1 Department of Obstetrics and Gynaecology, 2 Department of Clinical Epidemiology and Medical Technology Assessment and 3 Department of Medical Microbiology, Academisch Ziekenhuis Maastricht, PO Box 5800, 6202 AZ Maastricht, The Netherlands
4 To whom correspondence should be addressed. e-mail: jlan{at}sgyn.azm.nl
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: Chlamydia antibody test/diagnostic test/screening/serology/tubal infertility
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The first aim of this study was to compare five commercially available chlamydia IgG antibody tests in their accuracy to predict tubal factor subfertility, using laparoscopy as a reference. The performances of a species-specific MIF test and three ELISA tests were compared with findings obtained with a MIF test that had been used in our hospital for 10 years. Cross-reactivity with C. pneumoniae antibodies was evaluated in the five CAT tests as a possible cause of false positive CAT results. The second aim was to evaluate whether combinations of two tests could improve the predictive value of CAT.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Between June 1992 and September 2001, 315 successive patients underwent laparoscopy as part of their fertility work-up. All took part in this study. Patients who had undergone previous pelvic surgery (except for an uneventful appendectomy or Caesarean section) were excluded.
All laparoscopy reports were scored by two gynaecologists, who were blinded for the CAT results. For the sake of the study, tubal pathology at laparoscopy was defined as extensive periadnexal adhesions and/or distal occlusion of both tubes, since this has been shown to reflect chlamydia associated tubal disease most accurately (Land et al., 1998).
After thawing the cryopreserved sera of the participating patients, the five different CAT tests were performed. A number of patients had participated in a previous study (Land et al., 1998), but for the present study all serum samples were retested. Finally, data obtained at laparoscopy were correlated to the serological test results.
Serological methods
All assays and calculations were performed according to the manufacturers instructions. For each test, all sera were tested simultaneously using the same batch of test kits, to avoid inter-test variability.
MIF Biomerieux
Chlamydia trachomatis-spot IF test (Biomerieux). In this indirect fluorescent IgG antibody test, C. trachomatis L2 is used as a group antigen. A positive reaction is characterized by specific fluorescence of the C. trachomatis elementary bodies. For quantitative determination serial dilutions are made. In this study all slides were evaluated independently by two readers. In case of disagreement, the judgement of a third reader was decisive. In this study a third reader was needed in four cases.
MIF Labsystems
Chlamydia pneumoniae IgG/IgM micro-IF test (Labsystems, Finland). In this species-specific MIF test C. pneumoniae, C. trachomatis and C. psittaci elementary bodies are used as antigens. In order to diminish cross-reactivity, the immunological activity of chlamydia LPS in C. pneumoniae and C. trachomatis antigens has been reduced. In this study the test was used to detect anti-C. trachomatis and anti- C. pneumoniae IgG antibodies. A positive reaction is characterized by specific fluorescence of the respective elementary bodies. For quantitative determination serial dilutions are made. In this study all slides were evaluated independently by two readers. In case of disagreement, the judgement of a third reader was decisive. In this study a third reader was needed in three cases.
ELISA Labsystems
Chlamydia trachomatis IgG EIA (Labsystems). In this indirect enzyme immunoassay synthetic peptides are used, which are derived from MOMP of C. trachomatis (Närvänen et al., 1997). The optical density of the test plates is read in a spectrophotometer, and the signal to cut-off indices (SCI) are categorized as negative, equivocal, positive or highly positive.
pELISA Medac
Chlamydia trachomatis-IgG-pELISA (Medac, Germany). Medac has a LPS-based, genus-specific test (rELISA) and a species-specific ELISA (pELISA) on the market. The pELISA has been used in this study, and according to the manufacturers information it is based on a synthetic peptide from MOMP of C. trachomatis. The optical density of the test plates is read in a spectrophotometer, and the SCI are categorized as negative, equivocal, or positive.
ELISA Savyon
Sero-CT-IgG (Savyon, Israel). In this ELISA C. trachomatis species-specific peptides, derived from different serotypes are used. The optical density of the test plates is read in a spectrophotometer, and the SCI are categorized as negative, equivocal, or positive.
Statistical methods
The outcomes of the five tests were compared with the findings at laparoscopy in order to determine their prognostic value of tubal factor subfertility. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), likelihood ratio (LR), odds ratio (OR) and 95% confidence interval (CI) were calculated.
For comparison of LR of the results of the CAT tests, the 2-test was used. To test the difference between OR of the different tests, the bootstrap technique was used (Efron and Tibshiram, 1993
). Two-sided P < 0.05 was considered significant. A stepwise logistic regression analysis was used to select the best combination of tests. For evaluation of agreement between tests,
was calculated, and
> 0.7 was considered to indicate good agreement. For comparison of the distributions of C. pneumoniae titres in patients with a positive and negative chlamydia antibody test respectively, the MannWhitney U-test was used. P < 0.05 was considered significant.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
In 158 patients all five CAT tests were negative, 67 patients had one positive test, 25 patients had two positive tests, 18 patients had three positive tests, 20 patients had four positive tests and in 27 patients all five tests were positive. Table II gives the number of positive CAT tests in 315 patients. The likelihood of tubal factor subfertility was correlated to the number of positive tests. A statistically significant difference was found between the LR of three, four and five simultaneously positive tests as compared with the LR of one positive test only.
|
As a second strategy the CAT tests were divided into two groups: the ELISA tests, as they can be automated, and the more laborious MIF tests. Based on stepwise logistic regression analysis with the three ELISA tests in the first block and the two MIF tests in the second block, pELISA Medac was chosen as the first test, and MIF Labsystems as the second test (Table III). In retesting of initially positive sera, the serial set was considered positive if the second test was also positive, and in retesting of initially negative sera the serial set was considered negative if the second test was also negative. Using pELISA Medac on all serum samples, 62 patients (20%) had a positive test. After retesting these 62 sera by MIF Labsystems, 44 sera (14%) remained positive. An initial negative test was found in 253 patients (80%), and after retesting these 253 sera, 245 were negative (78%). The results of serial testing with the chosen set of CAT tests (i.e. pELISA Medac and MIF Labsystems) are given in Table III. The OR of both serial sets (14.2 and 9.2 respectively) were lower than the OR of MIF Labsystems as a single test (15.7; Table I).
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In the present study, CAT results were compared with the findings at laparoscopy, and therefore only women who had undergone a laparoscopy were included. Patients with low chlamydia antibody titres are less likely to have tubal pathology and were less likely to be included in the study, since many will have conceived before laparoscopy can be done. This verification bias (Mol et al., 1999) will influence predictive values of CAT, which are dependent on the prevalence of disease. Verification and selection bias is hard to prevent in clinical studies, however, unless one is prepared to perform the complete fertility investigation of a patient on a single day. Although the bias will result in overestimation of the LR, the overestimation will be similar for all tests investigated.
In comparing the five CAT tests, considerable variation was found in the number of patients with a positive test (Table I). An IgG titre 32 was found in 132 women using MIF Biomerieux and in 52 using MIF Labsystems, 53 had a positive test by ELISA Labsystems, 62 by pELISA Medac and 87 by ELISA Savyon. The tests with the highest number of positive test results had the lowest PPV, indicating the highest rates of false positive test results. False positive CAT results (i.e. positive CAT tests in patients without tubal pathology at laparoscopy) may be explained by cross-reactivity with C. pneumoniae antibodies, which can be found in
70% of subfertile women (Gijsen et al., 2001
). In this study a significant correlation was found between C. pneumoniae antibody titres and CAT results in four tests, in particular in MIF Biomerieux (P < 0.00001). This confirms the findings of our previous study using a C. pneumoniae IgG ELISA, in which we found cross-reactivity to occur between C. pneumoniae and C. trachomatis antibodies in MIF Biomerieux. Since the only CAT test (pELISA Medac) in which no significant correlation with the distribution of C. pneumoniae titres was found did not have the highest PPV, cross-reactivity with C. pneumoniae antibodies does not seem to be the only explanation for the high rates of false positive test results obtained in the CAT tests. Since not all women with chlamydia antibodies have tubal pathology at laparoscopy, it has been suggested that genetic factors in the host may also play a role, by modulating immune defence mechanisms and the development of late sequelae (Kinnunen et al., 2002
).
Comparing MIF and ELISA, the results (Table I) suggest that ELISA tests tend to have lower sensitivity and NPV, i.e. more false negative test results. In ELISA tests chosen for this study, specific synthetic peptides are used which are considered analogous to the serotype-specific antigenic determinants of MOMP of C. trachomatis (Närvänen et al., 1997). These serotype-specific determinants differ between tests, and may explain the differences found in test performances between ELISA tests of different manufacturers. Furthermore, tests based on highly specific peptides may be so specific that they are not able to detect all relevant antigens (Bas et al., 2001
). Variants of serotypes have been identified in urogenital isolates (Morré et al., 1998
) and mutations have been shown to occur in positions within MOMP (Dean et al., 2000
). Consequently, highly specific tests may not be able to identify all serotypes involved in chlamydia infection, and cause false negative CAT results (i.e. negative CAT tests in patients with tubal pathology at laparoscopy).
There are a few reports in the literature on the diagnostic accuracies of different antibody tests for C. trachomatis, in which the tests evaluated in our study have been included. In these studies either tubal pathology, or the direct demonstration of the micro-organism in the genital tract, have been used as reference standards for the serological tests. Each reference standard has its limitations, however. Since tubal pathology can be caused by other micro-organisms in addition to C. trachomatis, it is obvious that tests based on chlamydia antibodies will be imperfect in predicting all tubal pathology. Studies in which the direct detection of the micro-organism is used as a reference have also limited diagnostic accuracy (Chernesky et al., 1998; Bas et al., 2001
), since superficial infections may provide a poor stimulus for antibody formation. Paukku et al. (1998
) did not find a significant difference between the presence of IgG antibodies in 78 patients with tubal factor subfertility, using a modified MIF test and ELISA Labsystems. In a serological follow-up study of 16 women with C. trachomatis positive cervical swabs, ELISA Labsystems, ELISA Savyon and a MIF test have been used (Clad et al., 2000
). ELISA Labsystems was found to be the most sensitive test, and it was concluded that ELISA Savyon did not cover all chlamydia serotypes. Morré et al. (2002
) studied the IgG prevalences in 43 women with PCR positive cervical swabs and 106 PCR negative women. Results obtained by two in-house MIF tests were compared with results by ELISA Labsystems, pELISA Medac and ELISA Savyon. The authors concluded that the ELISA tests performed as well as the MIF tests.
In the present study the likelihood of tubal factor subfertility improved as the number of positive CAT tests in a patient increased (Table II). Compared with one positive CAT test, the LR improved significantly in patients in whom three, four or five positive tests were found. But from a clinical point of view, performing more than two CAT tests in patients is impractical and expensive. Therefore, we evaluated whether two serially performed tests could be of any benefit in predicting tubal factor subfertility. First, from a stepwise logistic regression analysis it was concluded that the diagnostic performance of the best single CAT test (i.e. MIF Labsystems) could not be improved by adding a second CAT test. Second, we evaluated the performance of a serial set of tests from a laboratory perspective, by constructing a model in which pELISA Medac was performed as the first test and MIF Labsystems as the second test. Although pELISA Medac was found to have a significantly lower OR (8.2) compared with MIF Labsystems (15.7), ELISA tests have the advantage over MIF tests of being less laborious. From our results it can be concluded that if pELISA Medac is performed as the first test on all samples, and all samples with positive test results (i.e. 20% of all samples) are retested with MIF Labsystems, the predictive value of the set (OR 14.2) is comparable to the predictive value of MIF Labsystems as a single test (OR 15.7). Cost-effectiveness analysis has to demonstrate which strategy is to be preferred: MIF Labsystems as a single test on all samples, or pELISA Medac on all samples and retesting of the positive ones with MIF Labsystems. Third, we hypothesized that two tests with poor agreement (low ) might have different antigenic properties, and might react with a greater number of serotypes and therefore complement each other. From the results presented in Table III it can be concluded that the low
set of tests (ELISA Labsystems and MIF Biomerieux) did not improve the predictive value of CAT significantly, compared with the predictive values of the single tests of which the set was composed. Therefore the hypothesis of complementary testing with the set of tests with low
had to be rejected.
The results of the present study show that there still is no excellent screening test for tubal pathology in subfertile women. In order to develop more accurate tests for the prediction of chlamydia-associated tubal pathology, future research should focus on the immunopathology of chlamydia infections. Evidence exists that patients with chronic, persisting chlamydia infections are particularly at risk for developing late sequelae. Although IgG antibodies are markers of a previous infection, they do not reflect an ongoing chronic inflammation properly. Candidates to be introduced into screening for tubal factor subfertility, in addition to specific C. trachomatis IgG antibodies, are anti-HSP60 and anti-LPS antibodies. HSP60 has been shown to play a prominent role in chronic inflammation and scarring (Claman et al., 1997), and anti-LPS antibodies might be indicators of ongoing chlamydia infection (Tuuminen et al., 2000
).
In conclusion, although ELISA tests have been claimed to be highly sensitive and specific, in the present study they were not superior in predicting tubal factor subfertility. Of the five CAT tests evaluated, MIF Labsystems had the best diagnostic performance, and among the three ELISA tests, pELISA Medac performed best. MIF Biomerieux had the largest number of false positive test results, probably due to cross-reactivity with C. pneumoniae antibodies. Combining two different CAT tests did not improve the predictive value for tubal factor subfertility. Health care evaluation from an economic perspective has to prove whether serial testing with the automated pELISA Medac as a first test, and retesting of all positive serum samples with the more laborious MIF Labsystems, is to be preferred to testing of all samples with MIF Labsystems only.
![]() |
Acknowledgements |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Chernesky, M., Luinstera, K., Sellors, J., Schachter, J., Moncada, J., Caul, C., Paul, I., Mikaelian, L., Toye, B., Paavonen, J. and Mahony, J. (1998) Can serology diagnose upper genital tract Chlamydia trachomatis infections? Studies on women with pelvic pain, with or without chlamydial plasmid DNA in endometrial biopsy tissue. Sex. Transm. Dis., 25, 1419.[ISI][Medline]
Clad, A., Freidank, H.M., Kunze, M., Schnoeckel, U., Hofmeier, S., Flecken, U. and Petersen, E.E. (2000) Detection of seroconversion and persistence of Chlamydia trachomatis antibodies in five different serological tests. Eur. J. Clin. Microbiol. Infect. Dis., 19, 932937.[CrossRef][ISI][Medline]
Claman, P., Honey, L., Peeling, R.W., Jessamine, P. and Toye, B. (1997) The presence of serum antibody to the chlamydial heat shock protein (CHSP60) as a diagnostic test for tubal factor infertility. Fertil. Steril., 67, 501504.[CrossRef][ISI][Medline]
Dean, D., Suchland, R.J. and Stamm, W.E. (2000) Evidence of long-term cervical persistence of Chlamydia trachomatis by omp1 genotyping. J. Infect. Dis., 182, 909916.[CrossRef][ISI][Medline]
Dowell, S.F., Peeling, R.W., Boman, J., Carlone, G.M., Fields, B.S., Guarner, J., Hammerschlag, M.R., Jackson, L.A., Kuo, C.C., Maass, M. et al. (2001) Standardizing Chlamydia pneumoniae assays: recommendations from the Centers for Disease Control and Prevention (USA) and the Laboratory Centre for Disease Control (Canada). Clin. Infect. Dis., 33, 492503.[CrossRef][ISI][Medline]
Efron, B. and Tibshiram, R.J. (1993) An Introduction to the Bootstrap. Chapman & Hall, New York.
Gijsen, A.P., Land, J.A., Goossens, V.J., Leffers, P., Bruggeman, C.A. and Evers, J.L.H. (2001) Chlamydia pneumoniae and screening for tubal factor subfertility. Hum. Reprod., 16, 487491.
Kinnunen, A.H., Surcel, H.M., Lehtinen, M., Karhukorpi, J., Tiitinen, A., Halttunen, M., Bloigu, A., Morisson, R.P., Karttunen, R. and Paavonen, J. (2002) HLA DQ alleles and interleukin-10 polymorphism associated with Chlamydia trachomatis-related tubal factor infertility: a casecontrol study. Hum. Reprod., 17, 20732078.
Land, J.A., Goossens, V.J. and Evers, J.L.H. (1998) How to use Chlamydia antibody testing in subfertility patients. Hum. Reprod., 13, 10941098.[Abstract]
Mannion, P.T., Mallinson, H. and Treharne, J.D. (1991) Serological diagnosis with the Chlamydia spot-IF test. J. Med. Microbiol., 35, 244248.[Abstract]
Mol, B.W.J., van der Veen, F. and Bosschuyt, P.M.M. (1999) Chlamydia antibody titres. Hum. Reprod., 14, 268270.
Morré, S.A., Ossewaarde, J.M., Lan, J., van Doornum, G.J.J., Walboomers, J.M.M., MacLaren, D.M., Meijer, C.J.L.M. and van den Brule, A.J.C. (1998) Serotyping and genotyping of genital Chlamydia trachomatis isolates reveal variants of serovar Ba, G and J as confirmed by omp1 nucleotide sequence analysis. J. Clin. Microbiol., 36, 345351.
Morré, S.A., Munk, C., Persson, K., Krüger-Kjaer, S., van Dijk, R., Meijer, C.J.L.M. and van den Brule, A.J.C. (2002) Comparison of three commercially available peptide-based immunoglobulin G (IgG) and IgA assays to microimmunofluorescence assay for detection of Chlamydia trachomatis antibodies. J. Clin. Microbiol., 40, 584587.
Närvänen, A., Puolakkainen, M., Hao, W., Kino, K. and Suni, J. (1997) Detection of antibodies to Chlamydia trachomatis with peptide-based species-specific enzyme immunoassay. Infect. Dis. Obstet. Gynecol., 5, 349354.
Ossewaarde, J.M., de Vries, A., van den Hoek, J.A.R. and van Loon, A.M. (1994) Enzyme immunoassay with enhanced specificity for detection of antibodies to Chlamydia trachomatis. J. Clin. Microbiol., 32, 14191426.[Abstract]
Paukku, M., Närvänen, A., Puolakkainen, M., Dreesbach, K., Tiitinen, A., Hao, W., Anttila, T.I. and Paavonen, J. (1998) Detection of Chlamydia trachomatis antibodies in 2 novel tests: rELISA and peptide EIA. Int. J. STD AIDS, 9, 604607.[CrossRef][ISI][Medline]
Peeling, R.W., Wang, S.P., Grayston, J.T., Blasi, F., Boman, J., Clad, A., Freidank, H., Gaydos, C.A., Gnarpe, J., Hagiwara, T. et al. (2000) Chlamydia pneumoniae serology: interlaboratory variation in microimmunofluorescence assay results. J. Infect. Dis., 181 (Suppl. 3), s426429.[CrossRef][ISI][Medline]
Punnonen, R., Terho, P., Nikkanen, V. and Meurman, O. (1979) Chlamydial serology in infertile women by immunofluorescence. Fertil. Steril., 31, 656659.[ISI][Medline]
Schachter, J., Cles, L., Ray, R. and Hines, P.A. (1979) Failure of serology in diagnosing chlamydial infections of the female genital tract. J. Clin. Microbiol., 10, 647649.[ISI][Medline]
Treharne, J.D., Darougar, S. and Jones, B.R. (1977) Modification of the microimmunofluorescence test to provide a routine serodiagnostic test for chlamydial infection. J. Clin. Pathol., 30, 510517.[Abstract]
Tuuminen, T., Palomäki, P. and Paavonen, J. (2000) The use of serologic tests for the diagnosis of chlamydial infections. J. Microbiol. Meth., 42, 265279.[CrossRef][ISI][Medline]
Wang, S.P. and Grayston, J.T. (1970) Immunologic relationship between genital TRIC, lymphogranuloma venereum, and related organisms in a new microtiter indirect immunofluorescence test. Am. J. Ophtalmol., 70, 367374.[ISI][Medline]
Submitted on March 14, 2003; resubmitted on May 14, 2003; accepted on August 19, 2003.