Use of an Adaptation of a Commercially Available P
http://www.100md.com
微生物临床杂志 2006年第2期
ABSTRACT
Trichomonas vaginalis PCR using reagents from a commercially available assay for Chlamydia trachomatis and Neisseria gonorrhoeae was evaluated for detection of infection in women and men attending a sexually transmitted disease clinic. Evaluations included three primer sets, endocervical swabs, vaginal swabs and urine, and various storage conditions. The TVK3/TVK7 primer set was optimal in our hands with sensitivities ranging from 69.5 to 96.8%. In all comparisons, T. vaginalis PCR performed better than routine diagnostics using microscopy for women and culture for men (P > 0.05). The assay performed well for all sample types tested, and vaginal swabs were stable for up to 7 days at ambient temperature. Using samples prepared for, and reagents from, the C. trachomatis-N. gonorrhoeae PCR assay allowed incorporation of T. vaginalis PCR diagnosis into routine clinical testing.
INTRODUCTION
Sexually transmitted infections (STI) are among the most common infectious diseases worldwide. The World Health Organization estimated that over 340 million cases of curable STI occurred in 1999 alone (39). Among these curable STI, Chlamydia trachomatis remains the most common bacterial cause of infection with an estimated 89 million annual cases while Neisseria gonorrhoeae is also common with over 60 million cases estimated to occur annually. However, the most common nonviral STI, worldwide, is caused by Trichomonas vaginalis with over 174 million estimated cases in 1999. Potential sequelae of these STI in females include pelvic inflammatory disease (19, 22), ectopic pregnancy, tubal factor infertility (6), adverse pregnancy outcomes (3, 31), and potentially increased risk of both transmission and acquisition of human immunodeficiency virus (HIV) (1, 2, 14, 15, 25, 30, 31, 35). Additionally, investigators have shown epidemiologic associations with chlamydia (21, 28) or trichomonas (5, 34, 36, 40) infection and subsequent cervical neoplasia and carcinoma. While some of these studies attempted to control for human papillomavirus (HPV) infection, the relationships are unclear and the potential role that trichomonas plays has yet to be fully evaluated. The use of T. vaginalis PCR diagnostics as a research tool in our laboratory has led to the finding that infection with trichomonas is associated with an increased risk for HPV persistence in adolescent women, which may lead to increased risk of cervical cancer (27).
Screening and control programs for chlamydia and gonorrhea have become increasingly feasible due to the commercial availability of nucleic acid amplification tests (NAAT) and consequent acceptability of less invasively collected specimen types (10, 26, 29, 38). Recommendations for STI control programs strongly support screening of men as well as women, especially those who may be asymptomatic and therefore not seeking treatment. Such programs have been established in many areas of the world and have resulted in substantial declines in the prevalence of these infections (8, 11).
Screening for trichomonas infection has been performed using direct microscopy or culture of the organism in most settings that include T. vaginalis as a component of a sexually transmitted disease (STD) control program. However, many programs do not include trichomonas infection screening and indeed, in the United States, this is not a reportable infection tracked by the Centers for Disease Control and Prevention. Recent reports of the impact of trichomonas infections on the risk of HIV transmission and acquisition suggest that more aggressive trichomonas infection screening measures may be warranted. Laga et al. (15) demonstrated a nearly twofold increase in the odds of HIV infection in women infected with T. vaginalis in an African cohort. Although this was not significant at = 0.05, the research group subsequently demonstrated a 1.7-fold increase in relative risk for HIV infection for women infected with T. vaginalis (14), suggesting that trichomonas infection may increase women's susceptibility to HIV. Using mathematical modeling, 40% of the new HIV infections in women in the United States may be attributable to increased risk associated with trichomonas infection (2). Studies have also suggested an increase in the risk of transmission during trichomonas infection as a result of increased viral shedding (9, 35). In both men with urethritis (23) and women with vaginal infections (35), HIV viral shedding was quickly reduced following appropriate treatment for T. vaginalis infection. These data suggest that incorporating trichomonas infection control into chlamydia and gonorrhea control programs may be useful not only for reducing upper reproductive tract infections and adverse outcomes of pregnancy but also as a prevention measure to reduce the spread of HIV (25).
We have adapted a commercially available CT/NG NAAT to allow amplification and detection of trichomonas DNA. Using this assay, we demonstrated the utility of a single sample from women or men for screening for all three organisms and compared the results of the T. vaginalis PCR assay to direct microscopy or trichomonas culture to establish the performance characteristics of this test.
MATERIALS AND METHODS
Population and samples. Specimens were collected from men and women attending the Bell Flower clinic, a county health department STD clinic in an urban Midwestern setting. All subjects provided consent for samples to be used in evaluation of novel diagnostic assays. Although not all specimens were collected for each study, samples were collected in the following order: female urine, vaginal swab, cervical or urethral swab for gonorrhea culture, cervical or urethral swab for chlamydia culture, and male urine. Vaginal swabs were placed into sterile, dry tubes or into tubes containing T. vaginalis culture medium. Swabs for chlamydia culture were collected in chlamydia transport medium (CTM). First-catch urine with volumes ranging from 15 to 90 ml was collected. For experiments involving archived samples, data were available from original clinical wet preparations, but no vaginal swabs were retained. For studies of freshly obtained samples, female urine was not obtained. For the comparison of TVA and TVK primer sets, female vaginal swabs were used for both microscopy and PCR. For comparison of storage conditions of vaginal swabs, cervical swabs in CTM were also used to define infection.
ATCC T. vaginalis isolates 30184, 30185, and 30188 were used as positive T. vaginalis controls in titration experiments. C. trachomatis strain H-UW4 and a clinical N. gonorrhoeae isolate, 9003-0025, were used in the competition assays.
T. vaginalis wet mount and culture. Women attending the STD clinic were evaluated for the presence of T. vaginalis by direct microscopy of a wet preparation. Motile, flagellated trichomonads were identified by clinicians following the routine clinical procedure. Men were not routinely screened for trichomonas infection as a part of clinical exams. Culture was performed in the laboratory using Diamond's medium for vaginal swabs and the In-Pouch culture system for male urine sediment. Urine was held at room temperature prior to transport to the laboratory to maintain the viability of T. vaginalis for culture. Ten milliliters of urine was centrifuged at 900 x g for 10 min, the supernatant was discarded, and the sediment was inoculated into the In-Pouch system. Cultures were read using a wet preparation slide on days 0, 1, and 3. Laboratory strains for determination of analytical sensitivity were grown in Diamond's medium and harvested by centrifugation at 900 x g for 10 min.
CT/NG PCR. C. trachomatis and N. gonorrhoeae testing was performed using the Amplicor C. trachomatis/N. gonorrhoeae assay (Roche Diagnostic Corporation, Indianapolis, IN). All testing was performed according to the manufacturer's package insert unless otherwise noted and has been described previously (32). Briefly, specimens were lysed and processed for amplification following protocols specific for urine or swab samples in CTM. Vaginal swabs were transported in dry, sterile tubes. One milliliter of molecular-reagent-grade water was added to the swab followed by vigorous vortexing for at least 10 seconds to elute the sample. Swabs were discarded, and from this point forward, samples were processed as if collected in CTM. Processed samples were added to Amplicor Master Mix (MM) reagent containing polymerase, deoxynucleoside triphosphates, an internal control sequence to verify amplification, and uracil-N-glycosylase to reduce potential carryover contamination. This MM was not modified and contained only CT/NG primers. Samples were amplified in a Perkin-Elmer 9600 Thermocycler, and amplified product was detected using oligonucleotide probes specific to an internal sequence of the target DNA. The probes were prebound to polystyrene wells, and the biotinylated primers of amplified sequence were detected using a standard 96-well enzyme-linked immunosorbent assay format (Fig. 1). The optical density was measured at 450 nm, and results were interpreted according to the package insert.
T. vaginalis PCR. Amplification of T. vaginalis DNA was performed using biotinylated primers specific for this organism added to the MM reagent. Three sets of primers, TVA (24), TVK (13), and ?-Tub (17), and their corresponding oligonucleotide probes were evaluated (Table 1) using the published cycling parameters. Although the MM contained CT/NG primers, the T. vaginalis primers required different amplification parameters (for TVK PCR: holding at 95°C for 5 min, followed by 35 cycles of 90°C for 1 min, 60°C for 30 s, and 70°C for 2 min, followed by holding at 70°C for 10 min). T. vaginalis DNA amplification was always performed independently of C. trachomatis/N. gonorrhoeae amplification (Fig. 1). Amplicor detection kits, in a C. trachomatis- or N. gonorrhoeae-specific 96-well format, are sold separately from the specimen processing and amplification reagents. In order to detect amplicon from 96 amplified samples, one C. trachomatis kit and one N. gonorrhoeae detection kit are used. The kits contain identical denaturation, hybridization, avidin-horseradish peroxidase conjugate, substrate, and stop reagents. The only difference between the kits is the 96-well probe-bound plate specific for the C. trachomatis or N. gonorrhoeae amplicons. The samples amplified for C. trachomatis/N. gonorrhoeae are denatured using the reagent from only one of the two kits, leaving the other bottle of reagent available for denaturation of T. vaginalis amplification tubes. Each 96-well kit contains sufficient reagent for 144 wells with the least volume in the conjugate reagent. Therefore, using one C. trachomatis and one N. gonorrhoeae detection kit, 96 samples could be tested for all three pathogens by pooling reagents (Fig. 1). Detection of amplified T. vaginalis DNA was performed using the reagents from Amplicor CT/NG kits with the following exceptions. Primer set-specific probes were designed using Oligo (Molecular Biology Insights, Cascade, CO) software and were bound without a capture tag to polystyrene wells (Immulon II; ThermoLabsystems, Franklin, MA) by incubation overnight in freshly prepared 1 M ammonium acetate. Excess probe was removed by rinsing with phosphate-buffered saline with 1 mM EDTA. Plates were blotted, allowed to air dry for 3 h, sealed, and stored at 4°C until use. This method was developed from the probe binding method used for genital ulcer disease multiplex PCR (20). The hybridization buffer included in the CT/NG kit was too stringent to allow binding of the T. vaginalis amplified product; therefore, we used a generic PCR-digoxigenin enzyme-linked immunosorbent assay hybridization buffer (catalogue no. 1717472; Roche Applied Science, Indianapolis, IN) instead of the buffer provided with the CT/NG detection kit for detection of T. vaginalis amplicons. In response to this change, the denaturation reagent from the CT/NG kit was reduced in volume from 100 to 80 μl/well for denaturation of the T. vaginalis amplicons. From this point forward, all reagents were from the CT/NG kit and were used exactly as described in the package insert.
Samples with an A450 below 0.300 were interpreted as negative, those above 0.800 were interpreted as positive, and those falling between these values were considered equivocal. Samples with equivocal results were retested in duplicate with at least two/three results over 0.500 required to report the test result as positive. Cutoff points were initially based on recommendations by Roche for the CT/NG assay and the multiplex genital ulcer disease PCR.
Analysis. Patients were classified as infected if the routine diagnostic method (wet preparation or culture) was positive or if two separate PCR results were positive. When multiple samples were available from the same patient, in order to be classified as infected if the routine test was negative, positive PCR results were required from both samples. Similarly, for a comparison of primer sets using the same processed sample, to qualify as an infection when the routine method was negative, both PCR primer results had to be positive. When the routine test was negative and only one PCR result was positive, the PCR was considered a false-positive result. Comparison of multiple assay results was performed using the Cochran Q statistic that tests the hypothesis that there is no difference in dichotomous results between multiple measures on the same individual. McNemar's 2 test was used for pairwise comparisons. Pairwise comparisons were performed for the overall results and separately for infected and uninfected participants to evaluate differences in sensitivity and specificity, respectively. Results were considered significant at the = 0.05 level.
RESULTS
Evaluation of T. vaginalis PCR using laboratory strains. Laboratory strains of C. trachomatis, N. gonorrhoeae, and T. vaginalis were loaded onto Dacron swabs to assess both the limit of detection of the T. vaginalis PCR assay and the effect of competition for resources in the event of coinfections. To evaluate the sensitivity of the assay, trichomonads, chilled to 4°C to reduce motility, were counted in a wet preparation using a hemacytometer. Since patient samples were eluted in water, all serial dilutions were performed in water to achieve the desired number of organisms. Concentrations tested were 0.54 x 101 to 6.0 x 105 organisms per reaction tube, 25% of which went into the amplification tube. Testing was performed with both the TVA and TVK primer sets. Amplification using the TVA primers was detectable with as few as five trichomonads, and the TVK primer set was positive below the theoretical concentration of one trichomonad per processed sample. These results may overestimate the analytical sensitivity of the T. vaginalis PCR since the calculations were based on counted trichomonads and do not account for any lysed organisms from which DNA may have been present in the preparations. The experiment was performed in duplicate with identical results.
The addition of T. vaginalis primers to the MM raised the concern that the presence of the C. trachomatis and N. gonorrhoeae primers, which are included in the commercially available amplification reagent, might result in a reduction of T. vaginalis amplification in the presence of C. trachomatis and N. gonorrhoeae coinfections. To evaluate this potential effect, swabs were loaded with approximately 35 trichomonads and chlamydial elementary bodies and gonococci were added in concentrations ranging from 0 to 900 inclusion-forming units and 0 to 960 CFU, respectively. Swabs were eluted into 1 ml of water and processed as described for vaginal swabs. After processing, a total of 1.25% of the original sample was loaded into the amplification tube. Samples prepared in this manner were then evaluated using the TVA and TVK primer sets. Both TVA and TVK primers amplified T. vaginalis DNA at all concentrations of C. trachomatis and N. gonorrhoeae. Since the cycling parameters for T. vaginalis amplification were different from those used for C. trachomatis and N. gonorrhoeae amplification, no assays were performed to assess amplification of these organisms' DNA under T. vaginalis cycling parameters.
Comparison of TVA and TVK primer sets using vaginal samples. Vaginal swab samples from 174 women were evaluated by T. vaginalis PCR using the same processed sample tested by both the TVA and TVK primer sets. The T. vaginalis PCR results were compared to clinically obtained wet preparation microscopy results. T. vaginalis infection was defined by a positive wet mount or by vaginal swab PCR results positive by both primer sets. Thirty-nine (22.4%) women were classified as infected with T. vaginalis. Of these, 18 were positive by both TVA and TVK, only three of which were positive by wet mount. One infection was positive by TVA and not TVK primers, while 13 infections were detected by TVK and not TVA primers. Seven were positive by wet mount alone. The sensitivity of wet mount using this definition of infection was 61.5%, compared to the performance of the TVA and TVK primer sets with sensitivities of 48.7% and 79.5%, respectively (Table 2). Analysis of the agreement of the results, taking multiple comparisons into account, indicated that at least one of the tests performed differently from the others (P = 0.003). In pairwise analysis, the sensitivity of the TVK assay was higher than that of the TVA primer set (P = 0.002). There was no difference between the specificities of the two primer sets (P > 0.05).
Comparison of TVK and ?-Tub primer sets using male urine. First-catch urine specimens were collected from 503 men. Urine samples were stored at ambient temperature (AT) prior to transportation to the laboratory to facilitate T. vaginalis culture and arrived in the laboratory within 16 h of collection. Immediately following arrival, T. vaginalis cultures were prepared from urine sediment and prior to processing for PCR specimens were subsequently stored at 4°C for up to 4 days after collection. All samples were tested for T. vaginalis using the TVK and ?-Tub primer sets, and the TVA primer set was used for discrepant analysis. Infection was defined by a positive culture or two PCR-positive results with independent primer sets. Twenty-five (5.0%) men were classified as infected with T. vaginalis. For comparison, in this population 27 (5.4%) were positive for C. trachomatis and 26 (5.2%) were infected with N. gonorrhoeae. The TVK primer set identified 24 (96.0%) of the infected men, and the ?-Tub primer set identified 23 (92.0%) while T. vaginalis culture identified only 14 (56.0%) (Table 3). The sensitivities of the three assays were statistically different (P = 0.001). In pairwise comparisons, the sensitivity of both PCR assays was statistically different from that of culture (both P values < 0.01). The sensitivity and specificity of the two PCR assays were equivalent (both P values > 0.05). Based on the findings from both male and female samples, all further studies were performed using the TVK primer/probe set as the assay of interest.
Comparison of female specimen types and storage using the TVK primer set. C. trachomatis and N. gonorrhoeae diagnostic tests are often performed on endocervical or urine samples. Therefore, these specimen types were evaluated for use with T. vaginalis PCR. Infection was defined by a positive routine diagnostic test or by positive PCR results from at least two samples (e.g., vaginal and endocervical). In these evaluations, the requirement for duplicate PCR-positive results was intended to control for laboratory error rather than confirm specificity since that had been evaluated using alternate primer sets in the previous experiments.
Samples and wet mount data were collected from 186 women, 41 (22.0%) of whom were classified as infected with T. vaginalis (Table 4). Infection was defined by a positive wet mount, culture, or both endocervical and urine PCR results being positive. In this population, C. trachomatis was identified in 28 (15.1%) and N. gonorrhoeae was found in 17 (9.1%) of the participants. There was no difference in the prevalence of C. trachomatis or N. gonorrhoeae as a function of the presence of T. vaginalis (all P > 0.05; data not shown). Wet mount identified 21 (51.2%) T. vaginalis infections while culture detected 25 (61.0%). T. vaginalis PCR using vaginal and endocervical swabs identified 37 (90.2%) and 36 (87.8%) of the infected women, respectively. At least one test was significantly different from the others (P < 0.001) with pairwise comparisons indicating that T. vaginalis PCR using vaginal or endocervical samples identified more positive samples than did culture (P = 0.004 and 0.007, respectively) and wet mount (all P < 0.001). There was no difference in sensitivity or specificity between vaginal and cervical samples for PCR (all P > 0.05).
Residual, archived samples stored at –70°C for approximately 5 years were available from 463 women for comparison of endocervical swabs and urine specimens (Table 4). Infection was defined by a positive wet mount recorded in the clinical chart for that visit or by both endocervical and urine PCR-positive results. Ninety-four (20.3%) women were infected. Endocervical PCR, urine PCR, and wet mount identified 96.8, 94.7, and 51.1% of infected women, respectively. There was a significant difference between wet mount and endocervical or urine PCR (P < 0.001). Sensitivity and specificity of the assay were similar for both endocervical and urine specimens (P > 0.05).
Stability of vaginal swabs. To assess the stability of vaginal swab samples, endocervical swabs in CTM and duplicate vaginal samples were collected. Storage at 4°C was compared to holding samples at AT. Additionally, storage of samples for 3 days was compared to storage for 7 days. Trichomonas infection was defined by a positive wet preparation performed at the clinic or by two separate specimens (e.g., a cervical and a vaginal sample or both vaginal swabs) being positive by T. vaginalis PCR. Duplicate vaginal samples were available from 288 women for the evaluation of storage temperature. T. vaginalis infection was identified in 67 (23.3%) (Table 4). Storage temperature did not affect the sensitivity or specificity of the assay (all P > 0.05). Duplicate vaginal samples were obtained from 217 women for comparison of storage for 3 or 7 days postcollection (Table 4). Of the 46 T. vaginalis cases identified in this cohort, 32 (69.5%) and 35 (76.1%) were positive following specimen storage for 3 or 7 days, respectively. The sensitivity and specificity of the assay were similar for the two storage conditions (all P > 0.05).
DISCUSSION
The ability to integrate this test into routine screening for C. trachomatis and N. gonorrhoeae using an adaptation of commercially available reagents makes molecular biology-based diagnosis of T. vaginalis feasible for many diagnostic laboratories. The use of samples collected by minimally invasive or noninvasive methods that have been stored for several days will allow population-based screening for T. vaginalis. In these evaluation studies, the sensitivity and specificity of the T. vaginalis PCR using the TVK primer set were high. The assay using the TVK primer set and probe sequence performed better than the TVA primer/probe set and equivalently to the ?-Tub assay. These findings were consistent with the primer evaluation performed by Crucitti et al., who found the TVK primer set to be optimal for detection of T. vaginalis DNA (4). We chose to use the TVK assay for logistic reasons such as a shorter amplification process while using the ?-Tub assay for confirmatory purposes. The cutoff points used were arbitrarily chosen based on the manufacturer's recommendation. A bimodal distribution of optical density values similar to that described by Kaydos et al. (12) was seen with the results of the TVK assay. Evaluation of over 7,200 samples showed only 1.2% in the equivocal range (data not shown). With repeat testing of equivocals, the proportion of those classified as positive increased from 2.4% to 71.4% as the value of the initial result increased from 0.300 to 0.800, suggesting that the equivocal range is appropriate. It should be noted that, in the absence of a means for determining the true rate of infection with T. vaginalis, the performance characteristics described here are only estimates based on the best available techniques. In some analyses (the evaluations of sample types and storage conditions) we used the method under evaluation, requiring two separate samples to be positive, as a means of defining infection with the understanding that this introduces bias by inflating the apparent sensitivity and specificity of the assay. However, in these comparisons, sensitivity and specificity were similar to the primer set comparisons. In these studies the question of interest was comparability of performance, and this was demonstrated.
The objective of this work was to develop a T. vaginalis diagnostic tool that offers improvement over currently used methods such as microscopy and culture. The choice of wet preparation microscopy for women was based on the clinical standard in place during these evaluations. The clinic was not routinely culturing for trichomonas, and men were not being tested at all. Male urine, rather than perhaps more sensitive samples such as urine and urethral swabs combined, was selected as the least invasive addition to the clinical procedures in place. Since cultures were not routinely performed, the In-Pouch package insert recommendation for length of culture was followed. Therefore, the methods for identification of trichomonas infection were either equivalent or superior to the routine clinical practices at the participating STD clinic. Although alternate sample types or variations on comparator testing might have improved the estimates of sensitivity and specificity of the TVK assay, such measures will rarely be available in routine STD control programs and were of less interest than comparison to basic diagnostic procedures.
It was common for T. vaginalis to be defined by a positive wet preparation while both PCR samples gave negative results. For example, 17.9% (7/39) of cases shown in Table 3 had positive wet mount and negative PCR results. Women were classified as infected according to the a priori definition, but this may not have been an accurate definition. Wet preparations were read at the clinic site, and the quality of the results was not subject to further analysis, quality review, or proficiency testing. It is possible that wet preparations may have been called positive based on the presence of a species of motile trichomonad that does not contain TVK primer binding sites. If the infection status was more loosely defined as requiring at least two positive results, the number of infections dropped to 31 (17.8%) and the sensitivity of the tests was 54.8, 61.3, and 96.8% for wet mount, TVA PCR, and TVK PCR, respectively. Therefore, the sensitivity of the assay may be underestimated for all comparisons. This approach is justified since we did not attempt to measure inhibition of the PCRs, which might also explain the occurrence of wet-mount-positive/PCR-negative samples. However, even using this potentially conservative approach, the T. vaginalis PCR provided improved case finding for both men and women, as has been reported by other investigators (4, 16, 17, 26, 33, 37, 38).
The assay appears to be useful for a variety of sample types in women: endocervical swabs, urine, or vaginal swabs. This was in contrast to the poor performance of female urine as a sample type for T. vaginalis PCR described by Lawing et al. (16). We used whole-urine aliquots that had been frozen whereas the Lawing study used urine sediment frozen in water. The difference in storage conditions may explain the difference in urine sensitivity between our results and the previous study (95% and 64%, respectively). Additionally, in studies using fresh urine samples (12, 18, 33) urine performed well as a sample type. Finally, the prevalence of T. vaginalis for each comparison cohort was similar, 20.3 to 23.3%, suggesting that the sample types are truly comparable and that the storage conditions are amenable to sending samples to a reference laboratory. Encouragingly, the prevalence from the repository of cervical and urine samples stored at –70°C for up to 5 years was also within this range, indicating that long-term storage of either of these sample types likely has little effect on the quality of the sample for T. vaginalis PCR.
As we demonstrated with mock samples using laboratory strains, in a cohort of women with a high prevalence of C. trachomatis (15.1%) and N. gonorrhoeae (9.1%), the T. vaginalis PCR performed well in the presence of these other STI. This was of particular concern since the Amplicor MM to which T. vaginalis primers were added contained C. trachomatis and N. gonorrhoeae primer pairs as well. Consistent with global data, the prevalence of T. vaginalis in these studies in women was substantially higher than that of C. trachomatis or N. gonorrhoeae. Not surprisingly, the number of cases identified in men was lower than that in women (5.0%), but in this population, the rates of C. trachomatis (5.4%) and N. gonorrhoeae (5.2%) were also lower. The low STD prevalence rates may be attributable to the handling of the urine in this study. Urine was held at AT until delivery to the laboratory in order to avoid cold shock to any live trichomonads. Approximately two-thirds of urine samples were transported to the lab within 4 h of collection while the remainder were held overnight at the clinic. This delay of up to 16 h between collection and culture with subsequent refrigeration of the urine was possibly responsible for the degradation of both live organisms for culture and DNA for PCR of T. vaginalis. However, there was no difference in the frequency of positive results based on overnight storage (data not shown; P > 0.05).
Although other methods for amplification detection, such as use of a digoxigenin reporter molecule, may improve the sensitivity of detection of amplified product and be of use in research settings, the assay described here provides an improvement to current routine practice that can be used with very minor modifications to an available diagnostic kit. Kaydos and colleagues (12) saw variability with an assay similar to the one that we have described. This variability prompted them to develop an alternative assay which performed more reliably in their evaluations. Our laboratory has been using the assay described here for more than 6 years with a high degree of probe plate lot-to-lot reproducibility. This may be due in part to the use of the newer version of the Amplicor CT/NG reagents and the substitution of a different hybridization buffer rather than the C. trachomatis stand-alone kit used in the study by Kaydos et al. In 66 evaluations of reproducibility performed in our lab, there have only been two (3%) occasions where the reagents did not perform satisfactorily (data not shown). In these two instances the probe-bound plates did not provide adequate binding and the correlation between previous and new lots was less than 0.80. This indicates that the assay reproducibility should be closely monitored, as variations in the plate lot, prior to probe binding, and adherence to the binding procedure can affect assay performance. On all other occasions, the correlations were high, with only three instances of lot testing with correlations below 0.94. This suggests that the assay is stable and reproducible over time.
Several investigators have reported excellent results of screening for C. trachomatis and N. gonorrhoeae using either patient-obtained or clinician-collected vaginal swabs (10, 16, 29, 38). Many of these studies have also used a NAAT for the detection of T. vaginalis to enhance the screening program. Indeed, more recent evaluations of LightCycler assays for detection of T. vaginalis suggest improved sensitivity (7). However, the studies published to date rely on DNA isolation procedures, in-house PCR amplification reagents, cumbersome detection methods, or specialized equipment that is restricted to research laboratories. We have described a system which uses commercially available reagents, with the exception of the primers and probe oligonucleotides, which can easily be added to the routine for detection of C. trachomatis and N. gonorrhoeae in a clinical laboratory. The CT/NG assay can be performed without deviation from the package insert since the T. vaginalis primers are not added during the amplification of the C. trachomatis/N. gonorrhoeae assay. The separate amplification and detection of T. vaginalis does not have an effect on the performance of the C. trachomatis and N. gonorrhoeae testing, making this a suitable assay for laboratories adhering to Clinical Laboratory Improvement Act regulations. The costs associated with this test are minimal since the gross excess reagent provided in the Amplicor CT/NG detection kits can be pooled to allow T. vaginalis testing. The major expense is the amplification reagent to which T. vaginalis primers are added.
The value of this type of assay for screening programs is improved case ascertainment compared to traditional methods such as microscopy and culture (37). Given the increasing evidence that T. vaginalis may play a significant role in the transmission of HIV, both in pandemic regions such as sub-Saharan Africa (1, 14, 15) and in the United States (2, 24), sensitive assays for detection of T. vaginalis in a clinical laboratory setting are clearly needed. The assay described here is designed for this type of application. The use of commercially available reagents and samples processed for C. trachomatis and N. gonorrhoeae testing will allow clinical laboratories to adopt this method and utilize the quality practices of the manufacturers to optimize the quality and reproducibility of the diagnostic results.
ACKNOWLEDGMENTS
We thank Diane R. Stothard for detection probe design and Nancy J. Smith for technical assistance.
REFERENCES
Buvé, A., H. A. Weiss, M. Laga, E. Van Dyck, R. Musonda, I. Zekeng, M. Kahindo, S. Anagonou, L. Morison, N. J. Robinson, and R. J. Hayes for the Study Group on Heterogeneity of HIV Epidemics in African Cities. 2001. The epidemiology of trichomoniasis in women in four African cities. AIDS 15:S89-S96.
Chesson, H. W., J. M. Blandford, and S. D. Pinkerton. 2004. Estimates of the annual number and cost of new HIV infections among women attributable to trichomoniasis in the United States. Sex. Transm. Dis. 31:547-551.
Cotch, M. F., J. G. Pastorek, R. P. Nugent, S. L. Hillier, R. S. Gibbs, D. H. Martin, D. A. Eschenbach, R. Edelman, J. C. Carey, J. A. Regan, M. A. Krohn, M. A. Klebanoff, A. V. Rao, G. G. Rhoads, et al. 1997.
Trichomonas vaginalis associated with low birth weight and preterm delivery. Sex. Transm. Dis. 24:353-360.
Crucitti, T., E. Van Dyck, A. Tehe, S. Abdellati, B. Vuylsteke, A. Buve, and M. Laga. 2003. Comparison of culture and different PCR assays for detection of Trichomonas vaginalis in self collected vaginal swab specimens. Sex. Transm. Infect. 79:393-398.
Gram, I. T., M. Macaluso, J. Churchill, and H. Stalsberg. 1992. Trichomonas vaginalis (TV) and human papillomavirus (HPV) infection and the incidence of cervical intraepithelial neoplasia (CIN) grade III. Cancer Causes Control 3:231-236.
Grodstein, F., M. B. Goldman, and D. W. Cramer. 1993. Relation of tubal infertility to history of sexually transmitted diseases. Am. J. Epidemiol. 137:577-584.
Hardick, J., S. Yang, S. Lin, D. Duncan, and C. Gaydos. 2003. Use of the Roche LightCycler instrument in a real-time PCR for Trichomonas vaginalis in urine samples from females and males. J. Clin. Microbiol. 41:5619-5622.
Herrmann, B., and M. Egger. 1995. Genital Chlamydia trachomatis infections in Uppsala County, Sweden, 1985-1993: declining rates for how much longer? Sex. Transm. Dis. 22:253-260.
Hobbs, M. M., P. Kazembe, A. W. Reed, W. C. Miller, E. Nkata, D. Zimba, C. C. Daly, H. Chakraborty, M. S. Cohen, and I. Hoffman. 1999. Trichomonas vaginalis as a cause of urethritis in Malawian men. Sex. Transm. Dis. 26:381-387.
Hsieh, Y. H., M. R. Howell, J. C. Gaydos, K. T. McKee, Jr., T. C. Quinn, and C. A. Gaydos. 2003. Preference among female Army recruits for use of self-administrated vaginal swabs or urine to screen for Chlamydia trachomatis genital infections. Sex. Transm. Dis. 30:769-773.
Katz, B. P., M. J. Blythe, B. Van Der Pol, and R. B. Jones. 1996. Declining prevalence of chlamydial infection among adolescent girls. Sex. Transm. Dis. 23:226-229.
Kaydos, S. C., H. Swygard, S. L. Wise, A. C. Sena, P. A. Leone, W. C. Miller, M. S. Cohen, and M. M. Hobbs. 2002. Development and validation of a PCR-based enzyme-linked immunosorbent assay with urine for use in clinical research settings to detect Trichomonas vaginalis in women. J. Clin. Microbiol. 40:89-95.
Kengne, P., F. Veas, N. Vidal, J. L. Rey, and G. Cuny. 1994. Trichomonas vaginalis: repeated DNA target for highly sensitive and specific polymerase chain reaction diagnosis. Cell. Mol. Biol. 40:819-831.
Laga, M., M. Alary, N. Nazila, A. T. Manoka, M. Tuzila, F. Behets, J. Goeman, M. St. Louis, and P. Piot. 1994. Condom promotion, sexually transmitted diseases treatment, and declining incidence of HIV-1 infection in female Zairian sex workers. Lancet 344:246-248.
Laga, M., A. Manoka, M. Kivuvu, B. Malele, M. Tuliza, N. Nzila, J. Goeman, F. Behets, V. Batter, and M. Alary. 1993. Non-ulcerative sexually transmitted diseases as risk factors for HIV-1 transmission in women: results from a cohort study. AIDS 7:95-102.
Lawing, L. F., S. R. Hedges, and J. R. Schwebke. 2000. Detection of trichomonosis in vaginal and urine specimens from women by culture and PCR. J. Clin. Microbiol. 38:3585-3588.
Madico, G., T. C. Quinn, A. Rompalo, K. T. McKee, Jr., and C. A. Gaydos. 1998. Diagnosis of Trichomonas vaginalis infection by PCR using vaginal swab samples. J. Clin. Microbiol. 36:3205-3210.
Mayta, H., R. H. Gilman, M. M. Calderon, A. Gottlieb, G. Soto, I. Tuero, S. Sanchez, and A. Vivar. 2000. 18S ribosomal DNA-based PCR for diagnosis of Trichomonas vaginalis. J. Clin. Microbiol. 38:2683-2687.
Moodley, P., D. Wilkinson, C. Connolly, J. Moodley, and A. W. Sturm. 2002. Trichomonas vaginalis is associated with pelvic inflammatory disease in women infected with human immunodeficiency virus. Clin. Infect. Dis. 34:519-522.
Orle, K. A., and J. B. Weiss. 1997. Detection of Treponema pallidum, Haemophilus ducreyi, and herpes simplex virus by multiplex PCR, p. 67-79. In R. W. Peeling and P. F. Sparling (ed.), Sexually transmitted diseases: methods and protocols. Humana Press, Totowa, New Jersey.
Paavonen, J., K. P. Karunakaran, Y. Noguchi, T. Anttila, A. Bloigu, J. Dillner, G. Hallmans, T. Hakulinen, E. Jellum, P. Koskela, M. Lehtinen, S. Thoresen, H. Lam, C. Shen, and R. C. Brunham. 2003. Serum antibody response to the heat shock protein 60 of Chlamydia trachomatis in women with developing cervical cancer. Am. J. Obstet. Gynecol. 189:1287-1292.
Paisarntantiwong, R., S. Brockman, L. Clarke, S. Landesman, J. Feldman, and H. Minkoff. 1995. The relationship pf vaginal trichomoniasis and pelvic inflammatory disease among women colonized with Chlamydia trachomatis. Sex. Transm. Dis. 22:344-347.
Price, M. A., D. Zimba, I. F. Hoffman, S. C. Kaydos-Daniels, W. C. Miller, F. Martinson, D. Chilongozi, E. Kip, E. Msowoya, M. M. Hobbs, P. N. Kazwmbe, and M. S. Cohen. 2003. Additional of treatment for trichomoniasis to syndromic management of urethritis in Malawi: a randomized clinical trial. Sex. Transm. Dis. 30:516-522.
Riley, D. E., M. C. Roberts, T. Takayama, and J. N. Krieger. 1992. Development of a polymerase chain reaction-based diagnosis of Trichomonas vaginalis. J. Clin. Microbiol. 30:465-472.
Rothenberg, R. B., J. N. Wasserheit, M. E. St. Louis, J. M. Douglas, et al. 2000. The effect of treating sexually transmitted diseases on the transmission of HIV in dually infected persons: a clinic-based estimate. Sex. Transm. Dis. 27:411-416.
Schwebke, J. R., and E. W. Hook III. 2003. High rates of Trichomonas vaginalis among men attending a sexually transmitted diseases clinic: implications for screening and urethritis management. J. Infect. Dis. 188:465-468.
Shew, M. L., J. D. Fortenberry, W. Tu, B. E. Juliar, B. E. Batteiger, B. Qadadri, and D. R. Brown. Association of condom use, sexual behaviors and sexually transmitted infections on genital human papillomavirus infection among adolescent women. Arch. Pediatr. Adolesc. Med., in press.
Smith, J. S., C. Bosetti, N. Munoz, R. Herrero, F. X. Bosch, J. Eluf-Neto, C. J. Meijer, A. J. Van Den Brule, S. Franceschi, R. W. Peeling, and IARC Multicentric Case-Control Study. 2004. Chlamydia trachomatis and invasive cervical cancer: a pooled analysis of the IARC multicentric case-control study. Int. J. Cancer 111:431-439.
Smith, K., K. Harrington, G. Wingood, M. K. Oh, E. W. Hook III, and R. J. DiClemente. 2001. Self-obtained vaginal swabs for diagnosis of treatable sexually transmitted diseases in adolescent girls. Arch. Pediatr. Adolesc. Med. 155:676-679.
Sorvillo, F., L. Smith, P. Kerndt, and L. Ash. 2001. Trichomonas vaginalis, HIV, and African-Americans. Emerg. Infect. Dis. 7:927-932.
Sutton, M. Y., M. Sternberg, M. Nsuami, F. Behets, A. M. Nelson, and M. St. Louis. 1999. Trichomoniasis in pregnant human immunodeficiency virus-infected and human immunodeficiency virus-uninfected Congolese women: prevalence, risk factors, and association with low birth weight. Am. J. Obstet. Gynecol. 181:656-662.
Van Der Pol, B., T. C. Quinn, C. A. Gaydos, K. Crotchfelt, J. Schachter, J. Moncada, D. Jungkind, D. H. Martin, B. Turner, C. Peyton, and R. B. Jones. 2000. Multicenter evaluation of the AMPLICOR and automated COBAS AMPLICOR CT/NG tests for detection of Chlamydia trachomatis. J. Clin. Microbiol. 38:1105-1112.
van der Schee, C., A. van Belkum, L. Zwijgers, E. van der Brugge, E. L. O'Neill, A. Luijendijk, T. Rijsoort-Vos, W. I. van der Meijden, H. Verbrugh, and H. J. Sluiters. 1999. Improved diagnosis of Trichomonas vaginalis infection by PCR using vaginal swabs and urine specimens compared to diagnosis by wet mount microscopy, culture, and fluorescent staining. J. Clin. Microbiol. 37:4127-4130.
Viikki, M., E. Pukkala, P. Nieminen, and M. Hakama. 2000. Gynaecological infections as risk determinants of subsequent cervical neoplasia. Acta Oncol. 39:71-75.
Wang, C. C., R. S. McClelland, M. Reilly, J. Overbaugh, S. R. Emery, K. Mandaliya, B. Chohan, J. Ndinya-Achola, J. Bwayo, and J. K. Kreiss. 2001. The effect of treatment of vaginal infections on shedding of human immunodeficiency virus type 1. J. Infect. Dis. 183:1017-1022.
Watts, D. H., M. Fazzari, H. Minkoff, S. L. Hillier, B. Sha, M. Glesby, A. M. Levine, R. Burk, J. M. Palefsky, M. Moxley, L. Ahdieh-Grant, and H. D. Strickler. 2005. Effects of bacterial vaginosis and other genital infections on the natural history of human papilloma virus infection in HIV-1-infected and high-risk HIV-1-uninfected women. J. Infect. Dis. 191:1129-1139.
Wendel, K. A., E. J. Erbelding, C. A. Gaydos, and A. M. Rompalo. 2002. Trichomonas vaginalis polymerase chain reaction compared with standard diagnostic and therapeutic protocols for detection and treatment of vaginal trichomoniasis. Clin. Infect. Dis. 35:576-580.
Witkin, S. S., S. R. Inglis, and M. Polaneczky. 1996. Detection of Chlamydia trachomatis and Trichomonas vaginalis by polymerase chain reaction in introital specimens from pregnant women. Am. J. Obstet. Gynecol. 175:165-167.
World Health Organization. 2001. Global prevalence and incidence of selected curable sexually transmitted infections: overview and estimates. World Health Organization, Geneva, Switzerland.
Zhang, Z. F., S. Graham, S. Z. Yu, J. Marshall, M. Zielezny, Y. X. Chen, M. Sun, S. L. Tang, C. S. Liao, and J. L. Xu. 1995. Trichomonas vaginalis and cervical cancer. A prospective study in China. Ann. Epidemiol. 5:325-332.
Division of Infectious Diseases, Indiana University School of Medicine, Indianapolis, Indiana,
1 Department of Internal Medicine, Emory University School of Medicine, Atlanta, Georgia(Barbara Van Der Pol, Coll)
Trichomonas vaginalis PCR using reagents from a commercially available assay for Chlamydia trachomatis and Neisseria gonorrhoeae was evaluated for detection of infection in women and men attending a sexually transmitted disease clinic. Evaluations included three primer sets, endocervical swabs, vaginal swabs and urine, and various storage conditions. The TVK3/TVK7 primer set was optimal in our hands with sensitivities ranging from 69.5 to 96.8%. In all comparisons, T. vaginalis PCR performed better than routine diagnostics using microscopy for women and culture for men (P > 0.05). The assay performed well for all sample types tested, and vaginal swabs were stable for up to 7 days at ambient temperature. Using samples prepared for, and reagents from, the C. trachomatis-N. gonorrhoeae PCR assay allowed incorporation of T. vaginalis PCR diagnosis into routine clinical testing.
INTRODUCTION
Sexually transmitted infections (STI) are among the most common infectious diseases worldwide. The World Health Organization estimated that over 340 million cases of curable STI occurred in 1999 alone (39). Among these curable STI, Chlamydia trachomatis remains the most common bacterial cause of infection with an estimated 89 million annual cases while Neisseria gonorrhoeae is also common with over 60 million cases estimated to occur annually. However, the most common nonviral STI, worldwide, is caused by Trichomonas vaginalis with over 174 million estimated cases in 1999. Potential sequelae of these STI in females include pelvic inflammatory disease (19, 22), ectopic pregnancy, tubal factor infertility (6), adverse pregnancy outcomes (3, 31), and potentially increased risk of both transmission and acquisition of human immunodeficiency virus (HIV) (1, 2, 14, 15, 25, 30, 31, 35). Additionally, investigators have shown epidemiologic associations with chlamydia (21, 28) or trichomonas (5, 34, 36, 40) infection and subsequent cervical neoplasia and carcinoma. While some of these studies attempted to control for human papillomavirus (HPV) infection, the relationships are unclear and the potential role that trichomonas plays has yet to be fully evaluated. The use of T. vaginalis PCR diagnostics as a research tool in our laboratory has led to the finding that infection with trichomonas is associated with an increased risk for HPV persistence in adolescent women, which may lead to increased risk of cervical cancer (27).
Screening and control programs for chlamydia and gonorrhea have become increasingly feasible due to the commercial availability of nucleic acid amplification tests (NAAT) and consequent acceptability of less invasively collected specimen types (10, 26, 29, 38). Recommendations for STI control programs strongly support screening of men as well as women, especially those who may be asymptomatic and therefore not seeking treatment. Such programs have been established in many areas of the world and have resulted in substantial declines in the prevalence of these infections (8, 11).
Screening for trichomonas infection has been performed using direct microscopy or culture of the organism in most settings that include T. vaginalis as a component of a sexually transmitted disease (STD) control program. However, many programs do not include trichomonas infection screening and indeed, in the United States, this is not a reportable infection tracked by the Centers for Disease Control and Prevention. Recent reports of the impact of trichomonas infections on the risk of HIV transmission and acquisition suggest that more aggressive trichomonas infection screening measures may be warranted. Laga et al. (15) demonstrated a nearly twofold increase in the odds of HIV infection in women infected with T. vaginalis in an African cohort. Although this was not significant at = 0.05, the research group subsequently demonstrated a 1.7-fold increase in relative risk for HIV infection for women infected with T. vaginalis (14), suggesting that trichomonas infection may increase women's susceptibility to HIV. Using mathematical modeling, 40% of the new HIV infections in women in the United States may be attributable to increased risk associated with trichomonas infection (2). Studies have also suggested an increase in the risk of transmission during trichomonas infection as a result of increased viral shedding (9, 35). In both men with urethritis (23) and women with vaginal infections (35), HIV viral shedding was quickly reduced following appropriate treatment for T. vaginalis infection. These data suggest that incorporating trichomonas infection control into chlamydia and gonorrhea control programs may be useful not only for reducing upper reproductive tract infections and adverse outcomes of pregnancy but also as a prevention measure to reduce the spread of HIV (25).
We have adapted a commercially available CT/NG NAAT to allow amplification and detection of trichomonas DNA. Using this assay, we demonstrated the utility of a single sample from women or men for screening for all three organisms and compared the results of the T. vaginalis PCR assay to direct microscopy or trichomonas culture to establish the performance characteristics of this test.
MATERIALS AND METHODS
Population and samples. Specimens were collected from men and women attending the Bell Flower clinic, a county health department STD clinic in an urban Midwestern setting. All subjects provided consent for samples to be used in evaluation of novel diagnostic assays. Although not all specimens were collected for each study, samples were collected in the following order: female urine, vaginal swab, cervical or urethral swab for gonorrhea culture, cervical or urethral swab for chlamydia culture, and male urine. Vaginal swabs were placed into sterile, dry tubes or into tubes containing T. vaginalis culture medium. Swabs for chlamydia culture were collected in chlamydia transport medium (CTM). First-catch urine with volumes ranging from 15 to 90 ml was collected. For experiments involving archived samples, data were available from original clinical wet preparations, but no vaginal swabs were retained. For studies of freshly obtained samples, female urine was not obtained. For the comparison of TVA and TVK primer sets, female vaginal swabs were used for both microscopy and PCR. For comparison of storage conditions of vaginal swabs, cervical swabs in CTM were also used to define infection.
ATCC T. vaginalis isolates 30184, 30185, and 30188 were used as positive T. vaginalis controls in titration experiments. C. trachomatis strain H-UW4 and a clinical N. gonorrhoeae isolate, 9003-0025, were used in the competition assays.
T. vaginalis wet mount and culture. Women attending the STD clinic were evaluated for the presence of T. vaginalis by direct microscopy of a wet preparation. Motile, flagellated trichomonads were identified by clinicians following the routine clinical procedure. Men were not routinely screened for trichomonas infection as a part of clinical exams. Culture was performed in the laboratory using Diamond's medium for vaginal swabs and the In-Pouch culture system for male urine sediment. Urine was held at room temperature prior to transport to the laboratory to maintain the viability of T. vaginalis for culture. Ten milliliters of urine was centrifuged at 900 x g for 10 min, the supernatant was discarded, and the sediment was inoculated into the In-Pouch system. Cultures were read using a wet preparation slide on days 0, 1, and 3. Laboratory strains for determination of analytical sensitivity were grown in Diamond's medium and harvested by centrifugation at 900 x g for 10 min.
CT/NG PCR. C. trachomatis and N. gonorrhoeae testing was performed using the Amplicor C. trachomatis/N. gonorrhoeae assay (Roche Diagnostic Corporation, Indianapolis, IN). All testing was performed according to the manufacturer's package insert unless otherwise noted and has been described previously (32). Briefly, specimens were lysed and processed for amplification following protocols specific for urine or swab samples in CTM. Vaginal swabs were transported in dry, sterile tubes. One milliliter of molecular-reagent-grade water was added to the swab followed by vigorous vortexing for at least 10 seconds to elute the sample. Swabs were discarded, and from this point forward, samples were processed as if collected in CTM. Processed samples were added to Amplicor Master Mix (MM) reagent containing polymerase, deoxynucleoside triphosphates, an internal control sequence to verify amplification, and uracil-N-glycosylase to reduce potential carryover contamination. This MM was not modified and contained only CT/NG primers. Samples were amplified in a Perkin-Elmer 9600 Thermocycler, and amplified product was detected using oligonucleotide probes specific to an internal sequence of the target DNA. The probes were prebound to polystyrene wells, and the biotinylated primers of amplified sequence were detected using a standard 96-well enzyme-linked immunosorbent assay format (Fig. 1). The optical density was measured at 450 nm, and results were interpreted according to the package insert.
T. vaginalis PCR. Amplification of T. vaginalis DNA was performed using biotinylated primers specific for this organism added to the MM reagent. Three sets of primers, TVA (24), TVK (13), and ?-Tub (17), and their corresponding oligonucleotide probes were evaluated (Table 1) using the published cycling parameters. Although the MM contained CT/NG primers, the T. vaginalis primers required different amplification parameters (for TVK PCR: holding at 95°C for 5 min, followed by 35 cycles of 90°C for 1 min, 60°C for 30 s, and 70°C for 2 min, followed by holding at 70°C for 10 min). T. vaginalis DNA amplification was always performed independently of C. trachomatis/N. gonorrhoeae amplification (Fig. 1). Amplicor detection kits, in a C. trachomatis- or N. gonorrhoeae-specific 96-well format, are sold separately from the specimen processing and amplification reagents. In order to detect amplicon from 96 amplified samples, one C. trachomatis kit and one N. gonorrhoeae detection kit are used. The kits contain identical denaturation, hybridization, avidin-horseradish peroxidase conjugate, substrate, and stop reagents. The only difference between the kits is the 96-well probe-bound plate specific for the C. trachomatis or N. gonorrhoeae amplicons. The samples amplified for C. trachomatis/N. gonorrhoeae are denatured using the reagent from only one of the two kits, leaving the other bottle of reagent available for denaturation of T. vaginalis amplification tubes. Each 96-well kit contains sufficient reagent for 144 wells with the least volume in the conjugate reagent. Therefore, using one C. trachomatis and one N. gonorrhoeae detection kit, 96 samples could be tested for all three pathogens by pooling reagents (Fig. 1). Detection of amplified T. vaginalis DNA was performed using the reagents from Amplicor CT/NG kits with the following exceptions. Primer set-specific probes were designed using Oligo (Molecular Biology Insights, Cascade, CO) software and were bound without a capture tag to polystyrene wells (Immulon II; ThermoLabsystems, Franklin, MA) by incubation overnight in freshly prepared 1 M ammonium acetate. Excess probe was removed by rinsing with phosphate-buffered saline with 1 mM EDTA. Plates were blotted, allowed to air dry for 3 h, sealed, and stored at 4°C until use. This method was developed from the probe binding method used for genital ulcer disease multiplex PCR (20). The hybridization buffer included in the CT/NG kit was too stringent to allow binding of the T. vaginalis amplified product; therefore, we used a generic PCR-digoxigenin enzyme-linked immunosorbent assay hybridization buffer (catalogue no. 1717472; Roche Applied Science, Indianapolis, IN) instead of the buffer provided with the CT/NG detection kit for detection of T. vaginalis amplicons. In response to this change, the denaturation reagent from the CT/NG kit was reduced in volume from 100 to 80 μl/well for denaturation of the T. vaginalis amplicons. From this point forward, all reagents were from the CT/NG kit and were used exactly as described in the package insert.
Samples with an A450 below 0.300 were interpreted as negative, those above 0.800 were interpreted as positive, and those falling between these values were considered equivocal. Samples with equivocal results were retested in duplicate with at least two/three results over 0.500 required to report the test result as positive. Cutoff points were initially based on recommendations by Roche for the CT/NG assay and the multiplex genital ulcer disease PCR.
Analysis. Patients were classified as infected if the routine diagnostic method (wet preparation or culture) was positive or if two separate PCR results were positive. When multiple samples were available from the same patient, in order to be classified as infected if the routine test was negative, positive PCR results were required from both samples. Similarly, for a comparison of primer sets using the same processed sample, to qualify as an infection when the routine method was negative, both PCR primer results had to be positive. When the routine test was negative and only one PCR result was positive, the PCR was considered a false-positive result. Comparison of multiple assay results was performed using the Cochran Q statistic that tests the hypothesis that there is no difference in dichotomous results between multiple measures on the same individual. McNemar's 2 test was used for pairwise comparisons. Pairwise comparisons were performed for the overall results and separately for infected and uninfected participants to evaluate differences in sensitivity and specificity, respectively. Results were considered significant at the = 0.05 level.
RESULTS
Evaluation of T. vaginalis PCR using laboratory strains. Laboratory strains of C. trachomatis, N. gonorrhoeae, and T. vaginalis were loaded onto Dacron swabs to assess both the limit of detection of the T. vaginalis PCR assay and the effect of competition for resources in the event of coinfections. To evaluate the sensitivity of the assay, trichomonads, chilled to 4°C to reduce motility, were counted in a wet preparation using a hemacytometer. Since patient samples were eluted in water, all serial dilutions were performed in water to achieve the desired number of organisms. Concentrations tested were 0.54 x 101 to 6.0 x 105 organisms per reaction tube, 25% of which went into the amplification tube. Testing was performed with both the TVA and TVK primer sets. Amplification using the TVA primers was detectable with as few as five trichomonads, and the TVK primer set was positive below the theoretical concentration of one trichomonad per processed sample. These results may overestimate the analytical sensitivity of the T. vaginalis PCR since the calculations were based on counted trichomonads and do not account for any lysed organisms from which DNA may have been present in the preparations. The experiment was performed in duplicate with identical results.
The addition of T. vaginalis primers to the MM raised the concern that the presence of the C. trachomatis and N. gonorrhoeae primers, which are included in the commercially available amplification reagent, might result in a reduction of T. vaginalis amplification in the presence of C. trachomatis and N. gonorrhoeae coinfections. To evaluate this potential effect, swabs were loaded with approximately 35 trichomonads and chlamydial elementary bodies and gonococci were added in concentrations ranging from 0 to 900 inclusion-forming units and 0 to 960 CFU, respectively. Swabs were eluted into 1 ml of water and processed as described for vaginal swabs. After processing, a total of 1.25% of the original sample was loaded into the amplification tube. Samples prepared in this manner were then evaluated using the TVA and TVK primer sets. Both TVA and TVK primers amplified T. vaginalis DNA at all concentrations of C. trachomatis and N. gonorrhoeae. Since the cycling parameters for T. vaginalis amplification were different from those used for C. trachomatis and N. gonorrhoeae amplification, no assays were performed to assess amplification of these organisms' DNA under T. vaginalis cycling parameters.
Comparison of TVA and TVK primer sets using vaginal samples. Vaginal swab samples from 174 women were evaluated by T. vaginalis PCR using the same processed sample tested by both the TVA and TVK primer sets. The T. vaginalis PCR results were compared to clinically obtained wet preparation microscopy results. T. vaginalis infection was defined by a positive wet mount or by vaginal swab PCR results positive by both primer sets. Thirty-nine (22.4%) women were classified as infected with T. vaginalis. Of these, 18 were positive by both TVA and TVK, only three of which were positive by wet mount. One infection was positive by TVA and not TVK primers, while 13 infections were detected by TVK and not TVA primers. Seven were positive by wet mount alone. The sensitivity of wet mount using this definition of infection was 61.5%, compared to the performance of the TVA and TVK primer sets with sensitivities of 48.7% and 79.5%, respectively (Table 2). Analysis of the agreement of the results, taking multiple comparisons into account, indicated that at least one of the tests performed differently from the others (P = 0.003). In pairwise analysis, the sensitivity of the TVK assay was higher than that of the TVA primer set (P = 0.002). There was no difference between the specificities of the two primer sets (P > 0.05).
Comparison of TVK and ?-Tub primer sets using male urine. First-catch urine specimens were collected from 503 men. Urine samples were stored at ambient temperature (AT) prior to transportation to the laboratory to facilitate T. vaginalis culture and arrived in the laboratory within 16 h of collection. Immediately following arrival, T. vaginalis cultures were prepared from urine sediment and prior to processing for PCR specimens were subsequently stored at 4°C for up to 4 days after collection. All samples were tested for T. vaginalis using the TVK and ?-Tub primer sets, and the TVA primer set was used for discrepant analysis. Infection was defined by a positive culture or two PCR-positive results with independent primer sets. Twenty-five (5.0%) men were classified as infected with T. vaginalis. For comparison, in this population 27 (5.4%) were positive for C. trachomatis and 26 (5.2%) were infected with N. gonorrhoeae. The TVK primer set identified 24 (96.0%) of the infected men, and the ?-Tub primer set identified 23 (92.0%) while T. vaginalis culture identified only 14 (56.0%) (Table 3). The sensitivities of the three assays were statistically different (P = 0.001). In pairwise comparisons, the sensitivity of both PCR assays was statistically different from that of culture (both P values < 0.01). The sensitivity and specificity of the two PCR assays were equivalent (both P values > 0.05). Based on the findings from both male and female samples, all further studies were performed using the TVK primer/probe set as the assay of interest.
Comparison of female specimen types and storage using the TVK primer set. C. trachomatis and N. gonorrhoeae diagnostic tests are often performed on endocervical or urine samples. Therefore, these specimen types were evaluated for use with T. vaginalis PCR. Infection was defined by a positive routine diagnostic test or by positive PCR results from at least two samples (e.g., vaginal and endocervical). In these evaluations, the requirement for duplicate PCR-positive results was intended to control for laboratory error rather than confirm specificity since that had been evaluated using alternate primer sets in the previous experiments.
Samples and wet mount data were collected from 186 women, 41 (22.0%) of whom were classified as infected with T. vaginalis (Table 4). Infection was defined by a positive wet mount, culture, or both endocervical and urine PCR results being positive. In this population, C. trachomatis was identified in 28 (15.1%) and N. gonorrhoeae was found in 17 (9.1%) of the participants. There was no difference in the prevalence of C. trachomatis or N. gonorrhoeae as a function of the presence of T. vaginalis (all P > 0.05; data not shown). Wet mount identified 21 (51.2%) T. vaginalis infections while culture detected 25 (61.0%). T. vaginalis PCR using vaginal and endocervical swabs identified 37 (90.2%) and 36 (87.8%) of the infected women, respectively. At least one test was significantly different from the others (P < 0.001) with pairwise comparisons indicating that T. vaginalis PCR using vaginal or endocervical samples identified more positive samples than did culture (P = 0.004 and 0.007, respectively) and wet mount (all P < 0.001). There was no difference in sensitivity or specificity between vaginal and cervical samples for PCR (all P > 0.05).
Residual, archived samples stored at –70°C for approximately 5 years were available from 463 women for comparison of endocervical swabs and urine specimens (Table 4). Infection was defined by a positive wet mount recorded in the clinical chart for that visit or by both endocervical and urine PCR-positive results. Ninety-four (20.3%) women were infected. Endocervical PCR, urine PCR, and wet mount identified 96.8, 94.7, and 51.1% of infected women, respectively. There was a significant difference between wet mount and endocervical or urine PCR (P < 0.001). Sensitivity and specificity of the assay were similar for both endocervical and urine specimens (P > 0.05).
Stability of vaginal swabs. To assess the stability of vaginal swab samples, endocervical swabs in CTM and duplicate vaginal samples were collected. Storage at 4°C was compared to holding samples at AT. Additionally, storage of samples for 3 days was compared to storage for 7 days. Trichomonas infection was defined by a positive wet preparation performed at the clinic or by two separate specimens (e.g., a cervical and a vaginal sample or both vaginal swabs) being positive by T. vaginalis PCR. Duplicate vaginal samples were available from 288 women for the evaluation of storage temperature. T. vaginalis infection was identified in 67 (23.3%) (Table 4). Storage temperature did not affect the sensitivity or specificity of the assay (all P > 0.05). Duplicate vaginal samples were obtained from 217 women for comparison of storage for 3 or 7 days postcollection (Table 4). Of the 46 T. vaginalis cases identified in this cohort, 32 (69.5%) and 35 (76.1%) were positive following specimen storage for 3 or 7 days, respectively. The sensitivity and specificity of the assay were similar for the two storage conditions (all P > 0.05).
DISCUSSION
The ability to integrate this test into routine screening for C. trachomatis and N. gonorrhoeae using an adaptation of commercially available reagents makes molecular biology-based diagnosis of T. vaginalis feasible for many diagnostic laboratories. The use of samples collected by minimally invasive or noninvasive methods that have been stored for several days will allow population-based screening for T. vaginalis. In these evaluation studies, the sensitivity and specificity of the T. vaginalis PCR using the TVK primer set were high. The assay using the TVK primer set and probe sequence performed better than the TVA primer/probe set and equivalently to the ?-Tub assay. These findings were consistent with the primer evaluation performed by Crucitti et al., who found the TVK primer set to be optimal for detection of T. vaginalis DNA (4). We chose to use the TVK assay for logistic reasons such as a shorter amplification process while using the ?-Tub assay for confirmatory purposes. The cutoff points used were arbitrarily chosen based on the manufacturer's recommendation. A bimodal distribution of optical density values similar to that described by Kaydos et al. (12) was seen with the results of the TVK assay. Evaluation of over 7,200 samples showed only 1.2% in the equivocal range (data not shown). With repeat testing of equivocals, the proportion of those classified as positive increased from 2.4% to 71.4% as the value of the initial result increased from 0.300 to 0.800, suggesting that the equivocal range is appropriate. It should be noted that, in the absence of a means for determining the true rate of infection with T. vaginalis, the performance characteristics described here are only estimates based on the best available techniques. In some analyses (the evaluations of sample types and storage conditions) we used the method under evaluation, requiring two separate samples to be positive, as a means of defining infection with the understanding that this introduces bias by inflating the apparent sensitivity and specificity of the assay. However, in these comparisons, sensitivity and specificity were similar to the primer set comparisons. In these studies the question of interest was comparability of performance, and this was demonstrated.
The objective of this work was to develop a T. vaginalis diagnostic tool that offers improvement over currently used methods such as microscopy and culture. The choice of wet preparation microscopy for women was based on the clinical standard in place during these evaluations. The clinic was not routinely culturing for trichomonas, and men were not being tested at all. Male urine, rather than perhaps more sensitive samples such as urine and urethral swabs combined, was selected as the least invasive addition to the clinical procedures in place. Since cultures were not routinely performed, the In-Pouch package insert recommendation for length of culture was followed. Therefore, the methods for identification of trichomonas infection were either equivalent or superior to the routine clinical practices at the participating STD clinic. Although alternate sample types or variations on comparator testing might have improved the estimates of sensitivity and specificity of the TVK assay, such measures will rarely be available in routine STD control programs and were of less interest than comparison to basic diagnostic procedures.
It was common for T. vaginalis to be defined by a positive wet preparation while both PCR samples gave negative results. For example, 17.9% (7/39) of cases shown in Table 3 had positive wet mount and negative PCR results. Women were classified as infected according to the a priori definition, but this may not have been an accurate definition. Wet preparations were read at the clinic site, and the quality of the results was not subject to further analysis, quality review, or proficiency testing. It is possible that wet preparations may have been called positive based on the presence of a species of motile trichomonad that does not contain TVK primer binding sites. If the infection status was more loosely defined as requiring at least two positive results, the number of infections dropped to 31 (17.8%) and the sensitivity of the tests was 54.8, 61.3, and 96.8% for wet mount, TVA PCR, and TVK PCR, respectively. Therefore, the sensitivity of the assay may be underestimated for all comparisons. This approach is justified since we did not attempt to measure inhibition of the PCRs, which might also explain the occurrence of wet-mount-positive/PCR-negative samples. However, even using this potentially conservative approach, the T. vaginalis PCR provided improved case finding for both men and women, as has been reported by other investigators (4, 16, 17, 26, 33, 37, 38).
The assay appears to be useful for a variety of sample types in women: endocervical swabs, urine, or vaginal swabs. This was in contrast to the poor performance of female urine as a sample type for T. vaginalis PCR described by Lawing et al. (16). We used whole-urine aliquots that had been frozen whereas the Lawing study used urine sediment frozen in water. The difference in storage conditions may explain the difference in urine sensitivity between our results and the previous study (95% and 64%, respectively). Additionally, in studies using fresh urine samples (12, 18, 33) urine performed well as a sample type. Finally, the prevalence of T. vaginalis for each comparison cohort was similar, 20.3 to 23.3%, suggesting that the sample types are truly comparable and that the storage conditions are amenable to sending samples to a reference laboratory. Encouragingly, the prevalence from the repository of cervical and urine samples stored at –70°C for up to 5 years was also within this range, indicating that long-term storage of either of these sample types likely has little effect on the quality of the sample for T. vaginalis PCR.
As we demonstrated with mock samples using laboratory strains, in a cohort of women with a high prevalence of C. trachomatis (15.1%) and N. gonorrhoeae (9.1%), the T. vaginalis PCR performed well in the presence of these other STI. This was of particular concern since the Amplicor MM to which T. vaginalis primers were added contained C. trachomatis and N. gonorrhoeae primer pairs as well. Consistent with global data, the prevalence of T. vaginalis in these studies in women was substantially higher than that of C. trachomatis or N. gonorrhoeae. Not surprisingly, the number of cases identified in men was lower than that in women (5.0%), but in this population, the rates of C. trachomatis (5.4%) and N. gonorrhoeae (5.2%) were also lower. The low STD prevalence rates may be attributable to the handling of the urine in this study. Urine was held at AT until delivery to the laboratory in order to avoid cold shock to any live trichomonads. Approximately two-thirds of urine samples were transported to the lab within 4 h of collection while the remainder were held overnight at the clinic. This delay of up to 16 h between collection and culture with subsequent refrigeration of the urine was possibly responsible for the degradation of both live organisms for culture and DNA for PCR of T. vaginalis. However, there was no difference in the frequency of positive results based on overnight storage (data not shown; P > 0.05).
Although other methods for amplification detection, such as use of a digoxigenin reporter molecule, may improve the sensitivity of detection of amplified product and be of use in research settings, the assay described here provides an improvement to current routine practice that can be used with very minor modifications to an available diagnostic kit. Kaydos and colleagues (12) saw variability with an assay similar to the one that we have described. This variability prompted them to develop an alternative assay which performed more reliably in their evaluations. Our laboratory has been using the assay described here for more than 6 years with a high degree of probe plate lot-to-lot reproducibility. This may be due in part to the use of the newer version of the Amplicor CT/NG reagents and the substitution of a different hybridization buffer rather than the C. trachomatis stand-alone kit used in the study by Kaydos et al. In 66 evaluations of reproducibility performed in our lab, there have only been two (3%) occasions where the reagents did not perform satisfactorily (data not shown). In these two instances the probe-bound plates did not provide adequate binding and the correlation between previous and new lots was less than 0.80. This indicates that the assay reproducibility should be closely monitored, as variations in the plate lot, prior to probe binding, and adherence to the binding procedure can affect assay performance. On all other occasions, the correlations were high, with only three instances of lot testing with correlations below 0.94. This suggests that the assay is stable and reproducible over time.
Several investigators have reported excellent results of screening for C. trachomatis and N. gonorrhoeae using either patient-obtained or clinician-collected vaginal swabs (10, 16, 29, 38). Many of these studies have also used a NAAT for the detection of T. vaginalis to enhance the screening program. Indeed, more recent evaluations of LightCycler assays for detection of T. vaginalis suggest improved sensitivity (7). However, the studies published to date rely on DNA isolation procedures, in-house PCR amplification reagents, cumbersome detection methods, or specialized equipment that is restricted to research laboratories. We have described a system which uses commercially available reagents, with the exception of the primers and probe oligonucleotides, which can easily be added to the routine for detection of C. trachomatis and N. gonorrhoeae in a clinical laboratory. The CT/NG assay can be performed without deviation from the package insert since the T. vaginalis primers are not added during the amplification of the C. trachomatis/N. gonorrhoeae assay. The separate amplification and detection of T. vaginalis does not have an effect on the performance of the C. trachomatis and N. gonorrhoeae testing, making this a suitable assay for laboratories adhering to Clinical Laboratory Improvement Act regulations. The costs associated with this test are minimal since the gross excess reagent provided in the Amplicor CT/NG detection kits can be pooled to allow T. vaginalis testing. The major expense is the amplification reagent to which T. vaginalis primers are added.
The value of this type of assay for screening programs is improved case ascertainment compared to traditional methods such as microscopy and culture (37). Given the increasing evidence that T. vaginalis may play a significant role in the transmission of HIV, both in pandemic regions such as sub-Saharan Africa (1, 14, 15) and in the United States (2, 24), sensitive assays for detection of T. vaginalis in a clinical laboratory setting are clearly needed. The assay described here is designed for this type of application. The use of commercially available reagents and samples processed for C. trachomatis and N. gonorrhoeae testing will allow clinical laboratories to adopt this method and utilize the quality practices of the manufacturers to optimize the quality and reproducibility of the diagnostic results.
ACKNOWLEDGMENTS
We thank Diane R. Stothard for detection probe design and Nancy J. Smith for technical assistance.
REFERENCES
Buvé, A., H. A. Weiss, M. Laga, E. Van Dyck, R. Musonda, I. Zekeng, M. Kahindo, S. Anagonou, L. Morison, N. J. Robinson, and R. J. Hayes for the Study Group on Heterogeneity of HIV Epidemics in African Cities. 2001. The epidemiology of trichomoniasis in women in four African cities. AIDS 15:S89-S96.
Chesson, H. W., J. M. Blandford, and S. D. Pinkerton. 2004. Estimates of the annual number and cost of new HIV infections among women attributable to trichomoniasis in the United States. Sex. Transm. Dis. 31:547-551.
Cotch, M. F., J. G. Pastorek, R. P. Nugent, S. L. Hillier, R. S. Gibbs, D. H. Martin, D. A. Eschenbach, R. Edelman, J. C. Carey, J. A. Regan, M. A. Krohn, M. A. Klebanoff, A. V. Rao, G. G. Rhoads, et al. 1997.
Trichomonas vaginalis associated with low birth weight and preterm delivery. Sex. Transm. Dis. 24:353-360.
Crucitti, T., E. Van Dyck, A. Tehe, S. Abdellati, B. Vuylsteke, A. Buve, and M. Laga. 2003. Comparison of culture and different PCR assays for detection of Trichomonas vaginalis in self collected vaginal swab specimens. Sex. Transm. Infect. 79:393-398.
Gram, I. T., M. Macaluso, J. Churchill, and H. Stalsberg. 1992. Trichomonas vaginalis (TV) and human papillomavirus (HPV) infection and the incidence of cervical intraepithelial neoplasia (CIN) grade III. Cancer Causes Control 3:231-236.
Grodstein, F., M. B. Goldman, and D. W. Cramer. 1993. Relation of tubal infertility to history of sexually transmitted diseases. Am. J. Epidemiol. 137:577-584.
Hardick, J., S. Yang, S. Lin, D. Duncan, and C. Gaydos. 2003. Use of the Roche LightCycler instrument in a real-time PCR for Trichomonas vaginalis in urine samples from females and males. J. Clin. Microbiol. 41:5619-5622.
Herrmann, B., and M. Egger. 1995. Genital Chlamydia trachomatis infections in Uppsala County, Sweden, 1985-1993: declining rates for how much longer? Sex. Transm. Dis. 22:253-260.
Hobbs, M. M., P. Kazembe, A. W. Reed, W. C. Miller, E. Nkata, D. Zimba, C. C. Daly, H. Chakraborty, M. S. Cohen, and I. Hoffman. 1999. Trichomonas vaginalis as a cause of urethritis in Malawian men. Sex. Transm. Dis. 26:381-387.
Hsieh, Y. H., M. R. Howell, J. C. Gaydos, K. T. McKee, Jr., T. C. Quinn, and C. A. Gaydos. 2003. Preference among female Army recruits for use of self-administrated vaginal swabs or urine to screen for Chlamydia trachomatis genital infections. Sex. Transm. Dis. 30:769-773.
Katz, B. P., M. J. Blythe, B. Van Der Pol, and R. B. Jones. 1996. Declining prevalence of chlamydial infection among adolescent girls. Sex. Transm. Dis. 23:226-229.
Kaydos, S. C., H. Swygard, S. L. Wise, A. C. Sena, P. A. Leone, W. C. Miller, M. S. Cohen, and M. M. Hobbs. 2002. Development and validation of a PCR-based enzyme-linked immunosorbent assay with urine for use in clinical research settings to detect Trichomonas vaginalis in women. J. Clin. Microbiol. 40:89-95.
Kengne, P., F. Veas, N. Vidal, J. L. Rey, and G. Cuny. 1994. Trichomonas vaginalis: repeated DNA target for highly sensitive and specific polymerase chain reaction diagnosis. Cell. Mol. Biol. 40:819-831.
Laga, M., M. Alary, N. Nazila, A. T. Manoka, M. Tuzila, F. Behets, J. Goeman, M. St. Louis, and P. Piot. 1994. Condom promotion, sexually transmitted diseases treatment, and declining incidence of HIV-1 infection in female Zairian sex workers. Lancet 344:246-248.
Laga, M., A. Manoka, M. Kivuvu, B. Malele, M. Tuliza, N. Nzila, J. Goeman, F. Behets, V. Batter, and M. Alary. 1993. Non-ulcerative sexually transmitted diseases as risk factors for HIV-1 transmission in women: results from a cohort study. AIDS 7:95-102.
Lawing, L. F., S. R. Hedges, and J. R. Schwebke. 2000. Detection of trichomonosis in vaginal and urine specimens from women by culture and PCR. J. Clin. Microbiol. 38:3585-3588.
Madico, G., T. C. Quinn, A. Rompalo, K. T. McKee, Jr., and C. A. Gaydos. 1998. Diagnosis of Trichomonas vaginalis infection by PCR using vaginal swab samples. J. Clin. Microbiol. 36:3205-3210.
Mayta, H., R. H. Gilman, M. M. Calderon, A. Gottlieb, G. Soto, I. Tuero, S. Sanchez, and A. Vivar. 2000. 18S ribosomal DNA-based PCR for diagnosis of Trichomonas vaginalis. J. Clin. Microbiol. 38:2683-2687.
Moodley, P., D. Wilkinson, C. Connolly, J. Moodley, and A. W. Sturm. 2002. Trichomonas vaginalis is associated with pelvic inflammatory disease in women infected with human immunodeficiency virus. Clin. Infect. Dis. 34:519-522.
Orle, K. A., and J. B. Weiss. 1997. Detection of Treponema pallidum, Haemophilus ducreyi, and herpes simplex virus by multiplex PCR, p. 67-79. In R. W. Peeling and P. F. Sparling (ed.), Sexually transmitted diseases: methods and protocols. Humana Press, Totowa, New Jersey.
Paavonen, J., K. P. Karunakaran, Y. Noguchi, T. Anttila, A. Bloigu, J. Dillner, G. Hallmans, T. Hakulinen, E. Jellum, P. Koskela, M. Lehtinen, S. Thoresen, H. Lam, C. Shen, and R. C. Brunham. 2003. Serum antibody response to the heat shock protein 60 of Chlamydia trachomatis in women with developing cervical cancer. Am. J. Obstet. Gynecol. 189:1287-1292.
Paisarntantiwong, R., S. Brockman, L. Clarke, S. Landesman, J. Feldman, and H. Minkoff. 1995. The relationship pf vaginal trichomoniasis and pelvic inflammatory disease among women colonized with Chlamydia trachomatis. Sex. Transm. Dis. 22:344-347.
Price, M. A., D. Zimba, I. F. Hoffman, S. C. Kaydos-Daniels, W. C. Miller, F. Martinson, D. Chilongozi, E. Kip, E. Msowoya, M. M. Hobbs, P. N. Kazwmbe, and M. S. Cohen. 2003. Additional of treatment for trichomoniasis to syndromic management of urethritis in Malawi: a randomized clinical trial. Sex. Transm. Dis. 30:516-522.
Riley, D. E., M. C. Roberts, T. Takayama, and J. N. Krieger. 1992. Development of a polymerase chain reaction-based diagnosis of Trichomonas vaginalis. J. Clin. Microbiol. 30:465-472.
Rothenberg, R. B., J. N. Wasserheit, M. E. St. Louis, J. M. Douglas, et al. 2000. The effect of treating sexually transmitted diseases on the transmission of HIV in dually infected persons: a clinic-based estimate. Sex. Transm. Dis. 27:411-416.
Schwebke, J. R., and E. W. Hook III. 2003. High rates of Trichomonas vaginalis among men attending a sexually transmitted diseases clinic: implications for screening and urethritis management. J. Infect. Dis. 188:465-468.
Shew, M. L., J. D. Fortenberry, W. Tu, B. E. Juliar, B. E. Batteiger, B. Qadadri, and D. R. Brown. Association of condom use, sexual behaviors and sexually transmitted infections on genital human papillomavirus infection among adolescent women. Arch. Pediatr. Adolesc. Med., in press.
Smith, J. S., C. Bosetti, N. Munoz, R. Herrero, F. X. Bosch, J. Eluf-Neto, C. J. Meijer, A. J. Van Den Brule, S. Franceschi, R. W. Peeling, and IARC Multicentric Case-Control Study. 2004. Chlamydia trachomatis and invasive cervical cancer: a pooled analysis of the IARC multicentric case-control study. Int. J. Cancer 111:431-439.
Smith, K., K. Harrington, G. Wingood, M. K. Oh, E. W. Hook III, and R. J. DiClemente. 2001. Self-obtained vaginal swabs for diagnosis of treatable sexually transmitted diseases in adolescent girls. Arch. Pediatr. Adolesc. Med. 155:676-679.
Sorvillo, F., L. Smith, P. Kerndt, and L. Ash. 2001. Trichomonas vaginalis, HIV, and African-Americans. Emerg. Infect. Dis. 7:927-932.
Sutton, M. Y., M. Sternberg, M. Nsuami, F. Behets, A. M. Nelson, and M. St. Louis. 1999. Trichomoniasis in pregnant human immunodeficiency virus-infected and human immunodeficiency virus-uninfected Congolese women: prevalence, risk factors, and association with low birth weight. Am. J. Obstet. Gynecol. 181:656-662.
Van Der Pol, B., T. C. Quinn, C. A. Gaydos, K. Crotchfelt, J. Schachter, J. Moncada, D. Jungkind, D. H. Martin, B. Turner, C. Peyton, and R. B. Jones. 2000. Multicenter evaluation of the AMPLICOR and automated COBAS AMPLICOR CT/NG tests for detection of Chlamydia trachomatis. J. Clin. Microbiol. 38:1105-1112.
van der Schee, C., A. van Belkum, L. Zwijgers, E. van der Brugge, E. L. O'Neill, A. Luijendijk, T. Rijsoort-Vos, W. I. van der Meijden, H. Verbrugh, and H. J. Sluiters. 1999. Improved diagnosis of Trichomonas vaginalis infection by PCR using vaginal swabs and urine specimens compared to diagnosis by wet mount microscopy, culture, and fluorescent staining. J. Clin. Microbiol. 37:4127-4130.
Viikki, M., E. Pukkala, P. Nieminen, and M. Hakama. 2000. Gynaecological infections as risk determinants of subsequent cervical neoplasia. Acta Oncol. 39:71-75.
Wang, C. C., R. S. McClelland, M. Reilly, J. Overbaugh, S. R. Emery, K. Mandaliya, B. Chohan, J. Ndinya-Achola, J. Bwayo, and J. K. Kreiss. 2001. The effect of treatment of vaginal infections on shedding of human immunodeficiency virus type 1. J. Infect. Dis. 183:1017-1022.
Watts, D. H., M. Fazzari, H. Minkoff, S. L. Hillier, B. Sha, M. Glesby, A. M. Levine, R. Burk, J. M. Palefsky, M. Moxley, L. Ahdieh-Grant, and H. D. Strickler. 2005. Effects of bacterial vaginosis and other genital infections on the natural history of human papilloma virus infection in HIV-1-infected and high-risk HIV-1-uninfected women. J. Infect. Dis. 191:1129-1139.
Wendel, K. A., E. J. Erbelding, C. A. Gaydos, and A. M. Rompalo. 2002. Trichomonas vaginalis polymerase chain reaction compared with standard diagnostic and therapeutic protocols for detection and treatment of vaginal trichomoniasis. Clin. Infect. Dis. 35:576-580.
Witkin, S. S., S. R. Inglis, and M. Polaneczky. 1996. Detection of Chlamydia trachomatis and Trichomonas vaginalis by polymerase chain reaction in introital specimens from pregnant women. Am. J. Obstet. Gynecol. 175:165-167.
World Health Organization. 2001. Global prevalence and incidence of selected curable sexually transmitted infections: overview and estimates. World Health Organization, Geneva, Switzerland.
Zhang, Z. F., S. Graham, S. Z. Yu, J. Marshall, M. Zielezny, Y. X. Chen, M. Sun, S. L. Tang, C. S. Liao, and J. L. Xu. 1995. Trichomonas vaginalis and cervical cancer. A prospective study in China. Ann. Epidemiol. 5:325-332.
Division of Infectious Diseases, Indiana University School of Medicine, Indianapolis, Indiana,
1 Department of Internal Medicine, Emory University School of Medicine, Atlanta, Georgia(Barbara Van Der Pol, Coll)