The PCR genotyping analysis, using msp-1 and msp-2 polymorphic markers, was performed to improve our knowledge on the genetic diversity of P. falciparum parasite populations in three regions with different malaria transmission pattern in Madagascar. So far, this information is currently lacking. Until now and excluding genotyping data obtained from clinical drug efficacy studies (performed to distinguish recrudescence from reinfections in enrolled patients presenting recurrences during their follow-up), only two studies, performed in 2000 and 2008, have reported such analysis [11, 19].
Overall, the number of different alleles for msp-1 and msp-2 found in the three sites (18 and 40, respectively) confirms a high level of malaria transmission in Madagascar. These numbers are comparable to those found in Africa such as in Nigeria, the Republic of Congo, the Central African Republic, the Equatorial Guinea or Senegal [20–24].
The K1-type for msp-1 and FC27-type for msp-2 were the most predominant alleles. These findings are consistent with data previously reported in Madagascar [11] and in different settings such as Africa (Nigeria [15, 22], Congo Brazzaville [25], Mauritania [26], Benin [27], Gabon [28, 29], Ivory Coast [30], Cameroon, [31], Ethiopia [32–35]), India [36] or Southeast Asia [37]. However, they contrast with recent reports from Myanmar [38] that showed that MAD20 and 3D7 were the most prevalent alleles.
The distribution of the msp-1 and msp-2 allelic families varied significantly between sites (Table 2). By msp-1 genotyping, half of the individuals had P. falciparum isolates with a single msp-1 allele. The proportion of isolates with more than one msp-1 allele was significantly higher in patients living in the Equatorial facies site, likely reflecting higher malaria transmission in this setting. This association was confirmed by a significantly highest proportion of polyclonal infections in Tsaratanana (Equatorial facies, 60.8%) compared to the two other sites (40.3% and 42.5%) (Table 3). Msp-2 allelic diversity was found to be more contrasted between sites: the type FC27 was largely more frequent than the 3D7 type in the Tropical zone, while in the Equatorial and the fringes facies, the 3D7 allele type was more predominant.
The number of msp-1 and msp-2 genotypes per isolate ranged from 1–4 and 1–3, respectively. Again, the means MOI calculated by msp-1 or combined msp-1 and msp-2 genotyping were found to be significantly higher in isolates from patients living in Tsaratanana (Equatorial facies) (1.92 and 2.52) compared to the means MOI in isolates from patients living in the two other sites (1.51 and 2.02 in Anjoma Ramartina; 1.50 and 2.25 in Antanimbary, respectively). These values are similar to those reported in some African countries, like Ghana [39], Congo Brazzaville [24], Ethiopia [32–35], but lower compared to Nigeria [22] and Gabon [40]. An association between the increase of the MOI and the age group was also observed. Particularly, the mean MOI (msp-1 and combined msp1/msp-2) of isolates obtained from patients aged from 5–15 years were higher in Tsaratanana (Equatorial facies) compared to Anjoma Ramartina (Fringes) and Antanimbary (Tropical facies). However, no such association was found between the mean MOI and the parasite density, counter to several reports [41].
However, these data confirm that the assessment of the MOI is a good measure of the malaria transmission intensity and this metric can be considered in Madagascar as a useful tool to evaluate the impact of vector control measures (Long-lasting bed nets and insecticides indoor sprayings) currently implemented over the country.
The genetic population measures did not reveal significant differences between sites. First, the He was found to be similar between study sites, age groups and parasite density groups. Second, the estimation of the fixation index (Fst) and the low proportion of shared genotypes between the study sites indicated that the circulation of the parasite population remains limited between sites.
Despite providing recent data on P. falciparum genetic diversity, this study presents several limitations. First, the main limitation was the use of msp-1 and msp-2 markers for genotyping that, as other markers based on DNA fragment size, could reduce the estimation of the genetic diversity. Nevertheless, msp-1 and msp-2 are usually considered as robust polymorphism markers and genotyping protocols are easy to perform in low-equipped laboratory such as those available in Madagascar. Second, the selected sites for sample collection was not designed for this study (but for assessing the efficacy of artesunate-amodiaquine in clinical drug efficacy study) which limits the opportunity to extrapolate the data. Third, as no direct malaria transmission measures such as the Entomological Inoculation Rate (EIR) were available in the selected sites, no opportunity to investigate association between these metrics and the genetic diversity data was possible.