Literature review > Issue_3 > Review on Verkooyen et al. 

In this paper Dr. Verkooyen and colleagues compare the performance of three new C. trachomatis (CT) synthetic peptide based EIA serologic assays. The same three assays (the EIA from Lab Systems in Finland, the SERO CT from Savyon in Israel, and the pELISA from Medak in Germany) were reported on in an article published in the Journal of Clinical Microbiology in 2002 [1]. That paper was reviewed by Dr. Stamm for the SDI website a year or so ago. In that study, performance of the assays was compared using the microimmunofluoresence test as the gold standard. Here, performance of the three assays was compared by testing stored sera from four groups of patients: Group 1 - 433 Dutch blood donors, Group 2 - 22 patients with well characterized C. pneumoniae infections, Group 3 - 324 Rotterdam sexually transmitted infections (STI) clinic patients who were positive by the Roche PCR assay for CT, and Group 4 - 100 patients from the same STI clinic who tested negative by PCR for CT. The use of these four groups provides interesting data, though a weakness of the study was the fact that the STI clinic patients apparently were not sequentially enrolled. This would not affect calculations of sensitivity but if Group 4 was selected in some way, bias could have been introduced that might have affected estimates of specificity. For the purposes of this review I assumed that this did not happen.

Overall there was little difference between the sensitivity and specificity estimates for the three assays. Based on the Group 3 data, sensitivity for the IgG assays varied from 68 to 75 percent, while for the IgA assays they ranged from 38 to 48 percent. Among the Group 4 patients apparent false positives for the IgG assays ranged from 26 to 38 percent and for the IgA assays from 14 to 17 percent. A strength of this study is that the Group 3 and 4 patients were defined as CT positive or negative based on an excellent gold standard, the Roche PCR assay. It would have been useful to have seen the data stratified by clinical diagnosis, i.e., for men with urethritis and women with cervicitis compared to those who had clinically unapparent chlamydial infections. In contrast to the findings from the STI clinic, seropositivity rates were uniformly low in Group 1 blood donors (5% for the IgA assays and approximately 6% for the IgG assays). Of greater interest was the fact that the antibody prevalence paralleled that of active infection when the data were stratified by age groups. Prevalence rates steadily decreased with age, which is in sharp contrast to what is observed for C. pneumoniae serologic results, where antibody prevalence rates steadily increase with age. This observation argues that the synthetic peptide CT antigens used in these assays do not cross react with C. pneumoniae. This hypothesis is further strengthened by the Group 2 data for patients with well characterized C. pneumoniae infections. Only slightly more of these patients were CT antibody positive than was observed in the normal blood donor group. Furthermore, despite the fact that C. pneumoniae antibody titers apparently rose in the Group 2 patients sequentially over time there was no change in the CT antibody titers. It would have been interesting to have seen these data graphed out for those few patients who had both CT and C. pneumoniae antibodies to clearly demonstrate this fact. Nonetheless, the authors successfully make the point that these three assays do not cross react significantly with C. pneumoniae, which has been a problem with other CT serologic tests.

In the discussion the authors make note of the relatively low sensitivity and specificity of these new assays in the STI clinic population, but then end by simply concluding that all three detect specific CT antibodies. Unfortunately they do not provide their views on the utility of CT serology in general or for the assays they studied in particular. This is an important issue, as most experts in the field do not believe that there is any role for serology in making the laboratory diagnosis of active CT infection. This is especially true today in the era of nucleic acid amplification tests (NAATs). It is surprising to me that commercial enterprises continue to produce serologic assays for CT diagnosis outside of North America. Here we have concluded that the role of CT serologic diagnosis is restricted to the research arena. The use of serologic assays for CT diagnosis is not recommended in the Center for Disease Control's recently published CT screening test guidelines [2]. The reasons for this are readily apparent from the data presented in this paper. Compared to a PCR assay in STI clinic patients, 30 to 35 percent of infected patients were missed by these new serologic tests. Furthermore, though the assays appear to be significantly more specific than older assays, 26 to 38 percent of PCR negative patients in the STI clinic population were positive. Of course this finding is not at all surprising, as STI clinic clients are more likely than the general population to have had a recent CT infection or to have had multiple past CT infections resulting in persistent CT specific antibody responses. It is clear that utilization of any one of these tests for CT diagnosis in an STI population in the Netherlands would have resulted in substantial under and over treatment. Therefore, I cannot think of a rationale for recommending these tests for use in deciding which patients should and should not be treated for chlamydial infections. This is especially true in the NAAT era. In the pre-NAAT era, a case for their use might have been made, but even then the sensitivity of these assays for active infection appears to be about the same as older nonculture tests such as antigen EIAs and direct gene probe assays [2]. Moreover, these non-culture CT tests have better specificities than do these serologic assays. Granted, NAAT tests are probably more expensive than serologic assays, but I would guess not by much. Moreover, the older nonculture assays would probably be no more expensive, if not cheaper, and at least would be more specific.

I have heard the argument made that sensitive and specific serologic assays might be useful for screening asymptomatic populations. However, NAATs, run on easy-to-obtain urine specimens, have many advantages over serologic tests including ease of specimen collection and much better performance profiles. Furthermore, it is not clear how these serologic assays would perform in asymptomatic CT infected individuals. Sensitivity of only 70 % in actively infected STI clinic patients suggests their performance in asymptomatically infected patients in the general population would be even worse.

In summary, it is not clear to me that these assays have applications other than research. Given the fact that they appear not to cross react with C. pneumoniae, they may have a role here but it is hard to believe that the manufactures will be able to make a profit based on such limited use. Therefore, there must be a market for such assays somewhere in the world and it would be of considerable interest to know how these assays actually are being used and what kind of arguments their protagonists would mount to justify their application. This might be an interesting subject for discussion at one or another of the international chlamydia meetings.

References:

1. Morre SA, Munk C, Persson K, Kruger-Kjaer S, van Dijk R. Meijer CJ, and van Den Brule AJ. Comparison of three commercially available peptide-based immunoglobulin G (IgG) and IgA assays to microimmunofluorescence assay for detection of Chlamydia trachomatis antibodies. Journal of Clinical Microbiology. 40(2):584-7, 2002 Feb.

2. Centers for Disease Control and Prevention. Screening tests to detect Chlamydia trachomatis and Neiseeria gonorrhoeae infections - 2002. MMWR 2002;51(No. RR-15): p. 6 

about SDI | newsletters | grants | publications | literature reviews

WHO Home - WHO Search - TDR Home - SDI Home - SDI Contact us
(c) WHO/OMS 2001