The Vaginal Microbiota, Human Papillomavirus Infection, and Cervical Carcinogenesis: A Systematic Review in the Latina Population

Background Latina women experience disproportionately higher rates of HPV infection, persistence, and progression to cervical dysplasia and cancer compared to other racial–ethnic groups. This systematic review explores the relationship between the cervicovaginal microbiome and human papillomavirus infection, cervical dysplasia, and cervical cancer in Latinas. Methods The review abides by the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines. PubMed, EMBASE, and Scopus databases were searched from January 2000 through November 11, 2022. The review included observational studies reporting on the cervicovaginal microbiota in premenopausal Latina women with human papillomavirus infection, cervical dysplasia, and cervical cancer. Results Twenty-five articles were eligible for final inclusion (N = 131,183). Forty-two unique bacteria were reported in the cervicovaginal microbiome of Latinas. Seven bacteria: Lactobacillus crispatus, Lactobacillus iners, Chlamydia trachomatis, Prevotella spp., Prevotella amnii, Fusobacterium spp. and Sneathia spp. were enriched across multiple stages of cervical carcinogenesis in Latinas. Therefore, the total number of reported bacteria includes four bacteria associated with the healthy state, 16 bacteria enriched in human papillomavirus outcomes, 24 unique bacteria associated with abnormal cytology/dysplasia, and five bacteria associated with cervical cancer. Furthermore, three studies reported significantly higher alpha and beta diversity in Latinas with cervical dysplasia and cancer compared to controls. Lactobacillus depletion and an increased abundance of L. iners in Latinas compared to non-Latinas, regardless of human papillomavirus status or lesions, were observed. Conclusions The identification of 42 unique bacteria and their enrichment in cervical carcinogenesis can guide future cervicovaginal microbiome research to better inform cervical cancer prevention strategies in Latinas. Supplementary Information The online version contains supplementary material available at 10.1007/s44197-024-00201-z.


Introduction
Cervical cancer is the fourth most common cancer affecting women worldwide [1].Incident rates vary among high-and low-income countries due to public health efforts targeting this preventable disease.As a result of population-based cancer screening and human papillomavirus (HPV) vaccination programs, cervical cancer incidence and mortality rates have largely declined in high-income countries [1].However, in Latin America and the Caribbean (LAC), HPV prevalence and cervical cancer mortality are among the highest in the world [2].By 2025, 126,000 Latinas are predicted to be diagnosed with cervical cancer in LAC-a 75% rise in the frequency of the disease from 2002 [2].
In addition, Latinas have the highest cervical cancer incidence of any racial-ethnic group and are 24% more likely to die from cervical cancer compared to non-Hispanic White women [7].This disparity could be exacerbated by systemic barriers that prevent Latinas from receiving adequate healthcare services, including HPV vaccination, cervical cancer screenings, and health education [2,8,9].
The premalignant precursor of cervical cancer, cervical intraepithelial neoplasia (CIN), is caused by persistent infection with high-risk HPV genotypes [10].Although over 90% of HPV infections are cleared, reinfection of HPV can occur, and persistent HPV infection is linked to carcinogenesis [11].The optimal environment that promotes HPV clearance is not completely understood; however, evidence shows that cervicovaginal microbiota play a dual role in HPV clearance or persistence and cervical cancer development and progression [10,12,13].
Two recent systematic reviews in non-Hispanic White women have (1) established a causal link between vaginal dysbiosis and cervical carcinogenesis [12] and (2) assessed community state types in women with HPV infection, cervical dysplasia, and cervical cancer [22].However, an individual analysis of specific cervicovaginal bacteria in Latina women who are particularly high risk for adverse gynecologic sequelae has not been conducted.The cervicovaginal microbiota of Latinas must be further investigated, given the significantly high rates of vaginal dysbiosis, HPV infection, and cervical cancer in this population [2, 3, 6-8, 21, 23, 24].

Objectives
The purpose of the present review was to identify bacteria reported in the cervicovaginal microbiome of Latinas relating to HPV infection, cervical dysplasia, and cervical cancer as well as better understand the role of the microbiome in these disease states worldwide.These data could provide novel insights and approaches to address health disparities in HPV infection, persistence, and cervical cancer morbidity and mortality in this historically understudied, underrepresented, and underreported population of women.

Methods
This systematic review followed the PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analyses) guidelines and was registered in PROSPERO (CRD42022367244) on 11/21/2022 [25].No similar systematic review or protocol was registered.

Eligibility Criteria, Information Sources, Search Strategy
Studies in which participants were pre-menopausal women identifying as Hispanic or Latina according to the U.S Census Bureau definition: "a person of Cuban, Mexican, Puerto Rican, South or Central American, or other Spanish culture or origin regardless of race" were included in the systematic review [26].Furthermore, studies were only included if participants were diagnosed with either HPV, dysplasia, and/or cervical cancer or were healthy controls for studying these disease states.In addition, studies were required to be observational and describe analyses related to cervicovaginal microbiota.Interventional studies, article reviews, or commentaries were excluded.Studies evaluating pregnant people or non-human subjects were also excluded.
The following electronic databases were searched from January 2000 through November 11, 2022: PubMed, EMBASE (Elsevier), and Scopus.A search of ClinicalTrials.govwas not applicable because the studies for this review were not interventional.The search strategy employed a combination of terms related to "vagina/cervix" and "microbiota."This review used the Latinx/Hispanic US Population Search Hedge developed by the Medical Library Association Latinx Caucus for this literature search [27].The search was restricted to articles published after January 2000 on "female(s)" with no language restrictions.A more detailed report of the search strategy can be found in Supplementary Table 1.All records were exported into the bibliographic software EndNote X9 to remove duplicates.The unique records were then transferred to a web-based systematic review software, DistillerSR [28].

Study Selection
Titles and abstracts of retrieved articles were screened independently by two authors (V.M., N.R.J.).Any discrepancies were included for full-text screening to determine eligibility.The same two authors (V.M., N.R.J.) independently screened full-text articles for eligibility, with discrepancies resolved by a third author (M.M.H-K).

Data Extraction
Two authors (V.M., N.R.J.) independently performed data extraction.A standardized data collection form in Distill-erSR was created to collect: (1) participant characteristics (i.e., age, menopausal status, and gender identity), (2) geographic characteristics, (3) study characteristics (i.e., recruitment site, sample collection method, number of participants) (4) clinical characteristics (i.e., HPV genotypes, indication of abnormal Papanicolaou (Pap) smears, histological stage/ grade, and comorbidities), and (5) microbiome methodology (i.e., identification methods for the microbiome and HPV, community state types, Lactobacillus dominance/ depletion, alpha and beta diversity, and bacteria associated with outcome groups).Discrepancies in data collection were resolved after discussion and reviewing the full-text article.The compiled data from the full-text extraction are provided in Supplementary Table 2.

Assessment of Risk of Bias
The Risk of Bias in Non-randomized Studies of Exposure (ROBINS-E) tool was utilized to evaluate the risk of bias for observational studies [29].The ROBINS-E tool evaluates each study according to seven domains of bias: (1) due to confounding, (2) arising from measurement of the exposure, (3) in selection of participants into the study (or into the analysis), (4) due to post-exposure interventions, (5) due to missing data, (6) arising from measurement of the outcome, and (7) in selection of the reported result.The risk of bias for each domain and overall bias for each study was consolidated into a summary table; see Table 1.

Data Synthesis
A data collection form was developed and completed for each article in DistillerSR.Data was synthesized in a tabulated form by DistillerSR and downloaded as an Excel spreadsheet (Supplementary Table 2).

Study Selection
The initial search yielded a total of 365 articles from three databases: PubMed (178), EMBASE (168), and Scopus (19).One article was included due to a similar and relevant title during the full-text upload of documents.Twenty-five conference abstracts and 17 duplicates were removed, resulting in 324 unique articles.A title-abstract screening was performed on 324 articles, leading to the exclusion of 220 articles.One hundred four full-text articles were assessed for eligibility.Seventy-nine studies were excluded for not meeting the inclusion criteria.Nearly, half of the excluded articles (39/79) were ineligible because they lacked information on the cervicovaginal microbiome.This search method yielded a final sample of 25 full-text articles; see Fig. 1 for the PRISMA methodological flowchart.

Study Characteristics
The key characteristics of the 25 included observational studies are featured in Table 2. Some longitudinal studies were included but most studies were cross-sectional.The types of cross-sectional studies featured include descriptive, population-based, and prevalence studies.No qualitative studies were included.The studies were published between 2004 and 2022.The geographical regions of the studies encompass North [ID# 1, 4, 5, 7-9, 16-18, 20, 21, 23, 25]

Quality Assessment
The quality of the included studies was assessed by performing the ROBINS-E risk of bias for each article (Table 1).
Out of the seven domains assessed, the domain with the highest risk of bias was considered the study's overall risk of bias [29].The first domain, bias due to confounding, had the highest reported biases.For this domain, the authors developed a list of major risk factors influencing cervicovaginal microbiota composition and referred to confounding variables included in a similar systematic review during this process [12].As a result, five major confounding factors influencing cervicovaginal microbiota composition were included: age [30,31], parity [32], hormonal contraceptive use [33,34], smoking [35], and STIs, including human immunodeficiency virus (HIV) [36,37], herpes simplex virus (HSV) [38,39], chlamydia [40,41], gonorrhea [42,43], and trichomoniasis [44,45].Seven articles did not control or adjust for the listed confounding variables contributing to their overall "very high" risk of bias score (

Synthesis of Results
Eighteen studies reported on bacteria associated with outcome groups (  Approximately 42 unique bacteria in Latinas were associated with a healthy microbiome, HPV infection, abnormal cytology/dysplasia, and/or cervical cancer outcome groups.Some bacteria were enriched across multiple stages of cervical carcinogenesis in Latinas, which is summarized in a Venn diagram; see Fig. 3

Principal Findings
This systematic review summarizes the currently available literature reporting on cervicovaginal bacteria associated with HPV infection, cervical dysplasia, and cervical cancer in North, Central, and South American Latinas.Of 324 studies, 25 articles met our inclusion criteria, revealing 42 unique cervicovaginal bacteria associated with HPV infection, cervical dysplasia, and/or cervical cancer in Latinas.Other significant findings related to alpha and beta diversity and Lactobacillus dominance or depletion in Latinas were also reported.

Comparison with Existing Literature
In our review, two bacteria, L. crispatus and L. iners, were identified in two articles reporting on the healthy cervicovaginal microbiome of Latinas.The literature suggests that L. crispatus is associated with HPV-resistance and clearance which was confirmed in our review on Latinas [46].L. iners, on the other hand, is not associated with HPV clearance and was consistently reported in women with bacterial vaginosis in a longitudinal study [46,47].In addition, our review revealed that in some studies, L. crispatus was reportedly enriched in health and abnormal cytology/dysplasia, while L. iners was enriched in health, HPV/Hr-HPV infection, and abnormal cytology/dysplasia in Latinas.These findings are supported by a systematic review of 29 articles which reports a generally high relative abundance of L. iners and a lower relative abundance of L. crispatus in non-Latina Fig. 2 Cervicovaginal Bacteria Associated With Cervical Carcinogenesis in Latinas.Differences and similarities in the cervicovaginal microbiome composition from (1) healthy state to (2) HPV infection, (3) abnormal cytology/dysplasia, and (4) cervical cancer are depicted.Panel 1 is a light green color indicating an association with vaginal health.The following disease conditions proceed in a pink-to-red gradient according to severity: Panel 2 is light pink for HPV infec-tion, Panel 3 is dark pink for abnormal cytology/dysplasia, and Panel 4 is bright red for cervical cancer.Enrichment or depletion of bacterial taxa associated with each stage is indicated with an up or down arrow, respectively.A question mark denotes differences in reports with regard to bacterial enrichment or depletion within the same panel.Bolded bacteria were reported in more than one study individuals with precancerous lesions and cervical cancer [48].Although, the review did not define specific values for "high" and "low" abundances of Lactobacillus species.Multiple studies have linked L. iners enrichment to HPV infection [22,49,50] and even cervical cancer [22,51].In contrast to women with L. iners-dominant microbiomes, those with L. crispatus-dominant microbiomes have a more stable microenvironment and are more likely to clear HPV infection [13,22,52].
Sixteen bacteria were enriched in the cervicovaginal microbiome of Latinas with HPV/Hr-HPV infection across nine studies in our review.Across studies related to HPV/ Hr-HPV infection, C. trachomatis and P. amnii were significantly enriched while four bacteria were consistently enriched, although not significant: Sneathia spp., L. iners, Prevotella spp., and Fusobacterium spp..Among the listed bacteria, C. trachomatis, P. amnii, and Prevotella spp.were enriched across eight studies in our review reporting on Latinas with HPV/Hr-HPV infection and abnormal cytology/dysplasia.C. trachomatis infection has been previously associated with HPV and Hr-HPV infection and persistence [53].In addition, HPV/C.trachomatis co-infection has been identified as a significant risk factor for cervical cytological abnormalities [54][55][56][57].Notably, a study based in Morocco revealed that women coinfected with HPV/C.trachomatis were three times more at risk of developing cervical abnormalities [57].Previous studies in non-Latina cohorts reveal an increased abundance of Prevotella spp.has been associated with HPV infection, persistent Hr-HPV infection, and cervical lesions compared to control groups [58][59][60].Moreover, Prevotella spp.have been associated with persistent CIN lesions and slower regression of disease [52].
Despite limited reports on P. amnii, it was the most abundant bacteria in women with low-grade SILs in a Chinese study cohort [60].
Abnormal cytology or dysplasia, such as abnormal Pap smears, SILs, cervical dysplasia, and CIN in Latinas, yielded associations with 24 unique bacteria.Across multiple studies in this category, C. trachomatis was significantly enriched, and seven additional bacteria were consistently enriched although not significant: Fannyhessea spp., G. vaginalis, L. crispatus, L. iners, Prevotella spp., P. amnii, and Sneathia spp..In our review, some conditions included depleted bacteria, while most reported enriched bacteria.For example, L. iners and L. crispatus were depleted in women with SILs, and Fannyhessea spp.were depleted in women with CIN.An enrichment of G. vaginalis was observed across three conditions: abnormal Pap smears, cervical dysplasia, and CIN.An abundance of G. vaginalis has been frequently reported in the vaginal microbiome of women with HPV and Hr-HPV infection and is also associated with HPV persistence [61,62].A study revealed that coinfections with HPV and G. vaginalis increased risk for SILs and cervical cancer [61].This may be due to the ability of G. vaginalis and other BVassociated bacteria to secrete sialidase, which is an enzyme that has been linked to an increased risk of cervical lesions [63].In our review, L. iners, L. crispatus, and Fannyhessea spp.had conflicting relative abundance reports in women with cervical dysplasia.Existing literature suggests that Fannyhessea spp.are frequently detected in women with highgrade CIN [64,65] and cervical cancer [64][65][66] in Chinese, South Korean, and Slovak cohorts.
Our literature review revealed that cervical cancer includes the enrichment of four bacteria and the depletion of  2) one in Latinas.The depleted bacterium in Latinas with cervical cancer was L. crispatus, which is generally associated with a healthy cervicovaginal microbiome [46].Fusobacterium and Sneathia spp.were enriched across three studies in our review on cervical cancer in Latinas.Fusobacterium spp.were enriched in Latinas with HPV/Hr-HPV infection and cervical cancer.Existing evidence demonstrates that Fusobacterium spp.abundance is strongly associated with HPV infection and high-grade dysplasia [22,52,58].Moreover, after examining 112 cervical cancer tumor tissues, Huang et al. (2020) suggest Fusobacterium spp.may be a potential diagnostic and prognostic biomarker for cervical cancer [67].Lastly, Sneathia spp.were enriched across HPV/Hr-HPV infection, abnormal cytology/dysplasia, and cervical cancer in Latinas in two included articles.study considered Sneathia spp. a microbiological marker of Hr-HPV infection, given that Sneathia spp.were detected three times more frequently in women with Hr-HPV infection [52].Moreover, previous studies on non-Latinas reported a significantly greater abundance of Sneathia spp. in women with cervical intraepithelial lesions, high-grade SILs, and cervical cancer compared to healthy controls [68,69].Furthermore, our lab demonstrated that Sneathia spp.exhibit potential oncogenic mechanisms based on the altered immunometabolic microenvironment in a 3D model of the human cervix [70,71].Although we were unable to run a meta-analysis, Sneathia spp.are microbes of interest and require further investigation in Latina cohorts and longitudinal studies considering they are enriched in all stages of cervical carcinogenesis.
Other significant findings related to alpha and beta diversity in Latinas were also reported.Two studies in our review reported higher species richness and diversity in Latinas with high-grade cervical intraepithelial neoplasia (CIN3) compared to low-grade cervical intraepithelial neoplasia (CIN1) and Latinas with SILs compared to non-SILs.Existing literature reports higher alpha diversity among women with abnormal cervical pathology, with a trend to increase the more severe the cervical lesion [72].In addition, a previous study reported higher alpha diversity among women with cervical cancer, which was confirmed in our review on Latinas [73].Our review also revealed a significantly higher beta diversity in women with SILs compared to non-SILs and women with cervical cancer compared to those with HPV infection and cervical dysplasia.Significant differences in beta diversity among non-Latina women with HPV infection, SILs, and cervical cancer compared to controls were also reported in a systematic review [74].Overall, our review revealed a trend toward increased alpha and beta diversity and cervical carcinogenesis progression in Latinas.
Regarding Lactobacillus dominance, two studies in our review reported L. iners, a transitional bacterium, as the most abundant among Latinas [47].Lastly, two studies in our review revealed significant depletion of Lactobacillus species among Latinas compared to non-Latinas.Previously, Ravel et al. (2011) demonstrated a racial-ethnic difference in microbiome composition as Black and Latina women harbor increased levels of diverse anaerobes and lower levels of health-associated, Lactobacillus species compared to White and Asian women who possess Lactobacillus-dominant vaginal microbiomes [75,76].In our review, we observed similar overall low levels of Lactobacillus and increased bacterial diversity across Latinas globally.
Our review featured studies from ten countries across North, Central, and South America and the Caribbean.Another review of women from Latin America and the Caribbean (LAC) revealed that early age at first sexual intercourse, number of sexual partners, and sexual behavior of the partner are associated with increased risk of genital HPV acquisition [77].Additional co-factors for cervical cancer in LAC include high parity, long-term use of oral contraceptives, high prevalence of smoking, and co-infection with the human immunodeficiency virus (HIV) [77].In addition to these risk factors, women in LAC face systematic barriers to healthcare such as uneven resource distribution, variable infrastructure, and healthcare service availability [77,78].Rural, low-resourced, and underserved populations in LAC are especially impacted and are less likely to have access to HPV vaccination, cervical cancer prevention, and screening [52].Studies along the U.S.-Mexican border also demonstrate the need for increased health education and awareness regarding the HPV vaccine and HPV/cervical disease diagnoses to prevent cervical cancer with screening, early diagnosis, and treatment [9,79].

Strengths and Limitations
To our knowledge, this is the first systematic review to report on the cervicovaginal bacteria in studies related to HPV infection, cervical dysplasia, and cancer in Latinas.The strengths of this review include strict criteria for inclusion, yielding only articles relevant to our research question.While our inclusion criteria were strict, our search generated a broad range of articles focused on our population of interest.By including the extensive Latinx search hedge, we were able to account for the racial-ethnic, geographic, and linguistic range the individual "Hispanic" search term lacks [27].
The limitations are those inherent to a systematic review, which includes relying on the information and quality of data available.Consequently, the varying study designs and general lack of homogeneity of reported analyses (e.g., few and dissimilar statistical tests, differing variables, etc.) failed to meet the basic criteria for unbiased meta-analytic methods.In an effort to update the review and reassess the potential for a meta-analysis, the search could be re-conducted after a few years following publication of additional studies.With more data available, significant associations of specific bacteria can be determined in Latinas with HPV, cervical dysplasia, and cervical cancer.Second, some studies adjusted for confounding factors and excluded individuals with STIs (i.e., C. trachomatis), while other studies included all patients, even those with STIs.This leads to a discrepancy in the inclusion or exclusion of individuals with C. trachomatis in the 25 included studies.As such, the C. trachomatis findings must be interpreted with caution.Hence, controlling for confounding variables involves balancing the need to include all individuals with the demand of strict bias criteria.Lastly, in order to establish a causal connection between the enriched bacteria and cervical carcinogenesis, longitudinal microbiome studies must be performed [80].Longitudinal microbiome studies and larger cohort studies, including Latinas will help determine the role of bacteria as drivers (influential disease-causing agents), passengers (less influential agents favoring the environment), or a consequence of disease in this population of women [80].

Conclusions and Implications
The systematic review identified 42 unique bacteria across 25 studies related to HPV infection, cervical dysplasia, and cervical cancer in Latinas.L. crispatus was enriched in healthy Latinas and those with abnormal cytology/dysplasia, while L. iners was enriched in multiple states ranging from healthy to abnormal cytology/dysplasia groupings.C. trachomatis, P. amnii, and Prevotella spp.were enriched in Latinas with HPV/Hr-HPV infection and abnormal cytology/dysplasia.Notably, Fusobacterium spp.were enriched in Latinas with HPV/Hr-HPV infection and cervical cancer.Importantly, Sneathia spp.were enriched across all stages in cervical cancer progression-HPV/Hr-HPV infection, abnormal cytology/dysplasia, and cervical cancer.In addition to more research on Latina populations, barriers related to social determinants of health and structural racism must be considered to improve outcomes against a preventable disease like cervical cancer amongst Latinas.Advanced public health efforts, including community-based participatory research projects with Latinas, can reduce health disparities in HPV infection and cervical cancer [81,82].In conclusion, future epidemiological studies must intentionally include Latina women in order to ultimately create primary or secondary preventative strategies for this susceptible population.

Fig. 1
Fig. 1 PRISMA Methodological Flowchart.The flowchart depicts the identification, selection, and final inclusion of articles.Records of excluded articles and reasons for exclusion are also included.a The total number of full-text articles excluded was n = 79.However, arti-

Fig. 3
Fig.3 Cervicovaginal Bacteria Enriched Across Stages of Cervical Carcinogenesis in Latinas.The quadruple Venn diagram highlights seven bacteria enriched across outcome groups, including healthy controls.Sneathia spp.and L. iners were associated with multiple outcome groups across cervical carcinogenesis.The numbers listed after each bacterial species correspond to the article reporting on the enrichment of the bacteria (reference Table 2)

Table 2
Study Characteristics and Key Findings of the Included Articles

Table 2 (
The data extraction table reports study authors and publication year, number of participants, number of participants in the outcome and control group, the age range of participants, the location of the study, the types of outcomes studied: HPV, abnormal cytology or dysplasia, and cervical cancer, sample collection method, bacteria identified in outcome or control groups, and other microbiome metrics, such as Lactobacillus