Background
Malaria remains a major health problem, with approximately 3.2 billion people at risk. In 2015, WHO reported approximately 214 million cases and about 438,000 deaths occurring in the world with the highest morbidity and mortality rates observed in Africa, especially among children under 5 years of age [
1]. The control of the disease particularly in low transmission settings is the key target for malaria elimination [
2] but such success is still a great challenge [
3]. One approach to reduce the disease incidence is to block the transmission. Primaquine, a 8-aminoquinoline effective for both transmission blocking of
Plasmodium falciparum and anti-relapse treatment against
Plasmodium vivax has been recommended for many years by the World Health Organization (WHO) [
4,
5]. However, primaquine has a major drawback which limits its widely use. The drug is known to cause acute haemolytic anaemia in individuals with glucose-6-phosphate-dehydrogenase (G6PD) deficiency [
6,
7].
G6PD is a key enzyme catalysing the first reaction in the pentose phosphate pathway and provides a reduced form of nicotinamide adenine dinucleotide phosphate (NADPH) to help cells to counterbalance oxidative stress [
8]. G6PD deficiency is a common, X-linked hereditary enzyme deficiency affecting approximately 400 million people worldwide [
9], mainly in malaria-endemic regions [
10]. Among the 186 mutations identified until 2012 in the
G6PD gene [
11], the variants 376A (G6PD type B), 376G (G6PD deficiency type A
+ and 202A (severe G6PD deficiency type A
−) are the most common ones. The deficiency types A
+ (moderately deficient) and A
− (severely deficient) constitute up to 90 % of reported
G6PD variants [
8,
12,
13]. Other mutations such as A542T, G680T or T968C have also been identified in parts of Africa and been suggested to contribute to G6PD deficiency. However, information available on these mutations is very scarce. For example, the T968C mutation has been reported to be common only in The Gambia [
14] and Senegal [
15]. Although, G6PD diagnostic enzyme tests are available, they are currently not widely used in most clinical studies, particularly because primaquine, is not regularly used in Africa. Any attempts to control malaria need to take into account also tertian malaria, which, although occurring rarely only in many parts of Africa, contributes to the world-wide malaria burden.
Previous case–control studies have reported an association of G6PD deficiency, in particular of the 202A variant, with an increased risk of severe malarial anaemia and a protection or reduced risk against cerebral malaria [
16,
17]. Therefore, it is interesting to assess the influence of this allele on the clinical presentation of severe malaria among the African children recruited into this study.
This study was a sub-study of a multicentre trial of artesunate conducted by the “Severe Malaria in African Children” (SMAC) consortium which assessed the non-inferiority of a simplified 3-dose regimen of intramuscular and intravenous artesunate [
18]. The aim was to determine using direct sequencing the frequency distribution of G6PD deficiency in malaria children from Africa. Some exploratory analyses have also been performed to investigate the effect of
G6PD genotypes on asexual parasitaemia, haemoglobin concentrations and severe malarial anaemia during admission.
Methods
Study design and participants
This study was a sub-study of the SMAC follow-up study. The SMAC study was an open-label, randomized, multicentre (Gabon, Malawi, Ghana, Kenya, and The Gambia), parallel-group, three-arm study to compare the anti-malarial activity and safety of three artesunate (ARS) dosing regimens in children with severe P. falciparum malaria. Patients were randomly assigned to one of three dosing regimens consisting of a total of 12 mg/kg parenteral ARS: (i) 2.4 mg/kg intramuscular on admission and at 12, 24, 48 and 72 h, (ii) 4 mg/kg intramuscular on admission and at 24 and 48 h, and (iii) 4 mg/kg intravenous on admission and at 24 and 48 h post admission. Parasitaemia was assessed by thick blood smears at 6 h intervals and prior to the each dose of treatment for at least 48 h following the first dose of study drug. Malaria occurs holoendemically and transmission rates in all study countries are high and perennial.
The present study involved participants of the SMAC study from Gabon, Ghana and Kenya. Three hundred and one children aged 6–120 months with a diagnosis of
P. falciparum infection (parasitaemia ≥5000 parasites/µL on initial blood smear) made using an alternative to conventional thick film examination (Lambaréné method) [
19] were randomly selected from the SMAC follow-up study [
18]. Blood samples (400 μL) from all participants were collected in heparinized tubes. Specimens were stored at −80 °C for subsequent molecular analyses.
Ethics statement
The study was conducted in accordance with Good Clinical Practices, and approved by authorities for each study site (the Regional Ethics Committee in Lambaréné (CERIL) for Gabon, Committee on Human Research, Publication and Ethics, Kwame Nkrumah University of Science and Technology (KNUST), Kumasi, for Ghana and the National Ethics Research Committee, Kenya Medical Research Institute (KEMRI) for Kenya). Children were enrolled into this study if a parent or guardian was willing to provide written informed consent in accordance with local practice.
G6PD genotyping
Genomic DNA was isolated using QIAamp DNA mini blood kit (Qiagen, Hilden, Germany). A 968 bp fragment of the
G6PD gene containing the polymorphisms 202G > A and 376A > G was amplified by PCR using primers 5′-GCCCCTGTGACCTCCCGCCA-3′ (forward) and 5′-GCAACGGCAAGCCTTACATCTGG-3′ (reverse). The main focus was directed only to these two variants although some other deficient genetic mutations such as A542T (Senegal 1 %, The Gambia 2.2 %), G680T (The Gambia 0 %, Senegal 0 %) and T968C (The Gambia 7.8 %, Senegal 10 %) have been reported at a substantially lower prevalences only [
14,
15], and might have been present in this study population. However, they seem not to be responsible for the prevalence of G6PD deficiency in all parts of Africa [
20]. Briefly, 10 ng of genomic DNA were added to a 20 µL reaction mixture containing 1 × PCR buffer (20 mM Tris–HCl pH 8.4, 50 mM KCl, 1.5 mM of MgCl
2), 0.125 mM of dNTPs, 0.25 mM of each primer and 1 U Taq DNA polymerase (Qiagen, Hilden, Germany). The PCR was run on a PTC-200 Thermal cycler (MJ Research, Waltham, USA). Thermal conditions after initial denaturation (94 °C, 5 min) were 35 cycles of 94 °C for 45 s, 65 °C for 1 min, and 72 °C for 1 min. PCR reactions were completed with a final extension step of 72 °C for 5 min. PCR products were visualized through electrophoresis on a 1.2 % agarose gel stained with SYBR green I in 1x Tris-electrophoresis buffer (90 mM Tris–acetate, pH 8.0, 90 mM boric acid, 2.5 mM EDTA).
Subsequently, PCR products were purified (Exo-SAP-IT, USB, Affymetrix, USA) and directly used as templates for DNA sequencing using the BigDye terminator v. 1.1 cycle sequencing kit (Applied Biosystems, Foster City, USA) on an ABI 3130XL DNA sequencer.
G6PD polymorphisms were identified by assembling the sequences with the reference sequence of
G6PD (NG_009015.2) gene using the Codoncode Aligner 4.0 software (
http://www.codoncode.com) and visually reconfirmed from their electropherograms.
Statistical analysis
Data were analysed by using GraphPad Prism v. 5.0 for windows (GraphPad software, San Diego, CA). The effect of G6PD genotypes was determined on initial parasitaemia and haemoglobin values using a multivariate regression model. The children were classified in groups of normal, intermediate (female heterozygous) and deficient (hemizygous males and homozygous females) individuals. To evaluate the effect of G6PD genotypes on haemoglobin concentrations, the model included adjustment for age, gender, centre, weight, temperature and parasitaemia. To investigate the G6PD effect on parasitaemia, parasite densities were log-transformed and the model included adjustment for age, gender, centre, weight, haemoglobin levels and temperature. For the construction of the multivariate regression model, a subjective model-building approach that excludes possible confounders such as gender, age and origin of study participants was applied. Kruskal–Wallis of One-way ANOVA with Dunn’s Multiple Comparison and Mann–Whitney tests were used to determine the differences among categories. The level of significance was set to a P value of 0.05.
Discussion
Glucose-6-phosphate dehydrogenase deficiency has raised in frequencies in malaria-endemic settings as a consequence of the evolutionary pressure exerted by malaria on the human genome [
16]. A plethora of previous studies have indicated and suggested a correlation between malaria endemicity and the occurrence of G6PD deficiency (reviewed in [
25]).
The main objective of this study was to assess the distribution of
G6PD genotypes among African children from three geographically countries presenting with severe malaria and participating in the SMAC clinical trial on different artesunate treatment regimens. While meanwhile many
G6PD variants have been described [
8,
11], the main focus was on the most relevant three variants (376A (G6PD type B, no deficiency), 376G (moderate G6PD deficiency type A
+) and 202A severe G6PD deficiency type A
−) in Africa [
26].
The prevalence of severe G6PD deficiency as determined genetically was higher compared to previous findings from Gabon with 17 % in the present study
versus 14 % reported earlier among males and 7 vs 2 % among females [
27]. In Ghana 28 % were observed in this study, compared to 9 % among males and 5 vs 3 % among females indicated previously [
28]. However, in Kenya the prevalence of severe G6PD deficiency among females was lower compared to previous reports with 0 vs 5 % and equal among males 8 vs 8 % [
29]. Regardless of the site, G6PD deficiency was considerably higher among males (22 %) compared to females (3 %), with an overall prevalence of 13 % across sites. This finding confirms that males are affected by this blood disorder rather than females and that G6PD deficient females are rather uncommon [
26].
A significant association of moderately and severely deficient
G6PD genotypes and haemoglobin levels according to the baseline data was observed. In fact, compared to children with the normal
G6PD genotype (haemoglobin median: 9.3 g/dL),
G6PD heterozygous (haemoglobin median: 6.6 g/dL) and deficient (haemoglobin median: 8.3 g/dL) children had 2.7 and 1 g/dL lower haemoglobin concentrations, respectively. The results contradict previous findings, which did not observe any association between
G6PD genotypes and haemoglobin levels [
30‐
32]. While in their study, May et al. [
32] found lower levels of haemoglobin in
G6DP deficient individuals, the association was not significant. An explanation could be the different designs of that and the present study. Here, patients with severe malaria were included, which implies high parasitaemia and haemoglobin concentrations <5 g/dL, whereas in the above cited study, the authors excluded patients with haemoglobin concentrations ≤7 g/L [
30] and recruited individuals only with uncomplicated malaria [
31] or who were asymptomatically infected [
32].
In the case of SMA, there was no difference of haemoglobin levels among children with different
G6PD genotypes. This may suggest that G6PD deficiency is not associated with this complication. However, this is in fact surprising especially as a significant difference of baseline haemoglobin values between G6PD normal and deficient individuals was observed. Moreover, in a large case–control study in Kenyan individuals, a significant increased risk to severe malarial anaemia associated with lower haemoglobin levels in G6PD deficient children with severe malaria at the time of hospital admission was found [
33]. The likely explanation could be due to the sample size which was rather small.
Although not significant, parasite densities were lower in G6PD deficient children than in G6PD heterozygous, compared to G6PD normal individuals. However, this trend is in agreement with previous works [
34,
35]. Conflicting results have been reported regarding parasitaemia and the various
G6PD genotypes. Other studies have indicated a significantly lower parasitaemia associated with female G6PD heterozygous individuals, suggesting a protective advantage by this genotype [
36‐
39].
In comparison to individuals with uncomplicated malaria, this study population was hyperparasitemic according to the criteria of severe malaria provided by WHO with lower parasitaemia in G6PD deficient individuals. However, either in uncomplicated or in severe/complicated malaria, G6PD deficiency was always associated with lower parasitaemia [
30,
33]. Although the underlying genetic mechanisms are not completely clear, mechanisms suggested are impaired growth of
P. falciparum parasites in G6PD deficient red blood cells [
40] and slow rates of parasite replication [
41,
42], more efficient clearance of infected red blood cells [
43], and lower abundance of
P. falciparum 6-phosphogluconolactonase mRNA in parasites from G6PD deficient individuals [
44] both uncomplicated and severe malaria.
Several limitations apply to this study. Although some of the observations are in agreement with previous findings, the study group is rather small. The focus was only on the
G6PD mutations 202G > A and 376A > G, although other variants such as A542T, G680T or T968C have also been reported to contribute to G6PD deficiency, albeit at far lower frequencies. The associations observed between
G6PD genotypes and the baseline clinical parameters at admission could in fact also be the result of different circumstances. First, only patients with severe malaria were included, which could explain the findings, especially as the
G6PD deficient allele 202A appears to confer protection against cerebral malaria and increases the risk of severe malarial anaemia [
16,
17]. Second, depending on the mechanism of protection, G6PD deficiency might be associated with delayed presentation to the hospital and be a plausible explanation of the differences observed in haemoglobin concentrations at first admission. Third, alpha-thalassaemia, another haemoglobinopathy which has been associated with lower haemoglobin levels in alpha-thalassaemic Nigerian children and adults [
45] cannot be ruled out as a confounding variable as it was not investigated in this study. The statistical analysis using multivariable regression model is in part subjective due to the adjusted variables added.
Authors’ contributions
TPV designed the experiments, supervised the experiments and data analysis. AAA, TA, BRO are involved in patient recruitment from respective study sites and are principal investigators for SMAC study. PGK contributed to study design and materials. CNN performed the experiments and drafted the manuscript. TPV and CGM contributed in writing the manuscript. All authors read and approved the final manuscript.