- Email: [email protected]
Relation of Biventricular Strain and Dyssynchrony in Repaired Tetralogy of Fallot Measured by Cardiac Magnetic Resonance to Death and Sustained Ventricular Tachycardia
Relation of Biventricular Strain and Dyssynchrony in Repaired Tetralogy of Fallot Measured by Cardiac Magnetic Resonance to Death and Sustained Ventricular Tachycardia Thomas J. Moon, MDa,b,*, Nadine Choueiter, MDa,b, Tal Geva, MDa,b, Anne Marie Valente, MDa,b, Kimberlee Gauvreau, ScDa,b, and David M. Harrild, MD, PhDa,b Ventricular strain and dyssynchrony can be measured in patients with repaired tetralogy of Fallot (TOF), but their effect on clinical outcomes is poorly understood. The purpose of this study was to investigate if ventricular strain and dyssynchrony measured by cardiac magnetic resonance feature tracking are associated with death and sustained ventricular tachycardia. Patients with TOF who died or had ventricular tachycardia (TOF case, n [ 16) were compared with age-matched patients with TOF with no adverse outcome (TOF control, n [ 32). For each patient, midventricular short-axis and 4-chamber cine steady-state free precession images were analyzed using cardiac magnetic resonance feature-tracking software. Peak left ventricular (LV) and right ventricular (RV) global circumferential and longitudinal strain and indexes of dyssynchrony were compared between groups. Compared with the TOF control group, median strain values were significantly lower for the TOF case group for both the LV (circumferential: 17% vs 23%, p [ 0.003; longitudinal: 13% vs 18%, p <0.001) and the RV (circumferential: 10% vs 16%, p [ 0.001; longitudinal: 11% vs 18%, p <0.001). In a multivariate model including strain and dyssynchrony parameters, RV and LV longitudinal strain were strongly associated with the adverse outcome (p [ 0.003 and 0.04, respectively; area under the curve [ 0.92). No differences in ventricular dyssynchrony were identified between the groups. In conclusion, patients with TOF in this cohort who experienced adverse outcomes had lower values of all strain parameters than those who did not, and impaired longitudinal strain of both ventricles was strongly associated with adverse clinical outcomes. Ó 2015 Elsevier Inc. All rights reserved. (Am J Cardiol 2015;115:676e680)
Despite excellent results of surgical management of tetralogy of Fallot (TOF) and good early clinical outcomes, the rates of exercise intolerance, major arrhythmias, and death increase substantially beginning in the third decade of life.1e3 Traditional factors linked to poor long-term outcomes in this population include older age at complete repair,3,4 longer time since surgery,5 prolonged QRS duration,6 and left ventricular (LV) or right ventricular (RV) dysfunction, measured by ejection fraction.7e9 Increasingly, strain and synchrony have been measured using speckle-tracking echocardiography to evaluate regional ventricular function allowing new insights into cardiac mechanics. Echocardiographic views of the RV, however, are often limited in patients with TOF, and RV circumferential strain frequently cannot be determined reliably by ultrasound-based techniques. Cardiac magnetic resonance (CMR) imaging overcomes most of these limitations to allow accurate and reproducible evaluation of the a Department of Cardiology, Boston Children’s Hospital, Boston, Massachusetts and bDepartment of Pediatrics, Harvard Medical School, Boston, Massachusetts. Manuscript received August 27, 2014; revised manuscript received and accepted December 1, 2014. This work was supported by the Higgins Family Noninvasive Cardiac Imaging Research Fund. See page 679 for disclosure information. *Corresponding author: Tel: (904) 633-4120; fax: (904) 633-4111. E-mail address: [email protected]fl.edu (T.J. Moon).
0002-9149/15/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjcard.2014.12.024
RV.10 Speckle trackingebased techniques have been applied to CMR data,11,12 and these CMR-based techniques have been termed “feature tracking.” The goals of this investigation were to use these new tools to (1) determine whether LV and RV strain measurements in patients with repaired TOF are associated with poor clinical outcomes such as death and sustained ventricular tachycardia (VT) and (2) evaluate whether indexes of dyssynchrony are related to the same clinical outcomes. Methods This was a single-center, retrospective, case-control design. The study was approved by the Committee on Clinical Investigation at Boston Children’s Hospital. Patients were identified for review from an internal institutional database. The clinical outcome in this study was death or sustained VT occurring after CMR imaging. The study group consisted of TOF case and TOF control patients. Inclusion criteria for the TOF case patients were (1) history of surgical TOF repair, (2) death or documented sustained VT, and (3) CMR from 2000 to 2011, with high-quality short- and longaxis cine images, before death or VT. TOF control patients were selected from a pool of patients with TOF with no adverse outcomes (defined previously). Case and control patients with 10-year interval between TOF repair and CMR evaluation were excluded, as were patients with www.ajconline.org
Congenital Heart Disease/Biventricular Strain and Dyssynchrony in TOF
677
Figure 1. Examples of segmental circumferential strain curves used for analysis. Time to peak and peak strain values are shown for the 6 segments of the contour; global peak strain is calculated as the average of the 6 values. Longitudinal strain values were calculated from the 4-chamber plane in a similar fashion.
Table 1 Demographic, electrocardiographic and cardiac magnetic resonance data Variable
TOF Case (n¼16)
TOF Control (n¼32)
p Value
Male/female 10/6 16/16 0.54 Age at surgical repair (years) 7 (0-31) 6 (0-49) 0.43 Age at CMR (years) 38 (9-62) 42 (13-64) 0.41 Time from complete repair 29 (6-40) 32 (13-46) 0.10 to CMR (years) Electrocardiographic characteristics QRS duration (ms) 158 (112-208) 152 (72-188) 0.14 (n¼16, 30) Heart rate (beats/minute) 81 (58-142) 72 (46-117) 0.02 CMR variables Body surface area (m2) 2.0 (0.8-2.4) 1.8 (1.1-2.4) 0.10 Pulmonary regurgitation 22 (0-52)* 32 (0-67) 0.71 fraction (%) RV end-diastolic volume 146 (75-247) 127 (56-287) 0.35 index (ml/m2) RV end-systolic volume 103 (40-195) 65 (27-192) 0.03 index (ml/m2) RV ejection fraction (%) 40 (21-52) 50 (33-64) <0.001 RV mass index (g/m2) 42 (24-81) 28 (0-56) <0.001 RV mass/volume (g/ml) 0.28 (0.14-0.67) 0.20 (0.14-0.41)† 0.01 LV end-diastolic volume 87 (61-139) 0.52 78 (63-209)z 2 index (ml/m ) LV ejection fraction (%) 51 (40-61)z 57 (34-71) 0.02 2 73 (40-102)z 53 (36-90) 0.01 LV mass index (g/m ) LV mass/volume (g/m2) 0.80 (0.36-1.4)z 0.61 (0.44-0.86) 0.03 Values shown are # (%) or median (range). * ¼ (n¼11). † ¼ (n¼31). z ¼ (n¼15).
pulmonary valve replacement after CMR. TOF control patients were individually matched 2:1 to TOF cases by date of birth (2 controls closest in age to the TOF case chosen), with the maximum allowable age difference between the matched case and control of 5 years.
Data extracted from the patients’ medical records included gender, surgical history, age at CMR, date and type of adverse outcome (in the case group), and QRS duration on electrocardiogram nearest in time to the CMR. For the TOF case group, the CMR used for analysis was the most recent study before the adverse outcome. For the TOF control group, the CMR was the most recent study available occurring at least 1 year before data analysis to provide a period of follow-up during which an adverse event could have occurred in this population. Details of the CMR protocol used in our laboratory for the assessment of patients with repaired TOF, including calculation of ventricular size and function and pulmonary regurgitation fraction, have been previously published.10 Briefly, CMR studies were performed on a commercially available 1.5-T whole-body scanner (GE Medical Systems, Milwaukee, Wisconsin; Philips Medical Systems, Best, The Netherlands), using the appropriate receiver coil for body size. CMR image analysis was performed using a commercially available workstation (Extended MR WorkSpace; Philips Medical Systems) and software (Mass and Flow; MEDIS Medical Imaging Systems, Leiden, The Netherlands). Feature-tracking analysis was performed on ventricular short-axis and 4-chamber cine images of both the RV and LV using commercially available software (Image-Arena VA 3.0 with Diogenes v. 1.1.02; TomTec Imaging Systems; Unterschleissheim, Germany). Starting at the LV midpapillary level, the short-axis slice used for analysis was the first one, proceeding in an apical direction, in which the RV myocardium was continuous (e.g., patch was no longer present). The LV was analyzed at the same level. For the 4-chamber analysis, a central slice was used. The featuretracking process was the same in short-axis and 4-chamber views. A splined curve was defined by a series of points placed manually along the endocardium at end-diastole. The analysis software then subdivided this curve into 48 points and 6 segments. The motion of these points was tracked in an automated fashion, and a peak strain value and time to peak strain for each of the 6 segments were calculated (Figure 1).
678
The American Journal of Cardiology (www.ajconline.org)
Table 2 Global strain Variable Left Ventricle Circumferential (%) Longitudinal (%) Right Ventricle Circumferential (%) Longitudinal (%)
Table 4 Multivariable analysis of factors associated with adverse outcomes TOF Case (n¼16)
TOF Control (n¼32)
p Value
17 (7-29) 13 (8-20)*
23 (9-30) 18 (10-27)
0.003 <0.001
10 (6-21) 11 (5-21)
16 (8-26) 18 (11-23)
0.001 <0.001
Including all variables: Odds Ratio RV ejection fraction (%) RV mass (g/m2)
Longitudinal RV average strain (%) Longitudinal LV average strain (%)
Left Ventricle Circumferential Standard deviation time to peak Maximum wall delay Cross-correlation delay Longitudinal Standard deviation time to peak Maximum wall delay Cross-correlation delay Right Ventricle Circumferential Standard deviation time to peak Maximum wall delay Cross-correlation delay Longitudinal Standard deviation time to peak Maximum wall delay Cross-correlation delay Interventricular Circumferential (RV - LV) time to peak Cross-correlation delay Longitudinal (RV - LV) time to peak Cross-correlation delay
TOF Control (n¼32)
p Value
39 (16-163)
43 (15-143)
0.84
96 (29-481) 82 (29-333)
112 (28-303) 75 (26-456)
0.89 0.82
80 (13-157)*
93 (12-252)
0.21
214 (26-400)* 270 (36-714)*
246 (24-730) 202 (0-975)
0.13 0.42
66 (12-165)
66 (16-135)
0.62
188 (22-474) 182 (22-725)
166 (30-344) 121 (28-825)
0.39 0.48
71 (26-230)
89 (40-184)
0.29
183 (57-530) 118 (33-588)
236 (92-466) 170 (26-756)
0.26 0.39
47 (0-185) 56 (0-203)
57 (-76-136) 33 (0-114)
0.65 0.16
23 (-146-151)* 53 (0-392)*
4 (-149-133) 34 (0-180)
0.24 0.31
(1.3, 6.1)
[ 5 g/m2 1.7 (1.1, 2.7) Area under ROC curve ¼ 0.94
Odds Ratio
Table 3 Comparison of dyssynchrony parameters TOF Case (n¼16)
2.8
0.008 0.02
Analysis limited to strain and dyssynchrony variables:
Values shown are jmedian (range)j. * ¼ (n ¼ 14).
Variable
Y 5%
95% CI p Value
95% CI p Value
Y 1%
1.6
(1.2, 2.3)
0.004
Y 1%
1.4
(1.0, 2.0)
0.03
Area under ROC curve ¼ 0.92 Outcome is occurrence of event (16 outcome events among 48 patients).
(“cross-correlation delay”), using the method described by Fornwalt et al.13 Finally, interventricular dyssynchrony was assessed using both difference in time to peak global strain (RV LV) and maximum cross-correlation delay between global strain curves for the RV and LV. Demographic, electrocardiographic, and CMR parameters (including indexes of strain and dyssynchrony) were compared between TOF case and TOF control patients. Absolute values of strain are reported, with median and maximum/minimum values listed unless otherwise noted. Categorical variables are reported as number and percentage. Comparisons between groups were made using Fisher’s exact test for categorical variables and the Wilcoxon rank sum test for continuous variables. Multivariate analysis was performed using logistic regression for the outcome occurrence of event (death or sustained VT), first considering all variables, and then including only strain and dyssynchrony variables. Stata 12.1 (StataCorp, College Station, Texas) was used for statistical analysis. A p value <0.05 was considered statistically significant.
Results
Values (ms) listed as median (range). * ¼ (n ¼ 14).
Peak global strain was calculated as the average of the 6 segmental values. Given the lack of a well-established reference method for calculating dyssynchrony, several different methods were used. The first was the SD of the time to peak strain of each of the 6 segments (referred to as “SD time to peak strain”). The second method identified the maximum time delay between the wall segment with the earliest time to peak strain and the segment with the latest time to peak strain (“maximum wall delay”). Next, cross-correlation analysis was used to calculate the maximum cross-correlation delay
The TOF case and TOF control groups included 16 and 32 patients, respectively. Among the TOF case group, 13 patients died after CMR and 3 patients had documented sustained VT. Demographic, electrocardiographic, and CMR data for the 2 groups are presented in Table 1. Global strain values and results of the dyssynchrony analyses for the 2 groups are presented in Tables 2 and 3, respectively. In a multivariate model including all candidate variables, the 2 variables most associated with an adverse event were lower RV ejection fraction and higher RV mass index (Table 4; area under the receiver operating characteristic curve ¼ 0.94). When considering only strain and dyssynchrony variables in the model, longitudinal RV and LV strain were highly associated with the adverse outcome (area under the receiver operating characteristic curve ¼ 0.92); discrimination does not differ significantly between this model and the one including all variables (p ¼ 0.61).
Congenital Heart Disease/Biventricular Strain and Dyssynchrony in TOF
Discussion In this study, we found that every measure of strain investigated (RV and LV, longitudinal and circumferential) was more impaired among patients with TOF who experienced death and VT compared with a TOF group who did not experience these outcomes. Differences in synchrony parameters between the 2 TOF groups were not identified. In a multivariate analysis of synchrony and strain parameters, longitudinal strain of both the LV and RV was highly related to the outcome of death or VT. The reasons for impaired LV and RV strain among patients with TOF are likely multifactorial and likely to be similar to the causes underlying depressed ventricular function in general. Decreased LV function is common among adult patients with TOF and has been shown to be related to duration from shunt to repair, presence of arrhythmia, and QRS duration.14 RV dilation and related dysfunction are frequently observed in this cohort as well, with impairments in function exacerbated by residual pressure-loading lesions (e.g., pulmonary valve or branch pulmonary stenosis).15 As well, a number of reports have described LV dysfunction in association with RV dilation or dysfunction, which is believed to be related to “ventricular-ventricular interactions.”8,16e18 Our results add to the list of factors known to be associated with late adverse outcomes in patients with TOF. Traditional CMR factors such as ventricular dysfunction, RV dilation, and RV mass-to-volume ratio have all been well documented as specific risk factors for poor outcomes.7e9,18,19 These findings were observed, as well, in our cohort. Reports of outcomes related to novel imaging techniques such as deformation analysis have begun to appear in the literature.20,21 For example, a recent echocardiographic study has identified LV longitudinal strain as a risk factor for arrhythmia and death in TOF.22 Using a CMR-based method capable of making high-quality measurements of deformation of both ventricles, we have not only confirmed this result but have also extended it by showing that impaired RV longitudinal strain is associated even more strongly with death and VT than reduced LV strain. Moreover, our multivariate model including only strain parameters was equally predictive of the outcome, compared with a model including all traditional CMR parameters. Previous reports of the presence and effect of dyssynchrony in repaired TOF have been variable. Fernandes et al23 did not identify a significant difference in dyssynchrony parameters in children and adolescents with repaired TOF compared with controls. Van der Hulst et al,24 however, have reported a delay in activation of the RV outlet in repaired pediatric patients with TOF, compared with normal controls. In our work, the TOF case (adverse outcome) group was not found to be significantly more dyssynchronous than the TOF control group, for any of the parameters investigated. In another previous study, however, Ortega et al20 described a relation between 1 parameter of dyssynchrony (time to peak circumferential strain) and adverse events in TOF. Once a patient with TOF has been identified to be “at risk,” the question remains regarding what, if anything, to do in an effort to mitigate that risk. Pulmonary valve replacement, either surgical or percutaneous, is a promising intervention, and many studies have investigated the potential
679
value of this procedure,25e27 with a recent report of improved LV strain after transcatheter valve replacement.28 Other patients may benefit from placement of an implantable cardioverter-defibrillator, but this intervention is not without its own risk.29 Moving forward, the myocardial parameters investigated in this work may prove useful in future efforts to develop more sensitive and specific stratification models for patients with TOF and help to identify those in greatest jeopardy, in whom the risk associated with further interventional procedures might be justified. This study was retrospective, with all the limitations inherent to this study design. The small number of patients with repaired TOF with adverse events after CMR may have limited our ability to detect subtle differences in dyssynchrony parameters. Because we chose to analyze short-axis slices in which the RV myocardium was continuous, we excluded the effect of infundibular patch on global strain and synchrony; this was done in a systematic fashion for both case and control cohorts, however, so any effect of this design choice would be expected to apply to both groups. The frame rate of the CMR cine images is modest and might reduce the ability of this method to identify differences between groups; again, however, this should apply equally to the case and control groups. Disclosures Tal Geva, MD is a consultant for Medtronic. All other authors have no conflicts of interest. 1. Apitz C, Webb GD, Redington AN. Tetralogy of Fallot. Lancet 2009;374:1462e1471. 2. Bacha EA, Scheule AM, Zurakowski D, Erickson LC, Hung J, Lang P, Mayer JE, del Nido PJ, Jonas RA. Long-term results after early primary repair of tetralogy of Fallot. J Thorac Cardiovasc Surg 2001;122:154e161. 3. Nollert G, Fischlein T, Bouterwek S, Bohmer C, Klinner W, Reichart B. Long-term survival in patients with repair of tetralogy of Fallot: 36year follow-up of 490 survivors of the first year after surgical repair. J Am Coll Cardiol 1997;30:1374e1383. 4. Murphy JG, Gersh BJ, Mair DD, Fuster V, McGoon MD, Ilstrup DM, McGoon DC, Kirklin JW, Danielson GK. Long-term outcome in patients undergoing surgical repair of tetralogy of Fallot. N Engl J Med 1993;329:593e599. 5. Alexiou C, Mahmoud H, Al-Khaddour A, Gnanapragasam J, Salmon AP, Keeton BR, Monro JL. Outcome after repair of tetralogy of Fallot in the first year of life. Ann Thorac Surg 2001;71:494e500. 6. Gatzoulis MA, Till JA, Somerville J, Redington AN. Mechanoelectrical interaction in tetralogy of Fallot. QRS prolongation relates to right ventricular size and predicts malignant ventricular arrhythmias and sudden death. Circulation 1995;92:231e237. 7. Ghai A, Silversides C, Harris L, Webb GD, Siu SC, Therrien J. Left ventricular dysfunction is a risk factor for sudden cardiac death in adults late after repair of tetralogy of Fallot. J Am Coll Cardiol 2002;40:1675e1680. 8. Geva T, Sandweiss BM, Gauvreau K, Lock JE, Powell AJ. Factors associated with impaired clinical status in long-term survivors of tetralogy of Fallot repair evaluated by cardiac magnetic resonance imaging. J Am Coll Cardiol 2004;43:1068e1074. 9. Knauth AL, Gauvreau K, Powell AJ, Landzberg MJ, Walsh EP, Lock JE, del Nido PJ, Geva T. Ventricular size and function assessed by cardiac MRI predict major adverse clinical outcomes late after tetralogy of Fallot repair. Heart 2008;94:211e216. 10. Geva T. Repaired tetralogy of Fallot: the roles of cardiovascular magnetic resonance in evaluating pathophysiology and for pulmonary valve replacement decision support. J Cardiovasc Magn Reson 2011;13:9. 11. Schuster A, Kutty S, Padiyath A, Parish V, Gribben P, Danford DA, Makowski MR, Bigalke B, Beerbaum P, Nagel E. Cardiovascular
680
12.
13.
14.
15.
16.
17.
18.
19.
The American Journal of Cardiology (www.ajconline.org) magnetic resonance myocardial feature tracking detects quantitative wall motion during dobutamine stress. J Cardiovasc Magn Reson 2011;13:58. Harrild DM, Marcus E, Hasan B, Alexander ME, Powell AJ, Geva T, McElhinney DB. Impact of transcatheter pulmonary valve replacement on biventricular strain and synchrony assessed by cardiac magnetic resonance feature tracking. Circ Cardiovasc Interv 2013;6: 680e687. Fornwalt BK, Arita T, Bhasin M, Voulgaris G, Merlino JD, León AR, Fyfe DA, Oshinski JN. Cross-correlation quantification of dyssynchrony: a new method for quantifying the synchrony of contraction and relaxation in the heart. J Am Soc Echocardiogr 2007;20:1330e1337. Broberg CS, Aboulhosn J, Mongeon FP, Kay J, Valente AM, Khairy P, Earing MG, Opotowsky AR, Lui G, Gersony DR, Cook S, Ting JG, Webb G, Gurvitz MZ; Alliance for Adult Research in Congenital Cardiology (AARCC). Prevalence of left ventricular systolic dysfunction in adults with repaired tetralogy of Fallot. Am J Cardiol 2011;107:1215e1220. Davlouros PA, Kilner PJ, Hornung TS, Li W, Francis JM, Moon JC, Smith GC, Tat T, Pennell DJ, Gatzoulis MA. Right ventricular function in adults with repaired tetralogy of Fallot assessed with cardiovascular magnetic resonance imaging: detrimental role of right ventricular outflow aneurysms or akinesia and adverse right-to-left ventricular interaction. J Am Coll Cardiol 2002;40:2044e2052. Cheung EW, Liang XC, Lam WW, Cheung YF. Impact of right ventricular dilation on left ventricular myocardial deformation in patients after surgical repair of tetralogy of Fallot. Am J Cardiol 2009;104: 1264e1270. Kempny A, Diller GP, Orwat S, Kaleschke G, Kerckhoff G, Bunck ACh, Maintz D, Baumgartner H. Right ventricular-left ventricular interaction in adults with tetralogy of Fallot: a combined cardiac magnetic resonance and echocardiographic speckle tracking study. Int J Cardiol 2012;154:259e264. Tobler D, Crean AM, Redington AN, Van Arsdell GS, Caldarone CA, Nanthakumar K, Stambach D, Dos L, Wintersperger BJ, Oechslin EN, Silversides CK, Wald RM. The left heart after pulmonary valve replacement in adults late after tetralogy of Fallot repair. Int J Cardiol 2012;160:165e170. Valente AM, Gauvreau K, Assenza GE, Babu-Narayan SV, Schreier J, Gatzoulis MA, Groenink M, Inuzuka R, Kilner PJ, Koyak Z, Landzberg MJ, Mulder B, Powell AJ, Wald R, Geva T. Contemporary predictors of death and sustained ventricular tachycardia in patients with repaired tetralogy of Fallot enrolled in the INDICATOR cohort. Heart 2014;100:247e253.
20. Ortega M, Treidman JK, Geva T, Harrild DM. Relation of left ventricular dyssynchrony measured by cardiac magnetic resonance tissue tracking in repaired tetralogy of Fallot to ventricular tachycardia and death. Am J Cardiol 2011;107:1535e1540. 21. Su HM, Lin TH, Hsu PC, Lee WH, Chu CY, Lee CS, Voon WC, Lai WT, Sheu SH. Global left ventricular systolic strain as a major predictor of cardiovascular events in patients with atrial fibrillation. Heart 2013;99:1588e1596. 22. Diller GP, Kempny A, Liodakis E, Alonso-Gonzalez R, Inuzuka R, Uebing A, Orwat S, Dimopoulos K, Swan L, Li W, Gatzoulis MA, Baumgartner H. Left ventricular longitudinal function predicts lifethreatening ventricular arrhythmia and death in adults with repaired tetralogy of Fallot. Circulation 2012;125:2440e2446. 23. Fernandes FP, Manlhiot C, Roche SL, Grosse-Wortmann L, Slorach C, McCrindle BW, Mertens L, Kantor PF, Friedberg MK. Impaired left ventricular myocardial mechanics and their relation to pulmonary regurgitation, right ventricular enlargement and exercise capacity in asymptomatic children after repair of tetralogy of Fallot. J Am Soc Echocardiogr 2012;25:494e503. 24. van der Hulst AE, Roest AA, Delgado V, Holman ER, de Roos A, Blom MA, Bax JJ. Relationship between temporal sequences of right ventricular deformation and right ventricular performance in patients with corrected tetralogy of Fallot. Heart 2011;97:231e236. 25. Kutty S, Deatsman SL, Russell D, Nugent ML, simpson PM, Frommelt PC. Pulmonary valve replacement improves but does not normalize right ventricular mechanics in repaired congenital heart disease: a comparative assessment using velocity vector imaging. J Am Soc Echocardiogr 2008;21:1216e1221. 26. Knirsch W, Dodge-Khatami A, Kadner A, Kretschmar O, Steiner J, Böttler P, Kececioglu D, Harpes P, Valsangiacomo Buechel ER. Assessment of myocardial function in pediatric patients with operated tetralogy of Fallot: preliminary results with 2D strain echocardiography. Pediatr Cardiol 2008;29:718e725. 27. Moiduddi N, Asoh K, Slorach C, Benson LN, Friedberg MK. Effect of transcatheter pulmonary valve implantation on short-term right ventricular function as determined by two-dimensional speckle tracking strain and strain rate imaging. Am J Cardiol 2009;104:862e867. 28. Tzemos N, Harris L, Carasso S, Subira LD, Greutmann M, Provost Y, Redington AN, Rakowski H, Siu SC, Silversides CK. Adverse left ventricular mechanics in adults with repaired tetralogy of Fallot. Am J Cardiol 2009;103:420e425. 29. Witte KK, Pepper CB, Cowan JC, Thomson JD, English KM, Blackburn ME. Implantable cardioverter-defibrillator therapy in adult patients with tetralogy of Fallot. Europace 2008;10:926e930.
Despite excellent results of surgical management of tetralogy of Fallot (TOF) and good early clinical outcomes, the rates of exercise intolerance, major arrhythmias, and death increase substantially beginning in the third decade of life.1e3 Traditional factors linked to poor long-term outcomes in this population include older age at complete repair,3,4 longer time since surgery,5 prolonged QRS duration,6 and left ventricular (LV) or right ventricular (RV) dysfunction, measured by ejection fraction.7e9 Increasingly, strain and synchrony have been measured using speckle-tracking echocardiography to evaluate regional ventricular function allowing new insights into cardiac mechanics. Echocardiographic views of the RV, however, are often limited in patients with TOF, and RV circumferential strain frequently cannot be determined reliably by ultrasound-based techniques. Cardiac magnetic resonance (CMR) imaging overcomes most of these limitations to allow accurate and reproducible evaluation of the a Department of Cardiology, Boston Children’s Hospital, Boston, Massachusetts and bDepartment of Pediatrics, Harvard Medical School, Boston, Massachusetts. Manuscript received August 27, 2014; revised manuscript received and accepted December 1, 2014. This work was supported by the Higgins Family Noninvasive Cardiac Imaging Research Fund. See page 679 for disclosure information. *Corresponding author: Tel: (904) 633-4120; fax: (904) 633-4111. E-mail address: [email protected]fl.edu (T.J. Moon).
0002-9149/15/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjcard.2014.12.024
RV.10 Speckle trackingebased techniques have been applied to CMR data,11,12 and these CMR-based techniques have been termed “feature tracking.” The goals of this investigation were to use these new tools to (1) determine whether LV and RV strain measurements in patients with repaired TOF are associated with poor clinical outcomes such as death and sustained ventricular tachycardia (VT) and (2) evaluate whether indexes of dyssynchrony are related to the same clinical outcomes. Methods This was a single-center, retrospective, case-control design. The study was approved by the Committee on Clinical Investigation at Boston Children’s Hospital. Patients were identified for review from an internal institutional database. The clinical outcome in this study was death or sustained VT occurring after CMR imaging. The study group consisted of TOF case and TOF control patients. Inclusion criteria for the TOF case patients were (1) history of surgical TOF repair, (2) death or documented sustained VT, and (3) CMR from 2000 to 2011, with high-quality short- and longaxis cine images, before death or VT. TOF control patients were selected from a pool of patients with TOF with no adverse outcomes (defined previously). Case and control patients with 10-year interval between TOF repair and CMR evaluation were excluded, as were patients with www.ajconline.org
Congenital Heart Disease/Biventricular Strain and Dyssynchrony in TOF
677
Figure 1. Examples of segmental circumferential strain curves used for analysis. Time to peak and peak strain values are shown for the 6 segments of the contour; global peak strain is calculated as the average of the 6 values. Longitudinal strain values were calculated from the 4-chamber plane in a similar fashion.
Table 1 Demographic, electrocardiographic and cardiac magnetic resonance data Variable
TOF Case (n¼16)
TOF Control (n¼32)
p Value
Male/female 10/6 16/16 0.54 Age at surgical repair (years) 7 (0-31) 6 (0-49) 0.43 Age at CMR (years) 38 (9-62) 42 (13-64) 0.41 Time from complete repair 29 (6-40) 32 (13-46) 0.10 to CMR (years) Electrocardiographic characteristics QRS duration (ms) 158 (112-208) 152 (72-188) 0.14 (n¼16, 30) Heart rate (beats/minute) 81 (58-142) 72 (46-117) 0.02 CMR variables Body surface area (m2) 2.0 (0.8-2.4) 1.8 (1.1-2.4) 0.10 Pulmonary regurgitation 22 (0-52)* 32 (0-67) 0.71 fraction (%) RV end-diastolic volume 146 (75-247) 127 (56-287) 0.35 index (ml/m2) RV end-systolic volume 103 (40-195) 65 (27-192) 0.03 index (ml/m2) RV ejection fraction (%) 40 (21-52) 50 (33-64) <0.001 RV mass index (g/m2) 42 (24-81) 28 (0-56) <0.001 RV mass/volume (g/ml) 0.28 (0.14-0.67) 0.20 (0.14-0.41)† 0.01 LV end-diastolic volume 87 (61-139) 0.52 78 (63-209)z 2 index (ml/m ) LV ejection fraction (%) 51 (40-61)z 57 (34-71) 0.02 2 73 (40-102)z 53 (36-90) 0.01 LV mass index (g/m ) LV mass/volume (g/m2) 0.80 (0.36-1.4)z 0.61 (0.44-0.86) 0.03 Values shown are # (%) or median (range). * ¼ (n¼11). † ¼ (n¼31). z ¼ (n¼15).
pulmonary valve replacement after CMR. TOF control patients were individually matched 2:1 to TOF cases by date of birth (2 controls closest in age to the TOF case chosen), with the maximum allowable age difference between the matched case and control of 5 years.
Data extracted from the patients’ medical records included gender, surgical history, age at CMR, date and type of adverse outcome (in the case group), and QRS duration on electrocardiogram nearest in time to the CMR. For the TOF case group, the CMR used for analysis was the most recent study before the adverse outcome. For the TOF control group, the CMR was the most recent study available occurring at least 1 year before data analysis to provide a period of follow-up during which an adverse event could have occurred in this population. Details of the CMR protocol used in our laboratory for the assessment of patients with repaired TOF, including calculation of ventricular size and function and pulmonary regurgitation fraction, have been previously published.10 Briefly, CMR studies were performed on a commercially available 1.5-T whole-body scanner (GE Medical Systems, Milwaukee, Wisconsin; Philips Medical Systems, Best, The Netherlands), using the appropriate receiver coil for body size. CMR image analysis was performed using a commercially available workstation (Extended MR WorkSpace; Philips Medical Systems) and software (Mass and Flow; MEDIS Medical Imaging Systems, Leiden, The Netherlands). Feature-tracking analysis was performed on ventricular short-axis and 4-chamber cine images of both the RV and LV using commercially available software (Image-Arena VA 3.0 with Diogenes v. 1.1.02; TomTec Imaging Systems; Unterschleissheim, Germany). Starting at the LV midpapillary level, the short-axis slice used for analysis was the first one, proceeding in an apical direction, in which the RV myocardium was continuous (e.g., patch was no longer present). The LV was analyzed at the same level. For the 4-chamber analysis, a central slice was used. The featuretracking process was the same in short-axis and 4-chamber views. A splined curve was defined by a series of points placed manually along the endocardium at end-diastole. The analysis software then subdivided this curve into 48 points and 6 segments. The motion of these points was tracked in an automated fashion, and a peak strain value and time to peak strain for each of the 6 segments were calculated (Figure 1).
678
The American Journal of Cardiology (www.ajconline.org)
Table 2 Global strain Variable Left Ventricle Circumferential (%) Longitudinal (%) Right Ventricle Circumferential (%) Longitudinal (%)
Table 4 Multivariable analysis of factors associated with adverse outcomes TOF Case (n¼16)
TOF Control (n¼32)
p Value
17 (7-29) 13 (8-20)*
23 (9-30) 18 (10-27)
0.003 <0.001
10 (6-21) 11 (5-21)
16 (8-26) 18 (11-23)
0.001 <0.001
Including all variables: Odds Ratio RV ejection fraction (%) RV mass (g/m2)
Longitudinal RV average strain (%) Longitudinal LV average strain (%)
Left Ventricle Circumferential Standard deviation time to peak Maximum wall delay Cross-correlation delay Longitudinal Standard deviation time to peak Maximum wall delay Cross-correlation delay Right Ventricle Circumferential Standard deviation time to peak Maximum wall delay Cross-correlation delay Longitudinal Standard deviation time to peak Maximum wall delay Cross-correlation delay Interventricular Circumferential (RV - LV) time to peak Cross-correlation delay Longitudinal (RV - LV) time to peak Cross-correlation delay
TOF Control (n¼32)
p Value
39 (16-163)
43 (15-143)
0.84
96 (29-481) 82 (29-333)
112 (28-303) 75 (26-456)
0.89 0.82
80 (13-157)*
93 (12-252)
0.21
214 (26-400)* 270 (36-714)*
246 (24-730) 202 (0-975)
0.13 0.42
66 (12-165)
66 (16-135)
0.62
188 (22-474) 182 (22-725)
166 (30-344) 121 (28-825)
0.39 0.48
71 (26-230)
89 (40-184)
0.29
183 (57-530) 118 (33-588)
236 (92-466) 170 (26-756)
0.26 0.39
47 (0-185) 56 (0-203)
57 (-76-136) 33 (0-114)
0.65 0.16
23 (-146-151)* 53 (0-392)*
4 (-149-133) 34 (0-180)
0.24 0.31
(1.3, 6.1)
[ 5 g/m2 1.7 (1.1, 2.7) Area under ROC curve ¼ 0.94
Odds Ratio
Table 3 Comparison of dyssynchrony parameters TOF Case (n¼16)
2.8
0.008 0.02
Analysis limited to strain and dyssynchrony variables:
Values shown are jmedian (range)j. * ¼ (n ¼ 14).
Variable
Y 5%
95% CI p Value
95% CI p Value
Y 1%
1.6
(1.2, 2.3)
0.004
Y 1%
1.4
(1.0, 2.0)
0.03
Area under ROC curve ¼ 0.92 Outcome is occurrence of event (16 outcome events among 48 patients).
(“cross-correlation delay”), using the method described by Fornwalt et al.13 Finally, interventricular dyssynchrony was assessed using both difference in time to peak global strain (RV LV) and maximum cross-correlation delay between global strain curves for the RV and LV. Demographic, electrocardiographic, and CMR parameters (including indexes of strain and dyssynchrony) were compared between TOF case and TOF control patients. Absolute values of strain are reported, with median and maximum/minimum values listed unless otherwise noted. Categorical variables are reported as number and percentage. Comparisons between groups were made using Fisher’s exact test for categorical variables and the Wilcoxon rank sum test for continuous variables. Multivariate analysis was performed using logistic regression for the outcome occurrence of event (death or sustained VT), first considering all variables, and then including only strain and dyssynchrony variables. Stata 12.1 (StataCorp, College Station, Texas) was used for statistical analysis. A p value <0.05 was considered statistically significant.
Results
Values (ms) listed as median (range). * ¼ (n ¼ 14).
Peak global strain was calculated as the average of the 6 segmental values. Given the lack of a well-established reference method for calculating dyssynchrony, several different methods were used. The first was the SD of the time to peak strain of each of the 6 segments (referred to as “SD time to peak strain”). The second method identified the maximum time delay between the wall segment with the earliest time to peak strain and the segment with the latest time to peak strain (“maximum wall delay”). Next, cross-correlation analysis was used to calculate the maximum cross-correlation delay
The TOF case and TOF control groups included 16 and 32 patients, respectively. Among the TOF case group, 13 patients died after CMR and 3 patients had documented sustained VT. Demographic, electrocardiographic, and CMR data for the 2 groups are presented in Table 1. Global strain values and results of the dyssynchrony analyses for the 2 groups are presented in Tables 2 and 3, respectively. In a multivariate model including all candidate variables, the 2 variables most associated with an adverse event were lower RV ejection fraction and higher RV mass index (Table 4; area under the receiver operating characteristic curve ¼ 0.94). When considering only strain and dyssynchrony variables in the model, longitudinal RV and LV strain were highly associated with the adverse outcome (area under the receiver operating characteristic curve ¼ 0.92); discrimination does not differ significantly between this model and the one including all variables (p ¼ 0.61).
Congenital Heart Disease/Biventricular Strain and Dyssynchrony in TOF
Discussion In this study, we found that every measure of strain investigated (RV and LV, longitudinal and circumferential) was more impaired among patients with TOF who experienced death and VT compared with a TOF group who did not experience these outcomes. Differences in synchrony parameters between the 2 TOF groups were not identified. In a multivariate analysis of synchrony and strain parameters, longitudinal strain of both the LV and RV was highly related to the outcome of death or VT. The reasons for impaired LV and RV strain among patients with TOF are likely multifactorial and likely to be similar to the causes underlying depressed ventricular function in general. Decreased LV function is common among adult patients with TOF and has been shown to be related to duration from shunt to repair, presence of arrhythmia, and QRS duration.14 RV dilation and related dysfunction are frequently observed in this cohort as well, with impairments in function exacerbated by residual pressure-loading lesions (e.g., pulmonary valve or branch pulmonary stenosis).15 As well, a number of reports have described LV dysfunction in association with RV dilation or dysfunction, which is believed to be related to “ventricular-ventricular interactions.”8,16e18 Our results add to the list of factors known to be associated with late adverse outcomes in patients with TOF. Traditional CMR factors such as ventricular dysfunction, RV dilation, and RV mass-to-volume ratio have all been well documented as specific risk factors for poor outcomes.7e9,18,19 These findings were observed, as well, in our cohort. Reports of outcomes related to novel imaging techniques such as deformation analysis have begun to appear in the literature.20,21 For example, a recent echocardiographic study has identified LV longitudinal strain as a risk factor for arrhythmia and death in TOF.22 Using a CMR-based method capable of making high-quality measurements of deformation of both ventricles, we have not only confirmed this result but have also extended it by showing that impaired RV longitudinal strain is associated even more strongly with death and VT than reduced LV strain. Moreover, our multivariate model including only strain parameters was equally predictive of the outcome, compared with a model including all traditional CMR parameters. Previous reports of the presence and effect of dyssynchrony in repaired TOF have been variable. Fernandes et al23 did not identify a significant difference in dyssynchrony parameters in children and adolescents with repaired TOF compared with controls. Van der Hulst et al,24 however, have reported a delay in activation of the RV outlet in repaired pediatric patients with TOF, compared with normal controls. In our work, the TOF case (adverse outcome) group was not found to be significantly more dyssynchronous than the TOF control group, for any of the parameters investigated. In another previous study, however, Ortega et al20 described a relation between 1 parameter of dyssynchrony (time to peak circumferential strain) and adverse events in TOF. Once a patient with TOF has been identified to be “at risk,” the question remains regarding what, if anything, to do in an effort to mitigate that risk. Pulmonary valve replacement, either surgical or percutaneous, is a promising intervention, and many studies have investigated the potential
679
value of this procedure,25e27 with a recent report of improved LV strain after transcatheter valve replacement.28 Other patients may benefit from placement of an implantable cardioverter-defibrillator, but this intervention is not without its own risk.29 Moving forward, the myocardial parameters investigated in this work may prove useful in future efforts to develop more sensitive and specific stratification models for patients with TOF and help to identify those in greatest jeopardy, in whom the risk associated with further interventional procedures might be justified. This study was retrospective, with all the limitations inherent to this study design. The small number of patients with repaired TOF with adverse events after CMR may have limited our ability to detect subtle differences in dyssynchrony parameters. Because we chose to analyze short-axis slices in which the RV myocardium was continuous, we excluded the effect of infundibular patch on global strain and synchrony; this was done in a systematic fashion for both case and control cohorts, however, so any effect of this design choice would be expected to apply to both groups. The frame rate of the CMR cine images is modest and might reduce the ability of this method to identify differences between groups; again, however, this should apply equally to the case and control groups. Disclosures Tal Geva, MD is a consultant for Medtronic. All other authors have no conflicts of interest. 1. Apitz C, Webb GD, Redington AN. Tetralogy of Fallot. Lancet 2009;374:1462e1471. 2. Bacha EA, Scheule AM, Zurakowski D, Erickson LC, Hung J, Lang P, Mayer JE, del Nido PJ, Jonas RA. Long-term results after early primary repair of tetralogy of Fallot. J Thorac Cardiovasc Surg 2001;122:154e161. 3. Nollert G, Fischlein T, Bouterwek S, Bohmer C, Klinner W, Reichart B. Long-term survival in patients with repair of tetralogy of Fallot: 36year follow-up of 490 survivors of the first year after surgical repair. J Am Coll Cardiol 1997;30:1374e1383. 4. Murphy JG, Gersh BJ, Mair DD, Fuster V, McGoon MD, Ilstrup DM, McGoon DC, Kirklin JW, Danielson GK. Long-term outcome in patients undergoing surgical repair of tetralogy of Fallot. N Engl J Med 1993;329:593e599. 5. Alexiou C, Mahmoud H, Al-Khaddour A, Gnanapragasam J, Salmon AP, Keeton BR, Monro JL. Outcome after repair of tetralogy of Fallot in the first year of life. Ann Thorac Surg 2001;71:494e500. 6. Gatzoulis MA, Till JA, Somerville J, Redington AN. Mechanoelectrical interaction in tetralogy of Fallot. QRS prolongation relates to right ventricular size and predicts malignant ventricular arrhythmias and sudden death. Circulation 1995;92:231e237. 7. Ghai A, Silversides C, Harris L, Webb GD, Siu SC, Therrien J. Left ventricular dysfunction is a risk factor for sudden cardiac death in adults late after repair of tetralogy of Fallot. J Am Coll Cardiol 2002;40:1675e1680. 8. Geva T, Sandweiss BM, Gauvreau K, Lock JE, Powell AJ. Factors associated with impaired clinical status in long-term survivors of tetralogy of Fallot repair evaluated by cardiac magnetic resonance imaging. J Am Coll Cardiol 2004;43:1068e1074. 9. Knauth AL, Gauvreau K, Powell AJ, Landzberg MJ, Walsh EP, Lock JE, del Nido PJ, Geva T. Ventricular size and function assessed by cardiac MRI predict major adverse clinical outcomes late after tetralogy of Fallot repair. Heart 2008;94:211e216. 10. Geva T. Repaired tetralogy of Fallot: the roles of cardiovascular magnetic resonance in evaluating pathophysiology and for pulmonary valve replacement decision support. J Cardiovasc Magn Reson 2011;13:9. 11. Schuster A, Kutty S, Padiyath A, Parish V, Gribben P, Danford DA, Makowski MR, Bigalke B, Beerbaum P, Nagel E. Cardiovascular
680
12.
13.
14.
15.
16.
17.
18.
19.
The American Journal of Cardiology (www.ajconline.org) magnetic resonance myocardial feature tracking detects quantitative wall motion during dobutamine stress. J Cardiovasc Magn Reson 2011;13:58. Harrild DM, Marcus E, Hasan B, Alexander ME, Powell AJ, Geva T, McElhinney DB. Impact of transcatheter pulmonary valve replacement on biventricular strain and synchrony assessed by cardiac magnetic resonance feature tracking. Circ Cardiovasc Interv 2013;6: 680e687. Fornwalt BK, Arita T, Bhasin M, Voulgaris G, Merlino JD, León AR, Fyfe DA, Oshinski JN. Cross-correlation quantification of dyssynchrony: a new method for quantifying the synchrony of contraction and relaxation in the heart. J Am Soc Echocardiogr 2007;20:1330e1337. Broberg CS, Aboulhosn J, Mongeon FP, Kay J, Valente AM, Khairy P, Earing MG, Opotowsky AR, Lui G, Gersony DR, Cook S, Ting JG, Webb G, Gurvitz MZ; Alliance for Adult Research in Congenital Cardiology (AARCC). Prevalence of left ventricular systolic dysfunction in adults with repaired tetralogy of Fallot. Am J Cardiol 2011;107:1215e1220. Davlouros PA, Kilner PJ, Hornung TS, Li W, Francis JM, Moon JC, Smith GC, Tat T, Pennell DJ, Gatzoulis MA. Right ventricular function in adults with repaired tetralogy of Fallot assessed with cardiovascular magnetic resonance imaging: detrimental role of right ventricular outflow aneurysms or akinesia and adverse right-to-left ventricular interaction. J Am Coll Cardiol 2002;40:2044e2052. Cheung EW, Liang XC, Lam WW, Cheung YF. Impact of right ventricular dilation on left ventricular myocardial deformation in patients after surgical repair of tetralogy of Fallot. Am J Cardiol 2009;104: 1264e1270. Kempny A, Diller GP, Orwat S, Kaleschke G, Kerckhoff G, Bunck ACh, Maintz D, Baumgartner H. Right ventricular-left ventricular interaction in adults with tetralogy of Fallot: a combined cardiac magnetic resonance and echocardiographic speckle tracking study. Int J Cardiol 2012;154:259e264. Tobler D, Crean AM, Redington AN, Van Arsdell GS, Caldarone CA, Nanthakumar K, Stambach D, Dos L, Wintersperger BJ, Oechslin EN, Silversides CK, Wald RM. The left heart after pulmonary valve replacement in adults late after tetralogy of Fallot repair. Int J Cardiol 2012;160:165e170. Valente AM, Gauvreau K, Assenza GE, Babu-Narayan SV, Schreier J, Gatzoulis MA, Groenink M, Inuzuka R, Kilner PJ, Koyak Z, Landzberg MJ, Mulder B, Powell AJ, Wald R, Geva T. Contemporary predictors of death and sustained ventricular tachycardia in patients with repaired tetralogy of Fallot enrolled in the INDICATOR cohort. Heart 2014;100:247e253.
20. Ortega M, Treidman JK, Geva T, Harrild DM. Relation of left ventricular dyssynchrony measured by cardiac magnetic resonance tissue tracking in repaired tetralogy of Fallot to ventricular tachycardia and death. Am J Cardiol 2011;107:1535e1540. 21. Su HM, Lin TH, Hsu PC, Lee WH, Chu CY, Lee CS, Voon WC, Lai WT, Sheu SH. Global left ventricular systolic strain as a major predictor of cardiovascular events in patients with atrial fibrillation. Heart 2013;99:1588e1596. 22. Diller GP, Kempny A, Liodakis E, Alonso-Gonzalez R, Inuzuka R, Uebing A, Orwat S, Dimopoulos K, Swan L, Li W, Gatzoulis MA, Baumgartner H. Left ventricular longitudinal function predicts lifethreatening ventricular arrhythmia and death in adults with repaired tetralogy of Fallot. Circulation 2012;125:2440e2446. 23. Fernandes FP, Manlhiot C, Roche SL, Grosse-Wortmann L, Slorach C, McCrindle BW, Mertens L, Kantor PF, Friedberg MK. Impaired left ventricular myocardial mechanics and their relation to pulmonary regurgitation, right ventricular enlargement and exercise capacity in asymptomatic children after repair of tetralogy of Fallot. J Am Soc Echocardiogr 2012;25:494e503. 24. van der Hulst AE, Roest AA, Delgado V, Holman ER, de Roos A, Blom MA, Bax JJ. Relationship between temporal sequences of right ventricular deformation and right ventricular performance in patients with corrected tetralogy of Fallot. Heart 2011;97:231e236. 25. Kutty S, Deatsman SL, Russell D, Nugent ML, simpson PM, Frommelt PC. Pulmonary valve replacement improves but does not normalize right ventricular mechanics in repaired congenital heart disease: a comparative assessment using velocity vector imaging. J Am Soc Echocardiogr 2008;21:1216e1221. 26. Knirsch W, Dodge-Khatami A, Kadner A, Kretschmar O, Steiner J, Böttler P, Kececioglu D, Harpes P, Valsangiacomo Buechel ER. Assessment of myocardial function in pediatric patients with operated tetralogy of Fallot: preliminary results with 2D strain echocardiography. Pediatr Cardiol 2008;29:718e725. 27. Moiduddi N, Asoh K, Slorach C, Benson LN, Friedberg MK. Effect of transcatheter pulmonary valve implantation on short-term right ventricular function as determined by two-dimensional speckle tracking strain and strain rate imaging. Am J Cardiol 2009;104:862e867. 28. Tzemos N, Harris L, Carasso S, Subira LD, Greutmann M, Provost Y, Redington AN, Rakowski H, Siu SC, Silversides CK. Adverse left ventricular mechanics in adults with repaired tetralogy of Fallot. Am J Cardiol 2009;103:420e425. 29. Witte KK, Pepper CB, Cowan JC, Thomson JD, English KM, Blackburn ME. Implantable cardioverter-defibrillator therapy in adult patients with tetralogy of Fallot. Europace 2008;10:926e930.