Potential of tumor-informed ctDNA as an early predictive indicator for relapse in advanced ovarian cancer.

Abstract

5574 Background: Prediction of relapse following firstline treatment in patients (pts) with high-grade serous ovarian cancer (HGSOC) remains a major challenge despite recent advances. Reliable markers for assessment of recurrence risk are urgently needed to tailor treatment strategies with circulating tumor DNA (ctDNA) emerging as a promising candidate. Methods: In this prospective feasibility study, pts with advanced HGSOC who underwent primary surgical and systemic treatment at two large-volume centers for gynecologic oncology were evaluated between July 2021 and September 2024. Whole genome sequencing was used to develop a personalized multiplex digital polymerase chain reaction fingerprint assay by identifying structural variants, single nucleotide variants and indels in FFPE tumor tissue. Longitudinal blood samples were collected perioperatively (preop, postop day 2 and 10), during firstline chemotherapy (cycle 1, 3 and 6 [c6]) and follow up. CA-125 levels were tested accordingly. For statistical analyses, chi squared, log rank tests and Kaplan-Meier method for PFS were applied as appropriate. Results: As of 21st January, 2025, a total of 31 pts have available samples from preop through c6 with completed ctDNA data. In this cohort, 11 recurrences (35%) have been diagnosed at a median clinical follow-up of 16.8 months [mo] (range 5.7-38.4 mo), median progression-free survival (PFS) was 11.8 mo (range 5.7-22.9 mo). At c6, levels of CA-125 were <35 kU/L in 25 (81%) and ≥35 kU/L in 6 (19%) of the 31 pts. Clearance of ctDNA was noted for 19 out of 31 pts (61%). 16 of these 19 pts (84%) had previous complete cytoreduction. While rates for recurrence did not align with CA-125 levels <35 kU/L (63.6%) and ≥35 kU/L (36.4%, P =0.075) at c6, a significantly lower recurrence rate was observed for patients with ctDNA clearance at c6 (4 of 19, 21.1%) compared to 7 of 12 patients with persistent ctDNA (58.3%, P= 0.034). In 21 patients with complete cytoreduction, five pts still had detectable ctDNA levels at c6. Three of these five pts had recurrence (60%), compared to two of 16 pts with ctDNA clearance (12.5%, P =0.023). Detection of residual ctDNA at c6 was strongly associated with an increased risk for recurrence in the overall cohort compared to pts with ctDNA clearance (HR: 5.78, 95%CI: 1.93 – 31.99, P =0.004). This effect appears to be more pronounced in pts with macroscopic complete cytoreduction, but was not seen in pts with residual tumor. Conclusions: Findings of this interim analysis underline the potential of tumor-informed ctDNA as a powerful tool for recurrence risk assessment in pts undergoing primary treatment for HGSOC. In contrast to CA-125, ctDNA evaluation at the time of completed firstline chemotherapy might serve as an early predictive marker for relapse. This information could help to develop patient-specific treatment strategies, especially in the subgroup of pts with complete macroscopic cytoreduction.