The data from the current study show a decrease in the prevalence of molecular markers associated with chloroquine resistance in Katete District, Eastern Zambia between 2012 and 2016, or nine and thirteen years post-chloroquine treatment withdrawal. Specifically, the prevalence of Pfcrt 76T and Pfmdr1 86Y both decreased from 1.8% and 2.1%, respectively, in 2012 to 0% in 2016. These data are consistent with the results of other studies conducted in different locations within Zambia [18, 22]. Additionally, these data further suggest that the withdrawal of antimalarial drug pressure from the parasite population seemed to have resulted in the re-expansion of the wild type parasites carrying the K76 and N86 residues on the Pfcrt and Pfmdr1 genes, respectively. This observed recovery of sensitive strains seems to have occurred over a long period of time as seen in the prevalence of resistant strains in the isolates collected in 2012.
Malawi, Zambia, and Zimbabwe are three African countries that have reported 100% prevalence of wild type codon K76 carrying P. falciparum parasites after official cessation of chloroquine use but also general trend towards restoration of chloroquine sensitivity has been reported for the continent [3, 23, 24]. Malawi was the first country to report the return of chloroquine-susceptible P. falciparum parasites nine years after the withdrawal of chloroquine as treatment [14]. The return of chloroquine-susceptible parasites must have resulted from the re-expansion of chloroquine-sensitive parasites that had a survival fitness advantage over the CQ-resistant strains [25].
Kenya reported that the frequency of the Pfcrt K76T mutation had decreased from 95–23% two decades post-chloroquine treatment withdrawal [26]. In Rwanda, the prevalence of wild type Pfcrt K76 was reported to be at 50% around 14 years after cessation of CQ use while Tanzania reported a similar K76 allelic prevalence between 85.5–93% ten years after the discontinuation of chloroquine from treatment guidelines with regional variabilities [27, 28].
The trends in the re-emergence of chloroquine-susceptible genotypes seen in Zambia, Malawi, and Zimbabwe can be attributed to the fact that chloroquine was completely withdrawn as treatment for uncomplicated falciparum malaria. The replacement for chloroquine, AL, has a different mode of action on the parasite and selects for chloroquine-sensitive P. falciparum [29, 30]. Furthermore, there was no chloroquine drug pressure on the P. falciparum population which allowed the more fit CQ sensitive parasites to thrive. This is not the case in Southeast Asia and South America, where the parasites carrying the mutation have become fixed in the population due to continued use of chloroquine in the treatment of vivax malaria [23, 31].
Malaria epidemiology may also have contributed to the re-emergence of chloroquine-susceptible P. falciparum. Malaria transmission intensity, which impacts both human host immunity and the rate of parasite recombination in the arthropod vector, contributes to the spread of Plasmodium parasite. Different models have shown that high malaria transmission intensity results in high host immunity due to repeated exposure to malaria infections. Additionally, there is a high rate of parasite recombination in high transmission areas when compared to low transmission areas due to multiplicity of infections. Eastern Province was classified as a high malaria transmission area at the time when the samples were collected. Therefore, the chloroquine-sensitive parasites which already have a high fitness advantage over the resistance parasites will quickly increase in the population once the drug pressure is removed. In low transmission areas, however, there are usually unique parasite population characteristics, and each individual receives one infectious bite with a single genotype. In such cases, this single genotype is taken up in the blood meal by an Anopheles mosquito. Consequently, during the sexual reproductive stage of the parasite in the mosquito midgut, there is little opportunity for genetic recombination resulting in a single fixed genotype in the parasite population [32].
This study had limitations and the results should be interpreted with caution, keeping in mind the following reasons; Firstly, the sample size for 2016 was smaller than that from 2012, which may have introduced a selection bias. Therefore, more studies should be conducted with a larger sample size to acquire more accurate estimates of prevalence. Secondly, this study was conducted in one province in Eastern Zambia, so the results cannot be generalized to the whole country because of differences in epidemiologic patterns in the countries. As such, prevalence cannot be compared between provinces, so further studies should examine potential differences between areas of low and high transmission status. Finally, not all of the samples that were considered positive by microscopy were also amplified by PCR, so there was a possibility of amplification bias, which could lead to an under- or over-estimation of the prevalence.