238 J Clin Pathol 1991;44:238-242 Automated microparticle enzyme immunoassays for IgG and IgM antibodies to Toxoplasma gondii J Clin Pathol: first published as 10.1136/jcp.44.3.238 on 1 March 1991. Downloaded from J W Safford, G G Abbott, M C Craine, R G MacDonald Increased awareness of the potentially Abstract devastating effects of toxoplasmosis and the Fully automated microparticle enzyme possibility that serological screening will gain immunoassays (MEIA) for the IMx wider acceptance8 demand the availability of immunoassay analyser were developed easily performed, accurate, and reliable to detect IgG and IgM antibodies to serological assays. We report the development Toxoplasma gondii. The IgG MEIA of two fully automated microparticle enzyme results are expressed in International immunoassays (MEIA), performed on the Units (IU) of IgG antibody interpolated IMx immunoassay analyser,9 for the detection from a six point calibration curve cover- of IgG and IgM antibodies to Tgondii. ing the range from 0 to 300 IUIml. Reproducible results were obtained from a calibration curve stored in the TOXO IGG MEIA instrument for at least one month. The qualitative IgM MEIA expresses results Polystyrene microparticles are coated with as an index using a single calibrator antigens extracted from the Rh strain of T included in each run. The Toxo IgG gondii'0 propagated in human foreskin MEIA and Toxo IgM MEIA were in 98% cells. When incubated with a patient's serum, and 97% agreement, respectively, with T gondii specific antibody reacts with the the reference assays used. Twenty four antigen coated microparticle solid phase. The sera can be completely processed in incubation mixture is transferred to a glass about 35 minutes. fibre matrix which captures the microparticles, allowing unbound serum components to be washed into a blotter located below the Infection with the obligate parasite Toxo- matrix. Affinity purified goat anti-human IgG plasma gondii is common throughout the (Kirkegaard and Perry, Gaithersburg, world. Fortunately, most infections range Maryland), coupled to alkaline phosphatase from asymptomatic to mild symptoms.' There (Boehringer and Mannheim, Indianapolis, http://jcp.bmj.com/ are, however, several conditions in which Indiana), followed by methylumbelliferyl acquired infection or reactivation of latent phosphate (MUP) is used to detect the infections can be life threatening, such as presence of IgG antibody bound to the pregnancy, after transplantation, and AIDS. Tgondii antigens on the solid phase. The rate Serological test results in each of these condi- of fluorescence development as MUP is tions, however, can provide important infor- dephosphorylated to methylumbelliferone is mation for improved patient management. measured by the front-surface fluorometer on September 24, 2021 by guest. Protected copyright. Infection during pregnancy often results in within the IMx.9 The Toxo IgG MEIA is fetal damage, causing a stillbirth or birth of an quantitative; results are interpolated from a infant with congenital toxoplasmosis, the six point calibration curve referenced to the effects of which may not be apparent, for World Health Organisation International months or even years.2 Mandatory prenatal Standard For anti-Toxoplasma serum. The serological screening in France has shown that six calibrators contain 0, 10, 50, 75, 150 and timely diagnosis by seroconversion, rising 300 IU IgG antibody/ml. IgG antibody titres, or the presence of IgM antibody3 followed by prenatal treatment can alleviate the effects of congenital toxo- Reproducibility of calibration curve after Infectious Diseases plasmosis.4 Tgondii antibody negative, im- storage and Immunology, munosuppressed recipients of heart, or heart Calibration curve storage would avoid the Diagnostic Division, and lung, transplants from antibody positive necessity of running a complete set of Abbott Laboratories, North Chicago, Illinois donors can acquire potentially fatal primary calibrators in every run; therefore a 60064, USA Tgondii infections. Serological test results can reproducibility study was performed to deter- J W Safford identify the susceptible patients, thus allowing mine if calibrator values could be stored G G Abbott M C Craine appropriate treatment to be started.5 Reactiva- within the software and used in subsequent R G MacDonald tion of latent Tgondii infection in patients runs. Two control sera and a four member Correspondence to: with AIDS is one of the most common causes panel were selected to represent each of the J W Safford of central nervous system complications in five line segments in the six point calibration Abbott Laboratories Building L-3, Departnent this group.6 Although serological test results curve. Calibrators, controls, and panels were 90D, 14th and Sheridan are not reliable indicators of acute toxoplas- tested nine times in duplicate over 35 days. Road, North Chicago, Illinois 60064, USA mosis in patients with AIDS,6 the absence of To compensate for minor day to day varia- Accepted for publication IgG antibody may rule out a diagnosis of tions, the stored calibration curve was normal- 31 October 1990 toxoplasma encephalitis.67 ised by an adjustment factor calculated for Automated IgG and IgM antibody assays 239 each run by dividing the value of a 50 IU/ml Reproducibility calibrator included in each run by the stored Inter- and intra-assay reproducibility were value of the 50 IU/ml calibrator. The IU/ml determined by performing 10 runs containing of the controls and panels were calculated by the index calibrator, a negative control, and two J Clin Pathol: first published as 10.1136/jcp.44.3.238 on 1 March 1991. Downloaded from point to point interpolation from the curve levels ofpositive sera in duplicate over 21 days. stored on the first day of the study after The index ofeach test sample was calculated by normalisation by multiplying the stored dividing the sample value by the value of a calibrator values by the adjustment factor. To single replicate of the index calibrator. determine if accurate adjustment could be obtained using only a single 50 IU/ml calibrator on each carousel, the adjustment Sensitivity and specificity factor was calculated using the first replicate The effectiveness of the HGG coated micro- of that calibrator instead of using the mean of particles in removing rheumatoid factor was duplicate values. tested by creating a series of rheumatoid factor positive sera containing increasing titres ofIgG antibody to Tgondii in the absence of IgM anti- Specificity and sensitivity Tgondii antibody. Dilutions of a serum highly Four hundred and twenty sera obtained from positive for IgG antibody (1500 IU/ml) were clinical laboratories in the United States and made in a negative serum. One volume of each France were tested by the Toxo IgG MEIA. of the dilutions was mixed with three volumes Results of less than 6 IU/ml were interpreted of a rheumatoid factor positive serum (titre 1/ as negative and those of 6 IU/ml or greater as 1280) resulting in a series of sera containing positive. The relative sensitivity and from 1 to 269 IU/ml IgG antibody to Tgondii specificity of the MEIA were determined by in the presence of rheumatoid factor. Each of comparison with results obtained with IgG the sera were tested by both the Toxo IgG and specific reference assays. Toxo IgM MEIA, as described previously. The primary reference used in the com- Four hundred and ninety five sera obtained parative study was the enzyme immunoassay from clinical laboratories in the United States, (EIA) Toxo-G EIA (Abbott Laboratories, France, Germany, Italy and Spain were tested North Chicago). The results of Toxo-G EIA by the Toxo IgM MEIA. Sera with an index of were expressed as an index derived by divid- 0-5 or greater were treated to adsorb any ing the test serum result by the result of the rheumatoid factor and retested as described low positive control; results of less than 0 5 previously. Results from the IgM MEIA were were interpreted as negative; results of 0-5 or compared with those obtained by the reference greater were interpreted as positive. Dis- assay Toxo-M EIA (Abbott Laboratories, cordant results between the Toxo IgG MEIA North Chicago). A Toxo-M EIA index of less and Toxo-G EIA were resolved with than 0-4 is negative, an index of0-5 or greater is the Enzygnost Toxoplasmosis/IgG assay positive, and the intermediate values (04- http://jcp.bmj.com/ (Behringwerke AG, Marburg, Germany). 0 499) are equivocal. Those sera giving discor- dant results between the MEIA and the reference EIA were tested by an immunosor- bent agglutination assay (ISAGA) specific for TOXO IGM MICROPARTICLE ENZYME IgM antibody to T gondii (Toxo-ISAGA, IMMUNOASSAY bioMerieux, France). The Toxo IgM MEIA is similar to the IgG on September 24, 2021 by guest. Protected copyright. MEIA in many respects, but with three prin- cipal differences. First, IgM antibody to IgG and IgM seroconversion Tgondiiwas detected using affinity purified goat Sera obtained from patients with serological anti-human IgM (Kirkegaard and Perry, evidence of primary exposure to Tgondii were Gaithersburg, Maryland) coupled to alkaline tested by both the Toxo IgG MEIA and Toxo phosphatase. Secondly, the results are IgM MEIA. Sera and Sabin-Feldman dye test qualitative and are expressed as an index results" were kindly supplied by Professor derived by dividing the patient value by the Thulliez, Institut de Puericulture de Paris, value of an index calibrator included in each Laboratorie de Serologie Neonatale, Paris, run. Finally, serum samples with an index of France. 0-5 or greater were adsorbed with microparti- cles coated with human gamma globulin Inter- and intra-assay reproducibility were (HGG) to remove rheumatoid factor. Samples calculated by one way analysis ofvariance using with an index of less than 0 5 are regarded as Statgraphics software package (STSC, Inc, negative without further testing. To adsorb the Rockville, Maryland). reactive samples, serum and microparticles were mixed and incubated at room temperature for five minutes, followed by five to 10 minutes Results centrifugation at 10 000 x g.