Materials and methods

STUDY POPULATION

Samples were collected from two population groups. The first population of 380 comprised 193 men with symptoms of urethritis and 187 high risk female commercial sex workers, who visited two STD clinics in Fukuoka, Japan, from April to December 1997. The second population group consisted of 91 high risk female commercial sex workers who attended one of the STD clinics from January to March 1998.

SAMPLE COLLECTION

For the first population group, first void urine (20–30 ml) was collected from male patients into sterile screw cap tubes and transported to the laboratory, where it was divided into three aliquots. The first aliquot (10 ml) was used for the IDEIA chlamydia and IDEIA PCE chlamydia (a newly improved EIA kit) tests (Dako, Ely, Cambridgeshire, UK); the second aliquot (8 ml) was used for the Amplicor PCR assay (Roche Molecular Systems, Branchburg, New Jersey, USA). The final aliquot was stored at −20°C for further evaluation of discrepant results.

For each woman, two endocervical specimens were obtained with a speculum by inserting a swab into the endocervix. Before sampling, the endocervix was cleaned with a swab to remove excess mucous. The swab was rotated several times before withdrawal. The first swab was placed into an Amplicor transport tube and the second into IDEIA transport medium, as provided with each kit. IDEIA chlamydia and Amplicor PCR specimen collection kits for swabs were used in accordance with each manufacturer's recommendation.

For assessment of vaginal specimens (second population group), four swab specimens, including one patient obtained vaginal swab, two clinician obtained endocervical swabs, and one clinician obtained vaginal swab, were collected from each woman. Initially, each woman was asked to obtain a vaginal swab specimen by inserting the swab about 3–5 cm into the vagina, rotating it several times, and removing it. The swab was placed into an IDEIA transport tube by a clinician. Then, a vaginal swab and two endocervical swab specimens (in that order) were obtained by a clinician using a speculum. The clinician obtained vaginal swab was placed into IDEIA transport medium. Of the two endocervical swabs, the first swab was placed into Amplicor transport medium and the second into IDEIA transport medium.

TESTING OF SAMPLES

In the first study group, first void urine and endocervical specimens were processed and tested by the IDEIA chlamydia test, the IDEIA PCE chlamydia test, and the Amplicor PCR assay, according to each manufacturer's instructions. For the second population group, patient obtained and clinician obtained vaginal swab specimens were assayed by the IDEIA PCE chlamydia kit, and endocervical swab specimens were assayed by the IDEIA PCE chlamydia kit and the Amplicor PCR test. Urine and endocervical swab specimens were stored at 2–8°C for up to three days until processed and measured with the Amplicor PCR test as described in detail in our previous study.² The results were interpreted according to instructions and quality control criteria provided by the manufacturer.

RESOLUTION OF DISCREPANCIES AND CONFIRMATORY TESTING

For evaluation of urine and endocervical specimens (first population group), a specimen was considered to be positive for C trachomatis infection if the IDEIA PCE test and the Amplicor PCR assay gave positive results. When there was a discrepancy between the IDEIA PCE and PCR results, nested PCR with a different plasmid target region from that of the Amplicor PCR test was performed as a confirmatory test. The first PCR amplification was performed using primers CT2 and CT5, as described previously.¹⁹ The reaction product was then amplified for a second time using primers CT7 (5`-GGATTTATCGGAAACC TTGA-3`) and CT8 (5`-CTTTCAATGG AATAGCGGGT-3`), with all other conditions remaining the same.¹⁹ The amplified product (10 μl) was analysed by electrophoresis on a 2% agarose gel. If a specimen was positive using the supplementary testing, combined with one other positive test result (IDEIA PCE or Amplicor PCR), the sample was confirmed as being positive for C trachomatis. After resolution of the discrepancies, relative sensitivity and specificity, positive and negative predictive values, and 95% confidence intervals were calculated. Statistical analysis of the data was also performed using the Pearson χ² test. A p value < 0.05 was considered significant.

For the evaluation of vaginal specimens (second population group), a woman was considered to be infected with C trachomatis if the Amplicor PCR or IDEIA PCE chlamydia test was positive for the clinician obtained endocervical swab. A woman who had a positive vaginal swab but negative endocervical swabs was considered to be positive if she was confirmed IDEIA PCE chlamydia positive using the IDEIA blocking test for vaginal swabs. The IDEIA blocking test was performed and results interpreted according to information provided by the manufacturer.

Results

URINE AND ENDOCERVICAL SPECIMENS

The results of the detection of C trachomatis in male and female specimens using IDEIA PCE were compared with those obtained by IDEIA and Amplicor PCR (table 1). Of 193 male first void urine specimens tested, 135 were negative and 46 were positive by IDEIA, IDEIA PCE, and Amplicor PCR. Twelve discrepant results were obtained. Of the 12 specimens, seven were positive according to IDEIA PCE and Amplicor PCR. Thus, these seven men were considered to be positive for C trachomatis. The remaining five specimens were confirmed as being positive using the nested PCR assay. After resolution of discrepancies, of the 193 male urine specimens tested, 58 (30.1%) were positive for C trachomatis and 135 (69.9%) were negative. Of 187 female endocervical swab specimens tested, 146 were negative and 34 were positive by IDEIA, IDEIA PCE, and Amplicor PCR. Seven discrepant results were obtained. Of the seven specimens, four were positive by IDEIA PCE and Amplicor PCR. Thus, these four specimens were considered to be positive for C trachomatis infection. Of the remaining three specimens, two were confirmed as being positive and one negative using nested PCR. After resolution of discrepancies, of the 187 endocervical specimens tested, 40 (21.4%) were considered to be positive for C trachomatis and 147 (78.6%) were negative. The relative sensitivity and specificity, 95% confidence intervals, and predictive values were then calculated according these results (table 2). The relative sensitivities of IDEIA, IDEIA PCE, and Amplicor PCR on male first void urine specimens were 79.3%, 91.4%, and 100%, respectively. The relative sensitivities of IDEIA, IDEIA PCE, and Amplicor PCR on female endocervical swab specimens were 85.0%, 95.0%, and 100%, respectively. There were no statistical differences between the sensitivities of the Amplicor PCR assay, the IDEIA PCE chlamydia test, and the IDEIA chlamydia test.

VAGINAL SPECIMENS

Of 91 women tested, 64 were negative for all four sample types collected and 20 were positive for all four sample types collected (two clinician obtained endocervical swabs for Amplicor PCR and IDEIA PCE, and one clinician obtained and one patient obtained vaginal swab for IDEIA PCE) (table 3). There were seven discrepancies among the sample types collected. Of the seven women, six were confirmed to be infected with C trachomatis because the Amplicor PCR assay or the IDEIA PCE test was positive for the clinician obtained endocervical specimens. The remaining woman was confirmed to be positive by the IDEIA PCE blocking assay using the patient obtained vaginal swab. After resolution of discrepancies, of the 91 women tested, 27 (29.8%) were found to be infected with C trachomatis and 64 (70.2%) were found not to be infected. The positivity rates for patient collected vaginal specimens and clinician collected vaginal specimens were 25.2% (23 of 91) and 23.1% (21 of 91), respectively; those for clinician collected endocervical swabs by PCR and IDEIA PCE were similar—both 27.5% (25 of 91). There was no difference between testing self collected or clinician collected specimens and between testing vaginal specimens and endocervical swabs.

Discussion

In Japan, commercial PCR or LCR assay kits are available as routine tests for the detection of C trachomatis. However, these DNA amplification tests are extremely costly⁵ compared with conventional EIAs for antigen detection. Furthermore, these tests require specialised facilities to reduce Amplicor PCR contaminants. More recently, a new generation dual amplified immunoassay IDEIA PCE chlamydia has become available, which has been shown to be diagnostically reliable when applied to genital swabs.¹⁸ The IDEIA PCE chlamydia test is 2.5 to five times more sensitive for the detection of C trachomatis elementary bodies than the conventional EIA test (IDEIA).²⁰ Currently, several studies have shown that analysis of urine specimens using DNA amplification methods is a possible alternative to the analysis of endocervical specimens for chlamydia diagnosis in women.^21,22 However, the sensitivity was lower when using female urine specimens than when endocervical specimens were used.^{10 21 22} The reason for this reduced sensitivity is that most women are infected with C trachomatis at the endocervix, a site remote from the urethra. Therefore, urine samples might not be suitable for the detection of endocervical infection. Moreover, the handling and laboratory processing of urine specimens is more difficult compared with endocervical or vaginal swab specimens. Recent publications have also shown that DNA amplification testing for chlamydia with patient obtained vaginal swabs is as sensitive as endocervical testing.^13–15 Patient obtained vaginal swabs seems to be a more suitable and less invasive method for the screening for C trachomatis than clinician obtained endocervical or vaginal specimens. To our knowledge, reports on C trachomatis detection in patient obtained vaginal swab specimens using an EIA test are very rare.

In the first part of our study, we compared the performance of IDEIA PCE with that of the IDEIA test and the commercially available PCR assay (Amplicor) in male first void urine and female endocervical samples. The positivity rates for IDEIA PCE on male first void urine and female endocervical swab specimens (urine, 27.5%; endocervical swab, 20.3%) were higher than those for IDEIA (urine, 23.8%; endocervical swab, 18.2%), and comparable with the Amplicor PCR assay (urine, 30.1%; endocervical swab, 21.9%). However, the results obtained in our study might not reflect the true clinical sensitivity of each test because samples were not tested by culture, and no allowance was made for amplification inhibitors, which might be present in some samples. Moreover, the discrepant analysis procedure used was only applied to the discrepant samples, and not the whole population tested, and this might have introduced some bias into the data analysis.²³ Other studies have reported EIAs to be less sensitive than amplification tests.^2–4 These EIAs are generally based on passive capture of chlamydia LPS and use conventional signal generation systems. The incorporation of dual immunoassay amplification technology into the IDEIA PCE test might explain why we obtained a comparable positivity rate to PCR. Moreover, the cost for each IDEIA PCE test is similar to IDEIA, but much lower than the Amplicor PCR assay. However, a large study might be required to assess the true clinical performance and value of the IDEIA PCE kit because the population size tested and number of positive samples in our study are not sufficient.

In the second part of our study, we evaluated the clinical importance of patient obtained vaginal swab specimens using a new EIA kit. The results demonstrated that testing self collected vaginal specimens in a Japanese population was as reliable as testing clinician collected specimens, and that testing vaginal specimens by IDEIA PCE chlamydia was an acceptable alternative to testing endocervical swabs by IDEIA PCE chlamydia or Amplicor PCR. The agreement between the positivity rates obtained for vaginal swabs and endocervical swabs was closer than has been reported for studies comparing urine specimens with endocervical swabs.^{10 21}

The prevalence rate of C trachomatis in the commercial sex workers tested was approximately 20–30%. This prevalence rate among these women is much higher than that in the general Japanese female population (approximately 5%).²⁴ Although the population tested was mainly asymptomatic, the prevalence was high because of the occupation of the population tested. In our city, female commercial sex workers are a major reservoir of STDs. To prevent the spread of C trachomatis infection to the general population, continuous close monitoring of C trachomatis infection among commercial sex workers is necessary. In this regard, patient obtained vaginal swabs using the IDEIA PCE test would be useful for the screening of C trachomatis among commercial sex workers and offers the potential for cost effective, reliable, and less invasive screening of high risk/prevalence female populations. However, our results might not be applicable to lower prevalence populations, such as those seen in family planning clinics, because the carriage of C trachomatis will be lower. Furthermore, a large study is required to evaluate the true clinical usefulness of patient vaginal swabs as an alternative to endocervical swabs.