Diagnosis of canine echinococcosis: comparison of coproantigen and serum antibody tests with arecoline purgation in Uruguay

veterinary parasitology ELSEVIER

Veterinary Parasitology 56 (1995) 293-301

Diagnosis of canine echinococcosis: comparison of coproantigen and serum antibody tests with arecoline purgation in Uruguay P.S. Craig a'*, R.B, Gasser b, L. Parada c, P. Cabrera d, S. Parietti a, C. Borgues e, A. Acuttis e, J. Agulla f, K. Snowden g, E. P a o l i l l o c aDepartment of Biological Sciences, University ofSalford, Salford M5 4WT, UK bDepartment of Veterinary Science, University of Melbourne, Werribee,Princes Highway, Vic. 3030, Australia CFundacion San Antonio de Padua, Av. Petrini 445, Sarandi del Yi, Durazno, Uruguay dFacultad Veterinaria, University de la Republica Montevideo, Uruguay eComision Departmental de Lucha Contra la Hidatidosis, Durazno, Uruguay fMinisterio de Salud Publica, CDLCH, Bulevar Espana 26 73, Montevideo, Uruguay •Department of Veterinary Pathobiology, Texas A&M University, CollegeStation, TX 77843, USA Accepted 12 April 1994

Abstract

Echinococcus granulosus is one of the most important and widespread of the helminth zoonoses. Diagnosis ofE. granulosus infection in dogs currently relies on arecoline dosing and detailed examination of the purge for adult worms. Two immunodiagnostic tests (ELISA) based on genus specific coproantigen detection or serum antibody (IgG, IgA and IgE) detection were compared against arecoline purgation for the detection of Echinococcus in naturally infected dogs in Uruguay. The coproantigen ELISA had a sensitivity of 76.9% compared with 34.6% for the serum IgG ELISA when assessed against 26 purge positive dogs (purge worm count range 1-4331 ). Coproantigen reactivity was positively correlated ( r = 0.65) to purge worm count, with a threshold at over 20 worms. There was no positive correlation of antibody levels with worm counts. In 26 matched Echinococcus positive dog samples, the overall sensitivity of serological detection increased to 69.2% when seroreactivity for IgA and IgE antibodies were included and to 96.2% for both coproantigen and antibody assays combined. The detection of current infection of individual dogs with E. granulosus by coproantigen ELISA has the potential to replace arecoline purgation, while specific serum antibody detection should be useful in assessing Echinococcus exposure in dog populations. * Corresponding author. 0304-4017/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSD10304-4017 (94) 00680-B

294

P.S. Craig et al. / Veterinary Parasitology 56 (1995) 293-301

Keywords: Echinococcus granulosus; Dog; Diagnosis-Cestoda; Coproantigen

1. Introduction

Human cystic echinococcosis (hydatidosis) is a serious helminthic infection with world-wide distribution particularly where sheep pastoralism is common (Schwabe, 1986). The causative organism, a taeniid tapeworm Echinococcus granulosus is transmitted cyclically and primarily between domestic dogs, which harbour the adult tapeworm (canine echinococcosis) and sheep, which act as intermediate host for the cystic larval stage (ovine hydatidosis). Humans and sheep may be infected (usually involving liver and/or lungs) after ingestion of infective eggs passed in the faeces of dogs. Human infection may occur through direct contact with dogs or from environmental contamination (Craig et al., 1988 ). The incidence of human hydatidosis in the UK is low with a maximum of 7 per 100 000 in the worst affected district of south Wales (Palmer and Biffin, 1987 ). In contrast, the incidence reaches 220 per 100 000 in the Turkana District of northwest Kenya (French and Nelson, 1982). Uruguay (population approximately 3 million) is also endemic for E. granulosus and it has been estimated that 500 hydatid patients undergo surgery each year (Purriel et al., 1973). Hydatid control programmes have been successfully implemented in several countries usually based on regular (6 weekly) praziquantel dosing of dogs coupled with targeted health education (Gemmell et al., 1986 ). Detection of the adult Echinococcus in dogs is of great importance in establishing prevalence data, for epidemiologic studies and in surveillance of control intervention (World Health Organization (WHO), 1981 ). Ante-mortem diagnosis of canine echinococcosis currently relies on the use of arecoline salts for purgation followed by careful examination of the purge for the presence of the small tapeworm (4-7 mm). Arecoline purgation, however, has several disadvantages, including variable sensitivity (Wachira et al., 1990), it is logistically difficult to carry out en masse, it is very time consuming, requires trained personnel, is biohazardous, may cause distress to some dogs and has a high failure rate (10-20%). The major advantage of arecoline purgation is that it is always 100% specific for Echinococcus as tapeworms are visualised directly. Echinococcus worm burdens in domestic dogs tend to be over dispersed and very high intensities (over 10 000 worms) may occur; however, the mean worm burden is usually 200-300 (Gemmell, 1990). Recent research has focused on the development of sensitive and specific immunodiagnostic tests for canine echinococcosis (Craig, 1993). Serological diagnosis based on the detection of adult E. granulosus antibodies (IgG) using a protoscolex antigen preparation was standardised in Australia using post-mortem confirmation and was initially reported to be 91.8% specific and 72.7% sensitive (Gasser et al., 1988 ). However, lower sensitivity (40%) was reported when nat-

P.S. Craig et al. / Veterinary Parasitology 56 (1995) 293-301

295

urally infected Turkana (Kenyan) dogs were tested (Jenkins et al., 1990). Using an alternative approach, genus specific coproantigen detection ELISAs were developed incorporating polyclonal antibodies against somatic (Allan et al., 1992) or excretory/secretory antigens of adult E. granulosus (Deplazes et al., 1992). Coproantigen specificity was high (over 96% ) with good sensitivity ( 87.5% ) and post-mortem worm burdens as low as 15-30 were detectable (Allan et al., 1992 ). An indirect diagnostic method based on coproantigen detection rather than serum antibodies is likely to be a better approach for detection of current Echinococcus infection in dogs. Coproantigen levels decrease rapidly in dogs within 6 days when experimental Taenia infections were treated with praziquantel (Deplazes et al., 1990). As part of a large community study of human hydatidosis in the Department of Durazno (Uruguay) an arecoline prevalence survey of canine echinococcosis was undertaken by a national dog dosing team (Comision Departmental de Lucha Centra la Hidatidosis). The results of arecoline purgation of dogs were used to validate and compare serum antibody detection and coproantigen detection ELISAs for the diagnosis of canine echinococcosis. 2. Materials and methods

2.1. Arecoline purgation Over a 7 day period (November 1991 ), domestic dogs from La Paloma town (and environs) Durazno, Uruguay were brought by their owners to a central roped testing area. Dogs were registered, washed, tethered and dosed orally with 2 nag kg-l of arecoline hydrobromide solution using a dosing gun. Purgation usually took approximately 30 min, although some dogs required a second or third arecoline dose (maximum). Following successful purgation, the complete purge was collected from the ground, boiled for 10 min in 5% formal saline and passed through a 150/an sieve. Purges were initially examined in the field using a fixed magnifying glass, according to standard procedures used by the Comision Departmental de Lucha Contra la Hidatidosis. Echinococcus positive dogs were treated with praziquantel (Praziquantel Injectable, Bayer). Owners with negative dogs received praziquantel tablets. In addition, all purges were re-examined in black-backed trays in a laboratory under a dissecting microscope. Total numbers of Echinococcus worms were counted, and the presence or absence of Taenia spp. and Dipylidium tapeworms noted.

2.2. Blood and faecal samples Venous blood was taken from the fore-leg (cephalic vein), allowed to clot and stored at 4°C. Serum was separated within 48 h and stored at - 20°C until tested. Approximately 1-2 g of solid faeces were collected prior to actual purge (or in the absence of purgation) and placed in a capped plastic tube. Faecal supernarants were prepared the same day as described by Allan et al. (1992). Briefly, stool samples were shaken vigorously with an equal volume (w/v) of 0.15 M

296

P.S. Craig et al. / Veterinary Parasitology 56 (1995) 293-301

phosphate buffered saline (PBS) containing 0.3% Tween 20 (Sigma), then centrifuged at 2000×g for 30 rain at room temperature. Faecal supernatants were stored at - 20 ° C in 0.5 ml aliquots.

2.3. Serum antibody ELISA The assay was optimised as described previously (Gasser et al., 1993 ). In brief, microtitration plates (Greiner, Germany) were coated overnight at 4°C with approximately 15/zg m l - ~ protoscolex somatic antigen diluted in bicarbonate-carbonate buffer, pH 9.6 (BCB) for IgG and IgA assays and 30 #g m1-1 for the IgE assay. Plates were washed three times with mouse tonicity phosphate-buffered saline, pH 7.2 (MT-PBS) containing 0.3% v/v Tween 20 (MT-PBS-T) and 50 /zl of MT-PBS-T containing 5% w/v skim milk powder (MT-PBS-TB) were added to wells. Fifty microlitres of serum, diluted 1:25 were incubated in MT-PBS-T for 1 h at 37°C. Plates were emptied, and washed three times with MT-PBS-T and 50 #1 affinity purified peroxidase conjugated sheep anti-dog IgG ( h + l chain) (Bethyl) diluted 1/6000 in MT-PBS-TB or goat anti-dog IgA (1:1000, Bethyl) or rabbit anti-dog IgE (1:50, Iatrics) were incubated for 1 h (IgG and IgA) or 3 h (IgE) at 22-24 ° C. Plates were subsequently washed three times with MT-PBST and three times with distilled water. Substrate solution containing 50/zl per well of 3,3',5,5'-tetramethylbenzidine (TMB) (KPL) was incubated for 30 min at 37 oC. The reaction was stopped by adding 50/11 of 1 M orthophosphoric acid. The absorbance at wavelength 450 nm was determined when a reference positive control serum had reached A450 of 1.0. The absorbance value of each serum was corrected back to a predetermined positive control A450 of 1.0 (McLaren et al., 1981 ). Positive-negative cut-off absorbance values of 0.40 (IgG), 0.35 (IgA) and 0.45 (IgE) were based on three standard deviations above the mean optical density (OD) value for uninfected and heterologous helminth species infected dogs (Gasser et al., 1993).

2.4. Coproantigen ELISA The assay was optimised as described previously (Allan et al., 1992). In summary, approximately 5 #g m l - ~of hyperimmune rabbit IgG against E. granulosus proglottis somatic extract diluted in BCB was used to coat ( 100/zl per well) microtitre plates (Immulon 1, Dynatech) overnight at 4 °C. Plates were washed three times with PBS/0.1% Tween (20 PBS-0.1% TW) and 'blocked' with 100/zl per well of PBS-0.3% TW for 1 h at room temperature. Fifty microlitres per well of faecal supernatant plus 50 #1 per well of heat inactivated foetal calf serum (FCS, Seralab) were added to each well and incubated (in duplicate) for 1 h, then washed three times in PBS-0.1% TW. One hundred microlitres per well of peroxidase conjugated rabbit IgG anti-E, granulosus proglottis somatic extract (diluted 1:200) was incubated for 1 h at room temperature and plates subsequently washed as above. Substrate solution was 5 aminosalicylic acid (Sigma) in 0.1 M

P.S. Craig et al. / Veterinary Parasitology 56 (1995) 293-301

297

phosphate buffer pH 6.0 containing 1 mM Na2 EDTA and 0.005% hydrogen peroxide. Absorbance at a wavelength of 450 nm was measured after 25 min (Dynatech MR500 microplate reader). Each plate incorporated known reactive and negative dog faecal supernatants and a positive-negative cut-off of 0.09 (i.e. mean _+3 SD) was employed based on uninfected and heterologous species infections (Allan et al., 1992 ). Non-endemic faecal samples were also used from ten UK dogs. 3. Results

3.1. Parasitological findings A total of 257 dogs were registered in La Paloma, of which 69 (26.8%) did not successfully purge following standard dosing with arecoline hydrobromide, or could not be dosed. Of the remaining 188 dogs, 48 (25.5%) were purge positive after laboratory examination for E. granulosus adult worms (range 1-4331 ). Sensitivity of purge examination for Echinococcus increased significantly using laboratory examination compared with the field diagnosis records. Following purge examination, 11.7% (22/188) were positive for Taenia spp. and 17.5% (33/188 ) for Dipylidium caninum.

3.2. Immunodiagnostic tests Matched samples from dogs, i.e. full purge and laboratory examination, with respective blood and stool specimens, were obtained from a total of 110 dogs (Table 1 ). Twenty-six (23.6%) of the 110 dogs were purge positive for Echinococcus and the numbers of worms counted for each dog purge are listed in Table 2. Based on these 26 Echinococcus infected dogs the coproantigen ELISA had a sensitivity of 76.9% (20/26). The serum IgG-ELISA had a sensitivity of 34.6% Table 1 Comparison ofcoproantigen and serum antibody (IgG, IgA, IgE) tests with arecoline purge positivity for Echinococcus granulosus in 110 dogs Test

Purge +

Purge-

Not purged/ doseda

Total

Coproantigen + CoproantigenIgG+ IgGIgA+ IgAIgE+ IgETotal

20 6 9 17 12 14 10 16 26

1 63 17 47 19 45 30 34 64

4 16 7 13 7 13 10 10 20

25 85 33 77 88 72 50 60 110

a Dogs that failed to purge, and dogs that were not dosed because of pregnancy, old age or illness.

P.S. Craig et al. / VeterinaryParasitology 56 (1995) 293-301

298

Table 2 Parasitologicaland immunodiagnostic data for 26 Echinococcus granulosus infected dogs

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Dog No.

Eg

Other

CAg

lgG

lea

IgE

470 504 527 515 464 383 380 400 398 456 528 516 477 460 520 545 415 521 529 522 531 535 543 556 502 503

1 1 2 2 2 2 3 3 4 4 14 17 17 19 20 22 27 69 71 95 123 182 243 2384 4331 +

D T,D T D T T T,D T,D T,D

0.143+ 0.0720.0750.230+ 0.0880.0880.135+ 0.0560.150+ 0.136+ 0.125+ 0.130+ 0.110+ 0.0360.140+ 0.105+ 0.160+ 0.170+ 0.172+ 0.136+ 0.206+ 0.116+ 0.137+ 0,202+ 0.300+ 0.140+

0.2790.430+ 0.2880.611+ 0.2190.2730.1910.629+ 0.2490.486+ 0.690+ 0.2160.2590.3590.3470.1950.590+ 0.980+ 0.413+ 0.2970.2170.3500.419+ 0.1690.3300.197-

0.2230.455+ 0.355+ 0.1450.2720.1740.2440.556+ 0.2050.539+ 2.25+ 0.2580.2230.2430.441+ 0.517+ 0.2820.899+ 0.878+ 0.2190.497+ 0.2600.703+ 0.2860.2680.466+

0.1170.499+ 0.07800.2480.1180.466+ 0.3860.585+ 0.502+ 0.3440.05300.495+ 0.06900.621 + 0.2920.231 0.661 + 0.1730.3680.3110.557+ 0.281 0.804+ 0.1760.3130.641 +

T,D D T,D T,D T,D T,D T,D T,D -

Eg, number ofE. granulosus worms in purge ( +, worm count not done); Other, presence or absence of Taenia spp. (T) or Dipylidium (D); CAg, coproantigen ELISA result for dog faecal supernatants; IgG, IgA and IgE antibodies in dog serum; +, positive result in ELISA ( > 3 SD above control means). Sensitivity of CAg ELISA, 76.9% (20/26); sensitivity of IgG ELISA, 34.6% (9/26); sensitivity of IgG/IgA/IgE ELISAs, 69.2% (18/26); sensitivity of CAg and antibody ELISAs combined, 96.2% (25/26). ( 9 / 2 6 ) , I g A o f 4 6 . 2 % ( 1 2 / 2 6 ) a n d IgE o f 38.5% ( 1 0 / 2 6 ) ( T a b l e 1 ). T h e sensit i v i t y o f t h e a n t i b o d y E L I S A i n c r e a s e d t o 6 9 . 2 % ( 1 8 / 2 6 ) w h e n s e r o p o s i t i v i t y for e i t h e r I g G , I g A o r IgE w a s c o n s i d e r e d , a n d w a s 9 6 . 2 % ( 2 5 / 2 6 ) f o r c o m b i n e d c o p r o a n t i g e n a n d a n t i b o d y ( I g G , A, E ) p o s i t i v i t y . T h e s t a t i s t i c a l a g r e e m e n t b e tween the coproantigen test and arecoline purgation was good (x=0.8) but a g r e e m e n t w a s l o w f o r s e r u m I g G a n t i b o d i e s ( x = 0.02 ). T h e r e w a s w e a k b u t sign i f i c a n t p o s i t i v e c o r r e l a t i o n ( r = 0.65 P < 0.01 ) b e t w e e n c o p r o a n t i g e n E L I S A O D v a l u e s a n d Echinococcus w o r m c o u n t s f r o m p u r g e s , b u t n o c o r r e l a t i o n ( r = 0 . 1 4 ) b e t w e e n s e r u m I g G - E L I S A O D v a l u e s a n d p u r g e Echinococcus w o r m n u m b e r s . I n a d d i t i o n t o t h e s e 26 Echinococcus p o s i t i v e d o g s , a n o t h e r five a n i m a l s o f t h e 1 10 w e r e c o p r o a n t i g e n p o s i t i v e b u t f a i l e d t o p u r g e ( o n e w a s p u r g e n e g a t i v e ) .

P.S. Craig et al. / Veterinary Parasitology 56 (1995) 293-301

299

Seventeen dogs were serum IgG positive but purge negative and another seven seropositives (IgG) failed to purge. A total of 49 dogs were positive for leA or IgE and negative on purge examination (Table 1 ). 4. Discussion

Accurate detection ofE. granulosus infection in the definitive host, the dog, is important to establish baseline data on prevalence, and in surveillance of hydatid control programmes (WHO, 1981 ). Low power microscopic examination of small intestine scrapings at necropsy is the best parasitological method for detection of Echinococcus; however, in most hydatid endemic regions dog post-mortem examinations are not possible. Arecoline purgation is currently the method of choice for ante-mortem diagnosis, but may be problematic for many reasons: it is very time consuming, its biohazardous nature and its variable sensitivity (Wachira et al., 1990). In the current study, 26% of dogs did not successfully purge. Indirect methods for diagnosis of canine echinococcosis have included egg identification from perianal swabs (Craig et al., 1988 ), antibody serology (Jenkins and Rickard, 1986; Gasser et al., 1988) and recently coproantigen tests (Allan et al., 1992; Deplazes et al., 1992). High genus specificity (over 95%) has been shown for both serologic (IgG antibodies against protoscolex antigens) and coproantigen ELISA tests, together with reasonable sensitivities (over 70%) for detection of natural canine echinococcosis (Gasser et al., 1988; Allan et al., 1992). To date there has been no direct comparison of serological and faecal antigen immunodiagnostic tests for detection ofE. granulosus in dogs. In the present study, post-mortem confirmation of immunodiagnostic results could not be carried out, and therefore the serum antibody and the coproantigen ELISAs were compared against arecoline purgation as the standard. Because arecoline diagnosis of canine echinocoecosis is not itself 100% sensitive (and sensitivity could not be determined in the current study), comparative analysis could only realistically be made with 26 samples from dogs which were proven Echinococcus purge positive. On this basis the sensitivity of the eoproantigen peroxidase ELISA was 76.9% and only 34.6% for the serum IgG-ELISA. The overall sensitivity of the serum antibody ELISAs was increased to 69.2% when positivity to either IgG, IgA or IgE was included. Gasser et al. (1993) previously reported the importance of including IgA and IgE antibody detection to increase sensitivity for detection ofE. granulosus in naturally infected dogs and dingoes, although a degree of geographic variability in sensitivity has been observed. The six false negative coproantigen stool samples came from dogs with low, less than 20 (i.e. individual worm numbers were 1, 2, 2, 2, 3, 19) total worm purge counts. Low purge counts for Echinococcus probably indicate low worm burdens. Coproantigen values were also significantly better correlated with purge worm counts than were serum IgG, IgA or IgE antibody values (respectively r=0.65 vs. r=0.14, P<0.005). The reasons for the low sensitivity of the serum IgG ELISA for diagnosis of canine echinococcosis in the Uruguay study corn-

300

P.S. Craig et al. / Veterinary Parasitology 56 (1995) 293-301

pared with previous Australian studies for circulating antibody detection (72.7% sensitivity, Gasser et al., 1988 ) or even with the Kenyan results (40% sensitivity, Jenkins et al., 1990) are not clear. Formation of circulating immune complexes, parasite induced immunosuppression, nutritional status or possible antigenic strain variation might be factors involved in lowering detectable antibody levels. Naturally infected canids in Australia, especially dingoes, frequently have high worm burdens of Echinococcus, and this may account for higher sensitivity of antibody serology in that country (Gasser et al., 1993 ). Although in experimental infections the threshold for detection of IgG antibodies is 1000 worms per dog, naturally infected dogs with burdens as low as 20 worms per dog can be detected (R.B. Gasser, unpublished data, 1989). Certainly, when combined the overall sensitivity of coproantigen and antibody immunodiaguostic tests was greatly increased (96.2%). Given the high specificity previously recorded for coproantigen detection using post-mortem data (Allan et al., 1992) a positive coproantigen result should be indicative of current infection. In experimental Echinococcus muhilocularis infection in dogs, coproantigens were detected by 5 days post infection (Deplazes et al., 1992). Therefore, the six 'no purge/purge negative' dogs that were coproantigen positive were probably infected with Echinococcus. However, seropositivity, in the absence of coproantigen positivity, suggests the possibility of recent or previous exposure. For population studies, it is useful to use serum antibody tests in combination to enable detection of dogs with previous infection eliminated spontaneously or by owners with drugs (praziquantel), or to rapidly screen dog populations for exposure. In conclusion, the coproantigen test has the potential to replace arecoline testing for the diagnosis of current infection in individual dogs, especially if its sensitivity can be improved. Detection of faecal antigen activity after stool fixation in 5% formalin is also a major advantage for coproantigen tests as stool samples may be stored for weeks at 4 ° C or room temperature (Allan et al., 1990; Deplazes et al., 1990). A number of dogs with very low Echinococcus burdens, however, may not be detected using the current coproantigen ELISA. Previous studies using post-mortem inspection of naturally infected dogs indicated a sensitivity threshold of around 15-30 adult E. granulosus based on a faecal antigen detection (Allan et al., 1992). The current study which was based on arecoline purgation and not post mortem analysis, also resulted in a similar lower limit of detection for the coproantigen ELISA of approximately over 20 worms per purge. We are currently undertaking biochemical analysis of Echinococcus coproantigen positive faeces in order to characterise parasite antigen (s) with a view to producing affinity purified polyclonal antisera or specific monoclonal antibodies to improve sensitivity of the Echinococcus coproantigen test and anticipate such a test will eventually replace arecoline purgation.

Acknowledgements We are extremely grateful to the diligence of the technicians of the dog dosing team (Uruguay Hydatid Commission ) from Durazno, and the hospitality of the

P.S. Craig et al. / Veterinary Parasitology 56 (1995) 293-301

301

people of La Paloma. This research was funded by the WeUcome Trust and the Fundacion San Antonio de Padua. We gratefully acknowledge the support from the Australian Research Council (ARC) and the British Council. Thank you to Jim Allan for comments on the manuscript.

References Allan, J.C., Craig, P.S., Garcia Noval, J., Mencos, E, Liu, D., Wang, Y., Wen, H., Zhou, P., Stringer, R., Rogan, M. and Zcyhle, E., 1992. Coproantigen detection for immunodiagnosis of echinococcosis and taeniasis in dogs and humans. Parasitology, 104: 347-355. Craig, P.S., 1993. Immunodiagnosis ofEchinococcus granulosas. In: F.L. Andersen (Editor), Compendium of Cystic Echinococcosis. Brigham Young University, Provo, Utah, in press. Craig, P.S., Macpherson, C.N.L., Watson-Jones, D.L. and Nelson, G.S., 1988. Immunodetection of Echinococcas eggs from naturally infected dogs and from environmental contamination sites in settlements in Turkana, Kenya. Trans. R. Soc. Trop. Med. Hyg., 82: 268-274. Deplazes, P., Gottstein, B., Stingelin. Y. and Eckert, J., 1990. Detection of Taenia hydatigena coproantigens by ELISA in dogs. Vet. Parasitol., 36: 91-103. Deplazes, P., Gottstein, B., Eckert, J., Jenkins, D.J., Ewald, D. and Jimenez-Palacios, S., 1992. Detection ofEchinococcus coproantigens by enzyme-linked immunosorbent assay in dogs, dingoes and foxes. Parasitol. Res., 78: 303-308. French, C.M. and Nelson, G.S., 1982. Hydatid disease in the Turkana District of Kenya II A study in medical geography. Ann. Trop. Med. Parasitol., 76: 439-457. Gasser, R.B., Jenkins, D.J., Paolillo, E., Parada, L., Cabrera, P. and Craig, P.S., 1993. Serum antibodies in canine echinococcosis. Int. J. Parasitol., 23: 579-586. Gasser, R.B., Lightowlers, M.W., Obendorf, D.L., Jenkins, D.J. and Rickard, M.D., 1988. Evaluation of a serological test system for the diagnosis of natural Echinococcus granulosas infection in dogs using E. granulosus protoscolex and oncosphere antigens. Aust. Vet. J., 65: 369-373. Gemmell, M.A., 1990. Australasian contributions to an understanding of the epidemiology and control of hydatid disease caused by Echinococcus granulosas--past, present and future. Int. J. Parasitol., 20:431-456. Gemmell, M.A., Lawson, J.R. and Roberts, M.G., 1986. Control ofechinococcosis/hydatidosis:present status of worldwide progress. Bull. W.H.O., 64: 333-339. Jenkins, D.J. and Rickard, M.D., 1986. Specific antibody responses in dogs experimentally infected with Echinococcusgranulosus. Am. J. Trop. Med. Hyg., 35: 345-349. Jenkins, D.J., Gasser, R.B., Zeyhle, E., Romig, T. and Macpherson, C.N.L., 1990. Assessment of a serological test for the detection ofEchinococcus granulosus infection in dogs in Kenya. Acta Trop., 47: 245-248. McLaren, M.L., Lillywhite, J.E. and Au, A.C.S., 1981. Indirect enzyme-linked immunosorbent assay (ELISA): practical aspects of standardization and quality control. Med. Lab. Sci., 38: 245-251. Palmer, S.R. and Biffin, A.H.B., 1987. The changing incidence of human hydatid disease in England and Wales. Epid. Inf., 99: 693-700. Purriel, P., Schantz, P.M., Boevide, H. and Mendoza, G., 1973. Human echinococcosis (hydatidosis) in Uruguay: a comparison of indices of morbidity and mortality 1962-1971. Bull. W.H.O., 49: 395--402. Sehwabe, C.W., 1986. Current status of hydatid disease: a zoonosis of increasing importance. In: R.C.A. Thompson (Editor), The Biology of Echinococcus and Hydatid Disease. George A1lan&Unwin, London, pp. 81-113. Wachira, T., Maepherson, C.N.L. and Gathuma, J.M., 1990. Hydatid disease in the Turkana District of Kenya VII. Analysis of the infection pressure on definitive and intermediate hosts of E. granulosus. Ann. Trop. Med. Parasitol., 84:361-368. World Health Organization, 1981. Guidelines for surveillance, prevention and control of echinococcosis/hydatidosis. J. Eckert, M.A. Gemmell and E.J.L. Soulsby (Editors), FAO/UNEP/WHO. WHO, Geneva, Switzerland.