Friday, December 23, 2016 Written by Ellen Jo Baron, Ph.D., D(ABMM), Prof. Emerita, Stanford University Director of Medical Affairs, Cepheid

History of Laboratory Testing for Sexually Transmitted Bacteria: From Old School to Next Gen

Sexually transmitted infections (STIs) are an unfortunate fact of life. Although at the beginning of the AIDS epidemic in the U.S., at least, the numbers of Neisseria gonorrhoeae (NG) and Chlamydia trachomatis (CT) temporarily dipped due to increased use of condoms, they seem to be resurging. In northern Europe, CT is again prevalent.

Early Tests and Limitations
The laboratory diagnoses of NG and CT were not always linked as they are now. In the 1960's when I first began working in clinical microbiology, laboratories were routinely performing Gram stains and cultures for NG, but CT testing was rarely requested or performed. Microbiologists, at least those trained before 1990, all learned that Gram stains for NG should not be performed on endocervical discharge, since there were other organisms (Acinetobacter, Veillonella, Moraxella, and even over-decolorized staphylococci) whose morphology could be confused with that of NG, resulting in false positive results. For women, culturing both an anal and a cervical swab resulted in significantly better recovery.1 Males without obvious urethral discharge dreaded the visit to the clinic because obtaining the appropriate sample necessitated insertion of a frighteningly large-appearing swab into the urethra. To enhance the possibility of recovering the occasional gonococci that were slightly more susceptible to vancomycin, both a chocolate agar and a selective medium (Thayer-Martin, Martin-Lewis, modified Thayer-Martin, etc.) were recommended.2 Cultures were best inoculated to media at the patient's bedside, thus JEMBEC (the John E. Martin Biological Environmental Chamber) plates with internal CO2-generating tablets were considered the most sensitive method.3 Chlamydia trachomatis was even more difficult to recover in culture since they are obligate intracellular pathogens. Bedside inoculation was impossible and organisms often were lost during transport to the distant laboratory capable of the highly sophisticated culture protocol.

CT Joins the Spotlight
Early in the 1980's the surprising prevalence of CT in the sexually active population of the U.S. became apparent. Not only was CT as prevalent or even more prevalent than NG, but it also could lead to pelvic inflammatory disease, infant pneumonia, and conjunctivitis, and significant downstream sequelae, including upper genital tract disease that could cause infertility. With the understanding of these potentially serious outcomes, particularly in women for whom the disease was often asymptomatic, the wider utilization of diagnostic tests gained a new urgency.

The first type of test that generated broad application was the enzyme immunoassay (EIA) for the antigen of CT. Compared to culture in a multi-pronged study, the Abbott Chlamydiazyme test was 62% sensitive and 99% specific.4 A relatively small number of laboratories had been able to perform cultures in McCoy cells, treated with cycloheximide to prevent them from dividing so that they would be more hospitable to the intracellular Chlamydia.5

Because CT cultures were performed in cells, they were relegated to the virology laboratory, and virology laboratories were anything but routine in the 1960's and 1970's. Chlamydia testing took a giant leap forward with the advent of fluorescent DFA stains by giving non-virology laboratories the option to perform a relatively rapid test, rather than waiting for a large batch of samples to accumulate to make the EIA cost-efficient. The first edition of Bailey and Scott's Diagnostic Microbiology that I co-authored, in 1986, described the Syva Micro-Trak direct fluorescent monoclonal antibody (DFA) stain for Chlamydia inclusions and elementary bodies. DFA tests were able to detect both the large inclusions and the elementary bodies, which were often abundant in discharge of patients with CT. And for the first time, a viable organism was not necessary, since the stain could be used to visualize both live and dead structures.

First Generation Molecular Tests for CT/NG
The introduction of nucleic acid detection methods revolutionized diagnosis of both CT and NG. GenProbe Pace was not enormously difficult to implement in routine clinical laboratories. This molecular method involved a hybridization step and subsequent signal amplification developed to selectively detect ribosomal nucleic acid (rRNA), present in more copies per cell and thus capable of slightly higher sensitivity than detection of a single chromosomal target. However, compared to a reference standard, the original Pace 2 assay had sensitivity only slightly higher than 75%.4 Nonetheless, this technology played an important role in establishing molecular methods as a more rapid alternative to culture.

The first PCR assay was commercialized by Roche (Amplicor), but the technical complexity of this test prevented its general introduction into many laboratories. The Abbott ligase chain reaction (LCx) assay, however, was easier to perform (still a relative term) and gained wide acceptance. A Washington state example illustrates the impact of the LCx test. It was reported that none of 70 laboratories surveyed were performing any nucleic acid amplification tests (NAATs) for Chlamydia in 1995. Just three years later, 23% of those laboratories were using the LCx assay.6 By 2003, the LCx test was probably the most common CT testing platform in the world. Although there were some sample-handling contamination problems, these were usually resolved with additional training, assiduous cleaning of the testing area of the laboratory, and testing continued. Many laboratories began testing the environmental surfaces surrounding the instrument and sample preparation area on a weekly basis to detect contamination problems.

My laboratory at Stanford was heavily invested in this technology, performing LCx assays almost every day for our large reference testing business. Then, in 2003, the test was suddenly withdrawn from the market. Almost without warning, hundreds of laboratories were forced to look for another alternative testing method. Although there had been a voluntary withdrawal of some of the test kits in 2001, subsequent lot-to-lot variation, resulting in invalid results and potential false positives, forced the removal of all remaining assays from the worldwide market.7 The removal of this test happened so fast that laboratories did not even have time to validate another test before their supplies ran out. A promised replacement product from Abbott failed to materialize, and laboratories moved on. Although some reference and public health laboratories adopted the GenProbe Aptima, most hospital laboratories ultimately brought in the Becton Dickinson Probe-Tec, again due to its comparative ease of use.

In 1995, urine was first used for testing CT. This was a breakthrough particularly for men, where the test was very sensitive, whereas for women, the tests of the time were slightly less sensitive with urine than with endocervical or vaginal swabs.8 Women were thus still required to visit a physician and have an endocervical swab obtained via a speculum examination. This was a major deterrent to broad utilization of diagnostic testing, especially in the more vulnerable populations. 1996 saw the first use of vaginal swabs for testing for CT.9 Subsequent assays have validated the superiority of self-collected vaginal swabs for both CT and NG.10,11

True PCR-based tests for CT and NG include the Roche COBAS Amplicor and COBAS 4800, and the Abbott system. The Abbott Realtime CT/NG was developed with PCR using a non-Taqman based real-time detection technology. BD Probe-Tec uses a rival technology, strand displacement amplification (SDA), which appears to be equally effective at amplifying target nucleic acids. Early PCR assays such as the COBAS Amplicor amplified genetic sequences and used a second assay for detection of the product, usually an immunoassay. Real-time PCR that enables direct detection of amplified products and a more rapid delivery of results has been the fastest growing choice for a platform since 2000. However, despite the manufacturer's claims of "new and improved," all these testing systems are based on technologies that are now over 20 years old, and many of them are really beginning to show their age. Thus, changes in the application of PCR technology are needed for real improvement.

False Positive Results – Still a problem after all these years
False positive results have been a persistent problem for NG testing, even years after the commercial introduction of NAATs. A 1999 study comparing Amplicor to an in-house nested PCR system showed that there were ~2% false negatives, but a more worrisome >15% of Amplicor-positive samples failed to confirm in the more specific assay.12 Similar results for some assays were reported by other authors.13 Such reports can become newsworthy, with potentially disastrous consequences for the laboratory once the news becomes public knowledge. Newspaper headlines in the New York Times (2004) proclaimed that five women from Hawaii received a report that they had gonorrhea, causing untold havoc in their personal lives, only to learn later that the laboratory test was falsely positive.14 Chlamydia testing is also not immune to problems. Just last year, an Indiana news channel report stated "Botched lab tests prompt Chlamydia scare" in Indianapolis after low-level contamination resulted in false positives for several patients.

Other sources of false positivity may arise because of the mischievous combination of bacterial promiscuity and the limitations of NAAT assay design. Neisserial organisms exchange genetic material via conjugation, and homologous recombination can result in the integration of the target gene of N. gonorrhoeae in the genome of a non-pathogenic neisserial species. This possibility was substantiated very recently in a study by Tabrizi et al. that challenged six NAATs with a range of Neisseria species likely to yield false positive results by NAAT. As expected, all tests did report false positive results for some species, with Probe-Tec and COBAS Amplicor showing the highest incorrect rates.15 Schachter and Chernesky proposed that the false positive results in Tabrizi's platform comparison study were due to low-level contamination during the testing process.16 In a letter to the editor, Tabrizi et al. responded that since they used doubly purified single colony picks, this was not the best explanation for their results.17

The reason given for sporadic false positives for NG in the current NAATs is cross-reactivity of the primers for genetically similar targets in Neisseria species other than gonorrhoeae. One such culprit is N. mucosa, and several studies of false positive results due to N. cinerea are well-known, some with devastating consequences to the falsely-labeled patients.12 This source of false positive results is inherent in the designs of the tests themselves, and no improvements in laboratory processes will prevent it. Many of the strains used in the Tabrizi study would be considered commensal flora in the throat or GI tract, suggesting that use of the NAATs on these specimen types would be especially problematic. Reporting a false positive result for any sexually-transmitted disease can have ruinous consequences, both for the patient and for the patient's contacts.

The batch-based testing approach of most modern NAAT formats creates another risk of internal or specimen-handling based contamination. If shipped together in the same container, samples can cross-contaminate each other. In addition, since samples are initially processed in an open system, there is a risk of carryover from strongly positive specimens into negative adjacent wells. This type of contamination can be quite sporadic and difficult to detect, but it is similar to the types of problems seen in laboratories performing mycobacterial liquid cultures, where specimen cross-contamination is a frequent hazard.

Besides random sample carryover during the specimen processing steps, today's large-batch instrument formats can lead to amplicon carryover within the instrument itself. Contamination with amplified target material is a well-known problem, so most laboratories dedicate a physically separate space for amplification product detection and have stockpiles of bleach on hand to help confront the seemingly inevitable. Either way, the manner in which we process specimens – in large batches with initial specimen transfer steps occurring above racks of open wells or tubes - leaves us open to the possibility of carryover-related false positives among current NAAT assays.

The most recent proficiency testing results from the College of American Pathologists (2012, HC6-A) shows a few laboratories, as many as 1% for one sample, reporting false positive results. This is worrisome for two reasons: (1) laboratories work very hard to do their best testing with proficiency samples, so real-world results are likely to be worse. And (2), as the true prevalence of NG in most U.S. populations being tested is also around 1%, it is thus conceivable that half of all positive test results are incorrectly false positives. To detect sporadic carryover problems, many more negative controls would need to be incorporated into the daily routine than are currently employed, thus decreasing the cost-effectiveness of batched testing platforms.

False Negative Results
Inhibition of amplification reaction can lead to false negative results, yet until recently, none of the commercial NAATs had the ability to detect inhibition coming from a clinical sample. Inhibition was pointed out as a significant problem for all of the commercial assays, but especially for the BD ProbeTec and Roche Amplicor assays for vaginal and urine samples where false negatives due to inhibition were above 20%.13 Unfortunately, without an internal process control, false negative results due to inhibition are difficult to discover. The good news is that manufacturers are moving toward correcting this problem. The newer BD assay has an internal inhibition control (ProbeTec Qx Amplified DNA assay product inserts). The GenProbe Aptima assay uses target capture prior to amplification, so although it is not completely immune to inhibition problems, it has overcome most of them. This has helped it become what many feel is the standard of the industry today.13


1. Bhattacharyya, M. N. and A. E. Jephcott (1974). "Diagnosis of gonorrhoea in women. Role of the rectal sample." Br J Vener Dis 50(2): 109-12.
2. Forbes, B. a. G., P. (1995). Processing speci
mens for bacteria. Manual of Clinical Microbiology. A. e. a. Balows. Washington, DC, ASM Press.
3. Symington, D. A. (1975). "Improved transport system for Neisseria gonorrhoeae in clinical specimens." J Clin Microbiol 2(6): 498-503.
4. Newhall, W. J., R. E. Johnson, et al. (1999). "Head-to-head evaluation of five chlamydia tests relative to a quality-assured culture standard." J Clin Microbiol 37(3): 681-5.
5. Stamm, W. E., M. Tam, et al. (1983). "Detection of Chlamydia trachomatis inclusions in Mccoy cell cultures with fluorescein-conjugated monoclonal antibodies." J Clin Microbiol 17(4): 666-8.
6. Battle, T. J., M. R. Golden, et al. (2001). "Evaluation of laboratory testing methods for Chlamydia trachomatis infection in the era of nucleic acid amplification." J Clin Microbiol 39(8): 2924-7.
7. Gronowski, A. M., S. Copper, et al. (2000). "Reproducibility problems with the Abbott laboratories LCx assay for Chlamydia trachomatis and Neisseria gonorrhoeae." J Clin Microbiol 38(6): 2416-8.
8. Van Der Pol, B., T. C. Quinn, et al. (2000). "Multicenter evaluation of the AMPLICOR and automated COBAS AMPLICOR CT/NG tests for detection of Chlamydia trachomatis." J Clin Microbiol 38(3): 1105-12.
9. Wiesenfeld, H. C., R. P. Heine, et al. (1996). "The vaginal introitus: a novel site for Chlamydia trachomatis testing in women." Am J Obstet Gynecol 174(5): 1542-6.
10. Hook, E. W., 3rd, K. Smith, et al. (1997). "Diagnosis of genitourinary Chlamydia trachomatis infections by using the ligase chain reaction on patient-obtained vaginal swabs." J Clin Microbiol 35(8): 2133-5.
11. Oh, M. K., C. M. Richey, et al. (1997). "High prevalence of Chlamydia trachomatis infections in adolescent females not having pelvic examinations: utility of PCR-based urine screening in urban adolescent clinic setting." J Adolesc Health 21(2): 80-6.
12. Farrell, D. J. (1999). "Evaluation of AMPLICOR Neisseria gonorrhoeae PCR using cppB nested PCR and 16S rRNA PCR." J Clin Microbiol 37(2): 386-90.
13. Chernesky, M., D. Jang, et al. (2006). "High analytical sensitivity and low rates of inhibition may contribute to detection of Chlamydia trachomatis in significantly more women by the APTIMA Combo 2 assay." J Clin Microbiol 44(2): 400-5.
14. Katz, A. R., P. V. Effler, et al. (2004). "False-positive gonorrhea test results with a nucleic acid amplification test: the impact of low prevalence on positive predictive value." Clin Infect Dis 38(6): 814-9.
15. Tabrizi, S. N., M. Unemo, et al. "Evaluation of six commercial nucleic acid amplification tests for detection of Neisseria gonorrhoeae and other Neisseria species." J Clin Microbiol 49(10): 3610-5.
16. Schachter, J. and M. A. Chernesky "Routine confirmation of positive nucleic acid amplification test results for Neisseria gonorrhoeae is not necessary." J Clin Microbiol 50(1): 208; author reply 209-10.
17. Tabrizi, S.N.Hjelmevoll, S.O.Garland, S.M.Unemo, M. Journal of Clinical Microbiology. 2012;50(1).