This study demonstrated that in young children and adolescents with dilated cardiomyopathy
(1) The detection of histologically proven focal myocardial fibrosis by LGE is uncommon
(found only in 16%) although LV dilatation was marked and LV function severely depressed,
(2) The majority of LGE positive patients had chronic myocarditis suggesting a relation
between histological signs of inflammation and LGE,
(3) Various patterns of LGE were found with transmural type of LGE being detected
in two of the 5 patients, which is not typical for non-ischemic forms of DCM and might
reflect embolic myocardial infarction.
The findings of our study are in contrast to previously reported CMR findings in adult
patients with non-ischemic DCM where LGE was detected more frequently (in up to 50%
of patients) and was associated with larger left ventricles, a higher degree of impaired
function, elevated biomarkers and arrhythmias resulting in worse overall outcome 9]–12], 20]. Accordingly, the presence of focal LGE in non-ischemic DCM in adults is considered
to be associated with contractile impairment that is thought to reflect the transition
from compensated to decompensated state representing the endpoint of adverse remodelling
with irreversible ventricular failure 9]. As the severity of LV dilatation and functional impairment in our study group was
comparable to that reported in series of adult patients, the low rate of macroscopic
fibrosis detected by LGE in our study is unexpected and the underlying reason for
this finding entirely unknown.
In general, limited data exists about the pathogenesis of myocardial fibrosis in neonates,
children and young adolescents with all types of cardiomyopathies including its assessment
by LGE 16], 21]–23]. Grosse-Wortmann and colleagues 24] reported 11 children and adolescents with DCM of various etiologies but found myocardial
fibrosis by LGE only in one patient. Although the severity of LV dilatation and dysfunction
was more pronounced in our study group, the prevalence of myocardial fibrosis by LGE
did not increase. Furthermore, as reported by this group, RV dimensions and function
differed widely among the patients illustrating that childhood DCM can present as
a biventricular disease although no relationship between RV and LV dimensions was
detected. However, the severity of LV dilatation negatively affected RV function probably
due to adverse ventricular-ventricular interactions while BNP levels were solely related
to LV dimensions.
Specific patterns of LGE have been reported in patients with ischemic and non-ischemic
DCM linking the type of LGE to a certain form of cardiomyopathy. In non-ischemic DCM
mid-wall striae of LGE is known to be the typical finding, although various patterns
such as focal patchy, subepicardial and diffuse LGE have also been described 9]–12], 25], 26]. Interestingly, the only patient in our study with mid-wall LGE had idiopathic DCM
and was the oldest patient in the study cohort (17Â years of age) while a midwall focal
patchy pattern of LGE and RV insertion site fibrosis were seen in one patient with
lymphocytic myocarditis, respectively. It has been described that myocarditic infiltrations
can persist after the acute phase of myocarditis, but tend to regress over time during
healing 17], 27]–29]. Conversely, a transmural pattern of fibrosis is not typically for myocarditis or
idiopathic DCM, rather than for ischemic heart disease, but was found in two patients
with myocarditis. These scars could be related to an embolic event that subsequently
caused myocardial infarction or might be associated with interstitial expansion due
to inflammation and/or edema as these patients were diagnosed for inflammatory cardiomyopathy.
However, discrimination of ischemic from non-ischemic lesions is crucial in these
patients and thus might influence therapeutic strategy and prognosis.
The underlying mechanisms for the development of fibrotic changes in non-ischemic
cardiomyopathies are related to different processes. Myocardial injury can be the
result of different pathogens and toxins causing inflammation, individual genetic
susceptibility, abnormal modulation of the immune system, permanent adrenergic activation
and metabolic dysregulation 30]–32]. Furthermore, increased wall stress due to LV enlargement can lead to microvascular
ischemia with subsequent myocyte necrosis which might explain why troponin levels
were found to be related to LV dimension and function in our population. The resulting
focal areas of replacement fibrosis are thought to be the basis for the detection
of LGE by CMR although LGE can also be the result of other forms of interstitial expansion.
Although good agreement between areas of LGE and the pathological location of macroscopic
fibrosis on autopsy has been reported 10], 11], the study by Schalla and colleagues showed that the presence of focal LGE in adult
DCM patients was more related to microscopic findings of inflammation rather than
fibrosis in endomyocardial biopsy 33]. This might explain why the majority (4 of the 5 patients) of our LGE positive patients
were found to have lymphocytic myocarditis. Otherwise, the low detection rate of LGE
in our population does not imply that myocardial fibrosis is absent in childhood DCM.
Indeed, only a few histopathological studies exist about childhood DCM, but those
also revealed fibrosis as a cardinal feature 34], 35]. This is in accordance with our findings, where analysis of endomyocardial biopsies
revealed a substantial degree of fibrosis in endomyocardial biopsies, both in patients
with and without LGE on CMR.
This suggests that the possible pathogenic factors and mechanisms leading to the phenotype
of LV dilatation and dysfunction considerably vary in younger patients. Although speculative,
the low incidence of myocardial fibrosis assessed by CMR might be related to differences
in myocardial remodelling in DCM between children and adults. Nishikawa et al. analysed
endomyocardial biopsies and found differences in the distribution of microscopic patterns
of fibrosis, revealing a higher myocarditic index and a higher incidence of bizarre
myocardial hypertrophy as well as discrete ultrastructural differences between adults
and children with DCM 34], 36] suggesting that factors and mechanisms causing myocardial damage may differ between
children and adults. Furthermore, differences in myocardial matrix structure and function
37] as well as in adrenergic receptor stimulation 38] have been described, which might have an impact on response to medical therapy thereby
influencing ventricular remodelling and prognosis.
Decision making whether cardiac transplantation is needed in the individual child
with DCM is crucial but is hampered by a substantial rate of spontaneous recovery
of ventricular dimension and function that has been described both in adult and childhood
DCM 3], 5], 6], 39], 40]. A recent multicentre study by Everitt et al. demonstrated a 2-year cumulative incidence
of recovery of normal LV size and function in 22% of patients 5]. One explanation for this phenomenon might be related to our findings of a less remodelled
myocardium which inherits a higher chance for recovery. These results can be supported
by the study of Masci et al. who found in adult patients with idiopathic DCM, that
the absence of LGE was a strong independent predictor of LV reverse remodelling at
2-years follow-up independent of LV size and function 13]. Nabeta and colleagues recently demonstrated that the estimation of fibrosis by baseline
CMR-LGE is superior in predicting reverse remodelling and outcome than the estimation
of fibrosis by baseline endomyocardial biopsy 41]. Due to the high number of patients in our study cohort who underwent pulmonary artery
banding or cardiac transplantation, procedures that affect the natural history of
this disease, only a minority of patients received exclusively medical heart failure
therapy not allowing any further statistical analysis of the predictive value of LGE
in children with DCM. However, detection of LGE seemed not to have an impact on the
occurrence of reverse remodelling or outcome which demonstrates that CMR LGE may not
be able to separate those patients with a high chance for recovery from those who
will be affected by an aggravation in ventricular function with a subsequent higher
risk for death and/or cardiac transplant. Whether myocardial fibrosis in childhood
DCM is a factor for evaluating innovative treatment options also taking advantage
of the suggested preserved cardiac regenerative capacity of the myocardium of young
children will require further investigations 42]–44].
Study limitations
The small study population and the retrospective design of the study were accompanied
by a selection bias towards patients presenting in an advanced stage of the disease
who were referred for cardiac transplantation might not represent the entire population
of children and adolescents with DCM. Therefore, the reported incidence of LGE as
well as the observed rate of spontaneous recovery and outcome might differ in reality.
Compared to previous studies, we observed a higher incidence of patients with myocarditis
which must be considered when interpreting the data. The reason for this might be
related to the fact that other causes for DCM, especially hereditary forms, were excluded
in order to analyse a homogenous patients cohort. While the underlying genetic causes
for childhood DCM are increasingly recognized 45], routine genetic testing for mutations known to cause DCM is not performed at our
institution which may have resulted in a certain selection bias.
Identification of LGE in very young patients may be limited by respiratory or motion
artifacts related to spontaneous breathing under sedation. Furthermore, spatial resolution
of the CMR sequence might be too low to detect small areas of local fibrosis in neonates
and infants thereby underestimating its real prevalence. However, in a previous study
in neonates and infants with ischemic DCM 21], LGE could be reliably assessed. Because no additional method besides LGE was used
to confirm histologically proven areas of myocardial fibrosis thereby potentially
missing important signs of adverse ventricular remodelling. Furthermore, the used
LGE technique cannot detect perivascular and interstitial myocardial fibrosis that
is frequently observed in non-ischemic DCM 46]. Further developments of T1 mapping sequences that enable reliable quantification
of the extracellular space in younger patients may potentially provide further and
earlier insights into adverse cardiac remodelling in pediatric heart failure.