Applicability of current NCCN Guidelines for pancreatic adenocarcinoma resectability: analysis and pitfalls.

Abstract

To test the applicability of National Comprehensive Cancer Network (NCCN v 3.2017) resectability criteria for pancreatic ductal adenocarcinoma (PDAC) in clinical practice, at a high-volume tertiary referral center. 102 consecutive patients (53 female; mean age 66.2years, range 34-90years) with biopsy proven, non-metastatic PDAC were evaluated by our multidisciplinary pancreatic cancer program between July 2013 and February 2016. Retrospective review of staging pancreatic CT angiography was performed, and radiographic features were categorized as conforming to or non-conforming to existing v 3.2017 definitions. Among 102 patients, 10 (10%) had CTA evidence of vascular involvement that did not conform to existing NCCN Guidelines. Six new scenarios of vascular involvement were identified. The remaining 92 patients presented with resectable (n=20 [22%]), borderline resectable (n=42 [45.6%]), or unresectable (n=30 [33%]) PDAC. Approximately half (n=21 [51%]) of borderline resectable patients' tumors demonstrated isolated venous involvement, whereas 39% had both arterial and venous involvement. A minority (11%) demonstrated only major arterial involvement. Assignment to unresectable status reflected both arterial and venous involvement (11, 37%), arterial involvement only (10, 33%) patients, and unreconstructible venous involvement in 9 (30%). In our experience, current NCCN resectability guidelines for PDAC do not accurately classify vascular involvement identified in approximately 10% of patients. Revision of the current guidelines could be helpful to clinical practice.