Detection of Plasmodium falciparum and Plasmodium vivax subclinical infection in non-endemic region: implications for blood transfusion and malaria epidemiology

In Brazil, human malaria is mainly caused by Plasmodium vivax, which accounts for more than 80% of identified cases; Plasmodium falciparum is responsible for 16.3%, and Plasmodium malariae is associated with a small number of cases [1, 2]. The majority of malaria is reported in the Brazilian Amazon region [3] where Anopheles darlingi is the primary vector in recently deforested areas [4].

The annual parasite index is typically employed to estimate the risk of occurrence of malaria in human populations that are exposed to infective Anopheles mosquito bites, in a specific time frame. In endemic regions, this index is spatially and temporarily variable, and its values can be low (0.1–9.9), moderate (10.0–49.9), or high (>50) [3], depending on ecological and environmental determinants that represent distinct risk in different social and economic contexts [5]. Areas of high endemicity have an elevated prevalence of individuals with subclinical infections [6]. In the Amazon region of Brazil, studies performed in Amerindians [7] and in riverine communities [8] in the state of Rondônia have shown the presence of long-term subclinical infections. Those individuals were infected with Plasmodium spp; however, most of them showed low levels of parasitaemia [9]. Subclinical infections also have been found in meso- and hypo-endemic regions [8, 10, 11]. The São Paulo municipality and nearby areas in southeastern Brazil are considered non-endemic regions; however, in 2013, 146 autochthonous cases of malaria were identified in São Paulo State, representing 22% of the total occurring outside the Amazon region. Moreover, there was a 256% increase compared to 2006 [12].

Plasmodium pathogens that cause malaria in humans are naturally transmitted by infectious Anopheles mosquito bites [13]. However, the protozoan can be accidentally transmitted by blood transfusion [14] and intravenous drug users [10]. The first case of transfusional malaria was registered in 1911 [15‐17]. Although uncommon, transfusional malaria can have serious clinical implications when undetected and not treated early [18, 19]. In the Amazon State, 0.3% of screened blood donors were found to be infected with P. vivax[20] when the diagnoses were based on an especially sensitive molecular probe, i.e., PCR technology. In Brazil, the laws governing the screening obligation for Plasmodium infection in blood donors from an endemic region are distinct from those adopted in non-endemic areas [21]. Furthermore, in non-endemic areas, all candidates for blood donation are subjected to an interview before they donate blood. Following the interview, potential blood donors can be either rejected or accepted for donation, without enforcement for a laboratorial screening for Plasmodium infection [21]. Considering both the circulation of Plasmodium spp. in non-endemic areas and the fact that there is no obligation for screening for Plasmodium, it is plausible to suppose that accidental transfusional Plasmodium transmission can occur, mainly by the presence of subclinical infection in humans.

Transfusion-transmitted Plasmodium parasites represent an important risk factor because they can cause severe malaria with a high fatality rate [22]. The chronic infection can persist in humans for a long period of time: up to two years or more for P. falciparum, up to seven years for P. vivax, and even for an entire lifetime for P. malariae[1]. Consequently, an asymptomatic human who is unaware of his/her status as a Plasmodium reservoir can transmit the parasites when donating blood. Usually, parasite concentration is low in asymptomatic individuals [8]; however, asymptomatic humans can both infect Anopheles females [9] and cause severe infection in blood recipients [22].

Employing a high sensitivity real-time PCR in a cross-sectional design, it was addressed 1) the prevalence of P. falciparum and P. vivax among blood donors at a public blood bank in the municipality of São Paulo, Brazil; 2) the spatial distribution of the blood donors; and 3) the potential association between the geographical origins of donors and the Plasmodium species.

A total of 1,108 individuals were selected in the present study. From this sample, 61 were exposed and had a subclinical malaria infection; 439 were exposed and did not have a subclinical malaria infection; 23 were non-exposed and had a subclinical malaria infection; and 585 were non-exposed and did not have a subclinical infection (Table 1). The overall prevalence of subclinical malarial infections among blood donors was 7.58 (%); prevalence among exposed individuals was 12.20 (%) and among non-exposed, it was 3.78 (%) (Table 1). The PR was 3.23 and showed a positive statistically significant association between exposure and subclinical malarial infection (CI 95% = 2.03, 5.13) (Table 1).

Of the subclinical malarial infections (84), 57 were P. falciparum, 25 were P. vivax, and two were mixed (P. vivax and P. falciparum). Considering only P. falciparum subclinical infections, 53 blood donors of the exposed group were positive for subclinical infection, 447 were exposed and did not have any evidence of infection, 4 individuals of the non-exposed group tested positive, and 604 tested negative for infection (Table 2). The total prevalence of subclinical P. falciparum infections in blood donors was 5.14 (%); the prevalence among exposed individuals was 10.60 (%); and among non-exposed, it was 0.66 (%) (Table 2). The prevalence ratio was 16.11 and showed a positive, statistically significant association between exposure and subclinical P. falciparum infection (CI 95% = 5.87, 44.21) (Table 2).

Considering the analysis of Plasmodium vivax subclinical infections, seven exposed participants had a subclinical infection, 493 were exposed without subclinical infection, 18 non-exposed had subclinical infection, and 590 non-exposed did not have subclinical infection (Table 3). The total prevalence of subclinical P. vivax infection among blood donors was 2.26 (%); prevalence among exposed individuals was 1.40 (%), and among non-exposed, it was 2.96 (%) (Table 3). The PR was 0.47 and showed a negative statistically non-significant association between exposure and subclinical P. vivax infection (CI 95% = 0.2, 1.12) (Table 3).

The 18 individuals with P. vivax subclinical infection in the non-exposed group (Table 3) were geo-referenced. These data showed that the majority of participants lived in residences near the forest fringes (Figure 1A), although these residences were in an urban area, i.e., the São Paulo Metropolitan Region. Residents infected with P. falciparum in the exposed region (Atlantic Forest) were from the mountains (mainly in the Juquitiba and São Lourenço da Serra municipalities). In both municipalities, forest cover is greater than 50%, and landscape fragmentation provides high levels of interaction between the forest and the urban, anthropogenic environment (Figure 2B). Individuals infected with P. vivax in the exposed region were from the coastal region of the Peruíbe (Figure 2C) and São Sebastião (Figure 2D) municipalities. These individuals inhabit houses located at the frontier of the Atlantic Forest and the urban landscape, which form a mosaic of human-modified and natural environments that facilitate contact between humans and sylvatic mosquitoes (Figures 2C and D).

Despite control and prevention measures, malaria has not been eliminated from non-endemic areas in southeastern Brazil. Autochthonous cases have been identified in several regions [33‐35], and human malaria is predominantly caused by P. vivax and its variants [36, 37]. Subclinical Plasmodium infections have been considered rare in Brazil; however, they have been detected and identified in endemic regions [38]. A retrospective analysis of human malaria in the State of São Paulo from 1985–2006 found 83 cases, four of which were subclinical infections [12]. Other studies have shown a low parasitaemia rate in humans who inhabit areas within the Atlantic Forest, with cases caused by P. vivax and P. malariae[36].

Recently, molecular tools have been employed to detect Plasmodium infection in human blood. Although PCR techniques do not represent a rapid test for malaria diagnosis, they allow detection of as little as 0.5–1 parasite/μL blood, making them 20 times more sensitive than thick blood smear microscopic examination [30]. PCR methods are also more sensitive in verifying mixed infections [39]. Additionally, real-time PCR allows pathogen quantification, reduces the risk of cross-contamination, provides accurate results [40], and identifies subclinical infections in humans [37] that might otherwise remain undetected by routine tests for malaria. Consequently, the new technology reveals a prevalence of subclinical infections in humans that is higher than previously thought [41].

Human reservoirs can maintain transmission in areas of low, moderate, or high endemicity because they can infect mosquitoes [42]. Despite the role of subclinical infection in maintaining malaria transmission, little is known about the factors that determine its occurrence and whether it is associated with human protective immune responses [41].

Blood transfusion is an important intervention in hospitals and health centres, but it also represents a route for transmission of parasites and pathogens that circulate in the blood, such as HIV, human Plasmodium, viruses, and others [43]. Consequently, several measures are being taken to improve blood recipient safety, including high-sensitivity screening tests. Considering only Plasmodium, after the collection of the blood, the parasite can survive for at least one week at 4°C, as well as in frozen erythrocytes [44]. Plasmodium transmission by transfusion usually occurs through whole blood and red blood cells and may occur less frequently through platelet concentrates, white blood cells, cryoprecipitate, and fresh frozen plasma [18, 45]. The blood bank is therefore an important target for controlling undesirable Plasmodium transmission in endemic [38, 46] and non-endemic regions [43, 47].

Here, subclinical malarial infections were detected from blood donors in the most important centre of public healthcare in the State of São Paulo. The presence of P. falciparum among blood donors and the high prevalence of these infections were not expected. In the present study, each blood donor was interviewed and denied having been in malaria endemic regions. Consequently, it is plausible to assume that all subclinical infections were autochthonous. It is noteworthy that all individuals reported here would be accepted for blood donation using the current protocol adopted in blood transfusion centres in the State of São Paulo, which includes a questionnaire and an interview to detect history of exposure to malaria, both applied by a qualified professional. Therefore, although the clinical importance of the findings described here remains to be investigated, the presence of asymptomatic Plasmodium infection in blood donors in non-endemic areas should be an important concern for medical services, health authorities, and blood recipient safety [14].

Plasmodium vivax and its variants as well as P. falciparum and P. malariae circulate in non-endemic areas, where they are responsible for only a few, mostly oligosymptomatic cases of malaria annually [1]. Because Plasmodium species can circulate in cycles that involve humans with low parasitaemia and sylvatic mosquitoes, and likely in non-human primate reservoirs [48, 49], malaria can be maintained indefinitely in silent cycles without the consequences of high-level epidemics. The estimated PR of subclinical infection found in blood donors in São Paulo clearly shows that P. falciparum is positively associated with forested areas. Juquitiba (30,239 inhabitants, 55.03 inhabitants/km²) and São Lourenço da Serra (14,874 inhabitants, 75 inhabitants/km²) are primarily rural locations embedded in a well-preserved and extensively forested area on the slope of Serra do Mar, where the vectors are mainly sylvatic mosquitoes from the subgenus Kerteszia of Anopheles. The human population from these areas was here defined as the exposed population whereas inhabitants of urban areas intercalated by forest fragments formed the non-exposed population. In the non-exposed population, the PR of P. vivax revealed a positive (PR = 0.47; 95% CI, 0.2, 1.12) between the parasite infection and urban inhabitants. However, this result may arise from the fact that the residences of the blood donors are situated in urban areas at the edge of forest fragments. Consequently, the presence of P. vivax is likely to be associated with forest fragments and involves both Kerteszia and Nyssorhynchus species, or even other Anopheles subgenera [48, 50]. The evolution of transmission of Plasmodium species that cause malaria in humans is a dynamic process that involves ecological, environmental, and climate factors as well as social, political, and economic determinants. The high prevalence of Plasmodium spp. subclinical infections in blood donors living either in forested areas or urban regions intermixed with forest fragments suggests that some evolutionary mechanisms are modulating the virulence of P. falciparum and P. vivax in a way that does not negatively affect survival of the parasite, humans, mosquitoes, and non-human primates that seem to be involved in malaria transmission in the Atlantic Forest. The dynamics of human malaria transmission in the Brazilian Amazon forest seems to involve similar evolutionary mechanisms that lead to subclinical infections in humans [42], and transmission cycles that involve humans and other mosquito species as vectors (in addition to An. darlingi) [51], and non-human primates, which were recently found to be infected with P. falciparum[52]. In the Atlantic Forest, Alouatta monkeys have been found to be infected by P. vivax and P. malariae[49], along with P. falciparum[48].

The scientific novelty and translational importance of the present results are two-fold: 1) They could represent a serious problem for the blood transfusion programme in the State of São Paulo; and 2) they could be evidence of evolutionary mechanisms of attenuation of malarial symptoms in humans.

The translational importance of the first aspect is obvious. Because some infected recipients may become asymptomatic carriers and constitute a reservoir for parasites, maintaining their transmission, further investigations are needed to better document the risk of transfusion-transmitted malaria in these regions and to establish whether a specific malarial infection test should be considered for blood donors who live in or near the Atlantic Forest. The second aspect is more subtle and needs more data before being established as a paradigmatic cornerstone for the evolution of malaria transmission dynamics. In terms of clinical and economic viewpoints, the Atlantic Forest’s dynamics of malaria transmission could be a real-world example of symptomatic malaria elimination.

The presence of subclinical P. falciparum and P. vivax infection in healthy blood donors from São Paulo a non-endemic region was described. This finding is a significant concern in the context of transfusion biosafety. Some infected recipients may be asymptomatic carriers, constituting a reservoir for parasites, maintaining their transmission and increasing the risk of transfusion-transmitted malaria in these regions. Additionally, prevalence ratio infection values showed a positive association between forest environment and P. falciparum infection, which was linked to residence in the mountain region of the Atlantic Forest domain. Moreover, P. vivax seemed to be associated with forest fragmentation because it was identified in residents of cities in forest-surrounded areas. These results indicate a link between the pathogen prevalence and the Atlantic Forest environment.