Plasmodium falciparum mutations pfcrt K76T and the dhfr/dhps “quintuple mutant” are molecular markers of resistance to chloroquine and sulfadoxine-pyrimethamine, respectively. During an epidemic of P. falciparum malaria in an area of political unrest in northern Mali, where standard efficacy studies have been impossible, we measured the prevalence of these markers in a cross-sectional survey. In 80% of cases of infection, pfcrt K76T was detected, but none of the cases carried the dhfr/dhps quintuple mutant. On the basis of these results, chloroquine was replaced by sulfadoxine-pyrimethamine in control efforts. This example illustrates how molecular markers for drug resistance can provide timely data that inform malaria-control policy during epidemics and other emergency situations.
Standard in vitro and in vivo studies of resistance to antimalarial drugs are time consuming and often difficult to interpret in endemic settings where transmission is ongoing . Furthermore, these methods are not feasible in emergency conditions, such as war zones, and for displaced populations . The pfcrt K76T mutation is a molecular marker of resistance to chloroquine . Age-adjusted ratios of the prevalence of this marker to the prevalence of in vivo resistance to chloroquine have been described as a practical means of using molecular markers to estimate levels of resistance to chloroquine (i.e., the genotype-resistance index [GRI]) . The prevalence of the Plasmodium falciparum genotype with the “quintuple mutant” is strongly associated with in vivo resistance to sulfadoxinepyrimethamine . This quintuple mutant includes the S108N, N51I, and C59R mutations in the dihydrofolate reductase gene (dhfr) and the A437G and K540E mutations in the dihydropteroate synthase gene (dhps). We report the first use of molecular markers of drug-resistant malaria in the management of a malaria epidemic.
Methods. The study was conducted in October 1999 in Kidal, a district of 42,500 people in the Sahara Desert in northern Mali. The inhabitants are nomadic Touareg and Arab cattle breeders. Malaria transmission is unstable and occurs near oases or after heavy rainfall. The region was emerging from a 10- year rebellion that resulted in a disruption of all government services, including health care. Although the average annual rainfall is 100 mm, Kidal received 280 mm of rain in August 1999. During the following September, a sharp increase in the number of cases of clinically suspected malaria (fever without any other apparent cause) was reported to the Malian Ministry of Health in Bamako (table 1).
Because land travel was not safe, staff from the Malian National Malaria Control Program (Bamako) and scientists from the Malaria Research and Training Center (Bamako) were flown to Kidal in a military plane. From 11 to 18 October 1999, epidemiological and entomological surveys were done to assess the epidemic, which was confirmed by microscopy to have been caused by P. falciparum. Stocks of chloroquine and sulfadoxinepyrimethamine in the public sector and in private pharmacies were measured. The lack of security prevented the use of conventional methods of measuring resistance to antimalarial drugs. Under military escort, a cross-sectional survey was conducted in 5 nomadic camps. Eighty-seven suspected cases of malaria were detected on the basis of clinical diagnoses. Thick and thin blood smears were obtained for microscopy, and blood samples from finger pricks were spotted onto filter-paper strips for molecular analysis. The prevalences of molecular markers for resistance to chloroquine and sulfadoxine-pyrimethamine were determined as described elsewhere [1, 3, 4]. Informed consent was obtained from patients or their parents or guardians.
Results. The inventory of drugs yielded 194,600 tablets of chloroquine, 1748 bottles of chloroquine syrup, and 4254 ampoules of injectable quinine but no sulfadoxine-pyrimethamine. P. falciparum infection was detected in 23 (27%) of 86 patients by microscopy (1 slide was lost) and in 57 (66%) of 87 patients by polymerase chain reaction (PCR) analysis. A week after the team returned to Bamako, initial molecular analysis of the filter papers was completed at the Malaria Research and Training Center. The wild-type K76 and mutant 76T forms of pfcrt were present in samples from 11 (19%) and 45 (79%) of the 57 infected patients, respectively; both the wild-type and mutant forms of pfcrt were detected in 1 infected patient (figure 1). According to the GRI model, these results would correspond to a rate of 27%–40% for in vivo resistance to chloroquine . Because the prevalence of the 76T form of pfcrt was higher than expected, results of the initial analysis were confirmed at the University of Maryland (Baltimore).
Only samples in which all 3 dhfr and both dhps mutations were successfully identified were included in the analysis. Amplification of all 5 of these markers was accomplished for 22 of the 57 samples that had PCR results positive for ⩾1 of the molecular markers. The inability to characterize all 5 mutations in some samples was attributed to very low parasite densities in PCR-positive, microscopy-negative samples. Among the 22 samples, the prevalences of the dhfr mutations S108N, C59R, and N51I were 45%, 35%, and 30%, respectively, and all 3 dhfr mutations were detected in 20% of the samples. For 1 patient, the 3 dhfr mutations and the dhps A437G mutation were detected. The dhfr/dhps quintuple mutant was not detected in any of the samples, suggesting that sulfadoxine-pyrimethamine would be effective in this setting.
By November 1999, the National Malaria Control Program was advised to provide sulfadoxine-pyrimethamine to health workers in Kidal and to request that sulfadoxine-pyrimethamine, rather than chloroquine, be used to treat cases of malaria in this area. In December 1999, the National Malaria Control Program reacted by shipping sulfadoxine-pyrimethamine to Kidal and by sending a technical note to local health authorities requesting that sulfadoxine-pyrimethamine be used to treat suspected cases of malaria in that health district.
Discussion. Molecular markers of resistance to chloroin areas of higher transmission . Although a set of dhfr/dhps mutations that does not include the full quintuple mutant may be sufficient to cause failure of sulfadoxine-pyrimethamine treatment, the essential role of dhps mutations in sulfadoxinepyrimethamine failure has been documented in a variety of settings in which malaria is transmitted. In a prospective study of in vivo sulfadoxine-pyrimethamine efficacy conducted in a mesoendemic, periurban setting in Mali, we documented prevalences of 5.4%, 12.3%, and 15.5% for dhfr mutations N51I, C59R, and S108N, respectively, but rates of only 1.1% for in vivo sulfadoxine-pyrimethamine parasitological failure and 0% for clinical failure . These results and those of other reports [9, 10] suggest that, in areas of unstable malaria transmission, the presence of dhps mutations in addition to dhfr mutations appears to be necessary to cause in vivo sulfadoxine-pyrimethamine failure. Even if the dhfr triple mutant was sufficient to cause sulfadoxine-pyrimethamine failure in the setting described in this report, its prevalence of 20% would predict a rate for sulfadoxine-pyrimethamine efficacy of <90% . Taken together, the data suggested that sulfadoxine-pyrimethamine, which is a second-line therapy in Mali, would be more effective than chloroquine in this setting. Artemisinin-based combination therapy (ACT) would have been likely to be even more effective, but recommendation of ACT was not feasible, given the lack of any such drugs in stock at the National Malaria Control Program at that time and the urgent need for intervention.
Assays for molecular markers of resistance to chloroquine and sulfadoxine-pyrimethamine are readily performed in many developing countries where malaria is endemic. The collection of blood-spotted filter papers as a source of parasite DNA is straightforward, and cross-sectional surveys can be completed in 1 day. As a result of intensive international efforts in building research capacity, an increasing number of countries in sub- Saharan Africa can now readily perform PCR assays for the detection of parasite gene polymorphisms. These methods may be useful not only in epidemics, when drug-resistance data are urgently needed, but also in war zones and refugee camps, where logistical or security concerns preclude other means of measuring resistance. Rapid surveys for molecular markers of drug resistance may also allow rapid estimations of drug resistance in prospective sites for refugee camps, as well as in areas being prepared for deployment of malaria-naive troops.
We thank the Malian National Malaria Control Program (Bamako), the Ministry of Defense of Mali (Bamako), and the health workers of the Kidal district in Mali.