New variant chlamydia: Limitations of diagnostic tests and importance of epidemiologic surveillance

Abstract

Introduction: The introduction of nucleic acid amplifi cation tests for the diagnosis of Chlamydia trachomatis infections in the 1990s represented the fi rst wholesale, practical substitution of traditional diagnostic methods in microbiology with this new technology. Prior to this, Chlamydia identifi cation in clinical specimens was often not routinely available or, if available, was dependent on tests with poor sensitivity. In addition, nucleic acid amplifi cation tests provide effi ciency of scale through batch testing, and they have proven to be nearly as reliable on urine specimens and self-collected vaginal specimens as they are on cervical and urethral specimens. Not surprisingly, with the introduction of reliable nucleic acid amplifi cation tests and expansion of Chlamydia testing services, the numbers of cases of genital C. trachomatis infections in the United States began to soar. Between 1990 and 2005, overall rates of Chlamydia infection in the United States went from ~ 150/100,000 population to ~ 340/100,000 population [1]. Several types of nucleic acid amplifi cation tests have been commercially introduced to detect Chlamydia. These include polymerase chain reaction (PCR), ligase chain reaction (LCR), and transcription-mediated amplifi cation (TMA). By their nature, the reliability of these tests depends on the selection of the appropriate target gene sequence and the conservation of that sequence over time. Some testing systems, including the Abbott m2000 (Abbott Diagnostics, Chicago, IL, USA) and the Cobas Amplicor/TaqMan48 (Roche Diagnostics, Basel, Switzerland), use the chlamydial cryptic plasmid as their target for amplifi cation. As the name implies, the actual function of this plasmid is unclear, but it provides several copies per cell and therefore results in increased sensitivity when compared with those tests that target chromosomal DNA [2]. However, what happens if and when, through random mutation, variant organisms appear that are lacking the sequence in question? It has long been recognized that plasmid-free strains of C. trachomatis exist in nature [3]. This did not seem to have any impact at the time on test system design. Apparently, however, now the clinical “chicken has come home to roost.” In Sweden, after a decade of rising C. trachomatis rates, it was observed that at least in one southern county in 2005 the rates declined by 25%. Rather than accept this as evidence of successful surveillance and disease management, investigators assumed there was a good chance that the diagnostic tests they were using were either faulty or that a change had occurred in the target for amplifi cation. After some study, a 377-bp deletion was found in the cryptic plasmid target region of a strain that initially had not been detected by routine testing. Had this occurred in the setting of anything less than effective surveillance, and the noted observation of an unexpected decline in prevalence after years of rising rates, it may not have come to light. Several studies and reports recently have highlighted this microbiologic phenomenon.