A review of mixed malaria species infections in anopheline mosquitoes

Where transmission of both vivax and falciparum malaria is high in parts of Oceania, mixed species infections in humans are common. With frequent infection from repeated inoculation it is not surprising that infections accumulate and so comprise a multiplicity of genotypes and species. In South-East Asia where transmission of malaria is low, seasonal, and unstable people typically receive one infected mosquito bite each year or less, yet a remarkably high proportion (30 to 50%) of acute Plasmodium falciparum malaria infections are followed soon afterwards by an infection with Plasmodium vivax. Indeed in endemic areas in this region, the main clinical complication of falciparum malaria is subsequent vivax malaria, far outweighing the incidence of recrudescent falciparum malaria [1‐5]. The intervals between presentation with falciparum malaria and the subsequent vivax episode are very similar to the intervals between acute vivax malaria and the first relapse [2, 5]. This suggests that the vivax malaria episode, which follows falciparum malaria, also results from a relapse. The very high rate of co-infection in South-East Asia has suggested that the two infections are acquired together despite the low transmission frequency.

Mixed species infections in patients are often under reported, with a tendency to over report the more dangerous P. falciparum. In endemic areas, where the prevalence of malaria revealed by sensitive PCR methods is much higher than evident from microscopy, high rates of mixed blood stage infection (~20%) have been reported, presumably because both infections are carried chronically [6, 7]. Lower rates of mixed species infection occur in symptomatic non-immunes presenting with acute falciparum malaria in low transmission settings. Even with sensitive PCR detection of low parasitaemia, the proportion of mixed infections detected does not approach the 30 to 50% required to explain the proportion of vivax malaria episodes, which follow P. falciparum malaria in these same patients. Simultaneous infection studies in the malaria therapy of neurosyphilis and in volunteers clearly showed that mixed infection could occur from simultaneous inoculation, that pre-erythrocytic development of P. falciparum was more rapid, and that in the blood stage infection P. falciparum tended to suppress P. vivax, so low level P. vivax parasitaemia might still be missed by current diagnostic methods [8‐10]. Plasmodium vivax may also suppress P. falciparum; up to 10% of acute vivax malaria episodes in Thailand are followed shortly after by P. falciparum infections without reinfection [3, 11]. Blood stage infections in which both parasites are detectable could either derive from simultaneous inoculation of the sporozoites of the two species from a doubly infected anopheline mosquito, or a recently acquired infection of one species might supervene a chronic infection with the other. As transmission intensities in most of South East Asia are very low (EIRs typically < 1/year) the probabilities of separate inoculations within a narrow time window are extremely low [12]. To resolve these questions and understand better the epidemiology of mixed species infection in different transmission settings we examined the published literature on immunological and molecular parasite species identification in anopheline vectors in malaria endemic areas.

The NLM PubMed was searched since 1987 for publications in English describing studies on anopheline vectors in malaria endemic areas in which malaria parasite identification had been performed using immunological or molecular methods. The search terms were; anopheles AND sporozoite OR oocyst; Subheadings: Asia OR Africa OR America. Studies of both trapped wild mosquitoes and artificially infected mosquitoes were included. After compilation the lists were checked manually. The location of study and year, the collection method, mosquito species, number of specimens, parasite stage examined (oocysts or sporozoites), parasite detection method, and the results of the species identification were tabulated. Investigators in the South East Asian region were contacted for further information. ELISA and more recently PCR methods were used generally for species identification. The sensitivity of these different methods were also reviewed.

In total, 42 studies were identified which were conducted in Asia and Oceania (11 were from Thailand) and 21 studies were conducted in other areas (14 were from Africa and 7 from Central or South America) between 1987 and 2011. Three of the American and three of the Asian studies involved blood feeding of anophelines and later dissection. In the remainder of the studies trapped wild anophelines were examined.

In areas where malaria transmission is stable and intense the majority of the human population is constantly infected, although older children and adults often have parasite densities below the level of microscopy detection. Many parts of Africa have this pattern of malaria transmission. Mixed genotype P. falciparum infections are common, but there are relatively few studies which have examined the prevalence of other infections. In Asia and Oceania, these areas of intense transmission tend to be very small focal forest or forest fringe areas, and much of the coastal part of the Island of New Guinea. In those areas where both P. falciparum and P. vivax are prevalent mixed species infections with both species are common but maybe also be below the level of microscopy detection [8]. Both infections coexist chronically [16]. Sensitive PCR methods reveal high rates of mixed species infection (~20%) [6] although it seems likely that the majority of mixed infections in humans are acquired from separate mosquito inoculations. Asymptomatic or oligosymptomatic individuals who are parasitaemic have declared themselves able to control the infections and so are much more likely to have a multiplicity of different genotypes present (albeit at low densities) and also mixed species infections.

In contrast in much of the remainder of Asia transmission is generally low and seasonal and, although P. vivax infections may still be asymptomatic, P. falciparum malaria is usually accompanied by illness and treatment seeking. Symptomatic infections are seen at all ages, although the burden of symptomatic vivax malaria is in young children. In this epidemiological context mixed species infections are found in 10% or less of acute infections, yet approximately one third of patients still experience vivax malaria following falciparum malaria. The high efficacy of initial falciparum malaria treatment against P. vivax and the time interval between initial presentation and the subsequent vivax malaria episode suggests that it is a relapse. This presumed relapse was thought to result from simultaneously acquired mixed species infection, but the evidence reviewed here does not support this hypothesis. The proportion of P. falciparum infected mosquitoes overall which also carried P. vivax in Asia and Oceania was only 6.6% overall. As expected most mixed infections in mosquitoes were found in studies on the island of New Guinea where transmission of both species is intense. None were found in Thailand where one third of P. falciparum malaria is followed by P. vivax. Thus, the proportion of mixed infections in anopheline mosquito vectors in Asia and Oceania is approximately one fifth of that expected. These findings suggest that simultaneous inoculation of the two malaria species does occur, and indeed is more frequent than would be expected by chance alone, but that the majority of mixed species infections do not derive from a single mosquito bite.

Several explanations are possible. The entomology sampling frameworks in these studies may not have been representative. The vectors might not have been adequately representative of those transmitting malaria. The laboratory methods may have been insensitive (Table 1) or not designed specifically to identify mixed species infections. These concerns were addressed to some extent in those studies in which some mixed infections were identified. Lack of sensitivity seems an unlikely explanation for the observations. As only a small fraction of the salivary gland sporozoites are inoculated by a biting anopheline mosquito- it seems unlikely that these would derive from very small subpopulations of different species sporozoites which were below the level of antigen or PCR detection.

If the mixed species infections are all acquired from simultaneous inoculation it would also requires very high rates of concomitant P. falciparum and P. vivax gametocyte carriage at infectious densities in the blood of the parasitaemic population. This may well occur in high transmission settings but is rare in low unstable transmission settings and is not supported by the mosquito feeding studies reported here. As P. vivax develops more rapidly than P. falciparum in the anopheline vector, it also demands mosquito longevity. If the two infections were acquired separately then the close and reproducible temporal association between the infections needs explanation. Although the distribution of infected mosquito biting undoubtedly clusters it would require a remarkable degree of association for one third of all P. falciparum inoculations to be accompanied within a day or two by a P. vivax inoculation -when average inoculation rates are less than one infected bite per person per year! Nevertheless mixed species infections in anopheline mosquitoes do occur in Asia significantly more frequently than expected by chance so it is clear that simultaneous transmission does occur and does contribute to mixed infections. It is likely to account for the significant proportion of acute symptomatic malaria in low transmission settings where both species are identified in the blood at presentation.

The data from Central and South America where transmission is generally low and unstable also showed low rates of mosquito infection. No mixed infections were found but as P. falciparum and P. malariae infection rates were also very low this negative finding has limited significance. There are surprisingly few data from Africa on species other than P. falciparum. Plasmodium vivax was identified in Ethiopia as expected. In Ghana where transmission is intense mixed infection was identified. Artificial infection studies in Guinea-Bissau suggested that mixed infection of mosquitoes would be frequent, but these data are too few to draw firm conclusions on the entomology of mixed malaria infection in Africa.

This review is almost certainly an underestimate of the available information. For example, much of the Central and South American literature is in Spanish or Portuguese and some of the African literature is in French. The search strategy employed may well have missed studies in English, which lacked the key words. Nevertheless because of the large apparent discrepancies between mixed species infection rates in the anopheline mosquito vectors and humans this comprehensive sample of the literature provides a convincing message. Mixed infections of anopheline vectors do occur and account for some of the mixed infections in humans. However in South East Asia where P. vivax infections follows P. falciparum malaria in one third of cases, the published data suggests that the majority of these mixed infections are not acquired by simultaneous inoculation. A more plausible explanation is that P. vivax relapse is stimulated by symptomatic falciparum malaria infection.