E234: Feasibility and Utility of Dynamic ctDNA Profiling in Patients with Pancreatic Adenocarcinoma in a Neoadjuvant Setting

Introduction: Circulating tumor DNA (ctDNA) is an emerging tool for the dynamic monitoring of the tumor genetic profile to guide personalized therapy and monitor disease burden. The role of ctDNA as a longitudinal biomarker for response to therapy in a neoadjuvant treatment (NAT) setting in pancreatic adenocarcinoma (PDAC) has not been evaluated. This pilot study aimed to evaluate the feasibility of detecting ctDNA in patients with resectable or borderline PDAC and assess changes in the ctDNA genetic profile in response to NAT.

Methods: Patients were enrolled between 8/2021 to 8/2023 for an IRB-approved prospective observational pilot study.  ctDNA was collected before the 1st, 3rd, 5th, and 7th cycles, intraoperatively, at postoperative day 6, and at the 3-month follow-up. Tumor tissues and liquid biopsy (peripheral blood) were subjected to NGS for mutations in 521 solid tumor-related genes evaluated for SNVs, indels, CNVs, and fusions, as well as TMB and MSI at a CLIA-certified commercial laboratory.  In parallel, CA 19-9, CEA, radiographic, and pathologic responses were collected.

Results: Six patients were enrolled, and 47 liquid biopsy samples were collected (Table 1). At the time of enrollment, 5 patients were radiographically staged as resectable and one patient was borderline resectable. All patients received FOLFIRINOX (FFX) during NAT, and one patient was switched to gemcitabine and nab-paclitaxel (GnP) due to disease progression. Pathogenic ctDNA mutations were detected in 66% (4/6) of patients. All patients (2) with tail lesions had detectable ctDNA. Dynamic changes in the presence of the driver genes (KRAS, SMAD4 and TP53) in ctDNA during NAT was detected in 3 patients (Table 1). One patient had a progression of disease and pathogenic PTPN11 and TSHR not present in the initial ctDNA sample appeared in ctDNA during cycle 5 when disease progression was confirmed radiographically. Normalization of CA 19-9 and pathological response to NAT were not associated with ctDNA clearance. All patients with detected ctDNA at the time of diagnosis had lymph node metastasis on surgical pathology.

Conclusions:

Prospective dynamic liquid biopsy in the NAT setting for resectable or borderline resectable PDAC patients is feasible and may represent a clinically useful approach for a subset of patients. The presence of pathogenic driver mutations in ctDNA at the time of diagnosis may predict early recurrence. Also, the detection of ctDNA may be more feasible in patients with tail lesions. The emergence of non-driver genetic mutations in ctDNA during disease progression warrants further investigation.