What Should We Cover Beyond Gross Disease With Radiation Therapy for Localized Pancreatic Ductal Adenocarcinoma (PDAC): Proposal of an Innovative and Standardized Clinical Tumor Volume (CTV) for PDAC of the Head – The Triangle

Abstract

<h3>Purpose/Objective(s)</h3> Although overshadowed by metastatic failure, patients (pts) with locally advanced pancreatic adenocarcinoma (LAPC) who undergo resection after neoadjuvant therapy (NAT) also experience high rates of locoregional failure (LF). Prior data from our institution indicate LF developed in > 30% of LAPC pts resected after NAT with multi-agent chemotherapy (MA-CT) and RT from 2016-2019 despite an R0 rate of 88%. In this cohort, gross disease and involved vessels were the target, raising the question whether field design modification could reduce LF. Indeed, elective treatment of tissue at risk of harboring microscopic residual disease after resection may be critical, but consensus regarding what constitutes the CTV for NAT does not exist. In the surgical literature, support is growing for extended dissection of the "Heidelberg Triangle" (HT), tissue between the celiac artery (CA), superior mesenteric artery (SMA), common hepatic artery (CHA), and portal vein (PV) that contains key perineural and lymphatic tracts at risk of harboring subclinical disease. Whether this concept can be applied to CTV field design is unclear. In investigation, we mapped LF locations in LAPC pts to assess if the LF distribution matches the anatomy of the Triangle volume (TV). <h3>Materials/Methods</h3> A total of 21 consecutive LAPC pts with involvement of the head or neck at diagnosis treated at our institution with MA-CT and stereotactic body radiation who developed LF after resection were reviewed. The TV starts superiorly at the CA take-off point, extends along the CA to the trifurcation, then extends laterally along the CHA and stops at the proximal proper HA. At this lateral boundary, the PV is also included and the TV extends posteriorly to the aorta along the inferior vena cava, including the SMA take-off until the crossing of the left renal vein. From the lateral boundary of the PV, the volume descends to the superior mesenteric vein (SMV) and covers the groove between the SMA-SMV. The inferior boundary is the branching of the first jejunal artery (JA) from the SMA. Descriptive statistics were generated to report the frequency of LF in relation to the TV and the frequency of critical vasculature involvement. <h3>Results</h3> In 21 consecutive pts, the median TV was 80.2 ± 30.1 cubic centimeters (cc) (27.7-110.3 cc). The LF centroid was located within the TV in 20 pts (95.2%). The SMA from take-off to the JA branch-point was the most common vessel involved at diagnosis (85.7%) and at failure (80.1%) with the next four being the CA (66.7%, 57.1%), SMV (61.9%, 66.7%), CHA (61.9%, 38.1%), and PV (38.1%, 47.6%). <h3>Conclusion</h3> The TV nearly universally encapsulates LF location and could serve as the basis for CTV design in the neoadjuvant setting. Future studies should investigate the feasibility of therapeutic dosimetry to this volume and ultimately whether treatment of the TV reduces LF.