Co-infected with trichomonas vaginitis increases the risk of CIN2-3 among HPV16 positive female: a large population-based study CURRENT STATUS:

Background: Evidences suggested that the vaginal microbiome played a functional role in the progression of cervical lesions in female infected by HPV. This study aimed at evaluating the influence of common vaginal infection on the carcinogenicity of hr-HPV. Methods: From January 15, 2017 to December 31, 2017, 310,545 female aged at least 30 years old had been recruited for cervical cancer screening from 9 clinical research centers in Central China. All the recruited participants received cervical cancer screening and vaginal microenvironment test by a high vaginal swab. Colposcopy-directed biopsy was recommended for female who were infected with HPV 16 and HPV 18 and other positive hr-HPV types through test had undertaken triage using liquid-based cytology, cases with the results ≥ ASCUS among them were referred to colposcopy directly. Results: Among 310,545 female, 6,067 (1.95%) were tested with positive HPV 16 and HPV 18, 18,297 (5.89%) were tested with other positive hr-HPV genotypes, cervical intraepithelial neoplasia (CIN) 1, CIN 2, CIN 3 and invasive cervical cancer (ICC) were detected in 861 cases, 377 cases, 423 cases and 77 cases, respectively. Candida albicans and Gardnerella were not associated with the progression of cervical lesions. Positive trichomonas vaginitis (TV) was correlated with hr-HPV infection (p<0.0001). Co-infection with TV increased the risk CIN 1 among female infected hr-HPV (OR with TV increased the among female infected with HPV 16 (OR 95% CI: Conclusions: Co-infection of TV and HPV 16 is an important risk factor for the progression of cervical lesions. of the results. These characteristics of the participants were presented as proportions. Correlations between specific vaginal infection and hr-HPV genotype and CIN/ICC were assessed using stepwise logistic regression after adjusting for all potential risk factors. ORs and 95% CIs were calculated to analyze the correlation between possible risk factors and the pathogenic infection. Data was analyzed with SPSS

Under normal conditions, vaginal flora is composed by more than 20 kinds of microbes [2,3] . Infection of female reproductive tract refers to a series of infectious inflammation caused by the destruction of the defense system by microorganisms, such as viruses and bacteria [4] . The genital tract infection leads to an imbalance of the vaginal flora, which inhibits or reduces the lactobacillus substantially,which may cause the decline of the clearance of hr-HPV. Candida spp, Gardnerella and TV are the most common vaginal infections. As a part of the human commensal flora, Candida spp always causes systemic and superficial infections [5] . Gardnerella vaginalis is considered as playing a key role in the pathogenesis of bacterial vaginitis (BV) [6] , and BV shows correlation with severity of cervical neoplasia in HPV-positive female [7] , but the aetiology and pathogenesis of BV are more complex, gardnerella vaginalis is thought as a potential founder organism. As a sexually transmitted infectious agent, TV is found to cause local inflammation, and it affects the clearance of hr-HPV and contributes to cervical lesion progression in several research, but it is controversial [8] . Some studies have demonstrated an correlation between vaginal infection and the carcinogenicity of hr-HPV, but most of them are small-scale researches and their conclusions are inconsistent [5,[9][10][11][12][13][14][15][16] . This research is with the largest scale and the most comprehensive investigation about the correlation between cervical hr-HPV infection, CIN/ICC and the vaginal microbiome.
Central China is one of the regions with the highest incidence of cervical cancer [17] , and the socialeconomic conditions vary greatly in different districts. In order to study ICC influenced by multiple socioeconomic factors, nine areas in Central China with significantly different economic levels were included in this study. This study aimed at evaluating the influence of common vaginal infection on the carcinogenicity of hr-HPV.

Methods
This project had been conducted in Central China from January 15, 2017 to December 31, 2017.
Female aged at least 30 years old were recruited to undergo the cervical cancer screening from 9 clinical research centers (Xiangzhou, Fancheng, Xiangcheng, Baokang, Nanzhang, Zaoyang, Yicheng, Gucheng and Laohekou) in Central China via media promotion and government notices. Female who received hysterectomy, were pregnant, without sexual history had been excluded. All the female had not been vaccinated against cervical cancer yet. All the participants received questionnaires, hr-HPV genotyping and vaginal microbiota examination. The screening process was based on the interim clinical guidance of ASCCP in 2015 [18] , female tested positive for HPV 16 and HPV 18 had been referred to colposcopy directly; female with other positive hr-HPV types through test had undertaken triage using liquid-based cytology, cases with the results ≥ ASCUS among them were referred to colposcopy directly. The protocol was approved by the Ethics Committee of Xiangyang Central Hospital.

Questionnaires
All the recruited participants were interviewed by a trained interviewer and filled out a questionnaire for the first time. The content of the questionnaire includes age, marital status, ethnicity, the highest level of education, whether in menopause, date of last menstruation, history of past HPV infection, family history of cancer, reproductive history (number of pregnancies and number of births), method of contraception (contraceptive, condom and intrauterine contraceptive device), number of lifetime sex partners and whether in poverty (whether being a poverty alleviation target).

HPV genotyping
All the recruited female underwent a gynecological examination. A cervical specimen was taken using a cervical brush, and a high vaginal swab was collected. COBAS4800 (Roche Molecular Systems, Alameda, CA) assay was used for typing HPV DNA. The COBAS 4800 HPV test detected a total of 14 hr-HPV types simultaneously: HPV-16 individually, HPV-18 individually, and pooled hr-HPV genotypes other than HPV 16 and 18 (31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68), in addition to a separate high b-globin control.

Vaginal microenvironment test
The high vaginal swab was processed for microscopic evaluation of vaginal microenvironment, including presence of Candida albicans, TV, gardnerella, clue cells and miscellaneous bacteria (vaginal bacteria except lactobacilli) density [19] . A vaginal wet mount was prepared for detection of Candida albicans. Candida was diagnosed using the KOH method (10% KOH), and vaginal trichomonas was examined according to conventional methods, the swimming trichomonas was observed under the microscope (x400). Gardnerella, clue cells, and miscellaneous bacteria density(MBD) were observed by Gram staining. The scoring method of miscellaneous bacteria density was listed as follows, no miscellaneous bacteria was scored as Ⅰ, 1-5 miscellaneous bacteria/OML as Ⅱ, 6-30 miscellaneous bacteria/OML as Ⅲ, 30 miscellaneous bacteria/OML as Ⅳ.

Thinprep cytologic test
According to manufacturer's instructions, slides for liquid-based cytology were prepared. Bethesda System 2001 terminology was used for reporting the results [20] . The following 5 different categories

Statistical analysis
Histological results (normal, CIN1, CIN2, CIN3 and ICC) were distributed among respective hr-HPV types and then evaluated. Univariate analysis was used for comparing whether the risk of any cervical lesions (tri categorical: CIN1, CIN2-3 and ICC) differed when female having vaginal infection (Candida albicans, Gardnerella and TV) or not. Odds ratios and 95% confidence intervals (CIs) were calculated to study the specific risk among different HPV types. Next, social-demographic and reproductive characteristics were taken into consideration so as to improve the accuracy of the results. These

Correlation between Candida Albicans, Gardnerella, TV and the progression of cervical lesions stratified by different hr-HPV types
Distribution of potential risk factors among CIN/ICC and normal participants is shown in S- Table 1.

Univariate analysis of the correlation between different vaginal infection (Candida albicans,
Gardnerella, TV micro-environment) and the histological results (three categories: CIN1, CIN2-3 and ICC) among different hr-HPV types is displayed in Fig. 2

Potential Risk Factors of hr-HPV mono-infection or co-infection with TV
To find the risk factors of hr-HPV infection, multivariate logistic regression analyses were utilized. The results are shown in Fig. 3, it was found that age ≥ 60 (p < 0.001), number of pregnancies ≥ 1 (p < Baseline features of the participants with hr-HPV and TV co-infection are shown in Table 3, primary education, intrauterine contraceptive device as a method of contraception and female with multiple sex partners were more likely to be infected with both TV and hr-HPV.

Discussions
According to the report "Global Cancer Statistics 2018", cervical cancer ranks the fourth in the incidence and mortality of female tumors worldwide, and the second in the incidence of female tumors in developing countries [1] . HPV infection is a necessary but insufficient condition for cervical cancer [21] , the number of cervical cancer patients is far lower than the number of HPV infection female [22] . With the development of microbiome, more and more studies have shown that the composition and changes of cervical-vaginal microenvironment are closely related to HPV invasion, persistent infection and the occurrence and development of cervical cancer [23,24] .
In our study, the prevalence of hr-HPV in this region was lower compared with the hr-HPV prevalence from other parts of China (from 9.03-16.8%) [17] . It was found that Candida spp was not associated with the progression of cervical lesions in this study, which was consistent with previous studies [5,13,14] . Gardnerella vaginalis was not the risk factors of CIN or ICC in our study, either. Gardnerella vaginalis was considered to possess characteristics which are important for the pathogenesis of BV [6,[25][26][27] , including production of sialidase [27] and vaginolysin, [26,28] . Although some research found that BV played a functional role in female with CIN or ICC [2,3] , the correlation between Gardnerella and the progression of cervical lesions had yet to be described. In this study, Gardnerella has had a low prevalence and was only detected in 1,050 (0.34%) female, and co-infected with Gardnerella didn't increase the risk of CIN or ICC. In the vagina of female without BV, Gardnerella vaginalis was also detected, although it had a lower abundance and prevalence [29,30] . There were enormous substantial genetic diversity existing within Gardnerella vaginalis [16,31] and virulence potential differentiate greatly between various genetic types/clades [15] , this may correlate with the result in our study.
Several researches had studied the relationship between TV, HPV and CIN/ICC, but their conclusions were inconsistent. Gweneth's study found that trichomonas vaginitis was associated with hr-HPV infection (specifically type 16) significantly, but no correlation between trichomonas and cervical lesions [10] . Ishita Ghosh's study found that the higher risk of cervical cancer observed in the female co-infected with HPV and TV was without any enhanced risk of CIN [5] . But both of them were smallscale studies. Rui-Mei Feng's pooled analysis found that current TV-positive female had an increased risk for hr-HPV infection compared with currently TV negative female. Both past and current TV-positive female had a decreased risk for CIN 2+ [9] , but hr-HPV was detected by Hybrid Capture 2 and HPV genotype can't be differentiated, TV was diagnosed by thinprep cytologic test in that research, and this method had a lower detection rate.
In this study, 5,683 among the overall participants were TV positive and 1,952 (34.35%) among these, were co-infected with hr-HPV. Co-infection with TV increased the risk of CIN 1 among female with hr-HPV, and increased the risk of CIN 2-3 among female with HPV 16. It was expected that TV was often associated with more severe inflammatory reaction and HPV persistence, and was likely closely related to the invasion, persistent infection of HPV 16. HPV including hr-HPV infections are mostly temporary, and the virus can be cleared through host immune responses spontaneously, only a small number of infections persisted and progressed to cervical cancer [32] . HPV 16 was found to be the most prevalent hr-HPV type with the highest risk among all hr-HPVs to progress to CIN2-3 and ICC [33] . Studies found that the outcome of HPV infection was closely related to the local microenvironmental of the cervix [23] . A lot of research have discovered that TV led to an imbalance of the vaginal flora, which inhibited or reduced the lactobacillus substantially [5] , and caused severe inflammation. Evidences were found that the continuous activation of inflammatory transcription factors can lead to cervical tissue damage, and thus improved the susceptibility and carcinogenic ability of HPVl6 [10,34] . Meanwhile, the severity of cervical lesions was correlated with the abundance and diversity of cervicovaginal flora positively and correlated with the number of lactobacilli negatively [23,35] . Unlike CIN 1, CIN 2-3 had been considered no longer reversible and required a more definite clinical treatment, the composition and characteristics of the cervical and vaginal flora were very different [36] . [2,3] . There may be some complex host immune response to TV and HPV 16 co-infection, but the hypothesis needed further research for validation. There were only 72 ICC in this study, so the correlation between TV, HPV and ICC were inconclusive.
Both TV and genital tract HPV infection are sexually transmitted disease (STD). In the districts with lower economic levels, many people went out to the developed areas to work and couples usually work in different places, and sex act extramarital is common [17] . Risk factors including primary education, intrauterine contraceptive device as a method of contraception and female with multiple sex partners significantly increased the odds of hr-HPV and TV co-infection. Compared with female with only 1 sexual partner, it was found that female with 2 or more lifetime sex partners had a higher risk of HPV and TV co-infection. Meanwhile, it was found that female who chose intrauterine contraceptive device as a method of contraception had a higher risk of co-infection, as compared with female who chose contraceptive or condom. Female who had primary education also were found to have higher risk of co-infection compared with higher educated female. As TV is a STD, and it was found that 2 or more lifetime sex partners and choosing intrauterine contraceptive device instead of condom were associated with co-infection, it was expected that the co-infection is likely closely related to the risky sex behaviors. Sexual interactions with multiple hosts were expected to contribute to the co-infection, because each host may be infected with one kind of special pathogen species [37,38] . The transmission and clinical progression of the sexually transmitted infectious diseases may be changed by sexual interactions [39,40] . Risk factors, including number of pregnancies ≥ 1, postmenopause, poverty, personal HPV history and miscellaneous bacteria density were associated with HPV infection significantly, but they were not related to co-infection significantly.
As it was known that, this research had the largest number of female who received COBAS human papillomavirus primary testing for cervical cancer screening up to now worldwide, and it was an multisite investigation with the largest scale about the correlation between cervical hr-HPV infection, CIN/ICC and vaginal microenvironment. The detection methods of hr-HPV and vaginal microenvironment in this study were highly sensitive and recognized internationally.
Of note, there were some limitations to our study. Because our study was a cross-sectional analysis of current hr-HPV and TV infections, it was not clear whether there was a persistent HPV infection among these female with HPV or not, this study will continue to following up on female with hr-HPV female and attempting to analyze the time and rates of HPV clearance. In addition, there are some participants who failed to be followed up for a variety of reasons in the screening process.
In one word, from this large population-based study, it was found that co-infection of TV and HPV 16 is an important risk factor for the progression of cervical lesions.

Ethics approval and consent to participate
This study received the verification of Medicine Ethics Committee Xiangyang Central Informed consent to participate in this study had been obtained from participants.

Consent to publish
Not applicable.

Availability of data and materials
All data generated or analysed during this study are included in this published article and its supplementary information files.

Competing interests
All authors declare that they have no competing interests. All authors have submitted the ICMJE Form

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download. SupplementaryBMC.pdf