In an increasing number of countries where malaria is endemic, the effectiveness of routinely administered antimalarials like chloroquine and sulfadoxine/pyrimethamine is severely reduced because of the multi-drug resistance of Plasmodium falciparum. This poses a serious problem in terms of treatment and prophylaxis. Yet, mainly for financial reasons, chloroquine remains the first-line drug for treatment of malaria in these countries.

On the other hand, travelers from Europe and other industrialised countries, who have contracted Falciparum malaria, are commonly treated, on their return home, with highly effective antimalarials (usually mefloquine, halofantrine or quinine) that are frequently not freely available for routine treatment in endemic areas. Therefore, drug resistance is as yet not a frequent problem in the treatment of Falciparum malaria in hospitals of industrialized countries. Infected European travelers and immigrants carry a wide variety of P. falciparum strains from all endemic areas. Data and parasite material gained from this population, if properly used, can help to predict the development of drug resistance in endemic areas.

Point mutations at several codons of Plasmodium falciparum genes have been associated with emergence of drug resistance to commonly used antimalarial drugs. In particular, resistance to antifolates (e.g. pyrimethamine/sulfadoxine) and chloroquine depends on point mutations at the Dihydrofolate Reductase (DHFR) and Dihydropteroate Synthetase (DHPS) or Pfcrt and Pfmdr1 genes, respectively 123 . Several protocols for the detection of relevant mutations have been developed and have been evaluated with samples from endemic areas. Although a clear correlation between parasite genotype and in vivo outcome could not be documented in all field trials, the results show convincingly that a selected parasite population, if properly used, can help to correlations between point mutations and phenotypic resistance distinguish 456789 . In order to move molecular assays for point mutations on resistance-related genes into the realm of applied tools for surveillance, we investigated a selection of P. falciparum isolates that were imported during the year 2001 into Europe to study the prevalence of resistance-associated point mutations at relevant codons. In particular, we tested isolates which were developing resistance to antifolates and chloroquine. The screening results were used for mapping prevalence of mutations and, thus, levels of potential drug resistance in endemic areas.

The study was established within the infrastructure of the European Network on Imported Infectious Disease Surveillance (TropNetEurop) which has been successfully providing surveillance data on imported malaria since 1999 10 . The network covers approximately 12% of all imported malaria cases in Western and Central Europe. TropNetEurop is designed to effectively detect emerging infections of potential regional, national or global impact at their point of entry into the domestic population. Sentinel Surveillance reporting is carried out by participating sites using a standardized and computerized reporting system. Immediate transmission of anonymized patient and laboratory data to the central database assures timely detection of sentinel events. The comprehensive collection of data on notifiable and non-notifiable infectious diseases in travelers makes it possible to identify needs for further surveillance and investigation and provides the potential for future case-control studies by identification of specific risk factors. The primary objectives of TropNetEurop are a) to construct and maintain a collaborative research network of clinical sites in Europe dealing with imported infectious diseases and b) to establish and maintain a clinical network for effective sentinel surveillance of imported infectious diseases in Europe. Membership is self-selected by participating centers and monitored by the steering committee of the network. Although the organization of the network does not guarantee a representative data collection for Europe, most referral centers in Europe are represented. From the beginning, malaria has been one of the major targets within this network of 38 clinical sites throughout 15 European countries.

Detection of resistance-related point mutation followed established protocols. During standard malaria testing by thick and thin blood film, 10 μl of full blood from each patient were dotted on Whatman 3 MM^® chromatography paper and air-dried at room temperature before initiation of treatment. DNA was prepared from the dried blood spots as previously described 11 . For detection of the single base change at codon 86 of Pfmdr1, a 330-basepair DNA fragment was amplified and followed by restriction-fragment-length-polymorphism (RFLP)-analysis 1213 . A nested PCR protocol was used to identify the K76T in the Pfcrt gene 814 . After amplification of a 145 bp fragment around the codon 76, alleles carrying the K76 or T76 codon were discriminated by ApoI-restriction. For the detection of polymorphisms on the DHFR and DHPS genes, a previously described nested PCR method was used for all samples 615 . A volume of 4 μl of PCR product was incubated with mutation specific restriction enzymes according to the manufacturer's instructions (New England Biolabs, Beverly, MA, USA).

Altogether, 337 samples were screened (table 1). The majority (n = 313; 92.9%) of these were imported from sub-Saharan Africa. Out of the 24 samples (7.1%) from other endemic areas, 14 were from Indonesia. As the small numbers of isolates from Asia and the Americas were only of limited value for the determination of prevalence rates of point mutations, we omitted these areas from further evaluation. The total distribution of mutation-related point mutations among all isolates is shown in table 2. The mutation on Pfcrt76_Lys/Thr has been identified as potentially crucial for developing resistance, while Pfmdr86_Asn/Tyr appears to play a supporting role 16 . The majority of samples from sub-Saharan Africa showed at least former mutation (figure 1). However, a surprisingly high proportion of samples from areas that are generally assumed to have a very high level of drug resistance showed none of the mutations 17 . In particular, 40.2% of all samples from West Africa lacked the crucial prerequisite to develop phenotypical resistance to chloroquine (figure 1). A combination of both mutations was only present in 37.9% of the tested samples (data not shown). This pattern became even clearer when testing for mutations that are associated with resistance to antifolates. Here, DHFR108_Ser/Asp has been identified as crucial mutation for resistance to group I antifolates, while DHFR51_Asp/Iso and DHFR59_Cys/Arg have a supporting function in enhancing the effect 1 . In similar fashion, DHPS437_Ala/Gly is viewed as prerequisite for resistance to group II antifolates. Thus, P. falciparum has to gain mutations on both genes in order to develop resistance to combined antifolates, e.g. sulfadoxine/pyrimethamine. In the random population of isolates gained from returning European travelers, the distribution of the relevant mutations across sub-Saharan Africa was comparatively low (figures 2 and 3). The combination of triple mutations on DHFR (108_Ser/Asp, 51_Asp/Iso and 59_Cys/Ar) was found in 39.2% of samples in West Africa, 30.8% in Central, 42.9% in East, and 33.3% in South Africa. The combination of DHPS437_Ala/Gly and DHPS 540_Lys/Glu, which is viewed as a necessary prerequisite for resistance to sulfonamides 9, was prevalent in 3.4% of the samples imported from West Africa, and 5.7% from East Africa, while no sample from either Central Africa, or from South Africa showed the combination. Thus, antifolates may still be effective in a comparatively high percentage of isolates from these areas.

This type of molecular surveillance has very little effect on treatment decisions for the individual travelers returning with Falciparum malaria from endemic regions. However, when used within a large clinical network, this method has an unsurpassed advantage. It has become possible to screen large numbers of isolates from malarious regions that are collected randomly from travelers and are transported back to Europe. This study shows that the continuous mapping of the patterns of resistance to crucial antimalarials can be performed. Data gained by fast and efficient molecular methods can be used as an early warning system for changes occurring in endemic areas, thus providing additional information that may be crucial for regional and international drug policy changes.