Author + information
- aBritish Heart Foundation Glasgow Cardiovascular Research Centre, Institute of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, United Kingdom
- bGolden Jubilee National Hospital, Clydebank, United Kingdom
- cMenzies Institute for Medical Research, University of Tasmania, Hobart, Australia
- dGoethe Institute for Experimental and Translational Cardiovascular Imaging, Frankfurt, Germany
- ↵∗Address for correspondence:
Prof. Colin Berry, British Heart Foundation Glasgow Cardiovascular Research Centre, 126 University Place, University of Glasgow, Glasgow G12 8TA, United Kingdom.
Secondary prevention of recurrent adverse cardiovascular events (MACE) is the top priority for clinicians treating patients following an acute myocardial infarction (MI). To this end, clinicians adopt the left ventricular ejection fraction (LVEF) as the principal imaging biomarker for prognostication, and LVEF is used to inform decisions for medical and device therapies (1).
Lagrangian strain is a relative measure of myocardial deformation [(L − L0)/L0]. In the context of LV geometry, 3 principal vectors: global longitudinal strain (GLS), global circumferential strain (GCS), and global radial strain (GRS) describe LV deformation mechanics (2). GLS is a superior prognostic biomarker compared with LVEF in the evaluation of ischemic and nonischemic cardiomyopathy (3,4), and GLS, GLS rate, and GCS rate have the same incremental prognostic value of LVEF when evaluated using echocardiographic speckle tracking following acute coronary syndromes (5,6). Feature tracking-derived strain parameters obtained from cardiac magnetic resonance (CMR) are inversely associated with infarct size and LV remodeling post-MI (7). CMR derived GLS also has prognostic value in dilated cardiomyopathy (8). However, to date, information on GLS measured with CMR and health outcomes post-MI have not been reported (Figure 1).
It is in this context in this issue of iJACC, that Eitel et al. (9) evaluated the prognostic value of CMR-derived left ventricular strain parameters in patients following an ACS (both ST-segment elevation and non−ST-segment elevation myocardial infarction [STEMI and NSTEMI]). They evaluated GLS in 1,107 patients and GCS/GRS in 1,085 patients. During 1-year follow-up, 76 patients experienced MACEs (6.9%). LVEF, 3 strain parameters, infarct size, and myocardial salvage index were significantly different between the MACE and non-MACE cohorts. The investigators reported excellent feasibility for strain computation (95%) and reproducibility of all of the strain measures.
The investigators addressed a number of critical questions. First, they demonstrated that GLS had the best prognostic value of the strain parameters (GLS, GCS, and GRS) using multivariate Cox regression. Second, GLS provided incremental value to the standard prognostic model of LVEF (C-statistic: 0.65 to 0.73; p = 0.04). Finally, using a multivariable Cox regression model, the investigators evaluated clinical and imaging biomarkers (LVEF, microvascular obstruction [MVO], GLS) and found independent predictive associations between GLS, age, Killip class on admission, and multivessel coronary artery disease. Perhaps surprising, LVEF and MVO were not independent predictors of MACE.
The NSTEMI population accounted for 44% of MACEs, and GLS and GRS were the only significantly different imaging variables in this population. This is a relevant finding, which highlighted the superiority of GLS over GCS in this population. These findings might be explained by the pathophysiology of NSTEMI, which typically involves subendocardial myocardial infarction, which is the location of the subendocardial longitudinally arrayed fibers. The results suggested that GLS derived from CMR might be useful for identification of higher risk patients post-NSTEMI. The result was all the more relevant considering that LVEF is usually normal in patients post-NSTEMI, which implied that GLS is a more sensitive biomarker of LV dysfunction (10). The relationships among GLS, infarct pathology, and prognosis merit further investigation, not least of which is because NSTEMI is the most common form of acute MI.
Can GLS be useful for risk stratification post-MI? Eitel et al. (9) addressed this question by assessing the prognostic value of GLS in patients stratified by convention to an LVEF that was either significantly reduced (≤35%) or not. The second group (EF >35%) accounts for 53 of 76 MACEs, and based on LVEF alone, this group would not qualify for defibrillator therapy. GLS approached statistical significance (p = 0.05) in dichotomizing this population into high- and low-risk subsets (event rates: 6.9% vs. 4.1%); however, the high number needed to treat would render device therapy prohibitively expensive in either group. With respect to further stratifying the cohort with EF ≤35%, the study was underpowered due to the small (n = 7) sample of patients with EF ≤35% and GLS ≤−16.4.
The results in this study place GLS in the spotlight, and extend relevant existing data for GLS and echocardiography to GLS derived with CMR (6). The incremental value of GLS in echocardiography in part may be due to the imputation of errors that arise as a result of geometrical assumptions when computing LVEF. This is not the case with CMR, which is the reference standard for LVEF. As such, the independent and incremental value of GLS in the current study supports the notion of that GLS derived using CMR is a more informative imaging biomarker than LVEF.
These findings need to be viewed in the context of certain limitations, including the strain computation protocol (only 2 long-axis views were used) and the lack of a validation cohort for the proposed cutpoints. In addition, precision of measurements is important, and consistent with previous reports, GCS had higher reproducibility compared with GLS.
The current study supported a compelling argument for the measurement of GLS in all patients post-MI. To make the transition from clinical trials to routine clinical practice certain issues need to be addressed. These include feasibility and reproducibility of strain measures outside of core laboratories and high volume units, and replication of these results and cutpoints in independent validation cohorts.
The application of discriminatory cutpoints is predicated upon improving the current intervendor differences (11) and heterogeneity in published normal references ranges (12). Because of the tremendous potential of feature tracking, a standardization initiative, which has been the case for speckle tracking, is an important next step (13).
For strain to transition from being a diagnostic biomarker to a theragnostic biomarker, new clinical evidence in therapeutic trials post-MI is desirable. In other words, GLS (or GCS) measured post-MI improves in association with established or novel therapies, and in turn, becomes a surrogate outcome by tracking (predicting) health outcomes in the longer term. Finally, randomized, controlled clinical trials are needed to determine the clinical usefulness of GLS-based clinical management in post-MI patients.
↵∗ Editorials published in JACC: Cardiovascular Imaging reflect the views of the authors and do not necessarily represent the views of JACC: Cardiovascular Imaging or the American College of Cardiology.
Dr. Mangion was supported by a Fellowship from the British Heart Foundation (FS/15/54/31639).
Dr. Berry was supported by a British Heart Foundation Centre of Excellence Award (RE/13/5/3017); and has a research agreement with Siemens Healthcare. Dr. Pathan has reported that he has no relationships relevant to the contents of this paper to disclose.
- 2018 American College of Cardiology Foundation
- Bristow M.R.,
- Kao D.P.,
- Breathett K.K.,
- et al.
- Pedrizzetti G.,
- Claus P.,
- Kilner P.J.,
- Nagel E.
- Stanton T.,
- Leano R.,
- Marwick T.H.
- Bertini M.,
- Ng A.C.,
- Antoni M.L.,
- et al.
- Hung C.L.,
- Verma A.,
- Uno H.,
- et al.
- Mangion K.,
- McComb C.,
- Auger D.A.,
- Epstein F.H.,
- Berry C.
- Eitel I.,
- Stiermaier T.,
- Lange T.,
- et al.
- Stokke T.M.,
- Hasselberg N.E.,
- Smedsrud M.K.,
- et al.
- Cao J.J.,
- Ngai N.,
- Duncanson L.,
- Cheng J.,
- Gliganic K.,
- Chen Q.
- Vo H.Q.,
- Marwick T.H.,
- Negishi K.