Left ventricular noncompaction (LVNC) has been extensively studied over the last years, and an increasing number of cases have been reported worldwide, with a large proportion comprising young and asymptomatic subjects, including athletes. The current epidemic of LVNC is likely the consequence of several causes, that is, the increased awareness of the disease and the refined cardiovascular imaging techniques. The current diagnostic methods, based uniquely on definition of morphologic findings, do not always resolve the overlap of a physiological myocardial architecture comprising a prominent trabecular pattern from a mild phenotypic expression of the real disease. Appropriate criteria for identification and management of LVNC in athletes have, therefore, become a novel challenge for cardiologists and sport physicians, who are required to solve the question of diagnosis and appropriate management in the setting of pre-participation cardiovascular screening. Indeed, although it is important to timely identify a true myocardial disease, to reduce the burden of adverse cardiac event in a young athlete, in contrast, a misdiagnosis of LVNC may lead to unwarranted restriction of the athlete lifestyle, with detrimental psychological, social, and economic consequences. This review report has been planned, therefore, to help physicians in diagnosing and managing athletes presenting with a morphologic pattern suggestive of LVNC with specific focus on criteria for advising sport participation.
Left ventricular noncompaction (LVNC) is a myocardial disorder phenotypically characterized by increased trabeculation of left ventricular (LV) chamber, typically 2-layered myocardium with a thin subepicardial compacted layer and a noncompacted thicker hypertrabeculated layer. Phenotypically, LVNC may present with various degrees of LV dilation and dysfunction. Recent observations suggest a larger prevalence of the disease than previously described; Paterick and Tajik reported that cases identified by echocardiographic examinations grew from 0.4% in 2010 to 1.0% in 2013. Greutmann et al reported similar experience, with a progressively increasing number of new diagnoses from 1984 to 2006 in a tertiary care referral center. The newly diagnosed cases comprise an increasing proportion of asymptomatic subjects, including athletes, with a mild phenotypic expression of the disease. Recently, Gati et al reported the occurrence of hypertrabeculation (defined as at least 3 trabeculations of at least 3 mm) in 18.3% of 1,146 athletes, with a subset of 8.1% athletes fulfilling the conventional echocardiographic criteria for LVNC. The increased trabecular pattern incidentally observed in an otherwise normal heart should not be considered sufficient for diagnosis of LVNC. Appropriate criteria for identification and management of LVNC in athletes have, therefore, become a novel challenge for cardiologists and sport physicians, who are required to solve this question in the setting of pre-participation cardiovascular (CV) screening.
Classification and Epidemiology
The World Health Organization and the European Society of Cardiology initially described LVNC as an unclassified cardiomyopathy. In contrast, the American Heart Association classified LVNC as a primary genetic cardiomyopathy. Inconsistency in the LVNC classification was related to uncertainty regarding the true nature of the disease, that is, whether it represents a primary cardiomyopathy or a phenotypic variant of other cardiomyopathies. LVNC may occur as a part of syndromes like Barth syndrome, myotonic dystrophy, mitochondrial myopathies, and zaspopathies, in association with congenital heart disease, or in neuromuscular disorders. LVNC has been identified in families with hypertrophic cardiomyopathy (HCM) or dilated cardiomyopathy (DCM), and recent genetic advances have demonstrated a common genetic pathway related to abnormal sarcomeric proteins, supporting the hypothesis that LVNC may be a phenotypic variant of other cardiomyopathies. A specific genetic background, however, and familiar occurrence have been identified in many cases of LVNC, suggesting that it may be considered as a primary myocardial disease. A recent classification of the myocardial diseases, named MOGE, proposed by Arbustini et al, is based on 5 characteristics of cardiomyopathies, that is, morphofunctional features (M), organ involvement (O), genetic inheritance (G), etiology (E), and heart failure stage (S). In the context of morphofunction characteristics, the LVNC pattern has been described as a distinct phenotype that may be present either in isolation or associated with DCM or HCM.
Prevalence of LVNC in the adult population ranges from 0.014% to 1.3%. Incidence of the disease in infants and children has been reported as 0.81 and 0.12 cases per 100,000 per year, respectively, representing ∼9% of all cardiomyopathies in childhood. Gender prevalence is heterogeneous in studies, and there is not a clear predominance gender related ; data on ethnic differences are scarce, but few studies seem to suggest increased prevalence in black subjects.
Pathogenesis
At the end of the fourth week of the gestational age, rich trabeculations develop from the endocardial layer with a centripetal growth. The trabeculations are aimed to increase the endocardial surface area, enabling efficient myocardial perfusion in the absence of coronary arteries. At the end of the eighth gestational week, parallel to the development of the coronary circulation, the trabecular pattern becomes compacted starting from the basal segments toward the apex and increasing the thickness of the epicardial layer. Some of the trabeculae coalesce to produce the papillary muscles, whereas the intertrabecular recesses collapse to form the capillary circulation. An arrest in the compaction process could explain persistent myocardial trabeculations and the double-layered appearance of the myocardial wall. This hypothesis explains why LVNC may be identified in fetuses and present at birth. However, identification of this condition in adult patients has raised the question whether an acquired phenotype of the disease may exist. In this regard, an interesting hypothesis is that the observed myocardial changes may represent the combined effect of genetic and environmental factors. Changes in sarcomere proteins may be genetically determined but might lead to morphologic LV changes only later in life, in a way similar to what occasionally occurs in HCM or DCM. Additionally, trabeculations have been observed in subjects with hypertensive heart disease, congenital heart disease, heart failure, sickle cell anemia, athletes, and pregnant women, all conditions characterized by an increased pressure or volume load. Therefore, it has been hypothesized that increased trabeculation in these conditions may represent an adaptation of the myocardial architecture to increased load conditions. This hypothesis may justify the increased prevalence of trabeculation described in trained athletes. An additional support comes from a recent study, where the morphologic cardiac changes occurring during the pregnancy (i.e., a physiological model of chronic increased preload) were evaluated; of note, 25% of 102 women developed increased trabeculations during pregnancy, with 8% fulfilling criteria for LVNC. Interestingly, the vast majority of women showed a complete regression of trabeculation in the post-partum period. Finally, D’Ascenzi et al recently reported longitudinal data on training-induced hypertrabeculation in basketball players. Compared with pre-training, athletes developed an increased trabecular pattern at peak of training, associated with electrocardiographic (ECG) repolarization changes that were more prominent in Afro-Caribbean athletes.
Genetics
Several studies investigated the genetic background of LVNC. A pathogenic mutation has been identified in up to 41% of the affected subjects, and a familial occurrence has been described in up to 64% of the probands. However, the prevalence of genetic mutations is difficult to assess, in that only selected candidate genes have been assessed (instead of whole-genomic sequencing), and usually only probands with more striking phenotypic changes have, so far, been tested for DNA anomalies. In adult patients, the autosomal dominant inheritance with incomplete penetrance is the usual technique of disease transmission. Hoedemaekers et al identified 11 genes mutations, with 6 of them related to sarcomere proteins ( MYH7, MYBPC3, TNNT2, ACTC, TPM1 , and TNNI3 ), 2 to calcium-handling proteins ( PLN and CASQ2 ), and the remaining 3 to LMNA, LDB3 , and TAZ genes. Specifically, mutation in myosin heavy chain ( MYH7 ) represents the most frequent defect, observed in 17% of mutation carriers.
Sarcomeric protein mutations have been reported in a significant proportion of patients with LVNC (29%). In detail, mutations in 5 sarcomeric proteins were described, including MYH7 in most cases, followed by MYBPC3 , ACTC1, TNNT2 , and TPM1 . Finally, 2 recent studies describe an association between the LVNC phenotype and mitral valve prolapse and bradycardia, with causative mutation in the HCN4 gene. These observations substantiate the concept that LVNC is a genetically heterogeneous disease with different mutations implicated. As in other cardiomyopathies, there is no close phenotype-genotype relation, and most of the identified mutations are not specific of LVNC but may also be present in other cardiomyopathies.
Genetic studies, so far, have not been carried out in athletes. Routine genetic testing in athletes with suspicious trabecular pattern is, however, controversial as the expected prevalence of positive mutations is very low and a negative test result does not exclude the disease. Therefore, genetic analysis should be limited to athletes with a positive personal or family history for cardiomyopathies (namely HCM, DCM, or LVNC) and/or a history of a cardiac event.
Clinical presentation and outcome
A classic triad of symptoms has been reported in patients with LVNC, including heart failure, tachyarrhythmias, and embolic events. However, the clinical profile of LVNC is broader and the outcome difficult to characterize, in consideration that several studies were retrospective in design and have often included small groups of patients and with short follow-up in most cases. Indeed, the outcome of LVNC seems extremely variable in relation to the age at diagnosis, the association with other congenital heart defects or syndromes, and the presence of symptoms at presentation. The disease diagnosed early in childhood (particularly in the first year of life) is associated with an unfavorable outcome, with high mortality rate (13%) cardiac dysfunction (62%), and arrhythmias (33%).
The large incidence of adverse events in adult population initially described by Oechslin and Jenni (death or transplantation in 48%, heart failure in 53%, and embolic events in 24% followed up for a period of 44 ± 40 months) was likely related to the selected study population, characterized by patients with advanced stage of the disease, dilated LV cavity, and severe LV dysfunction. More recent studies supported the concept that the absence of symptoms at the time of diagnosis is associated with subsequent favorable outcome. Specifically, Lofiego et al reported no major CV events in a subset of 17 asymptomatic patients with LVNC, compared with 31% complications in 48 symptomatic patients, followed up for a period of 46 ± 44 months. A large French registry, including 105 cases of LVNC followed up for 2.3 + 1.5 years, reported incidence of heart failure in 31%, death in 11%, cardiac transplantation in 9%, embolic events in 9%, and ventricular arrhythmias in 7%. In this study population, the major factors associated with incidence of complication were the presence of symptoms related to LV dysfunction (i.e., New York Heart Association class 3 or 4), the increased LV filling pressure, and the decreased ejection fraction (EF). Greutmann et al described the clinical outcome in a population of 115 patients with LVNC, including adolescents and adults, followed up for a period of 2.7 years. A favorable outcome was reported in 37 asymptomatic patients, with no CV events registered; conversely, of 95 symptomatic patients, the combined incidence of events, including death or transplantation, was 30%.
From a large cohort of 2,742 subjects free of clinically recognized CV disease enrolled in the Multi-Ethnic Study of Atherosclerosis study, those subjects with marked LV trabeculations, matching the criteria for diagnosis of LVNC, and no CV symptoms at the study entry developed no significant changes in LV volumes and function over almost a 10-year follow-up. Despite the limitation of being retrospective, this study highlights the concept that the incidental finding of LV hypertrabeculation in asymptomatic patients with low-pretest probability of CV events and a negative family history has a benign outcome.
Ventricular Arrhythmias and Sudden Cardiac Death
The most relevant concern regarding advising participation in competitive sport in athletes with suspected LVNC is the risk for ominous ventricular tachyarrhythmias and sudden cardiac death (SCD). The overall incidence of these complications in the published studies is listed in Table 1 . Sustained ventricular tachycardia occurs from 0% to 9%, with the highest figure in the Oechslin’s study (who included the more severely affected patients). Incidence of SCD ranges from 0.1% to 18% in the published studies. Brescia et al reported SCD in 6.2% of pediatric patients. As in previous study, most of them had abnormal LV dimensions and/or dysfunction (14 of 15 patients), and the only 2 patients without LV dysfunction developed tachyarrhythmias before the event.
| Author | Year | Sample size (n) | Age (years) | Gender (male) | Mean Follow- up (months) | Incidence of SCD ∗ | Incidence of Sustained VT † |
|---|---|---|---|---|---|---|---|
| Oechslin | 2000 | 34 | 42±17 | 74% | 44 | 18% | 9% |
| Murphy | 2005 | 45 | 37±17 | 62% | 46 | 2% | 0% |
| Aras | 2006 | 67 | 41±18 | 66% | 30 | 9% | 2% |
| Stollberger | 2007 | 86 | 52±16 | 76% | 51 | 4% | – |
| Lofiego | 2007 | 65 | 45±16 | 37% | 46 | 5% | 6% |
| Habib | 2011 | 105 | 45±17 | 66% | 28 | 0.1% | 6% |
| Greutman | 2012 | 132 | 41±17 | 35% | 32 | 9% | 4% |
A relevant consideration regarding the risk stratification for SCD in patients with LVNC is the observation that most of them were already symptomatic and usually presented heart failure symptoms, associated with the most severe morphologic phenotype. Therefore, death in most cases was not unexpected and was unlikely the first clinical appearance of the disease. Rather than being a mere issue of definition, this observation underlines the concept that SCD appears extremely rare in patients with LVNC, if asymptomatic and without LV dysfunction. If we focus specifically on incidence of SCD in athletes, it is worthy to underline that LVNC has not been reported in major epidemiologic studies as the primary cause of death. A few pathologic studies reported an overlapping phenotype of LVNC with dilated or HCM (thus underestimating the real prevalence), but even in the latest autopsy reports, LVNC has never been reported as primary or unique cause of SCD in athletes. Therefore, in young and asymptomatic patients with LVNC, including athletes, the actual risk of sports-related sudden death is extremely low.
Diagnostic methods and criteria
Echocardiography is the imaging technique of first choice to reach the diagnosis, in consideration of the wide availability and the relatively low cost of this technique. Three major echocardiographic criteria have been largely in use to reach diagnosis of LVNC: (1) X/Y ratio <0.5, measured in diastole from parasternal short-axis and apical views, where X is the compact layer and Y is the distance between epicardial surface and peak of trabeculation (Chin index ); (2) N/C >2.0 in systole, where N is the noncompacted layer and C is the compact layer, in association with evidence of intertrabecular recesses filled from LV cavity by color Doppler and predominant localization of noncompaction in the lateral, apical, or inferior walls and in the absence of coexisting cardiac disorders (Jenni index ); and (3) >3 trabeculations, apically to the papillary muscles, with same echogenicity of the myocardium and perfusion of the intertrabecular spaces from the LV cavity (Stollberger criterion ). Of the criteria so far in use, only those proposed by Jenni et al were validated by comparing the echocardiographic data with pathologic specimens in 7 patients who died or underwent cardiac transplantation. Despite the reported good accuracy for the diagnostic criteria, it appears that the specificity is still low.
Additional echocardiographic criteria have subsequently been developed to increase the diagnostic accuracy, also using the latest technology advances. Gebhard et al demonstrated that addition of the Jenni criteria to the thickness of compact layer (in noncompacted segments), with threshold value <8 mm in systole (in apical segments), helped the differentiation from normal hearts and proved to be accurate for LVNC diagnosis. Poscolieri et al showed, in a group of 22 asymptomatic athletes with LV hypertrabeculation suggestive for LVNC, that a reduced thickness of the compact layer <5 mm in diastole was able to identify subject with decreased LV systolic function. Therefore, it appears that measurement of the compact layer may have an additional value in the risk stratification of LVNC, supporting the hypothesis that when the compact layer is not substantially decreased, trabeculations are likely not to be associated with negative outcome, whereas when the compact thickness is markedly reduced, CV complications are more likely to occur. Additionally, a clear abnormality of the LV filling pattern should be considered as evidence for a pathologic myocardial disease and poor outcome. Indeed, in the study by Oechlin et al, all the evaluated patients had diastolic dysfunction.
Recently, evaluation by speckle tracking echocardiography showed that patients with LVNC present an abnormal twisting pattern characterized by rotation at the basal and apical level predominantly in the same direction, which is also known as “rigid body rotation.” This pattern is different from the normal LV twisting, characterized by counterclockwise basal and clockwise apical rotation, followed by end-systolic clockwise basal and counterclockwise apical rotation. Using 3-dimensional echocardiography, Caselli et al introduced the concept of the “trabeculated left ventricular volume” (TLV%), calculated as LV end-diastolic volume obtained by including and excluding the trabeculae in the cavity contour and normalized by the LV end-diastolic volume. In this analysis, patients with LVNC had ∼24% of LV cavity occupied by trabeculae, for example, fourfold increase compared with athletes and untrained subjects.
Because of its higher spatial resolution, CMR has been largely applied to the evaluation of the disease and 2 CMR-derived criteria are currently used. Petersen et al proposed a cut-off value of 2.3 for the noncompacted to compacted myocardium ratio (N/C) in diastole useful for differential diagnosis between LVNC and other conditions including normal hearts, athletes, and patients with different cardiomyopathies, with a sensitivity of 86% and specificity of 99%. Subsequently, Jacquier et al, introduced the trabeculated LV mass (calculated as global LV mass − compacted LV mass) and reported that the threshold of 20% for the trabeculated mass had a sensitivity of 94% and a specificity of 94% for diagnosis of LVNC. The value of the CMR, other than the mere morphologic assessment of LV trabeculation, is the capability for identifying the presence and extent of late gadolinium enhancement (LGE), as a surrogate marker of myocardial fibrosis. Nucifora et al reported positive LGE in up to 55% of patients with LVNC and showed that the extent of LGE was significantly related to abnormal clinical features (ECG abnormalities and tachyarrhythmias) and LV dysfunction. Evaluation of LGE is extremely important and should be included in all CMR studies, in that has the potential to identify patients with a pathologic substrate and at risk for ventricular tachyarrhythmias and eventually sudden death.
Electrocardiography
Electrocardiographic data are missing in most of the studies, and there are few reviews of the ECG patterns in LVNC. From the available observations, it appears that a normal ECG finding is rare in LVNC and observed in a minority from 6% to 13% of patients. Table 2 lists the ECG abnormalities reported in the reviewed literature. There are no specific changes that should raise specific suspicion for LVNC and certain patterns, such as R/S wave increased voltages suggestive for LV hypertrophy and overlap with ECG findings of trained athletes. However, certain ECG alterations, such as negative T waves >1 mm in depth in ≥2 leads (particularly, V4 to V6, II and aVF, or I and aVL), ST-segment depression ≥0.5 mm in depth in any lead, or pathologic Q waves (>3 mm in depth or >40 ms in duration) in ≥2 leads (except III and aVR) are more compatible for a cardiomyopathy. It is worthy to mention that a significant amount of black athletes (almost 13%) may show negative T waves in the anterior precordial leads that are not associated with structural disease. Because many black subjects may also show increased LV trabeculation, the association with negative T waves in the anterior leads may not uncommonly raise the problem of differential diagnosis with LVNC ( Figure 1 ).
| ECG abnormality | Prevalence |
|---|---|
| Atrial enlargement | 19% |
| Atrioventricular block I degree | 15% |
| Right Bundle Branch Block | 3-4% |
| Left Bundle Branch Block | 15-44% |
| Left ventricular hypertrophy | 18-41% |
| T wave inversion | 16-41% |
| ST segment abnormalities | 9-51% |
| WPW ∗ (only pediatric) | 8-17% |
| Q waves | 9% |
| Left axis deviation | 9% |
| Prolonged QT interval | 9-52% |
| Poor R wave progression | 7% |
Proposed management algorithm in athletes with suspected LVNC
Recent reports highlight the need for revised criteria of differential diagnosis between LVNC and normal morphologic variants, when a prominent LV trabecular pattern is found in otherwise normal heart. Figure 1 represents an example of LV hypertrabeculation, associated with abnormal ECG pattern, exemplifying the gray area of the differential diagnosis of LVNC. Resolution of this diagnostic dilemma may be difficult in the individual case: although it is important to timely identify a true myocardial disease, to reduce the burden of SCD in a young athlete, in contrast, misdiagnosis of LVNC may lead to unwarranted restriction of the athlete from competitions and raises unjustified clinical alarm, with detrimental psychological, social, and economic consequences.
We propose to address this problem through a clinical approach, based on the consideration that diagnosis of LVNC on purely morphologic criteria seems to be evanescent in terms of consistency (several, varying criteria are in place for echocardiography, and CMR) and appear to have low predictive value regarding the clinical outcome. Therefore, to solve the conundrum of appropriate diagnosis and risk stratification regarding the decision of sport participation, we recommend the algorithm shown in Figure 2 . What should be assessed in the scenario of an asymptomatic athlete with positive morphologic criteria for LVNC:
- (1)
Clinical and family history: Athletes should be carefully interrogated on conditions associated with LVNC, such as complex congenital defect, neuromuscular disorders, or a history of DCM or HCM in the relatives. A negative (or unavailable) family history does not exclude the LVNC diagnosis. We believe that when the patient’s or family history shows indicators for LVNC, HCM, or DCM, the genetic testing should be considered.
- (2)
Electrocardiography offers only an ancillary aid to the diagnosis, in that no changes are reported specific in patients with LVNC, and there is a potential overlap between the physiological ECG changes seen in athletes and the abnormal findings in patients with LVNC. However, the presence of markedly abnormal ECG pattern, in particular repolarization abnormalities, should raise suspicion for a pathologic substrate, and require additional investigation to solve diagnosis. Exercise test should be always performed, and when arrhythmias are found on electrocardiogram at rest or on exercise, an ambulatory ECG monitoring should be obtained. Indeed, recording of ventricular tachyarrhythmias (i.e., NSVT, frequent and/or polymorphic premature ventricular contractions) should be evaluated with caution and support the suspicion for a cardiomyopathy.
- (3)
LV systolic and diastolic functions: Impaired myocardial function is the strongest element to support the diagnosis of a myocardial disease; specifically, LVEF in athletes is, in our experience, usually >55% and never <50%. Therefore, we suggest the cutoff of EF <50% as threshold, useful to raise suspicion for the disease. In borderline cases, that is, when EF is in the lower range (50% to 54%), subclinical impairment of systolic function may be detected by a reduction in global longitudinal strain; it has been proposed that a global longitudinal strain >15% supports the evidence of a normal myocardial contraction, whereas values less than this cutoff may be the initial hallmark of a cardiomyopathy. Finally, diastolic function is always within normal limits in a normal heart.
