Despite several efforts using two-dimensional echocardiography and cardiac magnetic resonance in the diagnosis of left ventricular noncompaction (LVNC), there are no universally accepted diagnostic criteria. The aim of this study was to describe the extent of noncompacted myocardium using a new three-dimensional echocardiographic parameter.
Seventeen patients with diagnoses of LVNC on the basis of two-dimensional echocardiographic and clinical criteria, 26 Olympic rowing athletes, and 49 healthy volunteers underwent three-dimensional echocardiography. By offline analysis, left ventricular volumes, mass, ejection fraction, and sphericity index were calculated. Trabeculated left ventricular volume (TLV) was calculated as the difference between left ventricular end-diastolic volume obtained including and excluding the trabeculae in the cavity contour. TLV was also normalized by left ventricular end-diastolic volume (TLV%).
TLV and TLV% were significantly higher in patients with LVNC (33.7 ± 10.9 mL and 24 ± 7%) as opposed to controls (7.1 ± 2.2 mL, P < .001, and 6 ± 2%, P < .001, respectively) and athletes (8.0 ± 3.0 mL, P < .001, and 5 ± 2%, P < .001, respectively). In detail, on receiver operating characteristic curve analysis, optimal cutoff values of 15.8 mL for TLV and 12.8% for TLV% were determined for the identification of LVNC (area under the curve, 1.00; P < .001). Mild positive correlations of TLV and TLV% were found with sphericity index ( r = 0.294, P = .004, and r = 0.301, P = .004, respectively), and mild negative correlations were found with ejection fraction ( r = −0.454, P < .001, and r = −0.217, P = .038, respectively).
Because of high spatial resolution and accuracy in volumetric quantification, three-dimensional echocardiography allows accurate measurement of the extent of noncompacted myocardium and identification of patients with LVNC.
Over the past two decades, an increasingly large number of cases of left ventricular (LV) noncompaction (LVNC) have been described. This cardiac abnormality is characterized by multiple and prominent trabeculations and deep intertrabecular recesses communicating with LV cavity and localized most frequently at the apex of the left ventricle.
Several efforts have been carried out to identify optimal criteria for the identification of this pathology by echocardiography and cardiac magnetic resonance, but to date, there is no universal agreement on diagnostic criteria.
The aim of the present study was to describe morphologic, geometric, and functional characteristics of LVNC using three-dimensional echocardiography. By virtue of the high spatial definition of this technique, we hypothesized that assessment of the extent of LV trabeculations in three dimensions was possible for a comprehensive description of this cardiac abnormality. We therefore derived a three-dimensional echocardiographic method for the assessment of the extent of noncompacted myocardium which may be useful for the diagnosis of LVNC and to distinguish patients with LVNC from normal subjects.
This study was carried out at the Institute of Sports Medicine and Science, a division of the Italian National Olympic Committee and “Sapienza” University of Rome.
A population of 17 Caucasian patients, evaluated from June 2008 to December 2010, with diagnoses of LVNC on the basis of current clinical and echocardiographic criteria, were enrolled to participate the study. Echocardiographic criteria for the diagnosis of LVNC were considered the presence of two or more among the three commonly used criteria, namely, the Chin index, which is based on an X/Y ratio < 0.5 in diastole (where X is the distance from the epicardial surface to the bottom of the trabecular recess, and Y is the distance from the epicardial surface to peak of the trabeculae) ; the Jenni index, which is based on a ratio of noncompacted to compacted myocardium > 2 measured at end-systole ; and the Stöllberger criteria, which are based on the presence of more than three trabeculations protruding from the LV wall, apically to the papillary muscles, visible in a single imaging plane.
Additionally, to increase the pretest probability for this diagnosis, the echocardiographic findings of LVNC had to be confirmed by cardiac magnetic resonance imaging or accompanied by one or more of the following clinical criteria: documentation of a similar appearance in first-degree relatives; positive familiar history of sudden cardiac death or hypertrophic cardiomyopathy; associated neuromuscular disorders; complications such as systemic embolization, LV systolic dysfunction, and/or heart failure symptoms; and electrocardiographic abnormalities. According to Petersen et al. cardiac magnetic resonance criteria for the identification of patients with LVNC are considered the presence of a ratio of noncompacted to compacted myocardium of 2.3 in diastole.
Basic neurologic evaluation, including clinical history and physical examination, was carried out in each patient.
Atrial fibrillation was an exclusion criterion, because it made it impossible to perform three-dimensional echocardiographic acquisition with current technology.
A population of 26 elite athletes was randomly selected from the Italian Olympic rowing team and enrolled to participate the study during periods of intensive training. All the athletes have been involved in vigorous training protocols for ≥3 years.
Finally, a population of 49 healthy sedentary controls were also enrolled to take part in the study; they were volunteers among medical students of “Sapienza” University with ≤3 hours of regular exercise practice per week.
All subjects enrolled in the study underwent complete cardiologic evaluation, including medical history, physical examination, resting 12-lead electrocardiography, and two-dimensional and three-dimensional echocardiographic examinations. All subjects agreed to take part in the study, and the protocol was approved by institutional review boards.
Echocardiographic examinations were performed in each subject using a Philips iE33 (Philips Medical Systems, Andover, MA) equipped with an S3 probe (2–4 MHz) for two-dimensional and Doppler measurements and an X3-1 matrix-array transducer (1.9–3.8 MHz) for three-dimensional examinations. All acquisitions were performed in the echocardiography laboratory of the Department of Cardiovascular and Respiratory Sciences at “Sapienza” University by expert cardiologists, with specific training in three-dimensional echocardiography.
Two-dimensional assessment of LV cavity diameters, wall thickness, left atrium, and aortic root diameter were performed according to the current European Association of Echocardiography and American Society of Echocardiography criteria.
Transthoracic harmonic real-time three-dimensional echocardiography was performed from the apical four-chamber view using the full-volume technique. Wide-angle acquisition was performed, and a high frame rate (30 ± 4 Hz) was obtained using the seven-beat acquisition protocol, in which seven consecutive wedge-shaped subvolumes are consecutively sampled with a trigger to the R wave of the electrocardiogram.
At least three acquisitions were obtained for each subject and analyzed offline using commercially available software (QLAB version 7.0; Philips Medical Systems).
LV end-diastolic and end-systolic volumes, stroke volume, and ejection fraction were obtained using a semiautomatic border detection technique, as previously described. LV mass was calculated using a biplane measurement obtained from a three-dimensional data set, as previously reported and validated.
The three-dimensional echocardiographic sphericity index (3DSI) was calculated, according to Mannaerts et al. , as LV end-diastolic volume divided by the volume of a sphere, the diameter of which is the LV major end-diastolic long axis.
For the purpose of this study we specifically defined the trabeculated LV volume (TLV) as the difference between the LV end-diastolic volume measured by delineating the endocardial border at the bottom of the recesses and the end-diastolic volume delineated at the peak of the trabeculae.
In detail, the end-diastolic frame was identified, the gain level was optimized, and smoothing was reduced to visualize the endocardial border; by adjusting the cutting planes, the true apical four-chamber and two-chamber views were obtained, avoiding foreshortening, and then five reference points on the septal, lateral, anterior, and inferior aspects of the mitral annulus and at the apex were located. The software automatically delineated the endocardial contour of the left ventricle in the entire acquired data set. Subsequently, by rotating the long-axis cutting planes over 180°, the automatic tracing was reviewed by the operator, and the endocardial contour was modified to include or exclude the trabeculae and recesses ( Figure 1 ). The papillary muscles were excluded from tracing and included in the cavity. End-diastolic volume was first calculated by tracing the endocardial border at the interface between the noncompacted and compacted myocardium (the bottom of the recesses) and then by tracing the endocardial border at the peak of trabeculations; the TLV was obtained from the difference of the two values ( Figure 2 ).
The TLV as here defined included both trabeculae and blood inside the recesses and was expressed in milliliters. The TLV was also normalized by LV end-diastolic volume (TLV%) as the expression of the proportion of LV cavity occupied by trabeculae.
Offline analysis was performed by an investigator blinded to patients’ clinical data.
Continuous data are expressed as mean ± SD and categorical data as frequencies. Two-tailed P values < .05 were considered statistically significant. Differences between mean values for continuous variables were assessed using unpaired Student’s t tests or Mann-Whitney tests for variables with nonnormal distributions.
Differences in terms of age and body surface area were assessed using analysis of variance with multiple post hoc Bonferroni’s tests. Differences in terms of male proportion among groups were assessed using χ 2 tests.
Multiple Pearson’s simple correlations were calculated to test the association between TLV and TLV% and other parameters.
Receiver operating characteristic curve analysis was performed to identify optimal cutoff values for three-dimensional echocardiographic TLV and TLV% to identify patients with LVNC among our study population.
To assess the reproducibility of TLV and TLV%, measurements were repeated in a random sample of 10 patients with LVNC and 10 healthy controls by the same investigator (intraobserver variability) and by an additional reader (interobserver variability). Both were blinded to each other’s and patients’ data. Interobserver and intraobserver variability was assessed using intraclass correlation coefficients (ICCs) with 95% confidence intervals (CIs) and coefficients of variation (CVs; i.e., the absolute difference in percentage of the mean of repeated measurements).
Statistical analysis was carried out using SPSS version 15.0 (SPSS, Inc., Chicago, IL).
Demographic and Clinical Features
No difference in terms of age was detected between patients with LVNC (37 ± 21 years) and control subjects (26 ± 3 years, P = .077) or athletes (27 ± 2 years, P = .408). Athletes had higher body surface areas (2.06 ± 0.20 m 2 ) than both controls (1.81 ± 0.18 m 2 , P < .001) and patients with LVNC (1.83 ± 0.26 m 2 , P < .001), while no differences were identified between the last two. The proportion of men was higher among athletes ( n = 24 [92%], P = .007), while no differences were found between controls ( n = 12 [71%]) and patients with LVNC ( n = 28 [57%]).
Clinical characteristics of patients with LVNC are reported in Table 1 . Electrocardiographic abnormalities were common among patients with LVNC ( n = 13 76%) and included left-axis deviation ( n = 4), nonspecific repolarization abnormalities ( n = 4), complete left bundle branch block ( n = 2), frequent premature ventricular contractions ( n = 2), and Wolff-Parkinson-White pattern ( n = 1). Seven patients had positive family histories, consisting of similar appearance of LVNC in a first-degree relative ( n = 3), hypertrophic or dilated cardiomyopathy ( n = 2), and sudden cardiac death ( n = 2). Six patients had symptoms of congestive heart failure ( n = 2) or decreased ejection fraction ( n = 4); one young patient required heart transplantation. Nine patients underwent cardiac magnetic resonance, which confirmed the diagnosis of LVNC on the basis of a noncompacted/compacted ratio > 2.3. Three patients had cryptogenic stroke (documented by positive findings on cranial computed tomography) or transient ischemic attack. Finally, one patient had a neuromuscular disorder (esophageal achalasia).
|Patient||Gender||Chin index||Jenni index||Stöllberger criteria||Ejection fraction (%)||CMR||Positive family history||ECG abnormality||NMD||CHF||Stroke/TIA|