Utility of a Commercial Molecular Panel (REDPATH) in Pancreatic Cyst Fluid Analysis: A Single Center Experience

Abstract

Listen

Translate article

Background: Promising studies using molecular analysis to characterize pancreatic cysts as benign, pre-malignant, or malignant have led to the development of commercial molecular analysis panels, the role of which remains to be fully elucidated. Objective: To determine the clinical utility of commercial molecular analysis (RedPath) of pancreatic cyst fluid. Methods: Retrospective review of 105 consecutive EUS/FNA patients with pancreatic cysts referred to UC Davis Medical Center was completed. Among this group, 34 cases for which RedPath was obtained were used in the ensuing analysis. A CEA cutoff >192 ng/ml was used to classify cysts as mucinous. RedPath criteria for cyst classification as benign non-mucinous, indolent, or malignant included presence of a K-ras-2 mutation, two or more LOH mutations, or greatly elevated DNA quantity. Concordance was assessed using Cohen Kappa values. Simple probit regression was used to model clinical predictors of concordance. Sensitivity and specificity were determined using clinical consensus diagnosis as reference. Results: Fair agreement was noted between CEA and RedPath analysis (K=0.38). There were 24 concordant cases and 10 discordant cases. REDPATH did not predict more malignant pathology in any of the discordant cases. Regression analysis was used to assess whether clinical predictors such as family history of malignancy, presence of symptoms, diabetes, and cyst size were associated with discordance. No statistical significance was observed among these clinical predictors (p=0.69, p=0.98, p=0.30, p=0.63 respectively). No cases were identified in which RedPath results directly altered follow-up interval or referral to surgery. The results were used as supportive evidence in 62% of patients, to confirm surveillance recommendations. In the remaining 38% of patients, RedPath results were either unavailable or not addressed at follow-up. The sensitivity/specificity/accuracy of RedPath for diagnosing benign nonmucinous cysts was 100%/50%/75% and for benign mucinous cysts was 50%/100%/75%. Conclusion: CEA and RedPath analysis had fair concordance. Among discordant cases, RedPath neither predicted more malignant disease than CEA, nor did it alter follow-up intervals. RedPath was used mainly to corroborate CEA/EUS/cytology-based clinical decision on surveillance intervals. No clinical predictors were associated with discordant CEA and RedPath results. The high sensitivity of RedPath for benign nonmucinous cysts suggests a potential role for this modality as a reliable rule-out test for benign nonmucinous pathology. Further research should focus on cost effectiveness, and defining a specific role for its use in the management of pancreatic cystic neoplasms.