20 / 62
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 31, No 4, July/August 2020
182
AFRICA
Twenty-two (26.5%) of the subjects had a normal clinical
evaluation but an abnormal ECG. Within this group,
abnormal echocardiograms were found in five (22.7%). The
echocardiographic abnormalities found were three subjects
had DCMO, one had hypertensive heart disease with diastolic
dysfunction, and one had an incidental finding of a pericardial
effusion. The sensitivity and specificity of clinical and ECG
screening versus the defined gold standard of echocardiographic
diagnosis of cardiomyopathy (ILVNC/DCMO/HCM) are
depicted in Table 6.
Discussion
The main findings of this study are that family screening detected
the phenotype of dilated cardiomyopathy in 10.8% of subjects
with two-thirds of these individuals being asymptomatic. No
cases of ILVNC or HCM were detected. The second major
finding was that a screening strategy that utilised clinical
evaluation and an ECG was moderately sensitive in detecting
cardiomyopathy in comparison with cardiological screening,
which utilised echocardiography.
ILVNC is presumed to be a genetic disorder,
1
and
consequently, family screening has been advocated to detect
pathology in asymptomatic individuals. In this study, family
screening identified the phenotype of dilated cardiomyopathy in
nine (10.8%) previously undiagnosed individuals, of whom six
were asymptomatic, with three of these individuals belonging
to the same family. Despite some individuals having prominent
trabeculation, none of the individuals with the phenotype of
DCMO satisfied the criteria used in our study for the diagnosis
of ILVNC. Furthermore, no cases of HCM were identified.
Our findings differ from other family screening studies, which
found ILVNC in between 18 and 50% of subjects,
10-15
DCMO
in between 12 and 15%
16,17
and HCM in between 3 and 7% of
subjects.
16,17
These differences may be attributed to variations
in the diagnostic screening strategy employed, the population
studied, imaging techniques and criteria used, and referral bias
relating to this study.
The interplay between LVNC and DCMO is an important
consideration for the clinician. If the index case is presumed to
be ILVNC with or without a dilated cardiomyopathy phenotype,
family screening may reveal a DCMO phenotype without
ILVNC in screened relatives.
18-20
This most often arises in families
where there are sarcomeric gene mutations. The converse finding
of relatives with ILVNC phenotype discovered during family
screening where the index cases are DCMO has also been
described.
19
Hence the discovery of a
de novo
case of ILVNC
with either a dilated cardiomyopathy or a DCMO phenotype
may result in the discovery of diverse genotype–phenotype
manifestations when family screening is performed.
Several studies have highlighted the differences in detection
of affected family members based on the screening strategy
employed.
16,20
Echocardiographic screening has the advantage
of identifying the phenotype of ILVNC, DCMO or HCM in
individuals who are screened irrespective of whether they have
any cardiac symptoms. It has been suggested that up to 63% of
individuals with a phenotypic abnormality on routine screening
are asymptomatic. When only family history was used without
echocardiographic screening, 44% of individuals would not
have their phenotypic abnormality identified.
16
Furthermore,
since genetic abnormality has only been detected in 50% of
cardiologically screened confirmed cases of ILVNC,
16
it implies
that cardiological screening allows for more robust identification
of abnormality.
A major disadvantage not employing accompanying genetic
screening is that individuals with non-penetrance/reduced
penetrance may not be identified. Identifying individuals with
non-penetrance may require recurrent cardiac screening of
affected carriers, although the results of such a strategy have not
been adequately studied. Similarly, it is unknown whether repeat
cardiac screening is required in unaffected individuals from
families where the genetic abnormality is unknown.
This study comprised a cohort of adults over the age of 18
years. Several screening studies have included screening children
as well. A study in Australia on 314 children over a 10-year
period found ILVNC in 9.2%, HCM in 25.5% and DCMO in
58.6% of patients.
23
In a recent publication, which represents
the largest screening study conducted to date, van Waning
et
al
. found that mutations may be more common in children.
17
Therefore by excluding children, we may have underestimated
the prevalence of abnormality in our study.
A second issue relates to ethnicity since our cohort comprised
only individuals who were African. Ethnic differences may
result in various gene abnormalities and phenotypic expression
related to left ventricular remodelling. Therefore it may be that
African family members of individuals with either sporadic or
Table 4. Echocardiographic characteristics of
screened non-compaction relatives
Variable (
n
)
83
LVEDD (mm)
44.1 ± 5.5
LVESD (mm)
29.9 ± 5.4
Ejection fraction (%)
59.8 ± 6.2
End-diastolic volume (ml/m
2
)
88.4 ± 25.9
End-systolic volume (ml/m
2
)
36.3 ± 14.6
IVS (mm)
8.9 ± 2.0
Relative wall thickness (mm)
0.4 ± 0.1
Posterior wall thickness (mm)
8.5 ± 2.0
E wave (cm/s)
87.3 ± 22.8
A wave (cm/s)
66.2 ± 23.8
E/A (ratio)
1.6 ± 0.5
LVEDD: leeft ventricular end-diastolic diameter, LVESD: left ventricular end-
systolic diameter, IVS: interventricular septal diameter.
Table 6. Sensitivity and specificity of clinical and ECG screening
Sensitivity (%)
76
Specificity (%)
41.7
Positive predictive value (%)
88.5
Negative predictive value (%)
22.7
The likelihood ratio for a positive test
1.31
The likelihood ratio for a negative test
0.57
Table 5. Echocardiographic findings if a strategy of
clinical examination and ECG analysis were used
Subjects with normal clinical exam and a normal ECG,
n
(%)
61 (73.5)
Normal echo
54 (88.5)
Abnormal echo
7 (11.5)
Subjects with a normal clinical exam and an abnormal ECG,
n
(%) 22 (26.5)
Normal echo
17 (77.3)
Abnormal echo
5 (22.7)