CARDIOVASCULAR JOURNAL OF AFRICA • Vol 24, No 6, July 2013
AFRICA
219
patients with HIV-associated cardiomyopathy compared to those
with idiopathic dilated cardiomyopathy and heart transplant
recipients in Cape Town.
Methods
This was a case-comparison study in which the frequency of
myocarditis and cardiotropic viral infection in patients with
HIV-associated cardiomyopathy was compared with those with
idiopathic dilated cardiomyopathy without HIV infection (i.e.
negative controls for acquired immunodeficiency state) and
heart transplant recipients (i.e. positive controls for an acquired
immunodeficiency state). The endomyocardial biopsy study was
conducted in the cardiac catheterisation laboratory at Groote
Schuur Hospital in Cape Town from April 2002 to December
2007, before anti-retroviral therapy became widely available in
South Africa.
A minimum of four right ventricular endomyocardial biopsy
specimens were obtained for all patients through the right
internal jugular vein. The endomyocardial biopsy samples were
immediately placed in formalin (for light microscopy) and snap
frozen in liquid nitrogen (for the virological studies).
The study was approved by the Research Ethics Committee of
the Faculty of Health Sciences of the University of Cape Town,
and all participants gave written informed consent.
We recruited consecutive cases of HIV-positive and
HIV-negative patients with a new-onset echocardiographically
confirmed diagnosis of dilated cardiomyopathy (i.e. a left
ventricular ejection fraction
≤
45% and/or fractional shortening
<
25%, plus a left ventricular end-diastolic diameter
>
117% of
the predicted value for age and body surface area by applying
Henry’s formula)
10
or isolated left ventricular systolic dysfunction
(i.e. a left ventricular ejection fraction
≤
45% and/or fractional
shortening
<
25% plus normal left ventricular dimensions).
Patients with secondary causes of cardiomyopathy or left
ventricular systolic dysfunction (such as hypertension, valvular
heart disease, diabetes, coronary artery disease, or prior AIDS
defining opportunistic infection or malignancy) were excluded
from the study. Coronary artery disease was excluded by
coronary angiography in all cases with cardiomyopathy who
participated in this study.
The HIV-positive patients with unexplained cardiomyopathy
were categorised as HIV-associated cardiomyopathy, while
the HIV-negative cases were labelled as idiopathic dilated
cardiomyopathy. Heart transplant recipients who were undergoing
routine endomyocardial biopsy to monitor rejection were also
enrolled.
The diagnosis of myocarditis was based on the
immunohistological criteria of the World Heart Federation.
11,12
These criteria define an abnormal lymphocytic infiltrate as the
presence of
>
14 CD
3
+
lymphocytes per high-power field (and
up to four CD
68
+
macrophages included in this total amount)
(magnification,
×
40) with use of a microscope with wide-field,
10
×
eyepieces (diameter 550 m; area 2.4
×
105 m; Olympus
BH-2, Hamburg, Germany).
Acute myocarditis was defined as a significant lymphocytic
infiltrate with myocytolysis (i.e. necrosis or degeneration) on
light microscopy, while chronic myocarditis was defined as
a significant lymphocytic infiltration with no myocytolysis.
Patients with no infiltrating cells or insignificant lymphocyte
infiltrate (
<
14 leukocytes/mm
2
) and no myocytolysis were
classified as having no evidence of myocarditis.
Immunohistochemistry was used to define the types of
immune cells in the inflammatory infiltrate of myocarditis.
Formalin-fixed, paraffin-embedded tissues were sectioned
at 5
µ
m for routine, single-label immunohistochemistry. Cell
populations were characterised using antibodies specific for CD
3
(rabbit polyclonal, A 0452, DakoCytomation, Carpinteria, CA),
CD
4
(clone 1F6, VP-C318, Vector Laboratories, Burlingame,
CA), CD
8
(clone 1A5, VP-C325, Vector) and CD
68
(clone KP1,
M 0814, DakoCytomation).
Two methods were used for the detection of cardiotropic
viruses: viral antigen detection using immunohistochemical
methods and identification of viral genome by PCR or reverse
transcription (RT)-PCR. To evaluate the extent to which
cardiotropic viral infections may contribute to induction of the
myocardial inflammatory response in the heart tissue under
study, specific virological immunoperoxidase screening was
conducted for seven cardiotropic viruses using the following
virus-specific antigens: cytomegalovirus (Dako, Carpenteria,
California, M0854), adenovirus (Novocastra, NCL-ADENO),
herpes simplex virus types 1 and 2 (Dako, Carpenteria, California,
B0114 and B0116), HIV-1 p24 antigen (Dako, Carpenteria,
California, M0857), Epstein-Barr virus (Dako, Carpenteria,
California, M0897), enterovirus (Dako Carpenteria, California,
M7064), and parvovirus B19 (Dako Carpenteria, California,
B0091), along with reviews of haematoxylin and eosin (H&E)-
stained sections and special staining for histological evidence
of infective agents, such as
Toxoplasma gondii
,
Cryptococcus
neoformans
and mycobacteria.
Sections were deparaffinised in xylene and rehydrated
through graded ethanols, followed by blocking of endogenous
peroxidase by incubation in 3% hydrogen peroxide in phosphate-
buffered saline. Antigen retrieval in most cases consisted of
microwaving in citrate buffer. Antigen retrieval for CD
4
consisted
of microwaving in EDTA buffer (Lab Vision, Fremont, CA), for
adenovirus it consisted of five-minute digestion with proteinase
K (Dako, Carpenteria, California), and for CD
8
it consisted
of 20-min pressure cooker treatment in Trilogy solution (Cell
Marque, Hot Springs, AK). Sections were incubated with
primary antibody followed by an avidin–biotin block (Vector)
to block endogenous biotin, and sequential incubation with
biotinylated secondary antibody and horseradish peroxidase-
conjugated avidin (ABC Standard or ABC Elite, Vector), or
the EnVision polymer system (Dako, Carpenteria, California)
applied according to the manufacturer’s instructions.
Antigen–antibody complex formation was detected by use of
3,3
′
- diaminobenzidine (DAB) chromogen (Dako, Carpenteria,
California) and tissues were counterstained with Mayer’s
haematoxylin. Viral antigen was detected with viral antibodies
to cytomegalovirus, Epstein-Barr virus, herpes simplex virus,
parvovirus B19, adenovirus, enterovirus and HIV-1 (Dako,
Carpenteria, California; Novocastra, UK Laboratories) in the
three groups. Control lymph node tissue was used as positive
controls for each virus under study.
Statistical analysis
Groups were compared using the Mann-Whitney non-parametric
test, Kruskal-Wallis test,
c
2
test or Fisher’s exact test, as
