justus-Liebig-University Giessen, Kerckhoff-Heart-Centre, Benekestr. 2-8, 61231 Bad Nauheim, Germany; German Center for Cardiovascular Research (DZHK), Partner Site Rhine-Main, Bad Nauheim, Germany.
Andreas Schuster, Department of Cardiology and Pneumology, University Medical Center Göttingen, Georg-August University, Robert-Koch-Str. 30, 37075 Göttingen, Germany; German Center for Cardiovascular Research (DZHK), Partner Site Lower Saxony, Göttingen, Germany; FORUM Cardiology, An d. Ziegelei 1, 37124 Rosdorf, Germany.
Samuel Sossalla, Department of Cardiology, Campus Kerckhoff of the Justus-Liebig-University Giessen, Kerckhoff-Heart-Centre, Benekestr. 2-8, 61231 Bad Nauheim, Germany; German Center for Cardiovascular Research (DZHK), Partner Site Rhine-Main, Bad Nauheim, Germany; Department of Cardiology and Angiology, Medical Clinic I, University Hospital Giessen, Justus-Liebig-University Giessen, Klinikstr. 33, 35392 Giessen, Germany.
Andreas Rolf, Department of Cardiology, Campus Kerckhoff of the Justus-Liebig-University Giessen, Kerckhoff-Heart-Centre, Benekestr. 2-8, 61231 Bad Nauheim, Germany; German Center for Cardiovascular Research (DZHK), Partner Site Rhine-Main, Bad Nauheim, Germany; Department of Cardiology and Angiology, Medical Clinic I, University Hospital Giessen, Justus-Liebig-University Giessen, Klinikstr. 33, 35392 Giessen, Germany.
Author contributions
SB and AR designed the study protocol, performed data acquisition, performed statistical analyses and drafted the manuscript. JT was involved in conceptualization. JSW, SK and DS performed data acquisition/analysis. All authors revised the manuscript and participated in the scientific discussion during the study. All authors read and approved the final manuscript.
Consent
Not applicable.
Funding
The Kerckhoff Biomarker Registry (Bioreg) is funded by the Kerckhoff Heart Research Institute (KHFI) and the German Center for Cardiovascular Research (DZHK).
Ethics approval and consent to participate
The study was approved by the Ethics Committee of the University Hospital Giessen and complied with the Declaration of Helsinki. All individuals gave written informed consent before participating in the study.
Patient and public involvement
Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.
Data availability
Regarding data availability, we confirm that all relevant data are within the paper and all data underlying the findings are fully available without restriction and can be accessed at the Kerckhoff Heart Research Institute (KHFI) by researchers who meet the criteria for access to confidential data.
Lead author biography
Dr. Sören J. Backhaus is a consultant cardiologist at Kerckhoff Heart Centre Bad Nauheim, and senior lecturer at Justus-Liebig-University Giessen. He completed his cardiology training at University Medical Centre Göttingen, where he became consultant cardiologist and senior lecturer in Internal Medicine and Cardiology prior to the acceptance of his role in Bad Nauheim. He specializes in multimodality cardiovascular imaging with particular interest in cardiovascular magnetic resonance imaging in heart failure.
References
- 1. Smith SC Jr, Collins A, Ferrari R, Holmes DR Jr, Logstrup S, McGhie DV et al. Our time: a call to save preventable death from cardiovascular disease (heart disease and stroke). Circulation 2012;126:2769–75. [DOI] [PubMed] [Google Scholar]
- 2. Vrints C, Andreotti F, Koskinas KC, Rossello X, Adamo M, Ainslie J et al. 2024 ESC guidelines for the management of chronic coronary syndromes. Eur Heart J 2024;45:3415–537. [DOI] [PubMed] [Google Scholar]
- 3. Arbelo E, Protonotarios A, Gimeno JR, Arbustini E, Barriales-Villa R, Basso C et al. 2023 ESC guidelines for the management of cardiomyopathies. Eur Heart J 2023;44:3503–626. [DOI] [PubMed] [Google Scholar]
- 4. McDonagh TA, Metra M, Adamo M, Gardner RS, Baumbach A, Böhm M et al. 2021 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J 2021;42:3599–726. [DOI] [PubMed] [Google Scholar]
- 5. Zeppenfeld K, Tfelt-Hansen J, de Riva M, Winkel BG, Behr ER, Blom NA et al. 2022 ESC guidelines for the management of patients with ventricular arrhythmias and the prevention of sudden cardiac death. Eur Heart J 2022;43:3997–4126. [DOI] [PubMed] [Google Scholar]
- 6. Krishnasamy R, Isbel NM, Hawley CM, Pascoe EM, Burrage M, Leano R et al. Left ventricular global longitudinal strain (GLS) is a superior predictor of all-cause and cardiovascular mortality when compared to ejection fraction in advanced chronic kidney disease. PLoS One 2015;10:e0127044. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Schuster A, Backhaus SJ, Stiermaier T, Eitel I. Prognostic utility of global longitudinal strain in myocardial infarction. World J Cardiol 2018;10:35–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Ersbøll M, Valeur N, Mogensen UM, Andersen MJ, Møller JE, Velazquez EJ et al. Prediction of all-cause mortality and heart failure admissions from global left ventricular longitudinal strain in patients with acute myocardial infarction and preserved left ventricular ejection fraction. J Am Coll Cardiol 2013;61:2365–73. [DOI] [PubMed] [Google Scholar]
- 9. Marwick TH, Shah SJ, Thomas JD. Myocardial strain in the assessment of patients with heart failure: a review. JAMA Cardiol 2019;4:287–94. [DOI] [PubMed] [Google Scholar]
- 10. Gertz RJ, Lange T, Kowallick JT, Backhaus SJ, Steinmetz M, Staab W et al. Inter-vendor reproducibility of left and right ventricular cardiovascular magnetic resonance myocardial feature-tracking. PLoS One 2018;13:e0193746. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Eitel I, Stiermaier T, Lange T, Rommel K-P, Koschalka A, Kowallick JT et al. Cardiac magnetic resonance myocardial feature tracking for optimized prediction of cardiovascular events following myocardial infarction. JACC Cardiovasc Imaging 2018;11:1433–44. [DOI] [PubMed] [Google Scholar]
- 12. 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. [DOI] [PubMed] [Google Scholar]
- 13. Messroghli DR, Moon JC, Ferreira VM, Grosse-Wortmann L, He T, Kellman P et al. Clinical recommendations for cardiovascular magnetic resonance mapping of T1, T2, T2* and extracellular volume: a consensus statement by the Society for Cardiovascular Magnetic Resonance (SCMR) endorsed by the European Association for Cardiovascular Imaging (EACVI). J Cardiovasc Magn Reson 2017;19:75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Backhaus SJ, Metschies G, Billing M, Kowallick JT, Gertz RJ, Lapinskas T et al. Cardiovascular magnetic resonance imaging feature tracking: impact of training on observer performance and reproducibility. PLoS One 2019;14:e0210127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Backhaus SJ, Metschies G, Billing M, Schmidt-Rimpler J, Kowallick JT, Gertz RJ et al. Defining the optimal temporal and spatial resolution for cardiovascular magnetic resonance imaging feature tracking. J Cardiovasc Magn Reson 2021;23:60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Hicks KA, Tcheng JE, Bozkurt B, Chaitman BR, Cutlip DE, Farb A et al. 2014 ACC/AHA key data elements and definitions for cardiovascular endpoint events in clinical trials: a report of the American College of Cardiology/American Heart Association task force on clinical data standards (writing committee to develop cardiovascular endpoints data standards). J Am Coll Cardiol 2015;66:403–69. [DOI] [PubMed] [Google Scholar]
- 17. White HD, Norris RM, Brown MA, Brandt PW, Whitlock RM, Wild CJ. Left ventricular end-systolic volume as the major determinant of survival after recovery from myocardial infarction. Circulation 1987;76:44–51. [DOI] [PubMed] [Google Scholar]
- 18. Park JJ, Park J-B, Park J-H, Cho G-Y. Global longitudinal strain to predict mortality in patients with acute heart failure. J Am Coll Cardiol 2018;71:1947–57. [DOI] [PubMed] [Google Scholar]
- 19. Backhaus SJ, Lange T, George EF, Hellenkamp K, Gertz RJ, Billing M et al. Exercise stress real-time cardiac magnetic resonance imaging for noninvasive characterization of heart failure with preserved ejection fraction: the HFpEF-stress trial. Circulation 2021;143:1484–98. [DOI] [PubMed] [Google Scholar]
- 20. Schuster A, Schulz A, Lange T, Evertz R, Hartmann F, Kowallick JT et al. Concomitant latent pulmonary vascular disease leads to impaired global cardiac performance in heart failure with preserved ejection fraction. Eur J Heart Fail 2023;25:322–31. [DOI] [PubMed] [Google Scholar]
- 21. Rosch S, Kresoja K-P, Besler C, Fengler K, Schöber AR, von Roeder M et al. Characteristics of heart failure with preserved ejection fraction across the range of left ventricular ejection fraction. Circulation 2022;146:506–18. [DOI] [PubMed] [Google Scholar]
- 22. Backhaus SJ, Stiermaier T, Lange T, Chiribiri A, Uhlig J, Freund A et al. Atrial mechanics and their prognostic impact in Takotsubo syndrome: a cardiovascular magnetic resonance imaging study. Eur Heart J Cardiovasc Imaging 2019;20:1059–69. [DOI] [PubMed] [Google Scholar]
- 23. Backhaus SJ, Schulz A, Lange T, Schmidt-Schweda LS, Evertz R, Kowallick J et al. Real-time cardiovascular magnetic resonance imaging for non-invasive characterisation of heart failure with preserved ejection fraction: final outcomes of the HFpEF stress trial. Clin Res Cardiol 2024;113:496–508. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Schuster A, Backhaus SJ, Stiermaier T, Navarra JL, Uhlig J, Rommel KP et al. Left atrial function with MRI enables prediction of cardiovascular events after myocardial infarction: insights from the AIDA STEMI and TATORT NSTEMI trials. Radiology 2019;293:292–302. [DOI] [PubMed] [Google Scholar]
- 25. Schmermund BN, Rieth AJ, Rademann M, Borst PC, Kriechbaum S, Wolter JS et al. Abnormal left atrial strain by CMR is associated with left heart disease in patients with pulmonary hypertension. Circ Heart Fail 2025;18:e013480. [DOI] [PubMed] [Google Scholar]
- 26. Backhaus SJ, Metschies G, Zieschang V, Erley J, Mahsa Zamani S, Kowallick JT et al. Head-to-head comparison of cardiovascular MR feature tracking cine versus acquisition-based deformation strain imaging using myocardial tagging and strain encoding. Magn Reson Med 2021;85:357–68. [DOI] [PubMed] [Google Scholar]
- 27. Schuster A, Backhaus SJ, Stiermaier T, Kowallick JT, Stulle A, Koschalka A et al. Fast manual long-axis strain assessment provides optimized cardiovascular event prediction following myocardial infarction. Eur Heart J Cardiovasc Imaging 2019;20:1262–70. [DOI] [PubMed] [Google Scholar]
- 28. Wang Y-H, Sun L, Li S-W, Wang C-F, Pan X-F, Liu Y et al. Left ventricular global longitudinal strain using a novel fully automated method: a head-to-head comparison with a manual layer-specific strain and establishment of normal reference ranges. Int J Cardiol 2024;403:131886. [DOI] [PubMed] [Google Scholar]
- 29. Yang W, Xu J, Zhu L, Zhang Q, Wang Y, Zhao S et al. Myocardial strain measurements derived from MR feature-tracking: influence of sex, age, field strength, and vendor. JACC Cardiovasc Imaging 2024;17:364–79. [DOI] [PubMed] [Google Scholar]
- 30. Lim C, Blaszczyk E, Riazy L, Wiesemann S, Schüler J, von Knobelsdorff-Brenkenhoff F et al. Quantification of myocardial strain assessed by cardiovascular magnetic resonance feature tracking in healthy subjects-influence of segmentation and analysis software. Eur Radiol 2021;31:3962–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Japp AG, Gulati A, Cook SA, Cowie MR, Prasad SK. The diagnosis and evaluation of dilated cardiomyopathy. J Am Coll Cardiol 2016;67:2996–3010. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Regarding data availability, we confirm that all relevant data are within the paper and all data underlying the findings are fully available without restriction and can be accessed at the Kerckhoff Heart Research Institute (KHFI) by researchers who meet the criteria for access to confidential data.