Received 13 July 1994/Returned for modification 22 September 1994/Accepted 6 January 1995

Transcription

1 JOURNAL OF CLINICAL MICROBIOLOGY, Apr. 1995, p Vol. 33, No /95/$ Copyright 1995, American Society for Microbiology Evaluation of Risk and Diagnostic Value of Quantitative Assays for Anti-Toxoplasma gondii Immunoglobulin A (IgA), IgE, and IgM and Analytical Study of Specific IgG in Immunodeficient Patients JEAN MICHEL PINON, 1 * FREDERIQUE FOUDRINIER, 1 GENEVIEVE MOUGEOT, 2 CATHY MARX, 1 DOMINIQUE AUBERT, 1 OLIVIER TOUPANCE, 3 GERARD NIEL, 4 MARTIN DANIS, 4 PATRICK CAMERLYNCK, 5 GERARD REMY, 6 JACQUES FROTTIER, 7 DAMIEN JOLLY, 8 MARIE H. BESSIERES, 9 DOMINIQUE RICHARD-LENOBLE, 10 AND ANNIE BONHOMME 1 Laboratoire de Parasitologie, Eq 5 Institut National de la Santé et de la Recherche Médicale (Unité 314), 1 and Service de Nephrologie, 3 Centre Hospitalier Universitaire Hôpital Maison Blanche, Service des Maladies Infectieuses, Centre Hospitalier Universitaire Robert Debré, 6 and Departement d Information Medicale, IRF, Reims, Laboratoire de Parasitologie 2 and Service des Maladies Infectieuses, 7 Centre Hospitalier Universitaire Saint-Antoine, Paris, Laboratoire de Parasitologie, Centre Hospitalier Universitaire Pitié-Salpêtrière, Paris, 4 Laboratoire de Parasitologie, Centre Hospitalier Universitaire Hôpital du Bocage, Dijon, 5 Laboratoire de Parasitologie, Centre Hospitalier Universitaire Rangueil, Toulouse, 9 and Laboratoire de Parasitologie, Centre Hospitalier Universitaire, Tours, 10 France Received 13 July 1994/Returned for modification 22 September 1994/Accepted 6 January 1995 To determine their prognostic and diagnostic values for toxoplasmosis in immunodepressed subjects, we assayed immunoglobulin A (IgA) and IgE antibodies by means of immunocapture (IC) tests, with revelation done by using a suspension of T. gondii (ICT). We also carried out a simultaneous analytical study of IgG antibodies on cellulose acetate membranes by using the comparative immunological profile method and an enzyme-linked immunofiltration assay (ELIFA). A total of 1,238 samples (serum, cerebrospinal fluid, and aqueous humor from 318 patients) were tested. IgA and IgE antibodies were detected in all heart, kidney, and liver transplant recipients with clinical manifestations of toxoplasmosis; IgA was detected in the aqueous humor of a patient with chorioretinitis. In patients with AIDS-related toxoplasmosis, including the cerebral form, IgA and IgE antibodies or a significant modification of ELIFA IgG values were observed in 38, 19, and 25% of patients, respectively. IgM was detected by ICT only in 12% of patients and aided the diagnosis in 1 of 71 patients. IC tests for specific IgA and IgE alone and combined with ELIFA were in 39 and 46% of patients who developed clinical toxoplasmosis, respectively. In a serial study of 16 patients in whom at least one of these three tests was, a significant immunological signal sometimes preceded clinical onset by 1, 6, and even 17 months. Similarly, in a group of human immunodeficiency virus-infected patients with evidence of previous exposure to T. gondii but no clinical manifestations, IgA, IgE, and IgA and/or IgE antibodies were detected in only 11, 4, and 12% of patients, respectively. These two situations point to peripheral T. gondii reactivation. IgA and IgE emerged as interesting markers of the risk of toxoplasmosis in immunodepressed patients. They may also provide valuable assistance in the diagnosis of toxoplasmosis, especially because tests for specific IgM are disappointing. However, at least one in two patients with toxoplasmosis showed no detectable immunological reaction, suggesting that this polyisotypic approach should be combined with other noninvasive methods such as gene amplification. Toxoplasmosis is a serious complication of organ transplantation (23, 32) and AIDS (1, 12), resulting from primary infection or, more often, reactivation. Patients develop a variety of local or systemic forms (11, 12, 21, 32). The diagnosis is usually based on a combination of epidemiological, clinical, and radiological arguments, with laboratory tests playing a relatively minor role. In practice, diagnostic confirmation is based on a favorable clinical and/or radiological response to specific therapy. Pathological examinations of biopsy specimens by direct examination, culture, and molecular biology methods are rarely used (3, 10, 13). Furthermore, classical studies of specific immunoglobulin G (IgG) and IgM antibodies are disappointing (12, 34). However, immunocapture (IC) assays of specific IgA or IgE * Corresponding author. Mailing address: Département de Parasitologie, Hôpital Maison-Blanche, 45 rue Cognacq Jay, Reims, France. Phone: Fax: antibodies (Abs) and serial tests for specific IgG by enzymelinked immunofiltration assay (ELIFA) have been of value in the case of maternal-fetal toxoplasmosis (19, 26 28). In particular, detection of these isotypes can point to acute toxoplasmosis in pregnant women. We evaluated the same approach in immunodeficient patients, with two main aims. The first was to identify markers of toxoplasmosis correlating with the short- or mediumterm risk of the infection developing in immunodeficient patients. The second was to obtain immunological evidence to confirm the clinical diagnosis of toxoplasmosis. Because prophylaxis is now routinely prescribed, the present study was principally based on samples obtained in several centers between 1985 and 1990 and tested retrospectively in the same laboratory (Reims). MATERIALS AND METHODS Patients and samples. The samples (n 1,238) were obtained from 318 patients (ages, 18 to 55 years) who fell into three groups (Table 1). Group I (76 878

2 VOL. 33, 1995 Ig ISOTYPE MARKERS OF RISKS AND COMPLICATIONS 879 Group TABLE 1. Patient groups and characteristics of 318 immunodeficient patients No. of patients No. of samples Characteristic I Organ graft recipients A Kidney graft B 3 20 Heart transplant C 2 6 Liver graft II Initially toxoplasmosis-free, HIVinfected patients A With anti-t. gondii IgG B Without anti-t. gondii IgG III AIDS patients with clinical toxoplasmosis A Sampled before, during, and/or after onset of toxoplasmosis B Sampled only during and after onset of toxoplasmosis Total 318 1,238 organ graft recipients) was subdivided into group IA (71 kidney graft recipients; 284 serum samples), group IB (3 heart transplant patients; 20 serum samples), and group IC (2 liver graft recipients; 5 serum samples, 1 aqueous humor sample). Group II consisted of 171 initially toxoplasmosis-free human immunodeficiency virus (HIV)-infected patients with or without AIDS (546 samples) and was subdivided into group IIA (139 patients with anti-toxoplasma gondii IgG; 458 serum samples, 24 cerebrospinal fluid [CSF] samples) and group IIB (32 patients without anti-t. gondii IgG; 64 serum samples). Group III consisted of 71 AIDS patients (382 samples) with toxoplasmosis and was subdivided into group IIIA (38 patients who underwent serial sampling before, during, or after the onset of toxoplasmosis; 33 patients with encephalitis, 2 patients with chorioretinitis, 1 patient with pericarditis, and 2 patients with generalized toxoplasmosis; 228 serum samples, 16 CSF samples) and group (33 patients from whom samples were obtained only during and after the onset of toxoplasmosis; 31 patients with encephalitis, 1 patient with pulmonary toxoplasmosis, and 1 patient with generalized toxoplasmosis; 115 serum samples, 23 CSF samples). The diagnosis of toxoplasmosis, based on clinical, radiological, and/or laboratory findings (microscopic examination), was validated by a favorable outcome on specific therapy. IC tests. Quantitative IC assays for specific IgA, IgE, and IgM used a suspension of T. gondii for revelation (ICT) under standard conditions which were identical for the three isotypes (ICT-A, ICT-E, and ICT-M, respectively). As described elsewhere in detail (27, 28), microtiter plates (96-well Nunc Microwell Polysorp plates) were coated with monoclonal anti-, anti-ε, or anti- antibodies (MAbs; MAb CHAFI I.4, MAb ANAB16, and MAb Tibi 82; Argene; Biosoft, Varilhes, France). Each sample was distributed in three adjacent wells. After incubation and washing, 100, 150, or 200 l of a tachyzoite suspension (trypsin treated and formalin fixed; density, /ml) was added to one of the three sample wells. For each well, a score of 0 was given for total tachyzoite sedimentation; complete agglutination was given a score of 4. The cumulative score for the three wells could thus vary from 0 to 12. Given a lack of sufficient volume (aqueous humor), some samples were only tested in one well (score, 0 to 4). The reactions were read on a multipurpose spectrophotometer (Dynatech MR ) with specific software developed in our laboratory (Orkys, Reims, France) for automated reading, calculation, and transcription of the results. Scores of 2 or more were considered for sera in the ICT-A and ICT-E. Given the frequent interference with natural IgM, the ICT-M cutoff value was 6. Specific IgG was measured by means of high-sensitivity direct agglutination (HSDA) (37). The results are expressed in units per milliliter. Readings and calculations (equivalence between the last dilution, titer, or units) were automated, as with the ICT assays. Some samples were checked by an enzyme-linked immunosorbent assay (ELISA): Toxo (A, M, G) Ab EIA (Clonatec, Paris, France), or IMX Toxo, IgA, IgM, IgG (Abbott). ELIFA. When several samples of sufficient volume were available from the same patient, they were tested comparatively with each other by using the ELIFA comparative immunological profile method (CIP-ELIFA) (26). Briefly, this is a two-step procedure. The first consists of immunoelectrodiffusion on a micropore cellulose acetate membrane on which the soluble Toxoplasma antigen and three tested serum samples migrate by electrophoresis and electroendosmosis, respectively. The precipitation arcs are revealed by filtration of a solution of enzymecoupled anti-, anti-, or anti- antibody through the membrane. This method can be used to compare (by coalescence) the specificity and concentration changes of precipitating antibodies in the three samples. In this study we only characterized IgG. The Toxoplasma antigen with an approximate molecular mass of 68 to 70 kda (P70) was identified by MAb anti-p70 and then by a human reference serum recognizing this antigen. This antigen is the main component involved in ELIFA precipitation reactions in acquired or congenital toxoplasmic reactivation (25 28) with an IgG antibody increase. Statistical analysis. All comparisons of percentages were performed by the McNemar test for paired values between tests. Comparisons between groups were performed by Pearson s chi-squared test. RESULTS Specific IgA, IgE, IgM, and IgG in organ graft recipients (group I). Among the 71 renal graft recipients in group IA, 16 (25 serum samples) had no specific IgG, IgA, or IgE Ab; ICT-M was for 4 patients (values, 6); serial studies of the corresponding patients confirmed that these were natural IgM antibodies without seroconversion to the Toxoplasma antigen. In the remaining patients (n 55) there were three seroconversions, with IgA, IgE, and IgM positivity followed by the emergence of IgG Ab. One of these cases clearly corresponded to a primary infection, with the patient having a fever, altered general condition, and favorable outcome on pyrimethamine-sulfadiazine therapy. The other two seroconversions were subclinical, and the patients did not receive specific therapy. Among the other 52 patients in group IA, HSDA revealed a significant elevation ( 2 titers) of specific IgG Ab titers in eight patients (15%), with (i) no specific IgM, IgA, or IgE Ab response (one patient), (ii) an IgA Ab response (five patients, 9.5%) plus an IgM Ab response in two patients; and (iii) an IgM Ab response but no IgA Ab response (two cases). No clinical manifestations occurred in these patients, despite the absence of preventive therapy. In five other patients there was no variation in IgG Ab, but IgM Ab was detectable (above the cutoff in two patients). Apart from these three seroconversions, anti-t. gondii IgE was never detected in group IA. In group IB, high levels of specific IgA and IgE Abs were detected in the three heart transplant patients (scores, 9 to 12), and were associated with high levels of IgM Ab in two patients (scores, 11 and 12). These corresponded to two seroconversions and one immunological reactivation (increase in IgG or emergence of IgM, IgA, or IgE Abs); all of the patients developed severe clinical manifestations. In one of the two seroconverters a myocardial biopsy specimen revealed the presence of T. gondii. All three patients responded to treatment, but nonetheless retained very high specific IgA and IgE Ab levels for more than 7 months, an unusual observation. In group IC, high IgA and IgE Ab levels (7 to 12) were found with IgM Ab in both patients. The first patient developed classical toxoplasmosis with systemic manifestations 2 months after liver grafting. The second patient developed chorioretinitis a year after liver grafting while on immunosuppressive drugs; the coefficient of IgG load in the aqueous humor was highly significant (score, 5.7), with the presence in situ of IgA (value, 4/4) and IgM (value, 4/4) Ab but no IgE Ab. Specific IgA, IgE, IgM, and IgG in apparently toxoplasmosis-free HIV-infected subjects with or without AIDS (group II). Among the 139 patients in group IIA, 17 showed a significant increase (2 titers or more in the HSDA test) in IgG Ab titers in serum, while 10 had IgM, 19 had IgA, and 9 had IgE Abs. In fact, one of these patients had a typical seroconversion and in two patients toxoplasmic encephalitis was diagnosed retrospectively; in two other patients there was no clear correspondence between laboratory and clinical findings (mediastinal adenopathies in one patient and inguinal adenopathies in another, both of whom were lost to follow-up). Thus, apart from the

3 880 PINON ET AL. J. CLIN. MICROBIOL. Group TABLE 2. Positive results of CIP-ELIFA and IgA, IgE, and IgM ICT tests for HIV-infected or AIDS patients Total CIP-ELIFA variations IgA-ICT IgE-ICT No. (%) of patients IgM-ICT IgA-ICT and/or IgE- ICT IgA-ICT, and/or IgE-ICT, and/or IgM-ICT IgA-ICT, and/or IgE-ICT, and/or CIP-ELIFA IIA 136 a 16 a (11.70) 6 a (4.40) 7 a (5.10) 17 a (12.50) 17 a (12.50) IIIA (31.50) 11 (28.90) 7 (18.40) 5 (13.15) 12 (31.50) 13 (34.00) 16 (42.00) 33 6 (18.00) 16 b (48.50) 7 c (21.20) 4 (12.00) 16 (48.50) 16 (48.50) 17 (51.50) IIIA and ) (25.30) 27 (38.00) 14 (19.70) 9 (12.60) 28 (39.40) 29 (40.80) 33 (46.40) a Includes patients without seroconversion and two patients with encephalitis. b Including six patients at the cutoff. c Including one patient at the cutoff. patient who seroconverted and the two patients with encephalitis, HSDA revealed a significant increase in specific IgG Ab titers in 10.2% of patients; IgM, IgA, IgE, and IgA and/or IgE Abs were detected in 5.1, 11.7, 4.4, and 12.5% of patients, respectively (Table 2). Associations between the different immunoglobulin isotypes and their relationship with clinical manifestations and anti-pneumocystis carinii therapy, are given in Table 3. Among the 32 patients in group IIB (64 serum samples) in whom specific IgG Ab was never detected, 2 had IgM scores (score, 6) and 7 others had ICT-M values between 1 and 5.5. In a sample from one patient IgA Ab was present at the cutoff value (score, 2). In none of the patients in this group was IgE detectable in serum. TABLE 3. Association between specific IgG, IgA, IgE, and IgM and link with clinical manifestations or anti-p. carinii therapy in initially toxoplasmosis-free, HIV-infected patients with or without AIDS (group IIA) a Patient no. Manifestation or treatment IgG Ab (HSDA) b IgA Ab IgE Ab IgM Ab 1 Seroconversion 1 2 Inguinal nodes 1 3 None 1 4 None 1 5 None 1 6 None 1 7 None 1 8 Mediastinal nodes 1 9 None 1 10 None 1 11 None 3 12 None 3 13 None 3 14 None 1 15 None 3 16 None 3 17 Toxoplasmic 1 Encephalitis 18 Sulfamethoxazole 1 Trimethoprim Pentamidine 19 None 1 20 Toxoplasmic encephalitis 1 21 Sulfamethoxazoletrimethoprim 1 22 Pentamidine 1 a Specific IgG variation, n 17; for specific isotypes: n 19 for IgA, n 9 for IgE, and n 10 for IgM. b 1 and 3, increase and stability of specific IgG, respectively. Specific IgA, IgE, IgM, and IgG and CIP-ELIFA in AIDS patients from whom serial samples were taken before, during, or after the onset of toxoplasmosis (group IIIA). Group IIIA (38 patients) was subdivided into two groups according to the presence (group IIIA1) or absence (group IIIA2) of IgA, IgE, or IgM Ab or a significant variation in specific IgG in HSDA or CIP-ELIFA. In group IIIA1 (16 patients, 112 serum samples, 7 CSF samples), 12 patients had toxoplasmic encephalitis, 2 had chorioretinitis, 1 had disseminated toxoplasmosis, and 1 had pericarditis. An increase in specific IgG Ab, especially that recognizing antigen P70, was observed in 12 patients by CIP-ELIFA, with a significant increase in HSDA for only 9 patients (23% of the patients in group IIIA). Specific IgA and IgE Abs were detected in 11 and 7 patients, respectively, while IgM was above the cutoff for only 5 patients. In two patients with chorioretinitis, IgA and IgE Ab titers were particularly elevated (scores, 12). Twelve patients were by ICT-A and/or ICT-E. Sixteen patients had specific IgA and/or IgE and/or a significant variation in CIP-ELIFA; compared with the IgA or IgE tests, IgM Abs were of diagnostic value in only 1 of these 16 patients (Table 2). Relative to the date of onset of clinical toxoplasmosis, immunological markers were only contemporary in 2 patients; in 14 patients they preceded clinical onset by 1 month (n 3), 2 months (n 5), 3 months (n 1), 4 months (n 2), 6 months (n 2), and 17 months (n 1) (mean time, 3.6 months) (Table 4). Group IIIA2 (22 patients, 116 serum samples, 9 CSF samples) comprised 21 patients with of encephalitis and 1 with generalized toxoplasmosis. There were no immunological signs of reactivation, and anti-t. gondii IgG Ab was no longer detectable in one patient. It should be noted that the follow-up of these patients was very irregular, sometimes preceding clinical onset by less than 15 days and sometimes by periods of 4 months or more (up to 14 months) without immunological tests. Specific IgA, IgE, IgM, and IgG and CIP-ELIFA in AIDS patients studied during and after clinical onset of toxoplasmosis (group ). In group, which was composed of 33 patients, 17 (group 1) had at least one immunological sign of reactivation (elevation of IgG Ab titers by HSDA or CIP- ELIFA and/or the presence of specific IgA, IgE, and/or IgM Abs) (Table 5). In this group, 16 patients had toxoplasmic encephalitis and 1 patient had generalized toxoplasmosis. A significant variation in IgG Ab titers was observed by CIP- ELIFA in six patients and by HSDA in four patients (18 and 12%, respectively). IgM Ab titers were above the cutoff in 4 patients, while IgE and IgA Ab titers were in 7 and 16 patients, respectively (including 10 patients with ICT-A values of between 4 and 12 and 6 patients with ICT-A values at the

4 VOL. 33, 1995 Ig ISOTYPE MARKERS OF RISKS AND COMPLICATIONS 881 Patient no. TABLE 4. Specific IgG, IgA, IgE, and IgM and CIP-ELIFA in 16 AIDS patients sampled before or during and after clinical onset (group IIIA1) a IgG Ab (HSDA) b CIP-ELIFA IgA Ab IgE Ab IgM Ab Immunological anticipation/toxoplasmic complications ,5 mo/encephalitis mo/encephalitis mo/chorioretinitis mo/encephalitis Contemporary/encephalitis mo/encephalitis mo/encephalitis Contemporary/encephalitis mo/encephalitis mo/toxoplasmosis, generalized mo/encephalitis wk/pericarditis mo/encephalitis mo/encephalitis mo/encephalitis mo/encephalitis a Specific IgG increases, n 9 patients for IgG Ab (HSDA) and n 12 patients for CIP-ELIFA; immunocapture test results were found for 11 patients for IgA Ab, 7 patients for IgE Ab, and 5 patients for IgM Ab. b 1 and 3, increase and stability of specific IgG, respectively. cutoff). Simultaneous testing for IgM Abs provided no supplementary diagnostic evidence relative to IgA and IgE (Table 2). In the other 16 patients (group 2) no immunological signs were detected by any of the four methods, while 15 subjects had toxoplasmic encephalitis and 1 had pulmonary toxoplasmosis. Comparison of the different sensitivities and predictive values of laboratory diagnostic strategies in toxoplasmosis-free, HIV-infected patients (group II) and AIDS patients with toxoplasmosis (group III). The results of a comparison of the different sensitivities of laboratory diagnostic strategies in toxoplasmosis-free, HIV-infected patients (group II) and AIDS patients with toxoplasmosis (group III) are listed in Table 6. TABLE 5. Specific IgG, IgA, IgE, and IgM and CIP-ELIFA in 17 patients sampled during and after clinical onset (group 1) a Patient no. IgG Ab (HSDA) b CIP-ELIFA IgA Ab IgE Ab IgM Ab Toxoplasmosis c Encephalitis c Encephalitis c Encephalitis Encephalitis Encephalitis Encephalitis c Encephalitis c Encephalitis c Encephalitis c Encephalitis Encephalitis Encephalitis Encephalitis Encephalitis Encephalitis Encephalitis Generalized a AC Ab increases, n 4 patients for IgG Ab (HSDA) and n 6 patients for CIP-ELIFA; immunocapture test results were found for 16 patients for IgA Ab, including 6 patients at the cutoff; 7 patients for IgE Ab, including 1 patient at the cutoff; and 4 patients for IgM Ab. b 1, 3, and 2, increase, stability, and regression of specific IgG, respectively. c Positive at cutoff. The predictive values of a result by the ICT-A, ICT-E, ICT-M, ICT-A and ICT-E, ICT-A and/or ICT-E, and ICT-A and/or ICT-E and/or ICT-M tests were 0.614, 0.700, 0.500, 0.750, 0.609, and 0.591, respectively. CSF samples and IgA. We tested for IgA Ab in the CSF of 63 patients and detected IgA Ab in the CSF of 12 of them. Seven of these patients had toxoplasmic encephalitis (group III), while five had a negative diagnosis of toxoplasmosis (group II). The sample volume was inadequate to carry out additional tests, including tests of Ig load. DISCUSSION The diagnosis of toxoplasmosis remains difficult in immunodeficient patients (21). The use of noninvasive immunological tests has so far been very disappointing (12, 44). The potential value of specific IgA detection by ICT, first described in 1986 (27), has since been validated for use in the prenatal and neonatal diagnosis of toxoplasmosis (2, 6, 34). This isotype is also useful in the case of toxoplasmosis during pregnancy, in which it is a useful marker of recent infection on the basis of its kinetics and more rapid disappearance than IgM in ICT on the one hand and the absence of natural IgA on the other hand (24, 27). Specific IgE shows an even more rapid time course and, together with IgA, can be used for the specific dating of seroconversion (28), to assess the risk of maternal-fetal transmission, and to determine whether prenatal diagnosis is indicated. In graft recipients and HIV-infected patients, our results show that the detection of anti-t. gondii IgA or IgE by ICT and of IgG by CIP-ELIFA provides new information on the noninvasive management and diagnosis of toxoplasmosis. In the case of patients with liver and heart transplants, IgA or IgE detection was always in concordance with clinical toxoplasmosis after seroconversion or reactivation (28). These arguments were confirmed by the detection of T. gondii in situ and/or the response to treatment. In patients with chorioretinitis, local specific IgA and IgM synthesis is a major diagnostic criterion (11). Although the IgA or IgM load could not be determined owing to an inadequate sample volume, passive transfer of IgA or IgM can be ruled out by the absence of IgE in the aqueous humor, despite its presence in large amounts in

5 882 PINON ET AL. J. CLIN. MICROBIOL. TABLE 6. Comparison of sensitivity of laboratory diagnostic strategies in HIV-infected or AIDS patients a of ICT-A and/ or ICT-E and/or CIP- ELIFA of ICT-A and/or ICT-E and/ or ICT-M of ICT-A and/or ICT-E of ICT-M of ICT-E of ICT-A of CIP- ELIFA Group (no. of patients) IIA (136) 0.11 P P NS 0.12 P P 10 6 IIIA (38) NS 0.32 NS 0.28 NS 0.18 NS 0.13 NS 0.31 NS 0.34 NS 0.42 (33) III (71) P b NS NS a, comparisons for each test between groups IIIA and ; III, comparisons for each test between groups IIA and III; NS, not significant. b Comparisons between strategies and classical ICT-M. this patient s serum (28). Local synthesis of IgA Ab has recently been confirmed in other immunocompetent subjects with toxoplasmic chorioretinitis; in one patient tachyzoites were identified in the aqueous humor (22, 38). In renal graft recipients, specific IgA and IgE were simultaneously detected in three patients during seroconversion; the seroconversion in one patient was associated with clinical manifestations and the patient was treated successfully. The fact that no clinical manifestations occurred in the absence of specific treatment in the other two patients appears to confirm the relatively low risk of toxoplasmosis in kidney transplant patients (31) with adequate CD4 -cell counts (5). However, the possibility of this risk must not be overlooked (23). In AIDS patients with patent toxoplasmosis (group III), specific IgA, IgE, or IgG Ab synthesis was detected by CIP- ELIFA in 38, 19, and 25% of the patients, respectively. The combinations of ICT-A-E or ICT-A-E and CIP-ELIFA were valuable, with at least one test being for 39 and 46% of patients, respectively (Tables 2 and 6). These yields were far better than that of IgM (12%), detection of which was only of diagnostic value in 1 of 71 patients with encephalitis or patent toxoplasmosis. In addition, the significance of these IgM titers is unclear, because natural IgM has been detected above the cutoff value in serial samples from seronegative subjects. Interestingly, we observed seroreversion of IgG (HSDA, IFI 10 U/ml) in a patient with disseminated toxoplasmosis. So this situation, while exceptional (29, 44), is compatible with toxoplasmosis in patients with clinical signs, even when specific IgG is undetectable. The present study confirms the value of specific IgE, which has previously been detected in about 25% of subjects with toxoplasmic chorioretinitis or encephalitis (28, 43). In contrast, rates of detection of IgA are controversial, since they range between 4 and 38% in studies with large numbers of patients (4, 34, 36, 44). This may result from a number of factors. The use of tests based on different principles (ICT, ELISA, immunoblotting) is one such factor. Another is the revelation of IgA antibody activity by selective antigens (P30) or antigens with broad specificities (6, 9). The variety of soluble antigen preparations (reduced or nonreduced by sodium dodecyl sulfate) leads to variable results (14, 15, 37). The strong concentration of sera (1/25) sometimes used in immunoblotting adds to the confusion, with detection of natural IgA not being described previously (36). In addition, even with apparently identical techniques (ICT), the conditions (objective or optical) in which the results are read determines the choice of cutoff value; other factors are the source of T. gondii (commercial or laboratorymade), the origins of the RH strains, and tachyzoite pretreatment (trypsin, formalin, or acetone) (27, 28, 36, 37, 44). Regarding these last two factors, large differences have been found in the ultrastructural aspects of T. gondii and antigens recognized by specific IgA by transmission electron microscopy. In the absence of prior isotype separation, isotype competition phenomena were observed in ultrastructural immunohistochemistry, which might also be encountered in immunoblotting (18, 36). Despite all of these explanations it remains clear that more than one in two cases of clinical toxoplasmosis has no clear immunological corollary. In contrast, in patients in groups I and II, IgA Ab (9.5 and 11%, respectively), IgE Ab (0 and 4%, respectively), or an elevation in specific IgG Ab titers were observed, despite a lack of clinical manifestations. It is interesting that in the study of serial samples (group IIIA), these markers preceded clinical onset by 1 to 17 months, even if transiently (2 months for IgA, less than 4 weeks for IgE). Thus, the sampling intervals and dates relative to clinical onset can,

6 VOL. 33, 1995 Ig ISOTYPE MARKERS OF RISKS AND COMPLICATIONS 883 at least in part, be responsible for the diversity of results. Our experience on the immunological signal anticipation relative to the clinical onset of infection is similar to that of other reports on IgA (44) and both presumptive (7, 33) and confirmatory tests (10, 39) for Toxoplasma parasitemia. The detection of peripheral lymphocytes spontaneously secreting anti-p30 or anti-p68 antibodies (41) 5 to 15 months before the onset of encephalitis supports our observations. It should be noted that the IgG detected by CIP-ELIFA during immunological reactivation also recognizes antigens P68-70 located in rhoptria, which have been implicated in cell penetration by tachyzoites (18). Identical CIP-ELIFA results are observed in patients with congenital toxoplasmosis during reactivation after withdrawal of the cystostatic combination pyrimethamine-sulfadoxine (25). The pathogenic mechanism of toxoplasmic encephalitis usually leads to local cerebral cyst rupture (8). However, it seems logical that secondary cerebral and ocular localization (11) could derive from peripheral cyst reactivation via the bloodstream (with specific IgA, IgE, or IgG synthesis). Multiorgan involvement detected at necropsy may be interpreted as involvement at secondary sites but also as a peripheral starting point (35). It is also possible that this peripheral pathogenic route is modified by treatment (both anti-t. gondii and anti-p. carinii), because some of the drugs used against the latter are also active against the former (17, 20, 40). The morphological and histopathological patterns of brain lesions appear to have changed since 1990 with the use of new curative and prophylactic treatment strategies in patients with these two infections (35). This aspect should be taken into account in the comparison of our results, because most of our samples were obtained previously, between 1985 and The reasons for the disappointing anti-toxoplasma isotype detection rate involve immunological changes related to the HIV infection itself (1, 12, 17, 21). It should, however, be noted that other specific IgA Abs can be synthesized in patients with this syndrome; indeed, anti-hiv IgA is detected in two-thirds of children congenitally infected with the virus (42). Finally, the immunomodulatory effect of T. gondii itself, the use of inhibitory drugs, and the prolonged survival of AIDS patients may also play a role (5, 16). The present study of specific IgA or IgE by ICT and of IgG by CIP-ELIFA in organ graft recipients and HIV-infected patients provides new information. Even if the immunological criteria provided by a combination of these three isotype tests are still inadequate for diagnostic purposes, they are clearly more promising than the classical tests with the IgG-IgM combination (Tables 2 and 6; see also predictive values in Results section). In the case of clear clinical manifestations, detection of these Abs supports the diagnosis by revealing ongoing toxoplasmosis. The emergence of anti-t. gondii IgA or IgE several months before clinical onset makes them potential markers of a high risk of toxoplasmosis, especially in the case of patients with HIV infection (Table 4). Because they are sometimes transient, tests for these isotypes should be done every 2 months. In contrast to strictly cerebral reactivations, their synthesis could be induced during peripheral cyst activation, which would lead to secondary dissemination via the bloodstream. This last pathogenic mechanism would be sensitive to prophylactic and curative therapy (anti-t. gondii and anti-p. carinii), which is now given routinely. The inadequate diagnostic yield of the peripheral humoral immune response means that this approach should be combined with tests for local antibody synthesis or other noninvasive diagnostic means (21, 30). Gene amplification and PCR have opened up exciting new testing possibilities, but different results have been reported according to the target gene (P30 or B1), primers, and samples (venous blood, CSF, or brain biopsy) used (3, 7, 13, 33). The diagnostic and prognostic significance of a signal is not fully clear, because the detection of amplified microbial DNA does not necessarily reflect the presence of a living organism or an active infection. ACKNOWLEDGMENTS Financial support was provided by the Programme Hospitalier de Recherche Clinique (1994), Direction des Hôpitaux, Ministère des Affaires Sociale de la Santé et de la Ville, Paris, France. REFERENCES 1. Aurojo, F. G., and J. S. Remington Toxoplasmosis in immunocompromised patients. Eur. J. Clin. Microbiol. 6: Bessieres, M. H., C. Roques, A. Berrebi, V. Varre, M. Cazaux, and J. P. Seguela IgA antibody response during acquired and congenital toxoplasmosis. J. Clin. Pathol. 45: Burg, J. L., C. M. Grover, P. Pouletty, and J. C. Boothroyd Direct and sensitive detection of a pathogenic protozoan, Toxoplasma gondii, by polymerase chain reaction. J. Clin. Microbiol. 27: Darcy, F., F. Foudrinier, G. Mougeot, A. Decoster, A. Caron, C. Marx- Chemla, A. Capron, and J. M. Pinon Diagnostic value of specific IgA antibodies in AIDS patients with Toxoplasma infection: a bicentric evaluation. Immunol. Lett. 30: Darcy, F., and F. Santoro Toxoplasmosis, p In Parasitic infections and the immune system. Academic Press, Inc., New York. 6. Decoster, A., A. Caron, F. Darcy, and A. Capron IgA antibodies against P30 as markers of congenital and acute toxoplasmosis. Lancet ii: Dupouy-Camet, J., S. L. de Souza, C. Maslo, A. Paugam, A. G. Saimot, R. Benarous, C. Tourte-Schaefer, and F. Derouin Detection of Toxoplasma gondii in venous blood from AIDS patients by polymerase chain reaction. J. Clin. Microbiol. 31: Frenkel, J. K., and A. Escajadillo Cyst rupture as a pathogenic mechanism of toxoplasmic encephalitis. Am. J. Trop. Med. Hyg. 36: Hassl, A., and H. Aspock Antigens of Toxoplasma gondii recognized by sera of AIDS patients before, during and after clinically important infections. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. 272: Hoffin, J. M., and J. S. Remington Tissue culture isolation of Toxoplasma from blood of a patient with AIDS. Arch. Intern. Med. 145: Holland, G. N Ocular toxoplasmosis in the immunocompromised host. Int. Ophthalmol. 13: Holliman, R. E Toxoplasmosis and the acquired immune deficiency syndrome. J. Infect. 16: Holliman, R. E., J. D. Johnson, and D. Savva Diagnosis of cerebral toxoplasmosis in association with AIDS using the polymerase chain reaction. Scand. J. Infect. Dis. 22: Huskinson, J., P. N. Stepick-Biek, F. G. Araujo, P. Thulliez, Y. Suzuki, and J. S. Remington Toxoplasma antigens recognized by immunoglobulin G subclasses during acute and chronic infection. J. Clin. Microbiol. 27: Huskinson, J., P. Thulliez, and J. S. Remington Toxoplasma antigens recognized by human immunoglobulin A antibodies. J. Clin. Microbiol. 28: Israelski, D. M., C. Tom, and J. S. Remington Zidovudine antogonizes the action of pyrimethamine in experimental infection with Toxoplasma gondii. Antimicrob. Agents Chemother. 33: Koppen, S., T. Grünenwald, G. Jautzke, J. Gottschalk, H. D. Pohle, and B. Ruf Prevention of Pneumocystis carinii pneumonia and toxoplasmic encephalitis in human immunodeficiency virus infected patients: a clinical approach comparing aerosolized pentamidine and pyrimethamine/sulfadoxine. Clin. Invest. 70: Kumolosasi, E Thesis. University of Reims Champagne-Ardenne, Reims, France. 19. Le Fichoux, Y., P. Marty, and H. Chan Les IgA sériques spécifiques dans le diagnostic de la toxoplasmose. Ann. Pediatr. 34: Lindsay, D. S., B. L. Blagburn, J. E. Hall, and R. Tidwell Activity of pentamidine and pentamidine analogs against Toxoplasma gondii in cell cultures. Antimicrob. Agents Chemother. 35: Luft, B. J., and J. S. Remington Toxoplasmic encephalitis. J. Infect. Dis. 157: Marx-Chemla, C., E. Thelliez, F. Foudrinier, D. Aubert, A. Bonhomme, A. Ducasse, and J. M. Pinon Toxoplasma chorioretinitis asserted by detection of free tachyzoites and specific IgM in the aqueous humor. Presse Med. 22: Mejia, G., E. Leiderman, M. Builes, J. Henao, M. Arbelaez, J. L. Arango, and J. Borrero Transmission of toxoplasmosis by renal transplant. Am. J. Kidney Dis. 2:

7 884 PINON ET AL. J. CLIN. MICROBIOL. 24. Mirlesse, V., F. J. Jacquemard, and F. Daffos Toxoplasmose au cours de la grossesse. Diagnostic et nouvelles possibilités thérapeutiques. Presse Med. 22: Pinon, J. M., J. Poirriez, B. Leroux, D. Dupouy, C. Quereux, and J. P. Garin Early diagnosis and supervision of congenital toxoplasmosis. Presse Med. 16: Pinon, J. M., H. Thoannes, and N. Gruson An enzyme-linked immuno-filtration assay used to compare infant and maternal antibody profiles in toxoplasmosis. J. Immunol. Methods. 77: Pinon, J. M., H. Thoannes, P. Pouletty, J. Poirriez, J. Damiens, and P. Pelletier Detection of IgA specific for toxoplasmosis in serum and cerebrospinal fluid using a nonenzymatic IgA capture assay. Diagn. Immunol. 4: Pinon, J. M., D. Toubas, C. Marx, G. Mougeot, A. Bonnin, A. Bonhomme, M. Villaume, F. Foudrinier, and H. Lepan Detection of specific immunoglobulin E in patients with toxoplasmosis. J. Clin. Microbiol. 28: Porter, S. B., and M. A. Sande Toxoplasmosis of the central nervous system in the acquired immunodeficiency syndrome. N. Engl. J. Med. 23: Potasman, I., L. Resnick, J. B. Luft, and J. S. Remington Intrathecal production of antibodies against Toxoplasma gondii in patients with toxoplasmic encephalitis and the acquired immuno-deficiency syndrome (AIDS). Ann. Intern. Med. 108: Renoult, E., M. F. Biava, I. Aimone-Gastin, F. Aouragh, D. Hestin, L. Kures, and M. Kessler Evolution and significance of Toxoplasma gondii antibody titers in kidney transplant recipients. Transplant. Proc. 24: Salt, A., G. Sutehall, M. Sargaison, C. Woodwart, N. D. Barnes, R. Y. Calne, and T. G. Wreghitt Viral and Toxoplasma gondii infections in children after liver transplantation. J. Clin. Pathol. 43: Savva, D., J. C. Morris, J. D. Johnson, and R. E. Holliman Polymerase chain reaction for detection of Toxoplasma gondii. J. Med. Microbiol. 32: Stepick-Biek, P., P. Thulliez, F. G. Araujo, and J. S. Remington IgA antibodies for diagnosis of acute congenital and acquired toxoplasmosis. J. Infect. Dis. 162: Strittmatter, C., W. Lang, O. D. Wiestler, and P. Kleihues The changing pattern of human immunodeficiency virus associated cerebral toxoplasmosis: a study of 46 postmortem cases. Acta Europathol. 83: Sulahian, A., P. D. C. Nugues, Y. F. F. Garin, H. Pelloux, P. Longuet, B. Slizewicz, and F. Derouin Serodiagnosis of toxoplasmosis in patients with acquired or reactivating toxoplasmosis and analysis of the specific IgA antibody response by ELISA, agglutination and immunoblotting. Immunol. Infect. Dis. 3: Suzuki, Y., and J. S. Remington Importance of membrane-bound antigens of Toxoplasma gondii and their fixation for serodiagnosis of toxoplasmic encephalitis in patients with acquired immunodeficiency syndrome. J. Clin. Microbiol. 28: Thelliez, E., A. Ducasse, C. Marx-Chemla, E. Tabare, A. Segal, and J. M. Pinon Anterior chamber puncture and toxoplasmosis: a four year study. Bull. Soc. Ophthalmol. 4: Tirard, V., G. Niel, M. Rosenheim, C. Katlama, L. Ciceron, W. Ogunkolade, M. Danis, and M. Gentilini Diagnosis of toxoplasmosis in patients with AIDS by isolation of the parasite from the blood. New. Engl. J. Med. 324: Torres, R. A., M. Barr, M. Thorn, G. Gregory, S. Kiely, E. Chanin, C. Carlo, M. Martin, and J. Thornton Randomized trial of dapsone and aerosolized pentamidine for the prophylaxis of Pneumocystis carinii pneumonia and toxoplasmic encephalitis. Am. J. Med. 95: Vendrell, J. P., J. Reynes, M. F. Huguet, J. Ngou, C. Michard, N. Atoui, F. Pralong, M. Segondy, and A. Serre In-vitro synthesis of antibodies to Toxoplasma gondii by lymphocytes from HIV-1 infected patients. Lancet 342: Weiblen, B. J., F. K. Lee, E. R. Cooper, S. H. Landesman, K. Mcintosh, J. A. S. Harris, S. Nesheim, H. Mendez, S. I. Pelton, A. J. Nahmias, and R. Hoff Early diagnosis of HIV infection in infants by detection of IgA HIV antibodies. Lancet 335: Wong, S. Y., M. P. Hajdu, R. Ramirez, P. Thulliez, R. McLeod, and J. S. Remington Role of specific immunoglobulin E in diagnosis of acute Toxoplasma infection and toxoplasmosis. J. Clin. Microbiol. 31: Zufferey, J., A. Sugar, P. Rudaz, J. Bille, M. P. Glauser, and J. P. Chave Prevalence of latent toxoplasmosis and serological diagnosis of active infection in HIV- patients. Eur. J. Clin. Microbiol. Infect. Dis. 12: