Veterinary Parasitology 122 (2004) 273–286

Epidemiological aspects of filariosis in dogs on thecoast of Paraná state, Brazil: with emphasis on

Dirofilaria immitis

Larissa Reifura,∗, Vanete Thomaz-Soccolb,Fabiano Montiani-Ferreirac

a Pós Graduação em Ciencias Veterinárias, Setor de Ciencias Agrárias, Universidade Federal do Paraná(UFPR), Rua dos Funcionários no 1540, Curitiba, PR 80035-050, Brazil

b Departamento de Patologia Básica, UFPR, Setor the Ciencias Biológicas, Centro Politecnico,Curitiba, PR 81531-990, Brazil

c Departamento de Medicina Veterinária, UFPR, Rua dos Funcionários no 1540, Curitiba,PR 80035-050, Brazil

Received 31 March 2004; received in revised form 15 April 2004; accepted 6 May 2004

Abstract

The present study determined the prevalence and geographical distribution ofDirofilaria immitisand other filariae, from dogs in littoral areas of Paraná state, in Brazil. This survey spanned eightmonths, between 1998 and 1999, and was also designed to compare the efficacy of different tests fordiagnosis of heartworm infection in that area. Blood samples were collected from 256 native-owneddogs distributed along the Paraná coastal area. Five diagnostic procedures were used: direct smearexamination, the Knott’s modified test, filtration assay, and two heartworm antigen detection kits.A follow-up imaging exam was performed to support the heartworm diagnosis. The imaging di-agnosis included radiographic and ultrasonographic exams of six dogs that had positive results forthe heartworm antigen detection kits, but showed different microfilarial burdens. The presence andseverity of radiographic and ultrasonographic signs were compared with the results obtained in mi-crofilariae detection and antigen tests. Diagnostic parasitology results indicated that 31.25% of thedogs were microfilaremic. Three different microfilariae were recovered:D. immitis, Dipetalonemareconditum, and the third (mf3) was not identified.D. reconditum was the species with the highestprevalence: 22.6%. In general,D. immitis prevalence was 5.47% (28.57% occult infections), butit varied along the coast and the range was from 0 to 20%. No correlation could be establishedbetween the overall scores for microfilarial counts (small or large numbers) and the severity ofradiographic results or the likelihood of detecting filariae in the pulmonary artery using echocar-diography. The finding of a different type of microfilaria (mf) suggested the existence of a third

∗ Corresponding author. Present address: CMIB-CVM graduate student, Biomedical and Physical SciencesBuilding, Room 2209, Michigan State University, East Lansing, MI 48824, USA. Tel.:+1 517 355 6463x1550;fax: +1 517 353 8957.E-mail address: reifurla@cvm.msu.edu (L. Reifur).

0304-4017/$ – see front matter © 2004 Elsevier B.V. All rights reserved.doi:10.1016/j.vetpar.2004.05.017

274 L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286

species in Paraná state, whose prevalence was 4.68%. These results show that to obtain a reliablediagnosis of heartworm infection, antigen detection kits are indicated. Knott’s test or filtration shouldbe performed to confirm microfilaremia and not for diagnosis of heartworm infection. Imaging testssupport parasitology exams and add more about severity of infection. The northern areas, speciallyGuaraqueçaba and Ilha das Peças, presented the highest number of heartworm-infected dogs.© 2004 Elsevier B.V. All rights reserved.

Keywords: Dirofilaria immitis; Dipetalonema reconditum; Serology; Microfilariae; Epidemiology; Radiography;Ultrasonography; Dogs; Brazil

1. Introduction

Heartworm is a worldwide disease caused by the parasiteDirofilaria immitis. Canidsare the definitive hosts canids and mosquitoes from the family Culicidae act as interme-diate hosts or vectors.D. immitis is the most important filarial parasite of dogs. Adultsare normally found in the right ventricle and pulmonary arteries, causing extensive injury.They produce microfilariae after the prepatent period, which is about 6–7 months. At thisstage, microfilaria (mf) testing can be very important. However, occult heartworm infections(non-microfilaremic dogs) cause false-negative results if microfilaria tests are not combinedwith antigen tests. It is not rare to find dogs with occult infections presenting no clinicalsigns. In these cases, radiographic and ultrasonographic exams are helpful in evaluating thecardiac and pulmonary functions.

Areas with a hot and humid climate favor the mosquito population and heartworm trans-mission. Prevalence and geographic distribution ofD. immitis infection have been reportedall over the world, for example, in the US (Stromberg et al., 1995; Walters, 1995), Japan(Hatsushika et al., 1992), Spain (Guerrero et al., 1995; Aranda et al., 1998), Italy (Genchiet al., 1995; Cringoli et al., 2001) and in other countries (Rosa et al., 2002; Song et al.,2003). Amongst the 26 Brazilian states, heartworm prevalence is known in only eightstates: Recife (2.3%), Maranhão (43–46%), Alagoas (12.5%), Rio de Janeiro (16.1–49%),São Paulo (8.8%), Mato Grosso (2%), Santa Catarina (12–15%) and Rio Grande do Sul(1.1%) (Guerrero et al., 1989, 1992; Labarthe et al., 1997; Alves et al., 1999; Ahid et al.,1999; Araújo et al., 2003).

The importance of heartworm disease and the paucity of information about the prevalenceof canine filaria in the Paraná state inspired us to conduct this preliminary survey. We focusedon finding and distinguishing microfilariae by performing microscopic examination andantigen testing of fresh blood. We compared several microfilaria tests versus antigen testsand also observed the radiographic and ultrasonographic signs of some dogs presentingpositive results at time of antigen testing.

2. Materials and methods

2.1. Concentration area of the study

The present work was done on the lowlands of the Paraná coast where the climate ishot and humid (Troppmair, 1990). The selected areas were Antonina, Morretes, Matinhos,

L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286 275

Guaratuba, Guaraqueçaba, Pontal do Sul, Ipanema, Shangrilá, Praia de Leste, Ilha do Mel,and Ilha das Peças (Fig. 1).

2.2. Selection of dogs

A total of 256 dogs owned by native people were selected from October 1998 to May 1999.A questionnaire for owners included questions about the dog’s age, sex, type of hair coat,health status, and type of housing (indoor, outdoor, indoor–outdoor). Dogs were excludedfrom the study if they were younger than 1 year old, or receiving any microfilaricide oradulticide drug.

2.3. Sample collection (Paraná coastal area)

Approximately 5 ml of blood was withdrawn from the brachial vein and stored withEDTA, under refrigeration (2–4◦C).

Flyers were distributed during an oral presentation on dirofilariosis (including its courseand treatment), directed to the owners and the general public.

2.4. Sample testing (UFPR-Parasitology lab)

Samples were tested by direct blood smear examination, modified Knott’s test, filtrationtest, and by two heartworm antigen detection kits (WitnessHeartworm®a and Snap® canineheartwormb).

2.5. Diagnostic imaging analysis: radiography and ultrasonography (PUC-Radiologylab)

Thoracic radiographs and echocardiography were performed in six selected dogs diag-nosed with heartworm disease. The heartworm status (infected or not) was determinedusing heartworm antigen detection kits described above. Microfilaria burden was subjec-tively evaluated in these dogs using the microfilaria tests. One of these dogs was amicrofi-laremic and two presented a small number of microfilaria. A moderate or large number ofmicrofilaria was observed in the remaining three dogs. All animals were conscious whilebeing radiographed and ultrasonographed. Left lateral thoracic and dorsoventral (DV) ra-diographs were taken. Echocardiography was performed in the same group of dogs. Both 5and 7.5 MHz sector transducers were used to obtain several long axis and short axis views,from the right parasternal location.

2.6. Statistical methods

Dogs were considered to have heartworm infection only after results of at least oneantigen test were positive. Morphology and morphometry were used to diagnose infectionby filaria other thanD. immitis. The number of positive results was divided by 256 (totalnumber of dogs) and multiplied by 100 to find the prevalence of each parasite. Analyses

276 L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286

Fig. 1. Maps of Brazil, Parana state, and the lowland of Parana coast where the research was performed.

L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286 277

Fig. 2. Pictures obtained by microphotography of different microfilariae recovered from dogs of the Parana coastalarea. (A) Microfilaria motility representation as observed in blood smear; (B) microfilaria ofDipetalonema re-conditum; (C) microfilaria ofDirofilaria immitis; (D) third unindentified microfilaria (mf3), (B–D) were obtainedby either the modified Knott’s test or filtration test.

278 L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286

of all variables (dog’s age, sex, type of hair coat, health status, and type of housing) werecomputed using multivariable logistic regression.

3. Results

3.1. Results obtained by microfilaria tests

A total of 161 (62.9%) male dogs and 95 (37.1%) females were included. In a compar-ison of the tests that detect microfilariae the filtration test produced more accurate results(Table 1). The species of microfilaria found in this study and their morphology and mor-phometry obtained from direct smear, Knott’s test, and filtration test are presented inTable 2andFig. 2.

Fig. 3. Comparison of results obtained by different tests performed to detectDirofilaria immitis infection in dogsfrom the Parana coastal area (n = 256).

Fig. 4. Prevalence ofDipetalonema reconditum, Dirofilaria immitis, and the third non-identified microfilaria (mf3)diagnosed by identification of microfilariae or by detection of antigen in the blood of dogs from the Parana coastalarea (n = 256).

L.R

eifuretal./Veterinary

Parasitology122

(2004)273–286

279

Table 1Comparison of results from microfilaria (mf) tests performed in 256 blood samples of dogs from the Parana coastal area

Tests performed Microfilaremic dogs Microfilaria ofD. immitis Microfilaria of D. reconditum Microfilaria 3 (mf3)

Direct blood smear 31/256 (12.1%) – – –Modified Knott’s test 71/256 (27.7%) 9/256 (3.5%) 51/256 (19.9%) 11/256 (4.3%)Filtration test 80/256 (31.2%) 10/256 (3.9%) 58/256 (22.6%) 12/256 (4.7%)

280L

.Reifur

etal./VeterinaryParasitology

122(2004)

273–286

Table 2Results of diagnostic parasitology (fresh smear, modified Knott’sa and filtrationa) and morphology and morphometry differentiation of canine microfilariae of the Paranacoastal area (n = 256)

Criteria Dirofilaria immitis Dipetalonema reconditum Microfilaria 3 (mf3)

Number of positive samples 10 (3.9%) 58 (22.6%) 12 (4.7%)

Length measurement (�m)Range 306.2–340.2 236.2–286.0 304.4–338.9Mean 328.03± 10.56 258.50± 11.32 323.49± 12.73

Width (�m)Mean 6.7 4.1 6.0

Head morphology Tapered head Blunt head Tapered head

Head morphology Absent “tooth” Presence of “tooth” Presence of small “tooth”

Tail morphology Straight tail (inconsistent finding) “Button-hook” tail (inconsistent finding) Straight tail (inconsistent finding)

Amount in each sample Few to large numbers Small numbers Few to large numbersMotility Undulate in one place Undulate in one place/move across field Undulate in one place

a criteria given for lysate prepared using 2% formalin.

L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286 281

3.2. Results obtained by heartworm antigen tests

A total of 14 out of 256 dogs (5.5%) exhibited positive results to heartworm antigentesting. However, 4/14 (28.6%) were occult infections. All of the blood samples containingmicrofilariae ofD. immitis yielded positive results by at least one of the antigen tests

Fig. 5. Prevalence of filariae in dogs from 11 regions of the Parana coastal area, obtained after five diagnostic testswere performed (n = 256).

282 L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286

performed. None of the blood samples containing microfilaria ofDipetalonema reconditumor of mf3 showed positive results by the same antigen tests.

There were no significant differences among dogs regarding sex, age, type of hair coat,housing or health status and heartworm infection.

Considering only heartworm infection diagnosis, results were different for each testperformed and the antigen detection tests were the most efficient in identifying positivedogs (Fig. 3).

Fig. 6. Radiographic images obtained from a heartworm-infected dog. DV view showing enlarged right heart,subtle pulmonary alveolar and interstitial radiographic pattern, main pulmonary artery enlargement and peripheralpulmonary arterial abnormalities.

L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286 283

In summary, of the 256 dogs selected in the Paraná coastal area, 84 (32.81%) wereinfected by at least one species of filaria. The number of dogs infected byD. reconditumwas impressively higher than those infected by other filariae (Fig. 4).

3.3. Results obtained by area of study

Guaraqueçaba and Ilha das Peças, located to the north of Paraná state, had the highestpercentage (20.69 and 15%, respectively) of dogs infected withD. immitis, compared withthe other areas (Fig. 5).

3.4. Results obtained with thoracic radiographs and ultrasonography

All the six selected dogs presented at least one of the following distinctive radiographicsigns with varied degrees of severity. (1) Enlarged caudal pulmonary arteries with somedegree of tortuosity and ill-defined margins; (2) enlarged main pulmonary arterial segmentextending beyond the cardiac border on the DV view; (3) enlargement of the right side ofthe heart; (4) increased size of caudal vena cava; and (5) pulmonar parenchymal lesions

Fig. 7. Ultrasonographic image from the heart of a heartworm-infected dog. AD: right atrium, DV: right ventricle,AE: left atrium, VE: left ventricle. Right ventricular enlargement and presence of parallel hyperechoic bands nextto the right ventricular wall are observed.

284 L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286

(bronchointerstitial or alveolar pattern). A representative radiograph of an affected doganalyzed in the study is shown inFig. 6.

At ultrasonographic examination, several hyperechoic parallel linear bands were observedin the proximal right pulmonary artery lumen. These were identified as worm segments(filarid worm echoes) and were observed in only three dogs (Fig. 7). One of these threedogs had low microfilaria numbers and another was amicrofilaremic, as observed using themicrofilaria tests described above.

4. Discussion and conclusions

It was demonstrated in this study that at least two species of filarial parasites are presentin dogs from the Paraná state. The prevalence ofD. immitis infection is not uniformlydistributed across the Paraná littoral area, but regions of high prevalence do exist.

In general, the Paraná coastal area presented a lower prevalence of heartworm infection(5.47%) than most other Brazilian coastal areas. On the other hand, some Paraná areasalso appeared to a have high prevalence, such as 20.68% in Guaraqueçaba, 15% in Ilhadas Peças, 8.33% in Shangrilá, and 6.12% in Guaratuba. These areas have favorable cli-matic conditions, have several species of mosquitoes, most dogs living there do not receiveantiparasitic drugs, and some wild animals may serve as reservoirs of filarial worms. Inaddition, the Paraná coast is a summer vacation area for people from other cities and states;considering the dog transit from and into the Paraná coast, prophylaxis is recommended fordogs in this area.

The prevalence ofD. reconditum was 22.65%, with variations from one area to anotherranging from 0 to 45.0%. The prevalence of the third microfilaria (mf3) found in this surveywas 4.68%. Curiously, a survey performed in a nearby state (Santa Catarina) (Araújo et al.,2003), did not report detection ofD. reconditum or of mf3. Nevertheless, because of the highprevalence of filaria other thanD. immitis, as observed in the present study, it is necessaryto precisely identify the microfilariae (Table 2) for differential diagnosis. This is especiallyrecommended when diagnosis is done without confirmation by serological testing.

Among the three microfilaria tests used, microfilaremia was better confirmed by the fil-tration test, but the modified Knott’s test was better for microfilaria differentiation. Bothconcentration tests were markedly more effective than the direct blood smear. In regard todiagnostic method, heartworm infection should be confirmed by performing an antigen de-tection test, because these tests are more sensitive and because in the present work 28.6% ofthe cases were occult infections. The high percentage of occult infections is not uncommonand was previously reported by several authors (Rawlings et al., 1982; Grieve et al., 1986;andLabarthe et al., 1997).

Among the six dogs analyzed with diagnostic imaging techniques, no obvious correlationcould be detected between the numbers of microfilariae (small or large numbers) observed,using microfilaria tests, and the severity of radiographic signs and/or the likelihood ofdetecting filariae in the pulmonary artery using echocardiography.

These initial findings serve as baseline data for future studies related to heartworm man-agement in the Paraná coastal area. Hopefully, this data will encourage new surveys in orderto include more geographic areas and to identify the third microfilaria (mf3).

L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286 285

Acknowledgements

The authors thank all individuals who volunteered for technical assistance, CAPES forproviding financial support, the municipal governments of Pontal do Paraná, Guaratuba,Paranaguá and Matinhos, Merial for donation of Witness heartworm kits, Ubiratã A.T.da Silva for donation of the apparatus for the filtration test, Prof. Milton Andriguetto fordonation of the Paraná coastal map, and Dr. Alessandra Q. Augusto for performing thediagnostic imaging.

aAntigen test 1: WitnessHeartworm® canine heartworm antigen test kit, Synbiotics Corp.,San Diego, CA.

bAntigen test 2: Snap® canine heartworm antigen test kit, Idexx, Maine, USA.

References

Ahid, S.M.M., Oliveira, L., Saraiva, L.Q., 1999. Dirofilariose canina na Ilha de São Luıs, Norte do Brasil: Umazoonose potencial. Cad. Saúde Públ. Rio de Janeiro 15, 405–412.

Alves, L.C., Silva, L.V., Faustino, M.A., McCall, J.W., Supakonderj, P., Labarthe, N.W., Sanches, M., Caires, O.,1999. Survey of canine heartworm in the city of Recife, Pernambuco, Brazil. Memórias do Instituto OswaldoCruz, Rio de Janeiro 94 (5), 587–590.

Aranda, C., Panyella, O., Eritja, R., Castellà, J., 1998. Canine filariasis Importance and transmission in the BaixLlobregat area, Barcelona (Spain). Vet. Parasitol. 77, 267–275.

Araújo, R.T., Marcondes, C.B., Bastos, L.C., Sartor, D.C., 2003. Canine dirofilariasis in the region of ConceiçãoLagoon, Florianópolis, and in the Military Police Kennel, São José, State of Santa Catarina, Brazil. Vet. Parasitol.113, 239–242.

Cringoli, G., Rinaldi, L., Veneziano, V., Capelli, G., 2001. A prevalence survey and risk analysis of filariosis indogs from the Mt. Vesuvius area of southern Italy. Vet. Parasitol. 102, 243–252.

Genchi, C., Basano, F.S., Bandi, C., Di Sacco, B., Venco, L., Vezzoni, A., Cancrini, G. 1995. Factors influencing thespread of heartworms in Italy. Interactions betweenDirofialria immitis andDirofilaria repens. In: Soll, M.D.,David, H.K. (Eds.), Proceedings of the Heartworm Symposium’95. American Heartworm Society, Batavia, IL,1995, pp. 5–26.

Grieve, R.B., Glickman, L.T., Bater, A.K., 1986. CanineDirofilaria immitis infection in a hyperenzootic area:examination by parasitologic finds at necropsy and by two serologic methods. Am. J. Vet. Res. 47, 392–393.

Guerrero, J., Genchi, C., Vezzoni, A. 1989. Distribution ofDirofilaria immitis in Selected Areas of Europe andSouth America. In: Otto, G.F. (Ed.), Proceedings of the Heartworm Symposium’89. American HeartwormSociety, Washington, DC, pp. 13–18.

Guerrero, J., Ducos de la Hitte, J., Genchi, C. 1992. Update on the distribution ofDirofilaria immitis in Dogsfrom Southern Europe and Latin America. In: Soll, M.D. (Ed.), Proceedings of the Heartworm Symposium’92.American Heartworm Society, Batavia, IL, pp. 31–37.

Guerrero, J., Rodenas, A., Galindo, G., Florit F. 1995. The extension of the prevalence ofDirofilaria immitis inCataluña, Spain. In: Soll, M.D., David, H.K. (Eds.), Proceedings of the Heartworm Symposium’95. AmericanHeartworm Society, Batavia, IL, pp. 5–26.

Hatsushika, R., Okino, T., Ohyama, F., 1992. The prevalence of dog hearthworm (Dirofilaria immitis) infectionin stray dogs in Okayama, Japan. Kawasaki Med. J. 18, 75–83.

Labarthe, N., Almosny, N., Guerrero, J., Araújo, D.A.M., 1997. Description of the occurrence of caninedirofilariasis in the State of Rio de Janeiro, Brazil. Mem. Inst. Oswaldo Cruz 92, 47–51.

Rawlings, C.A., Dave, D.L., McCall, J.W., 1982. Four types of occultDirofilaria immitis infection in dogs. JAVMA180, 1323–1326.

Rosa, A., Ribicich, M., Betti, A., Kistermann, N., Cardillo, N., Basso, N., Hallu, R., 2002. Prevalence of caninedirofilariosis in the City of Buenos Aires and its outskirts (Argentina). Vet. Parasitol. 109, 261–264.

Song, K.H., Lee, S.E., Hayasaki, M., Shiramuzi, K., Kim, D.H., 2003. Soroprevalence of canine dirofilariosisSouth Korea. Vet. Parasitol. 114, 231–236.

286 L. Reifur et al. / Veterinary Parasitology 122 (2004) 273–286

Stromberg, B.E., Prouty, S.M., Schlotthauer, J.C. 1995. Six decades of Hearthworm in Minnesota. In: Soll, M.D.,David, H.K. (Eds.), Proceedings of the Heartworm Symposium’95. American Heartworm Society, Batavia, IL,pp. 5–26.

Troppmair H., 1990. Perfil fitoecológico do estado do Paraná. Boletim de Geografia. ISSN 0102-5198. UniversidadeEstadual de Maringá. Departamento de Geografia, ano 8, n. 1, pp. 67–79.

Walters, L. 1995. Risk factors for heartworm infection in northern California. In: Soll, M.D., David, H.K. (Eds.),Proceedings of the Heartworm Symposium’95. American Heartworm Society, Batavia, IL, pp. 5–26.