Mansonella ozzardi is a filarial parasite autochthonous in the American Continent, and its distribution is reported in the Caribbean Islands, Central and South America. In Brazil, it was first detected in blood samples from inhabitants of Manaus (Amazonas State) by Maria Deane in 1949. Thereafter, further studies were carried out in order to provide data about distribution and prevalence of this filarial nematode in our country. Thus, human infections are frequently found in the State of Amazonas-along Solimões, Purus, and Negro Rivers, its most important area of distribution (Batista et al. 1960). In this region, transmission is related to the blackflies Simulium amazonicum and Simulium argentiscutum whose distribution in the area frequently overlaps (Shelley et al. 1980; Moraes et al. 1985). Although the larval stages are found in blood samples of infected individuals, sometimes with high parasitic loads, the parasite is not considered of medical importance as its pathogenicity is still subject of controversy in the literature (Deane et al. 1953; Rachou 1957; Restrepo et al. 1962; Jörg 1983).

Experimental infection in Erythrocebus patas has been successful in obtaining adult worms of M. ozzardi (Orihel et al. 1993), but these forms collected from human beings are not well described and its real localization in human tissues is still obscure. The difficulty to obtain adult worms from this parasite turns its diagnose based only in the morphology of the larval stages-the microfilaria-that can be easily found into the peripheral blood of infected individuals.

Unlike M. ozzardi, Onchocerca volvulus has pathogenicity and distribution well established and human infections by this parasite are found among Yanomami Indians from border region with Venezuela in the States of Amazonas and Roraima (Shelley 2002).

Adult stages of O. volvulus may be found and extracted from subcutaneous nodules where they are dwelling and its larval stages lyes in to the skin but it was demonstrated that the parasite may be found in the peripheral blood of up one third of infected people in hyperendemic communities (WHO 1987). M. ozzardi presence have been reported to occur in samples of the skin (Moraes et al. 1983), and this was already cause of some misunderstandings in the past.

The standard methods employed for detecting M. ozzardi and O. volvulus infections in humans are, thus, based on the detection of the microfilarias in blood and skin samples, respectively, and the morphological discrimination of them. The characteristics of a typical microfilaria are well described (Orihel and Eberhard 1982) and some anatomic markers such as body length, cephalic space, shape of body nucleation, nerve ring, terminal nucleation of the tail, and tail space can be discerned through light microscopy studies (Kozek et al. 1983).

During a malaria survey conducted in some communities along Acre River, serum samples were collected and sent to our laboratory in order to evaluate its reactivity in an enzyme-linked immunosorbent assay test with Ov10, Ov11, and Ov16 recombinant proteins from O. volvulus. Surprisingly, 40% of the samples from a community named Vila Antimary (a non endemic area for O. volvulus) had positive reactions. Although these tests cannot be used to define the presence and level of infection, they are effective as an epidemiological tool to provide information about incidence of exposure and, therefore, transmission of the parasite (Bradley and Unnasch 1996). The possibility of O. volvulus presence along Acre River was subsequently investigated by Marilza Maia-Herzog and Anthony J. Shelley in the area and they found microfilarias in skin snips from human volunteers morphologically similar to M. ozzardi and with some characteristics movements of O. volvulus (personal communication). Unfortunately, some difficulties were faced in the field mainly due to the precarious conditions in the area and this material was lost during the process of fixation. However, some part of the collected skin snip was sent to Dr. Rory Post at the Natural History Museum in London for polymerase chain reaction examination, and parasite DNA studies showed only M. ozzardi in the area (Shelley 2002).

The present report involves an active filarial epidemiological survey carried out on the border of Amazonas and Acre States along the Antimary, Acre, and Purus Rivers and a light microscopy study of the microfilarias found in the area.

Study Area: Mapinguari (9°24′08″ S 68°08′51.6″ W), Andaraí (9°17′29.8″ S 67°45′44.0″ W), Porto Acre (9°35′42.8″ S 67°32′35.9″ W), Vila Antimary (9°04′01.0″ S 67°23′50.1″ W), and the Kamikuã Indian Village in Boca do Acre (8°45′07″ S 67°23′52″ W) were the selected communities along Antimary, Acre, and Purus Rivers (Fig. 1)-a conflicting taxonomic area for O. volvulus and M. ozzardi-whose distribution and occurrence had not been previously described.

Fig. 1 Geographical localization of the studied area

---

Venous blood samples collections were performed in a 10-ml sterile and disposable syringe and thick blood films were prepared at the moment of collection followed by dehemoglobinization, fixation in methanol and staining with Giemsa in order to detect M. ozzardi infections. To minimize false negative results, 1 ml of venous blood was deposited in a polystirene tube with 10 ml of a 2% formaline solution for Knott's Method (Knott 1939) and after a period of 12 h, thin films were prepared with the deposited sediment, fixed with methanol, and stained with Giemsa.

O. volvulus infections were searched through skin biopsies which were taken from scapula and iliac crest using a Holth^R Sclerocorneal Punch. Each skin snip was placed in a slide with distilled water for 6-12 h to permit microfilaria to migrate from the tissue. Then, the skin snip was removed and placed in numbered Eppendorffs with absolute ethanol (for further molecular analysis), and the slide was allowed to air-dry, fixed with methanol, and stained with Giemsa. In order to avoid losses during material preparation, distilled water and reagents were carried by our group to the field.

All specimens were returned to our laboratory at FIOCRUZ/Rio de Janeiro for accurate examination.

Blood and skin snips samples were collected from volunteers of Mapinguari (n = 67), Porto Acre (n = 129), Andaraí (n = 10), Vila Antimary (n = 56), and the Kamikuã Indian Village (n = 7).

The study design, including its ethical aspects, was reviewed and approved by an Ethical Comitee from FIOCRUZ (Parecer 281/05).

M. ozzardi microfilaria have an anterior region with a cephalic space that ends when the nuclear column begins, with an initial detailed nucleus followed by two that seems attached. Then, the nuclear arrangement proceeds with two or three paired nuclei until they reach some discontinuities, which correspond to locations as nerve ring and excretory vesicle (Fig. 2a). Posterior extremity with an initial discontinuity-corresponding to the anal vesicle-with nuclei apparently attached to each other, paired or even in single line but these structures do not reach the tip of the tail. On the contrary, seven to nine nuclei form a line and an annucleate region is attained-the caudal space that finishes in a hook shape (Fig. 2b).

Fig. 2 Comparison of typical M. ozzardi microfilaria and the atypical one found in blood samples from human volunteers in Vila Antimary. a Characteristic M. ozzardi microfilaria showing cephalic space and b detail of the tail. c Anterior region of atypical microfilaria with two paired nuclei at the anterior region and d detail of tail and caudal space (scale bars = 7 μm)

---

However, 12.5% of the volunteers infected and enrolled in agricultural activities from Vila Antimary presented besides M. ozzardi, an unusual form of microfilaria whose main difference could be noted at the anterior region (Fig. 2c). These forms have two paired nuclei after the cephalic space, followed by a single one, and afterwards begin the proceeding characteristic column of nucleus. Meanwhile, no differences could be noted at the caudal region; that is, they were very similar to the pattern of nucleation with a column of nucleus close to the parasite cuticle that elongate as long as they reach the tail tip (Fig. 2d).

Analysis of the material obtained through light microscopy study from the skin snips revealed no infections with O. volvulus (data not shown).

Epidemiological studies conducted by our group in the region started in 2006, and the goal was to show distribution and occurrence of filarial infections in communities along Antimary, Acre, and Purus Rivers. This is the first report of M. ozzardi in this specific region.

Up to now, Acre State presented the lowest rates as demonstrated by our results in Mapinguari and Porto Acre, but the prevalence increased along the neighboring of Amazonas State until to reach higher levels at Kamikuã Village, an Indian community along Purus River. The suspicious of an unidentified species of microfilaria transmitted by blackflies (Diptera, Simuliidae) in the Amazonia has already been noticed (Shelley 2002). Since the sample with atypical microfilaria was lost, our efforts in the present survey were concentrated in Vila Antimary at Acre River, where that previous detection was reported.

A preliminary trial of infected or uninfected material was performed, and the positive material was submitted to an accurate morphometric study where it was possible to detect this unusual microfilaria. This larval stage in the blood material was obtained with the concentration Knott's Method and differed from all microfilarias previously found. Studies concerned with morphological differentiation of M. ozzardi and O. volvulus microfilaria has been previously described (Post et al. 2003). Both filarian species present unsheated microfilaria, and there is a general concern that M. ozzardi microfilarias are smaller than those of O. volvulus, but the ranges frequently overlap when both are found in skin samples which were a reason of misleading diagnosis in the past (Moraes et al. 1983). In the present work, we reported M. ozzardi infections, and only the samples from Kamikuã Village, Vila Antimary, and Porto Acre were analyzed through morphometry. The values obtained are in accord with those previously reported for Haitian and Colombian M. ozzardi (Kozek and Raccurt 1983) prepared through Knott's Method but are slightly different from those achieved with Brazilian samples by Post et al. (2003) what may be related to factors such as source of the material, larval developmental stage, and procedures like drying, fixation, and staining of the smear.

The unusual microfilaria presented here has some characteristics of those of O. volvulus in the cephalic space and similarities in caudal space with M. ozzardi and its measures were closer to this last one. Comparison of this microfilaria with those of O. volvulus showed that there are some resemblance at the cephalic space, which is followed by two paired nuclei following this region but that similarity was absent at the posterior extremity, where a nuclear column with 7-8 nuclei are present in a regular and aligned arrangement and a detachment from cuticle are very perceptible (Fig. 3).

Fig. 3 Whole O. volvulus microfilaria, Scale = 10 μm (photomicrography from A. J. Shelley)

---

Some questions raised from these results: do these microfilarias belong to an unknown human filarial species or even to a species parasite of an animal? Are they morphological variations of M. ozzardi? Further morphological, ultrastructural, and molecular studies are being carried out in our laboratory in order to provide additional evidences capable to answer these questions.