A large number of SNPs related to resistance to different antimalarial drugs that have been used as first-line or preventive treatment have been evaluated in this study. Evolution of resistance to SP (pfdhfr, pfdhps), CQ and AQ (pfmdr1 and pfcrt) and artemisinin (pfk13) has been assessed taking into account the changes in public health treatment strategies.
Normally, when a country withdraws a given treatment due to drug resistance, the presence of sensitive parasites increases with respect to the resistant population over a given time period. However, it should also be borne in mind that misuse of treatments and the great plasticity of the parasite can make it possible for mutant parasites to increase and spread.
Numerous efficacy studies of antimalarial drugs were carried out in Equatorial Guinea between 1999 and 2019. Moreover, first-line treatments changed over time until 2008, when ACTs were introduced, the first being ASAQ. From 1990 to 1992, in vitro studies [44] found that CQ had a resistance of 16% and SP resistance was not particularly high (> 5%), thus meaning that SP could be used as a treatment in Equatorial Guinea. A subsequent in vivo study assessing the evolution of CQ and SP efficacy from 1992 to 1999 [45] found that CQ had 40% resistance and SP had 16% resistance, therefore CQ and SP were not used as treatment for malaria after 1999. Although CQ was withdrawn as a treatment in 1999, it can be seen that the frequency of resistance-related mutations in pfmd1, pfcrt and the combination pfmdr1 + pfcrt reached a maximum in 2001, after which it started to decrease. This may be because it is estimated that it usually takes two years for the parasite to show any change in response to treatment, which is why efficacy studies are conducted at least every two years. However, this may also be explained by the delay in implementing a uniform treatment strategy throughout the country. Resistance to CQ is linked to a mutation at codon 76 of pfcrt, which has proven to be the chief determinant of CQ resistance [46]. As such, the decreasing trend in this mutation and the haplotype 86Y/76T indicates that the withdrawal of CQ as a treatment was effective and has been maintained over time. As the selective pressure exerted by CQ on the parasite population has been removed, the frequency of the wild-type population has increased with respect to the mutant population. As noted by Achieng et al [47], what we would expect over time is a greater increase of wild pfcrt K76 and pfmdr1 N86 due to the longer withdrawal time of CQ and distribution of AL as treatment. It would therefore be of interest to revisit the mutational profile for the pfmdr1 and pfcrt genes of P. falciparum over a period of at least two years to see how this changed in the presence of AL.
The downward trend in these mutations means that the susceptible population is higher than the mutant population, thus meaning that the efficacy of other antimalarial drugs such as AQ and MQ is assured. However, other studies have shown that the reintroduction of CQ treatment might rapidly lead to the selection of mutant populations again [48].
Likewise, selection of the wild-type K76 codon in the P. falciparum CQ resistance transporter gene (pfcrt) has been associated with pressure from AL treatment (first-line of treatment in Equatorial Guinea from 2020) in a number of studies [49–54]. Indeed, dramatic increases in the prevalence of wild-type pfcrt K76 and pfmdr1 N86 have been associated with discontinuation of CQ and deployment of AL in western Kenya, although AL continued to be effective with these changes [47]. In Equatorial Guinea, there has been an evolution in this direction since 1999, i.e. a decrease in mutants in favour of an increase in wild populations. Thus, the wild genotype pfmdr1 N86 had a frequency of 86%, whereas that for pfcrt K76 was 100%, in 2019.
Although it was known from previous studies that there was 25% resistance to SP in Equatorial Guinea since 1999 [45], it was still used as a treatment, either alone or in combination, until 2008. In 2005 [3] a new study in which the combinations ASSP and ASAQ were tested was carried out in order to have two treatment alternatives in combination. Despite its slightly yet constantly increasing trend in resistance, SP was never discontinued as a treatment. These combinations were used in Equatorial Guinea from 2003 until 2008, when ASAQ, the first ACTs used in that country, was introduced. The use of SP in combination with AQ and AS has been widespread for many years, with the serious harm to resistance that its use entails. Accordingly, this has led to an increase in genetically resistant populations, as can be seen in the results section, and we can even see some SNPs that are fixed in the population, as is the case of pfdhfr 108N, which has a frequency of 100%. The use of these combinations has created a scenario with selective pressure in which some of the mutations not only increase and spread but also become fixed in the parasite population.
Importantly, the 164L mutation in pfdhfr, which is related to a significant resistance to SP [55], has not been detected in any of the years studied in this study. However, this mutation was previously detected in Equatorial Guinea, by us, in samples from 2013 [56]. In light of the above, the data obtained from the analysis of pfdhfr + pfdhps haplotypes over time reveal that there has been no real or total withdrawal of SP as a treatment. It can be seen that the partially resistant (IRNG) and fully resistant (IRNGE) haplotypes have been progressively increasing since 1999, reaching their peak in 2019, whereas the super-resistant ones (IRNGEG) have always had a low trend over time, never exceeding 2% and not being found in 2019. The data obtained are consistent with previous studies with samples from the island of Bioko, where the partially and fully resistant types were the most common and the frequency of the super-resistant type was very low [57].
Evidence for the misuse of SP in the mainland region of Equatorial Guinea, where it was found that 27.3% of children had received artemether in monotherapy, 13.8% SP and only 6.8% had received the official ACT treatment (ASAQ) [58], supports the hypothesis of the influence of incorrect treatments on the evolution of resistance. This is probably because the official first- and second-line treatments are not available countrywide. As SP is still used as a treatment and has not been reserved only for SP-IPT in the two main populations vulnerable to malaria, namely pregnant women (SP-IPTp) and children under 5 years of age (SP-IPTi), its efficacy has been compromised [17, 59].
Although resistance-related haplotypes to SP exhibit an upward trend, it has been observed that mutations at positions 540E (36.6%) and 581G (1.4%) are not sufficiently high to jeopardise the use of SP in IPT. Current WHO recommendations suggest that SP-IPTp should be discontinued if the frequency of 540E exceeds 50% and that of 581G exceeds 10% [57]. Based on current evidence, IPTp and IPTi remain effective in preventing the adverse consequences of malaria on maternal, foetal and infant outcomes in Equatorial Guinea. However, the implementation of control measures in the country should be maintained to avoid the spread of these mutations and the consequent reduction in the efficacy of IPTp.
The data obtained in this study are similar to those observed in countries bordering Equatorial Guinea, such as Cameroon and Gabon. In Gabon [60] for instance, the partially resistant haplotype appeared in 2014 with a frequency of 92.9% (compared with 85.2% and 81.1% in Equatorial Guinea in 2013 and 2016, respectively). Similarly, this haplotype is also the most frequent in Cameroon [61], this same haplotype is also the most frequent. Therefore, it seems that the distribution of parasites with resistant haplotypes to SP is quite homogeneous in the area.
The introduction of ACTS as a treatment for malaria was very effective in mitigating the threat of resistance to antimalarial treatments. In 2006 [5], an efficacy study was carried out to determine the efficacy of ASAQ, the first time that an ACT for treating uncomplicated malaria had been tested in Equatorial Guinea. As a result, ASAQ began to be used as first-line treatment in 2008, when the National Guidelines were changed, and AL as second line. Two years after the introduction of ACTs, in 2010, a new efficacy study was conducted for ASAQ [62] and its efficacy was found to be 95%, therefore its use as first-line treatment was maintained. However, it was difficult to maintain patient adherence [63] to this treatment due to side effects such as headache, nausea, tinnitus and fatigue. The presence of these side effects, and the lack of adherence to treatment, has led to the use of artemether as monotherapy in some areas of the country. Monotherapy is not permitted by WHO because it may favour the emergence of ART-resistant parasite populations, which could threaten the future efficacy of ACTs in the country.
The last efficacy study carried out in Equatorial Guinea in 2017/2018 [6], showed that the efficacy of ASAQ and AL was close to 95% and that no ART resistance was detected. Following the completion of this efficacy study, a new National Therapeutic Guideline for malaria was published in January 2020. This new guideline shows a change in the lines of treatment, recommending AL as the first line of treatment and ASAQ as the second, in order to facilitate patient adherence to treatment.
The study of pfk13 gene sequences carried out herein to determine the presence of mutations related to resistance to ART, and therefore to ACTs, allows us to ascertain whether this combination therapy is being used correctly in Equatorial Guinea. The development and spread of ART-resistant P. falciparum outside the Greater Mekong Subregion (GMS) poses a great challenge, particularly to sub-Saharan Africa, where in 2020 it accounted for 90% of global malaria cases and 95% of malaria deaths [1]. Genetic analysis of the whole genome sequences previously performed showed that the resistant isolates were classified as an African-specific group. This suggests that they may have originated in Africa and not through the migration process from GMS [64].
Current data from this study report a low prevalence (5%) for pfk13 mutations, both synonymous and non-synonymous, and none of these was among those associated with ART clearance delay in Southeast Asia. The allelic frequencies reported for Central, West and East Africa are generally less than 6% [65–67]. Our result is within this limit, because the frequency of the pfk13 mutation in Equatorial Guinea has increased in relative terms since 1999, reaching 5% in 2019. A study conducted in Cameroon, a country bordering Equatorial Guinea, revealed a high mutation rate of 15.1% for isolates containing at least one non-synonymous mutation [68].
The most common non-synonymous mutation (A578S) observed in Africa was detected in a sample from 2013, as well as in in two samples from the therapeutic study carried out in Equatorial Guinea (2017–2018) [6]. A similar study of pfk13 carried out in Equatorial Guinea detected that 2.04% of cases exhibited the non-synonymous A578S mutation [69]. The same mutation was detected in the same year (2013) in Cameroon and Gabon [32, 60], both of which border the mainland region of Equatorial Guinea. This mutation (578S) was detected in the 2017/18 efficacy study but not in the 2019 samples in the current study [6]. Moreover, it is the most common mutation in Africa, therefore it is likely that if we analysed a larger number of samples from 2019 it would also be detected. The non-synonymous mutation E612K (GAA to AAA) was detected in Cameroon in 2017 [37] and the same mutation appeared in our study, but synonymous (E612E, GAA to GAG), in a sample from 2001. Given that this mutation has been detected as non-synonymous in Cameroon and as synonymous in our study, it could be hypothesised that this point is an area of genetic instability. It could also indicate that this is the first step for a non-synonymous mutation to occur in Equatorial Guinea in the future. One interesting finding is the detection of the synonymous mutation C469C, which appears in one sample from 2013 and in two samples from 2019. It will be interesting to continue characterizing more isolates and to see if this mutation continues to appear, or if its frequency continues to increase over time. Surveillance will have to be established to see if in the future such a synonymous mutation could become non-synonymous and have clinical significance for ACTs resistance.
It is essential to continue to make correct use of the first- and second-line treatments (AL and ASAQ respectively) to avoid the appearance of new mutations, and good surveillance is essential to be able to quickly detect possible mutations from SEA that might be introduced into the country and, if they appear, to prevent them from spreading.
Taken together, the low frequencies of pfk13 mutant alleles found in Equatorial Guinea suggest that ART-resistant parasites are not under evolutionary selection in this country, thus reinforcing the assumption that such mutations are rare in Africa. Furthermore, none of the polymorphisms known to be involved in ART resistance in Asia has been associated with ART resistance in Africa. Therefore, local ART-resistant P. falciparum strains may emerge independently in Equatorial Guinea and in the African continent under constant drug pressure from ACT, possible misuse of these drugs if treatment guidelines are not followed, non-adherence to treatment, self-medication and the introduction of counterfeit drugs as is known to be occurring [2].
The effect observed on the evolution of parasites with mutations related to CQ resistance, which have decreased significantly compared to parasites from 20 years ago, indicates that avoiding pharmacological pressure by withdrawing treatment is one of the most important aspects affecting the increase of sensitive parasites compared to resistant parasites. Regarding to mutations in pfdhfr and pfdhps is important to establish intensive surveillance because the use of SP as a preventive treatment in pregnant women and children under 5 years of age could be at risk. As for ACTs, as recommended by WHO, treatments should only be administered when the presence of the parasite has been identified by a diagnostic method, avoiding unnecessary treatments. Compliance with the National Therapeutic Guidelines for malaria is mandatory to avoid the use of other treatments that have already been withdrawn and are no longer effective.
All of the above highlights the need for constant surveillance to detect resistance-related mutations early so that we can prevent them from spreading. Consequently, this will ensure that the population receives better health care and that treatment to cure malaria is adequate, as a complete cure is a benefit not only for the patient, but also for the whole community.