Preliminary review on the prevalence, proportion, geographical distribution, and characteristics of naturally acquired Plasmodium cynomolgi infection in mosquitoes, macaques, and humans: a systematic review and meta-analysis

Plasmodium spp. infections in humans have been identified with Plasmodium falciparum, P. malariae, P. vivax, P. ovale, and P. knowlesi [1]. Although P. falciparum and P. vivax are the two major malaria species that cause malaria in humans worldwide, the infection of P. knowlesi is a major cause of simian malaria in Malaysian Borneo [1‐4] and is also reported as a cause of simian malaria in other parts of Southeast Asia, including Indonesia [5], Laos [6], Vietnam [7, 8], and Thailand [9]. P. knowlesi and other simian malaria parasites, including P. fieldi, P. coatneyi, P. cynomolgi, and P. inui, mainly infect long-tailed (Macaca fascicularis) and pig-tailed macaques (Macaca nemestrina) [2].

P. cynomolgi has accidentally and experimentally been reported as a cause of human malaria [3‐5], but there was a case report showing this parasite as a naturally acquired human infection in a Malay woman from the east coast of Peninsular Malaysia who lives in an area where long-tailed macaques are present [6]. P. cynomolgi was first observed in 1907 in Macaca fascicularis collected in Java [7]. P. cynomolgi has morphological features similar to P. vivax, as shown by microscopy, including the asexual cycle (48 h), prepatent periods, and presence of hypnozoites, which can initiate relapses [6]. The identification of P. cynomolgi relies on the amplification of the 18S ribosomal RNA (rRNA) gene by nested polymerase chain reaction (PCR) sequencing. As P. cynomolgi is recorded as the most recently discovered simian malaria parasite infecting humans, its prevalence, proportion, geographical distribution, and characteristics remain unclear. The present study aimed to systematically review reports on naturally acquired P. cynomolgi in humans, mosquitoes, and macaques to provide relevant data for pre-emptive surveillance and preparation in the event of an outbreak of zoonotic malaria in Southeast Asia.

The successful transmission of zoonotic malaria is highly dependent on the bionomics and distribution of competent vectors and natural hosts. Humans and macaques live in and share overlapping spaces, particularly those who work in farming or agriculture near forests or tourists who come to visit areas where macaques are endemic. These close contacts and sharing of geographic distributions between humans and macaques and the presence of suitable vectors can lead to malarial disease transmission between host species. The increase in zoonotic malaria before the discovery of P. cynomolgi was primarily found in the occurrence of P. knowlesi malaria. P. knowlesi infection in humans is driven by multiple factors, including anthropogenic land-use factors leading to changes in the transmission pattern of the parasite between macaque reservoirs, vectors, and humans, as observed in Malaysian Borneo [24]. Deforestation for palm oil plantations or other clearing activities resulted in the loss of natural habitats for macaques, which led to their close contact with human settlements and increased incidence of P. knowlesi infection in humans [25].

The present study demonstrated a high prevalence and density of P. cynomolgi in macaques (47%), particularly in Indonesia [17] and Singapore [14, 17]. Interestingly, half of the P. cynomolgi infections in macaques were mixed infections with other Plasmodium species, and among the mixed infections, triple and quadruple infections were the most common types of mixed infections observed. It should be noted that mono-infection cases were predominantly in humans and Anopheles vectors, while an almost equal number of cases were reported in macaques. More interestingly, P. cynomolgi was likely to be a mixed infection with P. inui (90.4%), P. coatneyi (51.4%), and P. knowlesi (45.9%). The latter seems to support the similarity of transmission between P. cynomolgi and the more well-known zoonotic malaria caused by P. knowlesi. Zoonotic malaria caused by P. knowlesi infections was described and focused on in Sarawak, Malaysian Borneo, in 2004 [1], and later, it became a more common cause and was recognized as the fifth human malaria parasite throughout Southeast Asia [2]. Similar to P. knowlesi infection, naturally acquired P. cynomolgi infection in Malaysia was reported 10 years later in 2014 in endemic cases of people who lived in the same area with long-tailed macaques [6]. From that time to 2020, a tourist who traveled to Surat Thani Province, Thailand, was naturally infected with P. cynomolgi [26]. Then, naturally acquired P. cynomolgi infections in humans were detected in other parts of Southeast Asia, including Malaysia and Cambodia [18‐20]. Moreover, when comparing the pooled proportion of P. cynomolgi infections in three different hosts in the studies conducted before 2013 and after 2013, the pooled proportion of P. cynomolgi infections in mosquitoes was higher in studies conducted after 2013. Nevertheless, the pooled proportion of P. cynomolgi infections in macaques was lower in studies conducted after 2013. The results of these streamlined publication years suggested that P. cynomolgi infections in humans might be due to the proximity of suitable vectors and monkeys, leading to conditions favorable for interspecies transmission, as demonstrated in cases of P. knowlesi infection [11].

Figure 3 shows that regions with the presence of both vectors and natural hosts also reported human infection cases, as was the case in Malaysia. However, human transmission in Cambodia seemed to have been caused by the presence of macaques in the country and its proximity to southern Vietnam, which reported the most cases of infected Anopheles vectors. Although cases in macaques have been reported in the Philippines, Laos, Singapore, and Indonesia, there is a curiosity as to why there is a lack of reports on infected Anopheles mosquitoes and humans. The results of the present study indicated that in areas where a high prevalence of P. cynomolgi was seen in macaques, sustained public information and advocacy in the affected areas is still necessary even if public advocacies have already been performed since 2014 [27‐30].

Although a high prevalence of P. cynomolgi in macaques was demonstrated by the included studies, the prevalence of P. cynomolgi in mosquitoes was very low at 0–1%, and the proportion of P. cynomolgi compared to all Plasmodium species detected in mosquitoes was 18%. Therefore, the transmission of P. cynomolgi from macaques to mosquitoes should be limited by the bite rates, the susceptibility of Anopheles vectors, and the parasite density of the primate host. The contrast between a high prevalence of P. cynomolgi in macaques and a lower parasitemia of Anopheles mosquitoes should be due to the limited susceptibility of Anopheles vectors to harbor a lower parasitemia of P. cynomolgi in a specific environmental condition, which can lead to disease transmission, as changes in host preference, biting behavior, and adaptation of the mosquitoes to habitat changes could affect the transmission of zoonotic malaria in Southeast Asia. Although the proportion of P. cynomolgi infecting Anopheles mosquitoes was low, the importance of transmission through the infection of human hosts by the bite of the infected Anopheles mosquitoes should be monitored. In addition, although the pooled proportion of naturally acquired P. cynomolgi in human hosts demonstrated by the included studies was 1%, the high density of infected mosquitoes, such as in Northern Sabah (26%) [22], could lead to a high prevalence of naturally acquired P. cynomolgi in humans, as demonstrated in Northern Sabah (4%) [18].

P. cynomolgi infection in humans was observed to be asymptomatic and submicroscopic [18, 19]. This implied that the mono-infection of P. cynomolgi was very low in parasite density despite the mixed infection with P. vivax as demonstrated by Imwong et al. [19]. The morphological life cycle of relapse and the genetic similarities of P. cynomolgi and P. vivax could lead to undetected or undiagnosed P. cynomolgi cases in Southeast Asia. Most of the cases with suspected malaria admitted to the hospital were identified by the microscopic method, which has a low sensitivity and specificity to differentiate P. cynomolgi from P. vivax. In addition, the misidentification or missed detection of submicroscopic parasitemia by P. cynomolgi might lead to under-reported cases. Moreover, failure to treat P. cynomolgi malaria can lead to malarial recurrences, as observed in two individuals from Charkrya and Ou Treng, Cambodia [19]. Even in symptomatic infections of P. cynomolgi, such as in the mixed infection of P. cynomolgi and P. knowlesi, as reported in the study by Raja et al. [20], the parasite density ranged from 213 to 84,299 parasites per milliliter. Moreover, the most recent study with enrolled febrile patients at malaria clinics or local hospitals in Thailand demonstrated that most of the patients with P. cynomolgi mixed infection with P. vivax, P. falciparum or both P. vivax and P. knowlesi had parasitemia less than 10,000 parasites/μL (< 0.2% parasitemia) as demonstrated by Putaporntip et al. [31]. The low proportion of P. cynomolgi detected in humans by three studies [18‐20] might be due to the use of less-sensitive molecular protocols, as they used nested PCR amplifying the SSU rRNA gene for the detection of malaria parasites. PCR amplification of the malarial cytochrome oxidase gene is superior to PCR amplification of SSU rRNA, as cytochrome oxidase is a mitochondrial genome that is found at approximately 20–150 copies per parasite, whereas SSU rRNA is found at approximately 4–8 copies per parasite [32, 33].

The natural infection of P. cynomolgi in humans is similar to the infection of the natural host of long-tailed or pig-tailed macaques since mixed infections of more than one species in a single host were observed in both humans and macaques. Nevertheless, the dominance of P. knowlesi over P. cynomolgi, as demonstrated by Raja et al. [20], might be due to the differences in the developmental cycles of the two species. P. cynomolgi has both an exoerythrocytic cycle in the liver and erythrocytic cycle in red blood cells. This means that the incubation period for P. knowlesi is shorter by approximately 9–12 days than that of P. cynomolgi, which is between 15 and 37 days [7]. Furthermore, the erythrocytic cycle of P. cynomolgi (48 h) is longer than that of P. knowlesi (24 h) [34]. Therefore, P. cynomolgi mixed infection with P. knowlesi might occur simultaneously in a single mosquito bite harboring two parasite species. The simultaneous infection of P. cynomolgi with P. knowlesi in a single mosquito was demonstrated in a study by Chua et al. [22].

The present study had limitations. First, the number of studies identifying P. cynomolgi in humans and mosquitoes was limited, resulting in an imprecise estimate of the pooled prevalence or pooled proportion and of the clinical characteristics of human hosts. Second, the Northern Sabah study [18] demonstrated a lower NOS scale (2 stars) as poor comparability against the other studies which used molecular detection methods but retained in the present analysis. Therefore, the pooled proportion of P. cynomolgi infection in humans requires careful interpretation. Further studies and continued surveillance of P. cynomolgi throughout Southeast Asia and neighbouring regions may be necessary through the use of sensitive molecular protocols to obtain accurate and relevant data for the detection and reporting of this emerging zoonotic malaria. Moreover, it is essential to further explore the biology of P. cynomolgi, the possibility of relapses/recurrences, and asymptomatic infections, which have a direct impact on the epidemiology of malaria.

In conclusion, P. cynomolgi could continue to be a public health concern in Southeast Asian countries, as the natural habitat of the natural hosts and vectors and the evolution of the P. cynomolgi parasite itself could drive the transmission of this neglected but emerging disease. In addition, the environmental and climatic changes experienced worldwide could affect the transmission dynamics of P. cynomolgi as an emerging cause of malaria in humans. Further studies using molecular and multidisciplinary approaches to search for P. cynomolgi infection in humans, vectors, and natural hosts are necessary if human infections with P. cynomolgi do become public health concerns.