Early treatment ctDNA dynamics to predict response to chemotherapy in patients with metastatic pancreatic cancer: A prospective, observational pilot study.

e16338 Background: Response to chemotherapy by imaging has limitations in metastatic pancreatic cancer. Circulating tumor DNA (ctDNA) may offer a potential adjunct or alternative toward expedient response monitoring. This analysis seeks to use temporal ctDNA testing as a potential surrogate marker to provide lead-time treatment response data compared to standard-of-care CT scans in metastatic pancreatic cancer. Methods: In this prospective, observational, single-center trial, patients with metastatic pancreatic ductal adenocarcinoma starting a new systemic treatment regimen underwent ctDNA monitoring. CT imaging was obtained prior to and after 2 months of treatment, with response measured by RECIST 1.1 criteria, including target and nontarget lesions. High-resolution temporal ctDNA profiling using Signatera was performed on Week (W) 1 Day (D) 1, W1D3 or W2D1 (depending on 5FU or gemcitabine-based therapy), W3D1, W5D1, W7D1, and W9D1. CA 19-9 and CEA were recorded as well at these timepoints. Progression and survival events are continuing to be collected up to 12 months after the start of treatment. The primary objective was to compare ctDNA response with CT scan response at 2 months. Secondary objectives included comparing ctDNA at 2 months with earlier timepoints, as well as comparing ctDNA trends with CA 19-9 and CEA. Results: Between June and December 2023, 19 patients were enrolled. Sufficient tissue was available to perform tumor-informed ctDNA profiling in 12 of our 19 patients (63%) undergoing systemic therapy with 7 patients starting on first-line (1L) therapy and 5 on second-line (2L) therapy. At a median follow-up of 173 days, 8 of 12 patients (67%) had progression of disease (POD) with a median PFS of 112 days. At the first 2-month scan, 11 patients were noted to have stable disease with 1 having POD. The 11 stable patients had a 5.4% decrease in CA 19-9, a 16.7% decrease in CEA, and a 64.5% decrease in ctDNA at W9D1 compared to W1D1. The patient with POD had a 148.9% increase in CA 19-9, 70.0% increase in CEA, and 4.6% increase in ctDNA at W9D1 compared to W1D1. ctDNA clearance was noted in 3 of 12 patients, as early as W5D1 in 2 patients on 2L therapy and W9D1 in 1 on 1L therapy. At 2 months, median tumor size decrease on CT was 18.2% in the ctDNA clearance vs. 6.7% the non-clearance group (p = .013). Median PFS was numerically higher 177 days in ctDNA cleared vs. 112 days in non-clearance group. Additionally, 2 patients had a 1-log decrease in ctDNA at W5D1 without clearance. Of, note there was one patient who had a 1-log decrease in ctDNA at W7D1, but a 20-fold increase at W9D1 (unlike CEA and CA 19-9, which were stable at W9). This patient had subsequent POD 1 month later. Conclusions: Early follow up suggests that ctDNA clearance at W5 may predict 2-month imaging findings. Further follow-up is needed to assess long-term outcomes and temporal predictive value of ctDNA testing