“I’ve learned that I still have a lot to learn”—Maya Angelou
Left ventricular noncompaction cardiomyopathy (LVNC) remains a largely underinvestigated
and poorly understood diagnosis. The number of peer‐reviewed articles published on
LVNC has grown dramatically over the past decade. Clinicians and scientists around
the globe have advanced our understanding of the genetics, diagnostics, therapeutics,
and outcomes for adult and pediatric patients with LVNC. Yet, there continues to be
disagreement about diagnostic criteria, management, and classification of this complex
phenotype.
1
,
2
,
3
,
4
,
5
,
6
In this issue of the Journal of the American Heart Association (JAHA), Vaidya and
colleagues present data on identifiable clinical and imaging criteria that may predict
mortality in adults with LVNC.
7
The current report consists of 339 patients (median age, 47.4 years) with confirmed
LVNC, as diagnosed by either echocardiography or cardiac magnetic resonance imaging
(CMR). The median follow‐up was 6.3 years, during which time 69 patients died. On
multivariable Cox regression analysis, the authors found that age, left ventricular
ejection fraction (LVEF) <50%, and noncompaction extending from the apex to the mid
or basal segments were associated with all‐cause mortality. Not surprisingly, patients
with a formal diagnosis of LVNC had reduced overall survival compared with the expected
survival of an age‐ and sex‐matched US population. In addition, those patients with
noncompaction isolated to the apex of the left ventricle (LV) and those with an LVEF
>50% had similar survival to the general population. Overall, this is an important
addition to the existing literature and helps provide a partial framework for the
management of these patients.
LVNC remains a heterogeneous disease with multiple possible concomitant phenotypes.
We have described these previously and characterized the possible findings into 9
distinct subtypes.
8
Briefly, these subtypes are as follows: (1) the isolated or benign form of LVNC, (2)
the arrhythmogenic form of LVNC, (3) the dilated form of LVNC, (4) the hypertrophic
form of LVNC, (5) the “mixed” form of LVNC, (6) the restrictive form of LVNC, (7)
the biventricular form of LVNC, (8) the right ventricular hypertrabeculation with
normal LV form, and (9) the congenital heart disease form of LVNC. The authors thoughtfully
excluded patients with congenital heart disease from their analysis. However, the
other subtypes were not completely identified, which may impact some of the findings
being reported. The dilated form of LVNC (subtype 3) is characterized by depressed
systolic function, which is often accompanied by LV dilation. The outcome of this
group is similar to that in patients with isolated dilated cardiomyopathy. This provides
support to the finding of the investigators that an LVEF <50% was an independent predictor
of all‐cause mortality. Although not described in detail within the article, Table
5 reports the finding of “any right ventricular dysfunction” as a variable associated
with overall mortality (hazard ratio [HR], 1.98; 95% CI, 1.10–3.54). This may represent
patients with the biventricular form of LVNC. Historically, these patients are difficult
to diagnose by echocardiography and are typically identified by use of CMR. As only
118 subjects (35%) in the cohort in this study underwent CMR, it is worth considering
that the biventricular form of LVNC (subtype 7) may have been present but was not
identified.
An important consideration in interpreting the data from this recent report is recognizing
the reported experience in children. Brescia and colleagues reported on 242 children
diagnosed with LVNC at Texas Children's Hospital.
9
In this retrospective evaluation, the presence of cardiac dysfunction (LVEF <55%)
was strongly associated with mortality (HR, 11; P<0.001). The clinical presentation
and symptoms in this pediatric cohort are similar to those in the current report.
The presenting symptoms in the childhood cohort were as follows: (1) congestive heart
failure (25%), (2) abnormal cardiac examination (19%), abnormal ECG or chest x‐ray
film (16%), arrhythmia (10%), chest pain (9%), and syncope (5%). These findings reported
in children have some obvious commonalities with the Mayo Clinic data. ECG abnormalities
were present in 87% of the patients. Repolarization abnormalities were associated
with increased mortality (HR, 2.1; P=0.02). Eighty children had an arrhythmia that
resulted in increased mortality (HR, 2.8; P=0.002). During the evaluation period,
there were 15 cases of sudden cardiac death (6.2%). Nearly all patients who experienced
sudden death had abnormal cardiac dimensions or evidence of cardiac dysfunction. Notably,
no patient with normal cardiac dimensions and function without evidence of a proceeding
arrhythmia died. These findings mirror the report by Vaidya and colleagues in many
ways. The presenting symptoms in both groups are similar. This is remarkable considering
that many of the patients described from Texas Children's Hospital were too young
to reliably voice symptoms, such as chest pain. No patient with normal LV size or
systolic dysfunction in the absence of arrhythmia died. This is the same finding reported
in the Mayo adult cohort with the exception that arrhythmias were not identified as
an independent predictor of mortality.
The lack of robust arrhythmia analysis in the current report is a limitation and should
be recognized by those caring for children and adults with LVNC. The arrhythmogenic
form of LVNC (subtype 2) is an important component of longitudinal surveillance in
these patients. As noted above, the presence of an arrhythmia and/or a repolarization
abnormality resulted in increased mortality. Although identifying patients with normal
LV systolic function and isolated apical LV trabeculations may provide some comfort
to providers, these are not the only important phenotypic characteristics. The impact
of significant arrhythmias cannot be underestimated and mandates thoughtful surveillance.
The opportunity to accrue meaningful arrhythmia data is underleveraged in the current
management of cardiomyopathies, including LVNC. This is unfortunate as it would add
valuable information in the development of risk stratification instruments and inform
clinical decision‐making.
Although the data set provides typical clinical information about the LVNC phenotype,
the article fails to provide any genetic information. The authors appropriately recognize
the need for genotyping in this population. The focus of their study was to identify
clinical and imaging variables that are routinely available in practice. However,
in 2020, genetic testing is routinely available in clinical practice and should be
considered in the overall assessment of patients with LVNC. We have reported on the
importance of potential genetic triggers and genotype‐phenotype correlations in adult
patients with LVNC.
10
In our study, 190 adults from 174 families with concern of LVNC by echocardiography
were prospectively analyzed by CMR and whole exome sequencing. This provided the foundation
to attempt genotype‐phenotype correlations. We included 425 controls to assess for
genetic variants of interest (VOIs). In one of the largest reported CMR studies in
LVNC, we found 138 VOIs in 102 unrelated patients in 54 genes that have been previously
associated with LVNC or other cardiomyopathy phenotypes. VOIs were identified in 68
of 90 probands (76%) with LVNC and 34 of 84 probands (40%) with LV hypertrabeculation.
We also identified 0, 1, and ≥2 VOIs in 72, 74, and 28 probands, respectively. More
importantly, we found that the presence of an increasing number of VOIs in individual
patients correlated with several phenotypic markers, including the ratio of noncompacted/compacted
myocardium (P<0.001) and LVEF (P=0.01). Furthermore, the presence of sarcomere gene
mutations was associated with increased occurrence of late gadolinium enhancement
(P=0.004). A report from the Netherlands by van Waning and colleagues again documented
the importance of using genetic information in the risk stratification of children
and adults with LVNC.
11
On review of 327 unrelated cases of LVNC, the highest risk for cardiac events in both
age groups was related to LV systolic dysfunction in mutation carriers. Of note, mutations
in MYH7 had a low risk for major cardiac events. Li and colleagues also reported on
the importance of pathogenic mutations predicting adverse outcomes in an adult Chinese
cohort with LVNC.
12
These 3 reports underscore the important role that genotyping plays in risk stratification
for patients with LVNC. This also strengthens the hypothesis that LVNC is a distinct
and genetically triggered cardiomyopathy.
As Maya Angelou humbly noted, we are learning that we still have a lot to learn about
many things, including LVNC. Routine care for patients with LVNC continues to be greatly
confounded by the lack of consensus about the cause, diagnostic criteria, surveillance,
and management of this increasingly common diagnosis. Furthermore, over the past 3
decades, the ability to differentiate “benign” from “pathologic” has become increasingly
challenging given the morphologic spectrum and diverse populations described in the
literature. In a recent report from the PESA (Progression of Early Subclinical Atherosclerosis)
study, de la Chica et al found that vigorous physical activity was associated with
a higher prevalence of CMR‐detected LVNC.
13
This association was maintained using the Petersen, Jacquier, and Grothoff CMR criteria
for LVNC.
3
,
4
,
14
This reflects our limited understanding of the drivers of noncompaction but does suggest
that underpinning genetics may help differentiate those with “real disease.” The authors
are to be congratulated on their report. However, one must recognize that these identified
predictors are only a small piece in a large and complex puzzle.
Disclosures
None.