Prediction of Recurrence With KRAS Mutational Burden Using Ultrasensitive Digital Polymerase Chain Reaction of Radial Resection Margin of Resected Pancreatic Ductal Adenocarcinoma.

OBJECTIVE: Although complete surgical resection is the only curative method for pancreatic cancer, the radial resection margins of pylorus-preserving pancreaticoduodenectomy specimens might be underevaluated.

METHODS: KRAS mutation was assessed with droplet digital polymerase chain reaction on cells collected from the radial resection margins of 81 patients, and the results were compared with those of conventional pathologic resection margin (pRM) evaluation.

RESULTS: KRAS mutation was detected in 76 patients (94%), and molecular resection margin (mRM) positivity defined by a KRAS mutation rate of 4.19% or greater was observed in 18 patients (22%). Patients with mRM-positive had significantly worse recurrence-free survival (RFS) than those with mRM-negative in entire groups (P = 0.008) and in subgroups without chemotherapy or radiation therapy (all, P < 0.001). When combined pRMs-mRMs were evaluated, patients with combined pRM-mRM-positive (either pRM- or mRM-positive) had significantly worse RFS than those with combined resection margin-negative (both pRM and mRM negative) by univariate (P = 0.002) and multivariate (P = 0.03) analyses.

CONCLUSIONS: KRAS mutational analysis with ultrasensitive droplet digital polymerase chain reaction of the radial resection margin in pancreatic cancer patients who underwent pylorus-preserving pancreaticoduodenectomy can provide more accurate information on RFS by using alone or in combination with conventional pRM evaluation, especially in patients without chemotherapy or radiation therapy.