Trabeculated Myocardium in Hypertrophic Cardiomyopathy: Clinical Consequences

Aims: Hypertrophic cardiomyopathy (HCM) is often accompanied by increased trabeculated myocardium (TM)—which clinical relevance is unknown. We aim to measure the left ventricular (LV) mass and proportion of trabeculation in an HCM population and to analyze its clinical implication. Methods and Results: We evaluated 211 patients with HCM (mean age 47.8 ± 16.3 years, 73.0% males) with cardiac magnetic resonance (CMR) studies. LV trabecular and compacted mass were measured using dedicated software for automatic delineation of borders. Mean compacted myocardium (CM) was 160.0 ± 62.0 g and trabecular myocardium (TM) 55.5 ± 18.7 g. The percentage of trabeculated myocardium (TM%) was 26.7% ± 6.4%. Females had significantly increased TM% compared to males (29.7 ± 7.2 vs. 25.6 ± 5.8, p < 0.0001). Patients with LVEF < 50% had significantly higher values of TM% (30.2% ± 6.0% vs. 26.6% ± 6.4%, p = 0.02). Multivariable analysis showed that female gender and neutral pattern of hypertrophy were directly associated with TM%, while dynamic obstruction, maximal wall thickness and LVEF% were inversely associated with TM%. There was no association between TM% with arterial hypertension, physical activity, or symptoms. Atrial fibrillation and severity of hypertrophy were the only variables associated with cardiovascular death. Multivariable analysis failed to demonstrate any correlation between TM% and arrhythmias. Conclusions: Approximately 25% of myocardium appears non-compacted and can automatically be measured in HCM series. Proportion of non-compacted myocardium is increased in female, non-obstructives, and in those with lower contractility. The amount of trabeculation might help to identify HCM patients prone to systolic heart failure.

Abstract: Aims: Hypertrophic cardiomyopathy (HCM) is often accompanied by increased trabeculated myocardium (TM)-which clinical relevance is unknown. We aim to measure the left ventricular (LV) mass and proportion of trabeculation in an HCM population and to analyze its clinical implication. Methods and Results: We evaluated 211 patients with HCM (mean age 47.8 ± 16.3 years, 73.0% males) with cardiac magnetic resonance (CMR) studies. LV trabecular and compacted mass were measured using dedicated software for automatic delineation of borders. Mean compacted myocardium (CM) was 160.0 ± 62.0 g and trabecular myocardium (TM) 55.5 ± 18.7 g. The percentage of trabeculated myocardium (TM%) was 26.7% ± 6.4%. Females had significantly increased TM% compared to males (29.7 ± 7.2 vs. 25.6 ± 5.8, p < 0.0001). Patients with LVEF < 50% had significantly higher values of TM% (30.2% ± 6.0% vs. 26.6% ± 6.4%, p = 0.02). Multivariable analysis showed that female gender and neutral pattern of hypertrophy were directly associated with TM%, while dynamic obstruction, maximal wall thickness and LVEF% were inversely associated with TM%. There was no association between TM% with arterial hypertension, physical activity, or symptoms. Atrial fibrillation and severity of hypertrophy were the only variables associated with cardiovascular death. Multivariable analysis failed to demonstrate any correlation between TM% and arrhythmias. Conclusions: Approximately 25% of myocardium appears non-compacted and can automatically be measured in HCM series. Proportion of non-compacted myocardium is increased in

Introduction
Hypertrophic cardiomyopathy (HCM) is a genetic cardiac disease characterized by clinical and prognostic heterogeneity [1][2][3][4]. The variable phenotypic expression and incomplete penetrance [5] have constituted an obstacle to obtain a full understanding of the consequences of the disease. HCM is mainly caused by mutations in genes encoding sarcomeric contractile proteins [6,7].
Left ventricular non-compaction (LVNC) is defined by an increase of trabeculations in the left ventricular endo-myocardium. Adult appearance of trabeculated myocardium may mainly be a consequence of an arrest in the embryologic process of compaction.
Some authors have suggested that myocardial trabeculations can be acquired (although this hypothesis remains to be demonstrated) in response to an increase in cardiac load [8,9].
This entity was increasingly recognized with the development of high definition cardiac imaging. Although LVNC can be in isolation, an increase in hypertrabeculation often accompanies other genetic cardiomyopthies [10]. Despite early descriptions of LVNC in adults showed an increase in the rate of adverse outcomes, such as progression to heart failure, arrhythmias, and emboli [11,12], further analysis has showed inconsistent results [13][14][15].
The clinical relevance of the presence of hypertrabeculation in HCM is unknown. We aim to measure the left ventricular mass and proportion of trabeculation in a population of HCM patients.

Study Sample
Patients recruited from an Inherited Cardiomyopathy Clinic-meeting international HCM criteria with an available good quality CMR study-were included. The first CMR of each patient was used for the study. Clinical and outcome data prospectively collected in a database was available for analysis.
Patients were prospectively included, cardiac imaging and examinations were performed prospectively; symptoms, history of hypertension (HTN), regular physical exercise and arrhythmic events prior to the first evaluation were also recorded. HTN was defined following European Society of Hypertension and European Society of Cardiology recommendation [16].
All patients were offered a 24 h ECG-Holter and exercise test in order to complete sudden death (SD) risk stratification. Non-sustained ventricular tachycardia (NSVT) was assessed. Arrhythmic events and complications during follow up were also recorded. In the cohort, there were three resuscitated cardiac arrests (CA) and six appropriate implantable cardioverter defibrillator (ICD) discharges, which were computed as SD equivalent for survival estimates. Mean follow-up was 49.1 ± 37.1 months ( Figure 1).
All subjects gave their informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved

Trabecular Quantification
Magnetic resonance studies were performed in two hospitals with different scanners: SIGNA HDxt 1,5T: General Electric Systems, USA, post-processing software Advantage Workstation, AWA.3-08 and Achieva CV, Philips Medical Systems, Netherlands, (Philips Software workspace 2.6.3.2). A dedicated software tool for the automatic quantification and exact hyper-trabeculation degree of LVNC based on automatic delineation of the epicardial and endocardial borders of the LV and trabecular recesses was used [18]   Performance of the software was reviewed visually by one imaging cardiologist expert in order to rule out significant deviations. Additionally, delineation of the borders of 10 randomly selected cases was subjectively scored by two skilled cardiologists. The score went from 1 to 5, from large disagreement (1) to exact match (5). A total of 73 CMR slices were evaluated: 65 slices were scored with a five by the two and the remaining eight slices were classified with a four at least by one of the cardiology experts. The weighted kappa statistic showed an agreement between the two observers of 96.6% (kappa 0.76).

Trabecular Quantification
Magnetic resonance studies were performed in two hospitals with different scanners: SIGNA HDxt 1,5T: General Electric Systems, USA, post-processing software Advantage Workstation, AWA.3-08 and Achieva CV, Philips Medical Systems, Netherlands, (Philips Software workspace 2.6.3.2). A dedicated software tool for the automatic quantification and exact hyper-trabeculation degree of LVNC based on automatic delineation of the epicardial and endocardial borders of the LV and trabecular recesses was used [18]

Trabecular Quantification
Magnetic resonance studies were performed in two hospitals with different scanners: SIGNA HDxt 1,5T: General Electric Systems, USA, post-processing software Advantage Workstation, AWA.3-08 and Achieva CV, Philips Medical Systems, Netherlands, (Philips Software workspace 2.6.3.2). A dedicated software tool for the automatic quantification and exact hyper-trabeculation degree of LVNC based on automatic delineation of the epicardial and endocardial borders of the LV and trabecular recesses was used [18] (see Supplementary Material) ( Figure 2). Performance of the software was reviewed visually by one imaging cardiologist expert in order to rule out significant deviations. Additionally, delineation of the borders of 10 randomly selected cases was subjectively scored by two skilled cardiologists. The score went from 1 to 5, from large disagreement (1) to exact match (5). A total of 73 CMR slices were evaluated: 65 slices were scored with a five by the two and the remaining eight slices were classified with a four at least by one of the cardiology experts. The weighted kappa statistic showed an agreement between the two observers of 96.6% (kappa 0.76). Performance of the software was reviewed visually by one imaging cardiologist expert in order to rule out significant deviations. Additionally, delineation of the borders of 10 randomly selected cases was subjectively scored by two skilled cardiologists. The score went from 1 to 5, from large disagreement (1) to exact match (5). A total of 73 CMR slices were evaluated: 65 slices were scored with a five by the two and the remaining eight slices were classified with a four at least by one of the cardiology experts. The weighted kappa statistic showed an agreement between the two observers of 96.6% (kappa 0.76).

Statistical Analysis
Statistical analysis was performed using SPSS statistical software (version 21.0 IBM Corp. Chicago, IL, USA, 2012). Chi square, t-test, Anova and Pearson were used for comparison between groups where appropriate. Variables with p < 0.10 in univariate were included in the multivariate analysis. Additionally, those considered relevant to adjust the model were also included. Variables with redundant information or colinearity were excluded. Linear logistic regression was used for association between study variables (TM and TM%) and clinical variables. Kaplan Meier estimates and Cox regression analysis were used for survival analysis. p-values < 0.05 were considered statistically significant.

LV Trabeculation and Demographics
Despite both trabeculated (TM) and compacted myocardial (CM) mass were increased in males compared to females, females had significantly higher TM% (29.7% vs. 25.6%, p < 0.0001). Indexed CM remained increased in males compared to females while TM mass was similar in males and females (Table 3).

LV Trabeculation and Clinical Findings
TM% was not related with symptoms. Patients with dyspnea (NYHA III-IV) or syncope had similar TM and TM%. There was no difference in TM% in patients with or without AF. There were only 6 (2.8%) patients with sustained ventricular tachycardia who had a significantly lower TM% (19.8 ± 6.0 vs. 26.9 ± 6.3, p = 0.007).

LV Trabeculation and Clinical Findings
TM% was not related with symptoms. Patients with dyspnea (NYHA III-IV) or syncope had similar TM and TM%. There was no difference in TM% in patients with or without AF. There were only 6 (2.8%) patients with sustained ventricular tachycardia who had a significantly lower TM% (19.8 ± 6.0 vs. 26.9 ± 6.3, p = 0.007).
In total, 12 (5.7%) patients had at least one major event over a mean 49.1 ± 37.1 months of follow-up: three patients died suddenly, one had a resuscitated cardiac arrest and four had appropriate ICD therapies. There was one heart failure death, one stroke related death, one procedure related death, and one heart transplant.

Multivariable Analysis
Female gender, MWT, neutral pattern of hypertrophy, obstruction, and LVEF were associated with TM% on multivariable analysis which included morphologic and demographic predictors. Age, HTN, physical exercise, left atrium diameter, or the presence of LGE were not associated with TM% (Table 5) . Symptoms (syncope or dyspnea) were not associated with TM% in a model with demographic, main morphological variables and symptoms. AF and NSVT on Holter was neither associated with TM%.
On multivariable analysis TM% was not associated with cardiovascular (CV) death or equivalent. MWT and AF were the only two variables associated with CV death.
In the presence of CM mass calculated from CMR, MWT lost his value in the prediction of CV death. In this model, AF remained significantly associated (and a trend of TM%) ( Table 6). Severity of wall thickness, measured as MWT or CM mass were the strongest predictors of CV death and SD.

Discussion
This is the first study to systematically quantify LV trabeculation in an HCM population. The main finding from our study was that a clinical profile of patients with HCM with increased trabeculations may exist. Some of the features were predominant female gender, and non-obstructive forms with systolic impairment. There was no association between TM with age, ventricular volumes, or LGE. Higher degree of trabeculation in HCM was associated with LVEF below 50% which might precede development of systolic heart failure [19]. Degree of trabeculations do not seem to associate to an increase arrhythmic risk in HCM. A longer follow up is needed confirm these findings.
Jacquier et al. [20] manually measured the TM mass in a group of patients with isolated non-compaction, dilated cardiomyopathy, HCM, and controls. A threshold of 20% of TM was proposed as a good cutoff for the diagnosis of LVNC (S and E >93%). Sixteen LVNC patients had an average of trabeculated mass of 32%, compared to 11% to 12% of DCM, HCM and controls. In this paper, the 16 patients with HCM had a mean CM of 216 g, which is significantly higher to the one measured in our study (160.0 ± 62.0 g). On the contrary, trabeculated myocardial mass was lower in that paper compared to our estimation from 211 patients (28 g vs. 55.5 ± 18.7 g). In Jacquier's paper, papillary muscles were manually excluded from the TM, while in our automatic measurement this was included. Observed differences cannot solely be explained by different methodology [21].
The percentage of trabeculated myocardium showed in this large series of HCM (26.7 ± 6.4%) patients demonstrate a clear increase from that observed in controls, which suggests that LVNC or hypertrabeculation is part of the clinical phenotype of HCM. In an earlier paper from our group using the same software, a cutoff of 27% MT showed to have a good accuracy for differentiation of LVNC [22].
In contrast to the fractal analysis recently published by Captur et al. [23], based on tortuosity measured by pixelation of the line of the endocardial border, our software provides easy to understand clinical measurements of CM and TM mass. This computationally-assisted method could save valuable diagnostic time compared with traditional processing, thus minimizing the possibility of human error. We demonstrate also in this population of HCM patients, a good performance of the software for quantification of hypertrabeculation based on the automatic delineation of borders from CMR diastolic images [24].
Trabeculation in its different forms, including deep crypts and clefts has been suggested to be a pre-diagnostic (maybe congenital) feature in carriers of HCM associated mutations [25]. While other authors suggest an acquired mechanism [26], from observations of increases of trabeculation in high demanding conditions such as in athletes and pregnancy, the mechanism of how this occurs is however unknown. Moreover, a reversion of the phenotype with the normalization in cardiac load after delivery, has been shown in one of those studies [27]. We have failed to show any association between the degree of physical activity, presence of HTN, or obstruction in the magnitude of trabeculations in our series of HCM patients.
In contrast with an early paper of a dilated cardiomyopathy series [28], which suggested a higher rate of stroke in patients with LVNC, we also failed to demonstrate association between embolic events in higher trabeculated ventricles with HCM.
To finalize, we want to highlight the need of a quantification of trabeculation in order to define the limits between normality and abnormality and to establish the role of non-compaction in various cardiac conditions. Our software has demonstrated to be useful not only for automatic quantification of TM but also for CM. In this regard, in our study, CM mass measured by the software was the strongest predictor of the outcome above the traditional maximal wall thickness taken from echocardiography. The latter supports the use of global assessment of left ventricular CM and TM mass to replace focal measurements of wall thickness in risk algorithms.

Limitations
Differences in the characteristics of our series regarding distribution of the pattern of hypertrophy, value of LVH mass, proportion of obstruction or percentage of LGE might be due to the age of the cohort, which is relatively young (mean 44 years old), the inclusion of a~34% of relatives (as compared to probands) or the relatively high percentage of patients diagnosed through family screening (25%) or incidentally (28%). Further studies are warranted to verify our findings.
Despite the number of individuals being relatively large, only a few patients developed CV events. Survival analysis is then limited, and results should be taken carefully. Patients with devices were ruled out from CMR and some high-risk patients were excluded. Repeated CMR studies were not available and changes over follow-up in CM and TM could not be explored.
Genetic information was available in 88/140 (62.9%) of the probands included in the cohort. Genetic testing yield was 43/88 (48.9%). Further analysis of the association between genetic results and morphologic and clinical findings was not performed due to the incompleteness of the genetic information.

Conclusions
A significant proportion (~25%) of myocardium appears non-compacted and can be automatically measured in HCM series. There is a clinical profile of patients with HCM with increased trabeculations. Proportion of non-compacted myocardium is increased in female, non-obstructives, and in those with lower contractility. The amount of trabeculation might help to identify HCM patients prone to systolic heart failure. Amount of trabeculation does not seem to associate with arrhythmias.