It is important that all men who have sex with men (MSM) accessing sexual healthcare are tested for Neisseria gonorrhoeae (GC) and Chlamydia trachomatis (CT) at all anatomical sites where they may be at risk of infection. In the United Kingdom, culture is considered the diagnostic gold standard for GC at extragenital sites owing to its high specificity, while nucleic acid amplification tests (NAATs) are preferred for CT. However, sometimes rectal CT testing is only offered to symptomatic patients and often testing for pharyngeal CT infection is not performed at all because of both technical and financial concerns.1

In this issue of STI Ota et al present the results of a study examining the performance of culture and NAATs for the detection of CT and GC in extragenital sites in a high-prevalence group of MSM.2 Worryingly, very low sensitivity data for both CT and GC culture in rectal sites (21% and 41.4% respectively) and pharyngeal sites (0% for both agents) are reported. While the authors stress that their culture facilities may be suboptimal they are probably representative of many clinical settings and this highlights the problems with using culture as a diagnostic test. This is particularly relevant when using culture to detect STIs in extragenital sites where the infectious agents are fastidious in nature, present in low numbers and may be masked by high levels of commensal bacteria. To date the highest reported sensitivity data for culture in the rectum are 27% (CT) and 86.4% (GC), and in the pharynx they are 44% (CT) and 41% (GC), so even with the best culture facilities available a significant number of infections will go undetected.3 4 An undiagnosed reservoir of CT and GC infection in this high-risk group is unacceptable given the risk of onward transmission and the potential development of sequelae. Ultimately, low sensitivity and long turn-around times make culture an inappropriate and undesirable diagnostic or confirmatory test for CT and GC in rectal and pharyngeal sites, as cases will be missed. But what is the alternative?

A key message of the study by Ota et al is that NAATs detect CT and GC in extragenital sites with superior sensitivity compared to culture; however, it is the specificity of NAATs that has previously been questioned, particularly for GC. It is known that the different commercial NAAT platforms are not equal,5 and often the most striking difference in performance is the range of cross-reactivity observed in the GC component of the test with different commensal Neisseria species. While the Amplicor PCR (Roche) and the Probetec strand displacement assay (SDA, Becton Dickinson) have been reported to cross-react with other non-gonococcal Neisseria species to varying degrees,6 there have been no similar reports for either the Aptima Combo 2 (AC2) or the Aptima Monospecific GC test (AGC) (Genprobe). The main surprise of the study by Ota et al is that high specificities for both infections but particularly for GC, were achieved when testing rectal and pharyngeal specimens using the SDA and the AC2 tests. Indeed, it is probable that the actual specificity of the AC2 test was even higher than reported, as 100% of the positive AC2 results could be confirmed with the Aptima monospecific CT and GC kits, respectively, but despite this extra level of confirmation, a small number were recorded as false positives owing to the conservative definition of “true positive”. It was also an interesting finding that while there were equal numbers of AC2 and SDA positive specimens (11) which were not confirmed using the alternative manufacturer’s test: for the SDA specimens these were largely associated with positive pharyngeal GC results, which may be suggestive of cross-reactivity with commensal Neisseria species which are inevitability present in higher numbers at this site. Even when taking this into account and using a rigorous gold standard Ota et al report a high specificity for NAATs. So while NAATs were seen to have a small reduction in specificity, when compared with culture (98% vs 100%), they did detect 4.6 times more cases of infection in this group.

It is important when implementing any new diagnostic testing strategy to consider the patients being tested. Ota et al examined a high-risk, high-prevalence group of MSM attending for sexual healthcare. It is because of this high prevalence of infection that high positive predictive values (PPV) could be obtained and consequently it is the logical conclusion that NAATs should be the method of choice for testing extragenital sites in this group. However caution should be applied before extrapolating the same conclusions to other groups where the prevalence may be lower and also to other NAATs not included in this study. It is noteworthy that Schachter et al, when performing a head-to-head comparison of the NAAT platforms to test extragenital specimens, had to remove the Amplicor PCR test from the evaluation early owing to the high numbers of false-positive results in oropharygneal swabs.4

It has been recommended that an alternative way of ensuring high PPVs when using NAATs is to perform confirmatory testing on all positive specimens (especially GC positives), using a another NAAT with an alternative gene target.7 Using such an approach would greatly reduce the chance of misdiagnosis and should be encouraged where at all possible. However in the average laboratory environment, accurate confirmatory testing can be difficult to achieve. At the present time, Genprobe is the only platform with the capacity to carry out confirmatory testing and even then both the AC2 and the monospecific tests actually target different regions of the same gene. Most diagnostic laboratories do not actually have access to more than one NAAT platform and even if they do the potential incompatibility of different platform chemistries would require multiple specimens for accurate confirmation. Also differences in the analytical sensitivity of the different NAATs may make the interpretation of a true positive result when using confirmatory testing problematic and, consequently, patient management difficult.

Clearly, tackling the undiagnosed reservoir of infection in MSM groups will only be possible if all sexual health services are also able to overcome the financial hurdles and offer NAAT-based extragenital STI testing. Previous clinical audits have highlighted the inconsistent nature of testing in this patient group and the wide disparity of care received.1 Failure to perform STI testing at extragenital sites is a false economy as the prevalence of infection is often much higher in these sites than at the urethra. Ota et al report that over two-thirds of infections would have been missed if urethral-only screening had been performed. And given that recent studies have shown self-taken rectal and pharyngeal specimens are of equal quality to those taken by clinicians,8 the failure to test MSM for extragenital infection is hard to justify. Ultimately, however, it is only possible for clinics to offer such a service if they are fully supported by their local microbiology laboratory. To date none of the commercial NAAT platforms are approved by the Food and Drug Administration (FDA) or CE marked for use on extragenital specimens. This represents a real problem because as Ota et al have shown, the application of highly sensitive tests is essential in order to detect the majority of infections, but at the present time there is no strong financial incentive for the NAAT manufacturers to obtain approval for these specimen types. In the United Kingdom, at least, diagnostic microbiology laboratories are able to use CE-marked tests to screen unauthorised specimen types provided that they have sufficient evidence-based validation data to justify their testing approach. It is hoped that the study conducted by Ota et al will go some way to fulfilling this requirement and will also inspire local validation studies to be undertaken within UK MSM groups to ensure that similarly high positive predictive values can also be achieved.

In patient groups at highest risk of acquiring an STI, such as MSM, it is of paramount importance that adequate testing methods are available to detect infection at all anatomical sites where they may be at risk. Ota et al have provided timely and creditable validation data which, together with that provided by others, supports the use of NAATs for detecting extragenital CT and GC infections in high prevalence groups and discourages the use of culture.2^–4 8 It is possible that the long-term use of inadequate testing methods and the ongoing failure to offer testing at extragenital sites may be contributing to the high burden of infection in this patient group.