ctDNA informed management of metastatic pancreatic cancer.

Abstract

e16371 Background: Circulating tumor DNA (ctDNA) informs therapeutic decisions and its use in pancreatic cancer is emerging. We explore the adoption of ctDNA profiling in patients presenting to our pancreatic cancer clinic to help guide metastatic pancreatic cancer treatment. Methods: This was an IRB-approved retrospective cohort study of ctDNA profiling using the Signatera assay among patients with pancreatic cancer at Northwell Health between October 2021 and May 2024. Patients received ctDNA and CA 19-9 testing concurrently. The data presented here represent the subset of patients with metastatic disease receiving at least one Signatera test. Statistics and survival data were computed in Microsoft Excel and GraphPad Prism 10.4.0. The threshold for statistical significance used was an alpha of 0.05. Results: 124 ctDNA tests were completed among 32 patients with metastatic pancreatic cancer and at least one available ctDNA test. Median age at diagnosis was 66, 56% of patients were female, and 69% of patients identified as White, with the remaining 31% identifying as Hispanic, Asian, Black, or other. Gender and race were not associated with survival differences. The median number of per-patient tests was 4 (range 1-8). 72% of patients were initially tested prior to first line chemotherapy for their metastatic disease, 16% were tested during first line chemotherapy, and 12% were tested during second or later lines of chemotherapy. Median baseline ctDNA was 37 MTM/mL (0- 8,089 MTM/mL), while median baseline CA 19-9 was 368 U/mL (0-238,842 U/mL). Patients with a baseline ctDNA level below the median baseline ctDNA level had a statistically significant longer survival according to Mantel-Cox log-rank testing (17.3 vs. 10.0 months; HR = .4, p-value = .02), while CA 19-9 at baseline was not associated with survival (p = .37). In 16% of patients (N = 5), ctDNA impacted overall management by bolstering support for remission or relapse to help guide reimaging or continuation vs. maintenance of therapy, (N = 4), or by qualifying a patient for a trial (N = 1). Conclusions: ctDNA was a better prognostic biomarker than CA 19-9. Furthermore, it was clinically useful in several notable cases that helped providers select the most appropriate therapy or trial for their patients. However, because testing occurred at different timepoints along the cancer care continuum, significant heterogeneity may exist. Further studies as well as standardization of ctDNA testing protocols in practice will help elucidate best practices in applying ctDNA to improve patient outcomes.