Malaria transmission in Dakar: A two-year survey

Background According to entomological studies conducted over the past 30 years, there was low malaria transmission in suburb of Dakar but little evidence of it in the downtown area. However; there was some evidence of local transmission based on reports of malaria among permanent residents. An entomological evaluation of malaria transmission was conducted from May 2005 to October 2006 in two areas of Dakar. Methods Mosquitoes were sampled by human landing collection during 34 nights in seven places in Bel-air area (238 person-nights) and during 24 nights in five places in Ouakam area (120 person-nights). Mosquitoes were identified morphologically and by molecular methods. The Plasmodium falciparum circumsporozoïte indexes were measured by ELISA, and the entomological inoculation rates (EIR) were calculated for both areas. Molecular assessments of pyrethroid knock down resistance (Kdr) and of insensitive acetylcholinesterase resistance were conducted. Results From May 2005 to October 2006, 4,117 and 797 Anopheles gambiae s.l. respectively were caught in Bel-air and Ouakam. Three members of the complex were present: Anopheles arabiensis (> 98%), Anopheles melas (< 1%) and An. gambiae s.s. molecular form M (< 1%). Infected mosquitoes were caught only during the wintering period between September and November in both places. In 2005 and 2006, annual EIRs were 9,5 and 4, respectively, in Bel-air and 3 and 3, respectively, in Ouakam. The proportion of host-seeking An. gambiae s.l. captured indoors were 17% and 51% in Bel air and Ouakam, respectively. Ace 1 mutations were not identified in both members of the An. gambiae complex. Kdr mutation frequency in An. arabiensis was 12% in Bel-air and 9% in Ouakam. Conclusion Malaria is transmitted in Dakar downtown area. Infected mosquitoes were caught in two subsequent years during the wintering period in two distant quarters of Dakar. These data agree with clinical data from a Senegalese military Hospital of Dakar (Hospital Principal) where most malaria cases occurred between October and December. It was the first detection of An. melas in Dakar.


Background
In the sixties, the relative seriousness of malaria and the seasonal transmission of Plasmodium falciparum by Anopheles gambiae s.l. in Dakar were reported in different studies [1][2][3][4]. Immunity was acquired quite slowly during the first 20 years of life [5]. Malaria-infected individuals came from rural areas to Dakar and contributed to the start of transmission, which culminated in October-November after the end of the rainy period [6]. In the eighties, an entomological survey proved that malaria transmission occurred in Pikine in the suburb of Dakar and that Anopheles arabiensis was the main vector in this area [6]. In the nineties, malaria transmission persisted in Pikine and in the surrounding villages of Dakar, always with An. arabiensis as the vector [7]. During the same period, two parasitological and entomological studies conducted in two sanitary districts of Dakar showed that the prevalence of malaria was very low; a few An. arabiensis were caught, and none of them were infected by P. falciparum [8][9][10]. Based on these results, many practitioners thought that there was no malaria transmission in Dakar intra-muros (i.e. in the down-town area) and that the infections occurred in the suburbs or inland. Nevertheless, human malaria cases were reported in autochthonous people who had not been outside Dakar for at least one year, suggesting that malaria was transmitted in Dakar [10]. At the same time, French military doctors were confronted with malaria cases in expatriates, who were visiting a malaria endemic country for the first time, but had never left downtown Dakar. To assess the reality of malaria transmission in Dakar, an entomological evaluation was conducted at the French military bases of Dakar from May 2005 to October 2006, over two winters.

Location
Located at 14°40'20" North, 17°25'22" West (the westernmost point of Africa), Dakar, the capital city of Senegal, has 1,030,594 inhabitants and covers the major part of the Cap Vert Peninsula. The altitude does not exceed 104 m. The population of the Dakar area is estimated to be 2.45 million people, representing 20% of the Senegalese population. The estimated population density is 12,233 inhabitants/km 2 .
The study was conducted in two districts of Dakar: Bel-air in the east of the city and Ouakam in the west (Figure 1). Bel-air is a residential district with luxuriant vegetation and with many market gardens and water wells along a railway, which crosses the area. The French military camp is bordered by the sea on three sides and by the railway on the other side. Ouakam is also a residential district with individual and collective houses. The French military camp of Ouakam is bordered by the sea on the west side and by houses on the other sides. It is a dry area with little vegetation. Between the camp and the sea lie a market gardens area and two water wells. The waste water of the French camp flows into a network that irrigates all the market gardens.

Climate
The Cap Vert Peninsula is located in the Atlantic Sudan zone. Two distinct seasons exist: a hot and wet season from June to November (maximum average temperature 28.2°C in October) and a cool and dry season from December to May (minimum average temperature 20.4°C in February). The first rains generally occur at the end of June or the beginning of July, and the last occur at the beginning of October. In 2005 and 2006 (the period covered by the study), the average rainfall was 525 and 350 mm, respectively.

Field mosquito processing
Sampling by human landing of malaria vectors was carried out both indoors and outdoors. Collectors gave prior informed consent and received anti-malaria prophylaxis and yellow fever immunization. Collectors were organized in teams of two for each collection point. Replacement of workers within a team was done every two hours from 7:00 PM to 7:00 AM. Teams of collectors were rotated among the collection points on different collection nights to minimize sampling bias. Landing catches were performed at seven points (two places indoors and five outdoors) in two periods, Mosquitoes were recorded by the location and hours of capture. They were sorted by genera, and anopheline mosquitoes were identified morphologically following the Gillies and Coetzee keys and by software from Hervy et al [11,12]. Culicinae were identified morphologically following the Edwards keys [13]. All mosquitoes were stored individually in numbered vials with desiccant and preserved at -20°C until processing at the Medical Entomology Unit of the Institute for Tropical Medicine (IMTSSA), Marseille (France).

Laboratory mosquito processing
Heads and thoraces of anopheline females were tested by enzyme-linked immunosorbent assay (ELISA) for P. falciparum circumsporozoite protein (CSP) [14]. All females belonging to the A. gambiae complex caught during the dry season, a random sample of females caught during the rainy season, together with all CSP-positive anopheline, were identified by polymerase chain reaction (PCR) at the species and molecular forms levels [15]. Molecular characterizations of the Kdr and Ace1 mutations were carried out on these mosquitoes as previously described [16,17].

Data analysis
The human biting rate (HBR) was expressed as the number of female anopheline bites per human per night. The CSP index was calculated as the proportion of mosquitoes found to be positive for CSP. The entomological inoculation rate (EIR) was calculated as the product of the HBR and the CSP index of mosquitoes collected on humans. The An. gambiae s.l. biting activity is sufficient for transmission only during the end of the rainy period; the EIR calculated for this period will be considered as the annual EIR. Conformity of Kdr and Ace1 frequencies with Hardy-Weinberg expectations was tested using a Pearson chi-square test considered significant when P < 0.05. Endophagic rates were compared using a chi-square test.

Weather data
Rainfall data were graciously provided by the National Weather Agency.

Biting rates and biting behaviour of An. gambiae s.l
The biting activity at the two sites from May 2005 to May 2006 is shown in Figure 2. An. gambiae s.l. was present throughout the year in Dakar, but most of the specimens (98%) were caught between July and December. The peak of biting occurred in September and October, at the end of the winter period (rainy season): 67% and 87% of mosquitoes were caught during these two months in Ouakam  Figure 3. Indoors over 70% of biting occurred between 1:00 a.m. and 6:00 a.m. in Ouakam and between 2:00 a.m. and 7:00 a.m. in Bel-air; outdoors, over 70% of biting occurred between 1:00 a.m. and 6:00 a.m. in both locations.
The average number per catching point of host-seeking An. gambiae s.l. caught indoor and outdoor were 121 and 115, respectively in Ouakam and 115 and 540, respectively in Bel air. The proportion of host-seeking An. gambiae s.l. captured indoors were 51% and 17% in Ouakam and Bel air, respectively. (P < 0.0001, RR = 2.92 [2.37; 3.59], indicating that this species was more endophagic in Ouakam.

Kdr-w and Ace1 mutation frequencies in An. gambiae s.l
All CSP-positive mosquitoes per site and a random sample of 100 PCR-identified mosquitoes were tested for the Kd-wr and Ace1 mutations. Ace 1 mutations were not identified in any members of the An. gambiae complex either in Ouakam or in Bel-air. The Kdr-w mutation frequency in An. arabiensis was 12% in Bel-air and 9% in Ouakam (not a significant statistical difference). The genotypic frequencies are shown in Table 3. The populations were not at Hardy-Weinberg equilibrium at either site for kdr (P < 0.0001).

Discussion
Malaria transmission is a reality in Dakar. P. falciparuminfected Anopheles have been caught during two rainy periods consecutively in two quarters of the town.  [18,19]. In addition, neither Anopheles pharoensis nor An. gambiae molecular form M caught in Dakar were infected with P. falciparum. The aggressiveness of An. arabiensis was higher in Bel-air than in Ouakam, but EIRs were very high in the two quarters. An. arabiensis is more endophagic in Ouakam. This behavior allows for an easier access to a human blood meal that may explain the similar EIRs observed for the two quarters.
The seasonal character of transmission allows us to calculate mean annual EIR for the two years of study. In Bel-air, the mean annual EIR fell from 9.5 in 2005 to 4 in 2006.
In Ouakam, a similar mean annual EIR of 3 was estimated for 2005 and 2006. Our observations are consistent with the results of a meta-analysis of studies of malaria transmission in sub-Saharan Africa, which found a mean annual EIR of 7.1 in the city centers, with more than twothirds of the studies reporting EIRs < 4/year [20]. In 1996, the number of infective bites per person was estimated to 0.05/year for the central area of Dakar, i.e., one infective bite every 20 years [10]. This work was conducted on a very large area with 13 study sites but with only two catching points per site and human landing collections performed once every month. Ten years later, the risk of malaria transmission seemed to be 60 to 80 times higher. The peak of An. gambiae s.l. aggressiveness lasts for a little longer than one month (Figure 3). With a monthly rhythm of capture, the previous study could have missed this peak. Nevertheless, a modification of the entomological situation over ten years cannot be excluded. There is some evidence that anopheline species may be adapting to urban ecosystems. Adaptation of An. gambiae s.s. to urban aquatic habitats, such as water-filled domestic containers, has been observed in Accra, Ghana [21]. In addition, adaptations of the anopheline vectors to new breeding sites (tree holes, polluted water) are reported from many urban areas in Africa [22][23][24]. The impact of urbanization on the composition of the vector system and malaria transmission dynamics has been highlighted in many studies [25,26]. Urban farming provides ample aquatic habitats for mosquitoes, which are responsible for the persistence of anopheline populations in many African towns [27][28][29]. In Ouakam and Bel-air, as in other places in Dakar, market gardens are present with or without water wells (cement wells or traditional wells called "ceanes"). The urban area of Dakar contains more than 5,000 market-garden wells, which provide permanent sites for mosquito larvae, in particular An. Arabiensis [30]. The increase of the An. arabiensis population size during the 2005 winter is an argument to highlight the major role of temporary pools in malaria transmission in Dakar. Further specific study is necessary to understand the impact wells and urban farming on anopheline density and malaria transmission.
This situation questions the origins of these anopheline populations. Are there autochthonous populations from Dakar growing during the rainy season, or are they coming in from other areas of Senegal? Populations of An. arabiensis in West Africa are considered to be continuous throughout the year, with many individuals surviving through the dry season, perhaps in a physiologically altered state rather than through extinction or a severe bottleneck during the dry season, followed by re-colonization by a few individual survivors or immigrants in the subsequent rainy season [31]. In Barkedji, Senegal, Simard et al did not detect any difference in measures of genetic diversity and linkage disequilibrium between the dry and rainy seasons [32]. They concluded that, despite extreme minima in local density, malaria transmission in this area was due to autochthonous population of An. arabiensis. They also found a low differentiation between two populations, which were 250 Km apart, suggesting extensive gene flow across this distance. These results suggest that An. arabiensis maintains a large permanent deme over a large area. The situation in Dakar seems similar to that in Barkedji: a long dry season where no An. arabiensis (or only a few) are caught [33]. Further genetic studies will be necessary to confirm the hypothesis that malaria transmission is due to an autochthonous An. arabiensis population, and to assess the gene flow between the urban and rural populations of Senegalese An. arabiensis.
Pyrethroids are the main insecticide used in malaria vector control, including indoor residual spraying and impregnated materials (bednets, curtains, plastic sheeting). Pyrethroids have the advantage of acting very rapidly as insecticides, with both knockdown and lethal effects at dosages under the threshold of mammalian toxicity [34]. Since 1970s, pyrethroids have been extensively used in urban areas and for agricultural purpose in rural areas. Detected in the 1990s, knock-down resistance (Kdr) to pyrethroids and DDT of An. gambiae s.l is an increasing problem. Two mutations at the same locus in the voltagegated sodium channel are known to confer knock-down resistance to a wide range of pyrethroids and DDT [35][36][37]. These mutations were previously described in west and east Africa (Kdr-w and Kdr-e) in An. gambiae s.s. as well in An. arabiensis [38][39][40][41]. There are few studies of An. arabiensis insecticide susceptibility in the area of Dakar. In 1987, bioassays were conducted in two places in the suburbs of Pikine and of Thiaroye [42], where in vivo resistance to DDT was observed. The agricultural use of DDT in market gardens was incriminated. In 1999, a normal susceptibility of An. arabiensis was found in Dakar [43]. Our data show a Kdr-w frequency of 0.12 in Bel-air and 0.09 in Ouakam. The populations were not at Hardy-Weinberg equilibrium for Kdr in either site (P < 0.0001). This disequilibrium could be due to local selection pressure by the agricultural use of pyrethroids in market gardens.
The Kdr mutation has been shown to be closely associated with DDT and pyrethroid resistance in several An. gambiae populations (particularly the molecular S form) [35][36][37]. However, the role of Kdr in conferring resistance in An. arabiensis remains unclear [44,45] As a result, insecticide susceptibility tests should be carried out to assess physiological resistance levels in An. arabiensis in Dakar. However, this is the first report of the Kdr-w mutation in An. arabiensis in Dakar. Resistance to carbamate and organophosphate insecticides is also widespread in West Africa. The presence of an insensitive acetylcholinesterase in populations of An. gambiae s.s. of both forms was revealed by biochemical assays. The ACE-1-R mutation has also been detected in the two molecular forms of An. gambiae s.s in many West African countries [17,46]. This mutation has not been detected in An. arabiensis at present. Our results are consistent with these data.
In terms of effective vector control at the military camps and in town, choice of insecticide should depend on the results of susceptibility tests on An. arabiensis.
Malaria prevalence is very low in Dakar and its urban periphery [7,8,10]. Nevertheless, cerebral malaria is the first etiology of neuromeningeal diseases in Dakar [47]. Extensive genetic diversity was observed in P. falciparum isolates collected in Dakar [48][49][50]. Significant linkage disequilibrium was observed with microsatellite loci in urban parasites. Two non-exclusive hypothesis could explain the situation in Dakar: (i) a global non-panmictic structure of Dakar malaria population due to a high predominance of selfing; (ii) a structuration in subpopulations of several malaria foci in Dakar. The entomological findings of the present study are consistent with seasonal transmission leading to an increase of cases in the October end of the rainy season. There is a concern about the ori-gin of Plasmodium falciparum infections observed in Dakar during the rainy season. Are they due to autochthonous parasites transmitted during the winter or to parasites imported from the suburbs by commuters, as suggested by Vercruysse? [6] In Senegalese people employed by the French army, the first malaria cases are diagnosed in commuters at the beginning of the rainy season, and cases in residents occur later in the season. The movement of populations from the suburbs or rural areas to Dakar must be considered to understand malaria epidemiology in Dakar [51,52].
Urban malaria is considered to be an emerging problem in Africa. In 2003, 39% of Africa's people lived in urban settings; by 2030, 54% of Africans are expected to do so [53]. With the increase of people living in urban dwellings, it's important to develop and validate new approaches for rapid appraisal of malaria risk. The rapid urban malaria appraisal (RUMA) methodology has been develop to provide a cost effective tool to conduct assessment of the malaria situation in urban sub-Saharan Africa and to improve the understanding of urban malaria [54]. The only entomological point required is in this evaluation is the mapping of breeding sites. Considered very time-consuming, this task has been done only in Ouagadougou [27].
The results of this study confirm that the vector complex situation in African towns is always changing, with the description of An melas and An gambiae s.s. in Dakar. Entomological studies are long, difficult and require time and special expertise but they are indispensable to understand the dynamics of malaria transmission in urban settlements and to monitor the increase of the insecticide resistance in urban mosquitoes. The study showed a difference in host finding behaviour between the two quarters: An. arabiensis are more endophagic in Ouakam than in Bel air. This difference of behaviour has an impact on the malaria transmission level. With a lower biting activity during the transmission season, the Annual EIR in Ouakam remains as high than in Bel air. Indoor vector control measures will probably not have the same impact in the two quarters. The misuse of impregnated bed nets during this period is probably riskier in Ouakam.
Malaria is transmitted in Dakar. This seasonal transmission occurs only during the two last months of the rainy season. The transmission level could be very high. Further studies have to be conducted in other parts of Dakar to assess the risk of transmission, to understand the role of permanent and temporary pools, the impact of urban farming and to discover the origin of the Dakar anopheline populations.