© 2013, Wiley Periodicals, Inc. DOI: 10.1111/echo.12458



Utility of Speckle Tracking Echocardiography to Characterize Dysfunctional Myocardium in Patients with Ischemic Cardiomyopathy Referred for Cardiac Resynchronization Therapy Anna C. Kydd, M.D.,* Fakhar Khan, M.D.,* Deepa Gopalan, M.Sc.,† Liam Ring, M.D.,* Bushra S. Rana, M.D.,† Mohan S. Virdee, M.D.,† and David P. Dutka, D.M.* *Department of Medicine, Addenbrooke’s Hospital, University of Cambridge, Hills Road, Cambridge, United Kingdom; and †Papworth Hospital NHS Foundation Trust, Papworth Everard, Cambridge, United Kingdom

Background: Assessment of transmural scar at the site of latest mechanical activation is relevant to maximize outcomes in cardiac resynchronization therapy (CRT). Few studies have assessed the ability of speckle tracking echocardiography (STE)-derived short-axis strain to identify segmental myocardial scar, defined by contrast-enhanced cardiac magnetic resonance imaging (CMR), in patients referred for CRT. Methods: A total of 26 patients with ischemic cardiomyopathy who underwent preprocedure echocardiography and CMR were studied. Extent of transmural scar was assessed using contrast-enhanced CMR and corresponding peak segmental radial and circumferential strains were derived using twodimensional (2D) STE. Total left ventricle (LV) scar volume was compared with parameters of global strain. CRT response was defined as >15% reduction in LV end systolic volume (LVESV) at 6 months. Results: Speckle tracking short-axis strain analysis was technically possible in over 90% of LV segments. Applying a segmental radial strain cutoff value of 10% distinguished segments with >50% scar area with a high negative predictive value (98%). Global longitudinal strain < 5% predicted CRT response. Conclusions: Two-dimensional STE offers potential to characterize dysfunctional myocardium and define segmental scar offering an integrated imaging approach to guide LV lead placement for CRT. (Echocardiography 2013;00:1–8) Key words: speckle tracking echocardiography, scar, cardiac resynchronization therapy, myocardial dysfunction The use of echocardiographic parameters to enhance patient selection for cardiac resynchronization therapy (CRT) has largely focused on the assessment of dyssynchrony, and so far has been disappointing.1 Achieving an optimal left ventricular (LV) lead position, in viable myocardium, is important for successful CRT,2,3 and both regional and total LV scar burden have been shown to limit LV remodeling response.4 The role of preprocedure myocardial substrate characterization has increasing importance. Cardiac magnetic resonance (CMR) with delayed enhancement imaging following late gadolinium enhancement (LGE) is the gold standard for myocardial scar characterization, but general use is restricted by cost, availability, and limited to patients with preserved renal function. Address for correspondence and reprint requests: David P. Dutka, D.M., Department of Medicine, University of Cambridge, Level 6, ACCI Box 110, Addenbrooke’s Hospital, Hills Road, Cambridge, UK, CB2 0QQ. Fax: +44-1223-331505; E-mail: [email protected]

In comparison echocardiography is widely available, and may offer an alternative approach to tissue characterization, with potential to be incorporated into a widely utilized preassessment algorithm to enhance patient selection, LV lead placement and therefore outcomes following CRT. The technique of speckle tracking echocardiography (STE) can be used to assess segmental myocardial function with advantages over the use of more established tissue Doppler techniques as it is less affected by angle of insonation and translational motion. The magnitude of peak regional strain derived from STE is proposed as a marker of segmental viability.5 However, data regarding the clinical utility of published cutoff values and the comparison of circumferential and radial strain vectors is limited. Similarly, global strain values have shown potential to predict scar volume within the LV.6 We have previously demonstrated that peak segmental radial strain >10% at the site of the LV 1

Kydd, et al.

lead can be used to predict LV remodeling response following CRT and hypothesized that this was related to absence of significant scar within the paced segment.7 In patients with ischemic cardiomyopathy LV impairment may result from a complex combination of scar (both partial and full thickness) and regions of viable, but dysfunctional myocardium (hibernation and/ or stunning) despite adequate revascularization.8 In a proof-of-concept study we set out to assess whether peak two-dimensional (2D) segmental speckle tracking short-axis strain (radial and circumferential) could define myocardial scar segments (>50% segmental scar area), in a cohort of patients with ischemic cardiomyopathy undergoing CRT. Methods: Study Population: We studied 26 patients with advanced heart failure due to ischemic cardiomyopathy who underwent LGE-CMR prior to CRT. All patients were in sinus rhythm with impaired systolic function (LV ejection fraction [EF] 120 ms) and symptomatic limitation (New York Heart Association (NYHA) functional class III or IV) despite optimal medical therapy and serum creatinine 50% coronary stenosis on coronary angiography. All patients completed a baseline assessment including 2D echocardiography and were reassessed at 6 months. The study was approved by our local hospital review board. Echocardiography Image Acquisition and Analysis: Echocardiography was performed using a commercially available system (GE Vivid 7; GE Healthcare, Milwaukee, WI, USA). Images were stored in a digital cine loop format and analyzed offline (EchoPac Version 11; GE Vingmed Ultrasound, Horten, Norway) using a 16-segment model (no apical cap). Standard gray scale 2D and tissue Doppler images were obtained from the apical and short-axis LV views in accordance with American Society of Echocardiography Guidelines.9 Left ventricular EF and volumes were quantified using Simpson’s biplane method, with volumes and dimensions indexed to body surface area (BSA). Intraventricular mechanical dyssynchrony was assessed using anteroseptum to posterior wall delay by speckle tracking radial strain.10 Speckle tracking analysis was performed using images acquired over 3 cardiac cycles, frame rates 44–82 frames/s, and settings adjusted to 2

optimize endocardial definition. The endocardial border was traced just within the endocardium using a point-and-click technique in end-systole. A second larger concentric circle was automatically generated and adjusted near the epicardium, so that the area of interest included the entire myocardial wall. The image was then played to fine-tune tracking by visual assessment, ensuring that all wall segments tracked appropriately throughout the cardiac cycle and that the sectors defining each segment were adjusted appropriately. Peak strain was recorded for each segment with adequate tracking. In the presence of mechanical dyssynchrony, peak segmental strain may occur very early or very late (after aortic valve closure, AVC) in the cardiac cycle. For each individual segment peak strain, which could be either thickening or shortening, or lengthening or thinning, was recorded. Segmental radial and circumferential strain were assessed in the 6 basal and 6 mid-myocardial segments from parasternal short-axis views. Global radial and circumferential strain parameters were calculated as an average of the 12 basal and mid-myocardial segments. Global longitudinal strain was calculated as an average of 16 segments form the 3 apical long-axis views of the LV. The time taken to generate long-axis strain data was ~4–5 minutes, and to generate short-axis strain data ~4 minutes. The apical anterolateral and inferolateral segments were combined into an apical lateral segment, and the apical anteroseptal and inferoseptal segments were combined into an apical septal segment in keeping with a 16-segment model. The global strain was not recorded when speckle tracking data were unsatisfactory in more than 3 of the 6 myocardial segments in either apical- or short-axis views. A single investigator blinded to magnetic resonance imaging (MRI) data and clinical outcome performed echocardiographic analysis. All imaging studies were analyzed using the same 16-segment model and standardized anatomic markers, including right ventricular insertion point, to provide consistent segmental assignment. Reproducibility was assessed in 10 randomly selected cases for intra-observer variability; inter-observer variability was assessed in the same 10 cases by a second investigator experienced in strain analysis. Examples of segmental strain analysis and comparison with LGE-CMR are shown in Figure 1. MRI Image Acquisition and Analysis: Contrast-enhanced MRI was performed using a commercially available scanner (Siemens 1.5T Avanto, Erlangen, Germany) within 1 week of the echocardiographic study. Standard delayed enhancement images were acquired using a

Characterizing Dysfunctional Myocardium for CRT




Figure 1. Examples of differing scar distributions on CMR and accompanying segmental strain analysis. A. Demonstrates a patient with ischemic cardiomyopathy and previous anterior myocardial infarction with full thickness scar in the anteroseptal and anterior segments (arrows), corresponding low amplitude RS is seen in these segments (arrows). CS is also reduced in the anteroseptal segment, but less so anteriorly. B. Demonstrates a patient with dilated cardiomyopathy and no segments of scar or fibrosis on CMR. None of the myocardial segments demonstrate strain values below the derived thresholds for transmural scar (dotted line). There is significant mechanical dyssynchrony (anteroseptal to posterior wall delay >130 ms) with thickening of the inferolateral wall late in the cardiac cycle after aortic valve closure. C. Demonstrates a patient with ischemic heart disease and full thickness scar in the inferior segment and regions of nontransmural scar in the inferolateral wall (arrows). Peak RS falls below the threshold for full thickness scar in the inferior wall, peak strain in the remaining segments is greater than 7.4%. The inferior and inferoseptal segments demonstrate systolic stretch by CS. CMR = cardiac magnetic resonance.

segmented inversion recovery gradient echo sequence (repetition time [TR] 700 ms; echo time [TE] 4.9 ms; inversion time [TI] 220-270 ms: matrix 256 9 154; in-plane resolution 1.9 9 1.5 9 6 mm) 5 minutes after administration of contrast (Gadovist 0.2 mL/Kg; Bayer Healthcare, Leverkusen, Germany). The field of view was adjusted according to the patient’s body habitus, but was typically 330–400 mm. Late enhancement was visualized using an inversion recovery gradient echo sequence. The LV was divided into a short-axis stack in sequential 6 mm slices with a 4 mm inter-slice gap from the atrioventricular ring to the apex. A T1 scout was used to identify the optimum inver-

sion time to null the myocardium and the images analyzed (QMass MR 7.2 Delayed Signal Intensity (DSI) Analysis; Medis, Leiden, The Netherlands) to quantify the amount of scar tissue present. After manually tracing the endocardial and epicardial contours, positive hyperenhancement was defined using single threshold analysis (by full width half maximum method)11 and reported using a 16-segment model. This method is dependent on userselected areas of normal myocardium, and careful attention was taken to reference region selection. This approach has been previously shown to be the most reproducible technique for LGE quantification, with reported inter-class 3

Kydd, et al.

correlation coefficients >0.95 indicating minimal inter- and intra-observer variability.12,13 The following parameters were assessed: (1) the number of segments with transmural scar defined as hyperenhancement exceeding >50% of the myocardial wall thickness in any segment; (2) segmental hyperenhancement graded as 50% hyperenhancement area in comparison to segmental myocardial area; and (3) global LV scar volume expressed as a percentage of total myocardial volume. A segment with scar area of >50% was considered nonviable.14 CRT Implantation and Response: Cardiac resynchronization therapy was performed using a standard transvenous implantation technique. Both echocardiography and MRI data were available to the implanter, and the LV lead was preferentially targeted to the latest segment of activation, avoiding segments of low amplitude strain (50% scar was assessed using positive and negative predictive values. A receiver-operating characteristic (ROC) curve was used to assess the ability to distinguish between the presence and absence of >50% segmental scar in this cohort, with an optimal cutoff value selected to maximize the Youden index. Inter- and intraobserver reliability was reported using Pearson correlation. Statistical significance was considered when the P < 0.05 and analysis was performed using SPSS version 19 (SPSS Inc, Chicago, IL, USA). Results: The clinical characteristics of the 26 patients studied are given in Table I. Echocardiographic image quality was sufficient in all cases for inclusion in our study and enabled speckle tracking strain analysis in 93% (N = 291) segments for radial strain and 90% (N = 281) for circumferential strain. Multidirectional global strain values could be recorded for all subjects. All

TABLE I Baseline Clinical, Echocardiographic, and CMR Characteristics All Patients N = 26 Age (years) Male Diabetes mellitus QRS (ms) Echocardiography EF (%) LVEDV (mL) LVEDVi (mL/m2) LVESV (mL) LVESVi (mL/m2) STE AS-P delay (ms) MRI Scar volume% (% of total LV volume) Number of full thickness scar segments (per patient)

71  10 18(69) 11(42) 149  22 21.8 188 93 148 73 198


6.6 47 21 42 20 166

14.7  9 53

CRT Responders N = 12 71  10 7(58) 5(42) 149  22 24 184 92 140.7 70.5 212


6.6 48 24 43 23 95

13.4  6 4.6  3

CRT Nonresponders N = 14


72  9 11(78) 6(43) 146  23

0.8 0.4 0.95 0.73


6.4 47 21 43 19 11

0.13 0.7 0.8 0.4 0.6 0.07

15.7  11 64

0.52 0.3

20 192 94 155 75 164

CRT = cardiac resynchronization therapy; STE = speckle tracking echocardiography; LV = left ventricle; CMR = cardiac magnetic resonance; EF = ejection fraction; LVESV = LV end systolic volume; LVESVi = LV end systolic volume index; AS-P = anteroseptalposterior; MRI = magnetic resonance imaging. *P-value responders vs. nonresponders.


Characterizing Dysfunctional Myocardium for CRT

segments (N = 416) were analyzable in the CMR datasets. Scar Burden Defined by Gadolinium-Enhanced CMR: Of the 26 patients studied 19 (72%) had 1 or more segments with transmural scar on LGECMR. Transmural scar was present in 32% of myocardial segments; 49% of the apical segments (not included in echocardiographic assessment), and 27% of the nonapical segments. When nonapical segments were further examined according to segmental scar proportion 43% contained no scar (50% scar. The mean total LV scar volume was similar in CRT responders and nonresponders (13.4% vs. 15.7%, P = 0.52). Relationship between Segmental Scar and Peak Strain: The relationship between 2D short-axis strain and segmental myocardial scar on CMR is shown in Figure 2. Peak segmental radial strain was inversely related to segmental scar area. When stratified according to scar area 50% mean peak segmental strain values did not differ between segments containing 50% scar were reduced when compared with segments with 50% segmental scar area had a strong negative predictive value of 98%, but low positive predictive value of 14%. In this cohort the derived optimal cutoff value to define segmental scar was 5% was optimum to predict CRT nonresponse area under the curve (AUC 0.8; 95% CI 0.6–0.98; sensitivity 67%, specificity 90%).

TABLE II Segmental Strain Values by 2D Speckle Tracking According to Extent of Scar

Radial strain% (N) Circumferential strain% (N)

No Scar 50% Scar 4  9.6 (17) 1.86  8.5 (17)

P Value*

Post hoc Test No Scar vs. 1–49% Scar

Post hoc Test 1–49% Scar vs. >50% Scar

50% segmental area) scar. Electrical

Characterizing Dysfunctional Myocardium for CRT

conduction through scarred myocardial segments is slower and this may reduce the effect on any surrounding viable myocardium and therefore efficacy of CRT, and full thickness scar does not contract and there is no prospect of improvement following revascularization. If no alternative suitable pacing site is available, that is an adjacent segment with strain amplitude of >10% then further assessment should be considered to avoid a potentially detrimental outcome.24,25 In a population of patients with established CRT indications, however, the primary aim of this approach was to optimize implantation technique, rather than deny this important treatment modality. Limitations: We recognize that this is a relatively small retrospective observational study limited to patients with ischemic cardiomyopathy undergoing clinically indicated CMR. Overall CRT response rate was low which may reflect particularly advance heart failure in our patient population prompting additional CMR investigation. The relationship between segmental strain at LV lead location and CRT response could not be studied due to small patient numbers and the absence of segments with transmural or >50% scar area at the LV lead tip. The number of LV segments with a small extent of hyperenhancement (1–49%) may have affected sensitivity and specificity of the shortaxis strain parameters, and the ability of the technique to differentiate partial (50%) segmental scar. We also acknowledge that there is potential for misalignment in segmental data analysis for both echocardiography and CMR, and adopted a systematic approach to our analysis to minimize this. Conclusion: Speckle tracking echocardiography offers the potential to characterize the myocardium and define regions of the dysfunctional LV that are free from myocardial scar. Our findings suggest that in patients with significant LV dysfunction due to coronary artery disease, a peak radial strain >10% is indicative that there is absence of full thickness scar, and that the segment could provide the target for the LV lead during CRT. In contrast, reduced strain at the site of preferred LV lead placement indicates scar or dysfunctional myocardium that may limit CRT response. Acknowledgments: This work was supported in part by the Cambridge NIHR Comprehensive Biomedical Research Centre.

References 1. Chung ES, Leon AR, Tavazzi L, et al: Results of the predictors of response to CRT (PROSPECT) trial. Circulation 2008;117:2608–2616. 2. Becker M, Zwicker C, Kaminski M, et al: Dependency of cardiac resynchronization therapy on myocardial viability at the LV lead position. JACC Cardiovasc Imaging 2011;4:366–374. 3. Khan FZ, Virdee MS, Palmer CR, et al: Targeted left ventricular lead placement to guide cardiac resynchronization therapy: The TARGET study: A randomized, controlled trial. J Am Coll Cardiol 2012;59:1509–1518. 4. Ypenburg C, Roes SD, Bleeker GB, et al: Effect of total scar burden on contrast-enhanced magnetic resonance imaging on response to cardiac resynchronization therapy. Am J Cardiol 2007;99:657–660. 5. Becker M, Hoffmann R, Kuhl HP, et al: Analysis of myocardial deformation based on ultrasonic pixel tracking to determine transmurality in chronic myocardial infarction. Eur Heart J 2006;27:2560–2566. 6. D’Andrea A, Caso P, Scarafile R, et al: Effects of global longitudinal strain and total scar burden on response to cardiac resynchronization therapy in patients with ischaemic dilated cardiomyopathy. Eur J Heart Fail 2009;11:58–67. 7. Khan FZ, Virdee MS, Read PA, et al: Effect of low-amplitude two-dimensional radial strain at left ventricular pacing sites on response to cardiac resynchronization therapy. J Am Soc Echocardiogr 2010;23:1168–1176. 8. Bourantas CV, Nikitin NP, Loh HP, et al: Prevalence of scarred and dysfunctional myocardium in patients with heart failure of ischaemic origin: A cardiovascular magnetic resonance study. J Cardiovasc Magn Reson 2011;13:53. 9. Lang RM, Bierig M, Devereux RB, et al: Recommendations for chamber quantification: A report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005;18:1440–1463. 10. Suffoletto MS, Dohi K, Cannesson M, et al: Novel speckle-tracking radial strain from routine black-andwhite echocardiographic images to quantify dyssynchrony and predict response to cardiac resynchronization therapy. Circulation 2006;113:960–968. 11. Schuleri KH, Centola M, George RT, et al: Characterization of peri-infarct zone heterogeneity by contrastenhanced multidetector computed tomography: A comparison with magnetic resonance imaging. J Am Coll Cardiol 2009;53:1699–1707. 12. Flett AS, Hasleton J, Cook C, et al: Evaluation of techniques for the quantification of myocardial scar of differing etiology using cardiac magnetic resonance. JACC Cardiovasc Imaging 2011;4:150–156. 13. Gao P, Yee R, Gula L, et al: Prediction of arrhythmic events in ischemic and dilated cardiomyopathy patients referred for implantable cardiac defibrillator: Evaluation of multiple scar quantification measures for late gadolinium enhancement magnetic resonance imaging. Circ Cardiovasc Imaging 2012;5:448–456. 14. Kim RJ, Wu E, Rafael A, et al: The use of contrastenhanced magnetic resonance imaging to identify reversible myocardial dysfunction. N Engl J Med 2000;343:1445–1453. 15. Becker M, Lenzen A, Ocklenburg C, et al: Myocardial deformation imaging based on ultrasonic pixel tracking to identify reversible myocardial dysfunction. J Am Coll Cardiol 2008;51:1473–1481. 16. Chan J, Hanekom L, Wong C, et al: Differentiation of subendocardial and transmural infarction using


Kydd, et al.


18. 19.



two-dimensional strain rate imaging to assess short-axis and long-axis myocardial function. J Am Coll Cardiol 2006;48:2026–2033. Delgado V, van Bommel RJ, Bertini M, et al: Relative merits of left ventricular dyssynchrony, left ventricular lead position, and myocardial scar to predict long-term survival of ischemic heart failure patients undergoing cardiac resynchronization therapy. Circulation 2011;123:70–78. Nesser HJ, Winter S: Speckle tracking in the evaluation of left ventricular dyssynchrony. Echocardiography 2009;26:324–336. Kansal MM, Panse PM, Abe H, et al: Relationship of contrast-enhanced magnetic resonance imaging-derived intramural scar distribution and speckle tracking echocardiography-derived left ventricular two-dimensional strains. Eur Heart J Cardiovasc Imaging 2012;13:152–158. Roes SD, Mollema SA, Lamb HJ, et al: Validation of echocardiographic two-dimensional speckle tracking longitudinal strain imaging for viability assessment in patients with chronic ischemic left ventricular dysfunction and comparison with contrast-enhanced magnetic resonance imaging. Am J Cardiol 2009;104:312–317.

21. Iacoviello M, Puzzovivo A, Guida P, et al: Independent role of left ventricular global longitudinal strain in predicting prognosis of chronic heart failure patients. Echocardiography 2013;30:803–811. 22. Chalil S, Foley PW, Muyhaldeen SA, et al: Late gadolinium enhancement-cardiovascular magnetic resonance as a predictor of response to cardiac resynchronization therapy in patients with ischaemic cardiomyopathy. Europace 2007;9:1031–1037. 23. White JA, Yee R, Yuan X, et al: Delayed enhancement magnetic resonance imaging predicts response to cardiac resynchronization therapy in patients with intraventricular dyssynchrony. J Am Coll Cardiol 2006;48:1953–1960. 24. Ypenburg C, van Bommel RJ, Delgado V, et al: Optimal left ventricular lead position predicts reverse remodeling and survival after cardiac resynchronization therapy. J Am Coll Cardiol 2008;52:1402–1409. 25. Bleeker GB, Kaandorp TA, Lamb HJ, et al: Effect of posterolateral scar tissue on clinical and echocardiographic improvement after cardiac resynchronization therapy. Circulation 2006;113:969–976.

Utility of speckle tracking echocardiography to characterize dysfunctional myocardium in patients with ischemic cardiomyopathy referred for cardiac resynchronization therapy.

Assessment of transmural scar at the site of latest mechanical activation is relevant to maximize outcomes in cardiac resynchronization therapy (CRT)...
238KB Sizes 0 Downloads 0 Views