Prediction of functional recovery by cardiac magnetic resonance feature tracking imaging in first time ST-elevation myocardial infarction. Comparison to infarct size and transmurality by late gadolinium enhancement

International Journal of Cardiology 183 (2015) 162–170

Contents lists available at ScienceDirect

International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard

Prediction of functional recovery by cardiac magnetic resonance feature tracking imaging in first time ST-elevation myocardial infarction. Comparison to infarct size and transmurality by late gadolinium enhancement Sebastian J. Buss a, Birgit Krautz a, Nina Hofmann a, Yannick Sander a, Lukas Rust a, Sorin Giusca a, Christian Galuschky b, Sebastian Seitz a, Evangelos Giannitsis a, Sven Pleger a, Philip Raake a, Patrick Most a, Hugo A. Katus a, Grigorios Korosoglou a,⁎ a b

Department of Cardiology, University of Heidelberg, INF 410, 69120 Heidelberg, Germany TomTec Imaging Systems GmbH, Munich, Germany

a r t i c l e

i n f o

Article history: Received 14 October 2014 Received in revised form 23 December 2014 Accepted 4 January 2015 Available online 7 January 2015 Keywords: Myocardial infarction STEMI Cardiac magnetic resonance Left ventricular function Two dimensional strain imaging FTI Late gadolinium enhancement

a b s t r a c t Purpose: To investigate whether myocardial deformation imaging, assessed by feature tracking cardiac magnetic resonance (FTI-CMR), would allow objective quantification of myocardial strain and estimation of functional recovery in patients with first time ST-elevation myocardial infarction (STEMI). Methods: Cardiac magnetic resonance (CMR) imaging was performed in 74 consecutive patients 2–4 days after successfully reperfused STEMI, using a 1.5 T CMR scanner (Philips Achieva). Peak systolic circumferential and longitudinal strains were measured using the FTI applied to SSFP cine sequences and were compared to infarct size, determined by late gadolinium enhancement (LGE). Follow-up CMR at 6 months was performed in order to assess residual ejection fraction, which deemed as the reference standard for the estimation of functional recovery. Results: During the follow-up period 53 of 74 (72%) patients exhibited preserved residual ejection fraction ≥50%. A cut-off value of −19.3% for global circumferential strain identified patients with preserved ejection fraction ≥50% at follow-up with sensitivity of 76% and specificity of 85% (AUC = 0.86, 95% CI = 0.75–0.93, p b 0.001), which was superior to that provided by longitudinal strain (ΔAUC = 0.13, SE = 0.05, z-statistic = 2.5, p = 0.01), and non-inferior to that provided by LGE (ΔAUC = 0.07, p = NS). Multivariate analysis showed that global circumferential strain and LGE exhibited independent value for the prediction of preserved LV-function, surpassing that provided by age, diabetes and baseline ejection fraction (HR = 1.4, 95% CI = 1.0–1.9 and HR = 1.4, 95% CI = 1.1–1.7, respectively, p b 0.05 for both). Conclusions: Estimation of circumferential strain by FTI provides objective assessment of infarct size without the need for contrast agent administration and estimation of functional recovery with non-inferior accuracy compared to that provided by LGE. © 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Since the introduction of percutaneous coronary interventions (PCI) the clinical outcomes of patients with ST-elevation myocardial infarction (STEMI) have considerably improved [1,2]. However, despite modern interventional and pharmacologic therapeutic regimes, prognosis remains limited in a subset of patients with severe left ventricular dysfunction and chronic heart failure after myocardial infarction.

⁎ Corresponding author at: Dept. of Cardiology, Angiology and Pneumology, University of Heidelberg, Im Neuenheimer Feld 410, 69120 Heidelberg, Germany. E-mail address: [email protected] (G. Korosoglou).

http://dx.doi.org/10.1016/j.ijcard.2015.01.022 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.

Therefore, the identification of the extent and degree of contractile dysfunction and myocardial scar in patients with STEMI has important prognostic implications and may help tailoring current or future therapeutic regimes [3]. Several imaging techniques have been previously used for the evaluation of myocardial viability, such as echocardiography and nuclear scintigraphy (reviewed in [4]). However, limited echogenic windows and radiation exposure, respectively may limit the precision and serial applicability of these modalities. Cardiac magnetic resonance (CMR) imaging, on the other hand, is a non-invasive imaging technique that allows accurate assessment of myocardial function and viability with high spatial and temporal resolution and without radiation exposure for the patients [5–8]. Due to its tomographic nature, CMR allows for assessment of myocardial function and scar in identical

S.J. Buss et al. / International Journal of Cardiology 183 (2015) 162–170

myocardial slices, providing excellent reproducibility of the acquired results. In previous studies, we demonstrated the ability of strain encoded magnetic resonance (SENC) for the quantification of myocardial strain in patients with stable coronary artery disease (CAD) and after STEMI [9–11]. In the present study, we used the (feature tracking imaging) FTI algorithm, which can be applied to cine sequences from different vendors and field strengths and obviate the need for dedicated pulse sequences [12–16]. Using FTI we sought to investigate whether quantification of circumferential and longitudinal strains can be used to estimate infarct size by late gadolinium enhancement (LGE) and clinical outcomes in patients with first time STEMI. The results were compared to standard clinical and conventional CMR parameters such as LVejection fraction.

163

All procedures complied with the Declaration of Helsinki and were approved by our local ethic committee and all patients gave written informed consent. 2.2. Biochemical markers Cardiac troponin T (cTnT) was typically collected on presentation and at 72–96 h of follow-up to estimate the extent of myocardial damage [17–20] and was determined by a commercially available enzyme linked immunosorbent assay (Cardiac Reader, Roche, Mannheim, Germany). Creatine kinase was collected on presentation and serially up to the forth day of follow-up. Enzyme activity was determined in a Synchron LX-20® clinical chemistry system (Beckman Coulter, Krefeld, Germany) at 25 °C. 2.3. Cardiac catheterization Angiograms were performed in a standardized fashion, and at least two orthogonal views of every major coronary vessel and its side branches were obtained. The degree of stenosis was expressed as the percent reduction of the internal luminal diameter. TIMI flow grades were assessed visually as described previously after coronary angioplasty [21].

2. Methods

2.4. CMR examination

2.1. Patient cohort

Patients were examined in a clinical 1.5-T whole-body CMR-scanner Achieva system (Philips Medical Systems, Best, The Netherlands) using a dedicated cardiac phased-array receiver coil. A standardized protocol was followed 2–4 days after STEMI, aiming 1) at the assessment of baseline parameters of the left ventricle (LV-diameters, wall thickness, ejection fraction) using cine imaging, 2) the quantification of infarct size and infarct transmurality using late enhancement CMR and 3) the quantification of circumferential and longitudinal strains using feature tracking imaging (FTI) (details for cine imaging, LGE and FTI are provided in the supplementary Methods section). 6 months after STEMI CMR was repeated for the assessment of residual LV-ejection fraction (Fig. 1).

We prospectively enrolled 74 consecutive patients with first time STEMI who were admitted to chest pain unit of our department in the University Hospital Heidelberg. We included patients who had typical chest pain lasting N20 min within 12 h before their presentation and ST segment elevation of N0.2 mV in at least two contiguous ECG leads. Exclusion criteria were ECG signs or history of previous infarction. Thus, we included only patients with first-time acute myocardial infarction with a clearly identified culprit coronary vessel. Patients with severe hemodynamic compromise or patients requiring inotropic support were excluded from the study. In addition patients with other contraindications to CMR such as pacemaker and renal diseases (GFR b 30 ml/min/1.73 m2) and patients with claustrophobia were excluded from the study. All patients underwent cardiac catheterization and were treated with primary angioplasty and stent placement. CMR was performed at days 2–4 (baseline) and at 6 months after STEMI (Fig. 1).

2.5. Study end points Follow-up ejection fraction ≥50% (6 months after first time STEMI) was the primary endpoint of this study because this is an important clinical goal [21]. Secondary end

Fig. 1. Study flow chart and procedures during the study duration.

164

S.J. Buss et al. / International Journal of Cardiology 183 (2015) 162–170

point was the combined incidence of cardiac death, non-fatal recurrent myocardial infarction and new revascularization by PCI or CABG within 12 months of follow-up.

2.6. Statistical analysis Continuous variables were expressed as mean ± standard deviation for parametric parameters or as median with interquartile range (IQR) for nonparametric variables. For continuous variables, differences between two groups were compared using Student's ttest or Mann Whitney U test. In cases of more than two groups, differences were compared using ANOVA with Bonferroni adjustment for multiple comparisons. Categorical variables were expressed as counts and percentages and were compared by Chi-square test or Fisher exact test, respectively. Receiver operating characteristic (ROC) analysis was used in order to define the optimal cut-off values for myocardial strain, cTnT and quantitative LGE. Univariate and multivariate logistic regression analyses were performed to calculate hazard ratios (HR) and 95% confidence intervals (CI). For all analyses, a p-value of b0.05 was regarded as statistically significant. All tests were two-sided. Inter- and intra-observer variabilities for global longitudinal and circumferential strains and for LGE were assessed by repeated analysis of 30 randomly selected patients. Readings were separated by 8 weeks to minimize recall bias. The MedCalc 13.2 (MedCalc software, Mariakerke, Belgium) computer programs were used throughout.

3. Results 3.1. Clinical characteristics Data were prospectively collected in 74 patients (mean age = 57 ± 12 years, 54 (73%) males, 18 (24%) with diabetes, 72–96 h cTnT of 3.1 ± 1.8 μg/l). Baseline characteristics and CMR data are illustrated in Table 1. 72 of 74 (97%) patients achieved TIMI 3 flow after PCI in the infarct related artery.

Table 1 Demographic, clinical, laboratory and CMR data. Parameters

All patients (n = 74)

Clinical data Age (years) Male gender; n (%) Arterial hypertension Hyperlipidemia Smoking Diabetes mellitus Body mass index N30 kg/m2 NYHA class (0–4)

57 ± 11 54 (73%) 40 (54%) 38 (51%) 49 (66%) 12 (16%) 27 (36%) 1.2 ± 0.5

Laboratory data 72–96 h troponin T (μg/l) Admission creatine kinase (U/l) Peak creatine kinase (U/l) NT pro BNP (pg/ml) Admission creatinine (mg/dl)

3.1 ± 1.8 622 ± 1106 2385 ± 1715 1139 ± 812 0.8 ± 0.2

Baseline CMR LV ejection fraction (%) LVEDV (ml) LVESV (ml) Cardiac output (l/min) Cardiac index (l/min∗m2) LV mass (g) LV mass/BSA (g/m2) End-diastolic LV diameter (mm) End-systolic LV diameter (mm) Presence of LGE LGE as percent of LV-mass (%)

52.5 ± 9.7 170 ± 35 81 ± 30 6.1 ± 1.5 3.1 ± 0.6 156 ± 48 75 ± 26 52 ± 5 36 ± 7 74 (100%) 18 ± 9

Follow-up CMR LV ejection fraction (%)

55.4 ± 10.2

Cardiac medications ß-Blockers ACE inhibitors/AT II blockers Diuretics Statins

71 (96%) 71 (96%) 31 (42%) 73 (99%)

Data are presented as mean ± standard deviation or as proportions.

3.2. Association between regional strain & infarct transmurality by LGE Regional circumferential strain increased with increasing infarct transmurality (p b 0.001), was highly predictive of infarct transmurality ≥75% (AUC = 0.91, 95% CI = 0.90–0.93, p b 0.001) (Fig. 2A–B). Weaker albeit significant associations were observed for regional longitudinal strain (p b 0.001 for ANOVA and AUC = 0.74, 95% CI = 0.72–0.76, p b 0.001) (Fig. 2C–D). By comparison of the ROC curves regional circumferential strain exhibited significantly higher accuracy for the estimation of infarct transmurality (ΔAUC = 0.18, SE = 0.02, z-statistics = −7.1, p b 0.001). Lower regional circumferential strain in the inferior-lateral LV-wall by FTI can be appreciated in a patient with transmural inferior-lateral infarction in Fig. 3 and Online video. Conversely, regions of remote myocardium exhibited normal circumferential strain. 3.3. Associations between global infarct size by LGE, strain & troponin T Significant correlations were observed between LGE and 72–96 h cTnT (r = 0.67, 95% CI = 0.52 to 0.78, p b 0.001), LGE and follow-up ejection fraction (r = −0.70, 95% CI = −0.80 to −0.56, p b 0.001) as well as 72–96 h cTnT and follow-up ejection fraction (r = −0.53, 95% CI = −0.68 to −0.34, p b 0.001) (Fig. 4A–C). In addition, global circumferential strain was related to both LGE and follow-up ejection fraction (r = 0.75, 95% CI = 0.63 to 0.83 and r = − 0.71, 95% CI = − 0.80 to − 0.57, respectively, p b 0.001 for both). Associations were also observed for global longitudinal strain (r = 0.45, 95% CI = 0.24–0.61 for LGE and r = − 0.48, 95% CI = − 0.64 to − 0.28 for follow-up ejection fraction, respectively, p b 0.001 for both) (Fig. 5A–D). The observed associations were stronger for the global circumferential versus longitudinal strain both with the LGE (z-statistics = −2.9, p = 0.004) and with the follow-up ejection fraction (z-statistics = −2.2, 2.2, p = 0.03). 3.4. Prediction of the primary endpoint (follow-up ejection fraction ≥50%) During the follow-up period 53 of 74 (72%) patients exhibited preserved follow-up ejection fraction ≥50%. Only 2 (3%) patients showed severely impaired residual LV-ejection fraction b35%. ROC analysis demonstrated that LGE, 72–96 h cTnT and global circumferential & longitudinal strains predicted preserved follow-up ejection fraction ≥50% with AUC of 0.92, 0.83, 0.86 and 0.72, respectively, p b 0.01 for all (Fig. 6A–D). An overview of sensitivities, specificities and positive and negative predictive values for the primary endpoint is provided in Table 2. Comparison of ROC curves showed that LGE exhibited superior value for the prediction of preserved ejection fraction compared to global longitudinal strain (ΔAUC = 0.20, SE = 0.07, z-statistic = 3.1, p = 0.002). Global circumferential strain, on the other hand, was non-inferior to LGE (ΔAUC = 0.07, SE = 0.04, z-statistic = 1.8, p = NS) but superior to global longitudinal strain (ΔAUC = 0.13, SE = 0.05, z-statistic = 2.5, p = 0.01). 3.5. Uni- & multi-variate analyses By univariate analysis, 72–96 h cTnT, global circumferential and longitudinal strains and LGE were predictors of preserved follow-up LV function (Table 3A). By multivariate analysis, global circumferential strain and LGE both exhibited independent value for the prediction of preserved LVfunction, surpassing that provided by age, diabetes and baseline ejection-fraction (HR = 1.4, 95% CI = 1.0–1.9 and HR = 1.4, 95% CI = 1.1–1.7, respectively, p b 0.05 for both). Global longitudinal strain, on the other hand, was not independently predictive of the primary endpoint (Table 3B).

S.J. Buss et al. / International Journal of Cardiology 183 (2015) 162–170

A

B

C

D

165

Fig. 2. Circumferential strain was related to regional infarct transmurality by LGE (A) and predicted infarct transmurality ≥75% with high accuracy (B). Weaker associations were observed for longitudinal strain (C–D). Values are provided as mean ± 95% CI for longitudinal and circumferential strains.

Fig. 3. Patient with hypokinesia of the inferior-lateral wall (A–B) and transmural infarction by LGE (C). Using FTI analysis lower circumferential strain was seen in the corresponding inferior-lateral LV wall (D). The black arrow indicates the negative peak of the mean dotted strain curve, representing the average of all segments over the whole cardiac cycle for this patient. In the online video tracking of the LV endocardium by FTI can be appreciated.

166

S.J. Buss et al. / International Journal of Cardiology 183 (2015) 162–170

A

B

C

Fig. 4. LGE was associated with 72–96 h cTnT (A) and with follow-up ejection fraction (B). In addition 72–96 h cTnT was significantly related to follow-up ejection fraction (C).

3.6. Clinical outcomes during 12 months of follow-up During the follow-up period 13 patients reached the combined endpoint (n = 4 with recurrent non-fatal MI and 9 with repeated revascularization procedures by PCI). A trend was noticed for higher infarct size by LGE and reduced global circumferential strain in patients who reached the combined endpoint, without however reaching statistical significance (22.8 ± 9.3% versus 17.1 ± 9.8% for LGE, p = 0.06 and −19.8 ± 4.6 versus −21.7 ± 5.2, p = 0.21).

3) Global circumferential strain is a strong predictor of preserved ejection fraction during follow-up independent of age, diabetes and baseline LV ejection fraction and with similar precision to that provided by the clinically established LGE technique. Longitudinal strain, on the other hand, was not independently associated with follow-up ejection fraction. 4) Assessment of global circumferential strain by FTI may be a simple strategy for the risk stratification of STEMI patients with severely impaired renal function and contraindication for gadolinium administration.

3.7. Observer variability 4.1. Electrocardiogram and cardiac troponins Intra- and inter-observer variabilities were 7.8% and 10.8%, for the quantification of global circumferential and 8.6% and 12.7%, respectively for the quantification of global longitudinal strain. 4. Discussion We report on consecutive patients with first ST-elevation myocardial infarction and systematic assessment of LV deformation in terms of circumferential and longitudinal strains. Our results were related to baseline LV ejection fraction, infarct size estimation by cTnT and late gadolinium enhancement. Follow-up ejection fraction at 6 months was used as the standard reference for functional outcome (primary endpoint). The main findings of our study are: 1) The FTI algorithm enables the reproducible assessment of myocardial deformation in patients with myocardial infarction. 2) Regional circumferential strain is closely related to infarct transmurality by LGE. A weaker association was observed for regional longitudinal strain.

In clinical care of acute myocardial infarction, a number of techniques including the electrocardiogram and the release of cardiac troponins have been used to determine infarct size and potentially myocardial salvage. Although the electrocardiogram is a useful clinical tool in the primary diagnosis of acute STEMI, its role in determining the success of myocardial reperfusion is limited [21]. Cardiac troponins, on the other hand, are extremely useful in estimating infarct size in acute ischemic syndromes. Their incremental diagnostic value, compared to other enzymes is related to their cardio-specific isoform expression of troponin T and I, which allow for differentiation of cardiac and skeletal muscle injury providing a high myocardium-to-blood concentration gradient [22]. Due to their high sensitivity, specificity and accuracy, troponins are recommended by guidelines for the detection of myocardial injury in patients with acute coronary syndromes. In many randomized trials testing treatment options in such patients, patients with cTnT elevations experienced markedly higher cardiovascular event rates compared to troponin negative patients [23,24]. In our study, we demonstrated that cTnT assessed

S.J. Buss et al. / International Journal of Cardiology 183 (2015) 162–170

A

B

C

D

167

Fig. 5. Circumferential strain was related to both LGE and follow-up ejection fraction (A–B), while weaker associations were observed with longitudinal stain values (C–D).

between 72 and 96 h after STEMI was related to LGE and myocardial deformation and exhibited high sensitivity and accuracy for the prediction of preserved LV-function during follow-up. This is in agreement with previous studies, where we observed a close relation between cTnT and myocardial scar by LGE in patients with STEMI and non-STEMI [20]. In light of the practicability of a single cTnT value, this marker is in the meanwhile well established for the risk stratification. Particularly the measurement of cTnT on the 4th day after infarction has gained

wide clinical acceptance for the estimation of infarct size, because the measurement is easy and relatively inexpensive. 4.2. Earlier studies using ultrasound and CMR based deformation Echocardiography based deformation imaging was previously used for the analysis of myocardial stretching and compression, which was validated in experimental models using sonomicrometry [25]. By this

A

B

C

D

Fig. 6. ROC analysis demonstrated that LGE (A), 72–96 h cTnT (B), and global circumferential & longitudinal stains (C–D) predicted preserved follow-up ejection fraction ≥50% with overall AUC of 0.92, 0.83, 0.86 and 0.72, respectively, p b 0.01 for all.

168

S.J. Buss et al. / International Journal of Cardiology 183 (2015) 162–170

Table 2 Overview of sensitivities, specificities and positive and negative predictive values of LGE, 72–96 h cTnT and myocardial strain for the prediction of preserved LV-function. Cut-off value

Sensitivity Specificity PPV NPV AUC (%) (%) (%) (%)

Late gadolinium enhancement 23% 76% 72–96 h cTnT (μg/l) 2.6 95% Circumferential strain −19.3% 76% Longitudinal strain −12.6 50%

92% 69% 85% 91%

80% 56% 67% 67%

91% 97% 90% 83%

0.92 0.83 0.86 0.72

approach, echocardiographic strain provided assessment of myocardial viability, obviating the need of low-dose inotropic stimulation for the assessment of functional reserve [26]. Because infarct size and transmurality are both major predictors of cardiovascular death, reinfarction and congestive heart failure the notion of estimating such variables by resting echocardiography, which is a non-invasive imaging technique with high degree of practicability and cost-effectiveness, appears attractive [27]. However, echocardiography is inherently limited in patients with poor echogenic windows. In addition, echocardiographic tissue Doppler and strain imaging, are characterized by their dependence on the angle of insonation, allowing the assessment of only in-plane velocity vectors. CMR on the other hand, offers the advantage of 3-dimensional data acquisition, and generally provides higher reproducibility due to its tomographic nature. We and others have previously used both tagged CMR and Strain-Encoded MR (SENC) for the assessment of myocardial strain after infarction [9,28]. In this regard, such pulse sequences surpassed the ability of conventional cine imaging for the detection of myocardial viability [9]. However, both tagged CMR and SENC sequences have some limitations. Thus, conventional tagging suffers from low temporal resolution (b 30 frames/s), limiting its accuracy, especially in patients with higher heart rates. In addition, acquisition of tagged images often requires long series of breath holds, whereas tag fading during diastole limits the assessment of myocardial relaxation. SENC on the other hand, allows for precise estimation of longitudinal and circumferential strains, but requires dedicated pulse sequences, which are still not widely available with CMR scanners. Feature tracking imaging (FTI) was originally designed for the analysis of echocardiographic images and was previously validated in experimental and clinical settings [13,29,30]. Recently, FTI was applied with cine steady state free precession sequences exhibiting highly reproducible results for the assessment of myocardial strain compared to CMR tagging [13,14,31]. Importantly, using FTI, myocardial deformation can be assessed without the need for additional pulse sequences and post processing can be performed with different CMR vendors and field strengths [12]. In addition, tag fading during diastole is not an issue with FTI, which therefore allows for quantification of strain throughout the whole cardiac cycle with high signal-to-noise ratio. In our study, we showed to our knowledge for the first time in the current literature that circumferential strain based on the analysis of cine images by FTI can independently predict recovery of myocardial function in patients with first time STEMI. Importantly, all patients underwent baseline and follow-up CMR 6 months after infarction in order to assess residual LV-function, which is a major clinical goal. Circumferential rather than longitudinal strain exhibited similar precision for the prediction of the primary endpoint, compared to the established late gadolinium enhancement. This is in agreement with previous observations, where circumferential strain exhibited the highest sensitivity and specificity for the detection of myocardial dysfunction during inotropic stimulation and balloon-induced ischemia [32,33]. In addition, circumferential strain was previously shown to be a robust predictor of cardiac outcomes in 201 consecutive patients hospitalized for heart failure, surpassing the value of longitudinal strain and LV-ejection fraction [34]. Another echocardiographic study also demonstrated the noninferiority of circumferential strain compared to LGE for the prediction of global functional recovery, which is in close agreement to our

findings [35]. In the same line, a recent CMR study showed that circumferential strain has higher sensitivity and specificity for the differentiation of subendocardial versus transmural myocardial infarction, compared to longitudinal strain [36]. Although the prognostic value of circumferential compared to that of longitudinal strain is rated differently in the current literature [37,38], it seems that such differential effects between subendocardial and transmural infarctions on longitudinal and circumferential strains may be a consequence of the helical wrapping of cardiac fibers into three different anatomical layers, as described in previous experimental studies [39]. 4.3. Clinical implications Our findings may have potential clinical implications in light of the fact that myocardial strain measures can be performed together with all other standard CMR parameters during a single non-invasive examination, early after PCI and without additional time spent during scan acquisitions. The early identification of post-reperfusion microvascular dysfunction and impaired myocardial deformation may help in tailoring appropriate pharmacological interventions in patients with reduced global strain, who would otherwise have low likelihood to retain contractile function and higher risk for congestive heart failure in the long run. Thus, the assessment of myocardial deformation may help to identify patients, who would benefit from rescue PCI after unsuccessful thrombolysis and from aggressive medical therapy including the early and high-dose administration of ACE- and aldosterone-inhibitors or future approaches aiming at reduction of infarct size and left ventricular remodeling [40,41]. 4.4. Limitations The number of patients was relatively small. Thus, non-significant trends were observed for circumferential strain and LGE for our clinical endpoints. However, highly significant results were observed in terms of strain and LGE in patients with versus without preserved residual LV-function, which was our primary endpoint. In addition, feature tracking derived strain was not compared to CMR tagging or SENC in this study. However, such comparisons have been performed in

Table 3 Uni- and multi-variable analyses for the prediction of preserved follow-up ejection fraction. Parameter

HR (95% CI)

p-Value

A. Univariate analysis Age Male gender Diabetes mellitus Baseline ejection fraction (%) 72–96 h cTnT (μg/l) Circumferential strain (%) Longitudinal strain (%) Late gadolinium enhancement (%)

0.99 (0.95–1.03) 2.3 (0.58–9.0) 0.55 (0.14–2.2) 0.87 (0.81–0.94) 2.1 (1.4–3.1) 1.4 (1.2–1.6) 1.2 (1.1–1.4) 1.3 (1.2–1.5)

NS NS NS b0.001 b0.001 b0.001 0.005 b0.001

B (i). Multivariate analysis (LGE) Age Diabetes mellitus Baseline ejection fraction (%) Late gadolinium enhancement (%)

0.92 (0.83–1.02) 0.04 (0.002–1.02) 0.86 (0.73–1.02) 1.4 (1.1–1.7)

0.09 NS 0.09 b0.01

B (ii). Multivariate analysis (circumferential strain) Age 0.94 (0.88–1.02) Diabetes mellitus 0.17 (0.02–1.32) Baseline ejection fraction (%) 0.9 (0.80–1.06) Circumferential strain (%) 1.4 (1.0–1.9)

NS NS NS b0.05

B (iii). Multivariate analysis (longitudinal strain) Age 0.96 (0.89–1.02) Diabetes mellitus 0.27 (0.04–1.66) Baseline ejection fraction (%) 0.9 (0.76–0.95) Longitudinal strain (%) 1.1 (0.9–1.3)

NS NS b0.01 NS

S.J. Buss et al. / International Journal of Cardiology 183 (2015) 162–170

previous studies showing good agreement between CMR tagging and FTI [31]. Furthermore, analysis was mainly performed on a global level, because such an approach can more easily be translated to individual risk stratification in each patient. In addition, other noncontrast techniques such as native T1 mapping or T2 weighted imaging were not used in our study for differentiation between reversible and irreversible myocardial damage, which is a limitation. Finally, we focused on patients with first time STEMI, so that our results cannot be extrapolated in individuals with ischemic heart disease and recurrent infarctions. 5. Conclusions Circumferential strain, based on the analysis of cine-SSFP images by FTI, can independently predict preserved residual ejection fraction with non-inferior precision compared to that provided by the clinically established LGE imaging. Assessment of circumferential strain using this technique shows high reproducibility and may be an alternative to the use of low-dose inotropic stimulation for the assessment of functional reserve and to LGE imaging in patients with impaired renal function and contraindications to gadolinium administration. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijcard.2015.01.022. Competing interests

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

The authors declare that they have no competing interests. [21]

Acknowledgements We thank Daniel Helm, Angela Stöcker-Wochele, Miriam Hess, Vesna Vukovic and Birgit Hörig for their excellent technical assistance with the acquisitions of all CMR scans.

[22]

[23]

References [1] F. Zijlstra, A. Patel, M. Jones, C.L. Grines, S. Ellis, E. Garcia, et al., Clinical characteristics and outcome of patients with early (b2 h), intermediate (2–4 h) and late (N4 h) presentation treated by primary coronary angioplasty or thrombolytic therapy for acute myocardial infarction, Eur. Heart J. 23 (2002) 550–557. [2] J.K. Harrison, R.M. Califf, L.H. Woodlief, D. Kereiakes, B.S. George, R.S. Stack, et al., Systolic left ventricular function after reperfusion therapy for acute myocardial infarction. Analysis of determinants of improvement. The TAMI Study Group, Circulation 87 (1993) 1531–1541. [3] R. Bigi, A. Desideri, J.J. Bax, A. Galati, C. Coletta, C. Fiorentini, et al., Prognostic interaction between viability and residual myocardial ischemia by dobutamine stress echocardiography in patients with acute myocardial infarction and mildly impaired left ventricular function, Am. J. Cardiol. 87 (2001) 283–288. [4] T.H. Marwick, The viable myocardium: epidemiology, detection, and clinical implications, Lancet 351 (1998) 815–819. [5] B.L. Gerber, J. Garot, D.A. Bluemke, K.C. Wu, J.A. Lima, Accuracy of contrast-enhanced magnetic resonance imaging in predicting improvement of regional myocardial function in patients after acute myocardial infarction, Circulation 106 (2002) 1083–1089. [6] S. Kelle, E. Nagel, A. Voss, N. Hofmann, G. Gitsioudis, S.J. Buss, et al., A bi-center cardiovascular magnetic resonance prognosis study focusing on dobutamine wall motion and late gadolinium enhancement in 3,138 consecutive patients, J. Am. Coll. Cardiol. 61 (2013) 2310–2312. [7] D.P. Ripley, O.E. Gosling, L. Bhatia, C.R. Peebles, A.C. Shore, N. Curzen, et al., The relationship between the contralateral collateral supply and myocardial viability on cardiovascular magnetic resonance: can the angiogram predict functional recovery? Int. J. Cardiol. 177 (2014) 362–367. [8] F. Demirel, A. Adiyaman, J.R. Timmer, J.H. Dambrink, M. Kok, W.J. Boeve, et al., Myocardial scar characteristics based on cardiac magnetic resonance imaging is associated with ventricular tachyarrhythmia in patients with ischemic cardiomyopathy, Int. J. Cardiol. 177 (2014) 392–399. [9] M. Neizel, G. Korosoglou, D. Lossnitzer, H. Kuhl, R. Hoffmann, C. Ocklenburg, et al., Impact of systolic and diastolic deformation indexes assessed by strain-encoded imaging to predict persistent severe myocardial dysfunction in patients after acute myocardial infarction at follow-up, J. Am. Coll. Cardiol. 56 (2010) 1056–1062. [10] G. Korosoglou, S. Lehrke, A. Wochele, B. Hoerig, D. Lossnitzer, H. Steen, et al., Strainencoded CMR for the detection of inducible ischemia during intermediate stress, JACC Cardiovasc. Imaging 3 (2010) 361–371. [11] G. Korosoglou, G. Gitsioudis, A. Voss, S. Lehrke, N. Riedle, S.J. Buss, et al., Strainencoded cardiac magnetic resonance during high-dose dobutamine stress

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

169

testing for the estimation of cardiac outcomes: comparison to clinical parameters and conventional wall motion readings, J. Am. Coll. Cardiol. 58 (2011) 1140–1149. A. Schuster, G. Morton, S.T. Hussain, R. Jogiya, S. Kutty, K.N. Asrress, et al., The intraobserver reproducibility of cardiovascular magnetic resonance myocardial feature tracking strain assessment is independent of field strength, Eur. J. Radiol. 82 (2013) 296–301. K.N. Hor, W.M. Gottliebson, C. Carson, E. Wash, J. Cnota, R. Fleck, et al., Comparison of magnetic resonance feature tracking for strain calculation with harmonic phase imaging analysis, JACC Cardiovasc. Imaging 3 (2010) 144–151. D. Augustine, A.J. Lewandowski, M. Lazdam, A. Rai, J. Francis, S. Myerson, et al., Global and regional left ventricular myocardial deformation measures by magnetic resonance feature tracking in healthy volunteers: comparison with tagging and relevance of gender, J. Cardiovasc. Magn. Reson. 15 (2013) 8. S.J. Buss, K. Breuninger, S. Lehrke, A. Voss, C. Galuschky, D. Lossnitzer, et al., Assessment of myocardial deformation with cardiac magnetic resonance strain imaging improves risk stratification in patients with dilated cardiomyopathy, Eur. Heart J. Cardiovasc. Imaging (Sep 21 2014) pii: jeu181. [Epub ahead of print]. R.J. Taylor, F. Umar, W.E. Moody, C. Meyyappan, B. Stegemann, J.N. Townend, et al., Feature-tracking cardiovascular magnetic resonance as a novel technique for the assessment of mechanical dyssynchrony, Int. J. Cardiol. 175 (2014) 120–125. G. Korosoglou, N. Labadze, E. Giannitsis, R. Bekeredjian, A. Hansen, S.E. Hardt, et al., Usefulness of real-time myocardial perfusion imaging to evaluate tissue level reperfusion in patients with non-ST-elevation myocardial infarction, Am. J. Cardiol. 95 (2005) 1033–1038. A. Remppis, P. Ehlermann, E. Giannitsis, T. Greten, P. Most, M. Muller-Bardorff, et al., Cardiac troponin T levels at 96 hours reflect myocardial infarct size: a pathoanatomical study, Cardiology 93 (2000) 249–253. G. Korosoglou, N. Labadze, A. Hansen, C. Selter, E. Giannitsis, H. Katus, et al., Usefulness of real-time myocardial perfusion imaging in the evaluation of patients with first time chest pain, Am. J. Cardiol. 94 (2004) 1225–1231. E. Giannitsis, H. Steen, K. Kurz, B. Ivandic, A.C. Simon, S. Futterer, et al., Cardiac magnetic resonance imaging study for quantification of infarct size comparing directly serial versus single time-point measurements of cardiac troponin T, J. Am. Coll. Cardiol. 51 (2008) 307–314. G. Korosoglou, A. Haars, G. Michael, M. Erbacher, S. Hardt, E. Giannitsis, et al., Quantitative evaluation of myocardial blush to assess tissue level reperfusion in patients with acute ST-elevation myocardial infarction: incremental prognostic value compared with visual assessment, Am. Heart J. 153 (2007) 612–620. H.A. Katus, A. Remppis, S. Looser, K. Hallermeier, T. Scheffold, W. Kubler, Enzyme linked immuno assay of cardiac troponin T for the detection of acute myocardial infarction in patients, J. Mol. Cell. Cardiol. 21 (1989) 1349–1353. C.W. Hamm, C. Heeschen, B. Goldmann, A. Vahanian, J. Adgey, C.M. Miguel, et al., Benefit of abciximab in patients with refractory unstable angina in relation to serum troponin T levels. c7E3 Fab Antiplatelet Therapy in Unstable Refractory Angina (CAPTURE) Study Investigators, N. Engl. J. Med. 340 (1999) 1623–1629. D.A. Morrow, C.P. Cannon, N. Rifai, M.J. Frey, R. Vicari, N. Lakkis, et al., Ability of minor elevations of troponins I and T to predict benefit from an early invasive strategy in patients with unstable angina and non-ST elevation myocardial infarction: results from a randomized trial, JAMA 286 (2001) 2405–2412. S. Urheim, T. Edvardsen, H. Torp, B. Angelsen, O.A. Smiseth, Myocardial strain by Doppler echocardiography. Validation of a new method to quantify regional myocardial function, Circulation 102 (2000) 1158–1164. Y. Zhang, A.K. Chan, C.M. Yu, G.W. Yip, J.W. Fung, W.W. Lam, et al., Strain rate imaging differentiates transmural from non-transmural myocardial infarction: a validation study using delayed-enhancement magnetic resonance imaging, J. Am. Coll. Cardiol. 46 (2005) 864–871. K.C. Wu, E.A. Zerhouni, R.M. Judd, C.H. Lugo-Olivieri, L.A. Barouch, S.P. Schulman, et al., Prognostic significance of microvascular obstruction by magnetic resonance imaging in patients with acute myocardial infarction, Circulation 97 (1998) 765–772. M.J. Gotte, A.C. van Rossum, J.W.R. Twisk, J.P.A. Kuijer, J.T. Marcus, C.A. Visser, Quantification of regional contractile function after infarction: strain analysis superior to wall thickening analysis in discriminating infarct from remote myocardium, J. Am. Coll. Cardiol. 37 (2001) 808–817. B. Pirat, D.S. Khoury, C.J. Hartley, L. Tiller, L. Rao, D.G. Schulz, et al., A novel featuretracking echocardiographic method for the quantitation of regional myocardial function: validation in an animal model of ischemia–reperfusion, J. Am. Coll. Cardiol. 51 (2008) 651–659. J. Korinek, J. Wang, P.P. Sengupta, C. Miyazaki, J. Kjaergaard, E. McMahon, et al., Twodimensional strain—a Doppler-independent ultrasound method for quantitation of regional deformation: validation in vitro and in vivo, J. Am. Soc. Echocardiogr. 18 (2005) 1247–1253. A. Schuster, S. Kutty, A. Padiyath, V. Parish, P. Gribben, D.A. Danford, et al., Cardiovascular magnetic resonance myocardial feature tracking detects quantitative wall motion during dobutamine stress, J. Cardiovasc. Magn. Reson. 13 (2011) 58. H. Tanaka, Y. Oishi, Y. Mizuguchi, S. Emi, T. Ishimoto, N. Nagase, et al., Threedimensional evaluation of dobutamine-induced changes in regional myocardial deformation in ischemic myocardium using ultrasonic strain measurements: the role of circumferential myocardial shortening, J. Am. Soc. Echocardiogr. 20 (2007) 1294–1299. R. Winter, R. Jussila, J. Nowak, L.A. Brodin, Speckle tracking echocardiography is a sensitive tool for the detection of myocardial ischemia: a pilot study from the catheterization laboratory during percutaneous coronary intervention, J. Am. Soc. Echocardiogr. 20 (2007) 974–981.

170

S.J. Buss et al. / International Journal of Cardiology 183 (2015) 162–170

[34] G.Y. Cho, T.H. Marwick, H.S. Kim, M.K. Kim, K.S. Hong, D.J. Oh, Global 2-dimensional strain as a new prognosticator in patients with heart failure, J. Am. Coll. Cardiol. 54 (2009) 618–624. [35] E. Altiok, S. Tiemann, M. Becker, R. Koos, C. Zwicker, J. Schroeder, et al., Myocardial deformation imaging by two-dimensional speckle-tracking echocardiography for prediction of global and segmental functional changes after acute myocardial infarction: a comparison with late gadolinium enhancement cardiac magnetic resonance, J. Am. Soc. Echocardiogr. 27 (2014) 249–257. [36] N. Oyama-Manabe, N. Ishimori, H. Sugimori, M. Van Cauteren, K. Kudo, O. Manabe, et al., Identification and further differentiation of subendocardial and transmural myocardial infarction by fast strain-encoded (SENC) magnetic resonance imaging at 3.0 Tesla, Eur. Radiol. 21 (2011) 2362–2368. [37] M. Ersboll, N. Valeur, U.M. Mogensen, M.J. Andersen, J.E. Moller, E.J. Velazquez, 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. 61 (2013) 2365–2373.

[38] H. Motoki, A.G. Borowski, K. Shrestha, R.W. Troughton, W.H. Tang, J.D. Thomas, et al., Incremental prognostic value of assessing left ventricular myocardial mechanics in patients with chronic systolic heart failure, J. Am. Coll. Cardiol. 60 (2012) 2074–2081. [39] L.K. Waldman, Y.C. Fung, J.W. Covell, Transmural myocardial deformation in the canine left ventricle. Normal in vivo three-dimensional finite strains, Circ. Res. 57 (1985) 152–163. [40] A. Florian, A. Ludwig, S. Rosch, H. Yildiz, S. Klumpp, U. Sechtem, et al., Positive effect of intravenous iron-oxide administration on left ventricular remodelling in patients with acute ST-elevation myocardial infarction — a cardiovascular magnetic resonance (CMR) study, Int. J. Cardiol. 173 (2014) 184–189. [41] V. Bodi, J.M. Ruiz-Nodar, E. Feliu, G. Minana, J. Nunez, O. Husser, et al., Effect of ischemic postconditioning on microvascular obstruction in reperfused myocardial infarction. Results of a randomized study in patients and of an experimental model in swine, Int. J. Cardiol. 175 (2014) 138–146.