Author + information
- Received December 11, 2014
- Revision received April 22, 2015
- Accepted April 28, 2015
- Published online August 1, 2015.
- Mihaela-Silvia Amzulescu, MD,
- Michel F. Rousseau, MD, PhD,
- Sylvie A. Ahn, MS,
- Laurianne Boileau, MD,
- Christophe de Meester de Ravenstein, MS,
- David Vancraeynest, MD, PhD,
- Agnes Pasquet, MD, PhD,
- Jean Louis Vanoverschelde, MD, PhD,
- Anne-Catherine Pouleur, MD, PhD and
- Bernhard L. Gerber, MD, PhD∗ ()
- Division of Cardiology, Department of Cardiovascular Diseases, Cliniques Universitaires St. Luc and Pôle de Recherche Cardiovasculaire, Institut de Recherche Expérimentale et Clinique, Université Catholique de Louvain, Brussels, Belgium
- ↵∗Reprint requests and correspondence:
Dr. Bernhard L. Gerber, Division of Cardiology, Department of Cardiovascular Diseases, Cliniques Universitaires Saint Luc UCL, Avenue Hippocrate 10/2806, B-1200 Woluwe Saint Lambert, Belgium.
Objectives The purpose of this study was to evaluate the impact of hypertrabeculation and left ventricular (LV) myocardial noncompaction phenotype by cardiac magnetic resonance (CMR) on outcomes of patients with nonischemic dilated cardiomyopathy (DCM).
Background Myocardial trabeculations and noncompaction are increasingly observed in patients with DCM, but their prognostic impact remains unknown.
Methods We prospectively evaluated outcomes of 162 consecutive patients (102 men; age 55 ± 15 years; ejection fraction [EF] 25 ± 8%) with DCM undergoing CMR. The amount of noncompaction was quantified as noncompacted/compacted (NC/C) length in the long-axis view and as the ratio of NC/C mass in the short-axis view and compared against 48 healthy control subjects (age 60 ± 10 years).
Results Fifty-eight DCM patients (36%) had NC/C length ≥2.3, and 71 (44%) had NC/C mass greater than the 95% confidence interval (CI) of control subjects. NC/C length and NC/C mass did not correlate with any clinical, echocardiographic, or CMR parameters. Over a 3.4-year median follow-up, 29 patients experienced major adverse cardiovascular events (MACE) (12 cardiovascular deaths, 8 heart transplantations, 4 LV assist device implantations, and 5 resuscitated cardiac arrests or appropriate device shocks). Cox univariate analysis identified smoking, New York Heart Association functional class, blood pressure, LV and right ventricular end-diastolic and end-systolic volumes, LV EF, right ventricular EF, and late gadolinium enhancement as predictors of MACE. In multivariate analysis, only LV EF and late gadolinium enhancement were independent predictors of MACE-free survival (hazard ratio: 0.922, 95% CI: 0.878 to 0.967, p = 0.001 and HR: 1.096, 95% CI: 1.004 to 1.197, p = 0.04, respectively). Neither NC/C length nor NC/C mass had significant predictive value for MACE-free survival, either unadjusted or after adjustment for baseline variables. Also, there was no difference in cardioembolic event rate between groups with high and low NC/C length or mass.
Conclusions Cardiovascular outcomes of adult patients with nonischemic DCM do not appear to be influenced by the degree of trabeculation. This argues against a noncompaction phenotype designating a more severe form of DCM.
Left ventricular noncompaction (LVNC), is a recently recognized morphological abnormality of the left ventricle, characterized by LV dysfunction in the presence of excessive prominent trabeculations and deep inter-trabecular recesses (1,2). The definition of LVNC is purely morphological and is based on the noninvasive identification of thickened myocardium, with a 2-layered structure that consists of a thin, compacted epicardial layer and a much thicker, noncompacted endocardial layer.
LVNC was once considered a rare form of hereditary cardiomyopathy (3), with onset in early childhood, and was thought to result from the intrauterine arrest of the normal process of myocardial compaction causing a trabecular meshwork with deep endomyocardial spaces. However, this phenotype is now also increasingly reported in adult patients, probably because of increased awareness and the improved imaging capabilities of echocardiography and cardiac magnetic resonance (CMR).
To date, it remains uncertain whether LVNC is a distinct type of cardiomyopathy with different pathophysiology and outcome from other types of dilated cardiomyopathies (DCMs). Indeed, different echocardiographic (3,4) and CMR (5,6) diagnostic criteria have been proposed, and their sensitivity and specificity to accurately detect LVNC remain controversial (1,7). Highly variable clinical presentations have been described, ranging from totally asymptomatic disease to cardioembolic complications, ventricular arrhythmias, and heart failure. Thus, the prognosis of LVNC remains unknown (1).
Because it allows better visualization of trabeculations than echocardiography, CMR is currently considered the method of choice to identify LVNC (5). In the current study, we sought to evaluate whether the presence of the noncompaction phenotype and the amount of LVNC myocardium, assessed by CMR, influenced the prognosis of adult patients with nonischemic DCM. We therefore studied 162 consecutive DCM patients by CMR and evaluated the utility of noncompaction versus other parameters in univariate and multivariate survival analysis to predict overall and cardiovascular mortality and morbidity.
The study was approved by the institutional review board. From a prospective registry of all patients referred to our institution for performance of CMR for myocardial characterization with late gadolinium enhancement (LGE) between 2003 and 2013, we evaluated the follow-up of patients with a diagnosis of DCM based on World Health Organization/European Society of Cardiology recommendations, who had dilated LV volumes and ejection fraction (EF) <40% as measured by CMR. Exclusion criteria were ischemic etiology by coronary angiography or coronary computed tomography according to the criteria proposed by Felker et al. (8), and additionally, presence of subendocardial or transmural LGE suggestive of previous myocardial infarction (9). We also excluded patients with a poor life expectancy, such as patients with metastatic cancer, patients in palliative care, and patients hospitalized for terminal heart failure who were not considered candidates for heart transplantation. Patients with secondary cardiomyopathies (10) attributable to cardiotoxic chemotherapy, HIV, neuromuscular diseases, or coexisting severe primary valve disease and cardiomyopathies due to reversible causes such as endocrine disorders, hypertensive cardiomyopathy, and tachycardiomyopathy were also excluded. A total of 162 patients satisfied these criteria and constituted the final study population (Figure 1).
Clinical and echocardiographic data
Clinical data (including data on cardiovascular risk factors, comorbidities, and New York Heart Association functional class), electrocardiogram, laboratory tests, peak Vo2 exercise test, coronary angiography, and treatment were collected by review of medical records. Two-dimensional echocardiography data acquired within a median of 9 days (interquartile range [IQR]: 4 to 28 days) from the CMR examination and stored on a database server (Xcelera, Philips Healthcare, Best, the Netherlands) were analyzed by a single observer (M.-S.A.). Five echocardiography studies were unavailable for review. Diastolic function assessment (E/A mitral waves and E/e′ ratio) was available only for 127 patients and was missing in 30 patients because of atrial fibrillation or flutter (n = 10), sinus tachycardia (n = 19), or ventricular bigeminy (n = 1). Pulmonary artery systolic pressure was computed by adding the tricuspid regurgitation pressure gradient to the right atrial pressure estimated from the inferior vena cava dimensions and inspiratory collapse.
CMR studies were performed on 1.5-T or 3-T units (Intera-CV and Achieva, Philips Medical Systems, Best, the Netherlands) as reported previously (11). Briefly, 10 to 12 consecutive short-axis images covering the entire LV and 2-, 3-, and 4-chamber long-axis images were acquired with a cine steady-state free precession sequence to allow assessment of myocardial function and mass and quantification of noncompaction. Ten to 15 min after injection of 0.2 mmol/kg gadolinium contrast agent, LGE images were acquired by a 2- or 3-dimensional inversion recovery sequence in identical short- and long-axis slices.
CMR studies were analyzed with the freely available software Segment version 1.9 (Medviso, Lund, Sweden) (12) by the same observer. Right ventricular (RV) and LV volumes, mass, and EF were computed from the short-axis cine images by semiautomatic tracing of the endocardial and epicardial contours in the end-diastolic (ED) and end-systolic phases. Values were indexed to the body surface area (BSA).
LV trabeculations, identified as any endocardial wall contour irregularities present in the ED phase, distinct from papillary muscles and chordae, were measured by 2 previously described methods. In method 1 (Figure 2), according to Petersen et al. (5), on long-axis images, the lengths of compacted and noncompacted layers of the most trabeculated segment (excluding the true apex, segment 17) were measured in ED and reported as noncompacted to compacted (NC/C) length ratio. In method 2 (Figure 3), according to Jacquier et al. (6), on short-axis images in ED, LVNC mass was measured as the difference between the global LV mass (LV trabeculations included in the endocardial tracing) and compacted mass (LV trabeculations excluded from the endocardial tracing) and reported as NC/C mass ratio. The papillary muscles were each time excluded from the LV mass calculation.
Presence, localization, and extent of LGE were evaluated visually and quantified with a semiautomated algorithm (Segment software) (12) as reported previously (11). LGE patterns were characterized as mid-wall, epicardial, or patchy.
LA volume indexed to BSA was computed by the biplane area-length method on 4- and 2-chamber end-systolic cine images with Osirix software version 5.8 (Pixmeo, Geneva, Switzerland). Mitral regurgitation volume was computed as the difference between LV stroke volume by the Simpson method and phase-contrast aortic forward flow. Lastly, the presence of excessive RV trabeculations was evaluated visually in the short-axis stacks.
Normal values of CMR parameters and of NC/C length and mass were defined in a population of 48 healthy volunteers (n = 22 males, mean age 60 ± 10 years, mean BSA 1.8 ± 0.2 m2) without a history of cardiovascular disease and with unremarkable physical examination and electrocardiogram and normal computed tomography coronary angiography.
Follow-up and endpoint
Follow-up was performed by telephone contact of patients/family and physicians between January 2014 and August 2014 and by review of medical records for a total of 612 person-years. Follow-up was complete for all but 1 patient. The cause of death was categorized as cardiac or noncardiac. Cardiac death was defined as death attributable to congestive heart failure (i.e., death preceded by acute exacerbation or worsening of heart failure) or sudden death (i.e., unexpected, unwitnessed, or witnessed death in the absence of other apparent causes). Appropriate device intervention was defined as device shock or antitachycardia pacing delivered in response to a documented ventricular tachyarrhythmia.
The primary endpoint of our study was a composite endpoint of major adverse cardiovascular events (MACE) comprising cardiovascular death, heart transplantation, LV assist device implantation, resuscitated cardiac arrest, and appropriate device shocks. Secondary endpoints were all-cause mortality, cardiovascular mortality, stroke and embolic events.
Statistical analysis was performed with SPSS version 15.0 (SPSS Inc., Chicago, Illinois). Continuous variables are presented as mean ± SD or as median (IQR) and categorical variables as counts and percentages. Comparisons between groups were performed with a 2-sided Student t test, Wilcoxon-Mann-Whitney U test, and chi-square or Fisher exact test for categorical variables, as appropriate. Correlation was performed with the Pearson test. Values of p < 0.05 were considered statistically significant. Under the assumption that one-third of patients have high NC/C, our study had 80% power to detect a 153% higher event rate (i.e., 62% vs. 85% MACE-free survival) with α = 0.05.
On the basis of the 2 methods described above, the study population was divided into groups according to the degree of noncompaction, expressed as either the NC/C length or the NC/C mass. The cutoff values used to define the groups were NC/C length ≥2.3 (as proposed by Petersen) and NC/C mass ≥95% confidence interval (CI) (i.e., ≥31%) of control subjects.
The index date was the date of the CMR examination. The duration of follow-up was computed using the index date to the date of the first MACE or last clinical follow-up. Kaplan-Meier survival and Cox proportional hazards analyses were used to assess the relationship between the degree of NC by the 2 methods presented and MACE. Survival curves were compared with the log-rank (Mantel-Cox) test. Hazard ratios (HRs) are expressed as mean and 95% CIs. All significant univariate predictors of MACE were proposed for inclusion in multivariate forward and backward stepwise Cox models. The model with the lowest Akaike and Bayesian information criteria was retained. The predictive values of NC/C length and NC/C mass were evaluated both unadjusted and after adjustment for baseline predictors of MACE. Intraobserver and interobserver agreement for noncompaction measurement were tested in 15 randomly selected cases according to the Bland-Altman method and expressed as bias ± SD (95% CI) and intraclass correlation coefficients (ICCs).
Detection of noncompaction by CMR
Intraobserver and interobserver variability for measuring NC/C length were 0.06 ± 0.56 (95% CI: −0.54 to 1.67) and 0.11 ± 1.06 (95% CI: −1.00 to 3.13), with ICCs of 0.91 and 0.70, respectively. Intraobserver and interobserver variability for measurement of noncompacted mass were −0.98 ± 4.4% (95% CI: −4.27% to 13.2%) and −2.815 ± 3.37% (95% CI: −3.24% to 9.99%), with ICCs of 0.97 and 0.991, respectively.
Median (IQR) NC/C length and average NC/C mass in healthy volunteers were 1.5 (IQR: 1.1 to 2) and 17 ± 7%, respectively. Eight control subjects (17%) had NC/C length ≥2.3. Patients with DCM had significantly greater median (IQR) NC/C length (1.8 [IQR: 1.2 to 2.6]; p < 0.01) and average NC/C mass (31 ± 15%; p < 0.001) than volunteers. Noncompacted segments were mainly located at the mid and apical segments and the nonseptal regions and followed a similar distribution in patients as in control subjects.
Using the 2 previously described criteria for diagnosis of noncompaction (i.e., NC/C length ≥2.3 or ≥95% CI [≥31%] of NC/C mass of volunteers), we classified DCM patients into groups with a high or low degree of noncompaction. Respectively, 58 DCM patients (36%) had maximum NC/C length ≥2.3, whereas 71 (44%) had NC/C mass ≥31%. Forty-four patients (27%) presented with both criteria. Although the correlation between NC/C length and NC/C mass in patients was satisfactory (r = 0.62), the agreement of the 2 methods to classify patients into groups with high and low degrees of noncompaction was poor (κ < 0.2, p = 0.0001).
As shown in Table 1, except for higher systolic blood pressure and presence of hypertension in patients with NC/C mass <31% and a higher prevalence of atrial fibrillation in patient with NC/C length <2.3, there were no other statistical differences between the groups of patients with high and low NC/C length and NC/C mass in terms of clinical, electrocardiographic, or echocardiographic characteristics, severity of heart failure, or treatment. Also, except for average LV wall thickness and LV mass index (which were significantly higher in the NC/C mass <31% group than in the NC/C mass ≥31% group [p < 0.05]), there were no statistical significant differences between groups for any CMR parameters. RV trabeculations were statistically more prevalent in the groups with a higher degree of LV noncompaction (Table 2).
Neither NC/C length nor NC/C mass ratio of patients showed a significant correlation with any clinical, magnetic resonance, or echocardiographic parameter of disease severity. Intramural thrombi were observed in similar proportions in patients with high and low NC/C length or NC/C mass ratio.
Over a median (IQR) follow-up of 3.4 years (1.5 to 6.3 years), 29 patients experienced MACE: 12 cardiovascular deaths (7 sudden, 5 worsening heart failure), 8 heart transplantations, 4 LV assist device implantations, and 5 arrhythmic events (1 resuscitated cardiac arrest, 3 appropriate device shocks, and 1 antitachycardia pacing). In addition, 8 noncardiovascular deaths were recorded. Four patients had stroke, 1 in the NC/C mass ≥31% group and 2 in the NC/C length ≥2.3 group. Odds ratio for stroke in groups with high NC/C mass and length were 0.41 (95% CI: 0.04 to 4.0; p = 0.63) and 1.77 (95% CI: 0.24 to 12.9; p = 0.62), respectively. There were no other thromboembolic events recorded.
MACE occurred in 18 (17%) and 11 (19%) patients from the NC/C length <2.3 and NC/C length ≥2.3 groups and in 19 (21%) and 10 (14%) patients from the NC/C mass <31% and NC/C mass ≥31% groups, respectively. The MACE-free survival Kaplan-Meier curves did not show a statistically significant difference in outcome for the NC/C length and NC/C mass groups (Figure 4). Likewise, for the secondary endpoints (all-cause mortality and cardiovascular mortality), there were no differences between patients with high and low amounts of noncompaction, assessed as either NC/C length or NC/C mass. To further study the influence of noncompaction on prognosis, we also classified patients into tertiles of NC/C length and NC/C mass. As shown in Figures 5 to 7⇓⇓⇓, the groups with the highest degree of noncompaction did not experience worse outcomes than those in the lowest tertile of noncompaction.
Cox univariate predictors of MACE are listed in Table 3. NC/C length and NC/C mass, both as continuous and as binary parameters, were not associated with a worse prognosis, either unadjusted or after adjustment for baseline variables (HR: 1.08; 95% CI: 0.74 to 1.6, p = 0.88 for NC/C length and HR: 1.001, 95% CI: 0.98 to 1.03, p = 0.93 for NC/C mass). By multivariate analysis, only LV EF and the amount of LGE were independent predictors of MACE-free survival.
The findings of the present study can be summarized as follows. 1) Adult patients with DCM had an increased degree of myocardial trabeculations when assessed by CMR and frequently satisfied previously reported criteria for diagnosis of noncompacted cardiomyopathy. 2) The presence and amount of LVNC myocardium measured by CMR did not correlate with disease severity and did not predict subsequent cardiovascular mortality and morbidity. 3) The prognosis of our patients was influenced only by the magnitude of LV and RV remodeling and systolic dysfunction and by the presence and amount of LGE, but not by the noncompaction phenotype.
Diagnosis of noncompaction cardiomyopathy
At present, the diagnosis of LVNC cardiomyopathy is purely morphological and based on the detection of increased amount of trabeculated myocardium. So far, no consensus has been reached on a generally accepted and binding definition of LVNC. Indeed, different echocardiographic (3,4) and CMR (5,6) criteria exist for the definition of LVNC, with several proposed diagnostic cutoff values, all of which were established in small populations. For CMR, 2 different criteria were proposed: Petersen et al. (5) suggested measuring the length of noncompacted and compacted segments on long-axis steady-state free precession cine images in end diastole, at a site with the most prominent trabeculations, and defined a ratio of NC/C length ≥2.3 as abnormal. By contrast, Jacquier et al. (6) proposed measurement of the LV trabeculated mass as the difference between the global and compacted LV mass on consecutive short-axis images in end diastole. The latter authors proposed a cutoff of >20% as being able to distinguish LVNC from patients with a normal amount of LV trabeculation (as seen in DCM, hypertrophic cardiomyopathy, and healthy control subjects). In our study, the 95% CIs of the noncompacted mass of healthy volunteers were higher than in the data presented by Jacquier et al. (6), and we consequently used a higher cutoff value of noncompacted mass to define abnormal trabeculation. Although in our study, both methods had good intraobserver and interobserver agreement and acceptable correlation for measuring noncompacted myocardium, the classification into groups of patients with abnormal and normal noncompaction by the 2 approaches disagreed strongly.
When we applied the Petersen criteria (5) to our healthy volunteers, an NC/C length ≥2.3 was present in as many as 17% of them. This is consistent with a report of the Multi-Ethnic Study of Atherosclerosis (MESA), in which 44% of 306 healthy patients met a criterion of increased noncompaction (13). An additional problem with the definition of noncompaction is the increased prevalence of both LV trabeculations and criteria for LVNC in African patients with heart failure compared with Caucasians. Therefore, the current diagnostic criteria may lack accuracy and specificity and lead to an overdiagnosis of LVNC (7).
Moreover, it is still not clear whether LVNC is truly a distinct cardiomyopathy or an epiphenomenon of other cardiomyopathies. Indeed, the American Heart Association considers noncompaction cardiomyopathy as a new type of genetic cardiomyopathy (10). By contrast, according to a 2008 European Society of Cardiology position statement paper (14), LVNC remains an unclassified cardiomyopathy. Indeed, patients with LVNC show genetic heterogeneity (1) and different phenotypic expression. Also, the noncompaction phenotype has been observed in several cardiomyopathies of different origins, including DCM (15), neuromuscular disorders (16), and congenital heart disease (17), as well as in athletes (18) and healthy populations (13), and may even appear transiently during pregnancy (19). Although the definition of LVNC would clearly benefit from a genotype-phenotype correlation, mutations described in LVNC affect genes of proteins that link the extracellular matrix of the myocardial cell to the cytoskeleton, such as ZASP, alpha-dystrobrevin, and tafazzin (20,21) which overlaps with other forms of DCM.
Hypertrabeculation and LV noncompaction in DCM
Our study shows that an NC/C length ≥2.3 and an increased NC/C mass are common findings in DCM. The 2 suggested diagnostic criteria for LVNC were present in 36% and 44% of patients with DCM, respectively. This is in agreement with other recent works (7) that also reported that a high percentage of patients with heart failure have noncompaction. This calls into question whether noncompaction in DCM truly indicates a different type of disease. Indeed, LVNC in DCM could simply arise from greater visibility or more pronounced separation of trabeculations when the LV cavity dilates, or it could be the result of the remodeling process or the loading conditions of the LV. This is compatible with reports of reversal of LVNC phenotype after cardiac resynchronization therapy (22) and in the post-partum period of pregnant women (19).
In our study, the amount of noncompaction did not correlate with the extent of LV ventricular dysfunction and was related to neither clinical nor echocardiographic markers of disease severity. In contrast to some other reports (23), we did not find a correlation between the noncompaction phenotype and the presence of LGE. This argues against noncompaction phenotype indicating a different or more severe form of DCM, as some previous small studies have suggested (23).
Prognostic implications of noncompaction phenotype in DCM
To date, there have been contradictory reports regarding the prognosis of patients with LVNC, mainly because of the heterogeneity of the population studied and the lack of consensus in LVNC definition. Earlier studies (24), in particular in pediatric populations (3), in whom the disease may be associated with facial dysmorphism and Wolff-Parkinson-White syndrome, reported an increased mortality caused by heart failure, arrhythmia, and thromboembolic events. More recent reports in adult patients with noncompaction and LV dysfunction have been conflicting (25–28).
In our study, the cardiovascular mortality and morbidity of adult patients with nonischemic DCM were not influenced by the presence of higher degrees of noncompacted myocardium. This is in agreement with a recent report that demonstrated that hypertrabeculation in asymptomatic subjects in MESA was not associated with deterioration in LV volumes or function or worse outcome over a 10-year follow-up (29). Overall, we observed low mortality, arrhythmic events, and strokes, even in the group of patients with the highest degree of noncompaction. Our findings may somewhat be at odds with recent studies that suggested that greater degrees of noncompaction as assessed by magnetic resonance imaging predict a greater number of events (27,30). However, in our study, we included only patients with DCM and low EF, unlike previous reports that also included patients with preserved EF and compared the outcome of patients with or without noncompaction to patients with DCM. Moreover, it has been shown that the prognosis of patients with noncompaction is mainly affected by the presence of heart failure symptoms, LV dilatation, and systolic dysfunction (31). Indeed, in agreement with other works (9,32), we observed that parameters reflecting LV and RV remodeling and in particular LV dysfunction and the presence and extent of LGE were statistically significant predictors of outcome in our DCM population.
Our study demonstrated that increased degrees of trabeculation did not predict worst clinical outcomes, and hence, the noncompaction phenotype is not an important risk factor in patients with DCM. This argues against screening for LVNC and systematically measuring the NC/C ratio in DCM by CMR and against adopting a more aggressive therapeutic attitude in DCM patients in whom an LVNC phenotype is identified by CMR. By contrast, the fact that parameters of biventricular remodeling and function were the most important predictors of outcome in our population supports the use of these classic parameters for risk assessment and selection of medical and device treatment in DCM patients.
Although our study is the largest to date, the relatively low number of events limited statistical power, enabled only the detection of relatively large effects, reduced our ability to control for the sources of possible confounding variables, and exposed a potential risk of overfitting in multivariate models. Larger multicentric studies will be required to better understand the clinical implications of the noncompaction phenotype in DCM. The study included consecutive patients referred to our tertiary center, with a potential referral bias for CMR. Also, we excluded patients with EF >40% and those with neuromuscular diseases and other secondary cardiomyopathies (10), which may also display the LVNC phenotype. Therefore, the study conclusions cannot necessarily be extrapolated to these populations. Finally, our study conclusions only apply to trabeculation detected by CMR.
The cardiovascular mortality and morbidity of adult patients with nonischemic DCM are influenced by the magnitude of biventricular remodeling and dysfunction and by the presence and amount of LGE detected by CMR. By contrast, the prognosis did not appear to be influenced by the degree of LV myocardial trabeculation. This argues against the noncompaction phenotype being a more severe form of DCM.
COMPETENCY IN MEDICAL KNOWLEDGE: Hypertrabeculation and noncompaction phenotype are frequently observed by cardiac magnetic resonance imaging in patients with dilated cardiomyopathy. The presence of this phenotype was not associated with worse clinical outcome.
TRANSLATIONAL OUTLOOK: Because this is a relatively small study, larger studies are needed to confirm the impact of hypertrabeculation and noncompaction phenotype on outcome of patients with dilated cardiomyopathy. Our data suggest, however, that treatment of patients with dilated cardiomyopathy should not be influenced by the presence of an increased amount of trabeculation.
This study was supported with a grant from the Fondation Nationale de la Recherche Scientifique of the Belgian Government (FRSM 3.4557.02). The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- body surface area
- confidence interval
- cardiac magnetic resonance
- dilated cardiomyopathy
- ejection fraction
- hazard ratio
- intraclass correlation coefficient
- interquartile range
- late gadolinium enhancement
- left ventricle/ventricular
- left ventricular noncompaction
- major adverse cardiovascular event
- right ventricle/ventricular
- Received December 11, 2014.
- Revision received April 22, 2015.
- Accepted April 28, 2015.
- American College of Cardiology Foundation
- Thavendiranathan P.,
- Dahiya A.,
- Phelan D.,
- Desai M.Y.,
- Tang W.H.
- Oechslin E.,
- Jenni R.
- Chin T.K.,
- Perloff J.K.,
- Williams R.G.,
- Jue K.,
- Mohrmann R.
- Jenni R.,
- Oechslin E.,
- Schneider J.,
- Attenhofer J.C.,
- Kaufmann P.A.
- Petersen S.E.,
- Selvanayagam J.B.,
- Wiesmann F.,
- et al.
- Jacquier A.,
- Thuny F.,
- Jop B.,
- et al.
- Kohli S.K.,
- Pantazis A.A.,
- Shah J.S.,
- et al.
- Felker G.M.,
- Shaw L.K.,
- O'Connor C.M.
- Maron B.J.,
- Towbin J.A.,
- Thiene G.,
- et al.
- Barone-Rochette G.,
- Pierard S.,
- de Meester de Ravenstein C.,
- et al.
- Kawel N.,
- Nacif M.,
- Arai A.E.,
- et al.
- Elliott P.,
- Andersson B.,
- Arbustini E.,
- et al.
- Gati S.,
- Chandra N.,
- Bennett R.L.,
- et al.
- Gati S.,
- Papadakis M.,
- Papamichael N.D.,
- et al.
- Vatta M.,
- Mohapatra B.,
- Jimenez S.,
- et al.
- Bertini M.,
- Ziacchi M.,
- Biffi M.,
- et al.
- Oechslin E.N.,
- Attenhofer Jost C.H.,
- Rojas J.R.,
- Kaufmann P.A.,
- Jenni R.
- Lofiego C.,
- Biagini E.,
- Pasquale F.,
- et al.
- Stacey R.B.,
- Andersen M.M.,
- St Clair M.,
- Hundley W.G.,
- Thohan V.
- Zemrak F.,
- Ahlman M.A.,
- Captur G.,
- et al.
- Gulati A.,
- Ismail T.F.,
- Jabbour A.,
- et al.