Collaborative study on antigens for immunodiagnosis of schistosomiasis*

Bulletin of the World Health Organization, 60 (5): 729-753 (1982) Collaborative study on antigens for immunodiagnosis of schistosomiasis* K. E. Morr&H. DIXON1 Eight research laboratories in Europe and the United States ofAmerica were selected on the basis of having published data on Schistosoma mansoni and S. japonicum antigens to participate in a study of various antigen/test combinations for immunodiagnosis of schistosomiasis. The serum bank consisted of395 well documented serafromfour endemic areas in Brazil (2 areas), Kenya, and the Philippines. Altogther, 21 S. mansoni and four S. japonicum antigen and immunoassay combinations were evaluated. S. mansoni egg antigens yielded a higher combined sensitivity than adult worm antigens, irrespective of their purity, in active S. mansoni infections before and after specific treatment. Quantitative seroreactivity of characterized S. mansoni egg antigens showed good correlation with faecal egg counts in the 5-14 year age group. No correlation between morbidity related to S. mansoni and seroreactivity was observed in any test system. Three S. japonicum egg antigens showed high sensitivity and specificity in relation to thepresence or absence ofeggs in the stool. The quantitative seroreactivity ofthe character- ized S.japonicum egg antigens correlated directly with the intensity of S.japonicum infection in all age groups. The enzyme-linked immunosorbent assay (ELISA), using several differentprocedures, performed well with the antigens used in the study. The indium slide immunoassay (ISI), a simple qualitative visual test system using an S. mansoni egg antigen, demonstrated a high degree of sensitivity and specificity. The results did not indicate the superiority of any particular immunodiagnostic methodfor detecting antischistosome antibodies. This collaborative study is considered a first step towards developing and standardizing antigens for immunodiagnosis of schistosomiasis. Immunological tests for the diagnosis of schisto- somiasis have been in use for over 70 years. Initially, immunodiagnostic tests were designed to detect circulating antibodies and, therefore, could not dif- ferentiate between past and present infection, or Sponsored jointly by the Scientific Working Group on Schisto- somiasis of the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases and the Edna McConnell Clark Foundation. 1 On behalf of the participants listed in Annex 1, page 745. Requests for reprints should be addressed to Dr. K. E. Mott, Schisto- somiasis and other Helminthic Infections, World Health Organiz- ation, 1211 Geneva 27, Switzerland. infection with other parasites possessing cross- reactive antigens. These problems were discussed and clearly outlined in aWHO Memorandum in 1974 (1). While it has been generally accepted that improve- ment in the specificity of immunodiagnostic tests will depend on the availability of pure antigens, particular requirements for individual tests will depend on their intended use. For example, epidemiological studies have different test requirements than clinical studies or research, while field investigations require a simple, cheap, and easily applicable test. Since 1974, gel filtration, affinity and cation ex- 4224 729- K. E. MOTT & H. DIXON change chromatography, and other techniques have offered new opportunities for purification of antigens; more recently, monoclonal antibody tech- niques present new possibilities for antigen prepar- ation. Application of these and other research techniques has produced several new Schistosoma mansoni egg antigens, such as a "major serological antigen" (MSA) (2-4), CEF-6 (5, 6), and fractionated soluble egg antigen (SEA) (7), whose initial appli- cations in immunodiagnostic tests have been very promising (1). Purification of adult worm antigens has been generally directed towards excretory- secretory products (8) and circulating antigens (9). Further progress and evaluation have been limited by the lack of sufficient amounts of reagents, particu- larly antigens, for development of the immuno- diagnostic techniques. Improvement of the enzyme-linked immuno- sorbent assay (ELISA) (10) has been possible through the development of better standardized conjugates and more sensitive substrates (11). For instance, the introduction of methods for preparing conjugates with uniform enzyme-immunoglobulin complexes and free of unlabelled antibodies has led to better quantification and increased specificity. The use of the enzyme 3-galactosidase as a label has increased sensitivity, particularly when it is used with a fluoro- genic substrate. The standardization of plastic plates is also an important improvement. Although the indirect immunofluorescent assay (IFA) remains technically unchanged, the interpret- ation of the fluorescence patterns in the adult worms may provide an improved assessment of the state of the infection (12). Of the new immunodiagnostic tests that have appeared, the defined antigen substrate sphere (DASS) (13) has shown initial promise, and the indium slide immunoassay (ISI) (14) has potential for application under field conditions. In view of these advances, it was decided that an evaluation of available Schistosoma antigens was needed in order to assess the priorities for future re- search. A niulticentre collaborative study was there- fore established to examine the various antigens and their potential application in immunodiagnostic tests. OBJECTIVES OF THE COLLABORATIVE STUDY The antigens for immunodiagnosis of schisto- somiasis were evaluated by applying the various antigen/test systems to a serum bank composed of selected sera from endemic S. mansoni and S. japonicum areas and appropriate uninfected con- trols. The study was designed to assess: - sensitivity, i.e., the number of serologically positive sera among all sera from persons with S. mansoni or S. japonicum eggs in their stool at the time the sample was obtained; - specificity, i.e., the number of serologically negative sera from persons without S. mansoni or S. japonicum eggs in their stool at the time the sample was obtained; - ability to discriminate between lightly and heavily infected persons; - ability to discriminate between present and past infection; - correlation between serological result and the absence or presence of clinical morbidity; - the cost-effectiveness. MATERIALS AND METHODS Study design The collaborative study was designed and organ- ized by the organizing committee (see Annex 1). The serum bank was contributed by four centres in Brazil (2 centres), Kenya, and the Philippines. The participating laboratories were selected on the basis of their having published reports of defined antigens that had potential for application to the immunodiagnosis of schistosomiasis. Some of these laboratories agreed to prepare crude antigens that had been used previously in the immunodiagnosis of schistosomiasis. The participating laboratories reviewed and com- mented on questionnaires and data entry forms pre- pared by the organizing group prior to the receipt of the serum bank. The sera were stored and distributed throughWHO headquarters, Geneva, to the participating labora- tories in September 1980. Each participating labora- tory forwarded the results of the serological tests to WHO headquarters, Geneva, where data processing and analysis were carried out. The analysed results from each laboratory were returned one month before the evaluation meeting so that each laboratory could review its own results and prepare a presentation for the meeting. Staff in Geneva compared the results of the various participating laboratories and presented their analysis at the evaluation meeting in June 1981. Serum bank The serum bank was designed so that the serological results could be compared with corresponding para- sitological data. Sera from both untreated and pre- viously treated persons with or without S. mansoni or S. japonicum eggs in the stool were supplied. 730 IMMUNODIAGNOSIS OF SCHISTOSOMIASIS 731 Each blood donor was examined clinically and age, sex, faecal egg count, and treatment status recorded. Any liver or spleen enlargement was measured in centimetres below the costal margin. Quantitative parasitological diagnosis was based on at least one stool examination (2 slides) by the Kato-Katz tech- nique (15). Treatment with praziquantel was given in Kenya and the Philippines and with oxamniquine in Brazil. In general, post-treatment sera were obtained 6 months after treatment. The sera from uninfected persons collected in Brazil included 20 from persons in the Amazon region (a non-endemic area), 6 from foreigners (with a negative skin test), and 10 from Brazilians from endemic areas who were parasitologically negative in 20-40 Kato- Katz stool examinations and who had no skin reac- tivity to intradermal S. mansoni adult antigen. Sera from 11 Europeans were also included in the bank. The sera were collected by the staff of the collabor- ating centres. Blood samples of 15-20 ml were col- lected by venepuncture using a plastic syringe or vacuum tube and the blood was stored in a sterilized test tube with a rubber stopper. Whole blood was kept at room temperature for 6-8 h and at 4 °C overnight for complete clot retraction. The following day, the blood was centrifuged in the original tube, the serum was carefully removed and divided among individual serum storage tubes. In some instances, 10 g/litre sodium azide was added as a preservative. The sera were then frozen at -20 °C and shipped to WHO headquarters in containers with dry ice. All sera arrived in Geneva in the frozen state, and were stored at -20 °C for one month. For distri- bution to the participating laboratories, the sera were thawed at room temperature, and aliquots of 0.1-0.4 ml were placed into 3-ml polycarbonate tubesa using an automatic pipette. The sera were kept at 4 °C for up to 3 hours after transfer and then re- frozen at -20 'C. For shipment, the frozen sera were packed directly in an excess of dry ice and all sera arrived via airfreight in the frozen state at the partici- pating laboratories. A total of 248 sera originated from laboratories in Brazil, 99 from the Philippines, and 66 from Kenya. Only 50 of the sera from Kenya were examined in all laboratories. The results for 2 sera from Brazil were excluded because of lack of demographic infor- mation. The maximum number of serological results for any single antigen/test system combination among the participating laboratories was 395. Antigen and test combinations evaluated The following combinations of antigens and test systems were evaluated in the collaborative study (see Annexes 2 and 3). a NUNC, DR-4000 Roskilde, Denmark. Antigen (with identification number) Test Whole S. mansoni eggs (01) COPT Crude S.mansoni soluble egg antigen (SEA) (05) ELISA IHA Fractionated S. mansoni SEA (07) ELISA S. mansoni CEF-6 egg antigen (15) ELISA ISI S. mansoni egg fraction (02) RIA S. mansoni adult worm sections Rossman's fixative (13) IFA Cryostat (14 and 17) IFA Saline-extracted S. mansoni adult antigen (10) ELISA IHA RAST-IgE MDD TCA-extracted S. mansoni adult antigen (12) ELISA S. mansoni excretory-secretory antigen (11) ELISA S. mansoni purified glycoprotein fraction (09) ELISA IHA RAST-IgE MDD S. japonicum egg glycoprotein antigen (04) ELISA Crude S. japonicum SEA (06) ELISA IHA Fractionated S. japonicum SEA (08) ELISA Statistical analysis All the data from the participating laboratories were processed and analysed on an IBM 370 computer using an SPSS package for simple correlation analysis. The data obtained on the sera from un- treated and previously treated persons were analysed separately and combined. Overall correlation coefficients between quanti- tative serological results and faecal egg counts were ranked according to the antigen employed. For the five antigens showing the best correlation, coef- ficients were calculated by age and sex groups, and for treated and untreated groups according to age and sex. To ensure that high egg counts were not over- contributing to the low correlation coefficients, this procedure was re-run, disregarding the results for all sera from persons with more than 1200 eggs per gram. In general, this did not increase the correlation coef- ficient for any of the antigen/test combinations. Because of the gmall numbers in each age group, the last two procedures were not followed for S. japonicum antigens. Multiple regression analysis of the data was carried out to determine the contribution of egg count, age, and sex to the quantitative serological results. The equation fitted to the data was: K. E. MOTT & H. DIXON y = a + bix1 + b2x2 + b3x3 where y is the quantitative laboratory result expressed as either the optical density, or the RIA reading, or the log2 (titre), or the percentage reactivity for COPT; xi is logio (egg count + 1); x2 is the age in years; X3 is 1 for a male subject and 2 for a female subject; b1, b2, and b3 are the partial regression coefficients. The multiple regression coefficient, "multiple R ", was also computed and all coefficients were tested in order to determine if they were significantly different from zero. The significance of the coefficients was in- fluenced by the variance of the quantitative sero- logical data; therefore it was not possible to compare directly the coefficients of different antigen/test com- binations but it was possible to compare the signifi- cance findings. RESULTS Evaluation ofcontrol sera The serum bank contained 31 sera from Europeans and Amazon Indians who had never resided in an endemic area. The specificity of the antigen/test com- binations in these sera ranged from 10007 to 67.7%o (Table 1). Both egg and adult antigens demonstrated the highest specificity. Specificity was lower among sera from treated persons even though they had had multiple negative stool examinations for at least 3 months after treatment. Evaluation of antigens S. mansoni egg and adult worm antigens. A total of 21 S. mansoni antigen/test combinations were exam- ined in this study. Sensitivity was evaluated by exam- ination of all sera from untreated individuals who had eggs in their stool, as well as from those who were not cured after specific treatment. The specificity was evaluated in all sera from untreated persons who had consistently negative stool examination and sera from previously treated persons with negative stool examin- ations (Tables 2 and 3). The egg antigens, both crude and defined, had a higher combined sensitivity and specificity than the adult antigens both by simple ranking and by the J index for diagnostic tests (34). The adult worm antigens showed particularly low specificity in sera from previously treated persons who had negative stool examinations, whereas their sensi- tivity in these sera was similar to that of the egg antigens. The results of the microdouble diffusion test were recorded as either positive or negative; all other test systems gave rise to some equivocal results. S. japonicum egg antigens. The sensitivity of all three S. japonicum egg antigens was high (95.2- 10007o ) in sera from untreated individuals (Table 4). In sera from both treated and untreated persons, the fractionated SEA in the ELISA gave the highest sensi- tivity and specificity. Evaluation of the test systems The sensitivity of each test system was assessed on sera from untreated persons with eggs in their stools; specificity was assessed on sera from persons without eggs in their stools. Enzyme-linked immunosorbent assay (ELISA). Four different ELISA procedures and seven different S. mansoni antigens (three egg and four adult) were compared. In the sera from the untreated persons, the specificity and the rate of false positive readings were similar for both types of antigen (Table 5). In all sera, including those from previously treated persons, the egg antigens were generally associated with higher sen- sitivity (83-91 0o) and specificity (58-75.807o). Immunoassays using radioisotopes (RIA and IgE- RAST). The sensitivity and specificity of these assays were low in sera from treated and untreated persons (Table 6). The sensitivity and specificity of both purified protein adult antigen and the saline-extracted antigen were lower in the IgE-RAST than in the ELISA (Table 5) or IHA (Table 8). The RIA with egg fraction antigen gave a higher rate of false-positive readings in treated persons without eggs in their stools (47.1 0/0) than in untreated persons (12.207o). Indirect immunofluorescent antibody (IFA). Two antigens were used in the IFA tests-adult S. mansoni fixed in Rossman's fixative and cryopreserved adult S. mansoni-as adult worm sections. A positive reac- tion in the former test was indicated by fluorescence of the gut lining. In the latter test system, reactivity to both the gut and somatic structure was assessed. In untreated individuals, the specificity of all IFA tests ranged between 76.807o and 86.6%7o, and the rate of false positive reactions was low (12.2-22.0%o) (Table 7). In the sera from treated persons, the rate of false positive reactions was high in all tests (41.2-73.5%7o). Although the sensitivity assessed by combined gut and somatic reactivity in the cryopreserved parasite was high (91. I/o), the specificity was low (26.5 5o). Indirect haemagglutination test (IHA). The crude soluble egg antigen (SEA) in the IHA showed high sensitivity and specificity in the untreated sera (Table 8). In sera from previously treated persons the same antigen showed a much higher rate of false positive results. 732 IMMUNODIAGNOSIS OF SCHISTOSOMIASIS 733 Table 1. Specificity of S. mansoni and S.japonicum antigens Specificity (%) Antigen Antigen code Test Group 1 a Group 2b Group 3' Whole S. mansoni eggs Fractionated SEA S. mansoni Fractionated SEA S. japonicum TCA-extracted S. mansoni adult worm CEF-6 S. mansoni egg Crude SEA S. mansoni Purified glycoprotein fraction S. mansoni Crude S. mansoni adult worm extract Excretory-secretory S. mansoni S. mansoni GASP Purified glycoprotein fraction S. mansoni CEF-6 S. mansoni egg S. mansoni egg fraction S. japonicum egg glycoprotein Crude SEA S. mansoni Crude SEA S. japonicum Crude saline extract S. mansoni adult worm S. mansoni gut S. mansoni somatic Crude SEA S.japonicum Purified glycoprotein fraction S. mansoni Purified glycoprotein fraction S. mansoni Crude S. mansoni adult worm extract S. mansoni gut and somatic Saline-extracted adult S. mansoni 01 07 08 12 15 05 COPT ELISA ELISA ELISA ISI ELISA 09 MDD 10 1 1 13 09 15 02 04 05 06 10 14 17 06 IHA ELISA IFA IHA ELISA RIA ELISA IHA ELISA ELISA IFA IFA IHA 09 IgE RAST 09 ELISA 10 14& 17 10 IgE RAST IFA MDD a Combined results from a maximum of 11 sera from Europeans and 20 sera from Amazon Indians. b Results from a maximum of 20 sera from Amazon Indians. c Results from a maximum of 17 post-treatment sera from residents of S. mansoni-endemic areas with at least 1 5 negative Kato stool examinations. Qualitative tests. Of the 3 qualitative tests (circum- COPT (60.4%7) and they showed similar specificity oval precipitin test (COPT), indium slide immuno- (63.6%7o and 59.6%o, respectively). assay (ISI), and microdouble diffusion test (MDD)), the COPT and the ISI with the CEF-6 egg antigen had Relationship between seroreactivity and faecal egg the highest sensitivity and specificity and the lowest cont rate of false-positive reactions in sera from untreated count persons (Table 9). In sera from previously treated S. mansoni. The egg antigens in the ELISA detected persons, the ISI was more sensitive (83.0%o) than the over 90%7o of individuals with more than 400 eggs per 100 100 100 100 100 96.7 96.7 96.7 96.7 96.7 93.3 93.3 90.3 90.3 90.3 90.3 87.1 87.1 87.1 83.9 83.9 77.4 74.2 74.2 67.7 100 100 100 100 100 90 95 94.7 100 94.7 94.7 94.7 100 90 90 95 80 80 90 95 75 65 60 65 75.0 64.7 75.0 85.7 47.1 64.7 58.8 31.2 25 70.6 80 17.6 29.4 41.2 52.9 23.5 35.3 64.7 35.3 11.8 K. E. MOTT & H. DIXON Table 2. Sensitivity and specificity of S. mansoni antigen/test systems in sera from untreated persons Sensitivity b Specificity' Antigen a Test (%) Antigen a Test (%) 05 IHA 93.3 01 COPT 97.6 14+ 17 IFA 92.9 15 ISI 94.9 15 ELISA 91.7 05 ELISA 93.8 11 ELISA 91.2 07 ELISA 93.8 10 MDD 89.3 09 MDD 91.5 13 IFA 89.0 15 ELISA 90.0 10 ELISA 88.8 05 IHA 87.7 09 ELISA 88.2 09 IHA 87.7 07 ELISA 85.6 10 IHA 87.7 15 ISI 83.9 17 IFA 86.6 05 ELISA 83.7 11 ELISA 86.4 01 COPT 80.8 10 ELISA 84.1 12 ELISA 79.7 14 IFA 84.1 17 IFA 78.0 09 IgE RAST 84.1 09 IHA 71.2 12 ELISA 84.0 10 IHA 71.2 09 ELISA 80.5 14 IFA 70.9 13 IFA 79.0 02 RIA 65.9 10 IgE RAST 78.0 10 IgE RAST 64.0 14+ 17 IFA 76.8 09 IgE RAST 63.5 10 MDD 74.4 09 MDD 58.4 02 RIA 73.1 a Antigen code as given in Table 1. b Tested on 1 77-182 sera from persons with S. mansoni eggs in their stools. c Tested on 80-81 sera from persons with negative stool examinations, including 41 untreated persons from the Philippines. gram of faeces (Table 10). Both CEF-6 and fraction- ated SEA detected nearly all individuals with more than 200 eggs per gram. Both egg and adult worm antigens (except TCA-extracted adult antigen) in the ELISA detected at least 95% of individuals with more than 800 eggs/gram. In other test systems, RIA, IHA, and MDD, the adult worm antigens had a low sensitivity (42.9-79.507o). The combined gut and somatic IFA re- activity and crude saline-extracted adult worm antigen in the MDD were highly sensitive in all egg count groups, but their specificity has already been shown to be very low (Table 2). The ELISA with CEF-6 egg antigen showed the highest overall positive correlation between the op- tical density (OD) reading and faecal egg count in all S. mansoni sera (R = 0.66). Correlation coefficients were calculated by age and sex groups for the first five antigens (Table 11). The CEF-6 egg antigen was usually associated with the highest correlation coefficient for any particular group. Correlation coefficients were also calculated by age and sex according to treatment status. The quantita- tive seroreactivity to the CEF-6 egg antigen showed a high positive correlation with egg count in males in the 5-14-year age group, with or without prior treatment (Fig. 1) S. japonicum. The S. japonicum egg antigens tested in the ELISA and IHA were extremely sensitive in all egg count classes. The OD of the ELISA readings for all three S. japonicum egg antigens showed a good cor- relation with the intensity of infection, particularly in males aged 5-14 years (Table 12). The results from assays using S. japonicum fractionated SEA antigen (08) and S.japonicum egg glycoprotein antigen (04) had a higher correlation with intensity of infection than those using the crude SEA antigen (06), in all age groups, except in 5-14-year old males. Multiple regression analysis Most of the quantitative serological data showed a significant correlation with faecal egg count, i.e., as 734 IMMUNODIAGNOSIS OF SCHISTOSOMIASIS Table 3. Sensitivity and specificity of S. mansoni antigen/test systems in sera from treated persons Sensitivity b Specificity' Antigena Test (%) Antigena Test (%) 10 MDD 94.5 15 ELISA 77.3 05 IHA 94.3 05 ELISA 68.2 11 ELISA 90.9 01 COPT 63.6 10 ELISA 89.1 09 MDD 60.0 15 ELISA 88.7 15 ISI 59.1 09 ELISA 87.3 09 IHA 56.7 13 IFA 84.9 14 IFA 56.5 14+ 17 IFA 84.9 07 ELISA 47.8 15 ISI 83.0 10 IgE RAST 40.0 07 ELISA 83.0 10 IHA 36.7 05 ELISA 79.2 12 ELISA 31.8 17 IFA 77.4 02 RIA 30.4 02 RIA 71.7 09 IgE RAST 30.0 12 ELISA 67.9 17 IFA 26.1 09 IHA 67.3 09 ELISA 23.3 10 IHA 65.5 13 IFA 22.7 01 COPT 60.4 10 ELISA 20.0 09 IgE RAST 56.4 05 IHA 18.2 14 IFA 52.8 1 1 ELISA 13.6 10 IgE RAST 52.7 14+17 IFA 13.0 09 MDD 45.5 10 MDD 6.7 a Antigen code as given in Table 1. b Tested on 53-55 sera from persons with S. mansoni eggs in their stools. ' Tested on 22-30 sera from persons with negative stool examinations. Table 4. Sensitivity and specificity of S.japonicum egg antigens in sera from untreated persons Sensitivitya False negativesa Specificityb False positivesb S. japonicum antigen Test (%) (%) (%) M%) Fractionated SEA ELISA 100 0 89.5 10.5 Crude SEA ELISA 100 0 84.2 15.8 Crude SEA IHA 95.2 4.8 86.8 13.2 Purified glycoprotein ELISA 95.2 4.8 79.5 17.9 a Tested on 21 sera from persons with S. japonicum eggs in their stools. b Tested on 38-39 sera from persons with negative stool examinations. the egg count increased the quantitative seroreactivity increased (Table 13). The serological results with the egg antigens (02, 05, 07, 15) showed a negative corre- lation with age. The level of specific IgG antibody detected by the IHA (antigen 05) and RIA (antigen 02) tests decreased significantly with age. On the other hand, the level of specific IgE antibody detected by the RIA with adult antigens (antigens 09 and 10) increased significantly with age. The quantitative seroreactivity of egg antigen 05 in the ELISA was lower in males than in females, which may have contributed to a positive partial regression coefficient for sex in this test system. Except for antigen 14 in the IFA, all the tests showed a significant correlation with at least one of the variables tested. 735 K. E. MOTT & H. DIXON Table 5. Sensitivity and specificity of S. mansoni antigens in ELISA, in sera from untreated persons Sensitivity' False negativesa Specificity b False positives b Antigen (%) (%) (%) (%) CEF-6 egg 91.7 6.6 90.0 8.8 Fractionated SEA 85.6 10.0 93.8 4.9 Crude SEA 83.7 2.9 93.8 6.2 Saline-extracted adult 88.8 5.6 84.1 8.5 TCA-extracted adult 79.7 8.2 84.0 11.1 Purified glycoprotein adult 88.2 9.0 80.5 6.1 Excretory-secretory adult 91.2 4.9 86.4 11.1 a Tested on 178-182 sera from persons with S. mansoni eggs in their stools. b Tested on 80-82 sera from persons with negative stool examinations. Table 6. Sensitivity and specificity of S. mansoni antigens in RIA and IgE-RAST in sera from untreated persons Sensitivitya False negativesa Specificity b False positives b Antigen Test (%) (%) (%) (%) S.mansoniegg fraction RIA 65.9 9.3 73.1 12.2 Purified glycoprotein adult IgE RAST 63.5 32.6 84.1 11.0 Saline-extracted adult IgE RAST 64.0 30.9 78.0 14.6 a Tested on 178-182 sera from persons with S. mansoni eggs in their stools. b Tested on 80-82 sera from persons with negative stool examinations. Table 7. Sensitivity and specificity of S. mansoni adult antigens in the IFA in sera from untreated persons Sensitivitya False negativesa Specificityb False positivesb Antigen (%) (%) (%) 1%) Fixed adult 89.0 6.6 79.0 13.6 Gut cryopreserved adult 70.9 29.1 84.1 15.9 Somatic cryopreserved adult 78.0 21.4 86.6 12.2 Combined gut and somatic 92.9 5.5 76.8 22.0 a Tested on 182 sera from persons with S. mansoni eggs in their stools. b Tested on 82 sera from persons with negative stool examinations. Table 8. Sensitivity and specificity of S. mansoni antigens in the IHA test in sera from untreated persons Sensitivitya False negatives' Specificityb False positivesb Antigen (%) (%) (%) (%) Crude SEA 93.3 6.7 87.7 12.3 Purified glycoprotein adult 71.2 22.0 87.7 7.4 Saline-extracted adult 71.2 22.0 87.7 7.4 a Tested on 177-180 sera from persons with S. mansoni eggs in their stools. b Tested on 81 sera from persons with negative stool examinations. 736 IMMUNODIAGNOSIS OF SCHISTOSOMIASIS 737 Table 9. Sensitivity and specificity of S. mansoni antigens in qualitative tests in sera from untreated persons Sensitivitya False negativesa Specificityb False positivesb Test Antigen M%) M%) (%) M%) COPT Whole egg 80.8 11.0 97.6 2.4 ISI CEF-6 egg 83.9 13.9 94.9 2.5 MDD Purified glycoprotein adult 58.4 41.6 91.5 8.5 MDD Saline-extracted adult 89.3 10.7 74.4 25.6 a Tested on 177-182 sera from persons with S. mansoni eggs in their stools. b Tested on 79-82 sera from persons with negative stool examinations. Table 10. Sensitivity of S. mansoni antigen/test systems, according to intensity of infection, in sera from untreated persons No. of eggs/gram of faecesa 1-100 101-200 201-400 401-800 > 800 Antigen/test system (54-79) (18-24) (19-24) (10-14) (21-40) Egg (05)-lHA 88.5 87.0 100 100 100 Egg (15)-ELISA 87.3 87.0 100 100 95.1 Egg (07)-ELISA 75.6 78.3 91.7 100 100 Egg (05)-ELISA 75.3 77.3 87.0 92.9 100 Egg (01)-COPT 73.4 95.7 79.2 85.7 85.7 Egg (15)-lSI 79.7 82.6 95.8 78.6 87.5 Egg (02)-RIA 65.8 60.9 70.8 64.3 66.7 Adult (14+ 17)-lFA 91.1 87.0 100 92.9 95.2 Adult (10)-MDD 91.0 73.9 87.5 100 92.3 Adult (11)-ELISA 91.1 87.0 87.5 85.7 97.6 Adult (10)-ELISA 87.2 91.3 79.2 85.7 97.4 Adult (09)-ELISA 85.9 91.3 83.3 85.7 94.9 Adult (13)-IFA 91.1 87.0 83.3 85.7 90.5 Adult (12)-ELISA 74.7 87.0 83.3 78.6 83.3 Adult (1 7)-FA 68.4 78.3 79.2 85.7 92.9 Adult (10-lHA 71.4 73.9 62.5 71.4 74.4 Adult (14)-lFA 75.9 82.6 79.2 71.4 50.0 Adult (09)-lHA 72.7 73.9 54.2 71.4 76.9 Adult (09)-lgE RAST 62.8 56.5 62.5 42.9 76.9 Adult (09)-MDD 55.1 60.9 41.7 71.4 69.2 Adult (10)-lgE RAST 62.8 56.5 62.5 42.9 79.5 a Figures in parentheses give number of sera tested in each group. Seroreactivity and morbidity S. mansoni. The serum bank contained 178 sera from individuals from the S. mansoni endemic area who had no clinical signs, and 18 sera from persons with hepatosplenomegaly (liver more than 2 cm below the right costal margin or xiphoid and a palpable spleen). One-third of the group with hepato- splenomegaly had been treated previously. The sensi- tivity rates of the tests were similar in both groups. The sensitivity of the COPT and ISI with CEF-6 antigen was lower among those with low egg counts (1-100 per gram) and hepatosplenomegaly than in those in the same egg count class without clinical signs. 738 K. E. MOTT & H. DIXON Table 1 1. Correlation coefficients (R) between ELISA OD reading and logwo (egg count + 1) according to age and sex (S. mansoni) Males' Females Total Age group (years) Age group (years) Age group (years) Antigen/test 5-14 15-34 )35 5-14 15-34 > 35 5-14 15-34 35 (48-49) (55-56) (24-25) (35-36) (41-42) (23) (83-85) (96-98) (47-48) 05/ELISA 0.52 0.33 0.56 0.49 0.56 0.70 0.52 0.46 0.64 07/ELISA 0.64 0.43 0.27 0.40 0.72 0.61 0.54 0.58 0.52 09/ELISA 0.42 0.25 0.35 0.32 0.45 0.74 0.38 0.34 0.60 11/ELISA 0.35 0.15 0.36 0.24 0.60 0.61 0.31 0.40 0.53 15/ELISA 0.80 0.60 0.65 0.33 0.81 0.83 0.58 0.70 0.77 a Figures in parentheses give the number of sera tested in each group. The excretory-secretory antigen in the ELISA showed a significantly higher level of seroreactivity in individuals with hepatosplenomegaly than in those without a palpable liver or spleen (P < 0.008), al- though there was no difference between the mean age and faecal egg counts of the two groups. In all other test systems, the levels of seroreactivity to both egg 1.5 z a 0 I.O 0.5 and adult antigens were higher in individuals with hepatosplenomegaly, though the difference was not statistically significant (P > 0.05). S. japonicum. The serum bank contained 50 sera from persons from the S. japonicum endemic area with eggs in their faeces and no clinical findings, and * ,, 0 ** *0 v *0 00 *00 * * 00 * 0 0 * * 0. * 00 0 0 * 0 * _ 0 S:% A 0 0 0 * 0 00* 0 0 0 1.0 2.0 3.0 4.0 LOG 10 (EGG COUNT + 1) Fig. 1. Correlation between ELISA OD reading, with CEF-6 egg antigen, and S. mansonifaecal egg count in boys aged 5-14 years. IMMUNODIAGNOSIS OF SCHISTOSOMIASIS Table 12. Correlation coefficient (R) between ELISA OD reading and logio (egg count + 1) according to age and sex (S.japonicum) Males' Females Total Age group (years) Age group (years) Age group (years) Antigen/test 5-14 15-34 > 35 5-14 15-34 35 5-14 15-34 35 (12) (23) (21) (3) (19) (3) (15) (42) (24) 04/ELISA 0.91 0.59 0.62 - 0.61 0.86 0.75 0.60 0.65 06/ELISA 0.98 0.45 0.48 - 0.59 0.82 0.59 0.50 0.56 08/ELISA 0.95 0.61 0.59 - 0.73 0.99 0.76 0.66 0.65 a Figures in parentheses give the number of sera tested in each group. 15 sera from persons with hepatosplenomegaly (4 of the 5 persons who had zero egg counts had been treated 2-4 weeks prior to serum collection). The low specificity among those with hepatosplenomegaly was probably related to persistent antibody after treat- ment. The quantitative ELISA readings in persons with Table 13. Coefficients obtained by multiple regression analysis using logio (egg count + 1) (b1), age (b2), and sex (b3) (S. mansoni) Antigen/test bi b2 b3 R 1 O/RIA 0.289 0.0857a -0.594 0.31 6a 09/RIA 0.662 0.1 6630 -1.297 0.343° 15/ELISA 0.168a0 -0.0002 -0.053b 0.6710 07/ELISA 0.256° -0.0020 -0.054 0.568a O5/ELISA 0.1 34a 0.0003 0.013 0. 51 60 1 O/ELISA 0.096a 0.0025b - 0.089c 0.401° 09/ELISA 0.099a 0.0020 -0.093c 0.435a 01/COPT 3.198a 0.0628 -0.229 0.3880 13/IFA 0.560a 0.0176b -0.558b 0.324a 14/IFA 0.234 0.0165 -0.669 0.125 17/IFA 0.718a 0.0111 -0.038 0.286a 1 1 /ELISAd 0.966a 0.0054 -0.178 0.367a 1 1 /ELISA' 0. 109° 0.0014 -0.015 0.389a 12/ELISA 1.1 160 0.0148 -0.472 0.397a O9/IHA 1.048a 0.0337b -0.460 0.304° 1O/IHA 0.950a 0.0360b -0.650 0.296a 05/IHA 1.448a -0.0300c -0.095 0.546a 02/RIA 4.034C - 0.3389a -1.988 0.3100 a Significantly different from 0; P < 0.001 b Significantly different from 0; P < 0.05. c Significantly different from 0; P < 0.01 d Result expressed as 1o92 (titre). ' Result expressed as OD reading. S. japonicum infection and hepatosplenomegaly were higher than in those without hepatosplenomegaly; however, the egg counts in those with hepatospleno- megaly were also higher. Cross-reactivity between S. mansoni antigens and S. japonicum sera Qualitative results. The cross-reactivity between S. mansoni antigens and S. japonicum sera was over 7001o (Table 14). High levels of cross-reactivity occurred with two S. mansoni egg antigens, CEF-6 (81.0%) and fractionated SEA (81.0%), and the puri- fied glycoprotein adult antigen (95.3%o). These S. mansoni egg antigens appeared to be highly specific for schistosomiasis, since they reacted with few sera from egg-negative people from S. mansoni (7-12%o false positive) or S. japonicum (57o false positive) endemic areas. In order to evaluate further the cross-reactivity of S. mansoni antigens with sera from S. japonicum- infected persons, the reactivity of gut and somatic antigens in the IFA was assessed for S. mansoni and S.japonicum sera (Table 15). The reactivity to the S. mansoni somatic antigen was similar for those with S. japonicum or S. mansoni eggs in their stools (77.1 No and 77.9% respectively). However, the reacti- vity of the same S. japonicum sera to the gut antigen was low (54.2%), and the specificity of this antigen was high (91.7%). Quantitative results. Except in the COPT, all S. mansoni antigens showed significantly higher antibody binding than with sera from persons with S.japonicum eggs in their stools than with sera from persons without S. japonicum eggs (P < 0.01); the COPT results showed similar percentages of eggs with precipitins in both groups of sera. The S. mansoni egg fraction (antigen 02) showed significantly higher levels of antibody binding with sera from S. japonicum egg-positive persons than with 739 K. E. MOTT & H. DIXON Table 14. Results of the ELISA with S. mansoni antigens on sera from persons living in an S.japonicum endemic area Antigen Sensitivity' False negative' Specificityb False positiveb(% (%) (% (% CEF-6 81.0 19.0 94.9 5.1 Fractionated SEA 81.0 19.0 92.1 5.3 Crude SEA 76.0 19.0 92.1 7.9 Saline-extracted adult 81.0 9.5 89.7 5.1 Excretory-secretory adult 70.0 25.0 84.6 12.8 Purified glycoprotein adult 95.3 0.0 84.6 5.1 TCA-extracted adult 76.2 4.8 79.5 15.4 a Tested on 20-21 sera from persons with S. japonicum eggs in their stools. b Tested on 38-39 sera from persons with negative stool examination. Table 15. IFA reactivity of S. mansoni gut and somatic antigen with sera from people in S. mansoni and S.japonicum endemic areas Gut antigen Somatic antigen Positive Negative Positive Negative (%) (%) (%) (%) S. mansoni sera (269 samples) From egg-positive persons 66.8 33.2 77.9 21.3 From egg-negative persons 41.2 58.8 58.8 41.2 S. japonicum sera (60 samples) From egg-positive persons 54.2 45.8 77.1 18.8 From egg-negative persons 8.3 91.7 22.9 77.1 those from S. mansoni-infected persons (P < 0.001). The TCA-extracted S. mansoni adult antigen, the purified excretory-secretory glycoprotein, the somatic adult antigen, and both SEA antigens showed similar levels of reactivity in both groups. All other S. man- soni antigens had a significantly higher reaction with the homologous sera (P < 0.03). In 10 individuals with S. japonicum infection and hepatosplenomegaly, the CEF-6 antigen in ELISA was highly sensitive (90Vo); the same antigen in ISI gave negative results (0%). Both of these systems were highly sensitive and specific in persons with S. mansoni infection with or without hepatospleno- megaly. Cross-reactivity between S. japonicum antigens and S. mansoni sera Both the crude S. japonicum SEA and the fraction- ated SEA in the ELISA had low rates of reactivity with sera from individuals infected with S. mansoni (11.6% and 7.7%, respectively). Effect of treatment on seroreactivity Paired sera from 21 persons from Kenya with S. mansoni infection, collected before and 6 months after treatment, were tested in the ELISA with 3 egg antigens and 2 adult antigens (Table 16). The post- treatment sera gave a significantly lower mean OD reading with all antigens (P < 0.05). In the three paired sera from persons with no S. mansoni eggs in their stools after treatment, the adult antigens continued to detect significant levels of specific antibody. On the other hand, no specific anti- body to egg antigens was detected in two paired post- treatment sera. In unpaired sera from persons for whom no pre- treatment sera were available and who had no S. mansoni eggs in their stool 6 months after treat- 740 IMMUNODIAGNOSIS OF SCHISTOSOMIASIS Table 16. Serological reactivity in paired pre- and post-treatment S. mansoni sera Antigen/test Before treatment After treatment Statistical No. of system (OD) (OD) difference negative sera Egg antigens 15/ELISA 0.71 0.55 0.0001 2 05/ELISA 0.25 0.16 0.017 2 07/ELISA 0.86 0.71 0.027 2 Worm antigens 09/ELISA 0.63 0.54 0.017 0 12/ELISA 13.08 12.45 0.019 0 ment, the egg antigens (05, 07, 15) detected specific antibody in about one-third of the sera tested, and the adult antigens (09 and 12) detected antibody in 78% and 88% of sera, respectively. Cost evaluation The cost per 100 tests (excluding costs of labour and permanent equipment) in this collaborative study ranged from US $11 to US $520. The time required to perform 100 tests was between 4 and 30 man-hours. Differences in local conditions did not permit com- parison of labour or permanent equipment costs. The quantity of serum required was less than 0.05 ml for all tests and most were in the range of 0.01-0.02 ml. Comparison of quantities of antigen re- quired was not feasible since different standards were used by each laboratory. DISCUSSION This international collaborative study is the first recorded attempt to evaluate a range of both crude and purified egg and adult worm antigens for im- munodiagnosis of schistosomiasis. A well charac- terized serum bank with quantitative information on each donor regarding age, sex, faecal egg count, presence of other helminthic infection, morbidity, and previous specific treatment permitted a thorough analysis of the relationship of these variables to sero- reactivity. It is recognized, however, that this serum bank was organized from highly selected individuals, some of whom had high egg counts or severe morbidity. The age and sex distribution of the donors was not representative of a population from an en- demic area; nevertheless, they were persons whose sera would be examined in a diagnostic laboratory. The study also included sera from individuals with negative stool examinations who had had no possi- bility of exposure to schistosomiasis, and from persons living in endemic areas who were classified as uninfected after at least two Kato-Katz cellophane faecal thick smear examinations. These latter indi- viduals may have had a low intensity infection that was not detected by the parasitological examination, and this must be taken into account in interpreting the findings of the study. In the immunodiagnosis of parasitic diseases, the assessment of sensitivity and specificity is based on the results of the parasitological tests used to classify the serum donors as non-infected or infected (34). Al- though these tests vary in sensitivity and in the size of sample required, they constitute the accepted standard in classifying infected individuals according to their egg excretion levels. In areas with high preva- lence and high intensity of disease, most of the residents would be classified correctly by the current quantitative stool examinations. In areas of low prevalence and low intensity, the sensitivity of the parasitological test may be increased by increasing the faecal sample size and/or the number of examinations performed (34). If this is not done, the specificity of the serological tests would appear to be lower than the true value. The Kato technique, which was used to classify the donors in the present study, is insensitive in the low egg count range; the ideal antigen/antibody assay system would therefore be expected to have 100% sensitivity but lower specificity. In this study, the sensitivity of the S. japonicum egg antigens was close to lOOOo and the specificity was slightly lower. In general, S. mansoni egg antigens yielded a higher combined value for sensitivity and specificity than did worm antigens, irrespective of their degree of purity. The results did not indicate that any particular serological method for detecting anti- schistosome antibodies was superior. In view of this, 741 K. E. MOTT & H. DIXON choice of antigen and assay system will depend to a considerable extent on such factors as cost and avail- ability of antigen, and the reliability of the assay systems in laboratories other than those in which they were developed. Throughout the study, adult worm antigens in most test systems detected persistent antibodies in the sera of individuals who had been parasitologically cured, i.e., who had no eggs in their faeces after specific chemotherapy. Da Silva (35, 36) and Salih (37) also observed that antibodies to adult antigens persist in post-treatment sera, whether or not egg excretion has ceased. The ISI and MDD tests were evaluated for the first time as simple qualitative tests for immunodiagnosis of schistosomiasis. The ISI correlated very well with the ELISA using the same CEF-6 egg antigen, and de- tected nearly 90% of persons excreting over 100 eggs per gram of faeces. Further evaluation of the ISI under field conditions, in comparison with quantita- tive stool examinations, will indicate whether this technique can be useful for large-scale epidemio- logical studies. Several of the serological tests gave quantitative results indicating the level of antibody binding, and linear regression analysis was used to correlate these data with the faecal egg count. In general, there was little correlation, but the results for S. japonicum egg antigens were better than for S. mansoni egg antigens. S. japonicum antigens showed best correlation in young males (5-14 years of age). It is possible that primary infections occurred in this age group. The significant positive correlation between quanti- tative specific IgE levels to adult antigen and age was unexpected but suggests that further investigation is needed into the role of IgE in the human immune response to Schistosoma infection. In paired sera from the same individuals before and six months after treatment, all antigens showed a de- crease in the level of seroreactivity. However, the dif- ference between the pre- and post-treatment levels was significant only with CEF-6 and the S. mansoni SEA and with the purified glycoprotein and excretory- secretory antigens of the adult S. mansoni. Da Silva (35) and Capron (38) have previously shown an increase of antibody to adult antigen 2-4 weeks after treatment. It has also been shown that the antibody to adult worm persists in man for at least 6 months after treatment (38), and that, in experimental animals, the antibodies to egg antigens decrease below detectable levels within a few months. In the present study, the higher rate of seroreactivity to adult antigens in the sera from persons who had been treated up to 5 years previously extends these earlier observations and indicates that egg antigens will be more useful than adult antigens in defining cure of Schistosoma infec- tion. This conclusion is further supported by the greater post-treatment decrease in quantitative sero- reactivity measured in test systems using egg antigens than in those with adult antigens. None of the antigens in this study could discrimi- nate between sera from individuals with or without hepatosplenomegaly. The excretory-secretory adult S. mansoni antigen in ELISA showed a significantly higher quantitative seroreactivity in persons with hepatosplenomegaly than those without. A similar finding was reported by Santoro et al. (39) using adult S. mansoni worm antigens. Although not statistically significant, the other egg and adult S. mansoni antigens showed a higher level of seroreactivity in persons with hepatosplenomegaly. This trend was also observed by Goodgame et al. (40) in St. Lucia. Many sera from persons infected with S. japonicum reacted with S. mansoni antigens. In fact, for several of these antigens, sensitivity and specificity appeared to be equal to or even higher than in the homologous systems. The CEF-6 S. mansoni egg antigen in the ELISA reacted with most sera from S. japonicum- infected persons with hepatosplenomegaly; however, in the ISI, the same antigen was unreactive with these sera. The reasons for this loss of reactivity are un- known. As could be expected, however, the antibody titre of individual sera was lower with the heter- ologous than with the homologous antigen. In contrast, only a few sera from S. mansoni patients reacted with S. japonicum antigens and in- variably with low titre. Such one way cross-reactivity has been reported in man between S. mansoni and Trichinella spiralis (41). There are several possible ex- planations for this, such as factors inherent in the pre- paration of the different egg antigens, low density of the cross-reacting antigenic determinants in the S. japonicum antigen, or higher immunogenicity of the actual cross-reacting antigenic components of the S. mansoni antigen. Further systematic studies, including investigation of cross-reactivity with other intestinal helminth infections, are needed to explore these possibilities. At present, the main limitation to further develop- ment of immunodiagnostic techniques is the difficulty of large-scale production of uniform partially purified antigens at low cost. The study has provided cost estimates for a wide range of antigens and test systems which indicate that overall costs are still high. In certain endemic areas such as Japan, Puerto Rico, and St Lucia, schistosomiasis control efforts have reduced prevalence and intensity of infection to such low levels that the available parasitological tech- niques are becoming less cost-effective. This inter- national collaborative study is a first step towards developing and standardizing antigens for immuno- diagnostic tests for schistosomiasis for use in such areas. 742 IMMUNODIAGNOSIS OF SCHISTOSOMIASIS 743 RECOMMENDATIONS FOR FUTURE RESEARCH Work should continue on characterization and purification of antigens, and their comparison with standard crude antigens. Use should be made of all available technology, including monoclonal anti- bodies, to facilitate antigen purification, standardiz- ation, and quality control. Epidemiologically defined human sera are prerequisites for antigen quality control and standardization procedures. Inter- national collaborative efforts are needed between laboratories in the endemic areas and developed countries. In view of the results of the present study, S. mansoni egg and adult antigens should be evaluated in the field, in order to assess the correlation between serological and parasitological data. Current immunological techniques to detect circu- lating Schistosoma antigens have not been fully developed. As new technology becomes available, re- search in this area should be encouraged. Limited field testing of antigens and test systems that perform well in collaborative trials should be carried out. ACKNOWLEDGEMENTS The study was aided by Mr G. Brighouse, Centre de Transfusion de Sang, H6pital Cantonal Universitaire de Geneve, in preparing the serum bank; Mr E. Labbe, Mr E. Dayer, and Mr M. Cerdan in packing and shipping all the material to the par- ticipating laboratories; Mrs G. L. Bechet who prepared the documents for the evaluation meeting; Mrs Evelyn Borbor and Miss Jane Carter who prepared the final manuscript; and Mrs Mercedes Escribano-Matty who prepared the figure. RESUMt ETUDE COLLECTIVE SUR LES ANTIGENES DESTINES AU DIAGNOSTIC IMMUNOLOGIQUE DE LA SCHISTOSOMIASE Huit laboratoires de recherche des Etats-Unis d'Ame- rique et d'Europe, choisis en fonction de leurs publications sur les antigenes de Schistosoma mansoni et de S.japoni- cum, ont participe a cette etude. La banque de serums consistait en 395 serums bien documentes provenant de quatre regions d'endemie au Bresil (2 regions), au Kenya et aux Philippines. On a proc6d6 ainsi a l'evaluation d'un certain nombre de combinaisons de titrages immunolo- giques et d'antigenes, plus pr6cis6ment 21 avec S. mansoni et 4 avec S.japonicum. Les antigenes d'aeufs de S. mansoni ont temoigne d'une meilleure sensibilite que les antigenes de vers adultes, quelle que soit leur purete, lors d'infections evolutives a S. man- soni avant et apres traitement specifique. La r6activite sero- logique d'antigenes caracterises de ce type se correlait de facon quantitative avec la quantite d'ceufs determinee par examen coprologique chez les sujets de la tranche d'age 5-14 ans. Aucune correlation entre la morbidite due a S. mansoni et la seroreactivite n'a e observee dans l'un quelconque des systemes de titrage. Trois antigenes d'oeufs de S.japonicum ont presente une forte sensibilite et une specificite e1evee pour ce qui concerne la presence ou I'absence d'oeufs dans les selles. La seroreac- tivite des antigenes d'ceufs de S.japonicum caracterises presentait une correlation quantitative directe avec l'inten- site de l'infection a S.japonicum dans tous les groupes d'age. La methode immuno-enzymatique ELISA, applique selon diverses techniques, a donne de bons resultats avec les antigenes utilises dans 'etude en question. Le titrage immunologique sur lame a l'indium (ISI), qui constitue un systeme simple permettant une evaluation visuelle qualita- tive, s'est revele extremement sensible et specifique avec les antigenes d'ceufs de S. mansoni. Les resultats obtenus montrent qu'aucune methode d'immunodiagnostic particuliere l'emporte pour la recherche des anticorps antischistosomiens. Cette etude collective constitue une premiere etape en vue de mettre au point et de normaliser des antigenes destines au diagnostic immunologique de la schistosomiase. REFERENCES 1. Immunology of schistosomiasis: a WHO Memoran- dum. Bulletin of the World Health Organization, 51: 553-595 (1974). 2. PELLEY, R. P. ET AL. Schistosoma mansoni soluble egg antigens. I. Identification and purification of three major antigens and employment of radioimmunoassay for their further characterization. Journal of immu- nology, 117: 1553-1560 (1976). 3. HAMBURGER, J. ET AL. Schistosoma mansoni soluble egg antigens: Determination of the stage and species speci- ficity of their serologic reactivity by radioimmunoassay. Journal of immunology, 117: 1561-1566 (1976). 4. PELLEY, R. P. ET AL. Purified antigen radio-immuno- assay in serologic diagnosis of schistosomiasis mansoni. Lancet, 2: 781-785 (1977). 744 K. E. MOTT & H. DIXON 5. DUNNE, D. W. ET AL. Identification and partial purifi- cation of an antigen (wi) from Schistosoma mansoni eggs which is putatively hepatotoxic in T-cell deprived mice. Transactions of the Royal Society of Tropical Medicine and Hygiene, 75: 54-71 (1981). 6. MCLAREN M. L. ET AL. Serodiagnosis of human Schis- tosoma mansoni infections: Enhanced sensitivity and specificity in ELISA using a fraction containing S. man- soni egg fractions w, and al. Transactions of the Royal Society of Tropical Medicine and Hygiene, 75: 72-79 (1981). 7. HARRISON, D. J. ET AL. Immunoaffinity purification of Schistosoma mansoni soluble egg antigens. Journal of immunology, 122: 2210-2217 (1979). 8. ROTMANS, J. P. ET AL. Schistosoma mansoni: characterization of antigens in excretions and secretions. Experimental parasitology, 52: 171-182 (1981). 9. KELSOE, G. H. & WELLER, T. H. Immunodiagnosis of infection with Schistosoma mansoni: enzyme-linked immunosorbent assay for detection of antibody to circu- lating antigen. Proceedings ofthe NationalAcademy of Sciences ofthe United States ofAmerica, 75: 5715-5717 (1978). 10. ENGVALL, E. & PERLMANN, P. Enzyme-linked immuno- sorbent assay (ELISA). Quantitative assay of immuno- globulin G. Immunochemistry, 8: 871-874 (1971). 11. DEELDER, A. M. ET AL. Applicability of different anti- gen preparations in enzyme-linked immunosorbent assay for schistosomiasis mansoni. Americanjournal of tropical medicine and hygiene, 29: 401-410 (1980). 12. VAN HELDEN, H. P. T. ET AL. Are there stage-charac- teristic immunofluorescence patterns in schistoso- miasis? Transactions of the Royal Society of Tropical Medicine and Hygiene, 69: 309-311 (1975). 13. DEELDER, A. M. ET AL. Schistosoma mansoni: compari- son of the immunoperoxidase techniques, DASS and ELISA, for human diagnosis. Experimental para- sitology, 41: 133-140 (1977). 14. GIAEVER, I. Visual detection of carcinoembryonic anti- gen on surfaces. Journal of immunology, 116: 766-771 (1976). 15. KATZ, N. ET AL. A simple device for quantitative stool thick smear technique in schistosomiasis mansoni. Re- vista do Instituto de Medicina Tropical, 14: 397-400 (1972). 16. BOROS, D. L. & WARREN, K. S. Delayed hypersensi- tivity-type granuloma formation and dermal reaction induced and elicited by a soluble factor isolated from Schistosoma mansoni eggs. Journal of experimental medicine, 132: 488-507 (1970). 17. CARTER, C.E. & COLLEY, D.G. An electrophoretic analysis of Schistosoma mansoni soluble egg antigen preparation. Journal of parasitology, 64: 385-390 (1978). 18. CARTER, C. E. & COLLEY, D. G. Schistosoma japoni- cum soluble egg antigens: separation by con A chroma- tography and immunoaffinity purification. Journal of immunology, 18: 219-225 (1981). 19. CAPRON, A. ET AL. Structure antigenique des hel- minthes. Aspects immunologiques des relations h6te parasite. Pathologie et biologie, 16: 121-138 (1968). 20. DEELDER, A. M. ET AL. Schistosoma mansoni: demon- stration of two circulating antigens in infected hamsters. Experimentalparasitology, 40: 189-197 (1976). 21. NASH, T. E. Antibody response to a polysaccharide antigen present in the schistosome gut. I. Sensitivity and specificity. American journal of tropical medicine and hygiene, 27: 938-943 (1978). 22. DEELDER, A. M. & KORNELIS, D. A comparison of the IFA and the ELISA for the demonstration of antibodies against gut associated polysaccharide antigens in schistosomiasis. Zeitschrift fuir Parasitenkunde, 64: 65-75 (1980). 23. DEELDER, A. M. ET AL. Schistosoma mansoni: Charac- terization of two polysaccharide antigens and the immunological response to these antigens in mouse, hamster and human infections. Experimental para- sitology, 50: 16-32 (1980). 24. BOUT, D. ET AL. Immunodiagnosis of human parasitic diseases by the enzyme linked immunosorbent assay. In: Feldmann, G. P. et al., ed. INSERM symposium on immunoenzymatic techniques, Amsterdam, Elsevier- North Holland Publishing Company, 1975, pp. 175-182. 25. CARLIER, Y. ET AL. Evaluation of the enzyme-linked immunosorbent assay (ELISA) and other serological tests for the diagnosis of toxoplasmosis. Bulletin of the World Health Organization, 58: 99-105 (1980). 26. WIDE, A. ET AL. Diagnosis of allergy by an in vitro test for allergen antibodies. Lancet, 2: 1105-1107 (1967). 27. COLLEY, D. G. ET AL. Adoptive suppression of granu- loma formation by T lymphocytes and by lymphoid cells sensitive to cyclophosphamide. Cellular immunology, 46: 192-200 (1979). 28. HULDT, G. ET AL. Detection by immunofluorescence of antibodies to parasitic agents: use of class specific conju- gates. Annals of the New York Academy of Sciences, 254: 304-314 (1975). 29. BERGQUIST, N. R. & SCHILLING, W. G. E. E. Prep- aration of anti-human immunoglobulin for indirect fluorescent tracing of autoantibodies. In: Holborow, E. J., ed., Standardization in immunofluorescence, Oxford, Blackwell Scientific Publications, 1970, pp. 171-176. 30. FAGRAEUS, A. & BERGQUIST, N.R. The raison d'etre of standards in indirect immunofluorescence. Annals of the New York Academy of Sciences, 254: 69-76 (1975). 31. OLIVER-GONZALEZ, J. Anti-egg precipitins in the serum of humans infected with Schistosoma mansoni. Journal of infectious diseases, 95: 86-91 (1954). 32. YOGORE, M. G. ET AL. The circumoval precipitin (COP) test in schistosomiasis japonica. American journal of tropical medicine and hygiene, 17: 65-71 (1968). 33. OUCHTERLONY, 0. Antigen-antibody reactions in gels. Acta pathologica et microbiologica Scandinavica, 26: 507-515 (1949). 34. RuIz-TIBEN, E. ET AL. Intensity of infection with Schis- tosoma mansoni: its relationship to the sensitivity and specificity of serologic tests. American journal of trop- ical medicine and hygiene, 28: 230-236 (1979). IMMUNODIAGNOSIS OF SCHISTOSOMIASIS 745 35. DA SILVA, L. C. ET AL. Serum antibody changes after chemotherapy of patients with schistosomiasis mansoni. A statistical analysis. Revista do Instituto de Medicina Tropical de Sao Paulo, 17: 344-349 (1975). 36. DA SILVA, L. C. ET AL. Serum antibody changes in re- peated chemotherapeutic series in "parasitologically cured" patients with schistosomiasis mansoni. Revista do Instituto de Medicina Tropical de Sao Paulo, 18: 206-210 (1976). 37. SALIH, S. Y. ET AL. Detection of antibodies by enzyme- immunoassay in human Schistosoma mansoni infec- tions: A clinical and chemotherapeutic study. Tropen- medizin und Parasitologie, 29: 409-412 (1978). 38. CAPRON A. ET AL. Immunological studies in various types of schistosomiasis. Annals of the New York Academy of Sciences, 160: 863-879 (1969). 39. SANTORO, F. ET AL. Imuno-complexos na esquistosso- mose. II. Dosagem radioimunologica da ligacao do Clq-I125 ao IC. Revista do Instituto de Medicina Tropical de Sao Paulo, 18: 293-297 (1976). 40. GOODGAME, R. W. ET AL. Humoral immune responses in human schistosomiasis mansoni.Americanjournalof tropical medicine and hygiene, 27: 1174-1180 (1978). 41. ANDERSON, R. I. ET AL. Cross absorption studies per- formed with Schistosoma mansoni and Trichinella spiralis antigens in sera from patients with trichinosis. Experimental parasitology, 14: 323-329 (1963). 42. DRAPRON, R. & GUILBOT, A. T. Annales de technologie agricole, 11: 175 (1962). 43. Hsu, S. Y. ET AL. Schistosoma mansoni and S. japonicum: methylene blue test for the viability of schistosomula in vitro. Experimental parasitology, 41: 329-334 (1977). Annex I LIST OF PARTICIPANTS Organizing committee F. von Lichtenberg, Department of Pathology, Har- vard Medical School, Boston, USA A. Capron, Centre d'Immunologie et de Biologie Parasitaire, Institut Pasteur, Lille, France J. A. Cook, Edna McConnell Clark Foundation, New York, USA A. Davis, Parasitic Diseases Programme, World Health Organization, Geneva, Switzerland H. Dixon, Health Statistical Methodology, World Health Organization, Geneva, Switzerland K. E. Mott, Special Programme for Research and Training in Tropical Diseases, World Health Or- ganization, Geneva, Switzerland G. Torrigiani, Immunology, World Health Organiz- ation, Geneva, Switzerland Serum bank contributors Department of Parasitology, School of Public Health, University of the Philippines System, Manila, Philippines (E. G. Garcia) Centro de Pesquisas "Rene Rachou", Fundacao Oswaldo Cruz, Belo Horizonte, Brazil (G. Gazzi- nelli and N. Katz) Nucleo de Medicina Tropical, Universidade de Brasilia, Brasilia, Brazil (A. Prata) Tana River Project, Ministry of Health, Nairobi, Kenya (D. H. Smith) Participating laboratories Centre d'Immunologie et de Biologie Parasitaire, Institut Pasteur, Lille, France (A. Capron) Department of Parasitology, University of Leiden, Leiden, Netherlands (A. M. Deelder) Departamento de Biologia, Universidad de Puerto Rico, Rio Piedras, Puerto Rico (G. V. Hillyer) National Bacteriological Laboratory, Stockholm, Sweden (G. Huldt) Ross Institute of Tropical Hygiene, London School of Hygiene and Tropical Medicine, London, England (C. C. Draper) Winches Farm Field Station, St. Albans, England (M. J. Doenhoff) Department of Pathology, University of Chicago, Chicago, USA (R. P. Pelley) General Electric Research and Development, Schenectady, New York, USA (I. Giaever) Department of Pharmacology and Division of Geographical Medicine, Department of Medicine, Case Western Reserve University, Cleveland, Ohio, USA (J. Tracy) Department of Biology, Vanderbilt University, Nash- ville, USA (C. E. Carter) K. E. MOTT & H. DIXON Annex2 ANTIGEN PREPARATION Antigens from S. mansoni and S. japonicum eggs and S. mansoni adult worms were evaluated in this study, as well as whole S. mansoni eggs (for the COPT). This annex presents the laboratory protocols for antigen preparation: when the original published procedure was followed, only the reference is cited. CRUDE S. MANSONI EGG ANTIGENS Whole S. mansoni eggs (antigen 01) Schistosome eggs were obtained from the intestines of outbred general purpose mice infected with S. mansoni for seven weeks. The intestines were per- fused with cold saline (8.5 g/litre) until visibly clean, and minced in a Waring blender. The suspension was filtered through 50, 80, and 100 mesh sieves and cen- trifuged for 3 min at 200 g at room temperature. The tissue was removed from the supernatant and cold saline (17 g/litre) was added. Centrifugation and washing were repeated several times. An egg suspen- sion of approximately 100 eggs per drop was consti- tuted by adding 17 g/litre saline solution containing 2.0 g of sodium azide per litre, and stored at 4 'C. Crude soluble S. mansoni egg antigen (SEA) (antigen 05) The method of Boros & Warren (16) was used, as modified by Carter & Colley (17). FRACTIONATED S. MANSONI SOLUBLE EGG ANTIGEN PREPARATIONS S. mansoni eggfraction (antigen 02) The S. mansoni egg fraction was prepared using a modification of previously described procedures (4). Briefly, soluble egg glycoproteins were isolated by lectin affinity chromatography on Concanavalin A- Sepharose. The glycoproteins were then fractionated by ion-exchange chromatography on DEAE cellulose, elution being monitored with the monoclonal anti- body F5. Material eluting at a saline concentration of approximately 0.1 mol/litre was pooled. A 25-mg portion of this material was iodinated and gel-filtered through Sephadex G-200 in buffer containing 1 g of ovalbumin per litre. The material eluted with an apparent relative molecular mass (RMM) of between 68 000 and 200 000, corresponding to MSA-2c (R. P. Pelley, personal communication, 1982). Fractionated S. mansoni SEA (antigen 07) The method of isolation of this antigen has been described by Harrison et al. (7). CEF-6 S. mansoni egg antigen (antigen 15) This antigen was prepared according to the method of Dunne et al. (5) with the following modifica- tions. Eggs isolated from the liver and intestines of mice were mechanically homogenized in 17 g/litre NaCl solution and centrifuged at 20 000 g for 4 h at 4 'C. Aliquots of 4 ml of the resulting supernatant were applied to a CM-Sepharose column which had been equilibrated in starting buffer (0.01 mol/litre phos- phate buffer, pH 7.2, containing 8.5 g/litre NaCI). Under these conditions, only antigens ac and wl (those present in CEF-6) were absorbed on the column, and all the other egg components were eluted from the column with starting buffer at a constant flow of 20 ml/hour. When all non-absorbed material had been washed from the column, 0.01 mol/litre phosphate buffer containing 1 mol/litre NaCl was used to elute CEF-6. CRUDE S. MANSONI ADULT ANTIGENS Saline-extracted S. mansoni adult antigen (antigen 10) This antigen was prepared according to the method of Capron et al. (19). After removal from infected hamsters, the adult worms were washed 6-7 times in 8.5 g/litre NaCl solution and frozen at - 30 'C. For extraction, the worms were allowed to thaw at room temperature and 15 ml of 1 g/litre NaCl solu- tion was added to each gram of wet weight adult worms. The mixture was homogenized in a Virtis mixer in an ice bath three times for 5 minutes, and then placed in the freezer at -30 IC. Four successive "freeze-thaw" procedures were completed. Finally, the suspension was centrifuged at 40 000 g at 4 'C for 1 hour. The supernatant was dialysed against distilled water (one volume of antigen supernatant against 100 volumes of water) for 24 h, then lyophilized and stored at 4 'C. 746 IMMUNODIAGNOSIS OF SCHISTOSOMIASIS 747 TCA-extracted adult S. mansoni antigen (antigen 12) This antigen was isolated from adult S. mansoni worms, as described by Deelder et al. (20). Adult worms were isolated by perfusion of golden hamsters 48 days after infection with 1500 cercariae each. The worms were washed and homogenized in a glass homogenizer. The homogenate was then centri- fuged for 20 min at 25 000 g at 4 °C, and the super- nate was stored at 4 'C. The pellet was homogenized and centrifuged again and the supernate added to the first supernate. This solution was then treated with an equal volume of trichloroacetic acid (TCA) solution (150 ml/litre) and stirred with a magnetic stirrer for 20 min at room temperature. The mixture was centri- fuged for 10 min at 5000 g and the supernate was stored. The pellet was resuspended in distilled water and then treated again with TCA as described above. After centrifugation, the second supernate was added to the first one and this was then centrifuged for 15 min at 25 000 g at 4 'C. The supernate was then dialysed exhaustively against distilled water for at least 24 h, with 3 changes of the water. The antigen solution was then freeze-dried and stored at 4 'C until used. S. MANSONI EXCRETORY-SECRETORY ANTIGENS S. mansoni purified glycoprotein antigen (antigen 09) This glycoprotein antigen was isolated by the method of Capron (personal communication, 1981). A semi-purified extract of S. mansoni was obtained from the incubation of adult worms, or from the ex- traction of the soluble adult antigen of S. mansoni by immunoadsorption. The immunoabsorbant used was Concanavalin A, conjugated to Sepharose 4B (ConA-Sepharose) on 15-cm columns (diameter, 9 mm). a A 100-mg sample of the total antigenic extract was dissolved in 2 ml of acetate buffer, pH 6, and fixed on the column for a total of 80 min at room temperature. The column was then washed with 100 ml of acetate buffer, pH 6. The ConA-binding glycoproteins were eluted with 50 ml of 0.2 mol/litre methyl D-mannopyranoside buffer, dialysed for 24 h at 4 IC and concentrated once on Amicon PMlO,b and then ultrafiltered and concen- trated with collodion (collagen) tubes. The protein concentration was measured according to Lowry's method as modified by Drapron & Guilbot (42). a Item K9/ I5, Pharmacia Fine Chemicals AB, Box 175, Uppsala, Sweden. b Amicon BV, Mechelaarstraat 11, Oosterhout NB, Nether- lands. S. mansoni excretory-secretory antigen (ESA) (antigen 11) The preparation of an ESA with high relative mol- ecular mass was done according to the procedure of Rotmans et al. (8). Adult S. mansoni were collected from golden ham- sters by perfusion with a sterile Dulbecco's balanced salt solution, 48 days after percutaneous exposure of the hamsters to 1500 cercariae. About 8000 worms from 15 hamsters were first washed with the balanced salt solution and then with sterile culture medium H-l99.c These washings and the subsequent incu- bation were performed in a laminar flow cabinet in order to prevent contamination of the culture media. The worms were divided among 9 Erlenmeyer flasks, each containing 100 ml of H-199 medium supple- mented with penicillin (l10 U/litre), streptomycin (100 mg/litre), chloramphenicol (25 mg/litre) and miconazole (25 mg/litre). The concentrations of bi- carbonate and Hepes were 5 mmol/litre and 20 mmol/litre respectively. Glucose was added to a final concentration of 30 mmol/litre. After 1 hour's incubation, the pH of the medium was adjusted to 7.35 with 0.1 mol/litre sodium hydroxide solution. The mixture was incubated at 37 °C in the dark, in an atmosphere of92% air and 8% CO2. The culture fluid was mixed very gently on an orbital shaker. Every 2 days the worms were suspended in fresh medium and on the intermediate days the pH was adjusted to 7.3 by adding dilute sodium hydroxide solution. The vi- ability ofthe worms was checked by observation on an inverted microscope and by a methylene blue exclu- sion test (43). Cultures contaminated with micro- organisms were not used for the isolation of schisto- somal antigens. The pooled culture media were centri- fuged at 25 000 g for 20 min. Concentration and removal of small molecules was by ultrafiltration. After lyophilization, the ESA preparation was stored at -20 'C. Fractionation of ESA was carried out by gel filtra- tion on Ultrogel AcA-44d in a 1.6 x 80 cm column, at 4 'C. The gel was equilibrated with 0.05 mol/litre phosphate buffer, pH 7.2. Approximately 15 mg of ESA preparation was dissolved in 2 ml of phosphate buffer and centrifuged for 5 min at 15 000 g. The supernatant was applied to the column, and fractions of 5 ml were collected at a flow rate of 30 ml/hour. Absorbance at 280 nm was used to indicate the pres- ence of protein in the column eluates. The fractions containing the proteins, eluted with the void volume, were pooled and dialysed overnight at 4 IC against deionized water. After lyophilization the preparation was stored at - 20 'C. c Flow Laboratories Ltd., Irvine, Ayrshire, Scotland. d LKB-Produkter AB, S-161 26 Bromma, Sweden. 748 K. E. MOTT & H. DIXON S. MANSONI ADULT WORM SECTIONS S. mansoni gut-associated polysaccharide antigen (GASP) (antigen 13) Adult male worms were fixed in Rossman's fixative and embedded according to the method of Nash (21). S. mansoni gut and somatic antigens (antigens 14 and 17) Adult S. mansoni worms were harvested from mice infected 8-10 weeks previously with 100 cercariae. Directly after harvesting, 100 worms were embedded in a freezing compounde in 0.5-ml gelatin capsules, frozen, and kept at - 70 °C until sectioned. Sections of 5 zm thickness were cut in a Harris cryostat and placed directly on Cook microscope slides with three wells. Two sections were placed on each antigen spot. These slides were stored at - 20 IC until used. Before staining, the slides were allowed to dry at room temperature for 2 hours and were then fixed in acetone for 10 min. S. JAPONICUM EGG ANTIGENS Crude soluble S. japonicum egg antigen (SEA) (antigen 06) The procedure for preparation of this antigen has been described by Carter & Colley (17). Fractionated soluble S. japonicum egg antigen (antigen 08) The procedure for this defined S. japonicum egg antigen has been described by Carter & Colley (18). S. japonicum egg glycoprotein antigen fraction, GP-2 (antigen 04) This egg antigen was isolated according to the method of Tracy & Mahmoud (personal communi- cation 1981). Parasite eggs were recovered from the liver and in- testines of female CF-1 mice, exposed 9 weeks earlier to 50 cercariae of a Philippine strain of S. japonicum. S. japonicum SEA was prepared by grinding the eggs in phosphate-buffered saline (PBS), pH 7.4, at a concentration of 109 eggs/litre, as described for S. mansoni SEA (16). After centrifugation at 150 000 g for 2 hours, the supernatant was withdrawn and dia- e Miles Laboratories, Division of Ames, 1000 Lausanne 6, Switzerland. lysed overnight at 4 °C against 2 litres of PBS con- taining 10- 4 mol of MnCl2, 104 mol of CaCl2, and 0.2 g of sodium azide per litre. Dialysed S. japonicum SEA (40 mg of protein) was loaded onto a column (1.6 x 28 cm) of Concanavalin A-Sepharose 4B (Con A-Sepharose), equilibrated with the above buffer. The SEA was washed onto the column with three 1.0-ml portions of buffer and the column outlet clamped off. After 2 hours, the column was developed with 3 bed volumes of starting buffer to elute any unbound material. Next, weakly absorbed proteins were eluted with 3 bed volumes of a buffer containing 0.01 mol/litre of sodium phosphate, 0.5 mol/litre of NaCl, 0.1 mol/litre of D-glucose, 10-4 mol/litre of MnCl2, 104 mol/litre of CaCl2, and 0.2 g/litre of sodium azide (pH 7.4). The eluted glycoproteins were concentrated to approximately 2 ml in an Amicon type 12 stirred ultrafiltration cell fitted with a PM1O membrane. The Con A-Sepharose-purified glycoprotein frac- tion was subjected to gel filtration chromatography on a column (1.6 x 95 cm) of BioGelA l.5m, 200-400 mesh,f equilibrated at 4 °C with 0.05 mol/litre sodium phosphate buffer, pH 7.4, containing 0.2 g/ litre sodium azide. Fractions of 1.5 ml were collected at a linear flow rate of 3 cm/h. Of the three major protein peaks (absorbance at 280 nm) eluting from the column, the first, designated GP-1 (RMM approxi- mately 590 000) chromatographed between the marker proteins, bovine thyroglobulin (RMM = 675 000) and horse apoferritin (RMM = 475 000). The second peak, GP-2, displayed an apparent RMM of 245 000 and eluted between horse apoferritin and rabbit IgG (RMM = 155 000). The third peak, GP-3, chromatographed between the markers, bovine serum albumin (RMM = 68 000) and chymotrypsinogen A (RMM = 25 700) and had an apparent RMM of about 50 000. Further purification of these individual glyco- protein peaks was accomplished by subjecting each to a second gel filtration step on the same column. From a crude SEA preparation containing 40 mg of protein, approximately 240 ,g of GP-1, 560 ug of GP-2, and 350 ytg of GP-3 were recovered. Analysis of the purified glycoproteins by polyacrylamide gel elec- trophoresis revealed that GP-1 and GP-2 each con- tained a single major Coomassie blue-staining species (relative mobilities 0.05 and 0.16, respectively), which also stained with the periodic acid-Schiff reagent (PAS). Several protein species were found in GP-3. Of these, the major protein species displayed a relative mobility of 0.25. Each of the purified glycoproteins gave a single precipitin line by double immunodif- fusion against 16-week S. japonicum-infected mouse serum. Purified GP-1 and GP-2 displayed apparent RMMs of approximately 139 000 and 66 000, respec- f Bio-Rad Laboratories, 8152 Glassbrugg-Zurich, Switzerland. IMMUNODIAGNOSIS OF SCHISTOSOMIASIS 749 tively, when electrophoresed in 10070 polyacrylamide gels in the presence of sodium dodecylsulfate and 2-mercaptoethanol. The purified antigens were stored frozen in 0.05 mol/litre sodium phosphate, pH 7.4, at -70 OC. Annex 3 TEST SYSTEMS This annex presents the laboratory protocols of the different modifications of the test systems. ENZYME-LINKED IMMUNOSORBENT ASSAY (ELISA) ELISA procedure with antigens 05, 06, 07, and 08 The antigen preparation was diluted in 0.1 mol/ litre NaHCO3, pH 9.5, to a final concentration of 1 mg/litre. Aliquots of 1 ml of diluted antigen were added to each tube,a and incubated at 37 °C for 1 hour and then overnight at 4 'C. The antigen preparation was removed by aspiration and washed 3 times with 0.01 mol/litre NaPO4 buffer, pH 7.2, containing 8.5 g/litre NaCl and 0.5 ml/litre Tween 20 (PBS- Tween). Each wash lasted 5 min. The test sera were diluted 1:500 in PBS-Tween con- taining I g/litre bovine serum albumin (BSA).b Ali- quots of 1 ml of the diluted antisera were then added to each tube and incubated for 1 hour at 37 'C. Dupli- cate tubes were prepared for each test serum. The sera were removed by aspiration and washed 3 times as de- scribed above. The conjugate used was peroxidase-conjugated goat anti-human IgG fraction (H-chain specific),c which was diluted 1:500 in the same buffer as was used to dilute the antisera. Aliquots of 1 ml of diluted conjugate were added to each tube and incubated for 1 hour at 37 'C. The conjugate was removed by aspiration and the tubes washed as described above. The substrate was prepared by dissolving 40 mg of 2-phenylenediamine (o-phenylenediamine, OPD) in 100 ml of citrate-phosphate buffer, pH 5.0, (prepared by titrating 0.2 mol/litre Na2HPO4 in 0.1 mol/litre citric acid) and adding 0.5 ml of 300 ml/litre peroxide solution. Aliquots of 1 ml of substrate were then added to each tube and incubated for 1 hour at room temperature in darkness. The reac- tion was stopped by adding 0.05 ml of 2 mol/litre ° Falcon polystyrene tubes (12 x 75 mm), Becton Dickinson AG, 4002 Basel, Switzerland. b Miles Laboratories, Elkhart, IN, USA. c Cappel Laboratories Inc., Cochranville, PA 19330, USA. H2SO4 to each tube. The absorbance at 492 nm was determined spectrophotometrically. ELISA procedure with antigen 04 The antigen preparation was diluted in 0.15 mol/ litre NaCI solution to a final concentration of 0.2 mg/litre and 50 ul were added to each well of a 96- well microtitration plate." The plate was incubated in a humidified atmosphere at 23-25 °C for 16-18 h. The antigen solution was removed by inverting the plate and tapping on the edge of a sink. The plate was washed twice by flooding with PBS-Tween (PBS, pH 7.4, containing 0.5 g/litre Tween 20 and 0.2 g/ litre sodium azide), and after 2 minutes, the wash fluid was removed by inverting the plate and tapping. The plate was then flooded with 10 g/litre BSA sol- ution and incubated at 37 'C. After 1 hour, the BSA solution was removed and the plate was washed once with PBS-Tween. Any traces of wash fluid remaining in the wells were removed by aspiration. With a starting dilution of 1:10, 2-fold serum dilu- tions were prepared, 50-id samples were added to the wells, and the plate was incubated at 37 'C for 3 hours. The contents of each well were removed by as- piration and the plate was washed 3 times with PBS- Tween. After the final wash, traces of wash fluid were aspirated from each well. The conjugate used was alkaline phosphatase-con- jugated goat antibodies to human IgG. The antibody fraction was isolated from a commercially prepared antiserume by chromatography on a column ofhuman IgG coupled to Sepharose 4B. 1 mg of antibody pro- tein was conjugated to 3 mg of alkaline phosphatase (calf intestine, type VII),f using a one-step glutaralde- hyde method (10). The conjugate was diluted to 10 ml with 50 g/litre BSA and stored in 100-1l aliquots at -70 'C. The working dilution, determined by chequerboard titration, was 1:500. Aliquots of 50 1u of conjugate were added to each well and the plate was incubated at 25 'C for 16-18 h. The conjugate was d Dynatech Produkte AG, 8302 Kloten, Switzerland. Cappel Laboratories Inc., Cochranville, PA 19330, USA. Sigma Chemical Company, St Louis, MO 63178, USA. 750 K. E. MOTT & H. DIXON removed by aspiration and the plate was washed 3 times with PBS-Tween. Again, remaining traces of wash fluid were aspirated from the wells. Aliquots of 50 1l of substrate solution (1.2 g of 4-nitrophenyl phosphate per litre of sodium carbo- nate buffer (0.05 mol/litre, pH 9.8), containing 1 mmol/litre MgCl2) were added to each well and the plate was incubated at 37 'C. After 30-45 min, the plate was assessed visually against a white background. The incubation time was chosen such that a distinct difference was observed between the colour of the 1:640 dilution (pale yellow) and the 1:1280 dilution (virtually colourless) for the reference serum pool. This pool was prepared from the sera of 10 individuals who were chronically infected with S. japonicum. The endpoint titre of the reference was thus defined as 1:640. The remaining 11 test sera were assessed in an identical manner within 2 min and the endpoint titres recorded. Sera with a titre greater than or equal to 1:80 were scored as positive; those with an ELISA titre less than or equal to 1:20 were scored as negative, while those with a titre of 1:40 were scored as equivocal. ELISA procedure with antigen 15 CEF-6 S. mansoni egg antigen was used according to the method of McLaren et al. (6). A 1:50 dilution of each serum sample was prepared by adding 15 A1 of serum to 0.75 ml of PBS-Tween (0.5 ml of Tween 20 per litre of PBS). Antigen-coated platesg were pre- pared (results were based on four different batches of CEF-6 antigen) as described previously for soluble egg antigens. The stock antigen solution was diluted to a protein concentration of 500 ,g/litre in 0.05 mol/ litre coating carbonate buffer, pH 9.6 (159 g of Na2CO3 and 2.93 g of NaHCO3 per litre of water). Aliquots of 150 yd of diluted antigen were added to each well ofan ELISA microtitration plate, which was then covered and left overnight at room tempera- ture. Reference and control sera, prepared fresh for each test, were diluted 1:300 in PBS-Tween. The sera con- sisted of (a) a pooled reference positive serum for end- point determination; (b) a pooled control positive serum to assess reproducibility of test measurements; and (c) a pooled negative serum to check for non- specific background reaction. The antigen-coated plates were rinsed twice with washing solution (9 g ofNaCl and 0.5 ml ofTween 20 per litre of water), and reference and control sera were added to the plate in the appropriate wells. Aliquots of 120 Al ofPBS-Tween were added to the wells followed by 20 u1 of the 1:50 serum sample. The final serum di- lution was therefore approximately 1:300. The plates ' Type M129A, Dynatech Produkte AG, 8302 Kloten, Switzerland. were covered and incubated for 2 h at room tempera- ture. They were then rinsed 3 times in washing solu- tion, which was left on for 3 min in each cycle. A 1:2500 dilution of anti-human IgG peroxidase- labelled conjugateh was then prepared in PBS-Tween, and 150 gl added to each well. The plates were covered and incubated for 3 h at room temperature, and then the washing procedure was repeated. A peroxidase substrate was then prepared from 50 ml of distilled water, 24.7 ml of 0.1 mol/litre citric acid, 25.3 m! of 0.2 mol/litre Na2HPO4, 1 ml of stock OPD,' and 0.05 ml of 60 ml/litre H202. Aliquots of 150 ul of substrate were added to each well. The reaction in one PBS control well was stopped with 25 gl of 1 mol/litre H2SO4 and this was used as a blank. The optical densities of the positive reference samples at 492 nm were monitored progressively until a level of 0.75 was reached, at which point the reactions on the rest of the plate were stopped. Samples were considered to be negative if they had an OD less than 0.24, doubtful if the OD was between 0.24 and 0.26, and positive if the OD was greater than 0.26. ELISA procedure with antigen 12 TCA-extracted S. mansoni adult worm antigen (AWA) was used according to the method of Kelsoe& Weller (9), as modified by Deelder & Kornelis (22). ELISA microtitration trays were incubated for 8 h at 37 °C with 100 ,dof a solution of 150 mg ofAWA- TCA per litre of PBS (0.05 mol/litre, pH 7.8). The trays were then washed with 0.005 mol/litre PBS and coated overnight at 4 °C with 20 g/litre BSA in PBS (100 ul per well). The trays were washed with PBS and 2-fold serum dilutions in PBS (starting from a dilution of 1:25) were added; the plates were then incubated for 1 hour at 37 IC and washed again. A horseradish peroxidase-labelled sheep anti- human IgG (H + L) conjugatei (diluted 1:1000 in PBS containing 20 g/litre BSA) was added, and the plates were incubated again for 1 h at 37 'C. After incubation, the plates were washed and 5-aminosalicylic acid was added. The results of the reaction were determined by measuring the absorb- ance at 450 nm with an automated inverted measuring microscope (22). Positive and negative control sera were included in each tested set of sera. The test results were determined on the basis of the histograms ofthe absorbance values ofthe serum dilu- tions. The reciprocal value of the last serum dilution h Miles Laboratories, Elkhart, IN, USA. i 100 mg of 2-phenylenediamine mixed thoroughly with 10 ml of absolute methanol, and stored in the dark at 4 'C. This is stable for 1-2 weeks only. i Lot 14, Institut Pasteur, Paris. IMMUNODIAGNOSIS OF SCHISTOSOMIASIS 751 showing a significant difference from that of the blank (buffer) control was taken as the titre. ELISA procedure with antigen 11 High relative molecular mass excretory and secre- tory antigens were used in the ELISA, as described by Deelder et al. (13). Flat-bottom polystyrene microtitration plates were coated with the high relative molecular mass fraction ofESA (RMM greater than 200 000), by adding 100 Al of a solution of 50 mg high-RMM ESA per litre of 0.1 mol/litre sodium carbonate buffer, pH 9.6. The plates were incubated at 37 IC for 3 h and then stored overnight with the antigen solution at 4 'C. The tray was washed with 5 mmol/litre PBS, pH 7.8, and 2-fold serial dilutions of human sera (starting from 1:10) were prepared in 0.05 mol/litre PBS, pH 7.8. Trays were incubated at 37 'C for 45 min. After washing with 5 mmol/litre PBS, horseradish peroxi- dase-labelled sheep anti-human IgG (H + L) conju- gate (diluted 1:1000 in PBS containing 10 g/litre BSA) was added. Trays were incubated at 37 'C for 45 min. Excess conjugate was then removed by washing with 5 mmol/litre PBS. The amount of coupled conjugate was determined by incubating the trays with 5-aminosalicylic acid (0.7 ml/litre) and hydrogen peroxide (0.05 ml/litre). After 1 hour the staining reaction was stopped by the addition of 50 yd of 1 mol/litre NaOH. The results of the reaction were determined by measuring the absorbance at 450 nm with an automated inverted measuring microscope (22). Absorbance values were recorded at 1 cm pathlength and at 1:320 serum dilu- tion. For standardization of absorbance values, a standard positive serum (titre between 5120 and 10 240; A4soat 1:320 serum dilution, 0.92 ± 0.35) was included on each microtitration tray. The absorbance values, measured on one tray, were multiplied by a correction factor, calculated from the average absorb- ance values of the standard positive serum (at 1:320, 1:640, and 1:1280 serum dilution). For the determination ofthe background staining, a standard negative serum was included on every micro- titration tray. Titres were determined on the basis of the histograms of the absorbance values of the serum dilutions. The reciprocal value of the last serum dilu- tion showing a significant difference (0.05 absorbance unit) from that of the negative control serum was taken as the titre. ELISA procedure with antigens 09 and 10 The method described by Engvall & Perlmann (10) and modified successively by Bout et al. (24) and Carlier et al. (25) was used with saline-extracted S. mansoni worm antigen and purified glycoprotein fraction S. mansoni adult antigen. Aliquots of 1 ml of the antigen solution (concen- tration determined previously) were placed in poly- styrene tubes,k incubated for 3 h at 37 °C and then overnight at 4 'C. Tubes prepared in this manner were kept for up to one month at - 20 'C. Just before use, the antigen-coated tubes were washed 3 times with 0.01 mol/litre PBS (pH 7.2) con- taining 2 ml/litre Tween 20. After the addition of 1 ml of serum diluted 1:500 in 0.01 mol/litre PBS (pH 7.2) with 6 ml/litre Tween 20, the tubes were in- cubated for 4 h at room temperature and were then washed 3 times with the same buffer. Peroxidase-labelled sheep anti-human IgG (H + L) antibodies' were diluted 1:2500 in 0.01 mol/litre PBS (pH 7.2) containing 5 ml/litre Tween 20, and 1 ml of the diluted conjugate was pipetted into each tube. After mixing, the tubes were incubated overnight at 4 'C in the dark, and then washed 3 times with buffer. After successive washing, the amount of peroxidase fixed to the tube was determined by using hydrogen peroxide as the substrate and 2-dianisidine as the indi- cator. A solution containing 1 ml of 0.01 mol/litre PBS (pH 6.0), 1 ml of hydrogen peroxide diluted 1:100 in distilled water, 97 ml of distilled water, and 1 ml of a solution of 10 mg of 2-dianisidine in 1 ml of methanol was constituted and 1-ml aliquots added to each tube. After incubation for 1 hour at room temperature, the reaction was terminated by the addition of 50 ul of 5 mol/litre HCI. The optical density was determined at 405 nm against a substrate blank, and the results expressed as OD units. Three independent determi- nations were made for each sample, and the mean value was taken as the final result. ASSAYS USING RADIOISOTOPES RAST with antigens 09 and 10 Crude saline-extracted S. mansoni adult antigen and purified glycoprotein fraction S. mansoni adult antigen were used in a radioimmunosorbent test (RAST), as described by Wide et al. (26) using '25I-labelled anti-IgE.m Radioimmunoassay (RIA) with antigen 02 S. mansoni egg fraction was used in the RIA, as described by Pelley et al. (4). k Biomat, Hazebrouck, France. Institut Pasteur Production, Paris. m From Pharmacia Fine Chemicals AB, Box 175, Uppsala, Sweden. 752 K. E. MOTT & H. DIXON INDIRECT HAEMAGGLUTINATION (IHA) IHA procedure with antigens 05 and 06 Crude S. japonicum and S. mansoni soluble egg antigens and a fractionated soluble S. mansoni egg antigen preparation were used following the pro- cedure of Colley et al. (27), modified as described below. Sheep red blood cells (SRBC) were washed 3 times in PBS, pH 7.2. The cells were incubated overnight at 4 °C in 10 ml/litre glutaraldehyde prepared in PBS, and then washed 5 times with PBS. A 1:2000 dilution of 10 ml/litre tannic acid in PBS was then added and allowed to incubate for 15 min at 4 IC. The cells were washed 3 times in PBS, and then re- suspended in 20 times their own volume of PBS con- taining SEA at a concentration of 400 mg/litre. The cells were incubated at room temperature for 30 min. The SEA was then removed by washing the cells 3 times in PBS containing 5 ml/litre heat-inacti- vated fetal calf serum (PBS with FCS). The cells were stored in PBS with FCS, at a concentration of 100 ml/litre and diluted to a concentration of 2.5 ml/litre for use. To each of 21 wells of a U-bottomed microtitration plate,", 50 yd of PBS with FCS was added. To well 1, 50 1l of test serum was added; serial 2-fold dilutions were placed in wells 2-18; wells 19-21 received no serum. All 21 wells then received 50 ul of a solution of 2.5 ml/litre SEA-SRBC in PBS with FCS. The plates were incubated overnight and read the fol- lowing morning. The titre was taken as the reciprocal of the final dilution of serum showing haemaggluti- nation. IHA procedure with antigens 09 and 10 Crude saline-extracted S. mansoni adult worm pre- paration and purified S. mansoni glycoprotein frac- tion were used as follows: A 25 ml/litre solution of formalin-stabilized sheep red cells was collected by centrifugation. The super- natant was removed and the antigen solution (0.5 ml in 0.1 mol/litre glycol buffer, pH 8.2) and 10 ml/litre glutaraldehyde (0.3 ml in glycol buffer) added. The concentration of the adult worm antigen used in this sensitization procedure was 100 g per litre of saline solution (1 g/litre). The cells were incubated for 45 min at 56 'C, washed 4 times, and mixed with a solution of human albumin in glycol buffer (2 ml/litre) to a final concen- tration of 12.5 ml of sheep red blood cells per litre. All test sera were inactivated at 56 'C, for 30 min and absorbed with sheep red blood cells (50 yd of ' Linbro Chemical Co., New Haven, CT, USA. serum and 50 Al of red blood cells) for 30 min at room temperature. After centrifugation, the absorbed serum was withdrawn. Positive and negative reference sera were treated in the same manner. Two-fold serum dilutions (50 pl/well) were placed in a microtitration plate and 20 i1 of sensitized red blood cells were added; after incubation for 2 h at 4 °C, the wells were evacuated. The results were expressed as the reciprocal of the highest serum dilution showing a positive reaction. Under these conditions the sera tested were considered positive if the titre was greater than or equal to 80, equivocal if the titre was between 40 and 80, and negative if it was less than 40. INDIRECT FLUORESCENT ANTIBODY TEST (IFA) IFA procedure with antigens 14 and 17 After preparation of frozen parasite S. mansoni sections, the following procedure was used, as de- scribed by Huldt et al. (28). A polyvalent sheep anti-human IgG conjugate was used.' This was prepared as described by Bergquist & Schilling (29) and characterized as described by Fagraeus & Bergquist (30). The total protein equalled 5.7 g/litre, of which 1.3 g was directed against IgG. Molar F/P = 3.4. The optimal conjugate dilution was determined to be 1:30, by chequerboard titration, as described by Huldt et al. (28). The test results were analysed using a Leitz Dialux 20 fluorescence microscope adapted for substrate illumination with dark-field condenser and a KP490/ K510 filter combination. Two distinct fluorescence patterns could be observed on the worm sections: (1) staining of the cells lining the parasite gut (S. mansoni gut-associated antigens), and (2) staining of the worm parenchyma (S. mansoni somatic antigens). The titres reported represent the end-point for distinct fluorescent staining using a 2-fold serial dilution of the test serum. IFA procedure with antigen 13 This test was carried out according to the technique described by Nash (21), Deelder et al. (23), and Deelder & Kornelis (22). Male S. mansoni worms were fixed in Rossman's fixative and embedded in paraplast. Sections of 5 pm thickness were prepared and stored at room tempera- ture. Before use, the sections were rehydrated by placing the microscope slides in xylol for 1 hour at 37 °C, in xylol at room temperature for 15 min, and ° Prepared at the National Bacteriological Laboratory, Stock- holm, Sweden. IMMUNODIAGNOSIS OF SCHISTOSOMIASIS 753 by subsequent passage through a series of mixtures of ethanol and distilled water, at concentrations of 1000/, 800/o, 50'0o, and 307o ethanol. The slides were then dried and each section was circled with a waterproof marking pen. Aliquots of 10 1l of 2-fold serum dilutions in PBS (0.05 mol/litre, pH 7.8) were layered over the sections, starting with a dilution of 1:10, and the slides were incubated in a moist chamber for 45 min at 37 'C. They were then washed twice for 5 min in PBS and dried, and 10 Al of conjugate, diluted 1:50 in PBS with 1:10 000 Evans Blue, was added. The conjugate used was a swine anti-human IgM (Fc) labelled with fluor- escein isothiocyanate (FITC).P The slides were incu- bated for 45 min at 37 'C, washed twice for 5 min in PBS, and dried. They were then covered with glycerin-PBS (9:1) and a cover-glass and inspected under a Leitz MPV11 microfluorescence microscope equipped with a Ploemopak illuminator and filter combination I for FITC-fluorescence. A Leitz objec- tive ( x 25/0.55) and oculars (4 x ) were used. The reciprocal value of the last serum dilution showing a distinct fluorescence of the cells lining the gut was considered as the titre. Positive and negative control sera were also tested. Negative control sera showed a maximum titre of 10; a titre of 20 was con- sidered as equivocal. QUALITATIVE TESTS Circumovalprecipitin test (COPT) with antigen 01 This test was performed as described by Oliver- Gonzalez (31). COPT reactions were recorded as a percentage of reactive eggs and according to the type of activity (32). Indium slide immunoassay (ISI) with antigen 15 Using CEF-6 S. mansoni egg antigen, this test was run as described by Giaever (14) according to the following protocol, designed to detect approximately 10- 2 g/litre of antibody in the undiluted sera. The indium slides were cut into six pieces and placed in a wetbox (a plastic box containing a wet paper towel), and the antigens were applied. These were nominally diluted to approximately 50 mg/litre with physiological saline and applied in small drops of less than 1 Al to the slides using a 25-,l capillary tube. The slides were incubated in the wetbox with the antigen drops for at least 15 min, rinsed in tap water, and placed in a 1 g/litre BSA solution for a few minutes. The slides were taken directly from the BSA solu- P Nordic Immunologic Laboratories, Tilburg, Netherlands. tion and put into the serum to be tested, and incubated for 1 hour. The serum was diluted 1:10 with a Tris buffer, pH 8.2, to a final volume of 0.5 ml. For sera num- bered greater than 96, the dilution was 1: 12, to a final volume of 0.6 ml, and at the same time 1 g/litre of BSA was diluted into the buffer. This was because it was found that 0.5 ml sometimes did not cover the slides and that a few sera seemed to have antibody to BSA. After incubation the slides were removed from the serum, rinsed in running tap water, and blown dry with compressed air. The slides were read by eye and mounted on a plastic sheet for later study. Microdouble diffusion (MDD) Crude saline-extracted S. mansoni adult antigen and purified glycoprotein fraction S. mansoni antigen were used according to the method described by Ouch- terlony (33). In a water bath, 1 g of agaroseq was dissolved in 100 ml of Veronal buffer, pH 8.1, ionic strength 0.1, containing 160 g of sodium barbital, and 220 ml of 1 mol/litre HCI; the volume was made up to 10 litres with distilled water. The agarose solution was heated for 20 min per 100 ml of solution at a temperature not greater than 90 'C. Using a 20-ml pipette, 11.5 ml ofwarm agarose was distributed on an 8 x 8 cm glass slide which had been cleaned thoroughly with methanol and dried. The agarose was allowed to harden for 5 min and then placed in a humid chamber for 30 min. Sixty-four perforations, 4 mm in diameter, were made for the sera to be tested; 16 perforations, 1 mm in diameter, were made for the antigen. Each antigen well was surrounded equidistantly by 4 wells of sera. Aliquots of 15 1l of serum were pipetted into each serum well. The antigen wells were filled with 0.125 mg of antigen using a Pasteur pipette. The anti- gen solution was prepared by adding 2 ng of antigen to 50 ul of distilled water. The reagents were allowed to diffuse for 48 h at 4 'C in a humid chamber. The diffusion was termi- nated by covering with 50 g/litre sodium citrate solu- tion. The slides were washed in 8.5 g/litre NaCl solution for 48 h. They were then dried 4 times with filter paper at room temperature, and stained with amido black (524 ml of 13.6 g/litre sodium acetate, 425 ml of 6 ml/litre glacial acetic acid, and 1 g of amido black). The slides were decolourized by washing 3-4 times in 80 ml/litre glacial acetic acid. The presence of precipitin arcs indicated a positive reaction. q Indubiose A37, Industrie Biologique Franqaise, Gennevilliers, France.