URETHRITIS IS ONE OF THE most common conditions among men presenting at sexually transmitted disease (STD) clinics. In Scandinavia, gonococcal urethritis has diminished drastically during recent years. Consequently, most of the patients now present with nongonococcal urethritis (NGU). Chlamydia trachomatis is found in 20% to 35% of the NGU patients, but in the remaining two thirds of the patients, the etiology remains unclear.
Mycoplasma genitalium is gaining increasing interest as a pathogen in male urethritis. Several studies1–7 have shown a significant association between the presence of M. genitalium DNA and symptoms and/or signs of urethritis (recently reviewed by Jensen8) Since its first isolation reported in 1981,9,10 this fastidious and slow-growing mycoplasma species has proven very difficult to isolate by culture. Since the initial isolations, very few studies have reported successful isolation from the urogenital tract using cell culture to assist growth.7,11 Therefore, clinical studies need to rely on nonculture methods such as polymerase chain reaction (PCR).12,13 One of the main obstacles for increasing the knowledge about M. genitalium has been the lack of commercially available diagnostic tests and a common reference standard for evaluation of in-house tests. Furthermore, of the clinical studies that have been reported thus far, some have used swab specimens, whereas others have used first void urine (FVU), further complicating comparison. As part of a study on the prevalence of M. genitalium among patients attending a STD clinic, we decided to evaluate the efficacy of FVU as compared with swab specimens for detection of M. genitalium and C. trachomatis using in-house inhibitor-controlled PCR assays and subsequent confirmation of all positive findings with independent PCR assays. In a subset of the specimens, we aimed to estimate the effect of freezing of the FVU before and after sample preparation for PCR, which could have important implications in the interpretation of results obtained from archival specimens.
Materials and Methods
Patients and Samples
Between August 1997 and November 2001, 1937 male and 776 female patients attending the outpatient STD clinic at Huddinge University Hospital were enrolled in the study. Independent of the reason for attending the clinic and after informed consent, screening for urethritis, cervicitis, Neisseria gonorrhoeae, C. trachomatis, and M. genitalium was performed. The local ethical committee approved the study.
From the male patients, urethral smears were taken with a plastic loop and stained with methylene blue. A smear showing 5 or more polymorphonuclear leukocytes (PMNL) per high-power (1000×) microscopic field (hpf) in 5 or more fields was considered positive for the presence of urethritis.
After the smear was taken, a urethral swab specimen was collected for culture of N. gonorrhoeae. Secondly, an ear, nose and throat (ENT) cotton-tipped aluminium swab was inserted 3 to 4 cm into the urethra and subsequently placed in a tube with 1.8 mL of SP4 mycoplasma broth medium.11 This specimen was used for detection of M. genitalium and C. trachomatis by PCR. After the clinical examination, the patient was asked to collect 15 mL of the first voided urine; 11 mL was used for routine detection of C. trachomatis by PCR at the local diagnostic laboratory, and the remaining FVU was sent together with the urethral swab specimen by ordinary mail to the laboratory in Copenhagen.
For women, urethral and cervical smears were taken with the swab used for culture of N. gonorrhoeae. A urethral smear showing ≥5 PMNL/hpf was considered positive for the presence of urethritis and likewise, a cervical smear containing ≥30 PMNL/hpf was considered indicative of cervicitis. Second, urethral and cervical ENT swabs were collected in 2SP chlamydia transport medium12 for culture and later for PCR for C. trachomatis at the local laboratory. Third, the same sites were sampled with ENT swabs collected in SP4 for detection of M. genitalium and C. trachomatis by PCR in Copenhagen. After the clinical examination, 4 mL of the first voided urine was collected and sent together with the swab specimens to Copenhagen.
No information about the clinical status of the patients was given to the laboratory.
The same patient was allowed to contribute more than 1 specimen set. However, in the data analysis, only the specimen collected at the first visit was used in the evaluation, and specimens collected at later visits were analyzed separately regardless of the time between the visits.
Clinical information was retrospectively collected by chart review for M. genitalium PCR-positive patients to compare the diagnostic performance of the different specimen types in relation to clinical signs and symptoms. No attempts were made to compare M. genitalium- or C. trachomatis-negative patients.
Polymerase Chain Reaction Methods
Sample preparation for PCR was performed by mixing 100 μL of the specimen in SP4 medium with 300 μL of a 20% w/v Chelex 100 slurry (BioRad, Richmond, CA) in TE buffer (10 mmol/L Tris-HCl [pH 8.0], 1 mmol/L EDTA); the mixture was vortexed vigorously for 60 seconds and incubated at 95°C for 10 minutes. After a brief centrifugation, 10 μL of the supernatant corresponding to approximately 3 μL of the original specimen was used for PCR. Urine specimens were concentrated by centrifugation; 1.8 mL was centrifuged at 20.000 × g for 15 minutes. Most of the supernatant was discarded by careful aspiration with a Pasteur pipette taking care not to dislodge the pellet. The tubes were left to stand for at least 1 minute to collect the remaining urine at the bottom of the tube. Subsequently, the remaining urine was aspirated with a 200-μL pipette leaving only a minimal amount of urine in the tube. Three hundred microliters of Chelex 100 slurry was added and treated as the swab specimens; 10 μL of the supernatant corresponding to approximately 60 μL of the original specimen was used for PCR.
M. genitalium DNA was detected by a PCR with primers deduced from the 16 S rRNA gene sequence of M. genitalium.14 An internal control for inhibition (internal processing control [IPC]) was included to avoid false-negative results as a result of Taq polymerase inhibitors. All positive results were confirmed by a second PCR amplifying a part of the M. genitalium MgPa adhesin gene with primers MgPa-1 and MgPa-3 as previously described.12,14 The 2 assays were consistently capable of detecting <5 genome equivalents when purified M. genitalium DNA was added to PCR-negative clinical specimens.
Positive controls contained 5 and 50 genome equivalents of M. genitalium-purified DNA; negative controls were included in each run.
C. trachomatis DNA was detected by amplification of a sequence of the cryptic plasmid using primers CP24 (5′GGG ATT CCT GTA ACA ACA AGT CAG G) and CP27 (5′CCT CTT CCC CAG AAC AAT AAG AAC AC).15 The reaction conditions were 1× PCR buffer (SuperTaq; HT Biotechnology, Cambridge, UK) with 2.5 mmol/L MgCl2, 0.4 μmol/L of each primer, 125 μmol/L dATP, dGTP and dCTP, 250 μmol/L dUTP, IPC, and 0.5 U of Taq DNA polymerase (SuperTaq) in a total volume of 100 μL. Amplification was performed in an Omnigene (Hybaid, Teddington, UK) thermocycler running 40 cycles each consisting of a 94°C/30 seconds denaturation and a 60°C/60 seconds annealing and extension, and a final extension step at 72°C for 5 minutes. Positive results were confirmed with an inhibitor-controlled PCR detecting the 16S rRNA gene of chlamydia species modified from Pollard et al.16,17
Strict physical separation between PCR setup and post-PCR analysis laboratories were maintained as previously described.3 Sterile filter tips (ART; SDS, Falkenberg, Sweden) were used in all manipulations with the samples. Working surfaces in the PCR setup laboratory were regularly wiped with a 4% Diversol solution containing hypochlorite18 and exposed to ultraviolet light between sessions with the purpose of destroying contaminating DNA
Interpretation of Results
A patient was deemed positive if both the screening and the confirmatory PCR assay were positive. Both the swab and the FVU specimens were tested in the confirmatory assay if 1 of the specimens had been positive in the screening assay. Even if specimens belonging to the same patient were positive in the confirmatory assay, the specimen set was deemed discrepant to mimic the situation in which only 1 type of specimen was analyzed from each patient. A specimen was considered negative if the IPC was amplified. If the IPC failed to amplify, the assay was repeated with 5 and 2 μL of the pretreated specimen. Specimens that remained IPC negative with 2 μL were deemed inhibitory.
Evaluation of the Sample Preparation Method for Swab Specimens
Prompted by the rather low sensitivity of the M. genitalium PCR on urethral swab specimens, 15 urethral swab specimens were selected from male patients with an initial negative PCR result from the swab specimen but a positive result from the FVU. These specimens had been stored at -80°C for up to 5 years, and after thawing, they were subjected to a repeated standard swab sample preparation as described here as well as to a modified FVU sample preparation method. The modification consisted of using half the volume both of the SP4 medium and of the Chelex slurry. This was necessary as a result of the limitations in the amount of remaining SP4 medium. A subset of 100 M. genitalium concordantly negative swab specimens was also subjected to this modified sample preparation method.
Evaluation of the Impact of Freezing Specimens
A subset of 102 male and 22 female Chelex extracted FVU specimens prepared and tested immediately after receipt in the laboratory were tested after storage at -20°C for 1 to 18 months. From the same patients, a sufficient volume of the original FVU specimen that had been stored also at -20°C for 1 to 18 months was available from 68 male and 15 female patients. These specimens were thawed and subjected to sample preparation using the same methods as for the fresh specimens. In an attempt to estimate the reproducibility of the sample preparation method, 30 FVU specimens previously found M. genitalium-positive but not part of the current study were reexamined. The specimens had been stored at -20°C for up to 1 year. After thawing, 2 sample preparations were performed as per the standard protocol and tested in the 16S rRNA gene PCR assay. Only results obtained at the first test of the specimen were considered in the evaluation.
Fisher exact test was used to test for differences in proportions; McNemar’s test was used to compare paired proportions, but the exact P value of McNemar’s test statistic, based on the binomial distribution, was used (Liddell’s test); Cochran’s Q test was used for comparison of paired data with more than 2 groups, and the Mann-Whitney test was used to test for differences in continuous variables between groups. These tests were performed with the StatsDirect software version 2,2,7 except for Cochran’s Q test performed with SAS version 8.2, which was also used for logistic regression analysis using the Proc Logistic module.
The 1937 men attended the clinic for 2332 visits (range, 1–10). A complete specimen set comprising FVU and urethral swab specimen was obtained at 2202 appointments; in 1852 of these appointments, the patient attended at his first visit. Information regarding age was missing from 6 men, 1 positive for M. genitalium, 1 positive for C. trachomatis, and 4 men negative for both pathogens. Of the remaining 350 specimen sets, 223 were collected at visit 2, 76 at visit 3, 28 at visit 4, 14 at visit 5, and 9 at subsequent visits.
The 776 women attended the clinic for 865 visits (range, 1–5). A complete specimen set comprising FVU, a cervical and a urethral swab specimen was obtained at 828 appointments; in 753 of these appointments, the patient attended her first visit. Of the remaining 75 specimen sets, 57 were collected at visit 2, 12 at visit 3, 5 at visit 4, and 1 at visit 5.
Mycoplasma genitalium in Men
A total of 126 (6.8%) of the 1852 patients were positive for M. genitalium in a urine or urethral swab specimen at their first visit. M. genitalium-positive men had a median age of 27.8 years, not statistically different from the negative men who had a median age of 28 years (P = 0.7). The age-specific prevalence of M. genitalium is shown in Figure 1. The age distribution was significantly different from that of C. trachomatis-infected men. Information regarding previous episodes of NGU was available for 93 M. genitalium positive men of whom 12 reported an earlier episode. The median age of patients with previous NGU was 27.9 years compared with those experiencing their fist episode of urethritis who had a median age of 28 years. Thus, the different age distribution could not be attributed to patients with recurrent, longstanding complaints.
One hundred one were positive in both urine and urethral swab, 22 were positive in urine only, and 3 were positive in the urethral swab only (Table 2). Thus, the relative sensitivity for urine was 97.6% as compared with a urethral swab sensitivity of 82.5% (P = 0.0002). Nine (0.5%) urethral swab specimens were inhibitory in the M. genitalium PCR but were amplifiable in the C. trachomatis PCR, in which 1 was positive. The corresponding urine specimens were amplifiable and all were M. genitalium-negative. Twelve (0.6%) urine specimens were inhibitory in the M. genitalium PCR; 3 were also inhibitory in the C. trachomatis PCR and 1 was C. trachomatis PCR-positive. The corresponding urethral swab specimens were all M. genitalium PCR-negative (Table 1).
There was no statistically significant difference in the age distribution between men with discrepant (median, 25.7 years) and concordant (median. 28 years) specimens (P = 0.96).
Among the 25 M. genitalium PCR-discrepant specimen sets, 9 (36%) were C. trachomatis PCR-positive as compared with 12 (12%) of the 101 concordant specimen sets (P = 0.007).
Clinical information was available for 118 of the 126 M. genitalium-positive patients (21 with discrepant results and 97 with concordant results). Selected clinical variables are listed in Table 3; most importantly, men with discrepant specimen sets were less likely to present with self-reported (P = 0.03) or observed discharge (P = 0.02). Observed discharge was found in 27 (24%) of 114 men with M. genitalium infection not different from the 49 (21%) of 234 men with C. trachomatis infection when this information was available. However, among the 21 men with a dual infection by M. genitalium and C. trachomatis, only 1 (4,7%) had observed discharge. Thus, men with a dual infection were less likely to present with discharge than were those with M. genitalium infection alone (P = 0.02) or with C. trachomatis infection alone, although it not statistically significant (P = 0.08). Concordance between specimens positive for M. genitalium was analyzed in a multiple logistic regression model using exact tests and forward selection with the following variables: C. trachomatis infection status, the presence of observed discharge, urethritis, self-reported urethral itch, dysuria, use of tetracycline or macrolide antibiotics within the previous 3 months, and the continuous variables age (grouped as <26, 26–<32, ≥32), number of partners within the last 6 months (grouped as 1, 2, 3, ≥4), and transport time in days (grouped as <3, 3, 4, ≥5). The grouping of continuous variables was defined by limits creating similarly sized groups. In this analysis, only a positive C. trachomatis infection status remained statistically significantly associated with a discordant M. genitalium PCR result. For this variable, the odds ratio (OR) was 2.16 (95% confidence interval [CI], 1.51–11.51).
Considering all 2002 specimen sets collected from the 1852 men at any visit resulted in a total of 203 M. genitalium PCR-positive specimen sets, but did not change the conclusions regarding performance of urethral swab specimens compared with FVU (Table 2).
Chlamydia trachomatis in Men
A total of 246 (13.3%) of the 1852 patients were positive for C. trachomatis in urine or urethral swab specimen at their first visit. The C. trachomatis-positive men were significantly younger (median age, 25.2 years) than the C. trachomatis-negative men (median age, 28.4 years) (P <0.0001) and younger than the M. genitalium-positive men (P = 0.0003). The age-specific prevalence of C. trachomatis in men is shown in Figure 1. Of the 246 positive men, 216 were positive in both urine and urethral swab; 26 were positive in urine only, and 4 were positive in the urethral swab only (Table 2). Thus, the relative sensitivity for urine was 98.4% as compared with a urethral swab sensitivity of 89.4% (P <0.0001).
Inhibition of urethral swab specimens was not observed in the C. trachomatis PCR, and only 3 (0.2%) of the urine specimens were inhibitory in the C. trachomatis PCR; these 3 were also inhibitory in the M. genitalium PCR. The corresponding urethral swab specimens were C. trachomatis and M. genitalium PCR-negative (Table 1).
Men with discrepant specimens were older (median, 27.7 years) than were men with concordant specimens (median, 25 years) (P = 0.02).
Among the 30 C. trachomatis PCR-discrepant specimen sets, 2 (6.7%) were M. genitalium PCR-positive as compared with 19 (8.8%) of the 216 concordant specimen sets (P = 0.8).
Clinical information was available for 244 of the 246 C. trachomatis-positive patients (30 with discrepant results and 214 with concordant results). Selected clinical variables are listed in Table 3; most importantly, men with discrepant specimen sets were less likely to have urethritis (P = 0.007). As for the M. genitalium-positive men, the C. trachomatis-infected men with concordant specimen sets more often had reported or observed urethral discharge; for the men with chlamydia, this difference, however, did not reach statistical significance (P = 0.07). Concordance between specimens positive for C. trachomatis was analyzed in a multiple logistic regression model using exact tests and forward selection with the same variables as for the M. genitalium specimens, except that M. genitalium infection status was included and that the continuous variable “age” was grouped differently (grouped as <24, 24–<28, ≥28) because of the different age distribution of C. trachomatis-positive men. In the analysis, only the presence of urethritis remained statistically significantly associated with a concordant C. trachomatis PCR result. For this variable, the OR was 3.10 (95% CI, 1.38–6.99). Considering all 2002 specimen sets collected at any visit resulted in a total of 270 C. trachomatis PCR-positive specimen sets, but did not change the conclusions regarding performance of urethral swab specimens compared with FVU (Table 2).
Mycoplasma genitalium in Women
A total of 51 (6.8%) of the 753 women attending the clinic for their first visit were positive for M. genitalium in 1 or more of the FVU, cervical, or urethral swab specimens. M. genitalium-positive women were younger (median age, 23.3 years) than M. genitalium-negative women (median age, 25.8 years) (P = 0.015), but not different from the C. trachomatis-positive women (median age, 23.3 years). The age-specific M. genitalium prevalence is shown in Figure 2. Only 20 women (39%) were positive in all 3 specimen types (Table 5). Urethral swabs detected 29 (57%) of the infections, cervical swabs detected 36 (71%), and FVU detected 45 (88%) of the infected women (Table 4). Using Cochran’s Q test, the performances of the 3 different specimen types were statistically different (P = 0.002). By pairwise comparison using McNemar’s test, the cervical swab specimen detected more infections than the urethral swab; however, this difference did not reach statistical difference (P = 0.21). In contrast, the FVU specimen was significantly more efficient than both the urethral (P = 0.0009) and the cervical swab specimen (P = 0.049). Two women were positive in the urethral swab only, 3 in the cervical swab only, and 7 in FVU only. Thus, if using our sample preparation and PCR procedures, the FVU specimen should always be collected and could be supplemented either with a urethral swab or with a cervical swab specimen to increase sensitivity (Table 6). Two urethral swabs (0.3%) were inhibitory in the M. genitalium PCR compared with 1 (0.1%) cervical swab and 2 (0.3%) FVU specimens (Table 4). Eight (16%) of the women were simultaneously positive for C. trachomatis in 1 or more of the specimens. A positive C. trachomatis infection status showed a trend toward being associated with having an M. genitalium false-negative urethral swab specimen, because 6 of 21 patients with a negative urethral swab were C. trachomatis-positive compared with 2 of 29 with a positive urethral swab (P = 0.05).
Considering all 828 specimen sets collected from the 753 women at any visit resulted in a total of 61 M. genitalium PCR-positive specimen sets, but did not change the conclusions regarding performance of swab specimens compared with FVU (Tables 5 and 6).
Chlamydia trachomatis in Women
A total of 73 (9.7%) of the 753 women attending the clinic for their first visit were positive for C. trachomatis in 1 or more of the FVU, cervical, or urethral swab specimens. C. trachomatis-positive women were younger (median age, 23.3 years) than C. trachomatis-negative women (median age, 25.9 years) (P <0.0001). The age-specific C. trachomatis prevalence is shown in Figure 2. Only 51 (70%) were positive in all 3 specimen types (Table 5). Urethral swabs detected 57 (78%) of the infections, cervical swabs detected 63 (86%), and FVU detected 66 (90%) of the infected women (Table 4). Using Cochran’s Q test, the performance of the 3 different specimen types was only borderline statistically different (P = 0.057). By pairwise comparison, the cervical swab specimen detected more infections than the urethral swab, but using McNemar’s test, this difference did not reach statistical difference (P = 0.24). The FVU specimen was significantly more efficient than the urethral (P = 0.01) but not more efficient than the cervical swab specimen (P = 0.61). One woman was positive in the urethral swab only, whereas 6 and 4 were positive in the cervical swab only and in the FVU only, respectively. Thus, the cervical swab specimen should always be collected and could be supplemented with a urethral swab, or preferably with an FVU specimen to increase sensitivity (Table 6). Inhibition of the C. trachomatis PCR was not observed in any of the specimens.
Considering all 828 available specimen sets collected at any visit only added 2 C. trachomatis PCR-positive specimen sets, and thus, did not contribute to further information (Tables 5 and 6).
Evaluation of the Sample Preparation Method for Swab Specimens
Of the 15 M. genitalium-discrepant male urethral swab specimens used for evaluation of the efficiency of the sample preparation method, 5 (33%) were positive in the confirmatory PCR at initial testing. However, because they were initially negative in the screening assay, they were considered false-negative in the evaluation. After repeated sample preparation using the standard procedure, 2 were positive in the 16S rRNA gene PCR, whereas 9 were positive when the specimens were treated according to the modified FVU procedure (P = 0.02). Thus, concentrating the swab specimen by centrifugation might improve the sensitivity of this specimen type. Furthermore, when 100 urethral swab specimens from concordant-negative specimen sets were prepared with the modified FVU procedure, 2 additional positives were identified. When the original sample preparation of urethral and FVU specimens from these 2 patients were retested, both the urethral swab and the FVU were found positive either in the screening or in the confirmatory assay indicating that they were false-negative in the initial assay.
Evaluation of the Impact of Freezing Specimens
Male First Void Urine Specimens.
Of the 102 M. genitalium-positive FVU specimens subjected to Chelex extraction immediately after receipt in the laboratory, 96 (94%) remained positive after storage at -20°C for 1 to 18 months. When 68 of the corresponding original specimens were thawed and subjected to Chelex extraction, 5 were PCR-negative. These 5 specimens were among the 6 that tested PCR-negative after storage and repeated PCR testing. For 1 of the 6 specimens found false-negative after freezing the extracted specimen, the amount of urine was insufficient to retest. As judged by the intensity of the PCR amplicon in the primary assay, the specimens that became negative after storage contained very few genome copies of M. genitalium DNA. The reproducibility of the FVU sample preparation method was excellent because 29 of the 30 FVU specimens that were subjected to sample preparation in duplicate gave concordant results. The discordant specimen contained only 1 and 2 genome copies per 5 μL pretreated specimens as estimated by a quantitative PCR assay19 and on repeat testing in the 16S rRNA gene screening assay both preparations became positive. Two of the FVU specimens had become concordantly negative after storage. Thus, for male FVU specimens, there was no statistically significant difference in the sensitivity regardless if the specimens were tested after storage of the extracted specimen or of the FVU specimen. Furthermore, the lack of reproducibility for specimens found positive directly after sample preparation was no more pronounced than the lack of reproducibility among specimens extracted in duplicate. This variability was particularly a problem for specimens containing M. genitalium DNA close to the limit of detection.
Female First Void Urine Specimens.
All of the 22 female M. genitalium-positive FVU specimens subjected to Chelex extraction immediately after receipt in the laboratory were found to be repeatedly positive after storage at -20°C. Of the 15 corresponding original FVU specimens available for retesting, 11 (73%) remained positive after storage of the original FVU specimen at -20°C (P = 0.001). Thus, it appears that the optimal sensitivity on female urine could be obtained only if the specimens are subjected to sample preparation immediately after receipt.
The main purpose of the current study was to evaluate the relative performance of FVU compared with urethral and cervical swab specimens for detection of M. genitalium in men and women. Several studies addressing this question have been carried out for C. trachomatis with different testing methods,20–22 and because we used an in-house PCR method, our findings regarding this pathogen could be of less importance, but to our knowledge, this is the first attempt to clarify this issue for M. genitalium on a larger number of patients. For both pathogens, PCR tests with internal processing controls for detection of inhibition and amplification failures were applied. In addition, all positive results were confirmed with a pair of PCR primers targeting an unrelated sequence. In a routine diagnostic setting, the internal processing control is important for providing a high sensitivity, and the inclusion of a confirmatory assay increases the specificity of the tests,23 which is particularly important when dealing with sexually transmitted infections. The application of a confirmatory test, however, could slightly lower the sensitivity, because patients with a low burden of the infectious agent could be deemed negative because of sampling errors. Examples were seen in a few cases during the study, in which false-negative results of the confirmatory test were suspected as a result of the concurrent finding of a confirmed positive test in 1 of the other specimens examined. Likewise, false-negative results in the primary test were also seen as documented by the finding of a positive repeat test using the primary assay or a positive result of the confirmatory assay. The confirmatory assay was always performed on the complete set of specimens, but to make the evaluation of the performance of the different specimen types faithful to the situation in a routine setting, we decided to register only results that were positive in the primary assay and subsequently confirmed as true-positives.
In male patients, FVU was found to be superior to the urethral swab specimen detecting 97.6% and 98.4% of the M. genitalium- and C. trachomatis-infected patients, respectively. For detection of M. genitalium, a simultaneous C. trachomatis infection was found to be associated in a multiple logistic regression analysis with a discrepant specimen set, ie, a false-negative urethral swab. This was probably caused by the surprising lack of discharge in patients with a dual M. genitalium and C. trachomatis infection, the reason for which is unclear. The difficulties in detecting M. genitalium in specimens from patients coinfected with C. trachomatis and/or N. gonorrhoeae was also reported by Mena et al.24 They found that patients with negative PCR results as deemed from agarose gel electrophoresis but positive by Southern blot test, indicating either a low M. genitalium DNA load or partial inhibition, were more likely to be coinfected.
Apparently, the sample preparation method for swab specimens could be improved by concentrating the specimen by centrifugation. Obviously, the approach of retesting the negative swab specimen from a discrepant specimen set could introduce bias in the evaluation, and the finding of additionally 2 positive swab specimens of 100 presumed negative specimen sets after centrifugation of the swab specimens could simply reflect lack of reproducibility in specimens with an M. genitalium DNA load near the limit of detection. The lack of reproducibility in specimens with low DNA load was further substantiated by finding the corresponding FVU specimen positive on retesting. Clearly, there is room for improvement in the sample preparation method as recently documented by the finding that 20% of urogenital swab specimens contain <1 genome copy/μL of the pretreated specimen using our standard sample preparation method.19 However, the centrifugation step is time-consuming and increases the risk of sample-to-sample contamination, and therefore, it could be more rewarding to evaluate the effect of collecting the specimen in a smaller volume of transport medium. We chose the described method to have SP4 medium available for culture attempts and indeed, this could have biased the study because only approximately 3 μL of the swab specimen was analyzed as compared with 60 μL of the FVU. Thus, a proportional 20-fold concentration of the FVU occurred. However, because the amount of material collected by swab sampling relative to the flushing effect of urine is unknown, it is impossible to directly compare the 2 specimen types.
Detection of C. trachomatis was found to be influenced by the presence of urethritis in the multiple logistic regression model, in which patients with <5 PMNLs/hpf in the urethral smear were less likely to be positive in both the FVU and urethral swab specimen. In univariate analysis, patients with discrepant specimen sets were older, but this was apparently not an effect of age on the presence of urethritis, because patients with <5 PMNL/hpf had the same age distribution as those with urethritis (P = 0.6).
Inhibition was observed in the M. genitalium PCR assay in only 0.5% of the urethral swab specimen and in 0.6% of the FVUs. We have not systematically collected information regarding the percentage of specimens that were inhibitory in the first test round, but it appears that the M. genitalium assay was more prone to inhibition than the C. trachomatis assay in which only 0.2% of the FVU specimens were inhibitory. This difference could be the result of the different length of the amplicons in the 2 assays because inhibition tends to occur more often in PCRs amplifying longer target sequences.
In most studies evaluating FVU as a possible specimen for detection of C. trachomatis in women, only a cervical swab specimen has been used for comparison. However, at least by culture, it is well documented that some women are positive in the urethral specimen only.25 The use of an expanded reference standard, including different test methods and multiple samples per patient, will tend to decrease the apparent sensitivity of each specimen type. We, therefore, decided to test 3 specimens per woman. For both M. genitalium and C. trachomatis, FVU detected more infections than cervical swabs, and in particular, more than the urethral swab specimens. Because fewer women were included in the study, only 51 M. genitalium PCR-positive specimen sets were available for the evaluation. The FVU specimen could be shown to detect significantly more infections than the swab specimens. However, only 88% of the M. genitalium infections were diagnosed if FVU was used as the only specimen; this figure could be improved to 96% by testing a cervical swab specimen in parallel. For C. trachomatis, FVU detected significantly more infections than the urethral swab. However, the rate of detection could be increased from 90% to 99% by testing also the cervical swab specimen. Whether the cervical swab specimen could be transported in the urine specimen and used for detection of both pathogens, as has been shown to be an efficient and cost-effective approach for diagnosis of C. trachomatis infections,26 remains to be determined.
The present evaluation of the performance of different specimen types cannot be directly generalized to a screening situation in which FVU specimens are collected without previous clinical examination. For men, a urethral smear and a culture for N. gonorrhoeae were collected before the swab specimen for M. genitalium and C. trachomatis detection, which again was collected before the FVU. From a theoretical point of view, this sequence of sampling should have favored the urethral swab specimen, but it cannot be excluded that the urethral sampling for M. genitalium was adversely influenced by removing secretions with the preceding samplings. FVU, on the other hand, could have washed out secretions from the proximal urethra and therefore be affected to a lesser degree. In addition, deeper cell layers could have been exposed for flushing by the flow of urine by the repeated sampling, possibly increasing the detection rate in FVU. A study controlling for these variables, however, was not feasible with the relatively large population needed to document such an effect. In women, the same considerations would apply; 2 urethral swab specimens were collected before the M. genitalium and C. trachomatis swab for the current study. Then finally, the FVU was collected. Furthermore, the effect of the preceding speculum examination is indeed unpredictable, because it could have affected the performance of the FVU specimens in both a positive direction by contaminating the vulvar area with vaginal and/or cervical secretions, and in a negative way, if secretions and lubricants were removed after the examination.
New and very surprising knowledge about the epidemiology of M. genitalium infections was obtained. Although the number of patients studied did not allow for a detailed demographic description, the age-specific prevalence of M. genitalium infection in this population showed striking differences between the distribution of M. genitalium-positive and C. trachomatis-positive men. Whereas the age specific prevalence of C. trachomatis decreased from 22% in patients <20 years of age to 6% in men >40 years of age, the age-specific prevalence of M. genitalium infections remained almost constant between 4% and 8.5% in the different age groups. This age distribution is even more surprising considering the epidemiology of the infection in women in which 9.5% of women <20 years of age were positive as compared with 3.5% of women >35 years of age. Although the prevalence decreased by age, the trend was less marked than that seen for C. trachomatis, and actually, M. genitalium became more common than C. trachomatis in patients >30 years of age. It could be speculated that the decreasing prevalence with age in women reflects the decreasing cervical ectopy, but this hypothesis could not be tested from the current study. It could also be speculated that the relatively higher prevalence in the older age groups was caused by the tendency of M. genitalium to cause persistent or recurring infections. This hypothesis could not be unambiguously supported when the clinical histories of the male patients were compared. Men who reported previous episodes of NGU had the same age distribution as those reporting their first episode of NGU. However, 13% of M. genitalium-infected men had had a previous urethritis as compared with 6% of the C. trachomatis-infected men (dual infections excluded) (P = 0.03). Because no difference in the age distribution of men with and without previous urethritis could be shown, the difference in history cannot completely explain the different epidemiology.
The effect of freezing before and after sample preparation was studied on a subset of the male and female FVU specimens. Among the male specimens, 94% remained positive after storage of the Chelex extracted specimen at -20°C. Of the 6 false-negative specimens, 5 were available as FVU stored without preparation, and all were negative after repeated sample preparation. Because the specimens were only weakly positive at the primary testing, it is considered likely that the M. genitalium DNA was present in such a low amount and that the freeze–thaw cycle was sufficient to render them negative. Because 1 of 30 initially M. genitalium-positive FVU specimens produced discrepant results on repeated duplicate sample extraction, no significant effect of freezing of male FVU specimens could be documented. In contrast, female FVU specimens appeared to be unstable at -20°C because 27% of the specimens subjected to sample preparation after storage were found to be false-negative, whereas those extracted initially after receipt in the laboratory remained positive. The reason for the different stability of male and female specimens is not clear, and the number of female FVUs studied was small, but our findings certainly call for attention when archival FVU specimens are studied. By subjecting M. genitalium PCR-positive FVU specimens to sample preparation in duplicate, it was documented that the findings could not be attributed to lack of reproducibility of the sample preparation. Currently, we are studying the deterioration of M. genitalium-positive specimens using a recently developed quantitative PCR assay.
In conclusion, the current study has shown that using our standard sample preparation method, male FVU is superior to urethral swabs for detection of M. genitalium and C. trachomatis. In women, cervical swabs are needed to maintain a high sensitivity, but the urethral swab specimen, which by many women is considered painful, could be substituted with an FVU specimen. It appears to be essential for female FVUs, however, that they are subjected to a DNA extraction procedure before freezing to avoid loss of sensitivity.
The age-specific prevalence of M. genitalium infection in men was found to be less age-dependent than the C. trachomatis prevalence. This reason for this finding remains unclear but merits further studies.
1. Björnelius E, Lidbrink P, Jensen JS. Mycoplasma genitalium
in non-gonococcal urethritis—A study in Swedish male STD patients. Int J STD AIDS 2000; 11:292–296.
2. Horner PJ, Gilroy CB, Thomas BJ, Naidoo RO, Taylor-Robinson D. Association of Mycoplasma genitalium
with acute non-gonococcal urethritis. Lancet 1993; 342:582–585.
3. Jensen JS, Ørsum R, Dohn B, Uldum S, Worm AM, Lind K. Mycoplasma genitalium
: A cause of male urethritis? Genitourin Med 1993; 69:265–269.
4. Johannisson G, Enström Y, Löwhagen GB, et al. Occurrence and treatment of Mycoplasma genitalium
in patients visiting STD clinics in Sweden. Int J STD AIDS 2000; 11:324–326.
5. Janier M, Lassau F, Casin I, et al. Male urethritis with and without discharge: A clinical and microbiological study. Sex Transm Dis 1995; 22:244–252.
6. Maeda S, Tamaki M, Nakano M, Uno M, Deguchi T, Kawada Y. Detection of Mycoplasma genitalium
in patients with urethritis. J Urol 1998; 159:405–407.
7. Totten PA, Schwartz MA, Sjöström KE, et al. Association of Mycoplasma genitalium
with nongonococcal urethritis in heterosexual men. J Infect Dis 2001; 183:269–276.
8. Jensen JS. Mycoplasma genitalium: The aetiological agent of urethritis and other sexually transmitted diseases. J Eur Acad Dermatol Venereol 2004; 18:1–11.
9. Tully JG, Taylor-Robinson D, Cole RM, Rose DL. A newly discovered mycoplasma in the human urogenital tract. Lancet 1981; 1:1288–1291.
10. Tully JG, Taylor-Robinson D, Rose DL, Cole RM, Bove JM. Mycoplasma genitalium
, a new species from the human urogenital tract. Int J Syst Bacteriol 1983; 33:387–396.
11. Jensen JS, Hansen HT, Lind K. Isolation of Mycoplasma genitalium
strains from the male urethra. J Clin Microbiol 1996; 34:286–291.
12. Jensen JS, Uldum SA, Søndergård-Andersen J, Vuust J, Lind K. Polymerase chain reaction for detection of Mycoplasma genitalium
in clinical samples. J Clin Microbiol 1991; 29:46–50.
13. Palmer HM, Gilroy CB, Furr PM, Taylor-Robinson D. Development and evaluation of the polymerase chain reaction to detect Mycoplasma genitalium
. FEMS Microbiol Lett 1991; 61:199–203.
14. Jensen JS, Borre MB, Dohn B. Detection of Mycoplasma genitalium
by PCR amplification of the 16S rRNA gene. J Clin Microbiol 2003; 41:261–266.
15. Loeffelholz MJ, Lewinski CA, Silver SR, et al. Detection of Chlamydia trachomatis
in endocervical specimens by polymerase chain reaction. J Clin Microbiol 1992; 30:2847–2851.
16. Pollard DR, Tyler SD, Ng C-W, Rozee KR. A polymerase chain reaction (PCR) protocol for the specific detection of Chlamydia
spp. Mol Cell Probes 1989; 3:383–389.
17. Storgaard M, Ostergaard L, Jensen JS, et al. Chlamydia pneumoniae
in children with otitis media. Clin Infect Dis 1997; 25:1090–1093.
18. Prince AM, Andrus L. PCR—How to kill unwanted DNA. Biotechniques 1992; 12:358.
19. Jensen JS, Björnelius E, Dohn B, Lidbrink P. 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 2004; 42:682–692.
20. Sugunendran H, Birley HD, Mallinson H, Abbott M, Tong CY. Comparison of urine, first and second endourethral swabs for PCR based detection of genital Chlamydia trachomatis
infection in male patients. Sex Transm Infect 2001; 77:423–426.
21. Gaydos CA, Quinn TC, Willis D, et al. Performance of the APTIMA Combo 2 assay for detection of Chlamydia trachomatis
and Neisseria gonorrhoeae
in female urine and endocervical swab specimens. J Clin Microbiol 2003; 41:304–309.
22. Van Der PB, Ferraro DV, Buck-Barrington L, et al. Multicenter evaluation of the BDProbeTec ET system for detection of Chlamydia trachomatis
and Neisseria gonorrhoeae
in urine specimens, female endocervical swabs, and male urethral swabs. J Clin Microbiol 2001; 39:1009–1016.
23. Mahony JB, Luinstra KE, Sellors JW, Jang D, Chernesky MA. Confirmatory polymerase chain reaction testing for Chlamydia trachomatis
in 1st-void urine from asymptomatic and symptomatic men. J Clin Microbiol 1992; 30:2241–2245.
24. Mena L, Wang X, Mroczkowski TF, Martin DH. Mycoplasma genitalium
infections in asymptomatic men and men with urethritis attending a sexually transmitted diseases clinic in New Orleans. Clin Infect Dis 2002; 35:1167–1173.
25. Black CM, Marrazzo J, Johnson RE, et al. Head-to-head multicenter comparison of DNA probe and nucleic acid amplification tests for Chlamydia trachomatis
infection in women performed with an improved reference standard. J Clin Microbiol 2002; 40:3757–3763.
26. Airell A, Ottosson L, Bygdeman SM, et al. Chlamydia trachomatis
PCR (Cobas Amplicor) in women: Endocervical specimen transported in a specimen of urine versus endocervical and urethral specimens in 2-SP medium versus urine specimen only. Int J STD AIDS 2000; 11:651–658.