Prognostic value of circulating tumor DNA testing in surgically resected pancreatic cancer.

Abstract

4134 Background: The majority of patients with pancreatic cancer who undergo curative intent resection have recurrence, but optimal surveillance is unknown. Circulating tumor DNA (ctDNA) tests, custom-designed multiplex-PCR tumor assays unique to each patient’s tumor, may assist in monitoring for recurrence. Methods: We performed a single center retrospective review of patients with surgically resected pancreatic cancer who had at least 1 ctDNA test performed post-operatively between 8/1/21 and 11/13/23 with Natera Inc.’s Signatera assay. Primary outcomes were documented radiographic progression and death. Results: 74 patients were identified, of whom 36 were positive for ctDNA at any point during the study (median follow up from surgery 19.1 months) while 38 remained negative for ctDNA (15.3 months). There was no difference in age (≥ or < 70), sex, or race. Positive ctDNA at any time point was associated with advanced stage disease (p=0.013), positive resection margins (p=0.008), and having received neoadjuvant therapy (p=0.005). Those with any positive ctDNA were more likely to have radiographic progression (p<0.0001) or death (p=0.0005) by the end of the study. These differences remained when the population was examined in subgroups of first post-operative ctDNA positive, first ctDNA negative but at least one subsequent ctDNA positive, and all ctDNA negative. In Cox proportional hazard models, both any positive ctDNA test and positive first ctDNA test (first ctDNA limited to ≤120 days post-op) were associated with higher risk of radiographic progression and/or death compared to never testing positive for ctDNA and negative first ctDNA test, respectively (HR 12.6 (3.0-51.9), p=0.0005; HR 5.6 (1.4-22.8), p=0.02). Conclusions: Positive ctDNA tests, both in the immediate post-operative and subsequent settings, are associated with higher risk of radiographic progression or death in patients with resected pancreatic cancer. These data suggest that ctDNA testing provides independent, meaningful prognostic information superior to traditional pathologic risk factors and CA19-9 testing alone. These findings have implications for future clinical study design and decisions about surveillance and treatment. [Table: see text]