Association of Trichomonas vaginalis and Cytological Abnormalities of the Cervix in Low Risk Women

Objective Is Trichomonas vaginalis (TV) an inducing factor for the development of (pre-)cancerous lesions of the cervix? Design Cross sectional study. Setting Screening healthy Belgian women with low infection risk. Sample 63,251 consecutive liquid based cervical samples. Methods Real time quantitative PCR for presence of TV, 18 HPV types and Pap smear analysis of cytologic abnormalities. Main Outcome Measures Association of TV and HPV with cervix dysplasia Results The overall prevalence of TV DNA was 0.37%, of low risk HPV 2%, of high risk HPV 13.2%, and 8.8 % had cytological abnormalities. Both LR-HPV and HR-HPV were significantly associated with all cytological abnormalities. Presence of TV was associated with LR- and HR-HPV, ASC-US and HSIL, but not with other abnormalities. All women with TV and HSIL also had HR-HPV, while the latter was present in only 59% of women with TV and ASC-US. Amongst HPV negative women, TV was found in 1.3% of women with ASC-US, but only in 0.03% of women with normal cytology (OR 4.2, CL95% 2.1-8.6). In HR-HPV positive women, presence of TV increased the likelihood of cytological abnormalities somewhat (P=0.05), mainly due to an increase in ASC-US and LSIL, but not HSIL. Conclusions We conclude that TV infection is associated with both LR and HR-HPV infection of the cervix, as well as with ASC-US and HSIL. TV is a concomitant STI, but is not thought to be a co-factor in the causation of HSIL and cervical cancer. However, TV may cause false positive diagnoses of ASC-US.


Introduction
Both human papilloma virus (HPV) and Trichomonas vaginalis (TV) infection are common sexually transmitted infections (STIs) worldwide [1]. The increased risk of coexistent infections once a first STI is diagnosed is well known and leads to the recommendation to test for other genital infections simultaneously [2]. Despite this, it is not uncommon for women with HPV infections, nowadays readily detected by extended testing for cervical dysplasia and cancer, not to be tested for other STIs. TV is an under-recognized condition despite its clinical importance [3][4][5], probably due to its perceived rarity, the additional diagnostic effort requiring a fresh specimen for microcopy and/or culture, and its largely asymptomatic appearance in men and women [6,7]. Introduction of polymerase chain reaction (PCR)-based detection techniques for TV, 5 low-risk (LR) and 13 high-risk (HR) HPV types allowed us to determine the prevalence of these infections in our population and its relationship with HPV infection [8]. In some studies, TV is depicted as a risk factor for cervical dysplasia and cancer [9]. We assessed the possible role of TV in the pathogenesis of such precancerous lesions of the cervix. All clinical cervico-vaginal samples, received between 15 th  April and 1 st November 2008 by the Laboratory for Clinical  Pathology (labo RIATOL), Antwerp, Belgium, for cytologic examination and TV testing, were tested for presence of Trichomonas vaginalis and HPV DNA. The medical ethics committee of the University of Antwerp confirmed that ethical oversight is unnecessary for retrospective case studies such as this. All data were anonymized, no specific tests were performed outside routine practice, and there was no cost or additional risk to the patient.

Study population and used tests
Cervico-vaginal cells were collected with the Cervex-Brush Combi (Rovers Medical Devices B.V., Oss, The Netherlands). After collection, brush heads were transferred directly into a vial with BD-SurePath preservative fluid [10]. From the fluid containing the cellular material, a liquid based cytology sample was prepared with the robotic BD PrepStain Slide Processor, previously AutoCyte PREP System (BD Diagnostics -TriPath, Burlington, NC, USA) [11,12]. All slides are prescreened using BD-FocalPoint, a computerized scanning system for the primary screening of cervical smears (BD Diagnostics), followed by targeted microscopic interpretation of selected suspicious fields using BD-FocalPoint guided screening review stations. Cytology results were classified according to the Bethesda system 2001 [13,14] thus: negative for intraepithelial lesions and malignancies (NILM); atypical squamous cells of undetermined significance (ASC-US); atypical glandular cells (AGC); low-grade squamous intraepithelial lesions (LSIL); atypical squamous cells of undetermined significance, cannot exclude high-grade squamous intraepithelial lesions (ASC-H); and high-grade squamous intraepithelial lesions (HSIL). DNA isolation from the cellular pellet remaining after processing of the BD SurePath™ cytology specimen was performed as previously described [14,15]. The presence of HPV genotypes and Trichomonas vaginalis was determined using a multiplex TaqMan-based real-time quantitative PCR. The sequences for the primers and probes used for different type specific HPV qPCRs can be found in the paper by Micalessi and coworkers [16]. For detection of HPV genotypes type-specific sequences of viral E6 or E7 genes: 6 E6, 11 E6, 16 [14,15]. HPV types 6, 11, 53, 66 and 67 were also detected but were not considered as high-risk (LRHPV). The target for detection of TV was a 67base pair region of a repeated sequence of the TV genome (Gene Bank Accession Number L23861) [8,18,19]. Multiple validation tests under different conditions were done with 15 clinical samples with proven TV, and 150 proven negative samples before commencing large scale testing.
The analytic sensitivity of the different PCR assays ranged from 1 to 100 copies and was calculated using standard curves for 16 type-specific PCRs constructed with plasmids containing the entire genome of the different HPV types [15]. Real-time quantitative PCR for ß-globin was used to verify the quality of DNA in the sample and to measure the amount of input DNA [15,17].

Statistical analysis
Statistical analysis was performed with Chi-squared or Fisher exact test when appropriate for absolute numbers, and Student t test for normal distributed continuous variables.

Results
From 15th April to 1st November 2008, 64241 consecutive cervical cytology samples were processed in the laboratory. We excluded 966 (1.50%) follow up samples and 24 (0.04%) samples in which no human ß-globin DNA could be amplified.
When studying the correlation of LR and HR HPV and cytological abnormalities we found evidence of HPV infection in 7.8% in women without cytological abnormalities, and in 65.2%, 31.8%, 81.6%, 89.3% and 92.6% of women with ASC-US, AGC, LSIL, ASC-H and HSIL respectively. Both LR HPV and HR HPV were significantly associated with all cytologic abnormalities (ASCUS, ASC-H, AGC, LSIL and HSIL, p<0.0001, Table 1). TV infection was correlated with both LR HPV (OR 2.7, CL95% 1.5-5.0) and with HR HPV (OR1.7, CL95% 1.2-2.3). All women with TV and HSIL had HR HPV, while HR HPV was only present in 59% of women with TV and ASC-US.

Discussion
In this study, we confirm a correlation of TV with HPV and with cytological abnormalities of the cervix, indicating that STIs  are often coexist, even in a setting where the prevalence of TV is very low. This low prevalence makes the interpretation of the role of TV in the pathogenesis of cytologic abnormalities less evident in certain subgroups, despite high numbers of samples examined, a reliable diagnostic technique used (Real time PCR) and the completeness of the data obtained, which all add to the strength of the study. The correlation of LR HPV and HR HPV with different types of cytological abnormalities confirms expected HPV detection rates in cervical dysplastic lesions, as discussed in other studies. The prevalence of T. vaginalis is low in Flanders at 3.7 per 1000. Furthermore, we found that the age-dependent prevalence and epidemiology for HPV and TV were very different in this population, with a peak incidence of HPV infection at the age of 20-25, and of TV infection at 45-55 years [8], a finding repeated in several US centers [20,21]. Despite these differences, TV is clearly to be seen as an STI, as it is both correlated with HPV infection (both low and high risk) and with abnormal cytology of the cervix, a finding confirming evidence from other series [5,22,23]. As a consequence, some have speculated on the possible permissive role of TV in the pathogenesis of cervical intraepithelial lesions due to HPV [5,9].
With the present data, in a low prevalence setting, we could demonstrate that in all cases of HSIL where TV was discovered, HR HPV was also present as a causative agent. On the other hand, 36% (8/22) of the women with a combination of TV infection and ASC-US had no HPV. Although the prevalence of TV was higher in the HPV negative ASC-US group (1.3%) compared to the HPV positive (LR+ and HR+) ASC-US group (0.86%), this was not statistically significant. However since cytology reading was performed with knowledge of HPV status, presence of TV could still explain detection of atypical cells in the HPV negative ASC-US group. Also, in the group of HPV negative women, the infection rate of TV was 4 times higher in women with ASC-US than in cytological normal women. Furthermore, in HR HPV negative women, TV was never associated with HSIL or ASC-H, but was present in almost 1% of ASC-US and LSIL cases. These findings strongly suggest that TV can cause ASC-US like findings on cytology, without HR HPV or even any HPV present, and hence could lead to false positive diagnosis of ASC-US, enhancing the likelihood of overtreatment and unnecessary anxiety [24,25]. Other authors pointed out that TV can cause inadequate cytology readings, and pleaded for pretreatment of TV before performing cytology [26]. However, with the majority of TV infections being asymptomatic, and the accuracy of cytology in its detection being extremely poor and generally below 60% [27][28][29], the odds are high that presence of TV would continue to interfere with correct identification of low grade cellular atypia, especially ASC-US. Even taking into consideration an apparently better sensitivity for TV detection when liquid based media are used [30,31], it would be unlikely that cytology will enable TV detection in a sufficiently accurate way to prevent false positive readings, certainly not in high prevalence areas. Therefore we strongly suggest that real time PCR should be used to detect TV in the same sample used for cytology and HPV DNA detection. Also, the apparent relation of TV with HSIL (OR 3.2) is almost certainly due to co-infection with HR HPV in all cases, instead of being causative or even a co-factor for the development of HSIL. As only 1% of all HSIL were harboring TV, it seems very unlikely that these sporadic TV would have a significant role in the pathogenesis of HSIL. Still, if present together with HR HPV infection, TV infection was associated with a slightly higher rate of ASC-US+ abnormalities than when only HR HPV were present. As 4 out of 5 of these ASC-US+ lesions were low grade lesions (ASC-US and LSIL), we suppose the small increase in HSIL in combined TV-HR HPV patients may also be due to non-specific cellular aberrations caused by TV, as we found in one third of ASC-US in TV infected women without HPV infection.

Conclusion
We conclude that TV infection is associated with both low risk and high risk HPV infection of the cervix, as well as with the cervical cytological abnormalities ASC-US and HSIL. However, TV is unlikely to be involved in the causation or promotion of HSIL and cervical cancer. In women infected with TV, non-specific alterations of the epithelial cells may lead to a number of false positive diagnoses of ASC-US. In order to eradicate TV before making a final cytology diagnosis, we strongly advocate the use of real time PCR or systematic TV detection in liquid media before cytological examination.