The concentration of Mansonella ozzardi microfilariae in skin capillaries

M. A. P. MORAES’, A. J. SHELLEY’, A. P. A. LUNA DIAS~ AND C. J. MANCABEIRA SILVA~ ‘Universidade de Brasilia, Brasilia-D.F., Brazil; ‘British Museum (Natural History), London; 31nstituto

Oswald0 Cruz, Rio de Janeiro, Brazil; 4Ministt%o de Satide, SUCAM, Manaus, Brazil

Summary The densities of microfilariae of Mansonella ozzardi in different regions of the body of infected

individuals were studied. The density in venous blood closely approximated to that found in finger capillaries, but was significantly lower than the densities in capillaries of the scapular region and buttocks. A possible explanation for this concentration of microfdariae is the lack of glomi in these two sites, which consequently have a lower flow of blood than areas where glomi occur and where microfilarial densities are lower. No relationship was found between blood and skin densities in the buttocks and scapular region, probably due to the inadequacy of the skin snip method used for the detection of this species of filaria. It was concluded that the most acceptable detection method for M. ozzardi remains the sampling of finger capillary blood. It could not be decided, due to the paucity of data, whether any relationship exists between skin microfilarial densities and the biting sites of the vectors.

Introduction During investigations in 1975 into the reported

occurrence bv GUIMAR.&ES et al. 119751 of Onchocerca volvulus in &in snips taken fro’m Biazilian Ticuna Indians, MORAES et al. (1978) were only able to detect the presence of Mansonella ozzardi in a far larger sample taken from the same villages. Although this filaria was already known to be present in these Indians (RKHOU, 1957) its detection by the skin snip method was recognized only more recently (MORAES, 1976). Not only did these investigations show the absence of a focus of onchocerciasis in this locality but suggested the need for a study of the relationship between microfilarial densities of M. ozzardi in the skin and blood system of infected individuals.

The histological evidence of a concentration of microfilariae of M. ozzardi in the dermal capillaries of Brazilian Indians (MORAES, 1976; MORAES et al., 1978) emphasized the need for close examination of all microfilariae in skin snips taken in localities where both 0. volvulus and M. ozzardi occur as well as indicating a possible new detection method for M. ozzardi. The iatter point was discussed by NELSON & DAVIES (1976‘1 who found that M. ozzardi microfilar- iae were also concentrated in skin capillaries in infected individuals from Trinidad. They argued that if the microfilariae in infected individuals have a characteristic pattern of distribution in the skin as an adaptation to the preferential biting site of the vector, as occurs in several species of Onchocerca, skin snips taken from the more densely infected areas would be the preferred method of detection of the parasite and would explain the unexpectedly low infkction rates (based on blood samdintz) of children in Trinidad. however, NATHAN ei al.-11978) could not demons- trate such a skin distribution and found higher concentrations of microfilariae in the blood than in skin snips. NATHAN (1979) was also unable to establish-any relationship between microfilarial densi- ties of M. ozzardi in the skin and blood and the site of

preferential biting by the vector, which in Trinidad is Culicoides bhlebotomus (see NELSON & DAVIES. 1976; NATHAN,*~~~~, 1981s

I ,

MORAES et al. (1978) also believed that the concen- tration of M. ozzardi on microfilariae in skin capillar- ies could be an adaptation to the vectors’ feeding habits, which in the area of Brazil that they studied are two Simulium species (SHELLEY et al:, 1980). I f such a concentration effect does occur m the skin capillaries and is linked to vector feeding habits the following hypotheses should be correct: microfilariae will not be evenly distributed in the blood, being denser in capillary blood than in venous blood; microfilarial densities in capillary blood and skin snips will be related; and microfilarial densities in blood and skin snips will vary depending on the sample site in accordance with the preferred biting site on the body by the vector. A comparison of these densities is the subject of this paper.

Materials and Methods The work was carried out at Feijoal, a small Ticuna Indian

village situated on the margin of the River Solimbes in Amazonas state-mau reference: 4” 18’S 69” 33’W. This vil- lace was chosen since a orevious survev there showed a hieh inzidence of M. ozzardi and the absence of 0. volvulus in tie human population (MORAES et al., 1978).

The investigation was divided into two stages: the first in 1976 had as its objective the comparison of microfilarial densities in venous and capillary blood and the second, in 1978, compared the microNaria densities in capillary blood and skin snips taken from different regions of the body. During the first stage 127 people (62 men and 65 women) aged between 12 and 69 were samded. Caoillarv blood was tiken, using a lancet, from the fo&h fir&>-bf 6e left hand and randomly from either the left or right shoulder, and venous blood with a hypodermic syringe from the median cubital vein of the left arm. Two 20 mm3 samples of capillar blood were taken from each eite but a single 40 mm sample of venous blood was uSed. Samples were de-haemaglobinized, fixed and stained witb Giemsa and then examined for microfilariae. 223 Indians (115 men and 108

464 CONCENTRATION OF M. ozzardi MICROFILARIAE IN SKIN CAPILLARIES

women) aged between 9 and 75 years were used in the second stage of the investigation. Single 20 mm3 samples of capillary blood were taken from the fingers and shoulders of individuals as before, as well as from the upper part of the left or right buttock of 110 men. To measure microfilarial densities in the skin, a small skin snip 4 to 5 mm in diameter was removed from the shoulder of the 223 individuals and from the buttocks of 110 of the men. Skin snips, though randomly taken from the right or left hand side of the body, were always removed from the same side of the body as the respective capillary blood samples. No skin snips could be taken from the buttocks of the women due to the influence of a religious sect, the ‘Irmandade da Cruz’, which forbids the exposure of most of the body in females. Each skin snip was removed with a razor blade, placed in a drop of distilled water on a slide and repeatedly squeezed with a curved seeker to force out the microfilariae from the skin capillaries. This was necessary since microfilariae of M. ozzardi (in contrast to those of 0. volvulus) are less active and are found in capillaries rather than the intercellular fluid and hence are

less likely to leave the skin snips naturally. Skin smps were removed from the slides after 30 min by which time the drops of water on the slides were nearly dry, each skin snip then being preserved in 10% formalin to be weighed later. After drying, the drops were fixed and stained with Giemsa. The counting and identification of the microfilariae were carried out by a single person (MAPM), and infections expressed as number of microfilariae per mg of skin.

Results Of the 127 people examined during the first visit,

83.4% (106) were found positive for M. ozzardi in blood samples from either the median cubitai vein or capillaries in the finger or shoulder. Table I shows the variability in the detection of positive cases using blood from these three sources. There was no significant different between positivity (x2 = 1.1, P<30%) in the samples taken from the vein or finger

Table I-Infection rates and microMaria densities of M. ozzardi in venous and capillary blood of male and female Ticuna Indians (1976 investigation)

Venous blood

Finger capillary

blood

Shoulder capillary

blood

Males No. +/No. exam. (% infected) Total mff in samples Median mffiinfected (range) person

38/62 (61.2) 41162 (66.1) 48162 (77.4) 1476 1627 5515

38.8 (l-430) 39.6 (l-489) 114.8 (l-2575)

Females No. +/No. exam. (% infected) Total mff in sample Median mffjinfected person (range)

40165 (615) 46165 (70.7) 52165 (80.0) 778 1004 3235

19.4 (l-170) 21.8 (l-195) 62.2 (l-640)

Both sexes No. +/No. exam. (% infected) Total mff in samples Median mffiinfected person (range)

781127 (61.4) 871127 (68.5) 1001127 (78.7) 2254 2631 8750

28.9 (l-430) 30.2 (l-489) 87.5 (l-2575)

Infection rates using capillary blood based on two 20 mm samples for each site; microfilarial densities are based on the average numbers of mff from both samples.

Table II-Infection rates and microtilarial densities of M. ozzardi in capillary blood and skin snips of male and female Ticuna Indians (1978 investigation)

Finger Shoulder Shoulder Buttock Buttock capillary capillary skin capillary skin

blood blood snips blood snips

Males No. +/No. exam. (% infected) 641115 (55.6) 85/115 (73.9) 611115 (53.0) 881110 (80.0) 561110 (50.9) Total mff in samples 3298 15732 1219 20901 708 Median mff/infected (range) person 51.5 (l-383) 185.0 (l-2097) 19.9 (l-202) 237.5 (1-2621) 12.6 (l-67)

Females No. +/No. exam. (% infected) 54/108 (50.0) 66/108 (61.1) 511108 (47.2) - - Total mff in samples 2249 9428 429 - - Median mffiinfected (range) person 41.6 (l-738) 142.8 (l-2300) 8.4 (l-58) - -

Both sexes No. +/No. exam. (% infected) 1181223 (S2.9) 1511223 (67.7) 112/223 (50.2) - - Total mff in samples 5547 25160 1648 - - Median mff/infected person (range) 47.0 (l-738) 166.6 (l-2300) 14.7 (l-202) - -

Infection rates using capillary blood based on two 20 mm3 samples for each site; microfilarial densities are based on the average numbers of mff from both samples.

M. A. P. MORAES et al. 465

tip capillaries but blood from capillaries in the shoulder produced a significantly higher positivity (x2

body, rather than that previously suggested (NELSON & DAVIES, 1976; MORAES et al.. 1978) of a vector

= 6.7, P<l%) than in samnles taken from finger tip capillaries. As’ expected, individuals with low micro- filaraemias varied in positivity in the two samples of caoillarv blood taken from each site. Thus. of the 87 ne&le ‘detected as positive using linger capillary blood, only 67 showed the presence-of n&rofilariae in both 20 mm3 samoles: similarlv in blood from shoul- der capillaries only 8i individuals were positive in both samples compared to the over-all positivity of 100. Also, microfilarial density in each individual varied depending on the source of the blood sample (Table I). Median microfilarial density calculated from the positive cases detected using each method was similar in both venous blood and finger capillary blood samples but the density was considerably higher in samples taken from shoulder capillaries. Furthermore, -the median microfilarial density in shoulder canillaries was higher in the first 20 mm3 sample (112:4, range l-32OOj than in the second (80.2, range l-1950) but this difference was not seen in finger capillary blood (first sample-38.9, range l-465: second samnle 34.3, range l-514). Although the infection rate -was the-same for both men and women, median microfilarial densities were always higher in the men in all blood sampling methods (Table I).

parasite adaptation, is that it is correlated with the presence of arteriovenous shunts or glomi. These are present in the linger tip and to a lesser extent in the ear lobe and when open allow a greater blood flow from the arterial to the venous system. Because of this greater flow, microfilarial densities in the venous blood will annroximate those of the skin canillaries

Some of these findings were confirmed during the second investigation on a larger sample of Indians. Again positivity and microfilarial density was higher in samples of blood from shoulder capillaries than from finger tip capillaries (Table II). However, males showed slightly higher infection rates than females and again microlilarial densities were higher in males. These trends were also noted in skin snip samples taken from the shoulder. The infection rate in both males and females using this method was of the same order as that obtained by sampling linger capillary blood. However, median microlilarial density was considerably lower even than that obtained from using capillary blood from the finger. Samples of blood taken from skin capillaries of the buttocks of 110 of the men used in this investigation showed a higher infection rate and higher microfilaraemias than any of the other methods (Table II). A similar infection rate as that obtained using shoulder skin snips was found in skin snips from the buttocks but microfilarial densities were lower.

Discussion It was shown that microfilariae of M. ozzardi were

not evenly distributed in the blood, but no clear distinction between densities in venous and capillary blood could be made. Capillary blood from the finger was shown to contain the same density of microhlariae as venous blood, confirming the finding of NATHAN et al. (1978) in Trinidad. However, densities were more than three times higher when capillary blood from the shoulder was used and five times higher for blood from the buttocks, illustrating a concentration of mff in these two parts of the body. Higher concentrations of microfilariae of M. ozzardi were also found in capillary blood from the ear lobe compared to that from the finger by NATHAN & RACCURT (1979) in patients in Trinidad. A possible explanation for this concentration of microfilariae in certain parts of the

where the glomus occurs. Where it is absent, as in the back and buttocks, blood flow is lower and hence microfilariae tend to build UD. It is in fact difficult to obtain a blood sample from ihese parts because of the restricted blood supply.

No relationship was found between microfilarial densities recorded from capillary blood and skin snips taken from corresponding sites of the body. Similarly, NATHAN et al. (1978) and NATHAN (1979) working in Trinidad were also unable to demonstrate such a relationship but the former authors did not use skin snips that had been weighed or that were of standard size. It is probable that such a relationship does exist; if canillaries in the skin contain high densities of micr&ilariae it follows that skin snips should reflect this. That NATHAN et al. (1978), NATHAN (1979) and the present investigation failed to show this must suggest that the method used for sampling microfilar- iae in the skin was inappropriate. Though the skin snip method functions for microlilariae of 0. volvulus, which inhabit the tissue fluid and are relatively motile, it does not for M. ozzardi which are confined to the capillaries and are feeble movers. As our findings show, pressure on the skin snips to expel microfilariae from capillaries was insufficient, and anyway could not be standardized, to sample mic- rofilarial density accurately for M. ozzardi. It is also possible that the use of water for the incubation of skin snips added further to the inaccuracy. It would be necessary to digest the skin snip for accurate sampling of the microfilariae. Blood sampling there- fore remains the most effective method for detecting M. ozzardi infections and the best site for this is still the finger, even though the microfilarial density is not as high here as in other parts of the body; blood is easily obtained from this source and individuals are less reluctant to be tested when blood is taken from this site.

The present findings do not exclude the possibility of a relationship between parasite distribution in the host and the biting site of the vector. The results, though preliminary in this respect, do indicate that microlilarial densities are higher in the dorsal surface of the torso in nreciselv the region where the vectors at this locality 1.9. am&onicuk and S. argentiscutum) orefer to bite. It has vet to be tested if microfilarial densities are also high’in other bodv Darts such as the arms, legs and abdomen. Mapping of skin densities would rely on either sampling the capillary blood at different sites or sampling gut squashes of Simulium collected from the different parts of the body regions as DE LEON & DUKE (1966) and SHELLEY et al. (1979) used for 0. ~olvulus.

References De Leon, J. R. & Duke, B. 0. L. (1966). Experimental

studies on the transmission of Guatemalan and West African strains on Onchocerca voluulus by Simulium ochraceum, S. metallicuw and S. callidum. Transactions of

466 CONCENTRATION OF M. ozzardi MICROFILARIAE IN SKIN CAPILLARIES

the Royal Sociery of Tropical Medicine and Hygiene, 60, 735-752.

Guimaries, A., Mello, J. A. N. & Dourado, H. (1975). Novo focos de oncocercose no Estado do Amazonas. Presented at: XI Congress of the Brazilian Society for TyopicalMedicine, Rio de Janeiro, Brazil, February 1975.

Moraes, M. A. P. (1976). Mansonella ozzardi microfilariae in skin snips. Transactions of the Royal Society of Tropical Medicine and Hygiene, 70, 16.

Moraes, M. A. P., Ahnelda, M. M. R., Lovelace, J. K. & Chaves, G. M. (1978). Mansonella ozzardi entre indios Ticuna do Estado do Amazonas, Brasil. Boletin de la Oficinu Sanitaria Panamericana, 85, 16-25.

Nathan, M. B. (1978). Culicoides phledotomus, a vector of Mansonella ozzardi in coastal north Trinidad, West Indies. Transactions of the Royal Society of &pica1 Medicine and Hygiene, 72, 436-437.

Nathan, M. B. (1979. A-comparison of Mansonella ozzardi micro6laria densities in the blood and in skin snips from three areas of the body. Transactions of the RoyaLSociety of Trooical Medicine and Hveiene. 73. 338-340.

Nathan, k B. (1981). Transn&ion’of the human filarial parasite Mansonella ozzardi by Culicoides phlebotomus (Williston) (Diptera: Ceratopogonidae) in coastal north Trinidad. Bulletin of Entomological Research, 71,97-105.

Nathan, M. B., Bartholomew, C. F. & Tikasingh, E. S. (1978). The detection of Mansonella ozzardi microfilariae in the skin and blood with a nofe on the absence of

periodicity. Transactions of the Royal Society of Tropical Medicine and Hygiene, 72, 420-422.

Nathan, M. B. & Raccurt, C. (1979). Higher concentrations of microfilariae in capillary blood from the finger in Wuchereria bancrotii and Mansonella ozzardi infections. Transactions of the’Roya1 Society of Tropical Medicine and Hygiene, 73, 455-457.

Nelson, G. S. & Davies, J. B. (1976). Observations on Mansonella ozzardi in Trinidad. Transactions of the Roval Society of Tropical Medicine and Hygiene, 7i), 16-17.

Rachou, R. G. (1957). Distribuicio geografica das filarioses humanas no Brasil. Revista Brasileira de Malariologia e Doeqas Tropicais, 7? 79-100.

Shelley, A. J., Luna Dlas, A. P. A. & Moraes, M. A. P. (1980). Simulium species of the amazunicum group as vectors of Mansonella ozzardi in the Brazilian Amazon. Transactions of the Royal Society of Tropical Medicine and Hygiene, 74, 784-788.

Shelley, A. J., Pinger, R. R., Moraes, M. A. P. & Hayes, J. (1979). Concentrations of Onchocerca volvulus bv Simu- kum sanguineurn during feeding; use in mapping parasite distribution in the skin. 3oumal of Medical Entomology, 16, 48-51.

Accepted for publication 22nd November, 1982.

Laboratory Meetings

Publication of reports of demonstrations and posters shown at Laboratory Meetings of the Society will be reinstated from V’ol. 78 (1984).

Forms on which reports must be submitted will be issued in advance of the meetings. They should be completed and handed to the Convenor at the meeting and will be refereed. The maximum length for the text is 200 words, including references, and tables and illustrations cannot be accepted. Detailed information on the correct format will be printed on the reverse side of the form.

Late or overlong submissions will be rejected without reference to the author and will be omitted or cited by title only in the journal.