Background
Malaria is the vector-borne disease with the highest impact on the world's human population. In 2008, there were an estimated 243 million cases, leading to nearly 863,000 malaria-related deaths [
1]. Although malaria-endemic areas are mainly restricted to tropical and subtropical regions at present, several models nonetheless project the geographical expansion of potential malaria transmission over the next few decades, along with more substantial changes later in the century [
2]. Approximately 90% of clinically manifest infections of and practically all deaths from malaria are caused by
Plasmodium falciparum. Ranking second in importance is
Plasmodium vivax, which accounts for nearly 10% of global malaria incidence. Roughly 90% of all malaria cases occur in tropical Africa [
3].
Approximately 52 million people in Ethiopia are considered to be at risk of the disease [
4]. While 4 to 6 million clinical malaria cases are annually reported by the country's health facilities, the real number is estimated to be as high as 10 to 15 million [
5]. The major
Plasmodium species causing malaria in Ethiopia are
P. falciparum (about 60% of cases) and
P. vivax (about 40% of cases), with the former being the cause of the most severe clinical manifestations and most deaths [
6‐
8]. Malaria transmission follows a seasonal pattern (September-November), depending on the altitude and rainy season [
7,
9]. Epidemic malaria is frequent [
10], particularly in the highlands (1,000-2,000 m above sea level).
Anti-malarial drug resistance in
P. falciparum and
P. vivax is the most pressing problem confronting malaria control in many endemic countries [
11,
12].
Plasmodium falciparum has developed resistance to a series of drugs, and
P. vivax is resistant to chloroquine (CQ) and not primaquine-tolerant (PQ) [
13,
14]. Furthermore, in countries where intensive use has been made of sulphadoxine-pyrimethamine (SP), resistance to these drugs has appeared in
P. vivax populations [
15‐
20], though some authors view this as being a sign of innate resistance to sulphadoxine in such parasites [
21].
Current studies report high rates of therapeutic failure (72%) in some areas of Ethiopia [
22], due to the presence of SP resistance in the two principal species. In 2004, artemether-lumefantrine (AL) (Coartem
®) was introduced into the country as a first-line treatment, and the national guidelines have since prescribed this drug for uncomplicated malaria [
4]. Even so, not everyone is benefiting from this treatment because 85% of the population live in rural areas where access to basic health care is severely limited.
In general, the presence of high resistance to SP has stopped it from being used as a treatment in African endemic countries, and so nowadays it is only used, in combination with insecticide-treated mosquito nets, for Intermittent Preventive Treatment (IPT) in pregnant women and in areas where it has been given to children under five years of age has provided encouraging results [
23].
The study of point mutations (Single Nucleotide Polymorphisms - SNPs) in molecular markers is an extremely useful epidemiological tool, which enables the emergence and spread of mutant parasites to be monitored. Analyses of different molecular markers of resistance are currently used, namely: the T76 point mutation in the
P. falciparum pfcrt gene; the Y86, F184, C1034, N1042 and Y1246 point mutations in the
pfmdr1 gene [
24]; the L164, N/T108, I51 and R59 point mutations in the
pfdhfr gene; and the G437 and G581, A/F436, E540 and S613 point mutations in the
pfdhps gene [
12,
24‐
28].
Plasmodium vivax resistance to SP is associated with mutations that appear in the homologous genes,
pvdhfr (N117, R58 and T117) and
pvdhps (G383, G553) [
17,
18]. Mutations in
pvdhfr have been implicated in resistance to pyrimethamine
in vivo[
29]. Furthermore, it seems that
pvdhfr mutations at residues 117 and 58 have been observed to arise first when drug pressure is applied [
17,
18,
30]. The double and triple mutations, N117/R58 and N117/R58/L57, are associated with delayed parasite clearance following SP treatment [
17] and with therapeutic failure in many regions of south-east Asia [
17,
31,
32], Iran [
30], Pakistan [
33], India [
34], Colombia [
28] and Madagascar [
35].
However, the appearance of different combinations of L57/R58/N117/G383/G553 mutations in both the
pvdhfr and
pvdhps genes has been associated with SP treatment failure only in patients infected by
P. vivax and not in those co-infected by both species [
36,
37] in which no mutations have been found in the orthologous genes of
P. falciparum.
This study sought to determine the following: the prevalence of P. falciparum and P. vivax in patients with clinically suspected malaria, attending the Gambo General Rural Hospital in Ethiopia; and the prevalence of mutations in genes involved in resistance to different anti-malarial drugs, based on samples collected in different months across the period 2007-2009.
Discussion
In areas where P. falciparum and P. vivax co-exist, parasite-specific diagnosis and choice of effective treatment is crucial to prevent the emergence and spread of resistance. In malaria-endemic areas, the occurrence of P. falciparum/P. vivax co-infection is frequent and it is, therefore, common for P. vivax to have been exposed to treatment with SP.
In Ethiopia, not only are there wide interregional differences in the endemicity and transmission of malaria, but there is also a significant lack of information on the effectiveness of anti-malarial drugs. In 1999, the CQ treatment failure rate in the first two weeks still stood at 88% in the centre of the country [
42]. In 2005, SP therapeutic failure within two to four weeks of follow-up was 36% and 72% respectively. In this same year, however, treatment with AL yielded an appropriate clinical and parasitological response of 99% [
22].
With regard to the prevalence of species, a decrease of
P. falciparum mono-infection and an increase of
P. vivax mono-infection was observed in contrast to other results obtained by other studies previously conducted in Ethiopia [
6,
7,
43,
44]. The predominant species was
P. falciparum, which appeared in approximately 53% of cases, followed by
P. vivax with a prevalence higher than 30%, and a small proportion cases of co-infection by both species. In addition, two cases of infection by
Plasmodium ovale were detected. This finding is rather singular; in that
P. ovale has never been previously described in Ethiopia, and might be accounted for by the migration process from the western side of the continent to this country, or people returning after having gone to work in other areas of Africa with prevalence of
P. ovale. In the future, more studies about the prevalence of malaria species should be performed in Ethiopia, to know the factors are influencing in these changes of prevalence.
As with other studies in which high rates of therapeutic failure of CQ and SP have been detected, our results likewise show a high prevalence of these mutations linked to resistance in
P. falciparum, particularly T76 in
pfcrt (responsible for resistance to CQ) and the triple mutation in
pfdhfr (responsible for treatment failure in SP) (see Table
3 and Figure
2). Furthermore, the high rate of the
pfdhps double mutation, G437/E540, responsible for conferring a high degree of resistance to sulphadoxine and for therapeutic failure to SP in the presence of the triple mutation in
pfdhfr, has resulted in the
pfdhfr/pfdhps quintuple mutation being present in over 80% of cases. This corresponds to the low SP-treatment efficacy rates registered in different regions of Ethiopia [
22] and confirms the need for AL to be made available countrywide as the first-line treatment.
The
pfcrt T76,
pfmdr1 Y86 and
pfmdr1 Y1246 mutations are very useful molecular markers of CQ resistance in areas where resistance rates are low to mild [
45,
46]. This study showed high to moderate prevalence of
pfcrt and
pfmdr1 mutations, respectively, to be consistent with low CQ efficacy in Ethiopia [
42].
In the case of
P. vivax, there was a high prevalence of the
pvdhfr R58 and
pvdhfr N117 mutations. Furthermore, the
pvdhfr double mutation, N117/R58, appeared in 52 out of 57 cases infected by
P. vivax (taking the cases of
P. falciparum co-infection into account), and 42 out of 47 cases infected by this species alone. The
pvdhps G553 mutation always appeared in combination with
pvdhfr double mutation R58/N117, resulting
pvdhfr/pvdhps R58/N117/G553 mutation in two cases of
P. vivax mono-infection. This finding could explain a possible SP treatment failure in these patients, as has been associated in other studies [
36,
37]. Yet
pvdhfr T117, L57,
pvdhfr triple or
pvdhfr/pvdhps quadruple mutations were not in evidence, in the last case neither in
P. vivax mono-infection nor in
P. vivax-P. falciparum co-infection. Although these multiple mutations seem to be necessary for
in vivo resistance [
17,
18,
43], this finding nevertheless suggests widespread SP use, since the
pvdhfr mutations arise first under drug pressure [
17,
18,
43].
It should be taken into account that, when it comes to combined
P. falciparum/P. vivax infection, patients should be treated with a blood schizonticide (for both species) and a tissue schizonticide (for hepatic hypnozoites of
P. vivax). In the past, SP was used as treatment for
P. falciparum. As a result,
P. vivax came into contact with the same treatment when it was present in mixed infections, and mutations in different genes linked to SP resistance have thus appeared in this species. The mutations in the
pvdhps gene would be equally accounted for, i.e., here, only the G553 mutation, implicated in resistance to sulphadoxine and corresponding to the G581 mutation in
P. falciparum, displayed a very low prevalence and appeared in the presence of the
pvdhfr double mutation, N117/R58. In contrast, the G383 mutation, which corresponds to G437 in
P. falciparum, was observed in neither case. This strengthens the hypothesis that the asymmetric selection process of mutations observed in
P. falciparum is also applicable to
P. vivax[
21].
Samples co-infected by these two species showed mutations in either pfdhfr/pfdhps or pvdhfr, which would explain why treatment for both these species of Plasmodium could fail in such patients.
In view of the genotype results obtained, it would be advisable for a study of therapeutic efficacy to be conducted in this area, to ascertain whether the genotypic are the same as the phenotypic data, i.e., whether the parity mutation-resistance is produced in vivo.
Moreover, it would be advisable for future research to verify whether there were mutations in genes (such as atp6) related with resistance to artemisinin derivatives, the use of which as an anti-malarial drug has spread in Africa. Artemisinin-based combination therapy (ACT) (for example, using Coartem®) has come into use in Ethiopia, though it is proving difficult to implement in rural areas due to the continued use of drugs, such as SP and CQ, in combination with other medication, a practice that continues to favour the expansion of resistance.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
PM was involved in the molecular studies, interpretation of results and drafted the manuscript. AFM and VG helped to do molecular studies. AL helped to do molecular studies and carried out sample collection. JMR performed the statistical analysis and interpretation, and helped to draft the manuscript. FR helped to collect the samples in Gambo (Ethiopia). AB helped to draft the manuscript and PB conceived and funded the project, carried out sample collection, and drafted the manuscript. All authors read and approved the final manuscript.