Objectives: This study compared the sensitivity and specificity of culture and two nucleic acid amplification tests (NAATs): the BD Probetec ET system (PT) and the Aptima Combo 2 (AC2) in detecting Neisseria gonorrhoeae (GC) and Chlamydia trachomatis (CT) in pharyngeal and rectal specimens.
Methods: Male subjects were prospectively recruited at an MSM clinic in Toronto, Canada. Pharyngeal and rectal specimens were obtained for GC and CT culture, PT and AC2. Urine was also obtained for PT. A true positive was defined as: (1) positive culture, (2) positive PT and AC2 at the same site or (3) a single positive NAAT and detection of the same organism by any method at another site.
Results: 248 subjects were recruited. The prevalence of pharyngeal GC was 8.1%, rectal GC 11.7%, pharyngeal CT 2.0% and rectal CT 7.7%. The sensitivity of culture for pharyngeal GC and CT was 0%; 41.4% for rectal GC and 21.1% for rectal CT. The sensitivity of PT for pharyngeal GC, rectal GC, pharyngeal CT and rectal CT was 95.0%, 93.1%, 80.0% and 94.7%, respectively. The sensitivity of AC2 was 95.0% for pharyngeal GC and 100% at all other sites. Specificity was consistently above 98%.
Conclusions: PT and AC2 detected GC and CT with superior sensitivity compared to culture. They detected 73 pharyngeal or rectal GC and CT infections compared to 16 by culture, using a rigorous gold standard. NAATs should be the method of choice for the detection of GC and CT in extragenital sites in men who have sex with men.
Statistics from Altmetric.com
Nucleic acid amplification tests (NAATs) are the most sensitive tests available for the diagnosis of urogenital Neisseria gonorrhoeae (GC) and Chlamydia trachomatis (CT) infections in both males and females. As NAATs have become increasingly accepted for use in endocervical, male urethral, male urine1–5 and for some assays, female urine2 3 and self-collected vaginal specimens,6 studies have also been conducted to evaluate the use of NAATs for the detection of GC and CT in pharyngeal and rectal specimens, particularly in men who have sex with men (MSM). MSM are especially at risk because of high rates of unprotected anal intercourse and fellatio.7 8
While there are a number of published studies that have demonstrated the superior performance of NAATs over traditional culture in extragenital specimens,9–19 their impact has been limited by small sample size or the use of the now defunct ligase chain reaction assay.20 Furthermore, there has been only one other study utilising the currently available commercial NAATs for detection of GC and CT in extragenital sites.9
In the absence of Food and Drug Administration (FDA) approval of NAATs for these indications, the Centers for Disease Control and Prevention (CDC) in the United States21 and the Public Health Agency of Canada22 have continued to recommend culture techniques for the detection of GC and CT in extragenital sites despite their poor sensitivity.
In addition, concerns continue to be raised about the specificity of NAATs and their low positive predictive value (PPV) in low-prevalence settings.23 The CDC has addressed this issue by describing three confirmatory strategies in settings where the PPV is estimated to be less than 90%: (1) repeating the same NAAT on the same specimen, (2) performing a different NAAT (with a different target) on the same specimen or (3) performing a different NAAT using a separate specimen collected at the same time.23 Evaluation of these confirmatory strategies will need to occur in tandem with the evaluation of NAAT performance in extragenital sites.
This prospective study compared the sensitivity and specificity of culture, the BD Probetec ET system (PT) and the Gen-Probe Aptima Combo 2 assay (AC2) in the detection of GC and CT in pharyngeal and rectal specimens in MSM. In addition, laboratory and clinical parameters were compared for their ability to confirm positive NAAT results.
Male subjects presenting for symptomatic testing or asymptomatic screening for sexually transmitted infections (STI) were recruited prospectively at the Hassle Free Men’s Clinic in Toronto, Canada, between 1 December 2006 and 31 January 2008. Subjects were also eligible if they were presenting to receive treatment as an STI contact. Potential subjects were excluded if their sexual contact was exclusively with women or if they were less than 18 years of age.
This study was approved by the research ethics board at the Hospital for Sick Children in Toronto and written consent was obtained from all study participants.
Specimen collection and transport
Clinicians collected four pharyngeal and four rectal specimens from each subject for the following tests: (1) GC and CT testing by PT, (2) GC and CT testing by AC2, (3) GC culture and (4) CT culture. At each site, PT and AC2 specimens were obtained by inserting a BD ProbeTec ET Wet Swab and an Aptima Unisex Swab simultaneously. This was followed by a Starswab II charcoal swab for GC culture and then a Starswab Multitrans swab for CT culture. Pharyngeal specimens were obtained by swabbing both tonsillar pillars and the posterior pharynx. Rectal specimens were obtained by using an anoscope and swabbing the rectal mucosa at a depth of approximately 1 inch (2.5 cm) beyond the end of the anoscope. Finally, a urine specimen was obtained for GC and CT testing using PT, the system currently in use at the Public Health Laboratory, Ontario Agency for Health Protection and Promotion (formerly Ontario Ministry of Health and Long-Term Care). A urethral specimen was obtained for GC culture if the subject reported urethral discharge.
GC culture, CT culture and PT specimens were refrigerated until transport (at room temperature) to the Public Health Laboratory within 48 hours of collection. AC2 specimens were also refrigerated until transport to St Joseph’s Healthcare in Hamilton, Ontario and processed within the manufacturer’s 60-day window.
GC culture specimens were plated on to New York City agar and incubated in CO2 at 37°C and held for 72 hours. Identification was confirmed by characteristic colonial morphology and Gram stain; positive oxidase, negative ortho-nitrophenyl-β-galactoside (ONPG), evidence of CTA glucose fermentation and lack of CTA maltose or sucrose fermentation. The Accuprobe (Gen-Probe, San Diego, CA, USA) Neisseria gonorrhoeae culture identification test was used to confirm identification.
CT isolation involved inoculation of untreated McCoy cell monolayers in 96-well microtitre plates. A 1:20 dilution of amphotericin B (250 μg/ml) stock was added to each specimen before inoculation to prevent fungal contamination. Each specimen was vortexed and inoculated into two wells to achieve dilutions of 1:2 and 1:4 in growth medium. The inoculated microwell plates were centrifuged at 1600 g for 60 minutes at room temperature and then incubated for 60 minutes at 36°C in 5% CO2. Growth medium was then discarded and replaced with fresh medium. Plates were incubated for 48 hours at 36°C in 5% CO2. Monolayers were stained using the Pathfinder Chlamydia culture confirmation system (Bio-Rad Laboratories Inc, Hercules, CA, USA) that utilises a genus-specific direct fluorescent monoclonal antibody. Identification was confirmed by the presence of characteristic fluorescent intracellular inclusions.
The BD Probetec ET CT/GC/AC is a NAAT that is based on strand displacement amplification technology. A fluorescent-labelled detector probe is used to detect amplified GC and CT DNA. An inhibitor control was performed with every test. The Aptima Combo 2 is a NAAT that is based on transcription-mediated amplification and involves replication of specific regions of 16S rRNA in GC and 23S rRNA in CT. A fluorescent-labelled detector probe is used to detect rRNA amplification product sequences. All positive PT and AC2 results were repeated.
Analysis of test performance
A “true positive” was defined as: (1) positive culture, (2) positive PT and AC2 at the same site or (3) a single positive NAAT and detection of the same organism by any method at another site (that is, pharynx, rectum or urethra). Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated. For these calculations, a NAAT result was considered positive if confirmed positive on the second run.
The McNemar test was used to compare detection of GC and CT using culture against each NAAT. A sample size of 250 with an overall culture detection rate of 6.0% was determined to yield 61.0% power to detect a diagnostic odds ratio of 2 or more for NAAT compared with culture and a 97.6% power for an odds ratio of 3 or more.
Confirmation of positive NAAT results
The following confirmatory strategies were evaluated for all positive NAAT results on the first run: (1) repeating the same NAAT on the same specimen, (2) performing a different NAAT on a separate specimen taken simultaneously, (3) identifying a concomitant infection with the same organism at a different site and (4) reviewing the patient chart for documentation of a sexual contact infected with the same organism. During the chart review of positive results, symptoms were also recorded.
Finally, individual Aptima (Gen-Probe) tests for GC (AGC) and CT (ACT) were performed on the AC2 specimens of all cases with discordant AC2/PT results. The AGC and ACT assays use different targets from the AC2 assay.
A total of 248 subjects were recruited during the study period. Data collection was complete in 100% of the study participants. Using the study definition of a true positive, pharyngeal GC was detected in 20 subjects (8.1%), rectal GC in 29 (11.7%), pharyngeal CT in 5 (2.0%), rectal CT in 19 (7.7%), urethral GC in 17 (6.8%) and urethral CT in 12 subjects (4.8%). The true positive cases are summarised in table 1. All cases with positive culture had positive results by both PT and AC2. Inhibition was noted with PT in one pharyngeal specimen and in one rectal specimen.
There was one case of pharyngeal GC and CT co-infection and eight cases of rectal co-infection. Fourteen subjects had true positive GC infections in both the pharynx and rectum; five of these cases were also infected in the urethra. Three subjects had true positive CT infections in both the pharynx and rectum but none of them had urethral infection. Urine and urethral specimens would have identified seven (35.0%) of the true positive pharyngeal GC cases, 10 (34.5%) of the rectal GC cases, none of the pharyngeal CT cases and four (21.1%) of the rectal CT cases if performed alone.
None of the subjects with pharyngeal GC and CT had pharyngeal symptoms. Two of 29 subjects with rectal GC and one of 19 subjects with rectal CT had rectal symptoms.
Performance of culture, PT and AC2
The sensitivity, specificity, PPV and NPV of culture, PT and AC2 are summarised in table 2. Cultures performed poorly and failed to detect any cases of pharyngeal GC or CT. In the rectum, the sensitivity of culture was 41% for GC and 21% for CT. The sensitivity of PT for GC and CT ranged between 93–95% with the exception of pharyngeal CT where the sensitivity was 80%. The sensitivity of AC2 was 95.0% for pharyngeal GC and 100% at all other sites. The specificity of PT and AC2 was consistently above 98%.
Both PT and AC2 were superior to culture for detection of pharyngeal GC (p<0.001), rectal GC (p<0.001) and rectal CT (p<0.001) but not for pharyngeal CT.
Confirmation of positive NAAT results
Tables 3 summarises the performance of the confirmatory strategies for specimens that tested positive by PT and AC2 on the initial run. Re-testing confirmed 74 of 78 (94.9%) positive PT tests and 100% of AC2-positive tests. With respect to the use of a second NAAT performed on a second, simultaneously obtained specimen, 67 of 78 (85.9%) positive results were confirmed by the second NAAT for both PT and AC2. Table 4 summarises the discordance between PT and AC2. In the rectum, 93.8% and 88.2% of PT and AC2-positive results, respectively, were positive by the other NAAT, but in the pharynx, only 73.3% of PT-positive results and 81.5% of AC2-positive pharyngeal results were confirmed by the other NAAT. One hundred per cent of the AC2-positive/PT-negative results were confirmed positive by ACT/AGC testing.
This study population of MSM at an urban STI clinic demonstrated a high prevalence of pharyngeal, rectal and urethral GC and CT infections. These findings were consistent with those of Kent et al, who also documented a high prevalence of these pathogens at the same anatomical sites using molecular methods in patients from a similar MSM clinic in San Francisco.24
The results of this study reinforce the importance of obtaining pharyngeal and rectal specimens to identify extragenital GC and CT infections. Relying on urine and urethral specimens would have missed 72.0% of the pharyngeal infections and 71.0% of the rectal infections that were detected through site-specific testing.
PT and AC2 both detected GC and CT with superior sensitivity compared to culture. Schachter et al conducted a similar study but reported lower sensitivities for PT and AC2 in detecting pharyngeal GC (87.9% for PT, 87.9% for AC2 with 66 true positives); rectal GC (88.5% for PT, 92.3% for AC2 with 78 true positives) and rectal CT (89.1% for PT, 93.5% for AC2 with 46 true positives).9 The number of pharyngeal CT cases was too small to draw definitive conclusions about the relative sensitivity and specificity of each assay. We think, however, that a prevalence of 2.0% is relevant and that if applied to a high-risk population such as MSM, screening for pharyngeal CT is worthwhile.
With regard to the performance of the various confirmatory strategies examined, repeat testing was very effective, with 97.4% of all positive tests re-testing positive. It should be noted, however, that repeat testing actually demonstrates the reproducibility of an assay, rather than providing an additional independent testing method. Various GC NAAT assays including PT have been reported to cross-react with several non-pathogenic Neisseria species.25 Multiple positive results may occur in an individual patient owing to repeated cross-reactions with other Neisseria species, making repeat testing a less than optimal confirmatory strategy.
Use of a second NAAT performed simultaneously on a separate specimen, although also a CDC-recommended strategy, showed a lower rate of confirmation compared to repeat testing. Previously published data have suggested that while AC2 can be used to confirm PT results, the reverse leads to large numbers of unconfirmed AC2 results because of the lower sensitivity of PT.26 27 In our study, however, both PT and AC2 confirmed each other’s results 85.9% of the time. We observed heightened discordance between PT and AC2 in pharyngeal specimens, suggesting that a low inoculum of GC and CT might be the reason for NAAT results that did not confirm, in addition to causing a negative result by culture, which requires a higher inoculum for detection.
One hundred per cent of positive AC2 results were also positive by the AGC/ACT, suggesting that the discordant AC2-positive/PT-negative results were in fact, true positives. Unfortunately, it was not possible to perform additional testing on the original PT specimens that were PT-positive/AC2-negative because of the lack of availability of a second DNA target. This would have provided a more complete picture of the overall false-positive rate among PT-positive/AC2-negative results.
This study’s sample size was larger than most other comparable studies in the literature,13–19 and the high prevalence of both pharyngeal and rectal infections in this study sample permitted adequate power to statistically compare the detection of GC and CT by culture against detection using the two NAATs studied. It should be noted, however, that this study was not powered to compare detection by one NAAT against the other.
This study has some limitations. The yield of culture was low for both pathogens. Owing to the off-site location of the laboratory, transport time was undoubtedly a contributing factor. Refrigeration of primary specimens may also have affected the yield of GC although some evidence suggests that GC may survive under these conditions.28 The sampling order (NAAT swabs first, followed by GC culture and CT culture) could have also theoretically disadvantaged the yield of culture, although there appeared to be more than adequate testing surface area available as the surface area of the swabs was quite small relative to the area being tested. Overall, it is important to note that this study was conducted in a purely clinical setting and was integrated into the routine clinical procedures of a busy STD clinic. The poor yield of culture in this study reflects, to some degree, the practical challenges associated with culturing highly fastidious pathogens in a setting that does not have the advantages of specimen collection under optimal research conditions.
This study has documented the excellent sensitivity of two commercially available NAATs in detecting Neisseria gonorrhoeae and Chlamydia trachomatis in pharyngeal and rectal specimens and additionally, the poor performance of culture techniques in a clinical setting. We also demonstrated that when confirmation of a positive NAAT was rigorously pursued, nearly 100% were confirmed by one or more CDC-recommended methods, suggesting excellent specificity for both NAATs. It should be emphasised, however, that this study was conducted in an MSM population with high GC and CT prevalence. Confirmatory testing would be particularly important in populations (MSM and non-MSM) with a lower prevalence of extragenital GC and CT.
MSM are clearly a high-risk population in which rapid detection and expeditious therapy are critical to ensuring eradication and interruption of transmission. There is an urgent need to advocate for federal approval of molecular testing platforms that provide improved detection of Neisseria gonorrhoeae and Chlamydia trachomatis in extragenital sites in this high risk population.
Nucleic acid amplification testing detects Neisseria gonorrhoeae (GC) and Chlamydia trachomatis (CT) with superior sensitivity compared to culture.
Currently available commercial kits appear to be useful and reliable for extragenital testing for GC and CT by nucleic acid amplification.
We would like to thank Dr Max Chernesky and Dan Jang from St Joseph’s Healthcare, Shah Nawaz from the Public Health Laboratory, Ontario Agency for Health Protection and Promotion and Rajneesh Sikri, Sudesh Singh and Leo Mitterni from the Hassle Free Men’s Clinic for their assistance with this study. Assay kits, swabs and transport medium were provided in kind by Becton, Dickinson and Company and Gen-Probe Inc.
Contributors: SER was the principal investigator for the study. KVO was the lead author. All authors contributed to the design of the study. KVO, JJ, FJ and LT collected data. Analysis was performed by KVO and MS. All authors contributed to the manuscript.
Competing interests: None.
Ethics approval: This study received research ethics approval by the Research Ethics Board at the Hospital for Sick Children in Toronto, Canada.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.