Optimal Value of Mutant KRAS Circulating Tumor DNA for Predicting Prognosis and Monitoring in Patients with Pancreatic Adenocarcinoma: A Prospective Multicenter Cohort Study

Abstract Background The generalizability and applicability of circulating tumor DNA (ctDNA) KRAS mutations in patients with pancreatic adenocarcinoma (PAC) remain elusive. We aimed to propose a cut-off value of mutant KRAS (mKRAS) concentration in ctDNA and validate its clinical utility. Methods This prospective, multicenter cohort study enrolled 419 patients with PAC. mKRAS was determined using multiplex droplet digital polymerase chain reaction at baseline and longitudinal follow-up time points. The optimal cut-off value of mKRAS concentration for progression-free survival (PFS) established in the development cohort was validated for PFS and overall survival (OS) in the validation cohort. Results At the cut-off value of 160 copies/mL, a high mKRAS concentration was associated with shorter PFS (13.6 vs 6.4 months, P < 0.001) and OS (24.3 vs 11.2 months, P < 0.001) in the development cohort. The prognostic value of mKRAS concentration was verified in the validation cohort [hazard ratio (HR) 1.48, P = 0.011 for PFS; HR 1.85, P = 0.001 for OS]. The cut-off value also subclassified the prognosis of patients with normal carbohydrate antigen 19-9 (CA19-9). In resected PAC, the persistence of mKRAS ctDNA detection at 1 month after surgery was associated with a shorter PFS. Patients whose mKRAS persisted above 160 copies/mL during chemotherapy had the worst survival. Conclusions Measurement of mKRAS ctDNA before and during treatment confirms its clinical utility as an independent prognostic marker in patients with PAC. The clinical significance of mKRAS detection in plasma was also observed in patients with normal CA19-9. ClinicalTrials.gov Registration Number: NCT04241367