Asymptomatic parasitaemia in a malaria pre-elimination setting in the Putumayo area of the Ecuadorian Amazon basin Authors: Jacob van der Ende, M. Vanessa Dávila, Sandra Menting, Fabián E. Sáenz, Thomas Hanscheid, Henk D. F. H. Schallig, Martin P. Grobusch Published: 2025 ABSTRACT BackgroundEcuador is approaching malaria elimination, but its Amazonian Putumayo canton borders regions in Colombia and Peru with high transmission intensity. This ‘frontier’ setting represents a critical vulnerability where asymptomatic infections may sustain transmission, undetected by routine surveillance.MethodsIn a repeated cross-sectional study during the 2023 rainy season, we surveyed 293 asymptomatic individuals in four riverine communities in Putumayo. We used a tiered diagnostic approach: rapid diagnostic tests (RDTs), expert microscopy, and real-time qPCR targeting the 18S rRNA gene.ResultsNo infections were detected by RDT (0/293). Microscopy identified three P. vivax infections (1.0%, 95% CI 0.2–2.4%). qPCR detected five positive cases (1.7%, 95% CI 0.6–3.9%), revealing two additional submicroscopic infections. The failure of a subsequent species-specific PCR and high Cq values (mean 32.6) indicated very low parasite densities. Infections were found only in the mid-to-late rainy season, predominantly in males.ConclusionsMolecular surveillance revealed a low-density, asymptomatic malaria reservoir missed entirely by RDTs, and partially by microscopy. This hidden burden represents a critical blind spot for Ecuador’s elimination program. To achieve and sustain elimination, surveillance in such border regions must incorporate highly sensitive molecular tools and be supported by robust cross-border collaboration. Introduction Despite global progress in malaria control, the Amazon Basin remains a primary reservoir of transmission in the Americas. In 2023, the WHO Region of the Americas reported an incidence of 3.6 cases per 1000 population, well above the Global Technical Strategy target for 2025, and this aggregate figure conceals marked sub-national heterogeneity in Amazonian borderlands [1, 2]. Ecuador illustrates this mismatch between overall national gains and local persistence. The Ministry of Health has released funds by 2025 to finance RDTs, new microscopes and mosquito nets with the ambition to be malaria-free as soon as possible [3]. After reaching a historic low of 242 autochthonous cases in 2014, case numbers rebounded to 2175 in 2021 before declining to 604 in 2023 [1]. Transmission is now largely confined to specific foci, particularly in the Amazonian provinces of Pastaza and Sucumbíos, where malaria hospitalization rates can exceed those of some coastal areas [4, 5]. This residual transmission is shaped by cross-border epidemiology. In Peru, over 96% of cases occur in the Amazon, with Loreto as the epicenter; some Peruvian districts adjacent to the Ecuadorian border report incidence rates above 200 per 1000 inhabitants [6]. Likewise, Colombia’s Amazonian departments, including Putumayo (the Colombian department bordering with Ecuador), maintain persistent, hypoendemic transmission dominated by Plasmodium vivax [7]. This species accounts for approximately 88% of cases in Ecuador and 73% region-wide [1, 8]. Two reports from Ecuador confirm a predominance of 10 to 1 (P. vivax to P. falciparum), one from the Amazon region published over 25 years ago and more recently from the coastal region of Ecuador [9, 10]. Ecuador’s Putumayo canton exemplifies ‘frontier malaria’ [11]: a persistent transmission context shaped by porous borders, ecological continuity, and intense human mobility. The Putumayo and San Miguel rivers serve as transnational transport corridors, enabling near-continuous movement of people for economic, subsistence, and social purposes, which facilitates cross-border Plasmodium spp. circulation [12, 13]. Spatiotemporal analyses confirm that this mobility can sustain genetically identical parasite strains across distances exceeding 300 km over several months [12]. Extractive industries, mining, logging, and oil exploration, further amplify risk by attracting a mobile labor force into remote, under-resourced zones with limited health care coverage [14, 15]. This epidemiological challenge is compounded by a critical surveillance blind spot. Routine tools like microscopy and RDTs consistently underestimate the true malaria burden by failing to detect the low-density, asymptomatic infections that characterize pre-elimination settings, and which are only detectable by highly sensitive molecular methods [16, 17]. The presence of HRP2/3 gene deletions in the region may further compromise RDT-based surveillance for P. falciparum [1, 18]. In low-endemicity settings, this diagnostic gap is particularly dangerous, allowing residual transmission to persist undetected and potentially new outbreaks to occur [16, 19]. The border zone of Putumayo, marked by riverine connectivity and shared vector ecology with neighboring Colombia and Peru, is especially vulnerable to such hidden reservoirs. Yet, despite its strategic importance, no molecular epidemiological data have been generated from this area. To address this gap, this study aimed to assess malaria prevalence in four remote riverine communities in Putumayo canton (Fig. 1) using a tiered diagnostic approach (RDTs on site of sample collection, microscopy at the hospital laboratory, and qPCR later in the Amsterdam reference laboratory). The objective was to provide a direct estimation of the infectious reservoir, including the proportion of infections undetectable by routine methods, in order to inform the active surveillance strategies necessary for sustainable malaria elimination in Ecuador and the broader Amazon Basin [14].Fig. 1A Geographical location of Putumayo and B the communities visited and dates of when a positive sample was found B A Geographical location of Putumayo and B the communities visited and dates of when a positive sample was found B Methods Study design, setting, and participants This descriptive, repeated cross-sectional study was conducted in the Putumayo canton, a remote Amazonian border region in northeast Ecuador (Fig. 1). To account for seasonality, sampling was performed at three timepoints during the 2023 rainy season: start (May), mid (August), and end (November–December). We selected four communities based on their location: two border communities (El Litoral; the Puerto Rodriguez cluster) and two centrally-located communities (Singue; Puerto el Carmen). Although sampling was organized across four community clusters, three administrative sub-communities (Tres Fronteras, Buen Samaritano, Bajo Rodriguez) belong to the wider Puerto Rodriguez cluster, resulting in six named localities (Table 1). Table 1Details of communities visitedCommunitySamples collected N (month)N (%) of community sampledTotal inhabitantsDistance to Puerto el CarmenPuerto el Carmen48 (May), 35 (Aug), 22 (Dec)3.4%30900 kmEl Litoral10 (May), 10 (Aug), 17 (Dec)100%30250 km WestTres Fronteras47 (July)49%95250 km EastBuen Samaritano40 (July)8%495250 km EastBajo Rodriguez29 (Nov)6%447250 km EastSingue35 (Dec)58%5885 km SouthwestThe communities of Tres Fronteras, Buen Samaritano and Bajo Rodriguez all three belong to the larger community of Puerto Rodriguez Table 1Details of communities visitedCommunitySamples collected N (month)N (%) of community sampledTotal inhabitantsDistance to Puerto el CarmenPuerto el Carmen48 (May), 35 (Aug), 22 (Dec)3.4%30900 kmEl Litoral10 (May), 10 (Aug), 17 (Dec)100%30250 km WestTres Fronteras47 (July)49%95250 km EastBuen Samaritano40 (July)8%495250 km EastBajo Rodriguez29 (Nov)6%447250 km EastSingue35 (Dec)58%5885 km SouthwestThe communities of Tres Fronteras, Buen Samaritano and Bajo Rodriguez all three belong to the larger community of Puerto Rodriguez Details of communities visited The communities of Tres Fronteras, Buen Samaritano and Bajo Rodriguez all three belong to the larger community of Puerto Rodriguez Following announcements by community leaders, participants were recruited via convenience sampling. Eligible individuals aged three years or older who were asymptomatic at the time of sampling were invited to participate. Participants were considered asymptomatic if they reported no fever (or history of fever within the past 48 h) and exhibited no other acute signs and symptoms of malaria (e.g., fever, chills, headache, malaise). Written informed consent was obtained from all adult participants and from the parents or legal guardians of minors. According to protocol, individuals encountered during screening who would report ill with signs and symptoms suggestive of malaria would have been tested and if possible treated. Sample collection and laboratory procedures From each participant, a 3 mL EDTA blood sample was collected for a three-tiered diagnostic assessment: RDT, microscopy, and qPCR. On-site, a CareStart™ Malaria HRP2/pLDH (Pf/PAN) combo RDT (LOT RMRM −02571) was performed according to the manufacturer’s protocol. Results were read independently by two trained field staff, with a third reader resolving any discrepancies. For microscopy, thin and thick blood smears were prepared at the Hospital San Miguel laboratory. Thin smears were fixed with absolute methanol, and slides were stained with a 10% Giemsa solution (pH 7.0) for 15 min. Two expert microscopists independently examined each thick smear for at least 200 high-power fields (HPFs), corresponding to the common limit of detection (LOD) under normal circumstances of 20–50 parasites/µL. Species identification was performed on the corresponding thin smear for any positive case. For molecular analysis, dried blood spots (DBS) were prepared on Whatman 903 Protein Saver Cards (Whatman plc, Maidstone, UK), air-dried, sealed and shipped at ambient temperature with a transport time from Ecuador to the Amsterdam University Medical Centre of two days. DNA was extracted using an automated system (NucliSENS EMAG; bioMérieux) for qPCR analysis (CFX96 Touch Real-Time PCR Detection System; BioRAD). Samples were first screened with a real-time qPCR targeting the multi-copy Plasmodium 18S rRNA gene, with a low positive Plasmodium falciparum control of 10 parasites/µL (LOD). The results for the pan-Plasmodium assay were recorded as cycle quantification (Cq) values. A second, single nested endpoint PCRs were used to identify P. falciparum, P. vivax, P. malariae, and P. ovale in positive samples with low parasitemia. Data analysis and ethical considerations Prevalence and 95% exact binomial confidence intervals (CIs) were calculated for each diagnostic method. Due to the low number of positive cases, formal statistical testing was not performed, and data were analyzed descriptively. The study was approved by the ethics committee of Universidad San Francisco de Quito (CEISH; 0-IE02-JP184-2022-P22080M). Any individual who would have tested positive by RDT on-site was to be offered standard treatment according to national guidelines. Results Participant enrolment began on 14 May 2023, at the onset of the rainy season, and concluded 206 days later on 6 December 2023. Due to convenience of sampling, community representation varied substantially, from only 1% of the population in Puerto El Carmen (35/3,090) to 58% in Singue (35/60; Table 1). Among the 293 participants included, the mean age was 32 years; 33% were younger than 18 years, and 64% were male. All 293 rapid diagnostic tests (RDTs) targeting HRP2 and pLDH antigens were negative. This was confirmed by two independent, blinded readers, resulting in an observed RDT-based malaria prevalence of 0.0% (95% CI 0.0–0.0%). Accordingly, no treatment for malaria was prescribed. Microscopic examination of thick blood smears identified three positive samples, corresponding to a prevalence of 1.0% (95% CI 0.0–2.4%). Thin smear analysis identified all three infections as Plasmodium vivax, with parasite densities estimated at < 1 parasite/µL. Two reader discrepancies were resolved by a third, independent reader. Two microscopy-positive cases were male and one female, aged 5–15, 30–45, and > 45 years, respectively. Pan-Plasmodium 18S rRNA qPCR was positive in 5/293 samples, corresponding to a prevalence of 1.7% (95% CI 0.3–3.4%). The three microscopy-positive samples were also qPCR-positive. Two additional infections were detected only by qPCR, both in male participants, one a 6-year-old from Bajo Rodríguez and a 33-year-old from Singue. Cq values ranged from 31.1 to 33.8, with a mean of 32.6. The mean Cq value of the low Pf positive control was 33.55. All five qPCR-positive cases occurred during the middle (July–August) or end (November–December) of the rainy season; none were identified early in the season. In total, one infection was detected in a female participant and four in males. Age distribution included one child (5–15 years), two adults aged 30–45 years, and two individuals aged > 45 years. Geographical distribution was broad, with one case each detected in Puerto el Carmen, Bajo Rodríguez, Singue, Tres Fronteras, and El Litoral. Full demographic and diagnostic details of the five qPCR-positive individuals are shown in Table 2. Table 2Demographic and diagnostic characteristics of qPCR-positive malaria infectionsPatientAgeSexCommunityDate sampledCt valueMicroscopy153MTres Fronteras2023-07-2831.57Positive (1/µL)249MEl Litoral2023-08-1633.84Positive (1/µL)342FPuerto el Carmen2023-08-1831.09Positive (1/µL)46MBajo Rodríguez2023-11-2932.61Negative533MSingue2023-12-0633.79Negative(n = 5; all pan-Plasmodium 18S rRNA qPCR Cq < 34) Table 2Demographic and diagnostic characteristics of qPCR-positive malaria infectionsPatientAgeSexCommunityDate sampledCt valueMicroscopy153MTres Fronteras2023-07-2831.57Positive (1/µL)249MEl Litoral2023-08-1633.84Positive (1/µL)342FPuerto el Carmen2023-08-1831.09Positive (1/µL)46MBajo Rodríguez2023-11-2932.61Negative533MSingue2023-12-0633.79Negative(n = 5; all pan-Plasmodium 18S rRNA qPCR Cq < 34) Demographic and diagnostic characteristics of qPCR-positive malaria infections (n = 5; all pan-Plasmodium 18S rRNA qPCR Cq < 34) Single nested endpoint PCR assays targeting P. vivax, P. falciparum, P. ovale, and P. malariae were negative in all five qPCR-positive samples, including those that were microscopically confirmed as P. vivax. This may reflect very low parasite densities or sequence divergence affecting primer binding. Discussion This study provides the first molecular evidence of an asymptomatic malaria reservoir in Ecuador’s Putumayo canton. While the overall prevalence was low (1.7% by qPCR), our findings reveal a hidden burden of infection that routine surveillance methods would fail to detect. All these infections were missed by RDTs, and nearly half were missed by expert microscopy. This demonstrates a critical surveillance blind spot in a strategic border region, with significant implications for Ecuador’s malaria elimination goals. Our study’s central finding is the diagnostic cascade where qPCR detected more infections than microscopy, which in turn detected more than RDTs. The complete failure of RDTs is likely explained by their detection limits, typically above 100–200 parasites/µL, which is too high for the low-density infections we found [20, 21]. The fact that two cases were also microscopy-negative suggests parasite densities below the ~ 20–50 parasites/µL overall threshold of expert microscopy. However, as observed in our study, parasitaemias below the LOD can be detected occasionally, which usually occurs when rare or single parasites are detected and the number of visual fields screened is extended. These < 1 parasite/µL values are estimates extrapolated from parasite counts per number of microscopic fields observed, not absolute quantifications. This pattern is highly consistent with evidence from across the Amazon; studies in Colombia found that up to 90% of infections in low-transmission areas were asymptomatic and detectable only by molecular methods [17], while research in coastal Ecuador revealed 7.4% PCR positivity among microscopy-negative samples [9]. While the number of cases is too small for statistical analysis, the finding that four of five infections were in males aligns with regional patterns where occupational exposure, such as in mining or logging, often drives transmission [14, 15]. This trend reinforces that active surveillance strategies should consider targeting mobile, working-age populations who may have limited contact with traditional health services. The Putumayo and San Miguel rivers facilitate constant population movement, creating a single epidemiological zone where parasites circulate freely across borders (Fig. 1). This mirrors the situation on Ecuador’s southern border, where a binational initiative with Peru successfully eliminated malaria on the Ecuadorian side, demonstrating the potential for coordinated action [22]. The northern border, however, poses a distinct challenge due to the high malaria burden in the Colombian Pacific coast [18]. While Colombia’s highest malaria burden is concentrated on its Pacific coast, the persistent, hypoendemic transmission in its adjacent Amazonian departments creates a continuous source of parasite importation that threatens Ecuador’s gains [7, 8]. Microscopy identified P. vivax in the positive samples, reflecting the regional pattern where this species accounts for 88% of Ecuador’s malaria burden [1]. Plasmodium vivax is a major obstacle to elimination due to its ability to form hypnozoites that cause relapsing infections. With 12% of Ecuador’s 2023 cases documented as relapses [1], effective radical cure with primaquine or tafenoquine is paramount. The successful deployment of radical cure therapies for P. vivax is contingent upon overcoming logistical and regulatory hurdles that can limit the availability of essential medicines in front-line health facilities. Our findings add weight to the argument for shifting from passive case detection to active surveillance in foci like Putumayo. Strategies such as reactive case detection (RACD) or focal screening and treatment (FSAT) could effectively target this hidden reservoir [23, 24]. The observation that all infections were detected in the mid-to-late rainy season could help in timing these interventions to maximize yield. Implementing such strategies is challenging given Ecuador’s domestic malaria spending of just US$0.15 per person at risk, the lowest in the region [1]. However, cost-effectiveness analyses could demonstrate that a greater upfront investment in sensitive surveillance provides a strong return by preventing more costly outbreaks. This study has some limitations. The convenience sampling method resulted in variable community coverage, with differences not due to different levels of consent but to work-related or distance-related absences. Our prevalence estimates may not be generalizable across the entire Putumayo canton. Furthermore, the small number of positive cases precluded statistical analysis to better characterize populations at risk. Most importantly, the failure of our species-specific PCR assay to confirm the species of the qPCR-positive samples is a key finding. This is best explained by the extremely low parasite densities (mean Cq 32.6) falling below the assay’s detection limit. Our pan-Plasmodium assay targets the multi-copy 18S rRNA gene, which provides more DNA template per parasite and is thus inherently more sensitive than species-specific assays that often target single-copy genes [25]. Because our species-specific qPCR used only small volume of blood, its effective limit of detection was roughly 200–1 000 parasites mL⁻1; high-volume qPCR can reach 22 parasites mL⁻1 [26]. Our five Cq < 34 positives therefore likely represent just the tip of a larger, sub-microscopic reservoir that evades both RDTs and conventional PCR. Future studies in such settings may need to employ methods that increase the starting amount of target DNA, for instance by using larger blood volumes. Conclusions This study confirms that a low-density, asymptomatic malaria reservoir persists in Ecuador’s Putumayo canton, maintaining a source of transmission that is invisible to routine surveillance. Achieving and sustaining malaria elimination will require acknowledging this hidden burden and implementing active, molecular-based surveillance strategies with a low limit of detection, improving access to radical cure for P. vivax, and strengthening cross-border collaboration. While resource constraints are significant, economic analyses would likely show that the long-term cost of inaction far exceeds the investment required for proactive, sensitive surveillance to protect the hard-won gains against malaria.