Objectives: Current screening methods for ovarian cancer have failed to demonstrate a significant reduction in mortality. Uterine lavage catheters have been used to detect tumor-specific TP53 mutations from cells presumably shed from high-grade serous ovarian cancer, but this technique may not identify non-serous subtypes or early-stage disease. We aimed to pilot the combination of deep sequencing methods with an expanded gene panel to improve detection of both early stage and non-serous ovarian cancers. Methods: Lavage of the uterine cavity was performed in 35 consecutive patients undergoing surgery with preoperative concern for an ovarian malignancy. Duplex sequencing, an ultra-accurate error-correction sequencing approach, was used to deeply sequence extracted DNA from lavage samples (average duplex depth ~2500x) with a panel of candidate ovarian cancer driver genes including TP53, ARID1A, PTEN, PPP2R1A, CDKN2A, KRAS (whole genes), CTNNB1, PIK3CA, and BRAF (hotspots only). Tumor DNA was sequenced to identify driver mutations and compare with mutations found in the lavages. The overall mutation frequency in lavage DNA was calculated by dividing identified non-polymorphism mutant alleles by the total number of nucleotides sequenced in coding regions. Results: In total, lavage samples were collected from fourteen women with benign disease, thirteen with high grade serous carcinoma (HGSC), three with clear cell carcinoma, three with endometrioid carcinoma, one with granulosa cell carcinoma and one with carcinosarcoma. Thirteen women had stage I or II disease (including five with stage I or II high grade serous). Processed lavage samples yielded a median DNA of 596.5 ng. Tumor sequencing is ongoing, but of seven fully sequenced lavage/tumor pairs, the tumor-specific mutation was identified in four lavage samples: TP53 mutations found in two HGSC (stage IIb and stage III), an ARID1A mutation from a stage Ic3 clear cell carcinoma, and a PIK3CA mutation from a stage Ia endometrioid carcinoma. The tumor-specific mutation was not identified in lavage samples from two patients with endometrioid and one with clear cell carcinoma. Of 21 lavage samples that have presently undergone duplex sequencing, a total of 596 additional somatic mutations were identified in the nine genes. Lavages from patients with HGSC tended to have increased average mutation frequency of TP53 compared to patients with benign disease (TP53 2.2x10-6 vs 9.2x10-7, p=0.09). Conclusion: Sequencing at high depth and using an expanded gene panel allows for the identification of tumor mutations in uterine lavage samples, including some early stage and non-serous ovarian cancers, as well as identifying significant somatic mutational background. Larger studies are needed to confirm the clinical utility of this method, which may have potential to improve the early detection of ovarian cancer.

Citation Format: Talayeh Ghezelayagh, Jeanne Fredrickson, Enna Manhardt, Marc R. Radke, Brendan Kohrn, Heidi J. Gray, Renata R. Urban, Kathryn P. Pennington, John B. Liao, Kemi M. Doll, Elise J. Simons, Jennifer K. Burzawa, Barbara A. Goff, Paul Speiser, Elizabeth M. Swisher, Rosa Ana Risques, Barbara M. Norquist. Uterine lavage for the detection of ovarian cancer using an expanded gene panel [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr LB047.