Three-Year History of Transcription-Mediated Amplification-Based Trichomonas vaginalis Analyte-Specific Reagent Testing in a Subacute Care Patient Population

Abstract

A total of 7,899 specimens submitted for live clinical Trichomonas vaginalis analyte-specific reagent (ASR) screening from 2008 to 2010 were audited on the basis of patient gender, specimen source, molecular Neisseria gonorrhoeae and Chlamydia trachomatis results, and relative light unit (RLU) data yielded by T. vaginalis ASR. Only 1.4% of the screening was ordered by emergency department clinicians. The screening volume in 2010 was 126% higher than that in 2008. The proportions of annual female and male screening remained consistent throughout the 3-year interval (∼92 and 8%, respectively). Although 71.8 and 9.5% of screening was performed on endocervical and vaginal specimens, respectively, over the 3-year period, no significant difference was noted in the T. vaginalis detection rates (8.9 and 8.6%, P = 0.85). Increased T. vaginalis detection was derived from female urine specimens (12.6%) compared to female genital swabs (P = 0.0004). The proportion of female urine screening increased during the 3-year interval (P < 0.0002). T. vaginalis detection rate in males was 6.6%, with no difference between urethral and urine T. vaginalis detection (P = 0.53). The mean RLU value for 714 positive specimens was 3,971,441; analogous values for each female specimen source and combined male source testing showed no variance (P ≥ 0.29). Combined-gender T. vaginalis detection rate (9.1%) was significantly greater than those of C. trachomatis (5.9%) and N. gonorrhoeae (1.5%; P < 0.0002). Equivocal results presented at a rate of 0.4%. T. vaginalis ASR is an increasingly utilized assay that yields higher detection rates than other sexually transmitted infection etiologies in this community subacute care setting.

INTRODUCTION

Trichomonas vaginalis is globally recognized as one of the most common and curable etiologies of sexually transmitted infection (STI) (3). Despite current estimates of 180 million cases worldwide and 3 to 5 million cases occurring annually in the United States (32), trichomoniasis continues not to be a reportable disease to public health authorities. This status pervades clinical and public health in spite of data suggesting that T. vaginalis infection can be associated with deleterious pregnancy outcomes (7) and coinfection with Neisseria gonorrhoeae (13). Additional data have suggested an antecedent role for the organism in human immunodeficiency virus acquisition (19) and transmission (14, 22).

Advances in molecular diagnostics specific to T. vaginalis have yielded amplification assays that demonstrate increased sensitivity compared to culture-based (15, 25), microscopic (15, 24, 25), nucleic acid hybridization-based (1), and antigen-based (15) methods. Validation of the transcription-mediated amplification (TMA)-based T. vaginalis analyte-specific reagent (ASR; Gen-Probe, Inc., San Diego, CA) was completed by our laboratory in early 2007 on vaginal (23) and endocervical collections (24). TMA analysis of both specimen sources, in concert with confirmatory alternative target testing, revealed that the assay approximately doubled the T. vaginalis detection rate in a high-prevalence STI population compared to microscopic analysis of vaginal collections (wet mount) (24).

After the introduction of clinical T. vaginalis ASR screening, subsequent validation of female urine specimens, along with male urethral and urine specimens (12), was completed. The present report describes findings from a 3-year audit of clinical T. vaginalis ASR (subsequently marketed as the APTIMA Trichomonas vaginalis assay with U.S. Food and Drug Administration indications for female urine and genital swab testing) screening of a largely subacute care demographic. The reported data from this very large study set have the potential to impact primary care providers in terms of acknowledging true STI etiology prevalence, as well as selecting appropriate specimens for laboratory diagnosis.

(The results of this study were previously presented, in part, at the 111th General Meeting of the American Society for Microbiology, New Orleans, LA, 21 to 24 May 2011.)

MATERIALS AND METHODS

Setting.

A recent analysis of data from U.S. metropolitan statistical areas (MSAs) documents a long-standing trend of high STI prevalence in the Milwaukee (Wisconsin) metropolitan area (4). These data reveal a Milwaukee-Waukesha-West Allis MSA chlamydia rate of 672.2 per 100,000 population. This figure ranked the second highest in the United States and was 81.6% higher than the national cumulative MSA rate of 370.2 per 100,000 population. The same MSA had the second-highest gonorrhea rate among U.S. MSAs (328.5 per 100,000 population; nearly three times the national MSA total rate of 118.9 per 100,000 population). Wheaton Franciscan Laboratory serves five Milwaukee metropolitan hospitals (with one major urgent-care facility) and an ∼70-clinic metropolitan outpatient physician group in a three-county region of southeastern Wisconsin. T. vaginalis ASR was offered to clinicians as a stand-alone assay, as a reflexive test from negative wet mounts, or in conjunction with C. trachomatis and N. gonorrhoeae molecular screening.

Specimen collection.

Primary endocervical, vaginal, and male urethral specimens were deposited into APTIMA swab specimen transport tubes per APTIMA Combo 2 Assay (Gen-Probe) package insert protocol. In analogous fashion, ∼2-ml aliquots of first-void female and male urine were added to APTIMA urine specimen transport tubes. In some instances, vaginal discharges were sampled with a Dacron swab and placed into 1.0-ml 145 mM NaCl. These vaginal saline suspensions were held at ambient conditions prior to microscopy. When reflexive T. vaginalis ASR was requisitioned, 200-μl aliquots of primary suspensions demonstrating no motile trichomonads were transferred into specimen lysis tubes (23).

Primary molecular screening assays.

T. vaginalis ASR, detecting organism-specific 16S rRNA by utilizing the principles of target capture, TMA, and chemiluminescent hybridization protection, was previously validated on the aforementioned specimen sources in conjunction with a proprietary alternative target TMA-based confirmatory assay (Gen-Probe). In brief, target molecules were isolated via oligomer-specific capture onto the surface of magnetic particles. Unique primers, in tandem with reverse transcriptase, catalyzed construction of a double-stranded DNA binding site for RNA polymerase. After isothermal transcription, product detection was achieved by a kinetic assay that used a single-stranded DNA probe labeled with an acridinium ester molecule. Relative light units (RLU) of ≥50,000 were interpreted as positive; the remaining output of ≥20,000 RLU was classified as equivocal. Utilizing similar theory and protocol (10), the APTIMA Combo 2 Assay detected N. gonorrhoeae-specific 16S rRNA and Chlamydia trachomatis-specific 23S rRNA.

Audit.

In a Wheaton Franciscan Healthcare institutional review board-approved investigation, a 3-year audit of all primary clinical specimens subjected to T. vaginalis ASR was performed. Because of high specificity values associated with wet mount analysis (24), patients with positive wet mounts were excluded from further T. vaginalis ASR screening and data analysis. More than 98% of all specimens were additionally screened for C. trachomatis- and N. gonorrhoeae-specific RNA. The data pertinent to patient gender, specimen source, screening result, and T. vaginalis ASR-specific RLU were collected for consecutive 6-month intervals.

Statistical analysis.

The significance test of proportions determined whether changes in either rates of positive screening result or proportion of given specimen source were significant. The t test for independent samples determined whether differences in mean RLU associated with positive results were significant between specimen sources. The alpha level was set at 0.05 before the investigations commenced, and all P values are two-tailed.

RESULTS

Temporal increase in T. vaginalis ASR testing volume.

Subacute care facilities placed the vast majority of requisitions for T. vaginalis ASR screenings, with urban emergency departments contributing only 110 of 7,899 total requests (1.4%). Annually, between 91.6 and 92.5% of all requisitions for T. vaginalis ASR screening were performed on females (Fig. 1). Total screening volumes rose 42.1% from year 1 to year 2 of the audit, with an additional increase of 58.8% from year 2 to year 3. Toward the latter portion of the audit, the annual increase in female screenings (60.3% increase from year 2 to year 3) outpaced that increase in requisitions for male screenings (42.1%).

Fig. 1.

Frequency of T. vaginalis transcription-mediated amplification-based analyte-specific reagent performance per 6-month interval (), delineated by male (▪) and female (□) screening, during a 3-year clinical testing experience.

T. vaginalis ASR specimen source distribution.

In the first half of 2008, endocervical swabs and female urine comprised 78.4 and 5.0% of all specimens submitted for T. vaginalis ASR, respectively (Fig. 2). Over time, a significant shift (P < 0.0002) toward increased utilization of female urine was noted. By the second half of 2010, female urine accounted for 13.8% of specimens while endocervical swabs declined to 68.0% (Fig. 2). Gender-specific data revealed an 84.6 and 5.4% utilization of endocervical swabs and urine, respectively, in the first half of 2008. These values changed inversely by the end of the audit (73.5% cervical swabs, 14.9% urine; both P < 0.0002). No change in the proportion of vaginal swab submissions between the two intervals was noted (P = 0.26). No analogous trend was seen in male urine screenings, with relatively consistent percentages (5.5 to 6.6%) over the 3-year audit (P ≥ 0.33; Fig. 2).

Fig. 2.

Distribution of T. vaginalis transcription-mediated amplification-based analyte-specific reagent specimen sources, per chronological 6-month interval, during a 3-year clinical testing experience beginning in 2008. Shaded bars indicate endocervical specimens, vertically striped bars indicate vaginal specimens, and open bars indicate female urine specimens; solid bars indicate urethral specimens, and diagonally striped bars indicate male urine specimens.

T. vaginalis ASR operational characteristics, delineated by specimen source.

T. vaginalis ASR generated only 31 equivocal results in this population of nearly 8,000 specimens (0.39%; Table 1 ). No significant variation existed among specimen sources in generating these equivocal results. Median RLU per specimen source yielding positive screening results ranged from 4.32 million to 4.52 million. Tenth percentile data (Table 1) revealed that few such specimens yielded RLU output near a user-defined cutoff value. Due to the paucity of male urethral swab submissions, male urine and urethral swab data were combined into a gender-specific data set. On this basis, mean RLU per specimen source yielding a positive screening result did not differ significantly (P ≥ 0.29; Table 1).

Table 1.

Characterization of luminescence data output, delineated by specimen source, from a 3-year clinical T. vaginalis transcription-mediated amplification-based analyte-specific reagent testing experience

Specimen	n	Equivocal results (%)	RLUa
			Mean	10th percentile	Median	Range
Endocervix	5,671	25 (0.44)	3,933,989	956,549	4,426,962	53,792–6,172,227
Vagina	750	2 (0.27)	3,973,538	274,319	4,370,286	64,358–5,744,816
Female urine	856	3 (0.35)	4,097,909	2,534,578	4,320,966	86,974–5,674,339
Combined maleb	622	1 (0.16)c	4,098,849	2,055,508	4,515,039	90,625–5,434,705
Cumulative	7,899	31 (0.39)	3,971,441	1,313,117	4,402,870	53,792–6,172,227

^aRLU values greater than 50,000 constituted a positive result. RLU values were derived from specimens yielding a positive result. The RLU per specimen source exhibited no significant difference (P ≥ 0.29; t test of independent samples).

^bIncludes data from urethra (n = 145) and urine (n = 477) specimens.

^cObserved in one urethra specimen.

Rate of T. vaginalis-specific RNA detection.

Within a subset of 7,767 patients from whom molecular analysis for all STI agents was requisitioned, T. vaginalis detection rate was 9.1% (Table 2) and was statistically greater than rates for either C. trachomatis (5.9%) or N. gonorrhoeae (1.5%; both P < 0.0002). Detection rates from endocervical and vaginal swabs were 8.9 and 8.6%, respectively, and showed little variance (P = 0.85); however, these values were significantly lower than the detection rate from female urine (12.6%; P = 0.0004). T. vaginalis detection from females (9.3%) exceeded that of C. trachomatis (5.7%) and N. gonorrhoeae (1.4%; both P ≤ 0.0002). In contrast, the T. vaginalis detection rate in males (6.6%) did not significantly differ from that of C. trachomatis (8.8%; P = 0.17) but was increased over that of N. gonorrhoeae (3.2%; P = 0.006).

Table 2.

Detection of Trichomonas vaginalis-, Chlamydia trachomatis-, and Neisseria gonorrhoeae-specific RNA, per specimen source, in a subacute care population utilizing transcription-mediated amplification assays

Specimen	% Detection
	Trichomonas vaginalisa	Chlamydia trachomatis	Neisseria gonorrhoeae
Endocervix	8.9	5.7	1.3
Vagina	8.6	4.8	0.4
Female urine	12.6	6.0	2.6
Male urine	7.0	8.1	3.6
Urethra	5.5	11.0	2.1
Cumulative	9.1	5.9	1.5

^aT. vaginalis ASR screens from 132 patient encounters were excluded from calculation of specimen source-specific detection rates because transcription-mediated amplification-based screening for C. trachomatis and/or N. gonorrhoeae was not requisitioned.

DISCUSSION

The vast majority of T. vaginalis ASR screenings (98.6%) over the 3-year audit were requisitioned in a subacute care setting. Multiple factors may be involved in this phenomenon. In this setting, practitioners may not always have immediate access to equipment to perform wet mounts. Past data suggest that this delay may produce false-negative results due to attenuation of trophozoite motility (9). These microscopic analyses may be further compromised by issues related to training, proficiency, and competency of non-laboratorian observers (8, 29, 30). Utility of the T. vaginalis ASR as a reflex test in the context of a negative wet mount has been intimated (23). Such utilization may counteract the decreased sensitivity of wet mount analysis.

Patient diagnosis data were not available within the context of this audit to delineate symptomatic infection from asymptomatic organism carriage. It can be reasonably assumed that a smaller percentage of patients in this audit showed significant signs and symptoms of illness, given that these individuals largely did not present to an emergent or acute care facility. Moreover, some may have been subjects of routine STI screening. Notwithstanding, studies have estimated that 50 to 85% of individuals with T. vaginalis infection are asymptomatic, with many of these patients developing acute disease within 6 months (18, 31). Although the overall prevalence of T. vaginalis in women was approximately 10 to 15% less in this subacute care demographic compared to emergent care populations (16, 23, 24), significant T. vaginalis detection rates in both males and females (Table 2), as well as statistically increased T. vaginalis detection rates over those of C. trachomatis and N. gonorrhoeae provide evidence supporting consistent molecular screening for this agent. Indeed, appreciable T. vaginalis detection rates have been demonstrated in both symptomatic and asymptomatic populations (25, 27, 28).

An endocervical swab was initially the specimen of choice for female T. vaginalis ASR screening with a concomitant small percentage of urine specimens. Over time, these proportions shifted significantly as the overall percentage of female urine submitted for T. vaginalis ASR rose nearly 10% by the second half of 2010. This paradigm shift was not demonstrated in male populations over the course of the audit. From a technical standpoint, urine testing can potentiate generation of detectable amplicon with equivalent efficacy as genital specimens (Table 1). Furthermore, a 0.23% rate of equivocal result generated by combined-gender urine specimens contributes to the overall robustness of the assay. Benefits of paramagnetic target capture for removal of endogenous inhibitory substances, especially those found in urine, have been described (5, 11, 17, 21). Despite this phenomenon, the proportion of female-to-male T. vaginalis screening in this demographic did not change significantly over the 3-year interval (Fig. 1).

Urine specimens yielded a significantly higher percentage of T. vaginalis-positive results in women (12.6%) compared to either endocervical or vaginal specimens. A limitation inherent to this datum was this audit being based on live clinical testing experience in which a single specimen source was procured from each patient. Nye et al. (25) assessed the accuracy of TMA-based detection of T. vaginalis in women from whom vaginal, endocervical, and urine specimens were collected at the same health care encounter. Increased sensitivity of vaginal specimens (96.6%) over those of endocervical or urine specimens (89.8 and 87.5%, respectively) was reported in this population when a reference standard combining microscopy, culture, and molecular data was used.

Other studies have reported that sensitivity of vaginal swabs exceeds those of endocervical swabs or urine for detection of T. vaginalis (20, 33). However, past studies used DNA amplification methods that, in a general sense, have been shown clinically (6, 26) and in vitro (5, 17) to be less sensitive than TMA-based, target capture-supplemented methodology. Interestingly, the study published by Andrea and Chapin (1) focused on co-collections of vaginal swabs plus either endocervical swabs or urine. Similar to our findings, these authors demonstrated a urine T. vaginalis detection rate (11.5%; 22 of 191) that exceeded rates derived from endocervical and vaginal collections (3.2 and 5.2%, respectively). This suggests that urine may become a specimen of choice for T. vaginalis detection by highly sensitive TMA; however, additional studies must validate this hypothesis. Moreover, T. vaginalis detection rates from all female specimen sources, which greatly exceeded those for either C. trachomatis or N. gonorrhoeae, indicate that screening for this protozoan should be performed on a routine basis to identify the apparently large proportion of patients with asymptomatic infection.

No significant difference between T. vaginalis and C. trachomatis detection rates was observed in the large study set of 622 male specimens; however, the N. gonorrhoeae detection rate was significantly decreased from those of the other two STI etiologies. A paucity of studies has examined the role of laboratory diagnosis for management of trichomoniasis in men (2, 12, 25, 28). One of these studies (28) utilized DNA amplification technology to demonstrate detection of T. vaginalis in a greater proportion of asymptomatic males (51.4%) than in those with symptoms (23%; P = 0.009). A recent study described 42 males with laboratory-diagnosed trichomoniasis, with 41 of these determinations made by TMA-based methods (25). Twenty-six of these men (61.9% of the total cohort; 63.4% of men with a positive TMA screen) were characterized by a clinician as having an unremarkable presentation. A similar majority of the males with a positive TMA result (53.7%) self-reported as being asymptomatic.

In conclusion, subacute care practitioners have become pivotal in the diagnosis and treatment of trichomoniasis in both genders. T. vaginalis ASR presents an alternative to insensitive wet mount microscopy and can serve as a reflex test from negative wet mounts originating from settings that demonstrate sufficient competency in microscopy. As a result, practitioners now have the capacity for comprehensive assessment of symptomatic patients as well as routine screening of asymptomatic individuals in subacute care settings. An elevated recovery rate of STI etiologies, especially that of T. vaginalis, from urine specimens may augment this TMA-based strategy. The steady increase in testing volume over the 3-year audit suggests that clinicians indeed have expanded their focus within the realm of STI treatment and prevention.