Patterns of Locoregional Pancreatic Cancer Recurrence after Total Neoadjuvant Therapy and Implications on Optimal Neoadjuvant Radiation Treatment Volumes.

PURPOSE/OBJECTIVE(S): Neoadjuvant treatment for patients with localized pancreatic adenocarcinoma (PDAC) has improved survival duration. As survival increases, local disease control becomes even more important. We sought to understand the patterns of locoregional recurrence following total neoadjuvant therapy (TNT) and determine the impact of treatment volumes on recurrence.

MATERIALS/METHODS: Patients with PDAC managed with neoadjuvant chemotherapy and chemoradiation (TNT) followed by surgery who developed an isolated locoregional or simultaneously locoregional and distant recurrence were identified. Locoregional recurrences were individually contoured utilizing commercially available software. When available, original neoadjuvant dose distributions were registered to the scans on which the locoregional recurrences were contoured. Recurrences where then classified as in-field (> 95% of prescription dose), marginal (50-95% of prescription dose), or out of field (< 50% of prescription dose). Target volumes were created using four commonly utilized PDAC contouring guidelines to characterize the relationship of the local recurrence to the RT dose distribution.

RESULTS: Of 474 patients treated with TNT and surgery, 80 (17%) patients developed a locoregional recurrence with or without distant recurrence, visible on diagnostic imaging. Of the 80 patients, 56 (70%) had tumors in the pancreatic head; 46 (57.5%) were borderline resectable, 23 (28.8%) locally advanced, and 11 (13.6%) resectable. The most common initial neoadjuvant therapies were FOLFIRINOX (57.5%) and gemcitabine/nab-paclitaxel (18.8%). Chemoradiation included concurrent gemcitabine (47.5%) or 5-fluorouracil (26.3%). RT dose distributions were available for 38 patients; 22 (57.9%) had in-field failures, 9 (23.7%) marginal failures, and 7 (18.4%) out of field failures. Each published contouring atlas covered a relatively low percentage of recurrences, which are summarized in Table 1. Regions at particularly high likelihood of recurrence that were under covered on existing atlases included: aortic-diaphragmic junction, retro-pancreatic duodenal nodal basin, and the region to the right of the superior mesenteric artery (SMA).

CONCLUSION: We present the largest series (to our knowledge) of mapped locoregional recurrences for patients being treated with TNT in PDAC. These recurrences differ substantially from established atlases and highlight anatomical regions of highest priority for RT coverage. A novel visual contouring volume highlighting these regions will be presented which will strive to advance the use of RT in the TNT setting.