53 / 64
AFRICA
S51
CVJAFRICA • Volume 26, No 2, H3Africa Supplement, March/April 2015
populations,
10
occurred at a low frequency in both Cameroonian
14
and Tanzanian patients.
13
Nevertheless, in the
HMIP-2
sub-locus,
there was a much higher MAF of rs9389269 in Cameroonian
(0.18)
14
compared to the Tanzanian SCD patients (0.03).
13
This
observation could indicate a high degree of variation in the
MAF of this SNP among SCD patients in African population
groups.
16
Furthermore, studies in Cameroon and Tanzania lacked
power to replicate the association of a sub-locus (rs7482144) in
HBG2
(Table 1), which explained 2.2% of the variation in HbF
levels in African American patients.
8
This is likely to be due
to the absence of Senegal and Indian–Arab beta-globin locus
haplotypes that contain the rs7482144 in most Cameroonian
patients.
17
Similarly, a strong signal adjacent to the
HBB
cluster
recently detected in African-American patients at rs5006884
in
OR51B5/6
18
was not found to have significant association
in either Tanzanian
14
or Cameroonian SCD patients.
13
These
findings suggest that studies of multiple SCD populations in
Africa are warranted to improve our understanding of the
impact of human diversity on HbF expression in SCD.
19
The co-inheritance of alpha-thalassaemia and SCD
The co-inheritance of
α
-thalassaemia is associated with a milder
phenotype in patients with HbSS and S
β
0
thalassaemia, e.g.
higher haemoglobin level and lower stroke rate.
20
However, the
effect of
α
-thalassaemia is not all positive; pain and aseptic
necrosis may be higher.
21
In Cameroon, the co-inheritance of
α
-thalassaemia and
SCD was associated with late onset of clinical manifestations
and potentially increased survival in Cameroonian patients;
this could explain the much higher allele frequency of 3.7kb
α
-globin gene deletion among SCD patients than in controls.
22,23
In Tanzania, the co-inheritance of
α
-thalassaemia and SCD was
associated with a lower stroke risk.
24
These preliminary data indicate an urgent need to replicate
and expand genetic studies in many other African SCD
populations, including studies focused on loci that are linked to
stroke
25
and other cardiovascular conditions, to fully measure
the opportunities of their implementation to improve the care of
patients with SCD.
Addressing the burden of cardiovascular diseases in
SCD in Africa
Cardiovascular phenotypes in SCD include complications
involving the heart (e.g. heart failure), brain (e.g. stroke), lung
(e.g. pulmonary hypertension) and kidney (e.g. proteinuria).
Cerebrovascular disease is perhaps the most devastating
complication for children with SCD, including overt stroke,
transient ischaemic attacks, silent infarcts and neurocognitive
dysfunction. Longitudinal cohort data from the USA have
shown that between five and 10% of patients with SCD will
experience a clinically overt stroke during childhood.
26
The
prevalence of overt stroke in SCD in Africa may be higher than
that reported in high-income countries.
Overt stroke is a clinical diagnosis and should be easily
detected in any cohort of closely monitored SCD patients
.
Brain computerised tomography (CT) and magnetic resonance
imaging (MRI) are used to rule out haemorrhage or localise the
tissue/vascular pathological basis for the stroke event. Clinical
examination and CT scans identified a stroke prevalence of 6.7%
in Cameroon.
27,28
A study of children with SCD in Nigeria found
a stroke prevalence of 8.7%.
29
The prevalence of silent cerebral infarcts (SCI) and cerebral
vasculopathies has been shown to be even greater than overt
stroke risk: SCI occurs in 27% of this population before their
sixth, and 37% by their 14th birthdays.
30
SCI is diagnosed by
MRI, but has not been studied in Africa because of the limited
availability of MRI equipment. In fact SCI is not really silent,
as falling school performance and other signs of neurocognitive
dysfunction and change in personality/behaviour may all raise
suspicion for increased risk of overt stroke, and suspicion of
stroke with absence of motor or speech defect. SCI could be
better called covert cerebral infarction.
The lack of longitudinally monitored SCD cohorts in Africa
weakens incidence and prevalence estimates. Indeed, the cognitive
Table 1. Foetal haemoglobin association results for SNPs at the
BCL11A
,
HBS1L-MYB
and beta-globin loci in the
Cameroonian and Tanzanian sickle cell anaemia cohort
Locus
Genomic variations
HbSS Cameroon
(n
=
596
)
14
HbSS Tanzania
(n
=
1 124
)
13
SNP
Position on the
chromosome*
Allele
change
MAF Effect size
p-
value
MAF Effect size
p
-value
Chromosome 2
BCL11A
rs11886868 60720246
T
>
C
0.31
0.167
0.0129
0.26
–0.406 3.00E-30
BCL11A
rs4671393 60720951
G
>
A 0.3
0.201
0.0062
0.3
–0.412 3.90E-28
Chromosome 6
HBS1L-MYB
rs28384513 135376209
A
>
C
0.2
–0.3002
0.0002
0.21
–0.146 1.90E-04
HBS1L-MYB
rs9376090 135411228
T
>
C
0
NA
NA
0.01
0.471 1.60E-02
HBS1L-MYB
rs9399137 135419018
T
>
C
0.04
0.412
0.0086
0.01
0.668 8.30E-06
HBS1L-MYB
rs9389269 135427159
T
>
C
0.18
0.09561 0.2468
0.03
0.4
1.40E-05
HBS1L-MYB
rs9402686 135427817
G
>
A 0.03
0.1447
0.4437
0.06
0.342 1.60E-04
HBS1L-MYB
rs9494142 135431640
T
>
C
0.11
0.3391
0.0023
0.13
0.085 6.00E-02
Chromosome 11
HBG2
rs7482144
5276169
G
>
A 0
–0.05843
0.9076
0.01
0.562 1.60E-04
OR51B5/6
rs5006884
5373251
C
>
T
0.08
0.04163
0.7385
0.05
0.164 2.40E-02
NA, not applicable; monomorphic T for the entire sample; MAF, minor allele frequency; SNP, single-nucleotide polymorphisms.
*Chromosome, position on NCBI Build 36.1.