Vaginal tampons as specimen collection device for the molecular diagnosis of non-ulcerative sexually transmitted infections in antenatal clinic attendees

Self-inserted vaginal tampons for the molecular diagnosis of non-ulcerative STIs were evaluated. Cervical and vaginal swabs, tampons and urines were collected from 185 first-time antenatal clinic attendees. Cultures and nucleic acid amplification assays (NAA) were performed. The sensitivity of PCR on tampons for Trichomonas vaginalis was with 94% (CI 85-98%) significantly higher (P<0.001) than culture (50%, CI 38-62%) or urine (53%, CI 41-65%). Neisseria gonorrhoeae culture had a sensitivity of 64% (CI 36-86%), strand displacement assay (SDA) had a sensitivity of 79% (CI 49-94%) using tampon specimens, 57% (CI 30-81%) using endocervical swabs and 43% (CI 19-70%) using urines. There was no difference in sensitivity of SDA for Chlamydia trachomatis using tampon specimens, urine or endocervical swabs. The specificity approached 100% for all assays on all specimens. NAA on tampons for the detection of T. vaginalis, N. gonorrhoeae and C. trachomatis identified more infections than assays on swabs or urines. This reached statistical significance for T. vaginalis only.     

Evaluation of an enzyme immunoassay (Chlamydiazyme) with confirmatory test for the detection of chlamydial antigen in urine from men

First catch urine specimens from 312 male patients were examined for the presence of chlamydial antigen by an enzyme immunoassay (Chlamydiazyme). Positive results were repeated and confirmed using a blocking assay. In addition, urethral swabs were examined by cell culture for Chlamydia trachomatis. Discrepant results were further analysed by direct immunofluorescence (IF) of the spun urine deposit. Paired specimens were positive from 26 subjects, and negative from 276 subjects. Eight paired specimens were urethral culture positive, and urine EIA negative. Two specimens, urine EIA positive but urethral culture negative, were positive on direct IF. The sensitivity, specificity, predictive value of a positive result, and predictive value of a negative result for urine EIA against cell culture and/or direct IF were 77.8%, 100%, 100% and 97.2% respectively.     

Evaluation of the new nucleic acid amplification system for direct detection of Chlamydia trachomatis and Neisseria gonorrhoeae in women

The performance of a real-time DNA amplification assay, BD ProbeTec ET System (BDPT, BD Diagnostic Systems), to detect Chlamydia trachomatis and Neisseria gonorrhoeae on endocervical and oropharyngeal samples was evaluated. After obtaining informed consent, 364 endocervical, 363 urine and 247 oropharyngeal specimens were collected from 307 cases. The overall agreement rate of the BDPT and Amplicor (AMP, Roche) assays for the detection of C. trachomatis and N. gonorrhoeae in endocervical samples was 99.2% (361/364) for C. trachomatis and 99.5% (362/364) for N. gonorrhoeae. Assay of oropharyngeal swabs by the BDPT yielded 21 C. trachomatis positives, and 19 of them were C. trachomatis negative by the DNA probe assay (Gen-Probe PACE). The AMP assay showed that 16/19 (84.2%) of the BDPT +/DNA probe - samples were positive. The BDPT also yielded 21 N. gonorrhoeae positives, 15 of which were negative with the DNA probe. Additional testing showed that all 15 BDPT +/DNA probe - samples were positive by the established nested PCR method. Our data suggest that the performance of the BDPT is comparable to that of AMP for detection of C. trachomatis and N. gonorrhoeae in endocervical swab samples and that it may be a useful method for detecting of C. trachomatis and N. gonorrhoeae in oropharyngeal samples clinically.     

Evaluation of diagnostic efficacy of PCR methods for Chlamydia trachomatis infection in genital and urine specimens of symptomatic men and women in India

In India, given the scarce availability of sensitive and specific methods, Chlamydia trachomatis genital infections may lead to severe clinical complications when left undiagnosed or underdiagnosed. The present study was conducted to evaluate the diagnostic efficiency and feasibility of polymerase chain reaction (PCR) assays using genital and urine specimens from men and women in India. Genital swabs and urine specimens collected from 143 patients attending the sexually transmitted disease (STD) clinic, Government General Hospital, Chennai, were tested by culture and a plasmid based PCR. Culture was positive in 27 (18.9%) patients. PCR gave positive results for 46 (32.2%) cases using genital specimens, and the positivity rate in urine was 25.2%. Once the discordant results between culture and PCR had been resolved by using a major outer membrane protein PCR, the overall sensitivity, specificity, and positive and negative predictive values for the plasmid PCR in genital specimens were 100%, 98%, 95.7%, and 100%, respectively. Corresponding values for urine PCR were 81.8%, 100%, 100%, and 92.5%, respectively. The prevalence of confirmed C. trachomatis infection was 30.8% in this STD population. The results confirmed the need to use sensitive and specific molecular assays like PCR to prevent underdiagnosis of genital chlamydial infections and to facilitate better clinical management of this infection in India.     

Detection of Chlamydia trachomatis and Neisseria gonorrhoeae in urine samples of males and females by the strand displacement amplification (SDA) method

Urine samples collected from 422 males and 53 females visiting a clinic in Kawasaki City who were suspected to have sexually transmitted infection were tested for the detection of Chlamydia trachomatis and Neisseria gonorrhoeae by BD ProbeTecET (SDA method). The detection rates of C. trachomatis by the SDA method and polymerase chain reaction (PCR) method (control) were as high as 98.1% for C. trachomatis, and as high as 99.4% for N. gonorrhoeae, and the concordance rate of detection of both bacterial species was high. The detection sensitivity and specificity of the SDA method were 90.6 and 99.3%, respectively for C. trachomatis and 98.7% and 100% for N. gonorrhoeae, when PCR was used as the standard method. There were no differences in these results between males and females. The number of patients showing a discrepancy of the results obtained between the SDA method and the PCR method was 9 for C. trachomatis and 1 for N. gonorrhoeae, but the results of redetermination by the SDA method tended to coincide with those of the PCR method. Urine samples tested by the SDA method were positive for N. gonorrhoeae even in patients in whom the culture of secretions from the male urethra was negative for N. gonorrhoeae. Based on these results, the BD ProbeTecET (SDA method) was confirmed to have the equivalent capability to the PCR method for the detection of C. trachomatis and N. gonorrhoeae in urine samples.     

Performance of strand displacement amplification assay in the detection of Chlamydia trachomatis and Neisseria gonorrhoeae

This study is a critical analysis of certain amplification assays for detecting Chlamydia trachomatis and Neisseria gonorrhoeae infections which have demonstrated that the plasmid-free variant of C. trachomatis is frequently responsible for infection in our patients. Specifically, we evaluated the performance of the strand displacement amplification (SDA) assay in detecting either C. trachomatis or N. gonorrhoeae in 1,190 clinical samples, both urogenital and ocular, from 1,005 consecutive patients. The results obtained with the BDProbeTec ET System were compared with three referenced amplification methods for C. trachomatis (detecting the 16S rRNA gene, the omp1 gene and the plasmid of C. trachomatis) and with both the culture method as well as an amplification assay followed by genetic identification performed using the MicroSeq 500 16S ribosomal DNA-based system for N. gonorrhoeae. The sensitivity of SDA (76%) in detecting C. trachomatis is significantly low when compared with that of other molecular techniques employing 16S rDNA or omp1 as a target. The specificity of the methods for detecting C. trachomatis was excellent, ranging from 99.4 to 100%. Furthermore, the results of SDA in detecting N. gonorrhoeae also provided excellent results (100% specificity and sensitivity).     

Comparison of polymerase chain reaction and enzyme immunoassay (Chlamydiazyme) in detection of Chlamydia trachomatis from male urethritis]

A method using polymerase chain reaction (PCR) was compared to an enzyme immunoassay (Chlamydiazyme) for detection of Chlamydia trachomatis by testing a reference strain and clinical specimens. Two oligonucleotides based on sequences within the major outer membrane protein gene from C. trachomatis serovar L2 were used as primers for the PCR. A DNA fragment of 242 bp specific for C. trachomatis was amplified by the PCR, when DNA of greater than or equal to 10(2) C. trachomatis was used as template for the PCR. A chlamydial antigen was detected by Chlamydiazyme, when greater than or equal to 2.6 x 10(3) C. trachomatis were applied for the enzyme immunoassay. The PCR method was 26 times more sensitive than Chlamydiazyme in detection of C. trachomatis. The PCR method and Chlamydiazyme were carried out to examine 74 urethral swabs obtained from male patients with urethritis for detection of C. trachomatis. In 45 of 74 specimens, the DNA fragment of C. trachomatis was amplified by the PCR, and in 41 of 74, the chlamydial antigen was detected by Chlamydiazyme. The detection rate of the PCR method (60.8%) was higher than that of Chlamydiazyme (55.4%). The positive coincidence rate of the PCR method to Chlamydiazyme was 100% (41/41) and negative coincidence rate was 87.9% (29/33). The overall coincidence rate between the two methods was high (94.6%). Thus, the PCR method was more sensitive than Chlamydiazyme for detection of C. trachomatis and specific for diagnosis of chlamydial urethritis.     

Evaluation of amplicor chlamydia PCR and LCX chlamydia LCR to detect Chlamydia trachomatis in synovial fluid

OBJECTIVES: PCR has been successfully used in research for the detection of C. trachomatis DNA in synovial samples. However, each research laboratory has developed its own PCR, making inter-laboratory comparisons difficult. To allow for standardization we evaluated two commercially available amplification systems originally designed for the examination of urogenital samples (Roche Amplicor Chlamydia PCR and Abbott LCX Chlamydia LCR), using them to analyse spiked and clinical synovial fluid (SF) samples from reactive arthritis (ReA), undifferentiated arthritis (UA), and rheumatoid arthritis (RA) patients. We compared their sensitivity in assays of clinical SF samples with our in-house developed C. trachomatis specific nested PCR. METHODS: SF was spiked with purified C. trachomatis elementary bodies (EB) and analyzed by the commercial assays. Clinical SF samplesfrom ReA (n=21), UA (n=79) and RA (n=50) patients were examined by the two commercial assays and our in-house PCR. RESULTS: Using SF samples spiked with defined numbers of C. trachomatis EB, the sensitivity of the commercial tests was high and similar to published PCR sensitivity. In clinical SF specimens the commercial assays was also able to detect CT; however, the in-house PCR was more sensitive. Out of 10 PCR-positive SF samples Amplicor tested positive in only 4/10 and LCX in only 3/10. The in-house PCR detected chlamydial DNA in synovialfluidfrom 5/21 ReA (24%), 5/79 UA (6%) and in none of the 50 RA patients. CONCLUSION: Commercial amplification assays allow the detection of C. trachomatis in clinical specimens, although with a lower sensitivity than optimized PCR. Potential explanations are discussed.     

Usefulness of real-time PCR in detecting Chlamydia trachomatis and Neisseria gonorrhoeae in endocervical swabs and first-voided urine specimens

We evaluated performance of Abbott RealTime CT/NG assay (real-time PCR, Abbott Japan) for detect Chlamydia trachomatis and Neisseria gonorrhoeae by real-time PCR in 88 female patients with cervicitis symptoms seen at gynecological clinics and 100 male patients with urethritis symptoms seen at urological or dermatology clinics in Kitakyushu, Japan. Endocervical swab and first-voided urine (FVU) specimens were then collected from women and FVU specimens from men. Detection rates of C. trachomatis and N. gonorrhoeae by real-time PCR in the 3 types of specimens were compared to those by ProbeTec ET assay (ProbeTec, BD Diagnostic System). The overall positive concordance between real-time PCR and ProbTec were 97.1% (66/68) for C. trachomatis and 100% (33/33) for N. gonorrhoeae, C. trachomatis detection yielded 3 discordant results in endocervical specimens and 1 discordant result in male FVU by real-time PCR and ProbTec. Three of 4 reexamined using Aptime Combo 2 Assay (Fuji Rebio Inc.) were positive for C. trachomatis. Endocervical swab and FVU specimen results for C. trachomatis were discordant in 3 cases in real-time PCR and 4 in ProbeTec. Subjects with 2 or more positive endocervical awab results in female or male FVU specimens were assumed to be "true positive" for C. trachomatis. The sensitivities of real-time PCR for detecting C. trachomatis was 94.4% in endocervical swabs, 77.8% in female FVU and 97.4% in the male FVU. The sensitivities for real-time PCR for detecting N. gonorrhoeae was 100% in all 3 specimentypes. Abbott RealTime CT/NG assay was useful for detecting C. trachomatis using endocervical swabs or male FVU specimens and for detecting N. gonorrhoeae using endocervical swabs and all FVU specimens.     

Comparison of polymerase chain reaction and IDEIA Chlamydia in detection of Chlamydia trachomatis from first-voided urine of male urethritis patients]

We have reported a method for detection of Chlamydia trachomatis by polymerase chain reaction (PCR) with two oligonucleotides based on sequences within the major outer membrane protein gene from C. trachomatis serovar L2. In the previous report, in addition to treatment of the mixture of first-voided urine (FVU) sediment and 1 ml of urine with proteinase K. DNA purification by phenol extraction was necessary for preparation of template DNA for PCR. In this study, FVU sediment was suspended in 1 ml of Chlamydiazyme dilution buffer and a part of the suspension was treated with proteinase K for DNA extraction. The DNA extraction solution could be used as template for PCR without purification of DNA by phenol extraction. One hundred FVU specimens obtained from male urethritis patients were examined with the two methods (PCR and IDEIA) for detection of C. trachomatis. In 33 of 100 specimens, the DNA fragments of C. trachomatis was amplified by the PCR and in 32 of 100, the chlamydial antigen was detected by IDEIA. The positive and negative coincidence rate of the PCR to IDEIA were 93.8% (30.32) and 95.6% (65/68) respectively, resulting in a high overall coincidence rate at 95%. Thus, the improved method with PCR using FVU as a specimen is proved to be a useful, non-invasive diagnostic tool for diagnosis of chlamydial urethritis.     

A reappraisal of chlamydial and nonchlamydial acute non-gonococcal urethritis.

A cohort of 112 men presenting with acute non-gonococcal urethritis (NGU) was investigated for the presence of Chlamydia trachomatis. Men with 3 or more episodes of NGU in the preceding 12 months, or who had received treatment for NGU in the preceding 3 months were excluded. C. trachomatis was sought by examination of urethral smears by direct immunofluorescence, and by examination of the centrifuged deposit from a first pass urine (FPU) sample by direct immunofluorescence, IDEIA, and the polymerase chain reaction. Urethral samples from 48 men were positive for CT, and the FPU samples from an additional 7 men were positive by at least 2 assays. With such intensive investigation it is likely that those men identified as chlamydia-negative were genuinely free from the infection. The clinical history and response to treatment of those men who were chlamydia-positive were compared with those of the chlamydia-negative men. They differed in that a larger proportion of the chlamydia-positive men reported having had intercourse with more than one partner in the previous 3 months, and having had fewer previous episodes of NGU. Moreover, in contrast to some previous studies, after one week of treatment with doxycycline, a larger proportion (65%) of the chlamydia-negative men than the chlamydia-positive men (40%) was cured, although the difference was not sustained following later treatment.     

Mycoplasma genitalium in non-gonococcal urethritis--a study in Swedish male STD patients

Urethral swab specimens obtained from 101 men attending an STD clinic were examined for the presence of Mycoplasma genitalium by polymerase chain reaction (PCR) amplification. Fifty patients had non-gonococcal urethritis (NGU), and 51 patients were included as controls without urethritis. M. genitalium DNA was detected in 13 (26%) of the urethritis patients and in 5 (10%) of the control patients (P=0.06). No patient positive for M. genitalium had a simultaneous chlamydial infection. Thus, in the 36 patients with non-chlamydial NGU, the prevalence of M. genitalium infection was 36% (P=0.007 compared with controls). All patients with M. genitalium positive urethritis had a high grade urethritis defined as >10 polymorphonuclear cells per high power microscopical field. Compared with the control group, those with M. genitalium positive urethritis had more often had a history of urethritis than had those with chlamydial NGU or those with M. genitalium negative, non-chlamydial NGU.     

Simultaneous detection of Neisseria gonorrhoeae and Chlamydia trachomatis by PCR in genitourinary specimens from men and women attending an STD clinic

Neisseria gonorrhoeae and Chlamydia trachomatis are the two most common bacterial sexually transmitted infections that manifest primarily as urethritis in males and endocervicitis in females, though the infection may be asymptomatic especially in women. Since complications may occur in untreated symptomatic and asymptomatic infected individuals, early diagnosis and treatment of infected individuals is required to prevent severe sequelae and spread of these diseases. Recently molecular amplification assays like Polymerase Chain Reaction (PCR) and Ligase Chain Reaction (LCR) have been found to be highly sensitive and specific methods for detection of N. gonorrhoeae and C. trachonmatis not only in urethral and cervical specimens but also in urine. The objective of this study was to screen male and female Sexually Transmitted Disease (STD) clinic attenders, with and without symptoms suggestive of urethritis and cervicitis for presence of N. gonorrhoeae and C. trachomatis using a multiplex PCR based assay, to compare its performance with culture for N. gonorrhoeae and Direct Fluorescent Antibody (DFA) staining for C. trachomatis and also to compare the efficacy of PCR test performed on urine and genital swab specimens collected from this high risk group. Genital specimens and urine was collected from STD clinic attenders. N. gonorrhoeae and C. trachomatis was detected in genital specimens by culture and DFA respectively. Multiplex PCR was used to detect N. gonorrhoeae and C. trachomatis infection in both genital and urine specimens. Among men with urethritis, N. gonorrhoeae was detected in 70% by culture and 77% by PCR, while C. trachomatis as detected in 7.5% by DFA and 17.5% by PCR. Among females with endocervicitis, N. gonorrhoeae was detected in 7.7% by culture and 30.7% by PCR, while C. trachomatis was detected in 7.7% by DFA and in 15.4% by PCR. None of the asymptomatic males were positive for N. gonorrhoeae and C. trachomatis by conventional methods, while 43.9% were positive for N. gonorrhoeae and 7.5% for C. trachomatis by PCR. Fifty per cent of asymptomatic women were positive for C. trachomatis by PCR alone. We encountered PCR positive but culture/DFA negative results and also PCR negative but culture/DFA positive results. In view of this a single PCR test cannot be used for diagnosis and treatment of N. gonorrhoeae and C. trachomatis infection unless confirmed by a second test.     

Qualitative and quantitative aspects of the ligase chain reaction assay for Chlamydia trachomatis in genital tract samples and urines

The performance of the ligase chain reaction (LCR) assay for Chlamydia trachomatis was evaluated in a genitourinary medicine (GUM) clinic population. Its sensitivity was 100%, 91% and 95%, respectively, for cervical, vaginal and urine samples from 417 women, when compared with direct fluorescent antibody (DFA) staining of cervical samples, and 100% and 91%, respectively, for urethral and urine samples from 317 men, when compared with DFA staining of urethral smears. An enzyme immunoassay (EIA) was only 65% sensitive for cervical samples. Urethral swabs from a number of treated men remained LCR-positive when antigen was no longer detectable by DFA staining. An association between quantitative data from the LCR assay (i.e. the optical density of samples, measured in relation to internal controls and calibrators) and the antigen load of the samples, measured by DFA staining, indicated a lack of significant inhibition in the LCR assay in this study. This was probably due to freezing of the samples before testing. Diluting 20 LCR-positive urines with a range of antigen loads resulted in loss of positivity in 3, and a reduction in the signal in 13. The implications of the antigen load on the performance of detection assays for chlamydia-positive patients are discussed.     

Detection of Chlamydia trachomatis in first-voided urine sediments from male urethritis by polymerase chain reaction]

Chlamydia trachomatis was detected from first-voided urine sediments of 97 male patients with urethritis by polymerase chain reaction (PCR). Since urine and urinary sediments only treated with proteinase K inhibited DNA amplification by PCR, DNA was further purified by phenol extraction and concentrated. Two oligonucleotides based on sequences within the major outer membrane gene from C. trachomatis serovar L2 were used as primers. A DNA fragment of 242 bp specific for C. trachomatis was amplified by PCR and detected by agarose gel electrophoresis. The DNA fragment was amplified by PCR in all specimens of urine sediments from 50 patients with Chlamydiazyme-positive urethral swab. In 38 specimens of urine sediments from 47 patients with Chlamydiazyme-negative urethral swab, PCR was negative. The overall coincidence rate between the PCR for detecting C. trachomatis in first-voided urine sediments and Clamydiazyme in urethral swab was 90.7% (88/97). Detection of C. trachomatis from first-voided urine sediments by PCR was considered to be noninvasive and useful for the diagnosis of male urethritis due to C. trachomatis.     

Which test is best for chlamydia?

Nucleic acid amplification tests are now the tests of choice for diagnosing Chlamydia trachomatis infection. For the first time there are diagnostic tests for Chlamydia trachomatis that are more sensitive than tissue culture. Another major advantage is that they can be used with first-catch urine specimens and vaginal swabs. It is thus possible to test for genital chlamydial infection without using invasive specimen collection methods.     

Chlamydia trachomatis genital infection in apparently healthy adult population of Tamil Nadu, India: a population-based study

Since the epidemiology of Chlamydia trachomatis infection in apparently healthy population has not been studied in India, a population-based study was conducted in the state of Tamil Nadu, India in order to analyse the prevalence of genital chlamydial infections in the community and to implement control programmes. A representative sample was taken from three randomly selected districts by using the 'probability proportional to size' cluster survey method. Households were the basic units of clusters. Adults aged 15-45 years, pre-identified from the selected households were enrolled during the medical camps conducted for a major study on community prevalence of sexually transmitted diseases in Tamil Nadu. Blood and urine samples collected from the study subjects were tested by enzyme-linked immunosorbent assay (ELISA) for anti-chlamydial IgM antibodies and by the commercial Amplicor polymerase chain reaction (PCR) test for chlamydial DNA. The prevalence of anti-C. trachomatis antibodies determined by IgM-ELISA was 2.4% (95% CI 1.6%-3.2%). The prevalence of genital chlamydial infection determined by PCR was 1.1% (95% CI 0.5%-1.7%). Majority of the detected infections (68.8%) were asymptomatic. This is the first Indian report on the prevalence of genital chlamydial infections in the general population. It is concluded that this study provides evidence for a substantial burden of approximately 10 million asymptomatic genital chlamydial infection cases in the sexually active age groups in the general population of India.     

Molecular testing (strand displacement assay) for identification of urethral gonorrhoea in men: can it replace culture as the gold standard?

The objective was to evaluate the performance of Becton Dickinson's BD Probe Tec(TM) (BDPT) strand displacement amplification (SDA) test for the detection of Neisseria gonorrhoeae on urethral specimens from men with urethritis compared with conventional culture and to show that SDA improves the diagnostic yield of gonorrhoea infections (GC). Anonymized retrospective testing of stored urethral swab samples from men attending genitourinary services in East London was performed using SDA. The prevalence of GC culture positive infections in this sample was 20/152 (13%). The sensitivity, specificity, positive predictive value and negative predictive value for the BDPT-SDA system compared with culture were 100%, 95%, 77% and 100%, respectively. In this study population, the BDPT-SDA assay was a highly sensitive and specific test for the diagnosis of N. gonorrhoeae from urethral swabs in men. Therefore, SDA can be used to complement culture in the diagnosis of N. gonorrhoeae infection. No ethics committee approval was obtained as all samples were anonymized.     

Screening for genital chlamydia infection: DNA amplification techniques should be the test of choice

Our objective was to compare the sensitivities for the detection of Chlamydia trachomatis, of the ligase chain reaction (LCR) on first voided urine (FVU) specimens and enzyme immunoassay (EIA) on pooled endocervical/endourethral swabs from women and endourethral swabs from men. Men and women taking part in the UK chlamydia screening pilot were tested for chlamydia using LCR on a FVU. Patients attending genitourinary medicine clinics also had cervical and/or urethral swabs taken for chlamydia testing by EIA. In women, EIA on pooled swabs detected 575 of the 785 chlamydia positives and in men, EIA detected 209 of 351 positives. The sensitivity of EIA was 73% and 60% in women and men respectively. By using the EIA test, therefore, 27-40% of patients infected with chlamydia will be given a false negative result. We propose that it is unethical to use non-molecular testing in the future.     

Comparison of synovial tissue and synovial fluid as the source of nucleic acids for detection of Chlamydia trachomatis by polymerase chain reaction

Objective. Difficulties in detecting Chlamydia trachomatis in human joints by polymerase chain reaction (PCR) may be related to whether synovial tissue or synovial fluid (SF) is used as the source of DNA in PCR amplification. In this study, a new PCR assay was developed and used to compare chlamydial DNA in paired samples of SF and synovial tissue from patients with arthritis. Methods. The PCR assay targeted the ribosomal RNA operons, which are present in 2 copies on the C trachomatis chromosome. DNA from several relevant bacteria and chlamydial serovars was used for testing this screening system. The detection of chlamydial DNA in nucleic acid preparations from matched samples of SF and synovial tissue was compared by PCR assay. Samples were obtained from 55 patients, including patients with reactive arthritis, Reiter's syndrome, and other arthropathies. Results. Testing of the PCR screening system confirmed it to be highly specific and sensitive. Use of this assay to screen DNA from SF and synovial tissue samples showed that 29 (53%) of 55 synovial tissue preparations were positive for chlamydial DNA, but only 16 (29%) of the matched SF samples from these 29 patients were similarly positive. Five (9%) of 55 SF samples, but not their tissue counterparts, were positive for chlamydial DNA by PCR. Conclusion. Detection of chlamydial DNA in the joints of patients by PCR gives positive results more often when synovial tissue rather than SF is the source of target nucleic acids. Although synovial tissue is the source of choice for the most reliable determination of chlamydia in the joint, both synovial tissue and SF should be assayed if possible.