Normally, the left ventricle is smooth and compact in human adults and children. It has up to 3 normally prominent trabeculations. Non compactant cardiomyopathy (NCCM) was classified as a genetic cardiomyopathy in 2006 because it affects the myocardium mainly and it has a genetic etiology with an autosomal dominant inheritance mechanism [1]. It is characterized by abnormally prominent and, excessive trabeculations on the endocardial surface of the left ventricle with deep recesses, extending into the left ventricular wall, in direct communication with LV cavity.
NCCM can be presented by thromboembolic complication, Heart failure, Arrhythmia or sudden cardiac death. The range of thromboembolic events is 10-37% [2-4] including cerebrovascular accidents, transient ischemic attacks, pulmonary embolism, and mesenteric infarction. Thus, the diagnosis of NCCM is crucial in terms of family screening, anticoagulation and management of severe LV dysfunction.
Multiple imaging modalities specially transthoracic echocardiography( TTE) and cardiac magnetic resonance ( cardiac MRI ) were used for description, characterization , diagnosis and differential diagnosis of the non compact left ventricle. The two-dimensional echocardiography is the most commonly used technique because it is widely available and cheap. The prominent trabeculations can be misdiagnosed as false tendons, or LV apical thrombus .Sometimes the trabeculations may curve into left ventricular cavity in a manner that cannot be identified by the planar 2D-TTE cuts. Also, the diagnosis may be missed totally in some cases and other cases require serial 2D-TTE before clear diagnosis [5-7]. In contrary to the previous data which lead to underestimation of the diagnosis of NCCM by 2D-TTE, others found that, the dilated left ventricle may allow the endocardium to be visualized in details and over interpretation of normal trabeculations as hypertrabeculated left ventricle [8]. So, the prevalence of NCCM as well as the diagnostic accuracy of 2D-TTE cannot be fulfilled.
Three-dimensional echocardiography has the potential to provide better anatomic and functional details of cardiac structure and function compared with 2D-TTE, so Three-dimensional echocardiography (3D-TTE) is more accurate in diagnosis, characterization, and description of the extent of the affected myocardium. [7]. Abnormal excessive trabecular projections and intertrabecular recesses were visualized simultaneously. Also, the distinction between the noncompact and compact LV was easily visualized. In contrast to 2DE, 3DE allows sectioning in multiple planes and an unlimited number and angles of such planes can be used. Also, it allows tracking of trabeculations in multiple directions from base to apex [7], and allows direct visualization of mural thrombi within the recess which is challenging and usually misdiagnosed by 2D-TTE.
The cardiac MRI cannot be used as gold standard imaging technique because there are no agreement in-between the two modalities. In 2D-TTE, trabeculations are best seen at end diastole and the ratio of compacted to non compacted myocardium is measured in end systole as>2, While cardiac MRI diagnose the same ratio as >2.3 at end diastole [9]. In single cases, the excessive trabeculations may be only seen on cardiac MRI but not on 2D-TTE [10, 11]. Actually, cardiac MRI is helpful in differential diagnoses, to rule out other structural heart diseases, and to confirm the diagnosis in difficult echocardiographic cases [12].
The objective of the current study is to assess the potential of three-dimensional echocardiography, in the diagnosis of LVNC and to differentiate between the presence and absence of mural thrombi in the intertrabecular recess.
Patients and methods
We studied 43 patients, including 29 patients with isolated LVNC (group I) and 14 patients without LVNC (group II), retrospectively .Patients with isolated LVNC presented to our echo laboratory over 3 years period, started in May 2005.The age of the studied group ranged between 36-78 years (56.24+23.14 years).There was 9 females (33%) and 20 male patients in group I, and group II included 5 females (31%) and 11 male patients. The studied group A was subdivided into 2 subgroups according to the presence or absence of LV apical thrombus as follows:
Group IA: patients With LVNC, but no LV apical thrombus =23.
Group IB: patients with LVNC and apical thrombus=6.
Group II patient's including14 elite Athletes with athletes' heart came for routine check-up. This group was selected as a control group as there is a clear cut differentiation from LVNC in spite of sharing some morphological characteristics which can be misdiagnosed as LVNC.
Exclusion criteria:
Patients with ischemic cardiomyopathy or ischemic heart disease as documented by the history of myocardial infarction, coronary angiography with or without PCI, history of CABG operation, or positive stress thallium results.
Patients with congenital heart diseases, or associated congenital anomalies discovered during the echocardiographic diagnosis.
Children and age less than 18 years old.
Patients with documented HOCM, or infiltrative cardiomyopathy.
Difficult echo studies with > 4 out of 17 LV segments not well visualized.
Echocardiographic criteria of LVNC [13-15]:
Absence of other cardiac malformations (by definition).
Presence of excessive trabeculations > 3 mainly considered in the same image plane.
Blood flow in the intertrabecular recesses as detected by colored Doppler that communicates with ventricular cavity.
Involvement of the apex and / or apical and / or mid-ventricular segments of the lower and / or lateral wall. More than 80% of patients with this condition have a lack of maturity in one or more of these predilection sites.
The affected myocardium is dual layer, consisting of a thin outer layer and a epicedial thickened trabecular endocardial inner layer, and the ratio between the non-condensed and the condensed layer is > 2.0.
There is no international consensus on the final diagnostic criteria for the condition, and one or more of these echocardiographic criteria [13-15] can be found in other conditions or diseases affecting the left ventricle. Also, there is no international agreement on the degree of severity of the LVNC as mild, moderate and severe degrees.
Eighteen out of our studied group I patients had negative stress nuclear test for ischemic heart disease as they were referred based on their abnormal ECG. And the remaining 11 patients had no significant coronary artery disease in coronary angiography as they were referred based on ventricular tachycardia with LV systolic dysfunction to rule out ischemic cardiomyopathy. Six patients discovered during familial screening and they were asymptomatic in spite five patients had abnormal ECG and LV systolic dysfunction.
All study patients underwent an echocardiographic examination with the Vivid 7 Dimension ultrasound imaging system (General Electric, GE), which is equipped with a 3.5 MHz transducer for 2D-TTE, and a 4 MHz 4X matrix array transducer for live/real time 3D-TTE. The raw data sets were stored in the hard disc of the ultrasound system and later on offline analysis using an echo view software version 5.2 of the TomTec imaging system (Munich, Germany). The quality of the RT-3DE data set was graded according to image quality of the 17 LV segment model as good (0 to 2 segments not visible), moderate (3 or 4 segments not visible), or poor (_5 segments not visible).
Two-dimensional echocardiography:
A comprehensive 2D-TTE was done for all patients and LV systolic function was assessed using M-Mode measurement if no regional wall motion abnormality and Simpsons modified biplane view if there is any regional wall motion abnormality and all results checked by the eye ball assessment to avoid errors during measurements. Diastolic function assessed with pulsed Doppler with the sample volume at the tip of the mitral valve and confirmed by tissue Doppler technique. The wall thickness and LV dimensions was measured according to standard echo techniques and LV apex visualized for presence or absence of thrombus. Cardiac valves and pericardium was assessed to rule out associated cardiac abnormalities.
Three-dimensional echocardiography:
The three dimension data set were acquired with the patient suspending respiration for a few seconds (5-7seconds) during full volume acquisition. Care was taken to adequately center the LV in the scanning sector, and to minimize LV foreshortening by maximizing the distance between the apex and the base. Trabeculations were assessed by dividing the LV walls were divided according to the ASE 17 segment model into apical, middle and basal segments. The number of trabeculations arising from each of these walls was counted in the sequential short axis sections from apex to base, while cross checking with the apical 4-chamber and 2-chamber views visualized simultaneously on the computer screen. In this way, long trabeculations were counted once and duplicate counting of trabeculations was also avoided. Three-dimensional dataset was cropped using multiple angulations showing intramural thrombi among the recess with different echo density from the trabeculations which revealed uniformly homogenous surfaces. These 3D-TTE findings were helpful in making a confident diagnosis of thrombi existing among trabeculations.
Statistical analysis
All analyses were done using SPSS version 16 statistical package. Continuous variables were summarized as mean and standard variation and two-tailed student t-test was used. The number of patients expressed as percentage (%) for categorical variables and analyzed with fishers exact test. For all statistical assessment, p value <0.05 was considered statistically significant.
Results
Clinical Patient characteristics
Our study patients group I included 20 males (69%) and 9 Females (31%) with mean age 58+12.08 and 53.8+10.44 respectively (p=0.29, NS). The 23 patients in subgroup IA were not statistically different than group IB as regard risk profile, clinical presentation, or associated co-morbid conditions. The patients in group I main clinical presentation was palpitation 51.7%, due to both ventricular tachycardia 37.9% and atrial fibrillation 31% . Subgroup IA had no significant difference of either ventricular tachycardia or atrial fibrillation than subgroup IB . Although not statistically significant, the subgroup IB with LVNC and LVTH had significantly higher percentage of thromboembolic complication and presented more with SOB than subgroup IA patients. There are 7 patients with no cardiac symptoms 24.1 %, with 6 of them discovered during familial screening. Most of our studied group I patients had abnormal ECG with no significant difference between subgroups. Chronic kidney disease observed in 27.6% of our patients which is mostly due to hypertension which was detected in 58.6% of our studied group I patients.
Table 1 showed the clinical characteristics of the studied group I .
Table 2 showed the comparison of subgroup IA and IB.
Table 3 showed the characteristics of patients in group IB.
Comparison between two and three-dimensional echocardiography:
Among our studied group I patients, 2D-TTE diagnosed 26 cases with LVNC and 2 cases as LVTH, and 1 case with both LVNC and LVTH. Also 2 cases from group II was misdiagnosed as LVNC based on the number of LV trabeculations(4 and 5 respectively), the diagnosis later on corrected by three dimension echocardiography on the basis of the count number of trabeculations was within the normal range (3 in both cases0 with non compacted to compacted ratio were 1.2,and 1.3 respectively). Also, there were no clinical symptoms, no family history, normal 12 leads resting ECG, no arrhythmia in 24 hours holter monitor tests done for the patient in group II as a routine checkup. Also, both patients had normal diastolic function and systolic function was on the higher side of normal. Group I patients had depressed LV systolic dysfunction with EF less than 40% in 24 patients (82.2%) and the other 5 patients 17.8% with preserved LV function, had fair LV systolic function in the range of 40-50 %, but not normal LV systolic function. 2D-TTE had a sensitivity of 89.7% and specificity of 85.7% in the diagnosis of LVNC.
The trabeculation number assessed by 3D-TTE was significantly higher than the number assessed by 2D-TTE (mean+ standard deviation, p= <0.0001) as well as the ratio of compact to non compact myocardium (mean +standard deviation , P=0.012 ) . As regard The presence of mural thrombi the 2D-TTE was able to diagnose 3 patients in group IB, but diagnosed 4 cases as LVTH instead of LVNC in subgroup IA with a sensitivity of 50% and specificity of 86.96%.Although the 2D-TTE has a low sensitivity 33.33% as regard the presence or absence of mural thrombi within LV trabecular recess, it has a very good specificity 95.6% to rule in the diagnosis of the presence of both LVNC and LVTH. The trabeculations seen by 2D-TTE were mainly in the apical lateral and mid lateral wall of the LV with less trabeculations seen in the apical septal region. In contrary to 3D-TTE which detects the trabeculations mainly in apical septal, apical lateral and mid lateral wall respectively.
Figure 1 show the sensitivity and specificity of 2D-TTE in LVNC diagnosis.
Figure 2 show the sensitivity and specificity of 2D-TTE in LVNC plus LVTH diagnosis.
Discussion
The three dimensional echocardiography is more superior than the 2D-TTE in the diagnosis of LVNC, and in the differentiation between LVNC with and without mural thrombi in-between the LV trabeculations, provided that cautious diagnosis before labeling the patient. Among our group I 100% patients had diastolic dysfunction and abnormal systolic function (EF <40% in 82.7% of patients, EF=40-50% in 5 patients (17.8%), together with abnormal ECG in 93.1% of patients and after exclusion of normal variants and conditions morphologically mimicking LVNC as in group II patients.
Three Dimensional echocardiography is especially important in the diagnosis of LV thrombus inside the trabecular recesses which is very difficult for 2D-TTE as evident by its low sensitivity 33.3% in group IB patients as well as the diagnosis of 4 cases in group IA which later on ruled out by three dimensional echocardiography. This is in agreement with Chamoun AJ and colleagues [16], who stated that 2D-TTE images are mostly suggestive of the presence of LV apical thrombus due to LV apical foreshortening, but 3D-TTE is a good alternative to 2D-TTE especially in diagnosis, follow up and care of patients with suspected LVTH. Also, in case series by Dunacan K and colleagues [17] found that the 3D-TTE is of additional benefit to 2D-TTE for many reasons (1) comprehensive evaluation can be done even in a single apical dataset in short time even without holding the breath when the patient is not fully co-operative. (2) It detects the exact point of attachment of the thrombus to the LV wall, and (3) It can detect any echolucency in the thrombus indicating its lysis or embolisation and is very helpful in diagnosis and prognosis.
Our results showed that 3D-TTE is very helpful, in supplement to 2D-TTE, in the detection of segmental involvement of the trabeculation and of more accurate detection of the compact to non compact ratio which help in the rule out diagnosis in the 2 athletes in our study. This is in agreement with Bodiwala K and colleagues [18] who stated that the better assessment of segmental involvement, the characteristic honey comb appearance of the non compact myocardium as well as correct measurement of non compact to compact myocardial ratio ,all detected by 3D-TTE, are very important in the differential diagnosis of LVNC from other conditions like apical hypertrophic cardiomyopathy, left ventricular hypertrophy, left ventricular dysplasia, endocardial fibroelastosis, and intracardiac masses such as thrombus and tumors and thrombus.
The ability of 3D-TTE to view the LV trabeculations in multiple dimensions and at various angles is totally different than the 2 dimensional echo principles. This is the reason that 3D-TTE can detect significantly higher number of trabeculations in our study results and in agreement with other studies [17] which clarified the difference between the two modalities as regard comprehensive assessment of the trabecular number. This is due to the fact that the software algorism which allows the 3D-TTE to detect the points between the myocardial recesses and thus better assessment of trabecular anatomy. Thus 3D-TTE gives more confidence in diagnosis of LVNC before labeling the case.
Conclusion
Three- dimensional echocardiography is a valuable technique in the cautious diagnosis of LVNC and in differential diagnosis of the case from other conditions with good confidence .Also it is a good supplement for 2DTTE in the challenging cases of LVTH with LVNC.
Limitations of the study:
The absence of international agreement in diagnostic criteria and severity of LVNC and absence of gold standard diagnostic technique are the main limitations for our study.
The small number of cases of this relatively recent condition requires more studies to help for better understanding of the different aspects of the problem as regard therapy, follow up and prognosis of LVNC.
