Abstract LB381: Plasma microbiota as novel biomarkers of early epithelial ovarian cancer detection

Abstract Background: Ovarian cancer accounts for 2% of all female cancers in the United States yet it is the 5th leading cancer death among women. Advanced epithelial ovarian cancer (EOC) is associated with an overall survival rate of 30% but can be cured in up to 90% of cases if diagnosed early when the disease is still limited to the ovaries and fallopian tubes. While there is no effective screening method for EOC, developing novel and minimally invasive clinical tests is paramount to identify women with early disease as well as those who are at increased risk. Plasma-based microbiota signatures specific to EOC could become leading biomarkers of early disease in the field that complement existing clinical tools. Methods: We utilized an established institutional biobank to access 180 biobanked plasma samples from: (1) women with EOC; (2) women with non-EOC solid tumors; (3) women undergoing gynecologic-related procedures and non-EOC surgeries; and (4) healthy controls. Batch 1 included 96 plasma samples (n=24 per group) and Batch 2 included 84 samples (n=19 per group and 8 repeat EOC samples). Microbial profiles of plasma specimens were assessed using an amplicon-based sequencing method that targets the 16S rRNA gene in the bacterial genome to study bacterial phylogeny and taxonomy in diseased and healthy women. A stringent bioinformatic pipeline was applied to filter out suspicious bacterial contaminants at the genus level. Bacterial decontamination procedures in the pipeline included evaluating repeat plasma samples of EOC cases and sampling nucleic acid isolation kit reagents to serve as negative controls (only in Batch 2). Differential abundances of the plasma microbiota among the groups were assessed using analysis of compositions of microbiomes (ANCOM) as implemented in R software package ANCOM-BC. Results: The 16S rRNA gene sequencing identified, classified, and quantified 564 bacterial genera in the circulation of women diagnosed with EOC, non-EOC solid tumors, benign gynecologic conditions, and healthy controls. Bioinformatic decontamination analysis removed bacteria: (1) with differing bacterial abundances between batches; (2) with higher bacteria in negative controls; (3) with differing bacterial abundances in repeat samples; and (4) known as contaminants in the literature that resulted in the removal of 225 high-risk bacterial contaminants. A total of 339 high quality bacteria remained for evaluation. Because we did not collect negative controls in Batch 1, only Batch 2 plasma samples (n=84; 27 in EOC group; 19 in each other group) were included in the final analysis. Women with EOC, which were primarily of the high-grade serous ovarian carcinoma histologic subtype had significantly distinct differential abundances of genera Brevibacterium (p 0. 001), Chloronema (p 0. 001), Facklamia (p 0. 001), Zymomonas (p 0. 001), and Sutterella (p 0. 001) when compared to women with non-EOC solid tumors, benign gynecologic conditions, and healthy controls. Conclusions: These findings suggest that plasma EOC-associated bacteria could be exploited in the development of a microbial-based blood test for the diagnosis and early detection of EOC in routine practice settings. Citation Format: Diane E. Mahoney, Prabhakar Chalise, Dong Pei, Rachel Griffard, Trisha Home, Harsh Pathak, Shahid Umar, Andrew K. Godwin. Plasma microbiota as novel biomarkers of early epithelial ovarian cancer detection abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts) ; 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84 (7Suppl): Abstract nr LB381.