Geisler et al1 have typed strains of Chlamydia trachomatis from 102 individuals in Alabama who tested positive by culture and/or PCR at baseline, and again when they returned for treatment up to 60 days later, by sequencing the ompA gene. Ninety per cent of the subjects were female and 88% were African American. The most commonly identified genotype was E (28%), as has been found in other studies, followed by D/Da (23%) and J/Ja (19%). Subjects infected with genotype J/Ja were significantly more likely to be culture negative, suggesting that they might have a lower bacterial load than found with other genotypes and were more likely to clear infection spontaneously (p = 0.07).

Typing isolates of C trachomatis is potentially of great value. It could help to map sexual networks and to distinguish between treatment failure and reinfection in clinical trials. If associations could be found between particular strains and particular clinical findings, it could help to identify virulence determinants of C trachomatis, thereby increasing our understanding of the pathogenesis of C trachomatis infection.

The first typing method for C trachomatis, the micro-immunofluorescence test of Wang and Grayston, was developed more than 30 years ago.2 It was based on the ability of polyclonal antisera to distinguish 13 (later increased to 17) serotypes of C trachomatis. The serotype-specific epitopes were later shown to be on the major outer membrane protein, which is encoded by the ompA gene targeted by Geisler et al.1 Since the early 1990s, when PCR became widely available, many studies have been published in which genital and ocular strains of C trachomatis were genotyped following amplification of the ompA gene, either by sequencing or by restriction fragment length polymorphism analysis of the amplified product. The problem with these studies has been that ompA genotyping has not been sufficiently discriminatory to distinguish between persistent infection on the one hand and reinfection with a common genotype on the other. Moreover, most studies have used methodology that is not able to identify mixed infections. Perhaps because of these limitations, no consistent correlation has been found between the ompA genotype and clinical findings.3 4

Two new typing methods have been described recently. Klint et al5 have used multilocus sequence typing, targeting six variable genes identified through genome sequencing projects. Using this method, they were able to identify 32 genetic variants among 47 Swedish isolates, whereas ompA sequencing was only able to identify 12. Pedersen et al6 sequenced both the ompA gene and three loci with variable number tandem repeats; this method was also shown to be highly discriminatory. A recent study by Kari et al7 has identified, in a non-human primate model, genetic variations in six genes of C trachomatis that are associated with increased virulence among ocular strains. Future studies, using multilocus sequence typing and focussing on these variable genes, are likely to shed much light on the epidemiology and pathogenesis of C trachomatis infection in the near future.