We conducted a study in 5 provinces (Sucumbíos, Orellana, Napo, Pastaza, and Morona Santiago) in the Amazon rainforest region of Ecuador for which malaria slides were available. The Amazon region of Ecuador covers ≈40% of the area of this country, extends from the eastern Andes to the lowlands of the Amazon basin, and borders Colombia and Peru.

The study population was composed of mestizos and Kichwa, Shuar, and Achuar indigenous groups. Community health workers collected blood samples by finger prick from 7:00 am until 7:00 pm from persons suspected of having malaria. Malaria centers are located throughout the rainforest in an ongoing program for malaria control under the guidance of the Ministry of Public Health.

Thick and thin blood smears were stained with Giemsa and viewed by microscopy at 100× magnification under oil immersion for Plasmodium spp. parasites. We retrospectively screened all slides obtained during 2014–2015 for Mansonella spp. nematodes.

Although >5,000 stained blood smears were reviewed, only 2,756 slides could be read because of poor preservation. We examined thick and thin blood smears by using light microscopy and a 20× objective lens to detect microfilariae and a 40× objective lens for species identification. No epidemiologic data were available because we conducted a retrospective analysis.

Of 2,756 Giemsa-stained blood smears examined, we detected microfilariae of Mansonella spp. nematodes in 566 (20.5%). Microfilariae were unsheathed (average length 155 µm–212 µm) and had tapered, nonnucleated tails; anterior extremities ended in cephalic spaces. On the basis of these morphologic characteristics, we identified all filarial infections as M. ozzardi nematodes. Many microfilariae appeared damaged and partially destroyed. No microfilariae with the characteristics of M. perstans nematodes were observed. Infection rates between provinces ranged from 5.2% to 36.5%; infections were most prevalent in Morona Santiago, which borders Peru (Figure 1).

We then extracted DNA from all microscope slides positive for microfilariae. We scraped blood films off the slides into 70% ethanol. After microcentrifugation at 8,000 rpm for 2 min, we subjected supernatants to DNA extraction by using Chelex treatment (7), followed by proteinase K digestion. We performed a nested PCR according to the method of Tang et al. (6). This PCR amplifies the internal transcribed spacer region 1 of the rDNA gene of filarial species. The size of this region varies among O. volvulus, M. ozzardi, and M. perstans nematodes, and the PCR yields amplicons of different sizes for each species (6). Expected product sizes were 305 bp for M. ozzardi nematodes and 312 bp for M. perstans nematodes. PCR products after the second amplification were subjected to electrophoresis on 2% agarose gels (Figure 2).

We report a high frequency of autochthonous human mansonelliasis caused by M. ozzardi nematodes in the Amazon region of Ecuador. These parasites were identified by microscope identification of characteristic microfilariae and detection of species-specific DNA.

This study showed that human infections with M. ozzardi nematodes are highly prevalent throughout the Amazon region of Ecuador. High rates of circulating microfilariae are strongly suggestive of active local transmission, particularly because known vectors (biting midges of the genus Culicoides and black flies of the genus Simulium) are present in the surveyed region. Further entomologic studies are needed to identify specific Diptera species involved in local transmission of M. ozzardi nematodes in Ecuador.

Prevalence of M. ozzardi infection ranged from 5.2% to 36.5% in this study, similar to infection rates reported in bordering regions of Colombia and Peru, indicating that this region is probably a large focus of infection extending throughout several countries in the Amazon region (2). M. ozzardi infection rates were higher than those reported from Brazil, but considerably lower than those for the indigenous population of El Vaupés in Colombia, where the infection rate was 96% (8). However, in estimating prevalence by using stored, partially degraded blood smears and light microscopy, we might have underestimated the true prevalence. Use of FTA cards (Whatman, https://www.sigmaaldrich.com) has been demonstrated to be more sensitive for detection of M. ozzardi nematodes (9).

The morphologic features of microfilariae we observed were typical for M. ozzardi nematodes, and nested PCR confirmed detection of a DNA fragment (305 bp) known to be specific for this parasite. This assay used universal filariae PCR primers to amplify a variable portion of filarial parasite internal transcribed spacer 1 rDNA gene and enabled subsequent differentiation of species on the basis of the size of the amplified fragment (6). In addition, DNA was successfully extracted from scrapings of thick and thin dried blood films, even after long storage periods of 2–3 years.

We observed no microfilariae of M. perstans nematodes, and no positive samples showed DNA fragment sizes representative of this nematode (6). M. perstans nematodes have been reported only in the northern part of the Amazon rainforest from equatorial Brazil to the Caribbean coast of South America (1). In Colombia, M. perstans nematodes have been observed in a restricted area among the Curripacos Amerindians in the Comisaría del Guainia region bordering Venezuela (10), but have not been found in the Peruvian Amazon (1,2). However, further prospective studies using molecular techniques, are needed to clarify the epidemiologic status of this parasite.

In summary, we retrospectively identified human infection with M. ozzardi nematodes in the Amazon region of Ecuador. Our findings confirm that this parasite is endemic to this region.