Performance of the Gen-Probe Transmission-Mediated Amplification Research Assay Compared to That of a Multitarget Real-Time PCR for Mycoplas
http://www.100md.com
微生物临床杂志 2006年第4期
Division of Infectious Diseases, Johns Hopkins University School of Medicine, Baltimore, Maryland
Johns Hopkins University Department of Emergency Medicine, Baltimore, Maryland
National Institute of Allergy and Infectious Diseases, National Institutes of Health Bethesda, Maryland
ABSTRACT
Mycoplasma genitalium (MG) can cause nongonococcal urethritis and is potentially associated with urethritis, endometritis, and cervicitis. Several assays have been developed to detect MG. Molecular amplification assays for organism detection can be problematic due to the potential for false-positive and false-negative results. Confirmatory testing is often required in these situations, requiring additional time and resources. Use of multigene targets could integrate both detection and verification at lower cost. Utilizing two targets, the MgPa adhesion gene and the 16S rRNA gene, a multitarget real-time (MTRT) PCR for the detection of MG was developed. Samples from patients attending sexually transmitted disease clinics were collected in duplicate. Urine samples from males (n = 286) and self-collected vaginal swabs from females (n = 321) were analyzed by MTRT PCR for MG and the Gen-Probe transmission-mediated amplification (TMA) assay, which targets MG rRNA for detection (TMA-MG research use only). Utilizing the criteria of any two targets being positively amplified, the MTRT PCR had a sensitivity and specificity of 91.8% (101 positive samples/110 samples tested) and 99.5% (495/497), respectively, with a positive predictive value (PPV) of 98.1% (101/103) and a negative predictive value (NPV) of 98.2% (495/504). The Gen-Probe TMA-MG assay had a sensitivity, specificity, PPV, and NPV of 98.1% (108/110), 98.1% (488/497), 92.3% (108/117), and 99.5% (488/490), respectively. Comparison between the MTRT PCR and TMA-MG assay by kappa statistic analysis indicated that an overall kappa value was 0.941 (95% confidence interval, 0.907 and 0.976). Both assays demonstrated accuracy in the detection of MG from urine samples from male patients and self-collected vaginal swabs from female patients.
INTRODUCTION
Mycoplasma genitalium (MG) causes nongonococcal urethritis and has been associated with endometritis and cervicitis (14). MG has been implicated as a cause of several syndromes where etiology cannot be attributed to major sexually transmitted infections (STIs) (1, 13, 14, 20, 22, 28). Various PCR assays have improved clinical diagnostic capabilities with regard to MG (2-9, 14-18, 23, 25, 29-31). PCR can be advantageous in the detection of MG because it is a difficult organism to culture. PCR detection requires only DNA, alleviating the need for viable organisms. Real-time PCR methodology has further improved MG detection (3, 7, 8, 12, 14, 29). Real-time PCR couples simultaneous amplification and detection, resulting in shorter assay time and increased sensitivity and obviating the need for post-PCR processing that is required for traditional PCR assays.
There is concern when comparing new and older amplified tests for organism detection, since it may be difficult to determine a true positive (i.e., a "gold standard"). A secondary test is often utilized to confirm unique positive results (samples positive by one test and negative by another). Use of discrepant testing has been criticized, since it is often performed on a subset of specimens and may introduce statistical bias in terms of sensitivity and specificity (11, 19, 21).
Use of multiple gene targets can offer an alternative to discordant test analysis. Ideally, such an amplified test would consist of two primers and probe sets for simultaneous detection of different gene targets and could provide self-confirming results that could reduce, if not eliminate, the need for discrepant and confirmatory testing.
Multitarget real-time (MTRT) PCR for MG utilizing two alternate gene targets, the MgPa gene and the 16S rRNA gene (14, 29), was designed and used to study the performance of the GenProbe transmission-mediated amplification (TMA) research assay, which targets MG rRNA for detection in genital specimens from patients attending a sexually transmitted disease clinic. MgPa and the 16S rRNA gene were chosen as targets for MTRT PCR because the targets performed well in previous studies in terms of sensitivity and specificity and reproducibility (14, 29).
MATERIALS AND METHODS
Preparation of standards and controls. DNA was extracted from each MG standard (5 x 106 CFU/ml; gift strain from Lynn Duffy, University of Alabama at Birmingham, Birmingham, AL) by utilizing the Roche MagNA Pure LC robotic instrument (Roche Molecular Diagnostics, Indianapolis, IN). Control DNA dilutions ranged in concentration from 5 x 105 CFU/ml to 0.005 CFU/ml. These controls were utilized to determine the limit of detection for each target (MgPA and 16S rRNA) and were also utilized to gauge reproducibility. The MagNA Pure LC DNA Isolation Kit I (Roche) was used, and DNA extraction was carried out according to the instructions supplied for the MagNA Pure LC program, the DNA I Blood Cells High Performance Serum protocol. Negative controls were processed in an identical manner and consisted either of molecular grade H2O or carrier DNA, such as a noncomplementary oligonucleotide.
Patient specimen preparation. There were 615 people enrolled, with urine samples from male patients (n = 290) and self-collected vaginal swabs from female patients (n = 325), collected as part of a cervicitis and urethritis study at sexually transmitted disease clinics. Of the 615 people enrolled, 611 patients were tested, but only 607 patients were utilized for data analysis. The Johns Hopkins University Institutional Review Board and the Baltimore City Health Department approved this study; the study was funded by Gen-Probe Incorporated (San Diego, CA). For MTRT PCR, vaginal swabs from female patients were collected by patients before urine or cervical samples were collected and were shipped in a dry state. Dry swabs were rehydrated in 1 ml of Tris-EDTA buffer, of which 200 μl was removed for DNA extraction. Urine samples from male patients were collected before urethral swabs were collected. Samples (from male and female patients) were then extracted, utilizing the Roche MagNA Pure LC robotic instrument (Roche Molecular Diagnostics, Indianapolis, IN). DNA extraction was carried out according to the instructions supplied for the MagNA Pure LC program, the DNA I Blood Cells High Performance Serum protocol. For the TMA-MG research assay, the duplicate vaginal swabs collected in random order with the vaginal swabs for MTRT PCR and urine samples collected in specimen transport media (provided by Gen-Probe) were processed as for other TMA assays.
MTRT PCR primers and probes. Previously published sets of primers and probes were utilized in the further development of this assay (14, 29). For the MgPa gene target, the primers and probe were MgPa-355F (5'-GAGAAATACCTTGATGGTCAGCAA-3'), MgPa-432R (5'-GTTAATATCATATAAAGCTCTACCGTTGTTATC-3'), and MgPa-380 (5'-TET-ACTTTGCAATCAGAAGGT-MGBNFQ-3'), utilizing tetrabromosulfonefluorescein (TET) as the reporter and a minor groove-binding (MGB) nonfluorescent quencher (NFQ). The primers and probe for the 16S rRNA gene target were My-INS (5'-GTAATACATAGGTCGCAAGCGTTATC-3'), MGSO-2 (5'-CACCACCTGTCACTCGGTTAACCTC-3'), and the MgenP1 probe (5'-6FAM-CTGTCGGAGCGATCCCTTCGGT-MGBNFQ-3') (where 6FAM is a designation for 6-carboxyfluorescein), utilizing a fluorescein reporter and an MGB quencher. All primers and probes were synthesized by Applied Biosystems, Foster City, CA. The sequences of the previously published primers and probes remained unchanged; however, unique reporter molecules were assigned to each probe to ensure that there was a distinct and differentiable signal from each gene target.
PCR conditions. MTRT PCRs were performed with 80 μl of master mixture and 20 μl of template DNA. The master mixture contained 10 μl of 10x QIAGEN HotStarTaq polymerase buffer, 20 μl of 25 mM MgCl2, 8 μl of 25 mM deoxynucleoside triphosphates (dNTPs), 0.5 μl of each 50 μM primer and probe, 36 μl of PCR-grade H2O, 1 μl of QIAGEN Q solution, 2 μl of QIAGEN HotStarTaq polymerase, and 20 μl of template DNA. The 10x buffer, MgCl2, Q solution, and HotStarTaq polymerase were obtained from the QIAGEN HotStarTaq polymerase kit (QIAGEN, Valencia, CA). The dNTPs were obtained from the 100 mM dNTP set PCR Grade (Invitrogen). PCR-grade water was obtained from Quality Biologics Incorporated (Quality Biologics Incorporated, Gaithersburg, MD). PCR was performed using a 96-well plate format on the ABI 7900 HT Sequence Detection system (Applied Biosystems) with the ROX reference removed, under the following conditions: 50°C for 2 min and 95°C for 10 min, followed by 50 cycles, each consisting of 95°C for 15 s, 60°C for 1 min, and 72°C for 30 s. A cooling hold of 4°C for 2 min was added at the end of the cycling protocol for ease of plate handling.
GenProbe TMA testing. The duplicate urine samples from male patients and vaginal swabs from female patients were analyzed by target capture, amplification by TMA, and detection by the hybridization protection assay in a manner similar to procedures for other Gen-Probe Incorporated APTIMA assay family products (Gen-Probe APTIMA Combo2 package insert, IN0037-04 Rev A; Gen-Probe, Inc., San Diego, CA). Primers, probes, and target capture oligomers were designed by Gen-Probe to specifically target MG and were designed to be utilized with reagents that have the same formulation as APTIMA Combo2. Gen-Probe provided the reagents for this project as part of their ongoing research program. The specific MG reagents are not commercially available. The cutoff for positive reactions was set at 40,000 relative light units.
Specificity testing. Mycoplasma pneumoniae (ATTC 15492), Mycoplasma hominis (ATTC 14027), Mycoplasma pirum, Mycoplasma fermentans (PG18), Mycoplasma penetrans, Ureaplasma urealyticum (serotype 10), and Ureaplasma parvum (serotype 3) were evaluated for potential cross-reactivity with the primer and probe sets utilized in the MTRT PCR. Nucleic acid extraction was performed for PCR in the same manner as for the clinical samples. All organisms were evaluated at a high DNA concentration (>20 ng organism DNA/PCR). One reaction per organism was performed, and data evaluation was performed in the same manner as for the clinical samples. The primers and probes for the MgPA assay were previously evaluated against the following organisms: M. pneumoniae (FHT, Mac, M129-B8, M129-B170, and two clinical isolates), Mycoplasma hominis (PG21T, H34, H27, and three clinical isolates), Mycoplasma salivarium (PG20T), Mycoplasma buccale (CH 20247T), Mycoplasma orale (Patt and one clinical isolate), Mycoplasma fermentans (GT and S38), Mycoplasma faucium (DC 333T), Mycoplasma primatum (Navel), Mycoplasma pirum (Zeus), Mycoplasma lipophilum (Maby BT), Mycoplasma penetrans (GTU), Mycoplasma hyorhinis (GDL), Mycoplasma arginini (G230T), Mycoplasma gallisepticum (15302), Mycoplasma iowae (695), Mycoplasma imitans (4229), Mycoplasma testudinis (Hill), Mycoplasma alvi (Isley), and Acholeplasma laidlawii (AT), and Ureaplasma urealyticum (serotypes I [F. Black 7] and VIII [F. Black 960T]) (14). Primers and probes for the 16S rRNA target were previously evaluated against the following organisms: Mycoplasma buccale, Mycoplasma faucium, Mycoplasma fermentans, M. genitalium, Mycoplasma hominis, Mycoplasma lipophilum, Mycoplasma orale, Mycoplasma penetrans, Mycoplasma pirum, Mycoplasma pneumoniae, Mycoplasma primatum, Mycoplasma salivarium, Mycoplasma spermatophilum, Ureaplasma parvum, and Ureaplasma urealyticum (29).
Sample analysis. Positive and negative processing controls were extracted and included in every run of samples utilized for assay validation. Each sample was analyzed by both the MTRT PCR and Gen-Probe TMA-MG assays. True positives were defined a priori as any two positive amplifications from any three results: the Gen-Probe TMA-MG, 16S rRNA, or MgPA gene targets. Kappa statistical analysis was also performed to measure the degree of concordance between the two assays (27). Samples were considered equivocal if the test results did not repeat consistently upon multiple retesting.
RESULTS
The analytical sensitivity of the MTRT PCR was excellent, with positive controls containing 50 CFU/ml and 5 CFU/ml repeating 100% (6/6 samples) of the time for both the MgPa and 16S rRNA targets. Positive controls of 0.5 CFU/ml repeated 100% (6/6) of the time for the MgPa target and 83.3% (5/6) of the time for the 16S rRNA target. Positive controls of 0.05 CFU/ml and 0.005 CFU/ml were positive 100% (6/6) of the time for the MgPA target but were not positive for the 16S rRNA target. It is unclear why the MgPa and the 16S rRNA targets had different analytical sensitivities. There were attempts at additional optimization, including MgCl2 titrations and varying the amount of sample input; however, any changes that improved the analytical sensitivity for the 16S rRNA target also decreased the sensitivity for the MgPa target.
A brief examination of the statistical differences between targets and runs utilizing the generalized estimating equation method revealed that there were no statistical differences between runs for the MgPA target (P = 0.140) and slight differences between the runs for the 16S rRNA target (P = 0.035) (32). Overall, the assay was determined to be reproducible for qualitative screening for MG.
The MTRT PCR also did not cross-react with any of the organisms analyzed for specificity. A National Cancer Biological Institute BLAST search did not reveal cross-reacting primer sequences, and the sensitivity and specificity of the primer and probe sets utilized were analyzed and validated for specificity against a wide range of mycoplasmas and ureaplasmas in previous studies (14, 29).
Of the 615 participants, samples from 607 participants were utilized for analysis. Four samples were excluded from testing due to an insufficient sample, while the other four samples (three samples from male participants and one sample from a female participant) were excluded from analysis after equivocal results were obtained from multiple rounds of retesting by both assays (MTRT PCR and TMA-MG), in which the samples failed to resolve as true positives or true negatives. True positives were defined as any two positive amplifications from any three results; the Gen-Probe TMA-MG, 16S rRNA, or MgPA gene targets were utilized for determination of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). The MTRT PCR had a sensitivity and specificity of 91.8% (101/110 samples) and 99.5% (495/497), respectively, with a PPV of 98.1% (101/103) and an NPV of 98.2% (495/504). For men, sensitivity, specificity, PPV, and NPV were 88.8% (40/45 samples), 100% (241/241), 100% (40/40), and 97.9% (241/246), respectively. For women, the sensitivity, specificity, PPV, and NPV were 93.8% (61/65 samples), 99.2% (254/256), 96.8% (61/63), and 98.4% (254/258), respectively (Table 1).
The Gen-Probe TMA-MG assay had an overall sensitivity, specificity, PPV, and NPV of 98.1% (108/110 samples), 98.1% (488/497), 92.3% (108/117), and 99.5% (488/490), respectively. For men, sensitivity, specificity, PPV, and NPV were 100% (45/45 samples), 97.9% (236/241), 90.0% (45/50), and 100.0% (236/236), respectively. For women, the sensitivity, specificity, PPV, and NPV were 96.9% (63/65 samples), 98.4% (252/256), 94.0% (63/67), and 99.2% (252/254), respectively (Table 1).
There were 59 samples that were uniquely positive for the MgPA target and for the TMA-MG assay that were negative for the 16S rRNA target. These 59 samples failed to resolve as positive for the 16S rRNA target, despite retesting.
Comparison between the MTRT PCR and TMA-MG assay by kappa statistic analysis indicated an overall kappa value of 0.941 (95% confidence interval [CI], 0.907 and 0.976) while it was 0.937 (95% CI, 0.882 and 0.992) for men, and 0.944 (95% CI, 0.900 and 0.988) for women. All kappa analyses indicated "almost-perfect" agreement between the assays (27).
DISCUSSION
There were 110 positive samples identified, resulting in an 18.1% MG prevalence in the study population. Overall, both assays had high sensitivities and specificities at 91.8% (101/110 samples) and 99.5% (495/497) for the MTRT PCR assay and 98.1% (108/110) and 98.1% (488/497) for the Gen-Probe TMA MG assay.
Kappa analysis indicated an "almost-perfect" agreement between both assays with an overall value of 0.941 (95% CI, 0.907 and 0.976), indicating that either test is suitable for MG screening in research settings. These tests could be useful in future studies examining the associations of MG with urethritis, cervicitis, and other STIs.
There are several recently developed real-time methods available for MG detection (3, 8-10, 14, 26, 29, 31). Direct comparison of the sensitivity and specificity of the MTRT PCR or TMA-MG assay with the other published assays for MG detection is difficult because most assays do not include comparison with a gold standard. This is because MG culture is very difficult to perform, and there is no nucleic acid amplification test that is accepted as the gold standard for MG detection. However, in terms of limit of detection, both the MTRT PCR and TMA-MG assays are comparable with other published assays, whose limits of detection range from 1 genomic copy to 50 genomic copies (3, 8-10, 14, 26, 29, 31).
There is some limitation in the overall comparisons that can be made between the assays because different sample types were involved for each sex. Samples from male participants consisted of urine only, whereas samples from female participants were self-collected vaginal dry swabs, which potentially contain higher organism concentrations. Despite the differences in sample type, kappa analysis indicated "almost-perfect" agreement between the assays when examined overall and by sex.
Urine samples from male participants (n = 290) and self-collected vaginal swabs from female participants (n = 325) appear to be appropriate sample types for M. genitalium detection, although male urethral swabs and female cervical swabs were not evaluated in this study. Future studies should address these comparisons. Self-collected vaginal swabs shipped in a dry state could be particularly important for outreach programs targeting people without access to health clinics.
There were four samples (three samples from male participants and one sample from a female participant) that were excluded from analysis for failing to consistently repeat as positive or negative, despite multiple rounds of retesting. These results are most likely the result of low MG loads in the sample, resulting in sampling error due to Poisson distribution of the target in replicate samples.
It is of interest that 59 samples were positive only by the TMA-MG assay and the MgPA target and negative for the 16S rRNA target by the MTRT PCR. During assay development and validation, the 16S rRNA target had an excellent limit of detection (between 5 and 0.5 CFU/ml) and was determined to be a target that yielded reproducible results, based on generalized estimating equation analysis. Additionally, it is unclear why the MgPa and the 16S rRNA targets had different analytical sensitivities at <1 CFU/ml for the MgPa target and between 5 and <1 CFU/ml for the 16S rRNA target. Reoptimization of the assay to increase the sensitivity for the 16S rRNA target only served to decrease the sensitivity for the MgPa target. The primer and probe sets for each target were taken from previously published assays for MG (14, 29), and each set had different optimal annealing temperatures at 60°C and 66°C, respectively. The MTRT PCR utilized an annealing temperature of 55°C to emphasize sensitivity, and it may be the discrepancy between the annealing temperatures that accounts for different analytical sensitivities of each target. Or, the region of the 16S rRNA target of MG is a less-sensitive target choice for MG detection from clinical samples than other targets, such as MgPA (26). Future studies comparing the efficiency of 16S rRNA-based detection versus alternative targets for MG detection should resolve this issue. Although the MTRT PCR was not 100% self confirming, the ability of an assay to detect multiple targets simultaneously should be incorporated into future molecular detection assays, not only for MG but for other pathogens as well, because amplification of multiple, alternative targets from the same sample would increase confidence in assay results and decrease the need for costly and time-consuming confirmation testing.
There is growing evidence to suggest that M. genitalium does have a direct involvement as a STI in women and is also involved in urethritis and cervicitis (24). Additionally, there are suggested associations between M. genitalium, pelvic inflammatory disease, and infertility (24). If the potential links between MG and urethritis, cervicitis, pelvic inflammatory disease, and infertility are to be concretely established, improved diagnostics with high sensitivity and specificity for M. genitalium will be required, as culture methods are difficult.
The TMA-MG research assay performed very well in comparison with the MTRT PCR research assay and has the potential to provide clinicians with a commercially available and highly accurate diagnostic tool for future M. genitalium research. Because it does not require viable organisms and has the potential to be self confirming, the integrated MTRT research PCR represents a cost-effective method for M. genitalium detection and verification, especially in research studies.
REFERENCES
Bjornelius, E., P. Lidbrink, and J. S. Jensen. 2000. Mycoplasma genitalium in nongonococcal urethritis—a study in Swedish male STD patients. Int. J. STD AIDS 11:292-296.
Blanchard, A., W. Hamrick, L. Duffy, K. Baldus, and G. H. Cassell. 1993. Use of the polymerase chain reaction for detection of Mycoplasma fermentans and Mycoplasma genitalium in the urogenital tract and amniotic fluid. Clin. Infect. Dis. 17(Suppl. 1):S272-S279.
Blaylock, M. W., O. Musatovova, J. G. Baseman, and J. B. Baseman. 2004. Determination of infectious load of Mycoplasma genitalium in clinical samples of human vaginal cells. J. Clin. Microbiol. 42:746-752.
Busolo, F., D. Camposampiero, G. Bordignon, and G. Bertollo. 1997. Detection of Mycoplasma genitalium and Chlamydia trachomatis DNAs in male patients with urethritis using the polymerase chain reaction. New Microbiol. 20:325-332.
Cadieux, N., P. Lebel, and R. Brousseau. 1993. Use of a triplex polymerase chain reaction for the detection and differentiation of Mycoplasma pneumoniae and Mycoplasma genitalium in the presence of human DNA. J. Gen. Microbiol. 139:2431-2437.
de Barbeyrac, B., C. Bernet-Poggi, F. Febrer, H. Renaudin, M. Dupon, and C. Bebear. 1993. Detection of Mycoplasma pneumoniae and Mycoplasma genitalium in clinical samples by polymerase chain reaction. Clin. Infect. Dis. 17(Suppl. 1):S83-S89.
Deguchi, T., T. Yoshida, S. Yokoi, M. Ito, M. Tamaki, H. Ishiko, and S. Maeda. 2002. Longitudinal quantitative detection by real-time PCR of Mycoplasma genitalium in first-pass urine of men with recurrent nongonococcal urethritis. J. Clin. Microbiol. 40:3854-3856.
Dupin, N., G. Bijaoui, M. Schwarzinger, P. Ernault, P. Gerhardt, R. Jdid, S. Hilab, C. Pantoja, M. Buffet, J. P. Escande, and J. M. Costa. 2003. Detection and quantification of Mycoplasma genitalium in male patients with urethritis. Clin. Infect. Dis. 2003. 37:602-605.
Dutro, S. M., J. K. Hebb, C. A. Garin, J. P. Hughes, G. E. Kenny, and P. A. Totten. 2003. Development and performance of a microwell-plate-based polymerase chain reaction assay for Mycoplasma genitalium. Sex. Transm. Dis. 30:756-763.
Eastick, K., J. P. Leeming, E. O. Caul, P. J. Horner, and M. R. Millar. 2003. A novel polymerase chain reaction assay to detect Mycoplasma genitalium. Mol. Pathol. 56:25-28.
Green, T. A., C. M. Black, and R. E. Johnson. 1998. Evaluation of bias in diagnostic-test sensitivity and specificity estimates computed by discrepant analysis. J. Clin. Microbiol. 36:375-381.
Heid, C. A., J. Stevens, K. J. Livak, and P. M. Williams. 1996. Real time quantitative PCR. Genome Res. 6:986-994.
Horner, P. J., C. B. Gilroy, B. J. Thomas, R. O. Naidoo, and D. Taylor-Robinson. 1993. Association of Mycoplasma genitalium with acute non-gonococcal urethritis. Lancet 342:582-585.
Jensen, J. S., E. Bjornelius, B. Dohn, and P. Lidbrink. 2004. Use of TaqMan 5' nuclease real-time PCR for quantitative detection of Mycoplasma genitalium DNA in males with and without urethritis who were attendees at a sexually transmitted disease clinic. J. Clin. Microbiol. 42:683-692.
Jensen, J. S., M. B. Borre, and B. Dohn. 2003. Detection of Mycoplasma genitalium by PCR amplification of the 16S rRNA gene. J. Clin. Microbiol. 41:261-266.
Jensen, J. S., S. A. Uldum, J. Sondergard-Andersen, J. Vuust, and K. Lind. 1991. Polymerase chain reaction for detection of Mycoplasma genitalium in clinical samples. J. Clin. Microbiol. 29:46-50.
Luki, N., P. Lebel, M. Boucher, B. Doray, J. Turgeon, and R. Brousseau. 1998. Comparison of polymerase chain reaction assay with culture for detection of genital mycoplasmas in perinatal infections. Eur. J. Clin. Microbiol. Infect. Dis. 17:255-263.
Mahony, J. B., D. Jang, S. Chong, K. Luinstra, J. Sellors, M. Tyndall, and M. Chernesky. 1997. Detection of Chlamydia trachomatis, Neisseria gonorrhoeae, Ureaplasma urealyticum, and Mycoplasma genitalium in first-void urine specimens by multiplex polymerase chain reaction. Mol. Diagn. 2:161-168.
McAdam, A. J. 2000. Discrepant analysis: how can we test a test J. Clin. Microbiol. 38:2027-2029.
Mena, L., X. Wang, T. F. Mroczkowski, and D. H. Martin. 2002. Mycoplasma genitalium infections in asymptomatic men and men with urethritis attending the New Orleans Sexually Transmitted Disease Clinic. Clin. Infect. Dis. 35:1167-1173.
Miller, W. C. 1998. Bias in discrepant analysis: when two wrongs don't make a right. J. Clin. Epidemiol. 51:219-231.
Morency, P., M. J. Dubois, and G. Gresenguet. 2001. Aetiology of urethral discharge in Bangui, Central African Republic. Sex. Transm. Infect. 77:125-129.
Palmer, H. M., C. B. Gilroy, P. M. Furr, and D. Taylor-Robinson. 1991. Development and evaluation of the polymerase chain reaction to detect Mycoplasma genitalium. FEMS Microbiol. Lett. 61:199-203.
Ross, J. D. C. 2005. Is Mycoplasma genitalium a cause of pelvic inflammatory disease. Infect. Dis. Clin. N. Am. 19:407-413.
Sharma, S., R. Brousseau, and S. Kasatiya. 1998. Detection and confirmation of Mycoplasma pneumoniae in urogenital specimens by PCR. J. Clin. Microbiol. 36:277-280.
Svenstrup, H. F., J. S. Jensen, E. Bjrnelius, P. Lidbrink, S. Birkelund, and C. Gunna. 2005. Development of a quantitative real-time PCR assay for detection of Mycoplasma genitalium. J. Clin. Microbiol. 43:3121-3128.
Szklo, M. (ed.). 2000. Epidemiology beyond the basics, p. 343-404. Aspen Publishers, Inc., Gaithersburg, MD.
Totten, P. A., M. A. Schwartz, and K. E. Sjostrom. 2001. Association of Mycoplasma genitalium with nongonococcal urethritis in heterosexual men. J. Infect. Dis. 183:269-276.
Yoshida, T., T. Deguchi, M. Ito, S. I. Maeda, M. Tamaki, and H. Ishiko. 2002. Quantitative detection of Mycoplasma genitalium from first-pass urine of men with urethritis and asymptomatic men by real-time PCR. J. Clin. Microbiol. 40:1451-1455.
Yoshida, T., S. Maeda, T. Deguchi, and H. Ishiko. 2002. Phylogeny-based rapid identification of mycoplasmas and ureaplasmas from urethritis patients. J. Clin. Microbiol. 40:105-110.
Yoshida, T., S. Maeda, T. Deguchi, T. Miyazawa, and H. Ishiko. 2003. Rapid detection of Mycoplasma genitalium, Mycoplasma hominis, Ureaplasma parvum, and Ureaplasma urealyticum organisms in genitourinary samples by PCR-microtiter plate hybridization assay. J. Clin. Microbiol. 41:1850-1855.
Zeger, S. L., and K. Y. Liang. 1986. Longitudnal data analysis for discrete and continuous outcomes. Biometrics 42:121-130.(Justin Hardick, Julie Gil)
Johns Hopkins University Department of Emergency Medicine, Baltimore, Maryland
National Institute of Allergy and Infectious Diseases, National Institutes of Health Bethesda, Maryland
ABSTRACT
Mycoplasma genitalium (MG) can cause nongonococcal urethritis and is potentially associated with urethritis, endometritis, and cervicitis. Several assays have been developed to detect MG. Molecular amplification assays for organism detection can be problematic due to the potential for false-positive and false-negative results. Confirmatory testing is often required in these situations, requiring additional time and resources. Use of multigene targets could integrate both detection and verification at lower cost. Utilizing two targets, the MgPa adhesion gene and the 16S rRNA gene, a multitarget real-time (MTRT) PCR for the detection of MG was developed. Samples from patients attending sexually transmitted disease clinics were collected in duplicate. Urine samples from males (n = 286) and self-collected vaginal swabs from females (n = 321) were analyzed by MTRT PCR for MG and the Gen-Probe transmission-mediated amplification (TMA) assay, which targets MG rRNA for detection (TMA-MG research use only). Utilizing the criteria of any two targets being positively amplified, the MTRT PCR had a sensitivity and specificity of 91.8% (101 positive samples/110 samples tested) and 99.5% (495/497), respectively, with a positive predictive value (PPV) of 98.1% (101/103) and a negative predictive value (NPV) of 98.2% (495/504). The Gen-Probe TMA-MG assay had a sensitivity, specificity, PPV, and NPV of 98.1% (108/110), 98.1% (488/497), 92.3% (108/117), and 99.5% (488/490), respectively. Comparison between the MTRT PCR and TMA-MG assay by kappa statistic analysis indicated that an overall kappa value was 0.941 (95% confidence interval, 0.907 and 0.976). Both assays demonstrated accuracy in the detection of MG from urine samples from male patients and self-collected vaginal swabs from female patients.
INTRODUCTION
Mycoplasma genitalium (MG) causes nongonococcal urethritis and has been associated with endometritis and cervicitis (14). MG has been implicated as a cause of several syndromes where etiology cannot be attributed to major sexually transmitted infections (STIs) (1, 13, 14, 20, 22, 28). Various PCR assays have improved clinical diagnostic capabilities with regard to MG (2-9, 14-18, 23, 25, 29-31). PCR can be advantageous in the detection of MG because it is a difficult organism to culture. PCR detection requires only DNA, alleviating the need for viable organisms. Real-time PCR methodology has further improved MG detection (3, 7, 8, 12, 14, 29). Real-time PCR couples simultaneous amplification and detection, resulting in shorter assay time and increased sensitivity and obviating the need for post-PCR processing that is required for traditional PCR assays.
There is concern when comparing new and older amplified tests for organism detection, since it may be difficult to determine a true positive (i.e., a "gold standard"). A secondary test is often utilized to confirm unique positive results (samples positive by one test and negative by another). Use of discrepant testing has been criticized, since it is often performed on a subset of specimens and may introduce statistical bias in terms of sensitivity and specificity (11, 19, 21).
Use of multiple gene targets can offer an alternative to discordant test analysis. Ideally, such an amplified test would consist of two primers and probe sets for simultaneous detection of different gene targets and could provide self-confirming results that could reduce, if not eliminate, the need for discrepant and confirmatory testing.
Multitarget real-time (MTRT) PCR for MG utilizing two alternate gene targets, the MgPa gene and the 16S rRNA gene (14, 29), was designed and used to study the performance of the GenProbe transmission-mediated amplification (TMA) research assay, which targets MG rRNA for detection in genital specimens from patients attending a sexually transmitted disease clinic. MgPa and the 16S rRNA gene were chosen as targets for MTRT PCR because the targets performed well in previous studies in terms of sensitivity and specificity and reproducibility (14, 29).
MATERIALS AND METHODS
Preparation of standards and controls. DNA was extracted from each MG standard (5 x 106 CFU/ml; gift strain from Lynn Duffy, University of Alabama at Birmingham, Birmingham, AL) by utilizing the Roche MagNA Pure LC robotic instrument (Roche Molecular Diagnostics, Indianapolis, IN). Control DNA dilutions ranged in concentration from 5 x 105 CFU/ml to 0.005 CFU/ml. These controls were utilized to determine the limit of detection for each target (MgPA and 16S rRNA) and were also utilized to gauge reproducibility. The MagNA Pure LC DNA Isolation Kit I (Roche) was used, and DNA extraction was carried out according to the instructions supplied for the MagNA Pure LC program, the DNA I Blood Cells High Performance Serum protocol. Negative controls were processed in an identical manner and consisted either of molecular grade H2O or carrier DNA, such as a noncomplementary oligonucleotide.
Patient specimen preparation. There were 615 people enrolled, with urine samples from male patients (n = 290) and self-collected vaginal swabs from female patients (n = 325), collected as part of a cervicitis and urethritis study at sexually transmitted disease clinics. Of the 615 people enrolled, 611 patients were tested, but only 607 patients were utilized for data analysis. The Johns Hopkins University Institutional Review Board and the Baltimore City Health Department approved this study; the study was funded by Gen-Probe Incorporated (San Diego, CA). For MTRT PCR, vaginal swabs from female patients were collected by patients before urine or cervical samples were collected and were shipped in a dry state. Dry swabs were rehydrated in 1 ml of Tris-EDTA buffer, of which 200 μl was removed for DNA extraction. Urine samples from male patients were collected before urethral swabs were collected. Samples (from male and female patients) were then extracted, utilizing the Roche MagNA Pure LC robotic instrument (Roche Molecular Diagnostics, Indianapolis, IN). DNA extraction was carried out according to the instructions supplied for the MagNA Pure LC program, the DNA I Blood Cells High Performance Serum protocol. For the TMA-MG research assay, the duplicate vaginal swabs collected in random order with the vaginal swabs for MTRT PCR and urine samples collected in specimen transport media (provided by Gen-Probe) were processed as for other TMA assays.
MTRT PCR primers and probes. Previously published sets of primers and probes were utilized in the further development of this assay (14, 29). For the MgPa gene target, the primers and probe were MgPa-355F (5'-GAGAAATACCTTGATGGTCAGCAA-3'), MgPa-432R (5'-GTTAATATCATATAAAGCTCTACCGTTGTTATC-3'), and MgPa-380 (5'-TET-ACTTTGCAATCAGAAGGT-MGBNFQ-3'), utilizing tetrabromosulfonefluorescein (TET) as the reporter and a minor groove-binding (MGB) nonfluorescent quencher (NFQ). The primers and probe for the 16S rRNA gene target were My-INS (5'-GTAATACATAGGTCGCAAGCGTTATC-3'), MGSO-2 (5'-CACCACCTGTCACTCGGTTAACCTC-3'), and the MgenP1 probe (5'-6FAM-CTGTCGGAGCGATCCCTTCGGT-MGBNFQ-3') (where 6FAM is a designation for 6-carboxyfluorescein), utilizing a fluorescein reporter and an MGB quencher. All primers and probes were synthesized by Applied Biosystems, Foster City, CA. The sequences of the previously published primers and probes remained unchanged; however, unique reporter molecules were assigned to each probe to ensure that there was a distinct and differentiable signal from each gene target.
PCR conditions. MTRT PCRs were performed with 80 μl of master mixture and 20 μl of template DNA. The master mixture contained 10 μl of 10x QIAGEN HotStarTaq polymerase buffer, 20 μl of 25 mM MgCl2, 8 μl of 25 mM deoxynucleoside triphosphates (dNTPs), 0.5 μl of each 50 μM primer and probe, 36 μl of PCR-grade H2O, 1 μl of QIAGEN Q solution, 2 μl of QIAGEN HotStarTaq polymerase, and 20 μl of template DNA. The 10x buffer, MgCl2, Q solution, and HotStarTaq polymerase were obtained from the QIAGEN HotStarTaq polymerase kit (QIAGEN, Valencia, CA). The dNTPs were obtained from the 100 mM dNTP set PCR Grade (Invitrogen). PCR-grade water was obtained from Quality Biologics Incorporated (Quality Biologics Incorporated, Gaithersburg, MD). PCR was performed using a 96-well plate format on the ABI 7900 HT Sequence Detection system (Applied Biosystems) with the ROX reference removed, under the following conditions: 50°C for 2 min and 95°C for 10 min, followed by 50 cycles, each consisting of 95°C for 15 s, 60°C for 1 min, and 72°C for 30 s. A cooling hold of 4°C for 2 min was added at the end of the cycling protocol for ease of plate handling.
GenProbe TMA testing. The duplicate urine samples from male patients and vaginal swabs from female patients were analyzed by target capture, amplification by TMA, and detection by the hybridization protection assay in a manner similar to procedures for other Gen-Probe Incorporated APTIMA assay family products (Gen-Probe APTIMA Combo2 package insert, IN0037-04 Rev A; Gen-Probe, Inc., San Diego, CA). Primers, probes, and target capture oligomers were designed by Gen-Probe to specifically target MG and were designed to be utilized with reagents that have the same formulation as APTIMA Combo2. Gen-Probe provided the reagents for this project as part of their ongoing research program. The specific MG reagents are not commercially available. The cutoff for positive reactions was set at 40,000 relative light units.
Specificity testing. Mycoplasma pneumoniae (ATTC 15492), Mycoplasma hominis (ATTC 14027), Mycoplasma pirum, Mycoplasma fermentans (PG18), Mycoplasma penetrans, Ureaplasma urealyticum (serotype 10), and Ureaplasma parvum (serotype 3) were evaluated for potential cross-reactivity with the primer and probe sets utilized in the MTRT PCR. Nucleic acid extraction was performed for PCR in the same manner as for the clinical samples. All organisms were evaluated at a high DNA concentration (>20 ng organism DNA/PCR). One reaction per organism was performed, and data evaluation was performed in the same manner as for the clinical samples. The primers and probes for the MgPA assay were previously evaluated against the following organisms: M. pneumoniae (FHT, Mac, M129-B8, M129-B170, and two clinical isolates), Mycoplasma hominis (PG21T, H34, H27, and three clinical isolates), Mycoplasma salivarium (PG20T), Mycoplasma buccale (CH 20247T), Mycoplasma orale (Patt and one clinical isolate), Mycoplasma fermentans (GT and S38), Mycoplasma faucium (DC 333T), Mycoplasma primatum (Navel), Mycoplasma pirum (Zeus), Mycoplasma lipophilum (Maby BT), Mycoplasma penetrans (GTU), Mycoplasma hyorhinis (GDL), Mycoplasma arginini (G230T), Mycoplasma gallisepticum (15302), Mycoplasma iowae (695), Mycoplasma imitans (4229), Mycoplasma testudinis (Hill), Mycoplasma alvi (Isley), and Acholeplasma laidlawii (AT), and Ureaplasma urealyticum (serotypes I [F. Black 7] and VIII [F. Black 960T]) (14). Primers and probes for the 16S rRNA target were previously evaluated against the following organisms: Mycoplasma buccale, Mycoplasma faucium, Mycoplasma fermentans, M. genitalium, Mycoplasma hominis, Mycoplasma lipophilum, Mycoplasma orale, Mycoplasma penetrans, Mycoplasma pirum, Mycoplasma pneumoniae, Mycoplasma primatum, Mycoplasma salivarium, Mycoplasma spermatophilum, Ureaplasma parvum, and Ureaplasma urealyticum (29).
Sample analysis. Positive and negative processing controls were extracted and included in every run of samples utilized for assay validation. Each sample was analyzed by both the MTRT PCR and Gen-Probe TMA-MG assays. True positives were defined a priori as any two positive amplifications from any three results: the Gen-Probe TMA-MG, 16S rRNA, or MgPA gene targets. Kappa statistical analysis was also performed to measure the degree of concordance between the two assays (27). Samples were considered equivocal if the test results did not repeat consistently upon multiple retesting.
RESULTS
The analytical sensitivity of the MTRT PCR was excellent, with positive controls containing 50 CFU/ml and 5 CFU/ml repeating 100% (6/6 samples) of the time for both the MgPa and 16S rRNA targets. Positive controls of 0.5 CFU/ml repeated 100% (6/6) of the time for the MgPa target and 83.3% (5/6) of the time for the 16S rRNA target. Positive controls of 0.05 CFU/ml and 0.005 CFU/ml were positive 100% (6/6) of the time for the MgPA target but were not positive for the 16S rRNA target. It is unclear why the MgPa and the 16S rRNA targets had different analytical sensitivities. There were attempts at additional optimization, including MgCl2 titrations and varying the amount of sample input; however, any changes that improved the analytical sensitivity for the 16S rRNA target also decreased the sensitivity for the MgPa target.
A brief examination of the statistical differences between targets and runs utilizing the generalized estimating equation method revealed that there were no statistical differences between runs for the MgPA target (P = 0.140) and slight differences between the runs for the 16S rRNA target (P = 0.035) (32). Overall, the assay was determined to be reproducible for qualitative screening for MG.
The MTRT PCR also did not cross-react with any of the organisms analyzed for specificity. A National Cancer Biological Institute BLAST search did not reveal cross-reacting primer sequences, and the sensitivity and specificity of the primer and probe sets utilized were analyzed and validated for specificity against a wide range of mycoplasmas and ureaplasmas in previous studies (14, 29).
Of the 615 participants, samples from 607 participants were utilized for analysis. Four samples were excluded from testing due to an insufficient sample, while the other four samples (three samples from male participants and one sample from a female participant) were excluded from analysis after equivocal results were obtained from multiple rounds of retesting by both assays (MTRT PCR and TMA-MG), in which the samples failed to resolve as true positives or true negatives. True positives were defined as any two positive amplifications from any three results; the Gen-Probe TMA-MG, 16S rRNA, or MgPA gene targets were utilized for determination of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). The MTRT PCR had a sensitivity and specificity of 91.8% (101/110 samples) and 99.5% (495/497), respectively, with a PPV of 98.1% (101/103) and an NPV of 98.2% (495/504). For men, sensitivity, specificity, PPV, and NPV were 88.8% (40/45 samples), 100% (241/241), 100% (40/40), and 97.9% (241/246), respectively. For women, the sensitivity, specificity, PPV, and NPV were 93.8% (61/65 samples), 99.2% (254/256), 96.8% (61/63), and 98.4% (254/258), respectively (Table 1).
The Gen-Probe TMA-MG assay had an overall sensitivity, specificity, PPV, and NPV of 98.1% (108/110 samples), 98.1% (488/497), 92.3% (108/117), and 99.5% (488/490), respectively. For men, sensitivity, specificity, PPV, and NPV were 100% (45/45 samples), 97.9% (236/241), 90.0% (45/50), and 100.0% (236/236), respectively. For women, the sensitivity, specificity, PPV, and NPV were 96.9% (63/65 samples), 98.4% (252/256), 94.0% (63/67), and 99.2% (252/254), respectively (Table 1).
There were 59 samples that were uniquely positive for the MgPA target and for the TMA-MG assay that were negative for the 16S rRNA target. These 59 samples failed to resolve as positive for the 16S rRNA target, despite retesting.
Comparison between the MTRT PCR and TMA-MG assay by kappa statistic analysis indicated an overall kappa value of 0.941 (95% confidence interval [CI], 0.907 and 0.976) while it was 0.937 (95% CI, 0.882 and 0.992) for men, and 0.944 (95% CI, 0.900 and 0.988) for women. All kappa analyses indicated "almost-perfect" agreement between the assays (27).
DISCUSSION
There were 110 positive samples identified, resulting in an 18.1% MG prevalence in the study population. Overall, both assays had high sensitivities and specificities at 91.8% (101/110 samples) and 99.5% (495/497) for the MTRT PCR assay and 98.1% (108/110) and 98.1% (488/497) for the Gen-Probe TMA MG assay.
Kappa analysis indicated an "almost-perfect" agreement between both assays with an overall value of 0.941 (95% CI, 0.907 and 0.976), indicating that either test is suitable for MG screening in research settings. These tests could be useful in future studies examining the associations of MG with urethritis, cervicitis, and other STIs.
There are several recently developed real-time methods available for MG detection (3, 8-10, 14, 26, 29, 31). Direct comparison of the sensitivity and specificity of the MTRT PCR or TMA-MG assay with the other published assays for MG detection is difficult because most assays do not include comparison with a gold standard. This is because MG culture is very difficult to perform, and there is no nucleic acid amplification test that is accepted as the gold standard for MG detection. However, in terms of limit of detection, both the MTRT PCR and TMA-MG assays are comparable with other published assays, whose limits of detection range from 1 genomic copy to 50 genomic copies (3, 8-10, 14, 26, 29, 31).
There is some limitation in the overall comparisons that can be made between the assays because different sample types were involved for each sex. Samples from male participants consisted of urine only, whereas samples from female participants were self-collected vaginal dry swabs, which potentially contain higher organism concentrations. Despite the differences in sample type, kappa analysis indicated "almost-perfect" agreement between the assays when examined overall and by sex.
Urine samples from male participants (n = 290) and self-collected vaginal swabs from female participants (n = 325) appear to be appropriate sample types for M. genitalium detection, although male urethral swabs and female cervical swabs were not evaluated in this study. Future studies should address these comparisons. Self-collected vaginal swabs shipped in a dry state could be particularly important for outreach programs targeting people without access to health clinics.
There were four samples (three samples from male participants and one sample from a female participant) that were excluded from analysis for failing to consistently repeat as positive or negative, despite multiple rounds of retesting. These results are most likely the result of low MG loads in the sample, resulting in sampling error due to Poisson distribution of the target in replicate samples.
It is of interest that 59 samples were positive only by the TMA-MG assay and the MgPA target and negative for the 16S rRNA target by the MTRT PCR. During assay development and validation, the 16S rRNA target had an excellent limit of detection (between 5 and 0.5 CFU/ml) and was determined to be a target that yielded reproducible results, based on generalized estimating equation analysis. Additionally, it is unclear why the MgPa and the 16S rRNA targets had different analytical sensitivities at <1 CFU/ml for the MgPa target and between 5 and <1 CFU/ml for the 16S rRNA target. Reoptimization of the assay to increase the sensitivity for the 16S rRNA target only served to decrease the sensitivity for the MgPa target. The primer and probe sets for each target were taken from previously published assays for MG (14, 29), and each set had different optimal annealing temperatures at 60°C and 66°C, respectively. The MTRT PCR utilized an annealing temperature of 55°C to emphasize sensitivity, and it may be the discrepancy between the annealing temperatures that accounts for different analytical sensitivities of each target. Or, the region of the 16S rRNA target of MG is a less-sensitive target choice for MG detection from clinical samples than other targets, such as MgPA (26). Future studies comparing the efficiency of 16S rRNA-based detection versus alternative targets for MG detection should resolve this issue. Although the MTRT PCR was not 100% self confirming, the ability of an assay to detect multiple targets simultaneously should be incorporated into future molecular detection assays, not only for MG but for other pathogens as well, because amplification of multiple, alternative targets from the same sample would increase confidence in assay results and decrease the need for costly and time-consuming confirmation testing.
There is growing evidence to suggest that M. genitalium does have a direct involvement as a STI in women and is also involved in urethritis and cervicitis (24). Additionally, there are suggested associations between M. genitalium, pelvic inflammatory disease, and infertility (24). If the potential links between MG and urethritis, cervicitis, pelvic inflammatory disease, and infertility are to be concretely established, improved diagnostics with high sensitivity and specificity for M. genitalium will be required, as culture methods are difficult.
The TMA-MG research assay performed very well in comparison with the MTRT PCR research assay and has the potential to provide clinicians with a commercially available and highly accurate diagnostic tool for future M. genitalium research. Because it does not require viable organisms and has the potential to be self confirming, the integrated MTRT research PCR represents a cost-effective method for M. genitalium detection and verification, especially in research studies.
REFERENCES
Bjornelius, E., P. Lidbrink, and J. S. Jensen. 2000. Mycoplasma genitalium in nongonococcal urethritis—a study in Swedish male STD patients. Int. J. STD AIDS 11:292-296.
Blanchard, A., W. Hamrick, L. Duffy, K. Baldus, and G. H. Cassell. 1993. Use of the polymerase chain reaction for detection of Mycoplasma fermentans and Mycoplasma genitalium in the urogenital tract and amniotic fluid. Clin. Infect. Dis. 17(Suppl. 1):S272-S279.
Blaylock, M. W., O. Musatovova, J. G. Baseman, and J. B. Baseman. 2004. Determination of infectious load of Mycoplasma genitalium in clinical samples of human vaginal cells. J. Clin. Microbiol. 42:746-752.
Busolo, F., D. Camposampiero, G. Bordignon, and G. Bertollo. 1997. Detection of Mycoplasma genitalium and Chlamydia trachomatis DNAs in male patients with urethritis using the polymerase chain reaction. New Microbiol. 20:325-332.
Cadieux, N., P. Lebel, and R. Brousseau. 1993. Use of a triplex polymerase chain reaction for the detection and differentiation of Mycoplasma pneumoniae and Mycoplasma genitalium in the presence of human DNA. J. Gen. Microbiol. 139:2431-2437.
de Barbeyrac, B., C. Bernet-Poggi, F. Febrer, H. Renaudin, M. Dupon, and C. Bebear. 1993. Detection of Mycoplasma pneumoniae and Mycoplasma genitalium in clinical samples by polymerase chain reaction. Clin. Infect. Dis. 17(Suppl. 1):S83-S89.
Deguchi, T., T. Yoshida, S. Yokoi, M. Ito, M. Tamaki, H. Ishiko, and S. Maeda. 2002. Longitudinal quantitative detection by real-time PCR of Mycoplasma genitalium in first-pass urine of men with recurrent nongonococcal urethritis. J. Clin. Microbiol. 40:3854-3856.
Dupin, N., G. Bijaoui, M. Schwarzinger, P. Ernault, P. Gerhardt, R. Jdid, S. Hilab, C. Pantoja, M. Buffet, J. P. Escande, and J. M. Costa. 2003. Detection and quantification of Mycoplasma genitalium in male patients with urethritis. Clin. Infect. Dis. 2003. 37:602-605.
Dutro, S. M., J. K. Hebb, C. A. Garin, J. P. Hughes, G. E. Kenny, and P. A. Totten. 2003. Development and performance of a microwell-plate-based polymerase chain reaction assay for Mycoplasma genitalium. Sex. Transm. Dis. 30:756-763.
Eastick, K., J. P. Leeming, E. O. Caul, P. J. Horner, and M. R. Millar. 2003. A novel polymerase chain reaction assay to detect Mycoplasma genitalium. Mol. Pathol. 56:25-28.
Green, T. A., C. M. Black, and R. E. Johnson. 1998. Evaluation of bias in diagnostic-test sensitivity and specificity estimates computed by discrepant analysis. J. Clin. Microbiol. 36:375-381.
Heid, C. A., J. Stevens, K. J. Livak, and P. M. Williams. 1996. Real time quantitative PCR. Genome Res. 6:986-994.
Horner, P. J., C. B. Gilroy, B. J. Thomas, R. O. Naidoo, and D. Taylor-Robinson. 1993. Association of Mycoplasma genitalium with acute non-gonococcal urethritis. Lancet 342:582-585.
Jensen, J. S., E. Bjornelius, B. Dohn, and P. Lidbrink. 2004. Use of TaqMan 5' nuclease real-time PCR for quantitative detection of Mycoplasma genitalium DNA in males with and without urethritis who were attendees at a sexually transmitted disease clinic. J. Clin. Microbiol. 42:683-692.
Jensen, J. S., M. B. Borre, and B. Dohn. 2003. Detection of Mycoplasma genitalium by PCR amplification of the 16S rRNA gene. J. Clin. Microbiol. 41:261-266.
Jensen, J. S., S. A. Uldum, J. Sondergard-Andersen, J. Vuust, and K. Lind. 1991. Polymerase chain reaction for detection of Mycoplasma genitalium in clinical samples. J. Clin. Microbiol. 29:46-50.
Luki, N., P. Lebel, M. Boucher, B. Doray, J. Turgeon, and R. Brousseau. 1998. Comparison of polymerase chain reaction assay with culture for detection of genital mycoplasmas in perinatal infections. Eur. J. Clin. Microbiol. Infect. Dis. 17:255-263.
Mahony, J. B., D. Jang, S. Chong, K. Luinstra, J. Sellors, M. Tyndall, and M. Chernesky. 1997. Detection of Chlamydia trachomatis, Neisseria gonorrhoeae, Ureaplasma urealyticum, and Mycoplasma genitalium in first-void urine specimens by multiplex polymerase chain reaction. Mol. Diagn. 2:161-168.
McAdam, A. J. 2000. Discrepant analysis: how can we test a test J. Clin. Microbiol. 38:2027-2029.
Mena, L., X. Wang, T. F. Mroczkowski, and D. H. Martin. 2002. Mycoplasma genitalium infections in asymptomatic men and men with urethritis attending the New Orleans Sexually Transmitted Disease Clinic. Clin. Infect. Dis. 35:1167-1173.
Miller, W. C. 1998. Bias in discrepant analysis: when two wrongs don't make a right. J. Clin. Epidemiol. 51:219-231.
Morency, P., M. J. Dubois, and G. Gresenguet. 2001. Aetiology of urethral discharge in Bangui, Central African Republic. Sex. Transm. Infect. 77:125-129.
Palmer, H. M., C. B. Gilroy, P. M. Furr, and D. Taylor-Robinson. 1991. Development and evaluation of the polymerase chain reaction to detect Mycoplasma genitalium. FEMS Microbiol. Lett. 61:199-203.
Ross, J. D. C. 2005. Is Mycoplasma genitalium a cause of pelvic inflammatory disease. Infect. Dis. Clin. N. Am. 19:407-413.
Sharma, S., R. Brousseau, and S. Kasatiya. 1998. Detection and confirmation of Mycoplasma pneumoniae in urogenital specimens by PCR. J. Clin. Microbiol. 36:277-280.
Svenstrup, H. F., J. S. Jensen, E. Bjrnelius, P. Lidbrink, S. Birkelund, and C. Gunna. 2005. Development of a quantitative real-time PCR assay for detection of Mycoplasma genitalium. J. Clin. Microbiol. 43:3121-3128.
Szklo, M. (ed.). 2000. Epidemiology beyond the basics, p. 343-404. Aspen Publishers, Inc., Gaithersburg, MD.
Totten, P. A., M. A. Schwartz, and K. E. Sjostrom. 2001. Association of Mycoplasma genitalium with nongonococcal urethritis in heterosexual men. J. Infect. Dis. 183:269-276.
Yoshida, T., T. Deguchi, M. Ito, S. I. Maeda, M. Tamaki, and H. Ishiko. 2002. Quantitative detection of Mycoplasma genitalium from first-pass urine of men with urethritis and asymptomatic men by real-time PCR. J. Clin. Microbiol. 40:1451-1455.
Yoshida, T., S. Maeda, T. Deguchi, and H. Ishiko. 2002. Phylogeny-based rapid identification of mycoplasmas and ureaplasmas from urethritis patients. J. Clin. Microbiol. 40:105-110.
Yoshida, T., S. Maeda, T. Deguchi, T. Miyazawa, and H. Ishiko. 2003. Rapid detection of Mycoplasma genitalium, Mycoplasma hominis, Ureaplasma parvum, and Ureaplasma urealyticum organisms in genitourinary samples by PCR-microtiter plate hybridization assay. J. Clin. Microbiol. 41:1850-1855.
Zeger, S. L., and K. Y. Liang. 1986. Longitudnal data analysis for discrete and continuous outcomes. Biometrics 42:121-130.(Justin Hardick, Julie Gil)