Enzyme-Linked Immunoelectrotransfer Blot Test for Diagnosis of Human Hydatid Disease

JOURNAL OF CLINICAL MICROBIOLOGY, June 1992, P. 1557-1561 95-1137/92/61557-5$2./ Copyright ) 1992, American Society for Microbiology Vol. 3, No. 6 Enzyme-Linked Immunoelectrotransfer Blot Test for Diagnosis of Human Hydatid Disease MANUELA VERASTEGUI,l PEDRO MORO,' ANGELA GUEVARA,' TALIANA RODRIGUEZ,2 ELBA MIRANDA,' AND ROBERT H. GILMAN' 2,3* Laboratory of Parasitology, Department of Microbiology, Cayetano Heredia University, 1 and A. B. Prisma, Avenida Las Artes 36,2 Lima, Peru, and Department of International Health, The Johns Hopkins School of Public Health, 615 North Wolfe Street, Baltimore, Maryland 21253 Received 12 November 1991/Accepted 12 March 1992 Sera from 71 patients with surgically confirmed hydatid disease (which is caused by Echinococcus granulosus) were studied by an enzyme-linked immunoelectrotransfer blot (EITB) assay. Sera from patients either with other cestode infections or with another illness were used as controls. Results of the EITB test for hydatidosis were compared with those of the double-diffusion (DD5) test and an enzyme-linked immunosorbent assay (ELISA). In the EITB assay with bovine lyophilized hydatid fluid, three antigen bands of 8, 16, and 21 kda were diagnostically important. The sensitivity of the assay by using these antigen bands was 8%to for hepatic cysts, 56% for pulmonary cysts, and 56% for cysts located in multiple organs. In sera from controls, the specificity of the EITB assay was 1%. Cross-reactions to the 8-, 16-, and 21-kDa bands occurred, respectively, in 12, 4, and 4% of sera from patients with cysticercosis. No cross-reactions were noted in patients infected with Hymenolepis nana. The ELISA in which swine hydatid fluid was used as the antigen was as sensitive as the EITB test but was less specific (8%) and frequently cross-reacted with sera from patients with other cestode infections. The sensitivity of the DD5 test, which uses sheep hydatid fluid, was low (47%), but its specificity was as high as that of the EITB assay. However, in patients with cysticercosis, cross-reactions were observed in 23% of sera tested. Despite the higher sensitivity found with the EITB assay, 23% (n = 5) of the serum samples that were positive by the DD5 test were not detected by the EITB assay. The EITB assay offers greater sensitivity and specificity than do the ELISA and the DD5 test. The highest proportion of hydatid cases is detected when the EITB and DD5 tests are run simultaneously. Human hydatid disease caused by the cestode Echinococcus granulosus is a widely recognized health problem in sheep-raising areas of the Americas, principally in countries such as Uruguay, Argentina, Chile, and Peru (17). Serological tests are useful for diagnosing hydatid infections in people who live in areas of endemicity because of their low cost and ease of performance. Radiological techniques are often too costly or are not available in many areas where hydatidosis is highly endemic (4, 25). The double-diffusion (DD5) and the immunoelectrophoresis tests, both of which detect the arc 5 band, are the assays most commonly used because of their very low costs and relatively high specificities (12, 25). However, these tests are limited because they are not highly sensitive and crossreactions may occur when sera from patients with cysticercosis are tested (11, 13, 16, 24). Also, the present enzymelinked immunosorbent assays (ELISAs) are sensitive but not highly specific (8) and often cross-react with sera from patients with other cestode infections (8, 14, 2). A recent advance has been the introduction of enzymelinked immunoelectrotransfer blot (EITB) assays for the diagnosis of hydatid disease. EITB assays are reported to be more sensitive and specific than other serological tests (1, 11, 18, 19). In this report, we describe three antigenic bands that are useful for the serological diagnosis of hydatidosis. The results that were obtained by using these bands in the EITB assay were compared with those obtained by the DD5 test and ELISA. * Corresponding author. 1557 MATERUILS AND METHODS Patients with hydatid disease. A medical history was taken, a physical examination was performed, and sera were obtained from patients with tissue-confirmed hydatidosis. Information regarding the location of the cyst, the number of cysts (single or multiple), and cyst status (noncomplicated cysts [those from which crystalline fluid was aspirated] or complicated cysts [those that were infected or ruptured]) was also obtained. Patient sera. The following groups of patient serum samples were studied: (i) 71 serum samples from patients with tissue-confirmed hydatid disease collected from different hospitals in Lima, Peru, from 1988 to 1991; (ii) 3 serum samples from patients with tissue-confirmed cysticercosis (Taenia solium) collected from 1987 to 199; (iii) 32 serum samples from patients with Hymenolepis nana infection diagnosed by stool examination; and (iv) 45 serum samples from patients who served as negative controls (negative controls were divided in two groups: first, 32 serum samples from hospitalized patients in whom hydatidosis was excluded by abdominal ultrasonography and chest X ray and, second, 13 serum samples from patients with cystic images on radiology who were later surgically diagnosed as having congenital cysts, abscesses, or pulmonary cancer); (v) a pool of sera from 72 healthy North Americans (provided by the Centers for Disease Control, Atlanta Ga.) that were also included as negative controls for the ELISA and the EITB assay; and (vi) a positive control for the ELISA and the EITB assay consisting of a pool of serum prepared from 5 surgically confirmed patients with hydatidosis positive for arc 5 in the DD5 assay. Serum samples for all groups of patients were from different individuals.

1558 VERASTEGUI ET AL. Immunological assays. (i) EITB assay. A single lot of antigen for the EITB assay was prepared from crude bovine hydatid cyst fluid. Hydatid cysts were obtained from abattoirs in Lima. Cyst fluid was aspirated and placed in a beaker at 4 C to which.5 M of phenylmethylsulfonyl fluoride (Sigma Chemical Co., St. Louis, Mo.) was added at a proportion of 1/1. This fluid was then centrifuged at 3, x g for 1 min, and the supernatant was filtered through no. 3 filter paper (Whatman, Ltd., Maidstone, England). The filtered hydatid fluid was divided into aliquots and frozen at -7 C until it was lyophilized. The lyophilized antigen was stored at -2 C until use. Hydatid fluid from all three species tested (ovine, swine, and bovine) contained the three diagnostic bands. Bovine lyophilized hydatid fluid was found to give clearer bands, with lower background readings and better reproducibility in comparison with those for antigens from other hosts. The band intensities varied between lots of bovine hydatid fluid. There are antigenic differences in hydatid fluid obtained from different hosts (7, 9, 15). In the EITB assay and the ELISA, bovine and swine hydatid fluid, respectively, provided the best test results. The DD5 test was performed by following standard procedures (4). For the EITB assay procedure, the antigen that was prepared as described above was reconstituted with phosphate-buffered saline (PBS), which was prepared with.1 M dibasic sodium phosphate,.1 M monobasic sodium phosphate, and.15 M sodium chloride (ph 7.2); and its protein concentration was determined (2). The antigen was diluted with.1% sodium dodecyl sulfate-.25 (wt/vol) bromophenol blue-.25 M Tris-HCI (ph 8.), and the dilution was completed with 6% glycerol to give a final concentration of 1,ug of protein per ml; the solution was heated at 65 C for 1 to 2 min. The antigen that was prepared as described above was resolved by polyacrylamide gel electrophoresis as described elsewhere (22), with the only exception being the use of ultra-low-molecular-mass markers (Pharmacia Fine Chemicals, Piscataway, N.J.) to define the size of the 8-kDa band. When regular lowmolecular-mass standards (Bethesda Research Laboratories, Gaithersburg, Md.) were used, the 8-kDa band falsely appeared as a 12-kDa band. The steps used next were basically those described previously for the EITB assay (21, 22). Briefly, the separated proteins were electrotransferred to nitrocellulose sheets. Sheets were cut into 3-mm-wide strips and were immersed in a dilution (1/1) of human serum. Bound human antibodies were reacted to horseradish peroxidase-conjugated goat antibody to human immunoglobulin G at a dilution of 1/1, for 1 h (provided by the Centers for Disease Control). Finally, antibodies bound to the specific antigenic bands were visualized with 3,3'-diaminobenzidine (22). (ii) ELISA. Antigen for an ELISA was prepared from a single lot of crude swine hydatid fluid as described previously (23). Steps for the preparation of swine antigen were very similar to those described above for the EITB assay, but no phenylmethylsulfonyl fluoride was used and the filtered fluid was dialyzed against distilled water for 32 h at 4 C by using dialysis tubing membranes with a 3,5-molecular-weight cutoff (Spectra, Medical Industries, Los Angeles, Calif.) with five changes of distilled water, first with a 4-h interval between the first two changes and then with 8-h intervals between the last three changes. After dialysis, the protein concentration was determined and the fluid was lyophilized as a batch. A modification of a previously described ELISA (3) was J. CLIN. MICROBIOL. used to detect immunoglobulin G antibodies against hydatid fluid antigens in serum. Optimal amounts of reagents, including animal source and concentration of antigen used, were determined by checkerboard titration before the test was performed. Plates were covered throughout each step of the run. Polystyrene microtitration plates (Immulon I; Dynatech Laboratories, Chantilly, Va.) were incubated overnight with 1 p.1 of antigen (diluted in.6 M carbonate buffer [ph 9.6] to give a protein concentration of 1,ug/ml). Excess antigen was removed by washing the wells five times with PBS prepared for the ELISA, which was made up of.1 M dibasic sodium phosphate,.1 M monobasic potassium phosphate,.14 M sodium chloride, and.3 M potassium chloride (ph 7.4) containing.5% Tween 2; the plate was then blotted dry. One hundred microliters of 1% skim milk diluted in PBS containing.5% Tween 2 was used to block unbound polystyrene binding sites after incubation for 1 h at 37 C. Next, the wells were washed as described above and 1 RI of serum sample diluted 1/1 with PBS containing.5% Tween 2 was added to duplicate wells and left to incubate for 1 h at 37 C. Positive and negative pools of control sera were used in each plate run. After incubation, plates were washed as described above, and 1 p.1 of goat anti-human immunoglobulin G conjugated with peroxidase (Bio-Rad Laboratories, Rockville Centre, N.Y.) diluted 1/1, in PBS containing.5% Tween 2 was added to all wells and was incubated for 1 h at 37 C. Plates were washed; for each plate 1 p.1 of substrate solution containing 5 mg of o-phenylenediamine (Sigma), 5.9 ml of sodium citrate (.1 M), 4.9 ml of citric acid (.1 M), and 1 p.l of 3% hydrogen peroxide was added to each well. The reaction was allowed to proceed for 5 min at room temperature and was stopped with 25 p. of 4 M sulfuric acid. Immediately, plates were read at 49 nm with a Titertek Multiskan immunoassay reader (Flow Laboratories, McLean, Va.). The cutoff point for this study was determined by constructing sensitivity and specificity response tables and selecting the point at which the optimum specificity and sensitivity were achieved. An ELISA result was considered positive when absorbances were above a cutoff value of.241. The cutoff point for the ELISA was determined as described previously (3, 6). (iii) DD5 test. For DD5 test antigen preparation, sheep hydatid cyst fluid was aspirated and centrifuged at 3, x g for 1 min, and the supernatant was dialyzed and processed as described above for the ELISA antigen. A single lot of antigen was used for the study. To prepare control sera for the DD5 test, hyperimmune arc 5-positive sera were prepared from lyophilized hydatid fluid from swine. We used swine rather than sheep antigen for immunization in order to minimize reactions to nonparasitic host antigens. Lyophilized hydatid fluid from swine was reconstituted with PBS (ph 7.4) (prepared as described above for the ELISA but with no Tween 2 added) to give a concentration of 5 p.g of protein per ml. An equal volume of.5 ml of antigen was mixed with complete Freund adjuvant for the first antigen injection and then with incomplete Freund adjuvant for the next three antigen injections. Each antigen injection contained 25 p.g of protein, and the injection was given as a single intramuscular dose in the thigh at 2-week intervals. The side of the animal's body used for inoculation was changed each time. Sera were tested for the presence of the arc 5 band by the DD5 test prior to each immunization. Two weeks after the fourth immunization schedule, we noted a strong arc 5 band that was comparable to the arc 5 band of control sera provided by the Centro Panamericano de Zoonosis. Blood was then collected and

Voi- 3, 1992 21.) 1 6... 1 2 3 4 5-2 - 97.4-68 43-29 14.3 FIG. 1. EITB assay showing the three specific antigenic bands of 8, 16, and 21 kda. Bands above 21 kda either were not sensitive or were not specific and are not shown in detail. Lane 1. pool of sera from five surgically confirmed hydatid patients; lane 2, serum sample from a patient with cysticercosis; lane 3, serum sample from a patient with H. nana infection; lane 4, serum sample from a hospitalized patient with cholecystitis; lane 5, molecular mass standards (14 to 2 kda). centrifuged, and sera were divided into aliquots and kept at -2 C for later use. The DD5 test for hydatid infection detects the antigenantibody precipitation reaction denominated arc 5. The test was performed with simple agar as described before (4, 12). The crude lyophilized sheep antigen was reconstituted with buffered saline solution that was prepared by combining 9 ml of saline solution (.85%) with 1 ml of.1 M dibasic potassium phosphate. The solution was then taken to ph 7.4 by using a solution of.1 M monobasic potassium phosphate. The crude lyophilized sheep antigen was used at a concentration of 5 mg (dry weight) per ml. Other steps of the test were performed as described by Coltorti and Varela Diaz (4). RESULTS EITB assay. Sera from 71 patients with hydatid cysts were tested. Three antigenic bands located at 8, 16, and 21 kda were diagnostic for human hydatid disease caused by E. granulosus (Fig. 1). On basis of the presenec of one or more of the three bands, the sensitivity was 65% and the specificity was 1% (Table 1). Forty-one serum samples had all three bands (58%). When each antigenic band was considered separately, the 8-kDa band was the one that was found most frequently in patients with pulmonary cysts (Table 2). This band cross-reacted with sera from three of the patients with cysticercosis. The 16- and 21-kDa antigenic bands were more efficient at detecting antibodies in patients with hepatic cysts and cross-reacted less than did the 8-kDa band (Table 2). No cross-reactions were observed in patients with H. nana infections. We also noted the 38-kDa band described by Siracusano et al. (19). This band was present in only 23 patients with hydatid cysts (32%). The 38-kDa band cross-reacted with sera from 3 of 3 (1%) patients with cysticercosis but not TABLE 1. IMMUNOBLOT TEST FOR flydatidosis 1559 Sensitivity and specificity of the EITB assay for the diaunosis of hunlian hydatid disease Sensitivitv Cross-reactions Speciticity E1TB assay hand (h%)(hvp-ru n (kda) datid. Cvsticcrcosis Hymcnolcpiasis controls 11 = 71) PI = 3)" 32) = 43) 8 65 1() 1 16 59 3 1(( 21 58 3 () 1) All three hbands 58 1 ()1 Any or all hitnds 65 1 1 " Onc patient with cysticercosis glavc repeated indtctcrmina.ltc r-csuilts aind was cxclutdcdi from the stuidy. with sera from 32 paticnts with hymenolepiasis. The specificity was 1%r'. In 1 runs, a positive control serum specimen in which the three diagnostic bands were present was run by two different observers. All bands were detccted by both observers. No difference was noted in the runs performed by the same observer, but the band intensities varied between runs. Comparative sensitivity and specificity of the EITB assay, ELISA, and the DD5 test. Sera from 63 patients with hydatid cysts were tested in the three assays; 65%,. of the serum specimens had an 8-kDa band, 59%s had a 16-kDa band, and 55%c had a 21-kDa band, while only 47% had an -arc 5 that could bc detcctcd by the DD5 assay (Tablc 2). The DD5 assay detected 23% (5 of 22) of the cases not dctected by the EITB assay, whilc the EITB assay detccted 49% (16 of 33) of the cases not dctcctcd by the DD5 assay. The rcsults obtaincd by the two tcsts wcre significantly diffcrcnt (P =.2) as determined by the McNcmar tcst. The DD5 assay was 1% specific but cross-recacted with scra from 23% of the patients with cysticercosis. The ELISA, although ncarly as sensitivc (63% ) as the EITB assay, was less specific (8%). Furthermorc, cross-reactions commonly occurred in sera from patients infccted with cysticcrcosis or H. niania (Tablc 2). Effect of cyst status. Information about cyst status was obtained from 58 paticnts with surgically confirmcd hydatid disease. Scra from paticnts with complicated cysts tcnded to rcact more than did sera obtained from paticnts with noncomplicated cysts. In the EITB test, diffcrences in the sensitivity of the tcst between sera from paticnts with complicated cysts and those from patients with noncomplicated cysts for the 8-kDa band were 7 and 52%S., rcspectively, and for the 16- and 21-kDa bands the diffcrences were 62 and 43%, respectively. The incrcasc in antibody detcction in the scra of patients with complicated cysts compared with that in the sera of patients with noncomplicated cysts was also observed in the DD5 test (59 and 25%, respectivcly). DISCUSSION Wc identified, using the EITB assay. three diagnostic antigcns (8-, 16-, and 21-kDa bands), two of which were highly specific for E. grianulosst.v. The scnsitivity of the EITB assay for the two bands was higher than that of the DD5 tcst, regardlcss of the cyst sitc. Many cross-reactions werc obscrvcd with the ELISA, which makcs this tcst unsuitablc as a dcfinitivc diagnostic test. The highcr sensitivity and spccificity of the EITB tcst make it morc discriminating in arcas in which hydatidosis and cysticcrcosis are cndemic. The specific antigcns wc identificd sccmcd to correspond

156 VERASTEGUI ET AL. J. CLIN. MICROBIOL. TABLE 2. Comparative sensitivities and specificities of the EITB assay, ELISA, and DD5 test for the diagnosis of human hydatid disease No. (%) seropositive by: Parameter No. of EITB assay patients 8-kDa 16-kDa 21-kDa ELISA DD5 test band band band Sensitivity Pulmonary 32 18 (56) 15 (47) 14 (44) 18 (56) 14 (44) Hepatic 2 18 (8) 18 (8) 18 (8) 18 (8) 14 (7) Multiple 9" 6 (56) 5 (56) 4 (44) 5 (56) 4 (44) Ovarian 2 1 (5) Overall 63" 41 (65) 37 (59) 35 (56) 4 (64) 3 (48) Cross-reactions Cysticercosis Hymenolepiasis 26" 32 3 (12) 1 (4) 1 (4) 8 (31) 4 (13) 6 (23) Specificity 45 45 (1) 45 (1) 45 (1) 6 (8) 45 (1) " Includes three patients with hepatic and pulmonary cysts; one patient with liver, peritoneal, and lung cysts; four patients with hepatic and peritoneal cysts; and one patient with hepatic, pulmonary, and subcutaneous cysts. ' Eight serum specimens from patients with hydatidosis and four serum specimens from patients with cysticercosis were excluded because of an insufficient volume to run the DD5 assay. to some antigens described by other groups. The 8-kDa antigen that we identified is the same as that described by Maddison et al. (11). If standard low-molecular-mass standards, which are inaccurate below 14 kda, are used, the 8-kDa antigen incorrectly appears to be 12 kda in size. An accurate description of a band of this size can be made only when ultra-low-molecular-mass weight markers are used. The 12-kDa identified by Shepherd and McManus (18) would appear to be the 8-kDa band, since they used prestained high-molecular-mass markers to measure the molecular mass. The 8-kDa antigen is not entirely specific for echinococcosis (13). In this study, sera from patients infected with E. granulosus reacted not only with the 8-kDa antigen but also with two other low-molecular-mass antigens (16 and 21 kda). Both of these antigens appear to be more specific than the 8-kDa band. The 16-kDa band has been described by Shepherd and McManus (18) to be highly specific. The 2- and 48-kDa bands were purified by Al-Yaman and Knobloch (1); both bands were specific and achieved sensitivity levels of 1 and 96%, respectively. No details concerning the location of the cysts in the sera from patients confirmed to have hydatid disease were given, so it is difficult to evaluate the true sensitivity of their assay. Using a cruder antigen, we did not find a specific antigen band which migrated at 48 kda. Antigen 5 is a large thermolabile 62- and 52-kDa lipoprotein complex which, under reducing conditions, has both 38- and 2-kDa antigens (11, 18). Shepherd and McManus (18) noted that the 2-kDa antigen band cross-reacts with other related and nonrelated parasites. In contrast, in our study a 21-kDa band was highly specific and cross-reacted with sera from only 4% of patients with cysticercosis. Further investigations are needed to define whether our 21-kDa band is the minor subunit of antigen B. In a study done in Italy, a 39-kDa band which was derived from antigen 5 was 9% sensitive and 1% specific. Cross-reactions did not occur (19). In our study, the 38-kDa antigen was only 32% sensitive but 1% specific. However, sera from 1% of patients with cysticercosis cross-reacted with this antigen. Our sensitivity results were somewhat lower for hepatic cysts and higher for pulmonary cysts than those reported by Maddison et al. (11). It is important to emphasize that sensitivity varies with the location of the cyst (5, 8, 12, 25). Hepatic cysts appear to produce a much stronger immunological response than do pulmonary cysts (11, 12). It is important to take into account this wide difference in sensitivity when comparing the results of serological tests for hydatidosis. We recommend that the cyst location and its status be described in detail in serological studies of hydatidosis. Although the specificity of the blot was 1% for sera from the particular group tested, the true specificity for sera from individuals living in a zone where cysticercosis is endemic would be less than 1%. For example, in a zone where cysticercosis is highly endemic, approximately 1% of adult inhabitants are seropositive for cysticercosis (6). If the rate of cross-reaction is 12%, then for sera from 1 individuals, sera from 1 individual would give a false-positive result because of the cross-reacting antibody. Thus, the true specificity is approximately 99%. The EITB assay is the most sensitive serological assay for the detection of hydatidosis. It efficiently detects hepatic cysts but is less sensitive for the detection of pulmonary and noncomplicated cysts. In seroepidemiological studies in areas endemic for both hydatidosis and cysticercosis, it would be more suitable than the DD5 test because of its higher sensitivity and fewer cross-reactions. The EITB assay has the drawbacks of needing sophisticated equipment and highly trained personnel. Although the DD5 test is less sensitive and has a higher percentage of cross-reactions than the EITB test, the DD5 test may be more suitable for developing countries, because it is simple, needing little infrastructure and training to be performed (4). The ELISA is not as sensitive or as specific as the EITB test. However, the ELISA is cheaper and easier to perform. When large numbers of sera need to be tested, it may be useful to use the ELISA as a screening assay. Optimally, an optical density value which would detect over 95% of positive cases would be used as the cutoff value. All tests positive by the ELISA would then need to be confirmed by the DD5 and EITB tests.

VOL. 3, 1992 In our study, the DD5 test also detected cases missed by the EITB test, suggesting that hydatid disease is best detected when both assays are run. The DD5 test will remain a useful test until more-sensitive, specific, and equally low cost assays are developed. In a study in Turkana, Kenya, only 5% of cysts seen on ultrasonography were detected by an ELISA (1). Chromatographically purified antigens may increase the sensitivity of the EITB test for the detection of hydatidosis. Further investigations that examine the use of purified proteins as antigens for serological tests for hydatidosis are needed. ACKNOWLEDGMENTS This study was supported in part by Concytec, Peru, and RG-ER and PMP funds. The assistance of N. Perez-Palma, R. Castillo, J. B. Phu, D. Sara, and J. Naranjo is appreciated. We also appreciate the cooperation of the Cysticercosis Working Group in Peru and medical personnel from Hospital Cayetano Heredia and Hospital Almenara. The authors are members of the Clinical Coordinating Board of the Hydatid Group in Peru. Other members of the group are E. Gonzales, J. Lozano, and F. Armas. IMMUNOBLOT TEST FOR HYDATIDOSIS 1561 REFERENCES 1. Al-Yaman, F. M., and J. Knobloch. 1989. Isolation and partial characterization of species-specific and cross-reactive antigens of Echinococcus granulosus cyst fluid. Mol. Biochem. Parasitol. 37:11-18. 2. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254. 3. Castillo, R., P. Grados, C. Carcamo, E. Miranda, T. Montenegro, A. Guevara, and R. Gilman. 1991. Effect of treatment on serum antibody to Hymenolepsis nana detected by enzymelinked immunosorbent assay. J. Clin. Microbiol. 29:2-7. 4. Coltorti, E. A., and V. M. Varela Diaz. 1978. Detection of antibodies against Echinococcus granulosus arc 5 antigens by double diffusion test. Trans. R. Soc. Trop. Med. Hyg. 72:226-229. 5. Craig, P. S., E. Zeyhle, and T. Romig. 1986. Hydatid disease: research and control in Turkana. II. The role of immunological techniques for the diagnosis of hydatid disease. Trans. R. Soc. Trop. Med. Hyg. 8:183-192. 6. Diaz, F., H. Garcia, R. H. Gilman, M. Verastegui, V. C. W. Tsang, J. A. Pilcher, M. Lescano, A. Gonzales, M. Castro, C. Carcamo, G. Madico, and E. Miranda. Am. J. Epidemiol., in press. 7. Guevara, A. 199. Bachelor's thesis. Universidad Peruana Cayetano Heredia, Lima, Peru. 8. Iacona, A., C. Pini, and G. Vicari. 198. Enzyme-linked immunosorbent assay (ELISA) in the serodiagnosis of hydatid disease. Am. J. Trop. Med. Hyg. 29:95-12. 9. Janssen, D., M. De Wit, and P. H. De Rycke. 199. Hydatidosis in Belgium: analysis of larval Echinococcus granulosus by SDS-PAGE and Western blotting. Ann. Soc. Belge Med. Trop. 7:121-129. 1. MacPherson, C. N. L., E. Zeyhle, T. Romig, P. H. Rees, and J. B.. Were. 1987. Portable ultrasound scanner versus serology in screening for hydatid cysts in a nomadic population. Lancet ii:259-291. 11. Maddison, S. E., S. B. Slemenda, P. M. Schantz, J. A. Fried, M. Wilson, and V. C. W. Tsang. 1989. A specific diagnostic antigen of Echinococcus granulosus with an apparent molecular weight of 8 kda. Am. J. Trop. Med. Hyg. 4:377-383. 12. Mercado, R., A. Atias, B. Astorga, and M. Lorca. 1988. Reaccion de doble difusion en agar con deteccion del arco So en el diagnostico de la hidatidosis. Bol. Of. Sanit. Panam. 15:159-163. 13. Moro, P. L., R. H. Gilman, M. Wilson, P. M. Schantz, M. Verastegui, H. Garcia, E. Miranda, and the Cysticercosis Working Group in Peru. Trans. R. Soc. Trop. Med. Hyg., in press. 14. Pappas, M. G., P. M. Schantz, L. T. Cannon, and S. P. Wahlquist. 1986. Dot-ELISA for the rapid serodiagnosis of human hydatid disease. Diagn. Immunol. 4:271-276. 15. Rickard, M. D., and M. W. Lightowlers. 1986. Immunodiagnosis of hydatid disease, p. 217-249. In R. C. A. Thompson (ed.), The biology of Echinococcus and hydatid disease. George Allen & Unwin Ltd., London. 16. Schantz, P. M., D. Shanks, and M. Wilson. 198. Serologic cross-reactions with sera from patients with echinococcosis and cysticercosis. Am. J. Trop. Med. Hyg. 29:69-612. 17. Serra, I., and H. Reyes. 1989. Hidatidosis humana en cuatro paises de Sudamerica. Bol. Of. Sanit. Panam. 16:527-53. 18. Shepherd, J. C., and D. P. McManus. 1987. Specific and cross-reactive antigens of Echinococcus granulosus hydatid cyst fluid. Mol. Biochem. Parasitol. 25:143-154. 19. Siracusano, A., S. loppolo, S. Notargiacomo, E. Ortona, R. Rigano, A. Teggi, F. De Rosa, and G. Vicari. 1991. Detection of antibodies against Echinococcus granulosus major antigens and their subunits by immunoblotting. Trans. R. Soc. Trop. Med. Hyg. 85:239-243. 2. Sjolander, A., J. A. Guisantes, J. M. Torres-Rodriguez, and H. Schroder. 1989. The diagnosis of human hydatidosis by measurement of specific Ig E antibody by enzyme immunoassay. Scand. J. Infect. Dis. 21:213-218. 21. Tsang, V. C. W., J. A. Brand, and A. E. Boyer. 1989. An enzyme-linked immunoelectrotransfer blot assay and glycoprotein antigens for diagnosing human cysticercosis (Taenia solium). J. Infect. Dis. 159:5-59. 22. Tsang, V. C. W., K. Hancock, M. Wilson, D. F. Parmer, S. D. Whaley, J. S. McDougal, and S. Kennedy. 1986. Enzyme-linked immunoelectrotransfer blot technique (Western blot) for human T-lymphotropic virus type III/lymphadenopathy-associated virus (HTLV-III/LAV) antibodies. Immunology series no. 15. Procedural guide. Centers for Disease Control, Atlanta. 23. Varela Diaz, V. M., and E. A. Coltorti. 1976. Techniques for the immunodiagnosis of human hydatid disease. Series of Scientific and Technical Monographs no. 7. Pan American Zoonoses Center, Ramos Mejia, Buenos Aires, Argentina. 24. Varela Diaz, V. M., E. A. Coltorti, and A. D'Alessandro. 1978. Immunoelectrophoresis tests showing Echinococcus granulosus arc 5 in human cases of Echinococcus vogeli and cysticercosismultiple myeloma. Am. J. Trop. Med. Hyg. 27:554-557. 25. Varela Diaz, V. M., E. A. Guarnera, and E. A. Coltorti. 1986. Ventajas y limitaciones de los metodos inmunologicos y de deteccion por imagenes para el diagnostico de la hidatidosis. Bol. Of. Sanit. Panam. 1:369-386.