Sub-microscopic malaria cases and mixed malaria infection in a remote area of high malaria endemicity in Rattanakiri province, Cambodia: implication for malaria elimination

The development of sensitive PCR-based methods for malaria diagnosis has highlighted the low sensitivity of malaria microscopy for very low-level parasitaemia (<50 parasites/μl). The low sensitivity of microscopy may have at least two important consequences for malaria control and eradication efforts. First, asymptomatic sub-microscopic parasitaemia may serve as a reservoir for infection even when very efficient rapid diagnosis and treatment programmes have been implemented. Second, mixed infections may be overlooked when one species is present at low parasitaemia, and clinically or epidemiologically important interactions between species may be missed.

Current malaria control methods in Southeast Asia rely largely on early detection and treatment; control based on insecticide-treated bed nets (ITN) may be less effective in areas where malaria is primarily acquired not at home but by working age men engaged in mining, hunting, or logging in the forest. Since individuals with asymptomatic parasitaemia will not be identified by early detection and treatment programmes, they may continue to serve as a source of infection for vector mosquitoes, complicating control measures. It is clear that PCR-based methods can detect sub-microscopic parasitaemia [1‐4]; it remains to be determined how common such cases of parasitaemia are in the field under different ecological conditions and what effect they may have on transmission.

Accurate detection of mixed infections is important for both clinical and epidemiological reasons. The four Plasmodium species, which commonly infect humans, have somewhat different clinical characteristics. Plasmodium falciparum is the most lethal species, causing severe malarial anaemia and cerebral malaria and, consequently, the vast majority of malaria deaths. Plasmodium vivax and Plasmodium ovale may persist within the liver as hypnozoites causing relapses even after treatment with blood schizonticides. Plasmodium malariae may cause chronic asymptomatic parasitaemia and may lead to renal failure via unclear mechanisms [5]. Finally, Plasmodium knowlesi is a newly emergent malaria parasite in Southeast Asia [6, 7]. How these clinical characteristics are modified in mixed species infections is not clear. Voza et al in 2005 found a species-specific inhibition of cerebral malaria in mice coinfected with Plasmodium spp [8]. In human, according to several studies in Ivory Coast [9], Sri Lanka [10] and Thailand [11], the severity of infection may be modulated by mixed infections. In a study carried out in Vanuatu, P. falciparum appeared to be associated with a reduced P. vivax parasitaemia [12, 13], but the reverse effect has been observed in another study [14]. There seems to be a strong interaction between these two malaria species, and P. vivax co-infection has even been reported to decrease treatment failures of P. falciparum [15]. In other studies, P. falciparum - P. vivax mixed infections affected the clinical outcome in patients [13, 16]. Examples of P. malariae and P. ovale co-infections with P. falciparum protecting against malaria symptoms have also been reported [17]. Studies with a Plasmodium berghei and Plasmodium yoelii co-infection model in mice appear to confirm that there can be a strong cross-protective effect of mixed species infections [18]. Mixed infections may have important epidemiological effects. For example, if P. vivax parasitaemia is suppressed by co-infection with P. falciparum, then effective control of P. falciparum malaria in an area might be followed by an increase in P. vivax transmission.

Studies of mixed species infections based only on microscopy may underestimate their importance. The frequencies of minority species, such as P. malariae and P. ovale, are largely underestimated by microscopy [19‐21]. PCR-based methods are more sensitive [1‐4] and more readily detect mixed infections. Using PCR diagnosis, between one third and half of malaria infections in a study in Thailand were found to be mixed species infections [22].

The detection of Plasmodium species at very low parasitaemia is difficult and requires a molecular approach. The performance of a relatively simple and semi-automated technique has been evaluated here. It is a highly sensitive PCR-based diagnostic method using amplification of a fragment of the cytochrome b gene, followed by RFLP analysis using dHPLC to detect amplification and restriction products. Compared to the well established PCR typing of the 18S rRNA gene for species detection, this method appeared reliable, sensitive (multi copy number of the cytochrome b gene), specific and rapid (two PCR instead of five) and give the ability to amplified large number of species [27]. The dHPLC semi-automated analysis allowed the screening of a 96-well plate in 48 h and to diagnose 1,356 samples in less than one month.

Use of the RFLP-dHPLC method revealed a high prevalence of sub-microscopic infections. The overall prevalence detected by the RFLP-dHPLC method was twice that detected by microscopy (68.4% versus 30.7%). Interestingly, the proportion of sub-microscopic P. falciparum infections, was higher in older age groups. It is possible that acquired immunity favours maintenance of sub-microscopic, asymptomatic infections. This substantial population of adults with sub-microscopic, asymptomatic P. falciparum infections may represent a significant challenge to malaria control programmes. Such individuals will not seek treatment and even if they were included in a mass screening and treatment campaign, their parasitaemia would remain invisible to light microscopy or rapid diagnostic tests. Moreover, sub-microscopic P. falciparum gametocyte densities may contribute importantly to mosquito infections and thus to maintaining transmission [27]. In order to have a major impact on malaria transmission, mass screening and treatment campaigns may need to use molecular screening techniques capable of identifying sub-microscopic parasitaemia.

The RFLP-dHPLC method detected many more mixed species infections than did microscopy (Table 3). Indeed, 24.5% of all malaria infections in the Rattanakiri survey included more than one species. No association was found between the level of parasitaemia and the chance that the infection would be mixed. Co-infection with P. vivax and P. ovale was more common than would be expected based on the individual prevalences of these species (OR = 4.35, IC95 = 1.64 - 11.49, p = 0.003). This association could be attributed to cross-immunity that determines susceptibility to both infections [31] or to the exposure to infective bites of common vectors that transmit both species. An alternative explanation might be that both these species can produce relapses from hypnozoites. Similarly, P. falciparum and P. malariae infection were associated (OR = 1.61, IC95 = 1.12 - 2.32, p = 0.010). There was, however no association between P. vivax and P. falciparum infection.

Plasmodium falciparum and P. vivax infections were found at different prevalences in the different ethnic groups. Since the different groups lived in different individual villages, it is not possible to decide whether the observed differences reflect different genetic susceptibilities, as observed for some African ethnic groups [32] or simply to different local transmission intensities or other local factors.

Rapid diagnosis and treatment of malaria cases, vector control and protection against mosquito bites are the cornerstone of the malaria control strategy in Cambodia. Sub-microscopic infections may provide a reservoir of infection, which maintains transmission. This phenomenon has already been highlighted in Guinea Bissau [1] in Brazil [33] and Gabon [34]. The high prevalence of sub-microscopic parasitaemia within Rattanakiri Province, a site of higher malaria transmission in Cambodia, is here shown. Such infections may hamper elimination efforts, since they are asymptomatic and would not be efficiently detected by conventional diagnosis methods, i.e. microscopy or rapid diagnostic tests. Thus, the rapid and sensitive molecular diagnosis method developed here could be considered for mass screening and ACT treatment of inhabitants of low-endemicity areas of Southeast Asia to face the urgent need to eradicate malaria in Cambodia due to emerging artemisinin resistance in the country.

Ethical approval for this study was granted by the National Ethics Committee of the Kingdom of Cambodia.