Detection of high levels of mutations involved in anti-malarial drug resistance in Plasmodium falciparum and Plasmodium vivax at a rural hospital in southern Ethiopia

Background In Ethiopia, malaria is caused by Plasmodium falciparum and Plasmodium vivax, and anti-malarial drug resistance is the most pressing problem confronting control of the disease. Since co-infection by both species of parasite is common and sulphadoxine-pyrimethamine (SP) has been intensively used, resistance to these drugs has appeared in both P. falciparum and P. vivax populations. This study was conducted to assess the prevalence of anti-malarial drug resistance in P. falciparum and P. vivax isolates collected at a rural hospital in southern Ethiopia. Methods A total of 1,147 patients with suspected malaria were studied in different months across the period 2007-2009. Plasmodium falciparum dhfr and dhps mutations and P. vivax dhfr polymorphisms associated with resistance to SP, as well as P. falciparum pfcrt and pfmdr1 mutations conferring chloroquine resistance, were assessed. Results PCR-based diagnosis showed that 125 of the 1147 patients had malaria. Of these, 52.8% and 37.6% of cases were due to P. falciparum and P. vivax respectively. A total of 10 cases (8%) showed co-infection by both species and two cases (1.6%) were infected by Plasmodium ovale. Pfdhfr triple mutation and pfdhfr/pfdhps quintuple mutation occurred in 90.8% (95% confidence interval [CI]: 82.2%-95.5%) and 82.9% (95% CI: 72.9%-89.7%) of P. falciparum isolates, respectively. Pfcrt T76 was observed in all cases and pfmdr1 Y86 and pfmdr1 Y1246 in 32.9% (95% CI: 23.4%-44.15%) and 17.1% (95% CI: 10.3-27.1%), respectively. The P. vivax dhfr core mutations, N117 and R58, were present in 98.2% (95% CI: 89.4-99.9%) and 91.2% (95% CI: 80.0-96.7%), respectively. Conclusion Current molecular data show an extraordinarily high frequency of drug-resistance mutations in both P. falciparum and P. vivax in southern Ethiopia. Urgent surveillance of the emergence and spread of resistance is thus called for. The level of resistance indicates the need for implementation of entire population access to the new first-line treatment with artemether-lumefantrine, accompanied by government monitoring to prevent the emergence of resistance to this treatment.


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 malariaendemic 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][7][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][16][17][18][19][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][25][26][27][28].
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.

Study site and biological samples
The Gambo General Rural Hospital is situated in the province of West Arsi, Ethiopia, 245 kilometres southeast of the capital, Addis Ababa ( Figure 1), at an altitude of 2,250 metres (approximately 7,382 feet) above sea level. The hospital provides daily health care and covers nine peripheral health centres serving 23 villages, with a total population of 98,000. Patients may come from bordering areas, lying at lower altitudes.
The study was conducted during different calendar months across the period 2007-2009. At the hospital, a total of 1,147 blood samples were collected from patients (age range, 1 month to 75 years) with clinically suspected malaria (see Table 1). Two capillary blood specimens were taken from each patient, one for confirmation of Plasmodium infection by microscopic observation of thin and thick Giemsa-stained blood films, and the other on filter paper (Whatman © 3 MM) for molecular assay.
DNA extraction was performed using commercial kits (Speedtools tissue DNA Extraction Kit, Biotools), and Plasmodium species were subsequently confirmed by Seminested-Multiplex PCR [38]. The nested PCR-RFLP technique was used to analyse: codons 76 of pfcrt, 86 and 1246 of pfmdr1, 51, 59, 108 and 164 of pfdhfr and 436, 437, 540 and 581 of pfdhps from samples infected by P. falciparum [39]; and codons pvdhfr 57, 58 and 117 and pvdhps 383 and 553 from samples infected with P. vivax. The laboratory conditions and restriction enzymes have been described elsewhere [17,18,30]. As in other resistance studies [40], genes that showed mixed wildtype and mutant alleles were deemed to be mutant only when the amplification fragment was more intense for the latter. The study protocol was approved by the research review boards of the Institute of Health Carlos III (Madrid, Spain) and Gambo General Rural Hospital. Study participants with microscopy-confirmed malaria infection were treated in accordance with Ethiopian national guidelines [4,5]. A descriptive analysis was performed and estimates of prevalence along with their 95% confidence intervals (CIs) were obtained, using the method described by Newcombe [41].

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.

Conclusions
1. The prevalence of Plasmodium species is slightly different from the results obtained by other authors in previous studies conducted in Ethiopia, showing an increase of vivax malaria cases and the presence of P. ovale for the first time.
2. The prevalence of mutations linked to CQ and SP resistance detected in P. falciparum and P. vivax in southern Ethiopia is higher than expected.
3. Implementation of permanent surveillance of resistance molecular markers as an epidemiological tool will enable health professionals and authorities to be alerted to the possible emergence and spread of resistance to new lines of treatment, such as ACT.
4. It is essential that the entire population be provided with access to new first-line treatment with ACT, which should be given under government supervision to avoid the drug being mismanaged and so prevent the emergence of any new resistances.