Detection of Chlamydia trachomatis in genital swabs: comparison of commercial and in house amplification methods with culture.

AIMS: To evaluate the sensitivity of the Roche Cobas, Roche Amplicor plate kit, ligase chain reaction (LCR), and an in house polymerase chain reaction (PCR) by titration of purified elementary bodies (EB) and also to test 245 urethral and endocervical specimens for Chlamydia trachomatis by the four assays as well as conventional culture. STUDY DESIGN: EB titrations were run in duplicate in each commercial assay and six times in the in house PCR. Clinical samples were aliquoted and tested by each assay and were considered positive if C trachomatis was detected by two or more separate tests or if the sample was either culture or immunofluorescence positive. Major outer membrane protein (MOMP) specific primers were used as a confirmatory assay for the in house PCR. RESULTS: The in house PCR, Roche Cobas Amplicor, LCR, and Amplicor plate kit gave detection limits of approximately 1, 1-2, 2, and 2-4 EBs respectively. By the criteria described above for definition of a C trachomatis positive result in clinical samples we identified 23 true positives among the 245 clinical specimens. The in house PCR detected all 23 giving a sensitivity of 100% and a specificity of 98%. The Roche Cobas Amplicor, Roche Amplicor plate kit, and LCR detected 21, 19, and 19 of these respectively giving sensitivities of 87.5%, 82%, and 82% respectively and specificities of 99.5%, 99%, and 100% respectively. The culture gave a sensitivity of 78% and specificity of 100%. CONCLUSION: All four amplification assays had a greater sensitivity than the culture used routinely in this laboratory. The in house plasmid PCR had the greatest sensitivity and when combined with confirmation by immunofluorescence detected the greatest number of positives. This increased sensitivity is likely to have been achieved by the use of a DNA purification step and of nested primers in the amplification stage and their combined use in routine diagnostic assays for chlamydia might increase the frequency of C trachomatis detections. However, this assay is much less user friendly than the two semiautomated commercial assays investigated in this study.