312 J Clin Pathol 1998;51:312–315 Do IgA, IgE, and IgG avidity tests have any value in the diagnosis of toxoplasma infection in J Clin Pathol: first published as 10.1136/jcp.51.4.312 on 1 April 1998. Downloaded from pregnancy?

D Ashburn, AWLJoss, T H Pennington, D O Ho-Yen

Abstract infection is very small when maternal infection Aim—To determine the value of tests for occurs before conception, it is important to specific IgA, IgE, and IgG avidity in diag- make a precise estimation of the timing of nosing Toxoplasma gondii infection dur- maternal infection.2 Serological diagnosis of ing pregnancy. toxoplasma infection is usually based on Methods—In a retrospective study, current fourfold increases in IgG titre and demonstra- serological tests (dye test and three IgM tion of specific IgM.3 As rising titres are seldom assays with diVerent sensitivities) were demonstrated outside an antenatal screening compared with immunosorbent agglutina- programme,4 detection of specific IgM is often tion assays (ISAGA) for specific IgA and used to indicate acute infection. However, IgE and an IgG avidity enzyme linked detection of IgM may be protracted,3 and immunosorbent assay (ELISA). Patient therefore interpretation of results is dependent group 1 comprised six women with definite on the sensitivity of the IgM technique,56and or probable infection during pregnancy with the use of current tests it may not be pos- determined by congenital toxoplasmosis sible to determine the onset of maternal infec- or laboratory results. Group 2 comprised tion accurately. seven women infected during or before 11 In order to overcome these diYculties several pregnancies (two consecutive pregnancies new techniques have been developed. Immu- in two patients and three in a third). noglobulin A (IgA) is produced as part of the Results—One patient in group 1 serocon- acute phase response, and has been found to be verted during pregnancy. IgA ISAGA and of value in the investigation of cytomegalovirus avidity confirmed acute infection when infection.7 Toxoplasma specific IgA has been confirmatory IgM ELISA remained nega- measured both by enzyme linked immunosorb- tive. In five of six patients from group 1, ent assay (ELISA)48and immunosorbent agglu- IgA and IgE ISAGA and avidity confirmed tination assay (ISAGA),489 in the sera of acute infection. In group 2, the dye test patients with acquired primary infection. Fol- titre was raised in seven of 11 pregnancies lowing infection it is claimed that specific IgA http://jcp.bmj.com/ (six of seven patients). Specific IgM and production parallels that of specific IgM,10 or IgA were positive during all 11 pregnan- lags slightly behind IgM.9 Toxoplasma specific cies. IgE ISAGA was positive in only four IgE is also detected early after infection.11–13 of 11 pregnancies (three of seven pa- Results suggest that the duration of detectable tients), but negative results in the remain- specific IgE may be shorter than IgM or IgA.11–13 der may exclude acute infection. High While measurement of specific IgA and IgE may

avidity antibodies indicative of past infec- prove useful in improving the assessment of tim- on September 28, 2021 by guest. Protected copyright. tion were found in four of 11 pregnancies ing of infection, there is wide patient to patient (two of seven patients). variation in their detection.814 Measurement of Conclusions—Each test improved diagno- IgG avidity represents a diVerent approach to sis or timing of infection but no single test antibody studies, assessing the interaction be- was ideal. The IgA ISAGA was sensitive tween antibody and antigen.15 In early infection Scottish Toxoplasma and detected seroconversion. Positive IgE low avidity antibodies predominate, whereas Reference Laboratory, 16 Microbiology ISAGA and low avidity both confirmed later high avidity antibodies are more common. Department, infection, whereas negative IgE may ex- Measurement of IgG avidity has been applied to Raigmore Hospital clude acute infection. High avidity diag- several viral infections,15 and to toxoplasma NHS Trust, Inverness nosed past infection but persistence of low infection,15 17 18 in which it has been used to time IV2 3UJ, UK avidity reduced its value to diVerentiate maternal infection during pregnancy.19 D Ashburn AWLJoss acute and past infection. Further studies In this study, to determine whether newer D O Ho-Yen with larger patient groups are needed to techniques have any value in the diagnosis of determine the optimum diagnostic strat- toxoplasma infection during pregnancy, IgA Medical Microbiology egy. These techniques are valuable in and IgE ISAGAs and an IgG avidity assay Department, complementing existing tests. based on an in-house IgG ELISA were used. University of Aberdeen (J Clin Pathol 1998;51:312–315) The results were compared with those obtained Medical School, using current tests (dye test, two diVerent IgM Foresterhill, Aberdeen, Keywords: immunoglobulin avidity; toxoplasmosis; UK pregnancy ELISAs with diVerent sensitivity, and an IgM T H Pennington ISAGA).

Correspondence to: When a mother acquires primary Toxoplasma Methods Dr D Ashburn. gondii infection during pregnancy, congenital In a retrospective study, toxoplasma antibodies Accepted for publication infection may result, with severe consequences were measured in 43 sera from 17 pregnancies in 5 February 1998 for the fetus.1 However, as the risk of fetal 13 pregnant women. Sera had been tested using IgA, IgE, and IgG avidity tests in toxoplasma diagnosis 313

current tests: in-house IgG20 and IgM21 ELISAs, human IgE (å chain specific; 2.4 µg/ml) for the 22

and a micromodification of the dye test. Toxo- IgA and IgE ISAGA, respectively. Patients’ sera J Clin Pathol: first published as 10.1136/jcp.51.4.312 on 1 April 1998. Downloaded from plasma specific IgM positive results were con- (100 µl) diluted 1/100 in phosphate buVered firmed using a commercial assay, Toxonostika saline (PBS), pH 7.2, were added to triplicate ELISA-IgM (Organon Technika, Cambridge, wells. After incubation for two hours at 37°C UK). Those that were negative in the screening wells were washed with PBS plus 0.05% Tween test or not confirmed were also tested using the 20 (pH 7.2) (PBST) and then with PBS alone. more sensitive Toxo-ISAGA IgM (BioMerieux, Formaldehyde fixed tachyzoites (1.25 × 107/ Marcy l’Etoile, France). Patients were divided ml) were added to each of the triplicate (100, into two groups, determined by the presence of 150, and 200 µl) wells and incubated overnight congenital toxoplasmosis and the results of cur- at 37°C. Results were scored 0–4 in each well rent serological tests. relative to negative (no agglutination) and Group 1 consisted of six patients with positive (agglutination) controls. The scores of definite or probable infection during preg- triplicate wells were accumulated to produce nancy. The outcome of pregnancy was con- the ISAGA matrix (maximum 12). A matrix of genital toxoplasmosis in three cases. In two of < 2 was negative, 3–4 equivocal, 5–9 positive, the three cases there were rising dye test titres, and > 10 strong positive. raised titres (> 250 IU/ml) in the third, and IgG avidity was measured using the in-house IgM was confirmed positive in all. Two other IgG ELISA under conditions similar to a pub- cases were included on the basis of raised dye lished method.19 Flat bottomed microtitre test titres and confirmed IgM. One pregnancy plates (Virion B, Shield Diagnostics, Dundee, had a normal outcome, but was included UK) were coated overnight with toxoplasma because the mother had cervical lymphaden- antigen optimally diluted in a 10% solution of opathy, a raised dye test titre, and specific IgM ELISA coating buVer (Don Whitley Scientific, detectable in the latter stages of pregnancy. Shipley, W Yorkshire,UK). After aspirating the This combination of clinical and laboratory plates, they were dried at 37°C for four hours data indicates infection during pregnancy. The and stored at −20°C until use. Patients’ sera second pregnancy ended in fetal death which were diluted with PBST in fourfold steps to may have been caused by toxoplasma, but no produce five dilutions: 1/50, 1/200, 1/800, tissues were available for confirmation. The 1/3200, 1/12 800. The four highest dilutions remaining patient seroconverted during preg- (1/200–1/12 800) were added (100 µl per well) nancy, but again the outcome was normal. In to rows A–D and the four lowest to rows E–H. this patient IgM was not confirmed by Toxon- After incubation at room temperature for one ostika ELISA IgM, but was positive by the hour, the plates were washed as follows: rows more sensitive Toxo-ISAGA. A–D, 3 × 5 minutes with PBST; rows E–H 3 × Group 2 comprised seven women with past 5 minutes with urea wash solution (6 M urea in or possible acute infection during 11 pregnan- PBST). All wells were then washed a further

cies (two consecutive pregnancies in two three times with PBST. Conjugate— http://jcp.bmj.com/ women, and three in a third). These were fur- antihuman IgG–alkaline phosphatase (Sigma, ther subdivided into groups 2a and 2b. Poole, Dorset, UK)—was added (100 µl per Group 2a comprised women with only a sin- well) at optimal dilution. After incubation for gle pregnancy or the first of two or three one hour at room temperature, plates were consecutive pregnancies. Raised dye test titres washed with PBST and substrate added were measured in all four single pregnancies (Sigma phosphatase substrate) diluted in

and during the first pregnancy in two other ELISA substrate buVer (Don Whitley Scien- on September 28, 2021 by guest. Protected copyright. women. Toxonostika ELISA IgM was also tific). After 30 minutes, the optical density was positive in two of four single pregnancies and in measured at 405 nm using a Dynatech all three first pregnancies, but toxoplasma spe- MR-5000 plate reader (Dynex Technologies). cific IgM was detected by one or more of the Dilution curves were plotted for the urea methods used in all group 2a pregnancies. washed and PBST washed wells and avidity Specific IgM was detected early in all four sin- calculated as the ratio of titres measured gle pregnancies and—in the absence of symp- (urea:PBST) at the threshold of 0.2 absorb- toms or rising dye test titres—it was therefore ance units. Low, borderline, and high avidity not possible to determine whether infection was < 15%, 15–30%, and > 30% respectively. occurred before or after conception. Group 2b included the second pregnancy in Results three women and the third in one of these. Group 1 comprised six women with definite or Although the dye test titre was normal in the first probable evidence of T gondii infection during specimens taken during each of the consecutive pregnancy. In one patient, who seroconverted pregnancies, the Toxonostika ELISA-M was during pregnancy, toxoplasma specific IgM positive. Thus group 2 shows that similar was detected only by the BAM-ELISA and laboratory results may be obtained in both past Toxo-ISAGA. Positive IgA ISAGA results and infection and possible acute infection. low avidity IgG were also measured, but IgE All sera were tested using in-house immuno- ISAGA failed to diagnose acute infection in sorbent agglutination assays for specific IgA23 this patient. In the other five patients from and IgE.12 Briefly, round bottomed microtitre group 1, positive results in the IgA ISAGA, IgE plates (Immulon 1, Dynex Technologies, Bill- ISAGA, and low avidity IgG (table 1) con- ingshurst, UK) were coated with capture anti- firmed a diagnosis of acute infection. body (Dako, Copenhagen, Denmark), anti- In group 2a and 2b, the IgA ISAGA was human IgA (á chain specific; 2 µg/ml) or anti- positive during all pregnancies (table 1). The 314 Ashburn, Joss, Pennington, et al

Table 1 Serological results obtained during 17 pregnancies J Clin Pathol: first published as 10.1136/jcp.51.4.312 on 1 April 1998. Downloaded from Positive results in:

Raised dye test titre Toxonostika Group Patient No (>250 IU/ml) BAM-ELISA ELISA IgM Toxo-ISAGA M IgA IgE IgG avidity* 1 1+ ++ND++L 2+ + + ++ 3+ ++ND++L 4+ ++ND++L 5+ ++ND++L 6+ +−++−L

2 (a) Single/first 7 + + + ND + + B pregnancy 8 + + + ND + − H 9+ +−++−L 10+ −−++−H 11− ++ND+−B 12+ −+ND++L 13+ ++ND++H (b) Consecutive 11 − + + ND + − L pregnancies 12 − −+++−L − −−++−L 13− ++ND++H

Results shown are those in the first sample or post-seroconversion sample tested. +, positive; −, negative; ND, not done; *L, low (<15%); B, borderline (15–30%); H, high (>30%). IgE ISAGA was positive in only four of 11 previously.425 Our results of patients followed pregnancies from three of seven patients; three through consecutive pregnancies (group 2) of these were during the first or a single confirm that, like specific IgM, specific IgA pregnancy (group 2a) and the fourth from a positive results may persist for prolonged peri- consecutive pregnancy (group 2b). Negative ods. Therefore the IgA ISAGA was not able to results could possibly have been used to diVerentiate between acute and past infection. exclude acute infection in four of seven and This may reflect the enhanced sensitivity of three of four group 2a and 2b pregnancies, ISAGA over ELISA methods48; use of a less respectively. Avidity measurement was less sensitive IgA ELISA may be more appropriate useful, as high avidity antibodies were meas- to improve timing of infection and should be ured in only three of seven group 2a patients investigated. and in one of four group 2b pregnancies. Fur- In other studies,11 14 detection of specific IgE thermore IgG avidity did not mature from low was associated with toxoplasma infection to high avidity in two patients between their during pregnancy in 33 of 37 women. In our first pregnancy in group 2a and their second or study, toxoplasma specific IgE was detected in third in group 2b. In contrast to group 1 five of six patients from group 1, but not in the http://jcp.bmj.com/ patients, simultaneous positive results for IgM, patient who seroconverted, confirming the IgA, IgE, and IgG antibodies of low avidity results of the Toxonostika ELISA IgM. In our were measured in only one group 2a patient study, there was no evidence of congenital during her first pregnancy. These results toxoplasmosis in this patient and therefore suggest that infection occurred during this negative IgE ISAGA results may indicate a pregnancy. Therefore if all of the results from favourable outcome of infection during preg- all tests were used, a correct timing of the nancy. Further testing of more patients is nec- infection could be achieved. essary to confirm the prognostic value of a on September 28, 2021 by guest. Protected copyright. negative IgE ISAGA result. Specific IgE was Discussion not detected in eight of 11 pregnancies from Protracted detection of raised levels of specific five of seven patients from group 2. It has been IgM makes timing infection in pregnancy suggested that specific IgE remains detectable diYcult.1 Diagnosis of infection during preg- for less than four months after infection.11 nancy is therefore often based on comparison Therefore, negative results may be useful to of results from two or three IgM tests.3 In this exclude acute infection. However, because IgE laboratory, the less sensitive Toxonostika assays may lack sensitivity,11 12 results of the IgE ELISA IgM assay is most useful in the confir- ISAGA should be interpreted in conjunction mation of onset of infection within a period with results of other tests. < 1–2 trimesters.6 Patients were divided into Measurement of IgG avidity has previously two groups according to the likelihood of been applied to determination of duration of infection. Thus it was possible to evaluate the infection,18 19 26 and to identifying pregnancies three relatively new tests on their ability to at risk of congenital toxoplasmosis.19 One improve on the existing IgM based strategy. study19 included 11 women who seroconverted It has been claimed that specific IgA is during pregnancy. All had low avidity antibod- detected early after infection but also disap- ies, which in 10 of 11 women matured to high pears rapidly.24 25 In our study specific IgA was avidity within six months. The remaining detected in the first post-seroconversion sam- patient still produced low avidity antibodies ple in one patient, confirming its rapid appear- after 11 months. This is diVerent from the sec- ance after infection. However, the IgA ISAGA ond study in which the mean avidity result was was also positive in all other pregnancies from low both in sera collected less than 20 weeks patients in both groups. Persistence of raised and between 20 to 52 weeks after infection.26 levels of specific IgA has been reported The discrepancy between these studies is IgA, IgE, and IgG avidity tests in toxoplasma diagnosis 315

probably because in the first,19 results were pregnancy, although no single test was ideal.

correlated with the time elapsed since the last Therefore the value of specific IgA, IgE, or IgG J Clin Pathol: first published as 10.1136/jcp.51.4.312 on 1 April 1998. Downloaded from seronegative sample compared with the esti- avidity tests is to augment rather than replace mated time since infection.26 Nevertheless, in the current tests. both studies high avidity was measured in past infection. In our study, all patients from group We are grateful to the Scottish Home and Health Department 1, including one who seroconverted, had low (grant No K/MRS/50/C1828) and the European Network on avidity antibodies. Low avidity antibodies were Congenital Toxoplasmosis, working group on serological diagno- also measured in five of 11 pregnancies from sis (Biomed 1 programme PL921572 of the commission of the European Union), for funding. We would also like to thank Ms three of seven patients in group 2. As three of Debbie Gilham for secretarial assistance, and the users of the these were from group 2b, and infection was Scottish Toxoplasma Reference Laboratory for their support. therefore before pregnancy (second pregnan- cies of two patients and the third pregnancy of 1 Remington JS, Desmonts G. Toxoplasmosis. In: Remington another), the value of low avidity antibody JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant. Philadelphia: WB Saunders, 1990:89–195. measurement to confirm acute infection may 2 Desmonts G, Couvreur J. Toxoplasmosis in pregnancy and be less than previously suggested.19 26 Further- its transmission to the fetus. Bull N Y Acad Med 1974;50:146–59. more persistence of low avidity antibody, like 3 Joss AWL. Diagnosis. In: Ho-Yen DO, Joss AWL, eds. persistent IgM, may complicate diagnosis. Human toxoplasmosis. Oxford: Oxford University Press, These findings may indicate variation in 1992:79–118. 17 4 Patel B, Young Y, DuVyK,et al. Immunoglobulin-A detec- immune response. Persistent borderline avid- tion and the investigation of clinical toxoplasmosis. JMed ity antibody has also been reported in non- Microbiol 1993;38:286–92. 5 Skinner LJ, Chatterton JMW, Joss AWL, et al. The use of an pregnant patients with lymphadenopathy IgM immunosorbent agglutination assay to diagnose present for over six months.18 High avidity congenital toxoplasmosis. J Med Microbiol 1989;28:125–8. 6 Ashburn D, Evans R, Skinner LJ, et al. Comparison of rela- antibodies in four pregnancies, three of which tive uses of commercial assays for Toxoplasma gondii IgM were confirmed as IgM positive, were used to antibodies. J Clin Pathol 1992;45:483–6. diagnose past infection. Thus the IgG avidity 7 Crange MP, Kennedy D, Venters JL, et al. IgA responses during human cytomegalovirus infection in cardiac trans- ELISA has proved less useful in our studies plant recipients: concurrent detection of IgA and IgM anti- than previously described.19 26 This may be globulins. Serodiagn Immunother Infect Dis 1988;2:301–9. 8 Francis JM, Joynson DH. Duration of specific immu- attributed to the study size or to the patients noglobulin A antibody following acute toxoplasmosis as tested. Our inclusion of sera collected from determined by enzyme immunoassay and immunosorbent agglutination assay. Eur J Clin Microbiol Infect Dis 1993;12: three women with persistent IgM during 556–9. consecutive pregnancies may have introduced a 9 Bessières MH, Roques C, Berrebi A, et al. IgA antibody 19 26 response during acquired and congenital toxoplasmosis. J sampling bias absent from other studies. Clin Pathol 1992;45:605–8. Alternatively, the discrepancies may have been 10 Wong SY, Remington JS. Toxoplasmosis in pregnancy. Clin a result of diVerence between assay procedures, Infect Dis 1994;18:853–61. 11 Pinon JM, Toubas D, Marx C, et al. Detection of specific although this was not indicated by our results immunoglobulin E in patients with toxoplasmosis. J Clin in non-pregnant patients or comparison with a Microbiol 1990;28:1739–43. 27 12 Ashburn D, Joss AW, Pennington TH, et al. Specificity and commercial kit. usefulness of an IgE immunosorbent agglutination assay http://jcp.bmj.com/ The three women with persistent IgM for toxoplasmosis. J Clin Pathol 1995;48:64–9. 13 Montoya JG, Remington JS. Studies on the serodiagnosis of during consecutive pregnancies may represent toxoplasmic lymphadenitis. Clin Infect Dis 1995;20:781–9. 19 a group not previously studied. How fre- 14 Wong SY, Hajdu MP, Ramirez R, et al. Role of specific quently they occur is not known. However, immunoglobulin E in diagnosis of acute toxoplasma infec- tion and toxoplasmosis. J Clin Microbiol 1993;31:2952–9. they are an important group in which to assess 15 Hedman K, Lappalainen M, Söderlund M, et al. Avidity of the ability of relatively new tests to differentiate IgG in serodiagnosis of infectious diseases. Rev Med Micro 1993;4:123–9.

between acute and past infection. All the 16 Holliman RE. Recent developments in the diagnosis of toxo- on September 28, 2021 by guest. Protected copyright. consecutive pregnancies in group 2b were plasmosis. Serodiagn Immunother Infect Dis 1994;6:5–16. 17 Holliman RE, Raymond R, Renton N, et al. The diagnosis known to have had past infection, yet neither of toxoplasmosis using IgG avidity. Epidemiol Infect specific IgA, IgE, nor IgG avidity results 1994;112:399–408. 18 Joynson DHM, Payne RA, Rawal BK. Potential role of IgG suggested past infection in every case. There- avidity for diagnosing toxoplasmosis. J Clin Pathol 1990;43: fore, like specific IgM tests, these have to be 1032–3. 19 Lappalainen M, Koskela P, Koskiniemi M, et al. Toxoplas- interpreted with caution during pregnancy. mosis acquired during pregnancy: improved serodiagnosis In conclusion, the IgA ISAGA is sensitive based on avidity of IgG. J Infect Dis 1993;167:691–7. and can be used to detect seroconversion. 20 Joss AWL, Skinner LJ, Chatterton JMW, et al. Toxoplasmosis: eVectiveness of enzyme immunoassay Positive IgE ISAGA results and low avidity screening. Med Lab Sci 1989;46:107–12. antibodies were both able to confirm acute 21 Joss AWL, Skinner LJ, Moir IL, et al. Biotin-labelled antigen screening test for toxoplasma IgM antibody. J Clin Pathol infection. The absence of specific IgE in 1989;42:206–9. patients from group 2 may have been useful to 22 Williams KAB, Scott JM, Macfarlane DE, et al. Congenital toxoplasmosis: a prospective survey in the west of Scotland. exclude acute infection. Measurement of high J Infect 1981;3:219–29. avidity antibodies was also used to diagnose 23 Ashburn D, Joss AW, Ho-Yen D, et al. The value of toxoplasma specific IgA in diagnosis. J Clin Pathol 1995;48: past infection. However, persistence of low 689. avidity antibodies meant that the IgG avidity 24 Decoster A, Darcy F, Caron A, et al. Anti–P30 IgA antibod- ELISA was a less useful discriminator of ies as prenatal markers of congenital toxoplasma infection. Clin Exp Immunol 1992;87:310–15. current from past infection than previously 25 Foudrinier F, Marx-Chemla C, Aubert D, et al. Value of suggested. More conclusive interpretation of specific immunoglobulin A detection by two immunocap- ture assays in the diagnosis of toxoplasmosis. EurJClin the data may be limited by the small number of Microbiol Infect Dis 1995;14:585–90. patients tested, and further studies with a larger 26 Jenum PA, Stray-Pedersen B, Gundersen AG. Improved diagnosis of primary Toxoplasma gondii infection in early population are necessary to determine the pregnancy by determination of antitoxoplasma immu- optimum diagnostic strategy. However, each of noglobulin G avidity. J Clin Microbiol 1997;35:1972–7. 27 Ashburn D. The relevance of IgA and IgE assays, IgG avidity these techniques provided additional infor- and western blotting in the diagnosis of toxoplasma infection. mation to diagnose and time infection during Aberdeen: University of Aberdeen, 1996. (PhD thesis.)