Ovine echinococcosis

www.elsevier.com/locate/vetpar

Veterinary Parasitology 143 (2007) 112–121

Ovine echinococcosis

I. Immunological diagnosis by enzyme immunoassay

Antonio Gatti b, Angela Rosa Alvarez a, Daniel Araya b, Sergio Mancini b,Eduardo Herrero b, Graciela Santillan c, Edmundo Larrieu a,*

a Veterinary Faculty, UNL Pampa, Argentinab Health Ministry, Rio Negro Province, Argentina

c ANLIS/MALBRAN, Argentina

Received 5 April 2006; received in revised form 25 July 2006; accepted 3 August 2006

Abstract

Immunodiagnosis in sheep presents problems of sensitivity and specificity, limiting its applicability in surveillance systems. The

objective of this study was to develop a sensitive, specific and accessible technique for diagnosing cystic echinococcosis in naturally

infected sheep and to evaluate the validity of necropsy as a reference test. A total of 247 sheep were studied at slaughterhouses,

confirming the parasitological diagnosis with histology. Serum was processed with enzyme immunoassay (EIA) using three antigen

preparations: total hydatid liquid (LHT), purified fraction of LHT (S2B) and purified lipoprotein (B). Western Blot (WB) was used

as a control. EIA proved effective for differentiating Echinococcus granulosus from larval stage of Taenia hydatigena and intestinal

cestodes in all three antigen preparations. Serums from macroscopically negative sheep were reactive to EIA and positive with WB.

In the whole flock, sensitivity was 89.2% for LHT, 80.0% for S2B and 86.4% for B. Sensitivity in lambs was 78.6% for LHT, 75.0%

for S2B and 64.3% for B. Macroscopic diagnosis at the time of slaughter was found to have limitations as a reference test for

immunodiagnosis of cystic equinococcosis in sheep, so it was necessary to include histology and WB as reference tests. LHT was

the antigen preparation of greatest value and EIA proved to be a sensitive and specific technique, adequate for surveillance systems

and for evaluating control programmes.

# 2006 Elsevier B.V. All rights reserved.

Keywords: Cystic equinococcosis; Sheep; Immunodiagnosis; Enzyme immunoassay

1. Introduction

Cystic echinococcosis is a parasitic zoonosis

caused by an internal parasite of the phylum

Platyhelminthes, class Cestoda, order Cyclophyllidae,

family Taeniidae, genus Echinococcus and specie

granulosus.

The life cycle involves two mammalian hosts. The

definitive hosts are carnivores (especially dogs) in

* Corresponding author. Tel.: +54 2920 430007;

fax: +54 2920 430007.

E-mail address: [email protected] (E. Larrieu).

0304-4017/$ – see front matter # 2006 Elsevier B.V. All rights reserved.

doi:10.1016/j.vetpar.2006.08.022

which the adult or strobilocercus form is developed;

intermediate hosts are ungulates, especially sheep,

in which the larvae or metacestode forms develop

(Thompson and Mcmanus, 2001).

Information on immunological diagnosis is limited

in sheep. Indirect hemagglutination (Yong et al., 1978;

Conder et al., 1980; Bakos et al., 1985), double

diffusion (Yong and Heath, 1979; Conder et al., 1980)

and enzyme immunoassay (EIA) (Craig and Rickard,

1981; Yong et al., 1984; Lightowlers et al., 1984; Ming,

1986; Lloyd et al., 1991) have all been tried with little

success: none provided a consistent serological test for

livestock (Craig, 1997).

A. Gatti et al. / Veterinary Parasitology 143 (2007) 112–121 113

It has been pointed out that the main limitations to

immunodiagnosis in sheep are the cross reactions

observed with other taenias (larvae of Taenia hydati-

gena and Taenia ovis), appearance of false positives and

low sensitivity (Lightowlers and Gottstein, 1995; Craig,

1997; Eckert et al., 2001).

Necropsy is chosen as the best diagnostic test (Eckert

et al., 2001) and is also used as a reference test to estimate

the sensitivity and specificity of immunodiagnostic tests.

The development of a screening test is important for the

identification of cystic hydatid carriers when animals are

imported from endemic areas to areas free of infection

(Craig, 1997; Eckert et al., 2001), and for control

programmes identifying livestock farms with transmis-

sion foci or for epidemiological surveillance to assess the

prevalence of infection (Cabrera et al., 2003).

The objective of this study was to develop a specific,

sensitive and simple immunological diagnostic screen-

ing technique and to evaluate the efficiency of necropsy

as a reference test for immunodiagnosis tests.

2. Materials and methods

2.1. Working area and population

Naturally infected sheep were studied at the time of

slaughter in five abattoirs of the Province of Rio Negro

where the prevalence of infection is 18% (Larrieu et al.,

2001).

Sheep were classified according to their age into the

following groups: rearing lambs (R), 0–6 months old;

fattening lambs (F), over 6 months and under 24 months

old; adults (A), 24 months old and over.

A total of 247 sheep were studied out of which 85

were R (34%), 60 F (24.3%) and 102 adults (41.3%).

Regarding gender, 176 (71.3%) were males and 71

(28.7%) were females.

Blood samples (10 ml) were obtained from the

jugular vein at the time of slaughter and collected in

disposable plastic tubes which were marked with the

animal’s identification number.

Serum was obtained by centrifugation, refrigerated

at 5/8 8C and sent within 48 h to the laboratory where it

was stored at �20 8C until processed.

2.2. Enzyme inmunoassay (EIA) test

The EIA was developed as the screening test with

three different types of antigen preparations: total hydatid

liquid (LHT, obtained as described by Coltorti, 1986),

purified fraction of LHT (S2B, following Coltorti et al.,

1990) and purified lipoprotein (B, obtained as described

by Oriol et al., 1971). The National Institute of Micro-

biology ‘‘Carlos G. Malbran’’ provided the antigens.

For the EIA with LHT antigen (EIA.LHT), each

antigen vial was reconstituted with sufficient sensitizer

buffer (carbonate/bicarbonate, pH 9.6) to obtain an

optimal protein concentration as determined by the

Bradford Method (1976).

The plates (micro strips) were sensitized with 1 mg/

well of antigen. Each well was seeded with 50 ml of

diluted antigen, kept for 18 h in a moist chamber and

then washed three times, each for 4 min, with washing

buffer (PBS 0.3% Tween).

In each well, 100 ml of serum diluted 1/800 in PBS–

3% milk–0.3% Tween were added and left for 45 min at

37 8C. The liquid was then aspirated and washed three

times with washing buffer. Ovine conjugated anti-

gammaglobulin marked with peroxidase diluted 1/2500

in PBS–3% milk–0.3% Tween (100 ml) was added and

left for 45 min at 37 8C. The liquid was then aspirated

and washed three times with washing buffer.

ABTS solution (Sigma1) (100 ml), incubated 45 min

at 37 8C, was used as the developer substrate with the

addition of 100 ml of braking solution (SDS 1%).

Positive control serum was obtained from sheep with

hydatidosis confirmed by histology. Negative control

serum was obtained from sheep with no hydatidosis,

cysticercosis nor intestinal cestodes. The optimal

dilution of the control and problem serum was defined

through serial dilutions of antiserum and conjugated

antigen to obtain the maximum dilution that allows

differentiation of low positive, high positive and

negative control samples, using the same protocol in

each case. Problem and control serums were seeded and

duplicates of each serum were processed. Each micro

plate included three negative serums and one positive.

A vertical plate reader with 405 nm filter was used

for reading the samples. The cut-off value was defined

with receiver operating characteristic (ROC) curves

using an Excel spreadsheet according to the design by

Grainer (1995).

For EIA with S2B antigen (EIA.S2B), each antigen

vial was reconstituted with sensitizer buffer (PBS 7.2)

and the plates (micro strips) were sensitized with an

antigen concentration of 0.15 mg/well. Serum was

diluted 1/400.

For EIAwith B antigen (EIA.B), each antigen vial was

reconstituted with sensitizer buffer (PBS 7.2) and the

plates (micro strips) were sensitized with an antigen

concentration of 0.15 mg/well. Serum was diluted 1/400.

Control serums were seeded, read and the estimation of

the cut-off value was carried out in the same way as for

LHT.

A. Gatti et al. / Veterinary Parasitology 143 (2007) 112–121114

Table 1

Relationship within positive and negative EIA results to Echinococcus

granulosus with three antigen preparations and two cut-off values and

their relationship with positive and negative control test results

EIA/histology and WB Antigen

LHT 0.157/

0.200

S2B 0.073/

0.106

B 0.068/

0.078

+/+ 58/55 52/48 57/56

� 19/6 11/2 9/5

�/+ 7/10 13/17 13/10

�/� 163/176 171/180 168/176

Total 247 247 247

Sheep naturally infected. Rio Negro Province, 2003/2004.

2.3. Reference tests

Macroscopic diagnosis at necropsy was chosen as

the reference test. Confirmation of those resulting

positive was done by histology and the Western Blot

(WB) test.

The selected sheep were inspected at necropsy

looking for the presence of Echinococcus granulosus,

larvae stage of T. hydatigena, larvae stage of T. ovis and

intestinal cestodes.

Liver, lung, heart, kidney and omentum were

examined visually and through palpation and cuts, to

determine the presence of cysts in the parenchyma. Cuts

and palpation were also carried out in the masseter

muscle. Lastly, the small intestine was opened long-

itudinally. Hydatid cysts or cysts belonging to other

macroscopically atypical parasites were sampled

(n = 26 cystic of 20 sheep) at necropsy for histological

confirmation. Samples were maintained in 5% for-

maldehyde until sent to the laboratory.

Samples were stained with hematoxylin–eosine

(Jubb et al., 1993). Calcified cysts and/or contaminated

cysts were stained using the PAS technique (Jubb et al.,

1993).

Sheep that were positive to the screening test and

negative to the macroscopic diagnosis were studied

using the WB test. Proteins were separated using

SDSPage (Laemmli, 1970) with polyacrilamide gel of

0.75 mm thickness, in a Mini Protean II chamber from

Bio Rad. The problem samples were seeded together

with one positive and one negative control and

electrophoresis was carried out at 200 V, 60 mA and

30 W for approximately 1 h.

The samples were transferred to nitro-cellulose with

the corresponding buffer for 1 h at 100 V, 250 mA,

60 W.

The immuno-enzymatic reaction was performed

by blocking the nitro-cellulose with PBS Tween 0.5%

milk and stirring gently for 1 h. Then sheep serum

diluted 1/100 was added, incubated for another hour

while gently stirring, followed by three washes

with PBS 0.5% Tween. Lastly the samples were

incubated with anti-sheep IgG peroxide conjugate,

washed out with PBS 0.5% Tween and developed with

DAB.

Samples with five (52–67 kDa) or antigen B (8–12,

16 and 24 kDa) precipitation bands were classified as

positive.

Results from sheep that were EIA positive but

negative macroscopically at necropsy were compared to

expected prevalence for naturally infected sheep and to

data from flocks in Peru.

2.4. Estimates of the sensitivity and specificity of

enzyme inmunoassay

Sensitivity and specificity were estimated by normal

approximation with 95% confidence levels for each age

group (rearing lambs, fattening lambs and adults) and

for each antigen preparation (LHT, S2B and B).

Positive and negative predictive values were esti-

mated for an expected prevalence of 18%.

Calculations were carried out with Epidat 3.0

Software (Xunta of Galicia—PAHO/WHO).

3. Results

3.1. Immunological diagnosis with enzyme

inmunoassay (EIA)

Considering histology as the confirmation test for a

positive diagnosis together with the optical density

(OD), out of 247 sheep studied the cut-off values were

0.200 (IC95% 0.18–0.23) for EIA.LTH; 0.106 (IC95%

0.10–0.12) for EIA.S2B and 0.078 (IC95% 0.074–

0.087) for EIA.B.

When the WB test was incorporated as a control test

(ROC analysis), the cut-off values were 0.157 (IC95%

0.137–0.166) for EIA.LTH, 0.073 (IC95% 0.068–0.082)

for EIA.S2B and 0.068 (IC95% 0.039–0.077) for EIA.B.

The relationship between positive and negative EIA

results and negative and positive results to the control test

coincided in 94% of the cases with LHT, 91% with S2B

and 97% with antigen B and are presented in Table 1.

Using EIA.LHT, 77 sheep were classified as positive

with the OD 0.157 cut-off value and 61 sheep with the

OD 0.200. Using EIA.S2B, 63 sheep were classified as

positive with the OD 0.073 cut-off value and only 42

sheep with the OD 0.106. Finally using EIA.B, 66 sheep

were classified as positive with OD 0.068 and 61 sheep

with the OD 0.078.

A. Gatti et al. / Veterinary Parasitology 143 (2007) 112–121 115

Fig. 1. EIA capability to discriminate naturally infected ovines with Echinococcus granulosus, T. hydatigena, larval stage, intestinal cestodes and no

parasitized ovines. Rio Negro Province, 2003/2004.

A. Gatti et al. / Veterinary Parasitology 143 (2007) 112–121116

Fig. 2. Mean and standard error from ANOVA analysis in five ovine

groups. Rio Negro Province, 2003/2004.

The optical densities obtained from negative and

positive sheep with different parasites and their position

in relation to the cut-off values are presented in Fig. 1.

The global Kappa test comparing the three antigen

preparations gave a figure of 0.83, with no significant

differences being observed (Chi square of homogeneity

of Kappa; p: 0.47).

3.2. Reference test

Presumptive diagnosis of the presence of E.

granulosus, larval stage of T. hydatigena and intestinal

cestodes was performed on 44 sheep (17.8%), 18 sheep

(7.3%) and 11 sheep (4.5%), respectively. Of the total,

174 sheep (70.4%) were diagnosed as negative and no

case of larval stage of T. ovis was found.

Of the 23 cysts of 18 sheep sent to the laboratory with

a presumptive diagnosis of echinococcosis, 1 (4.3%)

was discarded (the definite diagnosis being caseous

lymphadenitis) and 22 (95.7%) were confirmed. Three

cysts of two sheep with no presumptive diagnosis of

cystic echinococcosis were classified as echinococcosis

(two complicated cysts and one calcified cyst). One of

these cases presented live oncosphere of E. granulosus

in the liver. Altogether, 15.4% (4/26 samples) of the

presumptive diagnoses were modified by histology.

Finally, considering the histological diagnoses as

control tests, 45 sheep were confirmed as having

echinococcosis, with 140 cysts being identified (average

3.1 cysts/sheep).

E. granulosus cyst locations were: 17 in the liver

(37.8%), 11 in lungs (24.4%) and 17 in both liver and

lung (37.8%). Gender distribution of E. granulosus was

statistically different ( p: 0.02) with 19 females (28.8%)

and 26 males (15.2%) showing the disease. Prevalence

increased significantly in relation with age (Chi square

lineal tendency p: 0.0007, OR 4.4 for adults).

A total of 24 WB tests were carried out, with 20

(83.3%) positive and 4 (26.7%) negative. Of the sheep

classified at necropsy as macroscopically negative to

equinococcosis and other parasites and showing titers of

EIA.LHT above the cut-off value, 16 (80%) proved

positive and 4 (20%) negative.

One sheep (100%), classified as E. granulosus

negative and intestinal cestodes positive at necropsy,

gave a titer above the cut-off value of EIA.LHT and was

classified as positive with the WB test.

All three sheep that were macroscopically negative

to echinococcosis at necropsy and had a EIA.LHT titer

below the cut-off value, but with EIA.B antigen higher

than the cut-off value, were classified as positive with

the WB test.

Finally, considering the histological diagnoses and

WB as control tests, 65 sheep (26.3%) were classified as

positive for E. granulosus and 182 (73.7%) were

classified as negative.

ANOVA analysis was performed to compare the

optical density in five groups of sheep: those negative at

necropsy and positive to WB (group 1); those positive at

necropsy, with positive histology and EIA.LHT (group

2); those negative both at necropsy and with EIA.LHT

(group 3); those with T. hydatigena larvae (group 4);

and those with intestinal cestodes (group 5). Significant

differences (F : 17.1 p: 0.00) were found between

groups. Individual comparisons using the protected

Fisher minimal significant differences test showed

significant differences between groups 2 and 3 ( p:

0.00), 1 and 2 ( p: 0.046) and 2 and 4 ( p: 0.0001); while

no significant differences were observed between

groups 3, 4 and 5 ( p > 0.05). Fig. 2 shows a summary

of the results obtained from ANOVA/DMS.

A comparative analysis of sheep positive to EIA and

negative to the macroscopic examination showed that

when the prevalence of infection in the flock increased,

the rate of positive EIA and macroscopically negative

cases increased proportionally, this tendency being

statistically significant (Chi square for lineal tendency

p: 0.00) (Table 2).

3.3. Simple diagnosis tests and ROC analysis

Using histology as the control test, the S2B antigen

preparation showed greatest sensitivity, both for thewhole

flock (82.2) and for lambs (62.5). Incorporating WB as a

control test, LHT antigen turned out to be the most

sensitive preparation for both flock (84.6–89.2) and lambs

(78.6 with two cut-off levels). In all cases, the older the

animals the more sensitive the tests proved to be (Table 3).

S2B antigen preparation showed the greatest speci-

ficity for immunodiagnosis of ovine echinococcosis

A. Gatti et al. / Veterinary Parasitology 143 (2007) 112–121 117

Table 2

EIA positive percentage from negative necropsy and its relationship with the echinococcosis sheepfold prevalence

Research area Sheepfold echinococcosis prevalence Percentage of sheep negative from

necropsy and positive to EIA

No. + % No. + %

Perua (endemic, with no control program) 212 152 71.7 60 47 78.3

Rio Negro (endemic, with control program) 247 45 18.2b 202 23 11.4

Raised sheep free of infectionc,d 43 0 0.0 42 1 2.3

36 0 0.0 36 0 0.0

Chi-square for lineal tendency p: 0.00.a Dueger and Gilman (2001); Dueger et al. (2003).b Larrieu et al. (2001).c Moro et al. (1997).d Moro et al. (1999).

Table 3

Sensitivity estimation of immunologic diagnoses through EIA with LHT, SB2 and B antigens and distinct cut-off values in different age groups,

applying histology and WB as control tests

Age group Antigen

LHT S2B B

0.200 0.157 0.106 0.073 0.078 0.068

0–6 months old 78.6 (76.7–80.5) 78.6 (76.7–80.5) 71.4 (69.5–73.3) 75.0 (73.1–76.9) 64.3 (60.6–68.2) 64.3 (60.6–68.1)

7–23 months old 100 (90.0–100) 100 (90.0–100) 80.0 (74.8–85.1) 90.0 (84.9–95.1) 80.0 (74.8–85.1) 90.0 (84.9–95.1)

24 months old and over 90.6 (89.0–92.2) 96.8 (95.3–98.3) 77.3 (76.1–78.5) 96.8 (95.3–98.5) 92.7 (91.4–93.9) 90.9 (89.7–92.1)

Total 84.6 (83.8–85.4) 89.2 (88.4–90.1) 72.7 (71.9–73.6) 80.0 (79.2–80.8) 84.5 (84.0–85.6) 86.4 (85.6–87.1)

Note: Value (CI 95%).

using histology as the control test, both for the whole

flock (94.1) and for lambs (96.1). When using WB as the

control test, S2B antigen preparation again showed

greatest specificity, both for the whole flock (99.4–93.9)

and for lambs (100–92.9). No difference in specificity

was observed for the older animals (Table 4).

The positive predictive value (for an expected

prevalence of 18%) was greater for S2B with both

histology as control test (75.3) and WB as control test

Table 4

Specificity estimation of immunologic diagnoses through EIA with LHT, S

applying histology and WB as control tests

Age group Antigen

LHT S2B

0.200 0.157 0.106

0–6 month old 94.7 (93.8–95.2) 89.5 (88.5–90.4) 100 (99.

7–23 month old 100 (100–100) 92.7 (91.8–93.7 100 (99–

24 month old and over 97.1 (96.4–97.8) 87.1 (86.4–87.9) 96.8 (9

Total 97.2 (96.9–97.5) 89.5 (89.2–89.8) 99.4 (9

Note: Value (CI 95%).

(96.4–74.2). On the other hand, the negative predictive

value was greater for S2B with histology as the control

test (96.0) and for LHT with WB as the control test

(96.2–97.7) (Table 5). Estimating the area under the

ROC curve, LHT has the greatest capacity to

discriminate (0.930 IC95% 0.86–0.97). The result of

comparing the areas beneath the ROC curve for the

three antigen preparations was not significant (Chi

square homogeneity test, p: 0.07).

B2 and B antigens and distinct cut-off values in different age groups,

B

0.073 0.078 0.068

1–100) 92.9 (92.1–93.9) 98.6 (97.8–99.3) 94.4 (93.6–95.1)

100) 98.0 (96.9–99.0) 94.0 (92.9–95.1) 88.0 (86.9–89.1)

6.1–97.6) 87.1 (86.4–87.9) 98.3 (97.5–99.2) 92.1 (91.2–92.9)

9.2–99.7) 93.9 (93.7–94.2) 97.2 (96.9–97.5) 92.8 (92.5–93.1)

A. Gatti et al. / Veterinary Parasitology 143 (2007) 112–121118

Table 5

Results from single diagnostic test and ROC analysis with different

antigenic preparations applying histology/WB as control tests

Antigen

LHT S2B B

Cut-off value 0.157 0.200 0.073 0.106 0.068 0.078

Sensitivity 89.2 84.6 80.0 72.7 86.4 84.5

Specificity 89.5 97.2 93.9 99.4 92.8 97.2

Predictive value

Positive (+) 67.9 89.8 74.2 96.4 72.5 86.8

Negative (�) 97.7 96.2 95.6 94.3 96.9 96.6

Area under ROC curve 0.930 0.860 0.890

4. Discussion

4.1. Screening tests for ovine cystic echinococcosis

In this study, EIA proved efficient, with the three

antigen preparations used, in discriminating sheep with

E. granulosus from carriers of other parasites (larval

stage of T. hydatigena and intestinal cestodes). When

choosing a cut-off value for maximum specificity by

means of ROC analysis, only one sheep with T.

hydatigena larvae was classified as positive for echino-

coccosis with LHT antigen, whereas with antigens B and

S2B no animal was classified incorrectly. When a cut-off

value for maximum sensitivity was applied, only one

animal gave a false positive with S2B antigen, two

animals with antigen B and five animals with antigen

LHT.

All three antigen preparations showed animals with

serology results above the cut-off level but which were

negative at necropsy (possible false positives). But most

of these animals proved to be positive when using the

Western Blot technique as the control test (80.9% with

LHT antigen, 91.7% with S2B and 94.7% with B).

Estimating the area under the ROC curve, LHT

discriminated best between positive and negative sheep,

although differences between the three antigen pre-

parations were not statistically significant ( p > 0.05).

Similar results were obtained when using the Kappa

global test, showing no statistical differences

( p > 0.05), and the crude estimated concordance was

high for all possible combinations (0.93–0.95), con-

firming that there was little difference between the use

of one or other of the three antigens.

Total antigen (LHT) obtained from hydatid cystic

liquid was the antigen preparation with greatest

sensitivity (89.2–77.8%, depending on the cut-off value)

while the highest specificity was reached with antigen

S2B (93.9–99.4%, depending on the cut-off value).

The greatest positive predictive value, for the

expected prevalence in the Province of Rio Negro,

was obtained with antigen S2B (96.4%); the greatest

negative predictive value with LHT (97.5%).

Analysis of these results shows that either LHT or

S2B antigens could be used to diagnose cystic

echinococcosis, although ease of production and

availability would point to LHT as the antigen of

choice. Antigen B is less appropriate.

In lambs, total antigen (LHT) also showed the

greatest sensitivity (78.6%) and antigen S2B the

greatest specificity (100%). These results are especially

valuable for applying immunodiagnosis in epidemio-

logical surveillance programmes; the identification of

transmission in the recent past (expressed as infection in

young animals) is an important component of control

programmes.

The optimum cut-off value by EIA.LHT in

epidemiological surveillance programmes would be

OD 0.157 (maximum sensibility and negative predictive

value), and in elimination programmes OD 0.200

(maximum specificity and positive predictive value).

Histological findings of oncospheres of E. granulo-

sus in the livers of naturally infected sheep which were

classified as negative macroscopically but as positive by

immunodiagnosis confirm the advantage of the enzyme

immunoassay for identifying recent infections when

compared to necropsy examination.

A similar increase in seroprevalence and in numbers

of cysts with age has been reported (Torgerson and

Heath, 2003) and was associated with the dynamics of

the metacestode and could explain the increase in

sensitivity and specificity with age seen in sheep in our

study.

Comparing previous studies, we note that initially

Yong et al. (1984), using native antigens from ovine

hydatid liquid with EIA in animals experimentally

infected with a variety of cestoda, attained sensitivity

and specificity of only 32 and 65%, respectively. Thus,

they did not recommend its application for cystic

echinococcosis surveillance programmes.

Ibrahem et al. (1996) attained sensitivity and

specificity of 36 and 93%, respectively, using antigen

from sheep hydatid liquid and of 71 and 96% using

antigens from camel hydatid liquid. Vargas et al. (2002)

using ovine hydatid liquid antigen, attained 83 and 75%

and Kittelberger et al. (2002) using protoscolex

antigens, attained 64.6 and 95%. None of these matched

the specificity and sensitivity attained in the present

study using EIA.

Ibrahem et al. (1996), using antigen B, attained

sensitivity and specificity of 57 and 93%, respectively,

A. Gatti et al. / Veterinary Parasitology 143 (2007) 112–121 119

using sheep antigen and of 90 and 99% using camel

antigen. Kittelberger et al. (2002), attained sensitivity

and specificity of 9.7 and 99.5% using antigen B of

ovine hydatid liquid; sensitivity attained in this study

using antigen B was 84.5/86.4% (depending on the

cut-off value) and specificity was 97.2/96.8%.

Thus, EIA is shown to be a specific and sensitive

technique for identifying the transmission of E.

granulosus from dogs to sheep, including lambs,

supporting its application for the diagnosis of cystic

echinococcosis in sheep. LHT would be the antigen

of choice due to its ease of production and availability.

4.2. Reference tests for serological diagnosis of

ovine echinococcosis

The present research shows that 15.4% of the

presumptive macroscopic diagnoses at necropsy were

modified by the histological confirmation test, detecting

false positive and false negative diagnoses in the

different age groups studied.

These results confirm previous studies carried out in

the Province of Rio Negro in which 37.2% of false

positive diagnoses identified at necropsy were caused

by unspecific granulomas, pseudo tuberculosis, emphy-

sema and fatty degeneration while 1.1% of false

negative diagnoses were due to small intra-parenchyma

cysts (Larrieu et al., 2001). This information coincides

with research carried out in Uruguay (Cabrera et al.,

1996) in which 26.1% of the sheep that were classified

as positive at necropsy were not hydatidosis at histology

but were mistaken with caseous lymphadenitis, larval

stage of T. hydatigena, white spot or abscesses. Small

intra-parenchyma cysts, especially in lungs, can be

unnoticeable at sanitary inspection. As a result, these

cases are falsely classified as negative at necropsy.

This study identified intra-parenchyma oncospheres

through histology and cysts of recent development

which were considered negative at the macroscopic

inspection (false negative at necropsy). This confirms

the limitations of macroscopic diagnosis at necropsy

and the validity of the histological test as a reference test

for the diagnosis of ovine cystic echinococcosis.

In order to evaluate the specificity and sensitivity

of the immunologic diagnosis in both naturally and

experimentally infected animals, previous studies

almost exclusively used serum panels where the animals

were classified as negative or positive according to the

macroscopic analysis at necropsy (Conder et al., 1980;

Craig and Rickard, 1981; Yong et al., 1984; Bakos et al.,

1985; Lightowlers et al., 1984; Ibrahem et al., 1996;

Moro et al., 1997, 1999; Kittelberger et al., 2002;

Vargas et al., 2002). Only one study used the Gram dye

to perform differential diagnosis with Corynebacterium

pseudotuberculosus (Dueger et al., 2003). This throws

some doubts on the interpretation of the above-

mentioned studies.

In contrast, the use of immunodiagnosis would make

it possible to classify as positive sheep carrying a recent

infection that would have been incorrectly classified as

negative at necropsy.

In areas with a high prevalence of the infection (e.g.

Peru) all animals can be considered to be exposed to

infection. Besides those found positive at necropsy,

71.7% of the animals negative at necropsy are detected

as serum positive (Moro et al., 1999; Dueger et al.,

2003). In areas of limited transmission (like the

Province of Rio Negro) that proportion is 11.4%; for

animals raised in areas free of infection the correspond-

ing rate is 0–2.3% (Moro et al., 1997, 1999). According

to the observations of this study, animals negative at

necropsy but positive at serology could be estimated to

be the product of different stages of the infection

process (oncospheres of a recent infection, recently

developing cysts, incipient hyaline cysts).

Yong et al. (1984) observed a positive antibody

response in three sheep experimentally infected with a

low infestation (10 eggs) of E. granulosus but failed to

find hydatid cysts at necropsy.

Serological conversion 10–14 days after inoculation

(Yong et al., 1984) confirms that animals with recent

infections will be classified as positive in serological

tests and as negative at necropsy.

There are, however, operative limitations to the use

of histological studies as reference tests. It is possible to

confirm all doubtful macroscopic lesions but it would be

laborious to check all the macroscopically negative

diagnoses in an attempt to detect recent infections,

information which would be particularly interesting for

a surveillance system, especially in lambs.

In this study, as an alternative, Western Blot was used

to check negative cases at necropsy that were positive to

the EIA screening test.

Western Blot, based on the identification of species-

specific antigen fractions of 8–12, 16 and 20–24 kDa,

was developed for immunodiagnosis of ovine echino-

coccosis and proved to have a 98.6–100% specificity

(Moro et al., 1997; Vargas et al., 2002) and no cross

reactions with other cestoda. The 8 kDa band was

present in all serum positive animals. WB sensitivity

has been reported as 91.4–97.6% (Dueger et al., 2003;

Vargas et al., 2002).

In this study, analysis of variance was carried out

comparing the optical densities of sera belonging to the

A. Gatti et al. / Veterinary Parasitology 143 (2007) 112–121120

following groups of sheep: those with macroscopically

diagnosed E. granulosus, those negative at necropsy but

positive to EIA and WB, those with other cestoda

infections, and those free of infection. This made it

possible to establish, with precision, the characteristics

common to sheep classified as positive by WB or

necropsy and the statistically significant differences

observed with sheep carriers of other cestoda and sheep

with no infection at all, confirming the specificity of the

WB diagnosis and its usefulness as a reference test.

The use of macroscopic diagnosis at necropsy as a

reference test for sensitivity and specificity estimations

has limitations, as demonstrated in this study.

Assays to be used as screening tests in sheep

populations should include confirmation of diagnosis

using either histology or highly specific serological tests

such as Western Blot.

Acknowledgements

Our acknowledgements go to Dr. Eduardo Guarnera

and Anıbal Franco for their advice on research design

and to Dr. Irma Sommerfelt for manuscript correction.

References

Bakos, E., Zurbriggen, M., Soni, C., Draghi, M., 1985. Diagnostico

serologico de la hidatidosis ovina mediante la hemaglutinacion

indirecta. Vet. Arg. 17, 624–627.

Bradford, M., 1976. Rapid and sensitive method for the quantitation of

microgram quantities of protein utilizing the principle of protein-

dye binding. A. Biochem. 72, 248–257.

Cabrera, P.A., Irabedra, P., Trindado, J., Agulla, J., Heinzen, T.,

Cardozo, M., Alvarez, M., Elola, S., Rista, L., Bondad, M., Haran,

G., Vinals, G., Mangeney, G., Sambran, Y., Valledor, S., Baralbar,

M., Morana, A., Orlando, D., 1996. Determinacion de la preva-

lencia de la hidatidosis ovina en playas de faena, Uruguay, 1994.

Hidatidosis (Uruguay) 2, 18–29.

Cabrera, P.A., Irabedra, P., Orlando, L., Rista, R., Haran, G., Vinals,

G., Blanco, M.T., Alvarez, M., Elosa, S., Morosoli, D., Morana,

A., Bondad, M., Sambran, Y., Heinzen, T., Chans, L., Pineiro, L.,

Perez, D., Pereyra, I., 2003. National prevalence larval echino-

coccosis in sheep in slaughtering plants. Ovis aries as an indicator

in control programmes in Uruguay. Acta Trop. 85, 185–281.

Coltorti, E., 1986. Standarization and evaluation of an enzymeimmu-

noassay as a screening test for the seroepidemiology of human

hydatidosis. Am. J. Trop. Med. Hyg. 35, 1000–1005.

Coltorti, E., Fernandez, E., Marguet, E., Scozzina, J., Guarnera, E.,

1990. Deteccion de portadores asintomaticos de quistes hidatıdi-

cos. Aumento de la especificidad del ensayo inmunoenzimatico.

Rev. Inst. Med. Trop. Sao Paulo 32, 275–284.

Conder, G.A., Andersen, F.L., Schantz, P.M., 1980. Immunodiagnos-

tic tests for hydatidosis in sheep: an evaluation of double diffusion,

immunoeletrophoresis, indirect hemagglutination, and intrader-

mal tests. J. Parasitol. 66, 577–584.

Craig, P.S., 1997. Immnunodiagnosis of Echinococcus granulosus and a

comparison of techniques for diagnoses of canine echinococcosis.

In: Andersen, F., Ouhelli, H., Kachani, M. (Eds.), Compendium on

Echinococcosis in Africa and in Middle Eastern Countries. Brigham

Young University Print Services, Utah, pp. 86–118.

Craig, P.S., Rickard, M., 1981. Studies on the specific immnunodiag-

nosis of larval cestode infections of cattle and sheep using antigens

purified by affinity chromatography in any enzyme-linked immu-

noabsorbent assay (ELISA). Int. J. Parasitol. 11, 441–449.

Dueger, E.L., Gilman, R.H., 2001. Prevalence, intensity and fertility of

ovine cystic echinococcosis in the central peruvian andes. Trans.

R. Soc. Trop. Med. Hyg. 95, 379–383.

Dueger, E.L., Verastegui, M., Gilman, R.H., 2003. Evaluation of the

enzyme-linked immunoelectrotransfer blot (EITB) for ovine

hydatidosis relative to age and cyst characteristics in naturally

infected sheep. Vet. Parasitol. 114, 285–293.

Eckert, J., Deplazes, P., Craig, P.S., Gemmel, M., Gottstein, B., Heath,

D.D., Jenkins, J., Kamiya, M., Lightowlers, M.H., 2001. Echino-

coccosis in animals: clinical aspects, diagnosis and treatment. In:

Eckert, J., Gemmel, M., Meslin, F., Pawlowski, Z. (Eds.), Manual

on Echinococcosis in Humans and Animals: a public health

problem of global concern, WHO/OIE, France, pp. 195–203.

Grainer, M., 1995. Two graph operating characteristic (TG-ROC): a

Microsoft Excel template for the selection of cut-off values in

diagnostic tests. J. Immunol. Methods 185, 145–146.

Ibrahem, M.M., Craig, P.S., Mcvie, A., 1996. Echinococcus granu-

losus antigen B and seroreactivity in natural ovine hydatidosis.

Res. Vet. Sci. 61, 102–106.

Jubb, K., Kennedy, P., Palmer, N., 1993. Pathology of Domestic

Animals, fourth ed., vol. II. Editorial Academic Press Inc.

Kittelberger, R., Reichel, M.P., Jenner, J., Heath, D.D., Ligthtowlers,

M.H., Moro, P., Ibrahem, M.M., Craig, P.S., Keefe, J., 2002.

Evaluation of three enzyme-linked immunoabsorbent assays

(ELISA) for the detection of serum antibodies in sheep infected

with Echinococcus granulosus. Vet. Parasitol. 110, 57–76.

Laemmli, U., 1970. Cleavage of structural proteins during the assem-

bly of the head of the bacteriophage T4. Nature 227, 680–685.

Larrieu, E., Costa, M., Cantoni, G., Alvarez, R., Cavagion, L.,

Labanchi, J., Bigatti, R., Araya, D., Herrero, E., Mancini, S.,

Cabrera, P.A., 2001. Ovine Echinococcus granulosus transmission

dynamics in the province of Rio Negro, Argentina, 1980–1999.

Vet. Parasitol. 98, 263–272.

Lightowlers, M.H., Gottstein, B., 1995. Echinococcosis/hydatidosis:

antigens, immunological and molecular diagnosis. In: Thompson,

R., Lymbery, J. (Eds.), The biology of Echinococcus granulosus

and hydatid disease. George Allen and Unwin, London, pp. 355–

396.

Lightowlers, M.H., Rickard, M., Honey, R., Obendorf, D., Mitchell,

G., 1984. Serological diagnosis of Echinococcus granulosus

infection in sheep using cyst fluid antigens processed by antibody

affinity chromatography. A. Vet. J. 61, 101–108.

Lloyd, S., Martin, S., Walters, T., Soulsby, E., 1991. Use of sentinel

lambs for early monitoring of teh South Powys hydatidosis control

scheme: prevalence of Echinococcus granulosus and some hel-

minths. Vet. Rec. 129, 73–76.

Ming, R., 1986. Application of enzyme linked immunabsorbent assay

to the diagnosis of echinococcus of sheep and cattle. A. Vet. Z. Sin.

17, 55–62.

Moro, P.L., Verastegui, M., Gilman, R.H., Falcon, N., Bernal, T.,

Gavidia, C., Malqui, V., Moro, M.H., Dueger, E.L., 1997. Enzime

linked immunoelectrotransferblot assay for diagnosis of hydati-

dosis in sheep. Vet. Rec. 140, 605–606.

Moro, P.L., Bonifacio, N., Gilman, R.H., Lopera, L., Silva, B.,

Takumoto, R., Verastegui, M., Cabrera, L., 1999. Field diagnosis

A. Gatti et al. / Veterinary Parasitology 143 (2007) 112–121 121

of Echinococcus granulosus infection among intermediate and

definitive hosts in an endemic focus of human cystic echinococ-

cosis. Trans. R. Soc. Trop. Med. Hyg. 91, 611–615.

Oriol, R., Williams, J., Perez Esandi, M., Oriol, R., 1971. Evaluation

of purified lipoprotein antigens of Echinococcus granulosus in the

immunodiagnosis of human infection. Am. J. Trop. Med. Hyg. 20,

569–574.

Thompson, R., Mcmanus, D., 2001. Aetiology: parasites and life

cycles. In: Eckert, J., Gemmel, M., Meslin, F., Pawlowski, Z.

(Eds.), Manual on Echinococcosis in Humans and Animals: a

Public Health Problem of Global Concern. WHO/OIE, France, pp.

1–19.

Torgerson, P.R., Heath, D.D., 2003. Transmission dynamics and

control options for Echinococcus granulosus. Parasitology 127,

143–158.

Vargas, D., Bonet, R., Campano, S., Chacon, T., Vidal, M., 2002.

Evaluacion epidemiologica de las tecnicas de Elisa y electroin-

muno transferencia en el diagnostico de la hidatidosis ovina en la

XI Region de Chile. Tesis Doctoral. Universidad Mayor, Chile.

Yong, W.K., Heath, D.D., Parmeter, S.N., 1978. Echinococcus gran-

ulosus, Taenia hydatigena and Taenia ovis evaluation of cyst fluids

as antigen for serodiagnosis of larval cestode in sheep. N. Z. Vet. J.

126, 231–234.

Yong, W.K., Heath, D.D., 1979. ‘Arc 5’ antibodies in sera of sheep

infected with Echinococcus granulosus, Taenia hydatigena and

Taenia ovis. Parasite Immnunol. 1, 27–38.

Yong, W.K., Heath, D.D., Van Knapen, F., 1984. Comparison of

cestode antigens in an enzyme-linked immunosorbent assay for

the diagnosis of Echinococcus granulosus, Taenia hydatigena and

Taenia ovis infections in sheep. Res. Vet. Sci. 36, 24–31.