INTRODUCTION
The factors influencing the rate of spread of antimalarial drug resistance, once it has emerged or been introduced in a given area, are still not defined. Theoretically, parasite resistance can emerge for any antimalarial drug, and the occurrence of de novo mutations and drug selection pressure (frequent drug use and long drug elimination half-life) are critical and essential prerequisites.1,2 Nevertheless, vector, parasite, and human host factors probably play an important role.3 Whether antimalarial drug resistance evolves faster in areas of high or low transmission has been a controversial subject since a study in Papua New Guinea reported that substantial inbreeding occurs in malaria parasite populations, even where malaria transmission is intense.4 Several theoretical models have attempted to explain this relationship, but they have generated conflicting predictions as to whether malaria transmission intensity has any role (direct or indirect) on the spread of parasite resistance.5–12 Resistance might spread faster: at the extremes of the transmission spectrum with a minimum at intermediate transmission intensity7,9; it could spread faster where transmission is low,13,14 where it is high,15 where transmission is low for only partial resistance and where it is high for full resistance,16 or transmission intensity has no effect in the early stages of the emergence of resistance.2 These hypotheses, although instructive, are not consistent because they depend on their inherent assumptions about the population biology and dynamics of Plasmodium falciparum.
The malaria parasite, Plasmodium, has an obligate sexual phase in the gut of the female anopheles mosquito that results into a diploid zygote that quickly undergoes a meiotic cell division. This process is associated with genetic recombination and assortment and has important theoretical and practical field implications for the molecular epidemiology and the population dynamics of Plasmodium.12 Meiosis could generate or break down novel parasite genotypes and would, if the genetic basis of drug resistance were polygenic, have important implications for the spread of drug resistance.
If resistance spreads faster in areas of low transmission, interventions aiming at reducing it further (insecticide-treated nets [ITNs], indoor residual spraying [IRS], or a malaria vaccine when available) could, in the later phases, have a negative effect on drug efficacy. A recent review of the theoretical models states that transmission intensity has no direct role in the evolution of drug resistance,17 but rather affects it indirectly through three mediators: 1) average number of parasite clones (clone multiplicity), linked to two antagonistic effectors: sexual recombination and intrahost dynamics; 2) risk of infection, modulating community drug use; and 3) immunity, modulating two effectors: therapeutic drug use and parasite biomass. Some “mediators” can be directly estimated, whereas others can only be estimated through their proxies. For example, the average number of parasite clones per infected host can be directly estimated in field studies18–20; the risk of infection by measuring either the parasite rate (PR) in children < 10 years old or by determining the annual entomologic inoculation rates (AEIR); the effect of immunity by comparing non-contiguous populations exposed to different forces of malaria transmission using predictors such as serological markers21; or the cumulative number of inoculations, if the clinical immunity is dependent on the size of the antigenic repertoire.22 With respect to the “effectors,” community drug use can be estimated by measuring drug metabolites in the blood or the urine,23 whereas therapeutic drug use by determining the proportion of malaria infections that progress to clinical disease (longitudinal follow-up)24,25; sexual recombination can be determined through the dissection of the mosquito gut.4 However, intrahost competition between the different parasite clones and the parasite biomass are difficult to measure, and their effect can only be explained using prima-facie epidemiologic observations.
The different and conflicting theoretical models were compared with data from epidemiologic studies that have investigated the impact of vector control or malaria transmission intensity and its mediators (average number of parasite clones, drug use, infection risk, and immunity) on the prevalence of drug resistance (in vivo treatment failure or the molecular markers linked to drug resistance).
MATERIALS AND METHODS
We searched the published literature in English in the National Library of Medicine through Pubmed and MEDLINE search engines for field studies that compared the prevalence of drug resistance (in vivo treatment failure or the molecular markers linked to drug resistance) in different geographical settings with varying malaria transmission intensity. In addition, studies that monitored trends in drug resistance after introducing interventions that impact malaria transmission intensity (insecticide-treated nets [ITNs] or curtains [ITCs] and indoor residual spraying [IRS]) and reviews and other reports were identified. Published reports were identified using key word searches such as malaria transmission intensity and drug resistance, complexity of malaria infections and transmission intensity, parasite diversity and drug resistance, in vivo efficacy and malaria transmission, and chemo-sensitivity and malaria transmission. Other key words were dihydrofolate reductase (dhfr), dihydropteroate synthase (dhps), Plasmodium falciparum multi-drug resistance gene 1 (pfmdr1), Plasmodium falciparum chloroquine resistance transporter (pfcrt) gene and malaria transmission, vector control and malaria drug resistance, and gene mutations and malaria transmission. In addition, the relevant cited bibliographies in the reports identified were reviewed.
RESULTS
Malaria transmission and community drug use.
At the time of the review, only one study had systematically measured drug use (estimated using the proportion of the population with detectable chloroquine [CQ] and sulphadoxine-pyrimethamine [SP] metabolites in the urine) and the intensity of malaria transmission.23 The level of community CQ and SP use was inversely related with transmission intensity (Figure 1).
Malaria transmission intensity and in vivo treatment failure.
Two studies, conducted in Uganda between 1999 and 2005, measured both the prevalence of in vivo treatment failure in different geographical areas and the intensity of malaria transmission (estimated using either the PR in children < 10 years old or the AEIR). The prevalence of CQ in vivo clinical treatment failure was higher in areas of both high and low transmission, whereas it was at its lowest at intermediate transmission.23,26 However, SP clinical treatment failure increased with increasing malaria transmission (Table 1; Figure 2). Because PR and AEIR data were available for the same sites, and in view of the limitations of the PR in differentiating parasite challenge, especially at very intense transmission, and the low AEIR detection threshold in areas of low transmission (Figure 3), we re-analyzed our previous dataset using a qualitative composite index that takes into consideration both the PR and AEIR instead of the PR alone to rank sites according to an ordered incremental ordinal scale of malaria transmission intensity. In areas where transmission intensity was low, and there were no major differences between the sites according to the AEIR (Mubende, Jinja, Kihihi, Kyenjojo; Figure 3A), the sites with a higher PR were ranked as having a higher transmission intensity. For intense transmission, where sites had similar PR (Apac, Arua, and Tororo; Figure 3B), the sites with a higher AEIR were ranked to have a higher transmission intensity. This approach was also applied to the data from the same sites reported by Francis and others.27 In addition, the prevalence of RI and RIII parasitologic failure has been used in an attempt to capture partial and full resistance as proposed by some models.16,17,26 For CQ, RIII prevalence was high in both low and high transmission settings, whereas the prevalence of RI was higher in low or intermediate transmission (Figure 4). Although SP RIII prevalence was generally low, it tended to be higher where transmission was intense, whereas RI was high at low or intermediate transmission (Figure 5). Treatment failure to combination therapy (CQ + SP and amodiaquine plus SP [AQ + SP]),27 adjusted for age and complexity of infection, decreased with increasing transmission intensity (Figure 6). In the low transmission sites (AEIR: 3–7 infective bites/person/year), the risk of treatment failure for CQ + SP varied from 44% to 73%, and from 11% to 38% for AQ + SP. In the high transmission sites (AEIR: 393–1,564 infective bites/person/year), it varied from 19% to 53% for CQ + SP and from 2% to 10% for AQ + SP. When considering only patients with infections carrying the mutations linked to SP (quintuple dhfr/dhps) and CQ (pfcrt 76) resistance, in the low transmission sites, the risk of treatment failure for CQ + SP varied from 57% to 83%, and from 14% to 40% for AQ + SP; in the high transmission sites, it varied from 25% to 44% for CQ + SP and from 5% to 11% for AQ + SP.
Malaria transmission intensity and gene mutations linked to drug resistance.
Two studies, both carried out in Uganda between 1999 and 2005, measured in different geographical areas both the prevalence of molecular markers linked to drug resistance and the transmission intensity (PR in children < 10 years old or AEIR). The frequency of the pfcrt 76 gene mutations (linked to CQ resistance) was highest at the extremes of the transmission spectrum, whereas the frequency of dhfr 59 mutation (a marker for SP resistance) increased with increasing transmission intensity (Figure 7).28 Moreover, the proportion of samples with at least one mutant strain for dhfr 59 (mixed or pure) was highest at the two sites with the highest transmission intensity.27 Similarly, the risk of carrying an infection with the dhfr 59 mutation was significantly higher at the two sites with the highest transmission intensity than in those of low transmission (range: 81–86% versus 57–77%; reported P < 0.001). However, there was no observed pattern for the prevalence of the dhps mutations because it was high at all sites, varying from 75% to 95% for the dhps Gly-437 mutation and from 74% to 92% for the dhps Glu-540 mutation.
Complexity of malaria infections and treatment efficacy.
The impact of complexity of infections on the antimalarial drug efficacy or the prevalence of molecular resistance makers was studied in Uganda.29 Patients infected with more than three strains had a 3-fold higher odds of treatment failure compared with those infected with one strain. Infections at the sites where transmission was more intense had a higher complexity.
Vector control interventions and falciparum-resistant malaria.
Three studies have assessed the impact of vector control on malaria drug resistance. In Tanzania, one village received permethrin-treated nets in 1998, whereas the other received deltamethrin-treated nets in 2001.30 The use of ITNs was associated with the re-establishment of the wild genotype for the dhfr codons 51, 59, and 108. In addition, the mean clone multiplicity of infections (MOIs) decreased significantly after the introduction of ITNs from 3.9 (range, 3.6–4.2) in 1998 to 2.9 (range, 2.4–3.5) in 1999 and 2.3 (range, 1.9–2.8) in 2000. In Zimbabwe, the risk of CQ treatment failure decreased 4-fold after 4 years of indoor residual spraying and was significantly lower than in the control villages.31 In the high altitude areas (thought to have generally low transmission), a 6-fold lower risk for CQ treatment failure compared with the low lying areas (generally thought to have high transmission) was observed. In Burkina Faso, no increase in the risk of CQ treatment failure or the prevalence of the gene mutations linked to CQ or SP resistance was observed in nine intervention villages that had received ITCs compared with nine control villages, suggesting that reducing transmission did not have a negative impact on the spread of drug resistance.32 All three intervention studies had a limitation because they did not collect drug use data. Therefore, it is difficult to explain whether the link between the vector control interventions was caused by changes in drug use patterns or some other factors. In the Tanzania study, complexity of infection was reduced in the intervention areas, which is a possible explanation for the link between interventions that reduce transmission intensity and the reduced prevalence of drug resistance.
DISCUSSION
Whether antimalarial drug resistance evolves faster in areas of high or low transmission has not been conclusively confirmed or refuted by empirical data. Theoretical predictions that drug resistance evolves faster in areas of low transmission cast some doubts on vector control strategies because they could, in the later phases, accelerate the spread of drug resistance. Although our review is limited by the few studies on this subject, it suggests that the prevalence and spread of both CQ and SP resistance (in vivo treatment failure and the gene mutations linked to resistance) was higher in high transmission areas. This implies that reducing malaria transmission is useful in limiting the spread of CQ and SP resistance and probably for that of other drugs with a similar genetic basis. In Zimbabwe, CQ treatment failure or the mutations linked to CQ resistance decreased after 4 years of IRS. In Tanzania, the prevalence of the dhfr 108, 51, and 59 wild genotypes increased after 2 years of ITNs distribution, whereas in Burkina Faso, ITCs did not have any negative impact on the prevalence and spread of drug resistance.
Areas with extremely low malaria transmission in Uganda show high CQ (in vivo treatment failure or the frequency of gene mutations), CQ + SP, and AQ + SP resistance. Treatment failure to these combinations decreased with increasing malaria transmission. Such pattern was not explained by the prevalence of genetic markers, largely because this was already high, but rather by the higher immunity and ability to clear parasites in individuals living in areas of high transmission. The few studies summarized in this paper do not allow definitive policy recommendations, but they provide important insights into the possible nature of the relationship between the malaria transmission intensity and drug resistance at population level. The data reviewed suggest that reducing malaria transmission may be useful, in most situations, to limit the spread and the prevalence of resistance, but at extremely low transmission, this may not be the case for some drugs. Presently, the World Health Organization (WHO) recommends that all children < 5 years old with fever in areas of intense malaria transmission should be treated presumptively (treatment without confirmed laboratory diagnosis). In the absence of recent malaria transmission intensity data in many countries, most of sub-Saharan Africa is assumed to have intense transmission. However, this review shows that one size might not fit all, even in one country, as the data from Uganda suggest. Consequently, the adjustment of antimalarial interventions to the epidemiologic context is an absolute necessity to get efficient malaria control. In areas of low transmission, presumptive treatment should be avoided through increasing the use of microscopy and introducing rapid diagnostic tests (RDTs) to reduce the drug selection pressure. However, health staff should be re-trained to avoid the present common practice of treating malaria cases even when the laboratory test is negative.33 In intense transmission settings, diagnosis (microscopy, RDTs) before treatment is preferable for adults, whereas for children < 5 years old, there is the urgent need of establishing the cost-effectiveness of such an approach. Furthermore, the choice of the drugs in combination therapy for low, intermediate, or high transmission settings may be important because of the role of elimination half-life in the selection of resistant parasites and in offering post-treatment chemoprophylaxis.
Malaria transmission intensity, drug elimination half-life, and spread of resistance.
The role of drug elimination half-life as an independent factor in the evolution of parasite resistance has been theoretically modeled.2 Drugs with a long elimination half-life, such as mefloquine and SP, have multiple therapeutic advantages. First, patient compliance is improved because these treatments are given either as a single (starting) dose or as a short regimen, which can be directly observed by the health staff. Second, residual drug levels during the post-therapeutic period offer a certain protection against recurrent parasitemia (recrudescence or new infections) for several weeks (~8 weeks for SP) and may help patients to recover from anemia, a major cause of ill health and death in areas of intense malaria transmission. However, these drugs are likely to exert undesirable drug selection pressure during the period their concentrations drop below a critical threshold that can still prevent re-infection by sensitive parasites but not by the partially or completely resistant ones.2 For example, in Kenya, a potent selective pressure for resistance has been observed even under conditions of supervised drug administration and optimal dosage.34 P. falciparum infections appearing between days 15 and 52 (the drug selection window) after SP treatment were more likely to be resistant to pyrimethamine in vitro. Therefore, drugs with a long elimination half-life benefit the individual, but they create a potent selective pressure at population that can accelerate the rate of evolution of resistance. It is possible that the higher prevalence of SP resistance in areas of intense transmission could be partly caused by its long elimination half-life, because in areas of intense transmission, the likelihood that an individual is re-infected 15–52 days after treatment in higher compared with low transmission settings. For example, at one of the Ugandan sites (Apac) with AEIR = 1,586 infective bites/person/year or an average of > 4 infective bites/person/night, there is a higher likelihood that individuals will be infected during the drug selection window for drugs with a long elimination half-life. The latter is not likely to be observed at a Ugandan site such as Mubende with AEIR = 4 infective bites /person/year. Therefore, the data for SP are very illustrative because they could be applicable, in areas of intense transmission, to artemisinin-based combination therapy or other combinations with a mismatch in drug elimination half-life of the partner drugs.
Malaria transmission intensity and community drug use.
Transmission intensity can influence the spread of drug resistance by affecting the frequency of drug use and the risk of selecting resistant parasites. In Uganda, community drug use was inversely related to intensity of transmission,23 a surprising finding considering that one would have expected a higher number of clinical/fever episodes in high rather than in low transmission areas. This can be explained by the acquisition of immunity and the dramatic reduction, after 10 years of age, of the risk for peripheral parasitaemia to evolve to a clinical attack.24,25
Malaria transmission intensity and average number of parasite clones.
Multiplicity of parasite clones in a single infection can influence the spread of drug resistance in two ways.35,36 If the genetic basis of resistance involves several mutations in different genes, drug resistance might spread faster because of the higher chances of inbreeding and of transmitting the entire combination of mutated genes to the progeny, where malaria infections are less polyclonal.4,37 On the contrary, the lower polyclonality would not have any effect if resistance is determined by a single mutation or several mutations in the same gene. However, mutant parasites are statistically more likely to occur in polyclonal infections. Indeed, the likelihood of treatment failure increases with increasing mean number of clones,29 the latter being related to increasing transmission18,20,28,29,38 until it reaches a plateau at very intense transmission. Consequently, extremes of transmission are likely to favor faster spread of drug resistance, if its genetic basis involves multiple genes, whereas only high, but not low, transmission would favor faster spread of drug resistance if its genetic basis involves one gene. Control interventions (treatment, ITNs, IRS) by reducing the mean number of clones can also reduce the spread of drug resistance if its genetic basis involves one gene or until a certain critical threshold if the genetic basis involves several genes.27,29 The data from Uganda summarized here for CQ support the theories based on sexual recombination and a multigenic basis for resistance to CQ, implying that resistance for such drugs could spread faster at very low or very high malaria transmission. The data for the relationship between resistance to combination therapy and transmission intensity are very scanty, but those for CQ + SP and AQ + SP from Uganda suggest that faster acquisition of immunity in areas of intense transmission is an additional factor in reducing the spread of drug resistance in areas with intense transmission. The genetic basis for resistance to combination therapy is likely to be modulated by several genes, and resistance is likely to spread in a pattern similar to that observed for CQ.
Transmission intensity and intrahost competition.
In areas of high transmission, intrahost competition between co-infecting parasite clones (intrahost dynamics) is probably an important factor.7 The generalized immunity model of intrahost competition predicts that resistance could spread faster in areas of high transmission.9 If intrahost competition plays an important role, the mutant parasite sub-population with the ability to survive drug treatment replaces the susceptible ones and this process is boosted by the highly polyclonal infections found in areas of intense transmission. The intrahost competition hypothesis is indirectly supported by field data from Uganda, Zimbabwe, and Tanzania on CQ and SP resistance (clinical and molecular markers) because this was high where transmission was intense or where there were no malaria interventions that interrupt transmission (IRS or ITNs).
Transmission intensity, drug resistance, and its genetic basis.
Mathematical models predict that if two or more genes encode resistance, the relationship between malaria transmission and spread of drug resistance exhibits “a valley phenomenon,” whereas if one gene encodes resistance, “a linear increasing function” is expected.7,9 The observations reported in this review show 1) increasing prevalence of SP resistance with increasing transmission intensity or a reduction of both CQ and SP resistance with interventions reducing transmission; 2) high CQ resistance at the extremes of malaria transmission (AEIR < 7 infective bites/person/year, or > 300 infective bites/person/year); 3) jigher CQ + SP and AQ + SP resistance at lower transmission intensities (AEIR < 7 infective bites/person/year). The observed differences can be explained by the genetic basis of drug resistance, CQ resistance being modulated by two or more genes and SP resistance predominantly by one gene,39–45 whereas combination therapy is likely to be modulated by multiple mutations on several genes.
Malaria transmission intensity and parasite biomass.
The emergence of drug resistance is primarily dependent on the de novo emergence of mutations whose probability is a function of the parasite biomass.1 A large parasite biomass is more likely to occur in areas of low transmission because of the lower host immunity.13,14 This is the likely explanation for the origin and rapid rise of multi-drug resistance in Southeast Asia. However, besides the observation that CQ and SP resistance seems to have originated at a few foci in areas of low transmission and spread through selective sweeps across the world,46–48 there is no other empirical evidence to support this theory. Parasite biomass is statistically attractive, but its measurement in field studies is difficult because of the multi-organ sequestration of parasites.
Malaria transmission intensity and impact of drug therapy on gametocyte carriage.
Malaria chemotherapy in Plasmodium malaria largely focuses on clearance of the asexual stages of falciparum in symptomatic patients. However, at a population level, drugs that reduce the carriage of gametocytes (the sexual stage responsible for infection of the mosquito vector) are likely to modulate the prevalence and spread of drug-resistant malaria depending on the coverage of drug use and the level of malaria transmission intensity. After the loss of CQ to resistance, the artemisinin-based combination therapies (ACTs) are presently the only available antimalarials that rapidly reduce both the asexual and sexual stages of P. falciparum. There is sufficient evidence that the ACTs reduce the prevalence and density of microscopically confirmed post-treatment gametocytemia,49,50 and subsequently, the prevalence of mosquito infection.49,51,52 Consequently, it has generally been believed that ACTs are likely to reduce malaria transmission at the population level,50,53 which could lead to a reduction in the transmission of resistant strains.50,53,54 Furthermore, it has been suggested that the probability of a mosquito being infected depends on the prevalence, duration, and density of viable gametocyte carriage in the human host, although additional humoral and leukocyte factors also affect transmissibility.55 The infectivity and transmission potential of a given antimalarial treatment is a function of blood gametocyte density and time, summed over the acute and all subsequent recrudescent infections. Therefore, what drives drug resistance within populations is the ratio of the transmission potential in drug-resistant compared with drug-sensitive infections.55 In view of the above, it has been proposed that a good malaria treatment policy should not only be based on therapeutic efficacy against asexual stages, but it should also consider transmission reduction. However, a theoretical model has recently raised concerns about this general belief that gametocytocidal activity is beneficial because it reduces the rate at which resistance evolves by reducing the transmission of resistant parasites. The model predicts that gametocytocidal activity could instead promote the spread of resistance because it reduces the transmission of drug-sensitive sexual forms to a greater extent than the drug-resistant ones, facilitating the spread of the latter.56 Nevertheless, in terms of magnitude, the model predicts that the increased rate of spread of resistance is relatively small if drug coverage is low or moderate, but the increase in spread could be substantial if drug use is widespread, and ultimately, gametocytocidal activity could be counterproductive by facilitating the rapid evolution of drug resistance.56 Whether this recent model will be supported by empirical data is a subject for study. However, in Southeast Asia, an area of low malaria transmission intensity, ACTs (because of their gametocyte-reducing effect) have resulted in a sustained decrease in malaria transmission and a decrease in the spread of resistance.53
Although none of the studies reviewed here reported gametocyte carriage or density, the high prevalence and spread of resistance to SP and the high frequency of mutations linked to its resistance in areas of intense transmission could be caused by its lack of gametocytocidal activity, leading to a high post-treatment gametocyte carriage and hence high transmission of drug-resistant forms. Although CQ is gametocytocidal, its sub-optimal efficacy and widespread use could have increased the potential for transmission ratio between resistant and sensitive forms, and the latter is likely to be more experienced in areas with intense transmission. The impact of widespread use of gametocyte-reducing drugs in areas of high malaria transmission in Africa, where treatment is often given without parasite-based diagnosis, is a subject for study. To ascertain whether this class of drugs will halt or facilitate the spread of the resistant forms to a greater extent than the sensitive ones requires field-based data. It will be very important to embed studies that assess the impact of wide-scale ACT deployment on the prevalence and spread of drug resistance in the long term, and such studies will be more informative if they collect both microscopically confirmed and sub-microscopic gameto-cytemia.57
Shedding some light on theoretical models.
The available data suggest that several models (intrahost competition, parasite biomass, potential for transmission ratio between resistant and sensitive forms, and sexual recombination) possibly apply, and a more complex pattern should be expected from the field data, depending on the genetic basis for drug resistance. This probably explains why mathematical models have previously generated conflicting predictions. The field evidence is consistent that the prevalence of resistance to CQ and SP monotherapy is invariably higher at intense transmission, but for CQ, it could paradoxically increase at extremely low transmission intensity, whereas data from one combination therapy study suggested that the prevalence of resistance is higher at very low transmission, largely because of low immunity. However, this study was conducted late in the evolution of drug resistance because mutations linked to resistance for the partner drugs in the combinations studied were saturated at all the sites.
Some caveats.
Only six studies (three in Uganda and one each in Burkina Faso, Tanzania, and Zimbabwe) have empirically studied the relationship between transmission intensity and spread of drug resistance, and therefore, the information available is extremely limited. In addition, most studies have used a relatively small number of sites (two, six, seven, and nine). Multiple sites distributed over different areas (possibly in different countries) would have been preferable because they would represent the substantial variation in mean exposure to infection across groups and, at the same time, avoid mixing and gene flow between sites.58
Furthermore, the application of PR in some studies to estimate falciparum malaria transmission intensity has its limitations because multi-country and multi-site studies have observed little variation in PR at very high EIR.59,60 The incidence of malaria infection in infants has at times been used as a proxy of transmission intensity. However, in one such study, the prevalence of infection did not increase when the daily EIR exceeded 1 infective bite/person/night.61 Some studies have used the AEIR, but it also has its limitations as well,62 especially when derived from human landing collections, because the frequency of contact between humans and vectors could vary with the individual collectors. In addition, mosquito sampling is normally conducted in a few households and extrapolated to whole villages or communities, yet parasite exposure may vary considerably in small geographical areas or even households.63,64 Moreover, the sampling from a few households, especially in low transmission areas, results in estimates with large confidence intervals. The AEIR may convey a degree of precision in the estimation of parasite challenge that is non-existent.62 The AEIR has been likened to the input into, whereas the PR likened to the output from, the “black box.” Within the “black box,” there are probably several variables, known and unknown, which modulate transmission. We re-analyzed our previous dataset using a qualitative composite index that takes into consideration both the PR and AEIR instead of the PR alone to rank sites according to malaria transmission intensity, and the pattern observed is consistent with our earlier observations.
Finally, the in vivo data from Uganda in the earlier (1999) study and the later (2004) study used different designs, durations of follow-up, and analyses.23,27 It is generally agreed that a shorter duration of follow-up underestimates treatment failure, especially for drugs with a long elimination half-life such as SP. Although it is important to have a follow-up of at least 28 days to estimate the efficacy of a given treatment and consequently define an appropriate antimalarial drug policy, a shorter follow-up is unlikely to invalidate our conclusions on the relationship between intensity of transmission and drug resistance. A longer follow-up has its own inherent limitations such as the interpretation of PCR results for which uniform standard operating procedures do not exist yet. It may be extremely difficult to interpret PCR results when infections are polyclonal, a common occurrence in areas of high transmission, and this may introduce a bias. The differences in resistance observed between high and low transmission areas for combination therapy in Uganda are likely to be modest, yet they indicate that resistance to combination therapy is higher in low transmission areas. Therefore, any misclassification, if it occurred, only served to diminish the differences and does not affect the conclusions of this review.
Conclusions, implications for malaria control, and future research.
We reviewed the studies that have empirically assessed the relationship between malaria transmission intensity and drug resistance and provided some explanations on the link between malaria transmission intensity or vector control interventions and the prevalence of resistance of P. falciparum to mono- (CQ and SP) or combination therapy (CQ + SP and AQ + SP). Although the number of available studies is limited, and the design of some does not allow final conclusions about causality, the data summarized give important insights.
The epidemiologic observations summarized here are indirect, but they offer field evidence that multiple factors/models (genetic basis of resistance, complexity of infection/sexual recombination, intrahost competition/dynamics, drug elimination half-life) play an important role in the link between malaria transmission intensity and drug resistance. The relationship between malaria transmission and SP resistance is fairly clear, whereas that for CQ and combination therapy resistance seems to be more complex. Drug selection pressure (drug use and drug elimination half-life) is probably the most important factor. It is important that all future studies assessing the relationship between malaria transmission intensity or vector control interventions and drug resistance collect robust data on drug use trends (population-based community drug use and therapeutic drug use) as well. Those that have done so have observed an inverse link between community drug use and malaria transmission intensity, yet the prevalence of in vivo drug resistance or the frequency of mutations linked to CQ or SP resistance did not entirely follow the same pattern. These epidemiologic observations could be explained by acknowledging that, besides modulating drug pressure, malaria transmission intensity modulates other factors and, depending on the genetic basis for resistance, affects the spread of drug resistance in different ways. Indeed, it is apparent that the prevalence resistance to CQ monotherapy could be higher at the extremes of the transmission spectrum and lowest in intermediate transmission areas, whereas the prevalence of SP resistance is likely to increase with increasing malaria transmission intensity. The relationship between the prevalence of resistance to combination therapy and transmission intensity is likely to be as complex as or even more complex than that for CQ, and studies that concurrently measure trends over time for the key mediators and effectors and other indices are urgently needed in non-contiguous areas with varying malaria transmission intensity. Presently, many countries are scaling up vector control interventions (ITNs, IRS, or a combination of both). Their coverage and impact on each other is likely to vary depending on transmission intensity, which offers an opportunity to conduct studies that measure trends within an implementation program that would otherwise be considered unethical. Nevertheless, the few studies available suggest that reducing malaria transmission intensity may be useful in limiting the spread and prevalence of resistance, except at very low transmission for some drugs. A more rational and judicious use of antimalarial drugs in populations exposed to low transmission coupled, in areas of high transmission, with interventions such as ITNs, IRS, or environmental management, is likely to have the additional benefit of slowing down the spread of falciparum-resistant malaria.
Malaria parasite prevalence (PR) (observed and predicted), annual entomologic inoculation rates, CQ and SP use, and CQ and SP treatment failure (RI and RIII), Uganda 1999 (adapted with some modifications from refs. 23, 49, and 50)
| Site | Sites ranked according to malaria transmission* | Observed PR | Predicted PR | AEIR | CQ use | CQ RI | CQ RIII | SP use | SP–RI | SP–RIII |
|---|---|---|---|---|---|---|---|---|---|---|
| * The AEIR and PR were used in a composite qualitative index to rank the sites. At low transmission intensity, the AEIR was inadequate in differentiating sites, and the PR was used. At high transmission, the PR was inadquate in differentiating sites and the AEIR was used. | ||||||||||
| Jinja | 1 (low) | 14.9 | 42.9 | 6 | 81.3 | 0 | 29.2 | 3.9 | 0.0 | 2.9 |
| Kanugu | 2 (low) | 43.3 | 42.7 | 6 | 68.1 | 0 | 4.1 | 4.7 | 0.0 | 0.0 |
| Mubende | 3 (low) | 57.3 | 40.3 | 4 | 57.7 | 13.3 | 3.3 | 1.7 | 23.3 | 2.9 |
| Kyenjojo (Kabarole) | 4 (low–medium) | 67.8 | 45.6 | 7 | 43.0 | 31.3 | 12.5 | 0.5 | 6.7 | 0 |
| Arua | 5 (high) | 80.5 | 87.6 | 397 | 32.6 | 25.5 | 2.0 | 1.7 | 6.0 | 2.0 |
| Tororo | 6 (high) | 90.6 | 91.2 | 562 | 40.5 | 14.8 | 5.5 | 0.4 | 16.7 | 0.0 |
| Apac | 7 (very high) | 79.5 | 100 | 1,586 | 32.0 | 0 | 17.5 | 0.4 | 1.7 | 6.7 |
Malaria transmission intensity and drug use. Adapted from ref. 23. Relationship between the parasite rate (index for malaria transmission) and community CQ and SP use, respectively. A, Community CQ use. B, Community SP use. Regression lines are based on ordinary least squares.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Malaria transmission intensity and drug use. Adapted from ref. 23. Relationship between the parasite rate (index for malaria transmission) and community CQ and SP use, respectively. A, Community CQ use. B, Community SP use. Regression lines are based on ordinary least squares.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Malaria transmission intensity and drug use. Adapted from ref. 23. Relationship between the parasite rate (index for malaria transmission) and community CQ and SP use, respectively. A, Community CQ use. B, Community SP use. Regression lines are based on ordinary least squares.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Variation of CQ and SP treatment failure (TF) with malaria transmission. Data from ref. 23 have been adapted, re-analyzed, and presented as a scatter plot to show the variation of CQ and SP treatment failure according the PR as the index for malaria transmission intensity. The outcome classification for treatment failure (without genotyping) was based on the World Health Organization (WHO, 1996) protocol with 14 days of follow-up.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Variation of CQ and SP treatment failure (TF) with malaria transmission. Data from ref. 23 have been adapted, re-analyzed, and presented as a scatter plot to show the variation of CQ and SP treatment failure according the PR as the index for malaria transmission intensity. The outcome classification for treatment failure (without genotyping) was based on the World Health Organization (WHO, 1996) protocol with 14 days of follow-up.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Variation of CQ and SP treatment failure (TF) with malaria transmission. Data from ref. 23 have been adapted, re-analyzed, and presented as a scatter plot to show the variation of CQ and SP treatment failure according the PR as the index for malaria transmission intensity. The outcome classification for treatment failure (without genotyping) was based on the World Health Organization (WHO, 1996) protocol with 14 days of follow-up.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Inadequacy of AEIR at low transmission intensity and of the PR at intense transmission. A, Sites are ranked according to AEIR. B, Sites are ranked according to the PR.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Inadequacy of AEIR at low transmission intensity and of the PR at intense transmission. A, Sites are ranked according to AEIR. B, Sites are ranked according to the PR.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Inadequacy of AEIR at low transmission intensity and of the PR at intense transmission. A, Sites are ranked according to AEIR. B, Sites are ranked according to the PR.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Parasitologic failure (RI and RIII) for CQ according to transmission intensity. Data from ref. 23 have been adapted and re-analyzed for total treatment failure (TTF), and RI and RIII as measures for partial and full resistance as proposed by some theoretical models.16,17 The parasitologic and clinical failure classifications are based on the World Health Organization (WHO, 1965 and 1996) protocols and 14 days of follow-up. The sites are ordered using a qualitative index that takes into consideration both the PR and AEIR to differentiate malaria transmission intensity at different sites. Trend lines are based on moving averages.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Parasitologic failure (RI and RIII) for CQ according to transmission intensity. Data from ref. 23 have been adapted and re-analyzed for total treatment failure (TTF), and RI and RIII as measures for partial and full resistance as proposed by some theoretical models.16,17 The parasitologic and clinical failure classifications are based on the World Health Organization (WHO, 1965 and 1996) protocols and 14 days of follow-up. The sites are ordered using a qualitative index that takes into consideration both the PR and AEIR to differentiate malaria transmission intensity at different sites. Trend lines are based on moving averages.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Parasitologic failure (RI and RIII) for CQ according to transmission intensity. Data from ref. 23 have been adapted and re-analyzed for total treatment failure (TTF), and RI and RIII as measures for partial and full resistance as proposed by some theoretical models.16,17 The parasitologic and clinical failure classifications are based on the World Health Organization (WHO, 1965 and 1996) protocols and 14 days of follow-up. The sites are ordered using a qualitative index that takes into consideration both the PR and AEIR to differentiate malaria transmission intensity at different sites. Trend lines are based on moving averages.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Parasitologic failure (RI and RIII) for SP according to transmission intensity. Data from ref. 23 have been adapted and re-analyzed for TTF. RI and RIII are proxies for partial and full resistance. The parasitologic and clinical failure classifications are based on the World Health Organization (WHO, 1965 and 1996) protocols and 14 days of follow-up. Ranking of sites takes into consideration the PR and the AEIR. Trend lines are based on moving averages.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Parasitologic failure (RI and RIII) for SP according to transmission intensity. Data from ref. 23 have been adapted and re-analyzed for TTF. RI and RIII are proxies for partial and full resistance. The parasitologic and clinical failure classifications are based on the World Health Organization (WHO, 1965 and 1996) protocols and 14 days of follow-up. Ranking of sites takes into consideration the PR and the AEIR. Trend lines are based on moving averages.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Parasitologic failure (RI and RIII) for SP according to transmission intensity. Data from ref. 23 have been adapted and re-analyzed for TTF. RI and RIII are proxies for partial and full resistance. The parasitologic and clinical failure classifications are based on the World Health Organization (WHO, 1965 and 1996) protocols and 14 days of follow-up. Ranking of sites takes into consideration the PR and the AEIR. Trend lines are based on moving averages.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Malaria transmission intensity and risk of treatment failure to combination therapy. Adapted from ref. 27. The relationship between malaria transmission intensity and risk of treatment failure (adjusted for age and complexity of infections) for (A) CQ + SP and (B) AQ + SP. The clinical failure classifications are based on the World Health Organization (WHO, 2003) protocol and 28 days of follow-up; the data were analyzed using a Cox proportional hazards model controlling for age and complexity of infection. The sites have been re-ordered according to an increasing malaria transmission intensity based on both the PR and AEIR.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Malaria transmission intensity and risk of treatment failure to combination therapy. Adapted from ref. 27. The relationship between malaria transmission intensity and risk of treatment failure (adjusted for age and complexity of infections) for (A) CQ + SP and (B) AQ + SP. The clinical failure classifications are based on the World Health Organization (WHO, 2003) protocol and 28 days of follow-up; the data were analyzed using a Cox proportional hazards model controlling for age and complexity of infection. The sites have been re-ordered according to an increasing malaria transmission intensity based on both the PR and AEIR.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Malaria transmission intensity and risk of treatment failure to combination therapy. Adapted from ref. 27. The relationship between malaria transmission intensity and risk of treatment failure (adjusted for age and complexity of infections) for (A) CQ + SP and (B) AQ + SP. The clinical failure classifications are based on the World Health Organization (WHO, 2003) protocol and 28 days of follow-up; the data were analyzed using a Cox proportional hazards model controlling for age and complexity of infection. The sites have been re-ordered according to an increasing malaria transmission intensity based on both the PR and AEIR.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Malaria transmission intensity and the pfcrt K76T or dhfr C59R allele frequency Adapted from ref. 28; the PR is the index for malaria transmission intensity; the allele frequency is derived by maximum likelihood estimation. A, pfcrt K76T allele frequency with the 95% confidence limits for both PR and the pfcrt 76T allele frequency. B, dhfr C59R allele frequency, with a superimposed regression line based on ordinary least square.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Malaria transmission intensity and the pfcrt K76T or dhfr C59R allele frequency Adapted from ref. 28; the PR is the index for malaria transmission intensity; the allele frequency is derived by maximum likelihood estimation. A, pfcrt K76T allele frequency with the 95% confidence limits for both PR and the pfcrt 76T allele frequency. B, dhfr C59R allele frequency, with a superimposed regression line based on ordinary least square.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Malaria transmission intensity and the pfcrt K76T or dhfr C59R allele frequency Adapted from ref. 28; the PR is the index for malaria transmission intensity; the allele frequency is derived by maximum likelihood estimation. A, pfcrt K76T allele frequency with the 95% confidence limits for both PR and the pfcrt 76T allele frequency. B, dhfr C59R allele frequency, with a superimposed regression line based on ordinary least square.
Citation: The American Journal of Tropical Medicine and Hygiene 77, 6_Suppl
Address correspondence to Ambrose O. Talisuna, Ministry of Health, Epidemiological Surveillance Division, PO Box 7272, Kampala, Uganda. E-mail atalisuna@afsat.com
Authors’ addresses: Ambrose O. Talisuna, Ministry of Health, Epidemiological Surveillance Division and East African Network for Monitoring Antimalarial Treatment (EANMAT). Paul Okello, PO Box 33861, Kampala, Uganda. Annette Erhart, Marc Coosemans, and Umberto D’Alessandro, Department of Parasitology, Prince Leopold Institute of Tropical Medicine, Nationalestraat 155, B-2000 Antwerp, Belgium, Telephone: 32-3-247-63-11, Fax: 32-3-247-63-59.
Acknowledgments: We are grateful to Dr J. G. Breman for stimulating the idea for this review and to the very extensive comments from two anonymous reviewers.
Financial support: The authors received no specific funding for this article. However, Ambrose O. Talisuna is supported by the Medicines for Malaria Venture (MMV) in collaboration with the Institute of Tropical Medicine, Antwerp, Belgium.
REFERENCES
- 1↑
White NJ, Pongtavornpinyo W, 2003. The de novo selection of drug resistant malaria parasites. Proc R Soc Lond B Biol Sci 270 :545–554.
- 2↑
Hastings IM, Watkins WM, White NJ, 2002. The evolution of drug-resistant malaria: the role of drug elimination half-life. Philos Trans R Soc Lond B Biol Sci 357 :505–519.
- 4↑
Paul REL, Packer MJ, Walmsley M, Lagog M, Randford-Cartwright LC, Paru R, Day KP, 1995. Mating patterns in malaria parasite populations of Papua New Guinea. Science 269 :1709–1711.
- 5↑
Cross AP, Singer B, 1991. Modelling the development of resistance of Plasmodium falciparum to anti-malarial drugs. Trans R Soc Trop Med Hyg 85 :349–355.
- 6
Curtis CF, Otoo LN, 1986. A simple model of the build-up of resistance to mixtures of anti-malarial drugs. Trans R Soc Trop Med Hyg 80 :889–892.
- 7↑
Hastings IM, 1997. A model for the origins and spread of drug-resistant malaria. Parasitology 115 :133–141.
- 8
Hastings IM, Mackinnon MJ, 1998. The emergence of drug-resistant malaria. Parasitology 117 :411–417.
- 9↑
Hastings IM, D’Alessandro U, 2000. Modelling a predictable disaster: the rise and spread of drug-resistant malaria. Parasitol Today 16 :340–347.
- 10
Hastings IM, 2001. Modelling parasite drug resistance: lessons for management and control strategies. Trop Med Int Health 6 :883–890.
- 11
Molyneux DH, Floyd K, Barnish G, Fevre EM, 1999. Transmission control and drug resistance in malaria: a crucial interaction. Parasitol Today 15 :238–240.
- 13↑
White N, 1999. Antimalarial drug resistance and combination chemotherapy. Philos Trans R Soc Lond B Biol Sci 354 :739–749.
- 14↑
White NJ, 1999. Delaying antimalarial drug resistance with combination chemotherapy. Parassitologia 41 :301–308.
- 15↑
Mackinnon MJ, Hastings IM, 1998. The evolution of multiple drug resistance in malaria parasites. Trans R Soc Trop Med Hyg 92 :188–195.
- 16↑
Prudhomme-O’Meara W, Smith DL, McKenzie FE, 2006. Potential impact of intermittent preventive treatment (IPT) on spread of drug-resistant malaria. PLoS Med. 3 :e141.
- 17↑
Hastings IM, Watkins WM, 2005. Intensity of malaria transmission and the evolution of drug resistance. Acta Trop 94 :218–229.
- 18↑
Arnot D, 1998. Unstable malaria in Sudan: the influence of the dry season. Clone multiplicity of Plasmodium falciparum infections in individuals exposed to variable levels of disease transmission. Trans R Soc Trop Med Hyg 92 :580–585.
- 19
Farnert A, Rooth I, Svensson A, Snounou G, Bjorkman A, 1999. Complexity of Plasmodium falciparum infections is consistent over time and protects against clinical disease in Tanzanian children. J Infect Dis 179 :989–995.
- 20↑
Paganotti GM, Babiker HA, Modiano D, Sirima BS, Verra F, Konate A, Ouedraogo AL, Diarra A, Mackinnon MJ, Coluzzi M, Walliker D, 2004. Genetic complexity of Plasmodium falciparum in two ethnic groups of Burkina Faso with marked differences in susceptibility to malaria. Am J Trop Med Hyg 71 :173–178.
- 21↑
Drakeley CJ, Corran PH, Coleman PG, Tongren JE, McDonald SL, Carneiro I, Malima R, Lusingu J, Manjurano A, Nkya WM, Lemnge MM, Cox J, Reyburn H, Riley EM, 2005. Estimating medium- and long-term trends in malaria transmission by using serological markers of malaria exposure. Proc Natl Acad Sci USA 102 :5108–5113.
- 22↑
Smith T, Charlwood JD, Kitua AY, Masanja H, Mwankusye S, Alonso PL, Tanner M, 1998. Relationships of malaria morbidity with exposure to Plasmodium falciparum in young children in a highly endemic area. Am J Trop Med Hyg 59 :252–257.
- 23↑
Talisuna OA, Langi P, Bakyaita N, Egwang T, Mutabingwa TK, Watkins W, Van Marck E, D’Alessandro U, 2002. Intensity of malaria transmission, antimalarial drug use and resistance in Uganda: what is the relationship between these three factors? Trans R Soc Trop Med Hyg 96 :310–317.
- 24↑
Rogier C, Diagne N, Fontenille D, Spiegel A, Trape JF, 1999. Plasmodium falciparum clinical malaria: lessons from longitudinal studies in Senegal. Parasitologia 41 :255–259.
- 25↑
Trape JF, Rogier C, 1996. Combating malaria morbidity and mortality by reducing transmission. Parasitol Today 12 :236–240.
- 26↑
World Health Organization, 1965. Resistance of Malaria Parasites to Drugs. Geneva: World Health Organization.
- 27↑
Francis D, Nsobya SL, Talisuna A, Yeka A, Kamya MR, Machekano R, Dokomajilar C, Rosenthal PJ, Dorsey G, 2006. Geographic differences in antimalarial drug efficacy in Uganda are explained by differences in endemicity and not by known molecular markers of drug resistance. J Infect Dis 193 :978–986.
- 28↑
Talisuna AO, Langi P, Mutabingwa TK, Van Marck E, Speybroeck N, Egwang TG, Watkins WM, Hastings IM, D’Alessandro U, 2003. Intensity of transmission and spread of gene mutations linked to chloroquine and sulphadoxine-pyrimethamine resistance in falciparum malaria. Int J Parasitol 33 :1051–1058.
- 29↑
Lee SA, Yeka A, Nsobya SL, Dokomajilar C, Rosenthal PJ, Talisuna A, Dorsey G, 2006. Complexity of Plasmodium falciparum infections and antimalarial drug efficacy at 7 sites in Uganda. J Infect Dis 193 :1160–1163.
- 30↑
Alifrangis M, Lemnge MM, Ronn AM, Segeja MD, Magesa SM, Khalil IF, Bygbjerg IC, 2003. Increasing prevalence of wild-types in the dihydrofolate reductase gene of Plasmodium falciparum in an area with high levels of sulfadoxine/pyrimethamine resistance after introduction of treated bed nets. Am J Trop Med Hyg 69 :238–243.
- 31↑
Mharakurwa S, Mutambu SL, Mudyiradima R, Chimbadzwa T, Chandiwana SK, Day KP, 2004. Association of house spraying with suppressed levels of drug resistance in Zimbabwe. Malar J 3 :35.
- 32↑
Diallo D, Sutherland C, Nebié I, Konaté A, Ord R, Pota H, Roper C, Ilboudo-Sanogo C, Greenwood B, Cousens S, 2007. Sustained use of insecticide treated curtains is not associated with greater circulation of drug resistant malaria parasites nor with higher risk of treatment failure among children with uncomplicated malaria in Burkina Faso. Am J Trop Med Hyg 76 :237–244.
- 33↑
Reyburn H, Ruanda J, Mwerinde O, Drakeley C, 2006. The contribution of microscopy to targeting antimalarial treatment in a low transmission area of Tanzania. Malar J 5 :4.
- 34↑
Watkins WM, Mosobo M, 1993. Treatment of Plasmodium falciparum malaria with pyrimethamine-sulfadoxine: selective pressure for resistance is a function of long elimination half-life. Trans R Soc Trop Med Hyg 87 :75–78.
- 35↑
Babiker A, Walliker D, 1997. Current views on the population structure of Plasmodium falciparum: implications for control. Parasitol Today 13 :262–267.
- 36↑
Walliker D, Quakyi IC, Wellems TE, McCutchan TF, Szarfman A, 1987. London WT, Corcoran LM, Burkot TR, Carter R. Genetic Analysis of the Human Malaria Parasite Plasmodium falciparum. Science 236 :1661–1666.
- 38↑
Slater M, Kiggundu M, Dokomajilar C, Kamya MR, Bakyaita N, Talisuna A, Rosenthal PJ, Dorsey G, 2005. Distinguishing recrudescences from new infections in antimalarial clinical trials: major impact of interpretation of genotyping results on estimates of drug efficacy. Am J Trop Med Hyg 73 :256–262.
- 39↑
Foote SJ, Kyle DE, Martin RK, Oduola AM, Forsyth K, Kemp DJ, Cowman AF, 1990. Several alleles of the multidrug-resistance gene are closely linked to chloroquine resistance in Plasmodium falciparum. Nature 345 :255–258.
- 40
Wellems TE, Panton LJ, Gluzman IY, Do RV, Gwadz RW, Walker-Jonah A, Krogstad DJ, 1990. Chloroquine resistance not linked to mdr-like genes in a Plasmodium falciparum cross. Nature 345 :253–255.
- 41
Djimdé A, Doumbo OK, Cortese JF, Kayentao K, Doumbo S, Diourte Y, Dicko A, Su XZ, Nomura T, Fidock DA, Wellems TE, Plowe CV, Coulibaly D, 2001. A molecular marker for chloroquine-resistant falciparum malaria. N Engl J Med 344 :257–263.
- 42
Fidock DA, Nomura T, Talley AK, Cooper RA, Dzekunov SM, Ferdig MT, Ursos LM, Sidhu AB, Naude B, Deitsch KW, Su XZ, Wootton JC, Roepe PD, Wellems TE, 2000. Mutations in the P. falciparum digestive vacuole transmembrane protein PfCRT and evidence for their role in chloroquine resistance. Mol Cell 6 :861–871.
- 43
Sidhu AB, Verdier-Pinard D, Fidock DA, 2002. Chloroquine resistance in Plasmodium falciparum malaria parasites conferred by pfcrt mutations. Science 298 :210–213.
- 44
Reed MB, Saliba KJ, Caruana SR, Kirk K, Cowman AF, 2000. Pgh1 modulates sensitivity and resistance to multiple antimalarials in Plasmodium falciparum. Nature 403 :906–909.
- 45↑
Sibley CH, Hyde JE, Sims PF, Plowe CV, Kublin JG, Mberu EK, Cowman AF, Winstanley PA, Watkins WM, Nzila AM, 2001. Pyrimethamine-sulfadoxine resistance in Plasmodium falciparum: what next? Trends Parasitol 17 :582–588.
- 46↑
Anderson TJ, Roper C, 2005. The origins and spread of antimalarial drug resistance: lessons for policy makers. Acta Trop 94 :269–280.
- 47
Roper C, Pearce R, Bredenkamp B, Gumede J, Drakeley C, Mosha F, Chandramohan D, Sharp B, 2003. Antifolate antimalarial resistance in southeast Africa: a population-based analysis. Lancet 361 :1174–1181.
- 48↑
Wootton JC, Feng X, Ferdig MT, Cooper RA, Mu J, Baruch DI, Magill AJ, Su XZ, 2002. Genetic diversity and chloroquine selective sweeps in Plasmodium falciparum. Nature 418 :320–323.
- 49↑
Targett G, Drakeley C, Jawara M, von Seidlein L, Coleman R, Deen J, Pinder M, Doherty T, Sutherland C, Walraven G, Milligan P, 2001. Artesunate reduces but does not prevent posttreatment transmission of Plasmodium falciparum to Anopheles gambiae. J Infect Dis 83 :1254–1259.
- 50↑
Price RN, Nosten F, Luxemburger C, ter Kuile FO, Paiphun L, Chongsuphajaisiddhi T, White NJ, 1996. Effects of artemisinin derivatives on malaria transmissibility. Lancet 347 :1654–1658.
- 51↑
Drakeley CJ, Jawara M, Targett GA, Walraven G, Obisike U, Coleman R, Pinder M, Sutherland CJ, 2004. Addition of artesunate to chloroquine for treatment of Plasmodium falciparum malaria in Gambian children causes a significant but short-lived reduction in infectiousness for mosquitoes. Trop Med Int Health 9 :53–61.
- 52↑
Sutherland CJ, Ord R, Dunyo S, Jawara M, Drakeley CJ, Alexander N, Coleman R, Pinder M, Walraven G, Targett GA, 2005. Reduction of malaria transmission to Anopheles mosquitoes with a six-dose regimen of co-artemether. PLoS Med 2 :e92.
- 53↑
Nosten F, van Vugt M, Price R, Luxemburger C, Thway KL, Brockman A, McGready R, ter Kuile F, Looareesuwan S, White NJ, 2000. Effects of artesunate-mefloquine combination on incidence of Plasmodium falciparum malaria and mefloquine resistance in western Thailand: a prospective study. Lancet 356 :297–302.
- 54↑
Hallett RL, Sutherland CJ, Alexander N, Ord R, Jawara M, Drakeley CJ, Pinder M, Walraven G, Targett GA, Alloueche A, 2004. Combination therapy counteracts the enhanced transmission of drug-resistant malaria parasites to mosquitoes. Antimicrob Agents Chemothe 48 :3940–3943.
- 55↑
Barnes KI, White NJ, 2005. Population biology and antimalarial resistance: The transmission of antimalarial drug resistance in Plasmodium falciparum. Acta Trop 94 :230–240.
- 56↑
Hastings IM, 2006. Gametocytocidal activity in antimalarial drugs speeds the spread of drug resistance. Trop Med Int Health 11 :1206–1217.
- 57↑
Bousema JT, Schneider P, Gouagna LC, Drakeley CJ, Tostmann A, Houben R, Githure JI, Ord R, Sutherland CJ, Omar SA, Sauerwein RW, 2006. Moderate effect of artemisinin-based combination therapy on transmission of Plasmodium falciparum. J Infect Dis 193 :1151–1159.
- 59↑
Beier JC, Killeen GF, Githure JI, 1999. Short report: entomologic inoculation rates and Plasmodium falciparum malaria prevalence in Africa. Am J Trop Med Hyg 61 :109–113.
- 60↑
Okello P, Van Bortel W, Correwyn A, Roelants P, Talisuna AO, D’Alessandro U, Coosemans M, 2006. Variation in malaria transmission intensity in seven sites throughout Uganda. Am J Trop Med Hyg 75 :219–225.
- 61↑
Charlwood JD, Smith T, Lyimo E, Kitua AY, Masanja H, Booth M, Alonso PL, Tanner M, 1998. Incidence of Plasmodium falciparum infection in infants in relation to exposure to sporozoite-infected anophelines. Am J Trop Med Hyg 59 :243–251.
- 62↑
Snow RW, Marsh K, 2002. The consequences of reducing transmission of Plasmodium falciparum in Africa. Adv Parasitol 52 :235–264.
- 63↑
Cattanni JA, Moir JS, Gibson FD, Ginny M, Paino J, Davidson W, Alpers MP, 1985. Small area variations in the epidemiology of malaria in Madang Province. P N G Med J 29 :11–17.
- 64↑
Mbogo CN, Snow RW, Khamala CP, Kabiru EW, Ouma JH, Githure JI, Marsh K, Beier JC, 1995. Relationships between Plasmodium falciparum transmission by vector populations and the incidence of severe disease at nine sites on the Kenyan coast. Am J Trop Med Hyg 52 :201–206.