Evaluation of five immunodiagnostic techniques in echinococcosis patients

Bulletin of the World Health Organization, 59 (5): 767-772 (1981) Evaluation of five immunodiagnostic techniques in echinococcosis patients A. K. CHEMTAI,1 T. R. BOWRY,2 & Z. AHMAD3 Double diffusion (DD), indirect haemagglutination (IHA), immunoelectrophoresis (IEP), latex agglutination (LA), andcomplementfixation (CF) tests were evaluatedforsensi- tivity and specificity in the diagnosis of 141 surgically-proven Turkana echinococcosis patients and 10 controls. The overall sensitivities for the tests were: IHA, 86.77%; LA, 53.3 %; CF, 63.3%; DD, 55. 0%; IEP, 55. 0%. LA and CF testsproduced a high number of falsepositive results; IHA gave afalse positive result in 10%o ofcases; nofalsepositives were obtained with IEP andDD. A combination ofthe latter three tests would therefore offer the best chance ofdetecting specific anti-Echinococcus antibodies, with an average sensitivity of 62. 7%. The possible reasons for the relatively high incidence offalse negative values are discussed. This paper presents the results of tests on the sensi- tivity and specificity of diagnostic techniques for human echinococcosis in 90 surgically-confirmed patients from Turkana district, Kenya, which has the highest incidence of the disease in the world (1, 2). Additional studies were carried out on a further 51 patients to evaluate the combined use of three tests. Results were obtained in male and female echino- coccosis patients of different ages, and grouped according to the stage of the infection and the condition of the cysts. MATERIALS AND METHODS Reference antigen and antisera A crude lyophilized hydatid cyst fluid antigen (20 mg) and antisera (1 ml) a were used as reference standards. Processing ofhuman hydatid cystfluid antigen (HHCF) HHCF was collected from hepatic, mesenteric, and abdominal cysts in echinococcosis patients under- going surgery, and processed as described elsewhere (3). Protein content was estimated using a modifi- I Postgraduate student, Institute for Molecular Biology, Univer- sity of Brussels, 1640 St-Genesius-Rode, Belgium. 2 Head, Immunology Section, Department of Human Pathology, University of Nairobi, Nairobi, Kenya. 3Biochemist, University of Nairobi, Nairobi, Kenya. a From the Pan American Zoonoses Center, Buenos Aires, Argentina. cation of the Lowry method (4). Dialysis was carried out against acetate buffer, and the fluid was lyophil- ized and used at a concentration of 100 g/litre in barbital buffer solution. Production ofhyperimmune rabbit serum Adult rabbits (3-4 kg) were inoculated with 0.5 mg of crude HHCF antigen in phosphate-buffered saline (PBS), emulsified with an equal volume of Freund's complete adjuvant. After 6 weeks, they were given the same amount of antigen in Freund's incomplete adjuvant. Blood samples were taken 10 days later and monitored for precipitating antibodies to the HHCF antigens by double diffusion (DD) and immuno- electrophoresis (IEP) tests. Both the antigen and the antisera were standardized using the reference standards. Serum samples Serum samples were collected from surgically- proven echinococcis patients in Lodwar and Kakuma hospitals in Turkana district, north-west Kenya. These samples were preserved in 0.015 mol/litre sodium azide and stored at -20 °C until tested. All the patients examined were screened for other parasitic infections. Negative control sera were also collected from the same district. Immunodiagnosis A panel of five immunodiagnostic techniques was employed to detect anti-Echinococcus antibodies in the patients' sera, as described below. 4113 767- 768 A. K. CHEMTAI ET AL. Some of the serum samples were examined indepen- dently in three other laboratories. This collaborative programme was carried out using a blind test pro- cedure and the results were communicated by the coordinator after receiving all the data. Immunoelectrophoresis. Conventional IEP was carried out as described elsewhere (5), using 12.0 g of agarose per litre of sodium barbital buffer (pH 8.6). Aliquots of 15 MA ofHHCF antigen were applied to the wells in the gel, and electrophoresis was carried out for 90 min at a constant current. A serum sample was then added to the troughs. After the precipitation bands had formed, the slides were washed in 0.15 mol/litre NaCl and the plates dried with filter papers. Finally, the slides were stained with a 0.5 g/litre solution of Coomassie blue in 1.75 mol/litre glacial acetic acid and 432 ml/litre ethanol, and then destained using 0.01 mol/litre trisodium citrate in the same concentration of ethanol and acetic acid. Double diffusion (DD). Agarose gel (12 g/litre) was prepared using PBS (pH 7.2), and antigen and anti- sera were applied to the wells, as described elsewhere (6). The plates were then incubated in a moist chamber and processed as described for the IEP slides. Indirect haemagglutination test (l-A). The IHA test was based on the tanned-cell technique (7) with a few modifications. Sheep red blood cells were coated with 4 g of HHCF antigen per litre, at pH 6.4. All the test sera were diluted 1: 256 (the diagnostic titre) and the agglutination reactions recorded. Complement fixation (CF) test. Serum samples from guinea pigs were used as the source of comple- ment. The serum was kept at -70 °C until used. The CF test system was then set up as described elsewhere (3, 8). By a checkerboard titration procedure, the optimal complement activity was found to be 1: 1000. After further standardization of the system the optimal antigen and antibody proportions were found to be 4.88 mg/litre (HHCF) and 1: 32. Latex agglutination (LA) test. A commercial preparation of Bacto-latex particles, 0.18 Mm diameter,b was diluted 1:4 with glycine buffer, pH 8.2. The latex particles were sensitized using 3 g of HHCF antigen per litre. Agglutination tests with positive control samples showed that the diagnostic titre was 1: 5. Agglutination reactions for the patients' serum samples were recorded using dilutions of 1: 5, 1:10, and undiluted sera. b From Difco Laboratories, Michigan, USA. RESULTS Standardization The strengths of the prepared HHCF antigen and the antisera raised in the rabbits were comparable with those of the reference standards. These were therefore used to check the antigen preparations used in the immunodiagnostic tests. The protein concentration in each patient's cyst fluid was between 0.23 g/litre and 1.02 g/litre. The concentration of proteins in the HHCF antigen used was 0.5 g/litre. Sensitivity and specifcity Table I shows the results of the five immunodiag- nostic tests on samples from 90 surgically-confirmed echinococcosis patients and 10 negative controls. The highest sensitivity obtained was 86.7%, with IHA, and the lowest was 53.3% (LA). In 32 patients all the tests were positive, while in 11 they were all negative. IHA was positive in 17 patients for whom all other tests were negative. There was no clear correlation for the remaining 33.3% of the cases. Although IHA had the highest sensitivity, the false positive results obtained indicated that other tests are also needed to confirm the diagnosis. LA and CFT were less discriminating in monitoring for human echinococcosis, producing a fairly high number of Table 1. Comparison of results of immunodiagnostic tests in 90 echinococcosis patients from Turkana district, Kenya Immunodiagnostic test No. of Percentage patients of patients IHA LA CFT DD IEP 32 35.6 + + + + + 11 12.2 5 5.6 + + + _ _ 3 3.3 + - + + + 2 2.2 + + - - - 5 5.6 + _ + - _ 6 6.7 + + + - + 17 18.9 + - - - - 5 5.6 + - + + - 3 3.3 + + - + + 1 1.0 - - + + - No. found positive 78/90 48/90 57/90 44/80 44/80 % 86.7 53.3 63.3 55.0 55.0 Negative control samples No. found positive 1/10 2/10 1/10 0 0 % 10 20 10 0 0 IMMUNODIAGNOSTIC TECHNIQUES IN ECHINOCOCCOSIS Table 2. Results of IHA, DD, and IEP tests on 51 echinococcosis patients No. of IHA DD IEP MeanClinical status No.atiet detectionCistpatients No. positive % No. positive % No. positive % rate 1%) Primary echinococcosis 40 30 75.0 22 55.0 22 55.0 61.7 Recurrent echinococcosis 11 9 81.8 7 63.6 7 63.6 69.7 Fertile cysts 19 16 84.2 11 57.9 11 57.9 66.7 Sterile cysts 32 17 53.1 13 40.6 13 40.6 44.8 Male patients 18 15 83.3 11 61.1 11 61.1 68.5 Female patients 33 26 78.7 20 60.6 20 60.6 66.6 Age group: 1-18 years 17 13 76.5 10 58.8 10 58.8 64.7 19-45 years 31 23 74.2 18 58.1 17 54.8 62.4 > 45 years 3 2 66.7 1 33.3 1 33.3 44.4 false positive results. The results of the studies on the additional 51 patients are shown in Table 2. The overall sensitivities in primary and recurrent Echinococcus infection were 61.7% and 69.7%, respectively. The breakdown of the results showed that the percentage of positive reactions varied according to she age of the patient, with 64.7% in the 1-18-year age group, 62.4% in the 19-45-year age group, and 44.4% in the group over 45 years old. Different values were also obtained for males and females (68.5% and 66.6%, respectively). Most of the patients examined had abdominal cysts, and 63.4% (26/41) gave positive results in the immunodiagnostic tests. The condition of the cyst also appeared to influence the detection rate of anti- Echinococcus antibodies in the serum. Of the patients with sterile cysts, 44.8% were positive, while of those with fertile cysts, 66.7% gave positive reactions. Table 3. Comparison of results of tests carried out in Switzerland and Kenya, on serum samples from patients and controls from Turkana district, Kenya Patient Switzerland KenyaPateno. Surgery Clinical statusno. 1IF" IHA IEP DD 1 - - - - Negative control 2 - - - - Negative control 3 ± - - - + Primary echinococcosis 4 + ± + + + Primary echinococcosis 5 ± ± + + + Primary echinococcosis 6 + + + + + Primary echinococcosis 7 + - - - Negative control 8 - - - - + Primary echinococcosis 9 - - - - Negative control 10 + + + + + Primary echinococcosis 11 - - - - Negative control 12 - - - - + Primary echinococcosis 13 ± - - - + Primary echinococcosis 14 - - - - + Primary echinococcosis 15 - - - - Negative control 16 + + + + + Primary echinococcosis 17 - - - - + Primary echinococcosis 8 IIF = Indirect immunofluorescent test 769 A. K. CHEMTAI ET AL. Table 4. Comparison of results of tests carried out in Argentina, France, and Kenya on serum samples from patients and controls from Turkana district, Kenya Patient France Argentina Kenya no. Surgery Clinical status IEP IHA DD IEP DD 15 - - Control 57 - - - - - Control 63 - - - - - Control 92 - + - - - Control 77 - - - - - Uterine fibroids 33 + + + + + + Primary echinococcosis 42 + + + + + + Primary echinococcosis 72 + + + + + + Primary echinococcosis 95 + + + + + + Primary echinococcosis 29 - - - - - + Primary echinococcosis 56 - - - - - + Primary echinococcosis 48 - + - - - + Primary echinococcosis 53 - + - - - + Primary echinococcosis 14 + + + + + + Recurrent 81 + + + + + + Recurrent 89 ± + - - - + Recurrent 17 + + + + + + Recurrent Collaborative studies Table 3 shows the results of a collaborative study on 17 Turkana patients. The percentage sensitivities obtained in the two laboratories were: IIF, 45.4%; IHA, 27.3%; IEP, 45.4%; and DD, 45.4%. Only the indirect immunofluorescence test (IIF) was positive in one negative control (patient no. 7). Table 4 shows data from three laboratories where sera from 17 Turkana patients were tested. No false positive reactions were obtained with the IEP and DD tests. IHA gave one false positive result (patient no. 92) at a titre of 1/80. The overall sensitivity of both IEP and DD was 58.3% (7/12); IHA was very sensi- tive, since in patients no. 48, 53, and 89 it gave highly significant titres of 1/640 when the IEP and DD tests both gave negative results. Only 2 patients (no. 29 and 56) had a low titre of 1/80, which was below the diagnostic titre, giving an overall percentage sensi- tivity for IHA of 83.3% (10/12). Thus there was complete agreement in the results of the four world laboratories on echinococcosis patients' examination. DISCUSSION Results have been presented on the performance of several immunodiagnostic tests for echinococcosis in patients from Turkana district. There is very little Echinococcus infection in other areas of Kenya. The average sensitivity of the five tests was 62.7%o, so that the diagnostic procedures employed would fail to detect anti-Echinococcus antibodies in 37.3%o of surgically-proven cases. The IHA test had an overall sensitivity of 86.7%o, but the average sensitivity of the other tests was rather low. Results obtained by the independent collaborating laboratories confirmed these low values, whereas results from Caucasian communities have produced false negative values of 10-20%o (9-11). The immunological aspects of the host-parasite relationship in patients from the Turkana district appear to differ from those seen in other areas. Work done on isoenzyme characterization and in vitro culture (C. Macpherson, personal communication, 1979) suggests the existence of a different strain of E.granulosus in Kenya. Infection of the Turkana people with this strain may change their immune responses and affect the diagnostic results. The high number of false negative results may also be a result of several other factors. The presence of inhibitors in the serum samples and the role of immune complexes ought to be considered. Although Richard-Lenoble et al. (11) found only 2 of 13 echinococcosis patients with positive sera to have immune complexes, our preliminary data indicate a much higher proportion with circulating immune complexes. Unlike the cysts in Caucasian populations, those found in Turkana patients were often very large, containing up to 4 litres of cyst fluid. This could mean 770 IMMUNODIAGNOSTIC TECHNIQUES IN ECHINOCOCCOSIS 771 that massive release of Echinococcus antigens occurs and immune complexes are formed readily in these patients. The effect of these circulating immune complexes on diagnostic techniques remains to be investigated. It is therefore important to interpret the immuno- diagnostic results in Turkana patients carefully, taking into consideration the immune status of the patient, possible strain differences in E. granulosus, and the presence of immune complexes and other inhibitors in the serum samples. No single test provides 1000o sensitivity and specificity for echino- coccosis and a combination of IHA, DD, and IEP appears to give the most reliable results. ACKNOWLEDGEMENTS We are very grateful to Professor A. Kungu for his assistance and encouragement and to Mr W. Emonyi for his valuable technical help. The project was supported by the African Medical Research Foundation in collaboration with the Ministry of Health, Nairobi, Kenya. We appreciate the help of Dr T. K. Siongok and Dr M. French, and wish to record special thanks to the collaborators in the study, Professor G. S. Nelson (Liverpool School of Tropical Medicine, Liverpool, England), Professor A. Capron (Institut Pasteur, Lille, France), Professor J. Eckert (Winterthurerstrasse 266, Zurich, Switzerland), Dr V. M. Varela-Diaz (Pan American Zoonoses Center, Buenos Aires, Argentina) and Dr I. Mann (FAO/UNEP/WHO, Nairobi, Kenya). The authors also acknowledge the secretarial assistance of Mrs D. W. Chege. RtSUM EVALUATION DE CINQ TECHNIQUES IMMUNODIAGNOSTIQUES CHEZ DES MALADES ATTEINTS D'ECHINOCOCCOSE Dans les regions ou l'echinococcose humaine sevit de facon endemique, il est urgent de pouvoir disposer d'epreuves immunodiagnostiques tant pour proceder A des etudes epidemiologiques que pour assurer le suivi des malades traites chirurgicalement. L'article presente les re- sultats d'etudes sur la sensibilite et la specificite de plusieurs techniques utilistes pour 1'examen d'echantillons de serum pr6leves chez des cas chirurgicalement confirmes provenant du district de Turkana, au Kenya, o4 l'incidence de l'echino- coccose est plus 6levee que partout ailleurs dans le monde. Dans l'ensemble, la sensibilite des tpreuves a ete la suivante: hemagglutination indirecte 86,7%; agglutination de particules de latex 53,3%; fixation du complement 63,3%; double diffusion 55%; et immunoelectrophorese 55%. La sensibilite moyenne etait de 62,7%, ce qui veut dire que ces techniques diagnostiques n'ont pas d6cele les anti- corps anti-Echinococcus chez 37,3% des malades positifs. La double diffusion et l'immunoelectrophorese n'ont donne aucun resultat faussement positif tandis que l'hemaggluti- nation indirecte, la fixation du complement et l'aggluti- nation du latex en ont donne respectivement 10%, 10% et 20%. C'est l'hemagglutination indirecte detectant les anti- corps agglutinants anti-Echinococcus, qui a pr6sent6 la sen- sibilite la plus elevee, tandis que les sensibilites moyennes des autres techniques, fondees sur la detection des anticorps pre- cipitants ou des anticorps activant le compliment, etaient plut6t faibles. Ces resultats ont ete confirmes par l'examen independant de certains des echantillons de serum dans des laboratoires en Suisse, en Argentine et en France. II convient, toutefois, d'interpreter avec circonspection ces resultats. II est important d'etudier I'etat immunitaire des malades, en particulier leurs reponses immunitaires cellulaires. II est aussi avere qu'il existe une souche locale differente de E. granulosus, dont il faudrait explorer la capacite de donner des antigenes ayant une valeur diagnos- tique et son effet sur les aspects immunologiques de la relation h8te-parasite. En outre, la presence d'immuncom- plexes et d'autres inhibiteurs seriques deceles chez certains des patients doit etre etudiee et leur importance pour l'immunodiagnostic demontree. A Turkana, par consequent, aucune epreuve n'a a elle seule eu une sensibilite et une specificite de 10001 et l'utili- sation combinee de l'hemagglutination indirecte, de la double diffusion et de l'immuno6lectrophor6se semble donner les r6sultats les plus tlables. Il faut, avant tout, conti- nuer a etudier la reponse immunitaire des malades atteints de E. granulosus. REFERENCES 1. O'LEARY, P. EastAfrican medicaljournal, 53: 540-544 (1976). 2. African Medical and Research Foundation. Third annual progress report, July 1978-June 1979. 3. VARELA-DIAZ, V. M. & COLTORTI, E. A. Techniques for the immunodiagnosis ofhuman hydatidosis. Buenos Aires, Pan American Zoonoses Center, 1976 (Technical rmonograph C.P. z-7). 772 A. K. CHEMTAI ET AL. 4. LOWRY, 0. H. ET AL. Journal of biological chemistry, 193: 265-275 (1951). 5. WEEKE B. In: Axelson, N. H. et al., ed., A manual of quantitative immunoelectrophoresis, Oslo, Univer- sitetsforlaget, 1973, pp. 15-35. 6. OUCHTERLONY, 0. Progress in allergy, 5: 1 (1958). 7. HERBER, W. J. In: Weir, D. M., ed., Handbook of experimental immunology, Vol. 1, Oxford, Blackwell, 1973. 8. LEVINE, L. In: Weir, D. M., ed., Handbook of experi- mental immunology, 2nd ed., Vol. 1, Oxford, Black- well, 1973. 9. M ATOSSIAN, R. M. Transactions ofthe Royal Society of Tropical Medicine and Hygiene, 71: 101-104 (1977). 10. MATOSSIAN, R. M. ET AL. Journal ofhelminthology, 53: 287-292 (1979). 11. RICHARD-LENOBLE, D. ET AL. Annals of tropical med- icine andparasitology, 72: 553-560 (1978).