Abstract C017: Race-Specific Patterns of Vaginal Microbiome and HPV Infections Predict Risk of Precancerous Cervical Lesions

Abstract

Abstract Background: Our previous research indicated that high-grade squamous intraepithelial lesions (HSIL), precursors to cervical cancer (CCa), are influenced by the vaginal microbiome (VMB) differentially by self-reported race. VMB enriched for certain anti-inflammatory Lactobacillus species confer protection against the risk of HSIL for white, but not Black women. Black women exhibit greater VMB diversity (associated with HPV tumorigenesis) and a lower prevalence of anti-inflammatory Lactobacillus species. But not all Black women with diverse VMB develop HSIL. Objective: Given HPV's essential role in cervical cancer, we aimed to assess the influence of sociodemographic, clinical, and microecological HPV attributes on HSIL risk within the context of the vaginal microbiome's variation by race. Methods: The VMB and HPV profiles of 1,168 women were analyzed alongside over 100 sociodemographic and clinical variables. Spearman’s correlation, Point-Biserial Correlation, chi-squared tests, and machine learning models, were employed to evaluate associations between sociodemographic, clinical, and microecological variables with HSIL, with a particular focus on the impact of race on these relationships. Results: The cohort was predominantly Black (72%), showing no significant race-based differences in HPV positivity (45%). However, HSIL incidence was nearly twice as high in Black (7%) compared to White women (4%). Women with HSIL had a higher prevalence of non-Lactobacillus-dominant VMB (68% vs. 51%) and were more likely to be HPV+ (73% vs. 44%) than those without HSIL. Co-infection with multiple HPV strains was associated with a higher risk of CIN3. Specifically, HPV16, HPV56, HPV58, HPV33, and HPV42 infections were significantly associated with higher CIN3 risks. The random forest model identified the relative abundance of Lactobacillus crispatus as the most essential factor in predicting CIN3 risk, followed by the number of HPV strains per sample. However, differential abundance analysis specifically indicated lower L. crispatus levels were associated with HSIL in white but not Black women. Conversely, HPV abundance was only associated with HSIL in Black women. Additionally, Black patients with 2 or more HPV co-infections had a higher likelihood of being diagnosed with CIN3+ compared to white counterparts. ML models highlighted different HPV-specific HSIL predictors by race: HPV number, and HPV 56, 16, and 33 in Blacks; sexual partners, HPV 90, 56, and 16 in whites. Conclusions: Our findings reveal race-specific interplays among individual risk factors, HPV infection, and VMB composition in HSIL development. The identification of key microbial and viral predictors highlights the importance of personalized approaches in preventing and managing HSIL, particularly among racially diverse populations. Citation Format: Katherine Y. Tossas, Bin Zhu, Katarzyna Tyc, Myrna Serrano, Jerome Strauss, Gregory Buck. Race-Specific Patterns of Vaginal Microbiome and HPV Infections Predict Risk of Precancerous Cervical Lesions [abstract]. In: Proceedings of the 17th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2024 Sep 21-24; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2024;33(9 Suppl):Abstract nr C017.