Abstract 3342: Congruence of tumor-specific ctDNA and radiographic recurrence in high-risk sarcoma

Abstract Background: Surveillance imaging remains the best method for detecting recurrence of high-risk sarcoma in patients off therapy. Whether tumor-specific ctDNA (circulating tumor DNA, Signatera) adds prognostic value to radiography remains unknown. We hypothesize that ctDNA conversion from negative to positive (mean tumor molecules per milliliter [MTM/ml]) will correlate with radiographic recurrence. Methods: We retrospectively assessed congruence of tumor-specific ctDNA and radiographs at serial timepoints in high-risk sarcoma patients 1) off therapy and on surveillance, with 2) scans showing no evidence of disease (NED) and a negative ctDNA +/- 28 days of those scans, and 3) ≥ 1 follow-up data point. Congruence was achieved if ctDNA and scans remained negative or if ctDNA conversion to positivity occurred prior to or ≤ 21 days post imaging recurrence. Results: 16 patients (9 male/7 female; median age 57yrs, Range 39-77) treated from 9/2021-9/2024 with bone or soft tissue sarcoma (STS) met study criteria. Histologies included 3 ewing sarcoma, 3 leiomyosarcoma (LMS), 3 liposarcoma (LPS), 1 each for osteosarcoma, sclerosing epithelioid fibrosarcoma (SEF), malignant pecoma, clear cell, undifferentiated pleomorphic, and other (2). 3 patients with metastatic disease (LMS, ewing, osteosarcoma) met study criteria. Prior therapies included surgery (14/16), radiation (7/16), and chemotherapy (4/16, 3/3 metastatic). Median follow-up time was 330.5 days (range 73-810) with 6 median assessments per patient (2-11). The median interval between the initial negative ctDNA and NED scan was 10 days (1-28) and 16 days (1-63) across all assessments. 10/16 patients (62.5%) remained NED on scans and 6 (37.5%) recurred, with ctDNA congruence in 9/10 (90%) and 5/6 (83.3%), respectively. 1/10 (10%) NED patients with myxoid LPS had a single positive ctDNA event (0.18 MTM/mL) with 2 subsequent negative events. All 3 metastatic patients recurred along with 1 SEF, 1 LMS, and 1 pleomorphic LPS. 5/6 positive ctDNA levels were detected 66 (pleomorphic LPS), 49, and 14 (both LMS) days before scans and 8 days post scans (ewing & SEF). An osteosarcoma patient with progression in pulmonary metastasis had 2 negative ctDNA levels before conversion (13.94 MTM/mL, 158 days) and required 2 lung biopsies prior to pathologic confirmation. Altogether, ctDNA was congruent in 14/16 (87.5%) cases, with sensitivity of 83.3% (35.9-99.6%) and specificity of 90% (55.5-99.8%). PPV and NPV were 83.3% (42.9-97.1%) and 90% (59.8-98.2%). Conclusions: Our findings show that tumor-specific ctDNA may be a useful prognostic sarcoma surveillance tool in high-risk bone and STS patients with local or metastatic disease. ctDNA positivity in 2 cases >6 weeks prior to scans suggests potential to detect early DNA shedding and the need to further examine temporal associations between ctDNA and scans in surveillance patients. Citation Format: Chahat Rana, Daniel Y. Wang, Amanda M. Dann, Nissy A. Alex, Sanjay Chandrasekaran. Congruence of tumor-specific ctDNA and radiographic recurrence in high-risk sarcoma. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3342.