53 / 92
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 27, No 4, July/August 2016
AFRICA
255
any of the study groups, although ST-segment elevation was
seen in leads V2–V3 in two of the HIV-positive HAART-
naïve patients. Early repolarisation abnormality, commonly
seen in blacks,
24,25
may be an explanation for this. In a similar
study involving 4 831 HIV-positive adults, including those
with hypertension and diabetes mellitus, T-wave inversion was
observed in 11.1% of participants and it was substantially
more common in women than men. Shikuma
et al
. found ECG
evidence of asymptomatic IHD in 10.9% of participants with
no known IHD.
23
Although the exact reason is not known, a
higher occurrence of myocarditis, coronary artery vasculitis and
ischaemia,
26
which cause cardiac abnormalities in HIV-positive
patients, may explain this higher prevalence.
Sinus tachycardia was seen more often in the HIV-positive
groups (8.6% in HIV-positive patients on HAART, 19.2% in
HIV-positive HAART-naïve patients) compared to 0% in the
controls. This may have been due to inter-current febrile illness,
anaemia, myocarditis and increased metabolic demand in HIV
patients.
27
These conditions induce autonomic dysfunction and
increased sympathetic activity. The higher prevalence of sinus
tachycardia in HIV-positive HAART-naïve patients may have
been due to their higher immunosuppression and the prevalence
of these identified factors. Sinus bradycardia, noted in 2% of the
controls and 0% of the HIV-positive patients, may have been due
to increased vagal tone, commonly seen in healthy young people.
28
First-degree heart block was found in both patients and
controls, although slightly more often in patients than in the
control group. Although currently out of phase, febrile illness
such as leptospirosis, which occurs commonly in HIV-positive
patients, has been reported to cause complete heart block.
29
The use of HAART, especially those containing PIs, in HIV
infection is known to cause prolonged QRS duration, first-
degree atrioventricular block and complete bundle branch
block. The conduction abnormalities seen in this study were not
unexpected since most of the patients had recurrent fever, and
were often not evaluated for leptospirosis. Moreover, some of
the HIV-positive patients on HAART in this study were on PIs.
Charbit
et al
.
30
reported a higher number of patients showing
prolonged QRS duration, first-degree atrioventricular block and
complete bundle branch block in those taking PIs.
Left ventricular hypertrophy (LVH) assessed by Sokolow and
Lyon criteria, as well as Araoye’s voltage criteria,
31
was found in
eight (11%) of the HIV-positive HAART-naïve patients but in
none of the HIV-positive patients on HAART or the controls.
Four HIV-positive patients on HAART, four HIV-positive
HAART-naïve patients and one control subject had LVH,
derived with Devereux’s formula.
We found left ventricular mass was increased in HIV infection
(Table 5). In similar studies to ours, Barbaro
et al
.
32
and Lipshultz
et al
.
33
reported increased LV mass in HIV-positive patients. On
the other hand, Martinez-Garcia
et al
.
34
found decreased LV
mass in asymptomatic HIV-infected patients, and Samaan
et al
.
35
found decreased LV mass among patients with AIDS wasting
syndrome.
The mechanisms by which these adverse effects on LV mass
occur in HIV-positive patients are not fully understood, but
are thought to be related to mitochondrial toxicity.
36
Many
studies have shown that HIV virions directly affect myocardial
cells and are associated with local release of cytokines and
other factors leading to inflammation, myocarditis and dilated
cardiomyopathy.
37
Also, increase or decrease in LV mass has been
suggested to be associated with opportunistic infections and
malnutrition,
35
therefore a lower nadir CD4 cell count has been
associated with higher LV mass index. Meng
et al.
38
reported
greater interventricular septal and posterior wall thicknesses
among patients exposed to PIs compared to those who were not
exposed.
The finding of higher numbers of HIV-positive patients with
LVH in our study was not surprising, given the pathogenesis
and sequela of HIV infection as well as the use of HAART.
Pewsner
et al.
and Devereux found LVH, assessed by ECG,
only in HIV-positive HAART-naïve patients, probably because
of the high specificity and low sensitivity of the ECG.
39,40
On
the other hand, Michael found four patients with LVH, using
echocardiography, in both HIV-positive patients on HAART
and the HIV-positive HAART-naïve group, and one patient in
the control group. This may have been because echocardiography
is much more sensitive than ECG.
41
QTc interval, corrected for heart rate using Bazett’s formula,
was more prolonged in our HIV-positive patients on HAART
and HIV-positive HAART-naïve patients than in the controls
(Table 4). The prevalence of prolonged QTc was 34.6% in
HIV-positive patients and 10.5% in the controls. A breakdown
of this showed a higher prevalence of 18.2% in HIV-positive
patients on HAART, compared to 16.4% in the HIV-positive
HAART-naïve group. This is similar to the 45% reported by
Okoye
22
and 34.7% reported by Ogunmodede.
42
Villa
et al
.
43
reported a high prevalence of prolonged QTc
interval of 65% in a highly selected cohort of HIV-positive
patients who had already developed autonomic dysfunction. On
the other hand, Kocheril
et al
.
44
reported a prevalence of 29%
in 42 AIDS patients in the absence of any known cause. QTc
prolongation in HIV-positive patients has been attributed to
electrolyte imbalance from poor nutrient intake and recurrent
diarrhoea, drugs including zidovudine,
45
protease inhibitors,
46
pentamidine,
46
halofantrin,
47
trimethoprim-sulfamethoxazole,
48
and autonomic dysfunction due to HIV-associated neuropathy.
43
Okoye
14
documented hypocalcaemia as the cause of QTc
prolongation in 35% of AIDS patients. In our study, the
higher QTc prolongation in the patients may have been due
to electrolyte imbalance from vomiting and diarrhoea, PIs,
Table 5. Comparison of echocardiographic parameters
measured across the groups using one-way ANOVA
Parameters
HIV-positive
on HAART
HIV-positive
HAART-naïve
Control
F
-value
p
-value
AO (cm)
2.71
±
0.40* 2.41
±
0.37 2.74
±
0.42* 21.363
<
0.001
LA (cm)
3.27
±
0.62
2.68
±
0.51 3.11
±
0.47 31.385
<
0.001
EDD (cm)
4.73
±
0.70* 4.41
±
0.55 4.75
±
0.42* 11.240
<
0.001
ESD (cm)
3.01
±
0.51
2.84
±
0.57 2.92
±
0.43 2.616 0.075
IVS (cm)
0.77
±
0.17* 0.85
±
0.17 0.78
±
0.15* 6.098 0.003
PW (cm)
0.82
±
0.16
0.87
±
0.17 0.82
±
0.13 2.878 0.058
EF (%)
68.95
±
12.43* 72.81
±
11.70 67.36
±
9.04* 6.223 0.002
FS (%)
36.77
±
9.81 36.51
±
8.64 37.77
±
6.53 0.623 0.537
LVM (g)
141.94
±
49.75 138.61
±
48.53 131.26
±
31.55 1.540 0.216
LVMI (g/m
2
) 79.95
±
26.25 77.55
±
25.91 72.37
±
16.52 2.760 0.065
*Duncan
post hoc
multiple comparison test indicating means for groups in
homogenous subsets (means not significantly different).
AO
=
aorta; LA
=
left atrium; EDD
=
end-diastolic diameter; ESD
=
end-systolic
diameter; IVS
=
interventricular septum; PW
=
posterior wall of left ventricle;
EF
=
ejection fraction; FS
=
fractional shortening; LVM
=
left ventricular mass;
LVMI
=
left ventricular mass index.