Post on 11-Aug-2021
transcript
Long-Term Outcomes of Childhood Left Ventricular Non-Compaction
Cardiomyopathy: Results from a National Population-Based Study
William Y. Shi, MBBS, PhD1,2,3; Margarita Moreno-Betancur, PhD2,3; Alan W. Nugent,
MBBS4; Michael Cheung, MBChB5; Steven Colan, MD6; Christian Turner, MBBS7; Gary
F. Sholler, MBBS8; Terry Robertson, MBBS9; Robert Justo MBBS9; Andrew Bullock,
MBBS10; Ingrid King2, Andrew M. Davis, MBBS5;2,3,5 Piers E. F. Daubeney, DM
MBBS11,12*; Robert G. Weintraub, MBBS2,3,5,*; for the National Australian Childhood
Cardiomyopathy Study
1. Department of Cardiac Surgery, Royal Children’s Hospital, Melbourne, AUSTRALIA
2. Murdoch Children’s Research Institute, Melbourne, AUSTRALIA
3. University of Melbourne, AUSTRALIA
4. Department of Pediatrics, University of Texas Southwestern, Dallas, TX, USA
5. Department of Cardiology, Royal Children’s Hospital, Melbourne, AUSTRALIA
6. Department of Cardiology, Boston Children’s Hospital, MA, USA
7. Department of Cardiology, Children’s Hospital at Westmead, Sydney, AUSTRALIA
8. Department of Cardiology, Women’s and Children’s Hospital, Adelaide, AUSTRALIA
9. Department of Cardiology, Mater Children’s Hospital, Brisbane, AUSTRALIA
10. Department of Cardiology, Princess Margaret Hospital, Perth, AUSTRALIA
11. Department of Paediatric Cardiology, Royal Brompton Hospital, London, UNITED
KINGDOM
12. National Heart and Lung Institute, Imperial College, London, UNITED KINGDOM
* Equal senior co-authors
Total word count: 3432 words
Corresponding author:Robert G. WeintraubDepartment of CardiologyRoyal Children’s HospitalMelbourne, Victoria, AUSTRALIARobert.Weintraub@rch.org.au
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
252627282930313233
ABSTRACT
Background: Long-term outcomes for childhood left ventricular non-compaction
(LVNC) are uncertain. We examined late outcomes for children with LVNC enrolled in a
national population-based study.
Methods: The National Australian Childhood Cardiomyopathy Study includes all
children in Australia with primary cardiomyopathy diagnosed <10 years of age between
1987 and 1996. Outcomes for LVNC subjects with a dilated phenotype (LVNC-D) were
compared to those with dilated cardiomyopathy (DCM). Propensity-score analysis was
used for risk factor adjustment.
Results: There were 29 subjects with LVNC (9.2% of all cardiomyopathy subjects) with
a mean annual incidence of newly diagnosed cases of 0.11 per 100,000 at-risk persons.
Congestive heart failure was the initial symptom in 24 (83%) of 29 subjects, and 27
(93%) had a dilated phenotype (LVNC-D). The median age at diagnosis was 0.3
(interquartile interval 0.08 – 1.3) years of age. The median (interquartile interval)
duration of follow-up was 6.8 (0.7-14.1) years for all subjects and 24.7 (23.3 – 27.7)
years for surviving subjects. Freedom from death or transplantation was 48% (95% CI 30
– 65%) at 10 years after diagnosis and 45% (95% CI 27-63%) at 15 years. By competing
risk analysis, 21% of LVNC subjects were alive with normal LV systolic function and
31% were alive with abnormal function at 15 years. Propensity-score matching between
LVNC-D and DCM subjects suggested a lower freedom from death/transplantation at 15
years after diagnosis in the LVNC-D subjects (LVNC-D: 46% (95% CI 26-66%) vs.
DCM: 70% (95% CI 42-97%), p=0.08). Using propensity-score inverse probability of
treatment weighted Cox regression, we found evidence that LVNC-D was associated with
a greater risk of death or transplantation (HR 2.3, 95% CI 1.4-3.8, p=0.0012).
Conclusions: Symptomatic children with LVNC usually present in early infancy with a
predominant dilated phenotype. Long-term outcomes are worse than for matched children
with DCM.
2
12
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
CLINICAL PERSPECTIVE
What is new?
Long-term outcomes for children with left ventricular non-compaction (LVNC)
are uncertain.
The NACCS uniquely represents the longest and most complete longitudinal
cohort study of childhood cardiomyopathy.
This population-based study defines the incidence, presentation and long-term
outcomes for LVNC diagnosed during childhood.
Risk-adjusted analyses in this study showed that subjects with LVNC and dilated
physiology tended to experience worse survival than matched subjects with
dilated cardiomyopathy, with a two-fold higher risk of death or transplantation.
Clinical implications:
Our findings support consideration of LVNC as a distinct cardiomyopathy
phenotype in pediatric patients.
Children with LVNC who develop heart failure at an early age may benefit from
specialist care delivered in paediatric heart failure centres
3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
INTRODUCTION
Left ventricular (LV) non-compaction (LVNC) is an increasingly recognized type of
cardiomyopathy (1) characterized by the presence of an extensive trabeculated
myocardium separated into 2 distinct compacted and non-compacted layers (2, 3). The
American Heart Association classifies it as a genetic cardiomyopathy caused by arrested
myocardial development (2, 4, 5) although on occasion cases have been reported to
develop postnatally in response to alterations of preload (6, 7). The European Society of
Cardiology lists LVNC as an unclassified cardiomyopathy (8).
In childhood LVNC, different phenotypes have been observed including dilated,
hypertrophic, restrictive, isolated (not associated with abnormal physiology) and that
associated with congenital heart disease, or in combination (1, 9). Whereas isolated
LVNC is thought to have a relatively good prognosis, (10), LVNC with a dilated
phenotype is associated with worse short-term outcomes (1, 11).
Long-term outcomes for children with LVNC remain uncertain. In addition, it is unclear
whether outcomes are determined solely by the extent and severity of cardiac
dysfunction, or whether the presence of LVNC confers additional adverse prognostic
information.
We examined late outcomes for children with LVNC enrolled in the National Australian
Childhood Cardiomyopathy Study (NACCS), a population-based, longitudinal cohort
study with long-term follow-up extending to 30 years (12-14)
4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
METHODS
The NACCS study includes all children in Australia with primary cardiomyopathy who
were diagnosed at 0 - <10 years of age, between 1 January 1987 and 31 December 1996.
Local institutional ethics committee approval was obtained from participating centres.
The methodology has been detailed previously (12, 15, 16). In brief, study subjects were
enrolled through site visits undertaken from 1997-2007 by the same three investigators
visiting all 9 pediatric cardiac centers and an additional 12 hospitals caring for children
with cardiac conditions (Appendix 1). The data that support the findings of this study are
available from the corresponding author upon reasonable request.
Study subjects were identified from multiple sources, including local cardiology
databases, echocardiographic records and ICD-9 CM medical record codes. Children with
structural heart disease and progressive neuromuscular disorders were excluded.
Examination of centralized records compiled by the Australian Bureau of Statistics
indicated that there were no children with sudden death as their initial symptom who
were diagnosed at autopsy. Ethics committee approval was obtained from each
participating institution.
Prospective follow-up was arranged for any subjects who were not being followed
regularly. Study questionnaires were used to extract standardized clinical and
echocardiographic data obtained during follow-up.
Definitions
Cardiomyopathies were classified during the course of site visits according to the existing
World Health Organization cardiomyopathy classification (5) by a single observer after
reviewing all relevant investigations, including all available cardiac imaging(12).
Study subjects were classified based on echocardiographic findings as having LVNC if
there was the characteristic morphological appearance comprising: 1) multiple
trabeculations, 2) deep intertrabecular recesses seen on color flow and 3) a 2-layered
structure of the myocardium with ratio of non-compacted to compacted myocardium of
5
1
23
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
>2:1 in systole (17). An experienced pediatric pathologist with cardiac expertise
reviewed available pathological specimens, including cardiac histology.
A family history of cardiomyopathy was defined by the presence of at least one affected
first- or second-degree relative with LVNC identified from case notes, or from familial
screening.
The presence of congestive heart failure was based on signs and symptoms recorded by
the attending physician. Sudden death was defined as a sudden and unexpected death in a
child within 4 hours of new symptoms (14).
Serial echocardiograms at presentation, those closest to 3, 6, 12 months after diagnosis
and then at yearly intervals during follow-up were reviewed by a single observer and
measurements of LV dimensions, wall thickness and fractional shortening (FS) in those
without regional wall motion abnormalities were expressed as Z-scores based on body
surface area or age in the case of FS (18, 19).
LVNC with dilated pathophysiology (LVNC-D) was defined if the LV was dilated with
reduced systolic function (FS Z-score < -2.0 and/or LV ejection fraction (EF) < 45%).
Similar echocardiographic criteria were used to classify DCM subjects in the absence of
LVNC (13). Restrictive pathophysiology in LVNC subjects was diagnosed from
echocardiography and/or cardiac catheterization in subjects with impaired diastolic
function and normal LV size with preserved systolic function (12). A hypertrophic
phenotype was defined by otherwise unexplained LV free wall or septal hypertrophy
(diastolic wall thickness Z score >2) (15).
Statistical analysis
Statistical analyses were performed using IBM SPSS Statistics Version 23 and R
statistical package version 3.3.1. Incidence rates were calculated from age-specific
population at risk between 1987-97 (20, 21). Continuous variables were described as
median (interquartile interval) and compared, where relevant, using the Wilcoxon sum-
6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
rank test. Categorical variables were described through proportions and compared using
Fisher’s exact test.
Analysis of long-term outcome
Our primary endpoint was the combined endpoint of death or transplantation from the
time of presentation. The endpoints of death and transplantation were combined into a
composite primary endpoint as cardiac transplantation was not widely utilized in the early
part of the study.
The Kaplan-Meier method was used to estimate long-term transplantation-free survival
probabilities, and the log-rank test was used for unadjusted comparison of these between
LVNC-D and DCM subjects.
Normalization of LV function in LVNC-D subjects was considered to have occurred on
the first occasion when the LV end-diastolic dimension (LVEDD) Z score was < 2.0 and
the LV FS Z score was > -2.0. Cumulative incidence curves were constructed for the
competing events of 1) normalization of LV function, 2) death or transplantation before
normalization and 3) remaining alive, transplant-free and without normalization. The
cumulative incidence curves were constructed so as to represent the first-encountered
competing event.
Due to the relatively low case numbers, only univariable Cox regression analysis was
performed to examine the association between variables and the combined outcome of
death or transplantation. Variables analyzed in a univariate fashion were: age at
presentation, gender, body surface area, presentation with congestive heart failure,
consanguinity, family history of cardiomyopathy, baseline LVEDD Z-score and baseline
FS Z-score.
The current value and the current rate of change of echocardiographic measurements
(LVEDD Z score and FS Z-score) were both analyzed as time-dependent variables. The
current rate of change was approximated as the difference between the two most recent
7
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
measurements before the event divided by the time between the measurements.
Normalization of LV function was also analyzed as a time-dependent variable.
Propensity-score analysis
Cases with missing values were excluded from regression and propensity-score analyses.
Two separate propensity-score methods were used to risk-adjust for differences in clinical
profiles between LVNC-D and DCM subjects: Propensity-score matching and inverse
probability of treatment weighting. A propensity-score was generated for each case by
performing a logistic regression with LVNC-D binary indicator as the dependent
variable. Baseline clinical and investigative co-variables were used. These were: age at
presentation, gender, year of presentation, body surface area, congestive heart failure at
diagnosis, consanguinity, family history of cardiomyopathy, LVEDD Z-score and FS Z-
score on the first available echocardiogram.
For propensity-score matching, subjects were matched 1 to 1 on their propensity-score
without replacement using the greedy matching method with a caliper width equal to 0.2
of the standard deviation of the logit of the propensity-score. In the matched sample,
paired t-tests were used for continuous data, whereas McNemar’s test, which compares
discordance of 2 dichotomous outcomes, was used to compare categorical variables.
Similarly, long-term freedom from death/transplantation was compared using the test
proposed by Klein and Moeschberger (22).
For the propensity-score inverse probability of treatment weighting analysis (IPTW), the
contribution of each individual to Kaplan-Meier estimates or Cox regression is weighted
by the inverse of the probability that they belong to the corresponding group, which is
estimated from the propensity scores obtained from the aforementioned logistic
regression. Intuitively, these weights generate a pseudo-population in which both groups
to be compared are balanced with respect to the measured confounders. This enables the
comparison of these groups and estimation of the population-average (marginal) effect of
LVNC-D on the combined endpoint of death or transplantation, expressed as a hazard
ratio. The analysis was performed according to best practice recommendations. (23)
8
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
RESULTS
Incidence rate of LVNC
During the 10-year period there were 29 newly diagnosed cases of LVNC, comprising
9.2% of the NACCS study population. The mean annual at-risk population (number of
Australian children aged 0-10 years) during this period was 2,532,400 (20, 21) leading to
a mean annual incidence of newly diagnosed cases of 0.11 (95% CI 0.08-0.16) per
100,000 persons at risk. Table 1 shows mean annual incidence according to age at
presentation. The highest incidence was in the first year of life (0.83 (0.51-1.30) per
100,000) representing a more than 5-fold increase compared to subjects aged >1 year at
time of presentation. During this same time period there were 175 newly diagnosed cases
of DCM who did not have LVNC (ref 13)
Presentation and clinical characteristics
Table 2 shows demographics, clinical characteristics at presentation and principal
echocardiographic findings of those diagnosed with LVNC. Of the 29 cases, 27 had
dilated phenotype. One subject had restrictive phenotype and one subject with Barth
syndrome had isolated LVNC without associated cardiac dysfunction. Twenty of 29
(69%) of subjects were males. Associated anomalies in 9 subjects included Barth
syndrome in 7, left bronchial stenosis in 1 and bile duct hypoplasia requiring liver
transplantation in the remaining subject. Two first cousins, both with consanguineous
parents, were found to have a truncating mutation in the ALPK3 gene and one subject
with Barth syndrome had a Complex I respiratory chain enzyme deficiency.
Echocardiography was available in all 29 cases. Additional confirmation of LVNC was
available from characteristic findings on left ventricular angiography in 12 subjects and
direct examination of the heart (explant or autopsy) in 11.
The most common symptom at diagnosis was congestive cardiac failure in 24 (83%)
cases; two subjects were diagnosed on routine family screening, including 1 with Barth
syndrome who has never developed symptoms. The remaining 3 subjects had a diagnosis
of LVNC made on the basis of feeding difficulties, an abnormal chest X-ray and onset of
9
12
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
supraventricular tachycardia in a premature infant. Further clinical details are presented
in Table 2.
Light microscopic findings from autopsy, explantation at the time of transplantation or
endomyocardial biopsy were abnormal in 19/21 available cases and included myocyte
hypertrophy, nuclear irregularity and endocardial fibroelastosis.
Long-term outcomes
There were no subjects lost to follow-up. The median (interquartile interval) duration of
follow-up was 6.8 (0.7-21.7) years for all cases and 21.4 (11.9-24.5) years for survivors.
Of the 27 subjects with dilated phenotype, 19 died or underwent heart transplantation
(including 5 cases of sudden death). The subject with restrictive physiology died during
follow-up while the subject with Barth syndrome and isolated LVNC remains alive with
normal cardiac function 24 years after diagnosis. The two related subjects with an
ALPK3 mutations progressed from a dilated to a hypertrophic phenotype.
Normalization of left ventricular systolic function occurred in 8 of 29 (28%) subjects, all
with a dilated phenotype, including of 4 of 7 subjects (57%) with Barth syndrome and
both subjects who developed a hypertrophic phenotype. Two of 8 subjects who had
normalization of left ventricular function later died suddenly, two redeveloped a dilated
phenotype with one undergoing cardiac transplantation 14 years later. Two have normal
LV size but with reduced function at late follow-up, while the remaining two continue to
have normal LV size and systolic function at late follow-up.
Freedom from death or transplantation was 69% (95% CI 49- 83%) at 1 year, 52% (33-
68%) at 5 years, 48% (30-65%) at 10 years and 45% (27-63%) at 15 years. Figure 1
shows survival for all cases and Figure 2 shows the cumulative proportion of cases
experiencing each competing endpoint after presentation, for all 29 cases. At 15 years
after diagnosis, 21% of LVNC cases were alive with normal LV function while 31% were
alive without transplantation or normalization of LV function.
10
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Of 8 subjects in the NACCS registry with Barth syndrome, 7 had LVNC. Of these, 2
died (both with sudden death), 1 underwent transplantation, 2 have normal LV systolic
function at latest follow-up, and 2 are alive with a persisting dilated phenotype. There
was a nominal association towards Barth syndrome patients having better survival at 15
years (Barth: 71% (39-100%) vs. non-Barth: 36% (16-57%), p=0.08).
Table 3 shows the association between clinical factors and the combined endpoint of
death or transplantation determined by univariable Cox regression. Favorable prognostic
variables comprised male gender, presence of a positive family history and a higher FS Z
score at any time during follow-up. A larger LVEDD Z score during follow-up was
associated with worse prognosis. Familial LVNC and male gender were unrelated to
survival in the absence of Barth syndrome.
Of the 29 cases of LVNC, long-term therapy with an angiotensin converting enzyme
inhibitor (ACEI) was used in 13, a beta-blocker in 6 and either an ACEI or beta-blocker
was used in 16.
Comparison between LVNC-D and DCM
Compared to the 175 cases with isolated DCM, cases with LVNC-D were more likely to
be male, and of lower weight at presentation. Baseline echocardiographic measurements
were similar (Table 4). Unadjusted freedom from death or transplantation was similar
between LVNC-D and DCM groups (Figure 3A).
Propensity-score matching yielded 24 matched pairs. The clinical and echocardiographic
characteristics amongst matched cases are presented in Table 5. Amongst matched pairs,
there was a nominal association (p=0.08) for subjects with LVNC to have worse
transplant-free survival compared to those with DCM at 15 years as shown in Figure 3B.
Long-term therapy with an ACEI or a beta-blocker was used in 12 of 24 (50%) matched
LVNC-D cases, and 11 the 24 (46%) of the matched DCM cases.
11
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Propensity-score analysis using inverse probability of treatment weighted Cox regression
showed that LVNC was associated with a greater risk of death or transplantation (HR 2.3,
95% CI 1.4-3.8, p=0.0012) (Figure 3C).
12
1
2
3
4
DISCUSSION
The NACCS uniquely represents the longest and most complete longitudinal cohort study
of childhood cardiomyopathy with a median follow-up of survivors of nearly 25 years.
Ongoing follow-up of this cohort has provided novel information for long-term outcomes
for childhood cardiomyopathy (13, 14).
Our population-based study defines the incidence, presentation and long-term outcomes
for LVNC diagnosed during childhood. In particular, long-term outcomes for children
with LVNC associated with a dilated cardiomyopathy phenotype were worse than those
for matched children with DCM.
Incidence
Previously considered an uncommon cardiomyopathy type, LVNC accounted for nearly
10% of all childhood cardiomyopathy cases in the NACCS study. The incidence of newly
diagnosed cases in our study was 0.11/100,000 at risk subjects per year, with the highest
incidence being during the first year of life, similar to children with dilated and
hypertrophic cardiomyopathy from the NACCS registry. The overall proportion of cases
with LVNC in our study was similar to a single-center study from Texas Children’s
Hospital (24), and somewhat greater than the 4.8% reported from the North American
Pediatric Cardiomyopathy Registry (PCMR) (1). In another study from the UK, 3% of
children aged <16 years with myocardial disease resulting in new onset heart failure were
diagnosed with LVNC (25). The higher proportion of LVNC subjects in our study may
reflect systematic case classification by a single observer.
Inheritance
Multiple modes of LVNC inheritance have been described, X-linked recessive or
autosomal dominant are most common, while autosomal recessive and mitochondrial
inheritance have also been described. Our study provides further evidence for a genetic
basis for LVNC, with 31% of subjects having a positive family history of
cardiomyopathy, which is comparable to published reports of 16-44% (24, 26-29). A
13
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
significant proportion of these subjects had Barth syndrome, accounting for the male
preponderance.
Long-term outcomes for LVNC
In our study, long-term outcome was worse than previously described in some other
series, with only 45% of subjects alive and transplant-free 15 years after presentation.
Our findings contrast with a 10-year transplant-free survival of 60-86% in studies from
Toronto and Texas (24, 27), and 93% in a study from Japan (26). The higher mortality in
our population-based study most likely reflects a high proportion of sick, young infants,
many of whom died within one year after presentation. The low mortality in the Japanese
study may conversely reflect a high proportion of asymptomatic cases as a result of a
systematic childhood screening program in Japan.
Rates of death or transplantation for subjects in our study were also higher than those
reported by the PCMR (1). The PCMR study comprised children with variable LVNC
phenotypes, including a greater proportion without cardiac dysfunction. Outcomes for
children with LVNC and dilated physiology in the PCMR were also poor with a hazard
ration for death or transplantation of >6 compared to subjects with isolated LVNC.
In our study, risk factors on univariable regression analysis for death/transplantation
comprised females, sporadic LVNC (non-familial) and worse LV systolic function. The
sole case with restrictive cardiomyopathy died. As in other studies, severity of systolic
dysfunction was the most important predictor of survival at any time during follow-up.
The association towards better survival in Barth syndrome patients is unexpected. We
postulate that this may be related to more systematic screening and earlier diagnosis of
LVNC given the presence of an underlying syndrome.
Comparison with dilated cardiomyopathy
The 15-year transplant-free survival of 45% for LVNC subjects is worse than the 56%
20-year survival reported from the NACCS for children with dilated cardiomyopathy
14
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
(13). Propensity-score analyses in this study showed that subjects with LVNC and dilated
physiology tended to experience worse survival than matched subjects with dilated
cardiomyopathy, and that LVNC-D was associated with a two-fold higher risk of death or
transplantation.
The frequency of utilization of remodeling therapy was similar in both the LVNC-D and
DCM groups, suggesting that better outcomes for the matched DCM were not due to the
use of remodeling heart failure therapy. However our study was not designed to examine
the impact of any particular heart failure therapy.
Our findings contrast with studies of adult subjects with DCM and LVNC, such as
Amzulescu and colleagues (30), who examined 162 subjects with dilated cardiomyopathy
undergoing cardiac MRI. Those who met MRI criteria for LVNC had outcomes similar to
those of remaining subjects, leading the authors to argue against a non-compaction
phenotype being a more severe form of dilated cardiomyopathy. These contrasting
findings most likely reflect a different spectrum of etiologies for both conditions in adult
subjects.
Higher rates of death or transplantation in children with LVNC, when compared to those
with isolated DCM, may reflect a different genetic basis and in some cases the absence of
a reversible etiology such a lymphocytic myocarditis, which is present in a significant
proportion of children with DCM (12, 13). We also speculate that the spectrum of genetic
etiologies responsible for childhood LVNC may differ from those of affected adults,
leading to a more severe clinical course for those who develop heart failure at a young
age. Rates of sudden death in LVNC subjects were comparatively high and an undulating
phenotype or fluctuating LV systolic function (24) were observed in a significant
proportion of subjects. Taken together, these findings support consideration of LVNC as
a distinct cardiomyopathy phenotype in pediatric patients.
The use of circulatory support and adult-based heart failure therapies in children with
dilated cardiomyopathy has increased in recent years. Routine screening combined with
15
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
cascade genetic testing may contribute to earlier diagnosis of cardiomyopathy in at-risk
family members. These developments are most likely responsible for the improved
outcomes noted in children with DCM who are receiving care in specialized pediatric
heart failure centers in the past 3 decades (31). Children with LVNC, who develop
symptoms at an early age and have worse outcomes than those with DCM, might
therefore benefit similarly from specialist care delivered in high volume pediatric centers.
Limitations
Our population-based study had a high proportion of symptomatic young subjects with
dilated physiology. Outcomes from the present study cannot be extrapolated to children
who have LVNC with other phenotypes, and to those without symptoms diagnosed
during routine family screening, or as a result of genetic testing. Genetic and
mitochondrial testing were not routinely available for all subjects during the study period,
and the proportion of cases with one of these etiologies may therefore have been
underestimated. The limited number of cases precluded a multivariable analysis of risk
factors for death or transplantation. Furthermore, propensity-score analyses, as any other
type of analysis, are inherently unable to account for unmeasured variables, including
more sophisticated parameters of left ventricular function.
CONCLUSIONS
In our population-based study, LVNC accounted for 9.2% of all cardiomyopathies
diagnosed in Australia during the first decade of life. Most subjects were diagnosed
during infancy with a dilated phenotype, following symptoms of congestive heart failure.
Survival free from transplantation at 15 years after diagnosis was 45%. Transplant-free
survival was worse for subjects with the greatest degree of left ventricular systolic
dysfunction at any time during follow-up, and subjects with LVNC and a dilated
phenotype had worse outcomes than matched subjects with DCM.
Funding sources
Supported by Grant 98001 from Royal Children’s Hospital Research Foundation, Grants
G98M0159, G04M1586, G05M2151, G07M3180 from National Heart Foundation of
16
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Australia, NACCS grant from Australia and New Zealand Children’s Heart Research
Centre. Dr William Shi is supported by the Royal Australasian College of Surgeons
Foundation for Surgery Peter King/Heart Foundation Research Scholarship in addition to
the University of Melbourne Viola Edith Reid and the RG and AU Meade Scholarships.
Dr Daubeney’s research was supported by the Biomedical Research Unit at the Royal
Brompton Hospital. The Murdoch Children’s Research Institute is supported by the
Victorian Government's Operational Infrastructure Support Program.
Disclosures
No conflict of interest disclosure
17
1
2
3
4
5
6
7
8
9
10
Table 1: Incidence of LVNC by age at presentation
Age at presentation (years) 0-<1yrs 1-<2yrs 2-<5yrs 5 -<10 yrs Total
Number of cases during 10-year study period (%) 21 (72.4) 3 (10.3) 4 (13.8) 1 (3.4) 29
Average annual at-risk population 253,954 253,685 761,737 1,262,971 2,532,347
Annual incidence rate/100,000 person-years at risk 0.83 0.12 0.053 0.0079 0.11
18
1
2
3
4
Table 2: Clinical and echocardiographic characteristics of all LVNC subjects
Clinical characteristics n=29
Age – median years (IQR)* 0.3 (0.1-1.3)
Male – n (%) 20 (69)
Year of presentation
1987-1991 17 (59)
1992-1996 12 (41)
Weight – median kg (IQ interval) 5.1 (3.7-9.0)
Dilated phenotype – n (%) 27 (93)
Parental consanguinity†– n (%) 4 (14)
Barth syndrome– n (%) 7 (24)
CHF at diagnosis – n (%) 24 (83)
Family history cardiomyopathy – n (%) 9 (31)
Baseline LVEDD Z-score – median (IQ interval)* 4.6 (0.2-6.0)
Baseline LVESD Z-score – median (IQ interval)* 7.0 (3.3-8.1)
Baseline FS Z-score – median (IQ interval)* -10.7 (-12.1 to -9.1)
Median (IQ interval) duration of follow-up in years 6.8 (0.7 – 24.0)Median (IQ interval) duration of follow-up in survivors in years 24.7 (23.3 – 27.7)
Number of subjects with death/transplant 20
Number of deaths 14
Number of transplants 6*On first available echocardiogram; BSA: body surface area; CHF: congestive heart failure; FS:
fractional shortening; IQ interquartile; LVEDD: left ventricular end-diastolic diameter; LVESD:
left ventricular end-systolic diameter; SD: standard deviation.
†There were 4 subjects with parental consanguinity including 2 siblings
At diagnosis, the median non-compacted-to-compacted myocardial thickness ratio in systole,
measured from echocardiography at the site of maximal involvement, was 2.6 (interquartile
interval 2.1-3.2). In diastole it was 2.3 (2.0-2.5).
Data were incomplete in 2 subjects, who were excluded from propensity-score analyses.
19
1
2
3
4
5
6
7
8
9
10
Table 3: Univariable predictors of death or transplantation after diagnosis of LVNC (n=29)
Variable HR 95% Confidence interval p value
Male* 0.39 0.15-1.00 0.049
Age at presentation (per year) 1.10 0.87-1.40 0.43
Age at presentation (per month) 1.01 0.99-1.03 0.44
Presentation with CHF 1.83 0.54-6.27 0.34
Family history of CMǂ 0.12 0.03-0.53 0.005
Consanguinity 0.61 0.14-2.67 0.51
Barth Syndrome 0.35 0.10-1.21 0.097
Baseline BSA 3.83 0.32-46.54 0.29
Baseline FS Z-score† 0.83 0.67-1.03 0.092
Baseline EDD Z-score† 1.05 0.89-1.24 0.55
LVNC ratio systole 2.33 0.88-6.14 0.088
LVNC ratio diastole 1.70 0.96-1.04 0.071
Current FS Z-score†‡ 0.78 0.66-0.91 0.002
Current EDD Z-score†‡ 1.25 1.02-1.54 0.03
Rate of change in FS Z-score†‡ 1.01 0.96-1.06 0.76* Values after exclusion of Barth syndrome subjects; Male: HR 0.28 (0.03-3.07), p=0.30; Family history: n/a† - per unit Z score‡ Variables treated as time-varying covariates in univariable Cox regression analysis
20
123
4567
Table 4: Comparison of clinical characteristics between LVNC with dilated
phenotype (LVNC-D) and those with isolated DCM.
LVNC-D DCM
Clinical characteristics n=27 n=175 p value
Age – median years (IQ interval) 0.3 (0.07-1.4) 0.6 (0.2-1.7) 0.26
Male – n (%) 19 (70) 77 (44) 0.013
Year of presentation
1987-1991 16 (59) 74 (42) -
1992-1996 11 (41) 101 (58) 0.14
Weight – median kg (IQ interval) 5.1 (3.6-9.0) 7.1 (4.3-11.3) 0.047
BSA (m2) – median (IQ interval) 0.3 (0.2-0.5) 0.4 (0.3-0.5) 0.061
Consanguinity 4 (15) 14 (8) 0.27
CHF at diagnosis 22 (81) 163 (93) 0.058
Family history of cardiomyopathy 8 (30) 26 (15) 0.092
Cardiothoracic ratio on CXR 66 (60-72) 65 (61-70) 0.75
Echocardiography
LVEDD Z-score – median (IQ interval) 4.8 (0.5-6.0) 4.4 (2.5-6.3) 0.50
LVESD Z-score – median (IQ interval) 7.2 (3.9-8.2) 6.5 (4.5-8.4) 0.75
FS Z-score – median (IQ interval) -11.2 (-12.1 to -9.9) -10.5 (-12.2 to -8.7) 0.46
BSA: body surface area; CHF: congestive heart failure; FS: fractional shortening; IQ
interquartile; LVEDD: left ventricular end diastolic diameter; LVESD: left ventricular
end systolic diameter; SD: standard deviation.
Data were missing in 2 LVNC subjects and 33 DCM subjects and these cases were
excluded from propensity-score analysis.
21
1
2
3
4
5
6
7
8
Table 5: Clinical characteristics of 24 propensity-score matched pairs.
LVNC-D DCM
Clinical characteristics n=24 n=24 p value
Age – median years (IQR) 0.3 (0.2-1.8) 0.5 (0.3-1.1) 0.88
Male – n (%) 17 (71) 16 (67) >0.99
Year of presentation
1987-1991 14 (58) 14 (58) -
1992-1996 10 (42) 10 (42) >0.99
Weight – median kg (IQR) 5.4 (4.0-9.5) 7.7 (5.6-9.5) 0.54
BSA (m2) – median (IQR) 0.30 (0.25-0.46) 0.39 (0.30-0.47) 0.55
Consanguinity 3 (13) 4 (17) >0.99
Presentation with CHF 18 (75) 19 (79) >0.99
Family history cardiomyopathy 8 (33) 7 (29) >0.99
Echocardiography
LVEDD Z-score – median (IQR) 4.8 (0.5-6.0) 4.0 (1.9-5.2) 0.86
LVESD Z-score – median (IQR) 7.2 (3.9-8.2) 6.1 (4.3-8.0) 0.92
FS Z-score – median (IQR) -11.2 (-12.1 to -9.9) -9.7 (-12.7 to -8.4) 0.48
BSA: body surface area; CHF: congestive heart failure; FS: fractional shortening;
LVEDD: left ventricular end-diastolic diameter; LVESD: left ventricular end-systolic
diameter; The c-statistic of the propensity-score model was 0.70.
22
1
2
3
4
5
FIGURE 1: Long-term freedom from death or transplantation of all 29 LVNC subjects
FIGURE 2: Cumulative incidence curve of competing outcomes after diagnosis of LVNC (n=29). Competing events: Normalization of LV function (blue), and death or transplantation before normalization (black). The probability of remaining alive, transplant-free and without normalization is also shown (red).
23
1
2
34567
89
10
1112
FIGURE 3: Comparison of probability of freedom from death/ transplantation at 15 years between subjects with LVNC-D (n=27) and DCM (n=175): A) Unadjusted (full sample); B) Adjusted using propensity-score matching (matched pairs); C) Adjusted using inverse probability of treatment weighted (full sample).
24
12345
67
REFERENCES
1. Jefferies JL., Wilkinson JD., Sleeper LA, Colan SD, Lu M, Pahl E, et al. Cardiomyopathy
Phenotypes and Outcomes for Children With Left Ventricular Myocardial Noncompaction:
Results From the Pediatric Cardiomyopathy Registry. J Card Fail. 2015;21(11):877-84.
2. Freedom RM., Yoo SJ, Perrin D, Taylor G, Petersen S, Anderson RH. The morphological
spectrum of ventricular noncompaction. Cardiol Young. 2005;15(4):345-64.
3. Lipshultz SE, Cochran TR, Briston DA., Brown SR, Sambatakos PJ, Miller TL., et al.
Pediatric cardiomyopathies: causes, epidemiology, clinical course, preventive strategies and
therapies. Future Cardiol. 2013;9(6):817-48.
4. Jenni R., Oechslin E. N., van der Loo B. Isolated ventricular non-compaction of the
myocardium in adults. Heart. 2007;93(1):11-5.
5. Maron BJ., Towbin JA., Thiene G, Antzelevitch C, Corrado D, Arnett D, et al.
Contemporary definitions and classification of the cardiomyopathies: an American Heart
Association Scientific Statement from the Council on Clinical Cardiology, Heart Failure and
Transplantation Committee; Quality of Care and Outcomes Research and Functional Genomics
and Translational Biology Interdisciplinary Working Groups; and Council on Epidemiology and
Prevention. Circulation. 2006;113(14):1807-16.
6. Finsterer J, Stöllberger C, Wegmann R, Janssend LA. Acquired left ventricular
hypertrabeculation/noncompaction in myotonic dystrophy type 1. Int J Cardiol. 2009;137(3):310-
13.
7. Gati S., Papadakis M., Papamichael N. D., Zaidi A., Sheikh N., Reed M., et al. Reversible
de novo left ventricular trabeculations in pregnant women: implications for the diagnosis of left
ventricular noncompaction in low-risk populations. Circulation. 2014;130(6):475-83.
8. Elliott P, Andersson B, Arbustini E, Bilinska Z, Cecchi F, Charron P, et al. Classification
of the cardiomyopathies: a position statement from the European Society Of Cardiology Working
Group on Myocardial and Pericardial Diseases. Eur Heart J. 2008;29:270-6.
9. Towbin JA. Left ventricular noncompaction: a new form of heart failure. Heart Fail Clin.
2010;6(4):453-69, viii.
25
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
10. Towbin JA, Lorts A, Jefferies JL. Left ventricular non-compaction cardiomyopathy. The
Lancet. 2015;386(9995):813-25.
11. Brescia ST., Rossano JW., Pignatelli R, Jefferies JL, Price JF, Decker JA., et al. Mortality
and sudden death in pediatric left ventricular noncompaction in a tertiary referral center.
Circulation. 2013;127(22):2202-8.
12. Nugent AW, Daubeney PE., Chondros P., Carlin JB., Cheung M., Wilkinson LC., et al.
The epidemiology of childhood cardiomyopathy in Australia. N Engl J Med. 2003;348(17):1639-
46.
13. Alexander PM., Daubeney PE., Nugent AW., Lee KJ, Turner C, Colan SD, et al. Long-
term outcomes of dilated cardiomyopathy diagnosed during childhood: results from a national
population-based study of childhood cardiomyopathy. Circulation. 2013;128(18):2039-46.
14. Bharucha T, Lee KJ., Daubeney PE., Nugent AW, Turner C, Sholler GF, et al. Sudden
death in childhood cardiomyopathy: results from a long-term national population-based study. J
Am Coll Cardiol. 2015;65(21):2302-10.
15. Nugent AW, Daubeney PE, Chondros P, Carlin JB., Colan SD., Cheung M., et al.
Clinical features and outcomes of childhood hypertrophic cardiomyopathy: results from a
national population-based study. Circulation. 2005;112(9):1332-8.
16. Daubeney PE, Nugent AW, Chondros P., Carlin JB., Colan SD, Cheung M., et al.
Clinical features and outcomes of childhood dilated cardiomyopathy: results from a national
population-based study. Circulation. 2006;114(24):2671-8.
17. Jenni R, Oechslin E, Schneider J, Attenhofer JC., Kaufmann PA. Echocardiographic and
pathoanatomical characteristics of isolated left ventricular non-compaction: a step towards
classification as a distinct cardiomyopathy. Heart. 2001;86(6):666-71.
18. Sluysmans T., Colan S. D. Theoretical and empirical derivation of cardiovascular
allometric relationships in children. J Appl Physiol. 2005;99(2):445-57.
19. Colan SD, Parness IA., Spevak PJ., Sanders SP. Developmental modulation of
myocardial mechanics: age- and growth-related alterations in afterload and contractility. J Am
Coll Cardiol. 1992;19(3):619-29.
26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
20. Estimated resident population by sex and age, states and territories of Australia, June
1987 to June 1992. Australian Bureau of Statistics. Canberra: 1993 (catalog no. 3201.0).
21. Population by age and sex, Australian states and territories, June 1992 to June 1997.
Australian Bureau of Statistics. Canberra: 1997 (catalog no. 3201.0).
22. Austin PC. Propensity-score matching in the cardiovascular surgery literature from 2004
to 2006: a systematic review and suggestions for improvement. J Thorac Cardiovasc Surg.
2007;134:1128-35.
23. Austin PC, Stuart EA. Moving towards best practice when using inverse probability of
treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in
observational studies. Stat Med. 2015;34:3661-79.
24. Pignatelli RH, McMahon CJ, Dreyer WJ, Denfield SW, Price J, Belmont JW, et al.
Clinical characterization of left ventricular noncompaction in children: a relatively common form
of cardiomyopathy. Circulation. 2003;108(21):2672-8.
25. Andrews RE., Fenton MJ., Ridout DA., Burch M. New-onset heart failure due to heart
muscle disease in childhood: a prospective study in the United kingdom and Ireland. Circulation.
2008;117(1):79-84.
26. Ichida F, Hamamichi Y, Miyawaki T, Ono Y, Kamiya T, Akagi T, et al. Clinical features
of isolated noncompaction of the ventricular myocardium: long-term clinical course,
hemodynamic properties, and genetic background. J Am Coll Cardiol. 1999;34(1):233-40.
27. Wald R, Veldtman G., Golding F., Kirsh J, McCrindle BW, Benson L. Determinants of
outcome in isolated ventricular noncompaction in childhood. Am J Cardiol. 2004;94(12):1581-4.
28. McMahon CJ., Pignatelli R. H., Nagueh S. F., Lee V. V., Vaughn W., Valdes S. O., et al.
Left ventricular non-compaction cardiomyopathy in children: characterisation of clinical status
using tissue Doppler-derived indices of left ventricular diastolic relaxation. Heart.
2007;93(6):676-81.
29. Saleeb SF, Margossian R, Spencer CT, Alexander ME, Smoot LB, Dorfman AL, et al.
Reproducibility of echocardiographic diagnosis of left ventricular noncompaction. J Am Soc
Echocardiogr. 2012;25:194-202.
27
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
30. Amzulescu MS, Rousseau MF, Ahn SA, Boileau L, de Meester de Ravenstein C,
Vancraeynest D, et al. Prognostic Impact of Hypertrabeculation and Noncompaction Phenotype in
Dilated Cardiomyopathy: A CMR Study. JACC Cardiovasc Imaging 2015;8:934-46.
31. Singh RK, Canter CE, Shi L, Colan SD, Dodd DA, Everitt MD, et al. Survival Without
Cardiac Transplantation Among Children With Dilated Cardiomyopathy. J Am Coll Cardiol
2017;70:2663-73.
28
1
2
3
4
5
6
7
8
APPENDIX
The National Australian Childhood Cardiomyopathy Study comprises the following
medical institutions and physicians:
- The Princess Margaret Hospital for Children, Perth:
o Dr L D’Orsogna, Dr A Bullock, Dr J Ramsey, Dr D Kothari
- The Women’s and Children’s Hospital Adelaide:
o Dr T Robinson, Dr M Richardson, Dr G Wheaton
- The Mater Children’s Hospital, Brisbane:
o Dr D Radford, Dr C Whight, Dr R Justo, Dr C Ward
- The Children’s Hospital at Westmead, Sydney:
o Dr G Sholler, Dr R Hawker, Dr S Cooper, Dr K Lau, Dr M Sherwood
- The Sydney Children’s Hospital:
o Dr O Jones,
- The John Hunter Children’s Hospital, Newcastle:
o Dr G Warner
- The Royal Children’s Hospital, Melbourne:
o Professor J L Wilkinson, Dr T Goh, Dr B Edis, Professor D Penny, Dr G
Lane
- The Monash Medical Centre, Melbourne:
o Professor S Menahem, Dr L Fong
- Other participating physicians include Dr C Semsarian, Dr A Gailbraith, Dr R
Jeremy, Dr R Fryda, Dr P Robinson and Dr L Lee
29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24