Duodenal/ampullary invasion is a poor prognostic factor for neoadjuvant-treated pancreatic ductal adenocarcinoma patients

Abstract

Presenter: Katelyn Smith BA | University of Pittsburgh Medical Center Background: The current 8th edition of the American Joint Committee on Cancer (AJCC) T-staging system for pancreatic ductal adenocarcinoma (PDAC) is largely based on tumor size. In contrast, previous editions have not only incorporated tumor size but have also included extrapancreatic extension of the primary tumor. This modification in T-staging was done in part to address concerns of pathologic reproducibility in defining tumor spread into the surrounding peripancreatic soft tissue. While removing extrapancreatic extension may facilitate a more reproducible system, relying on tumor size alone may not account for other prognostic and less contentious pathologic features, such as duodenal/ampullary invasion. Moreover, considering the recent shift in clinical practice to include neoadjuvant therapy, defining and validating prognostic parameters for posttherapy pancreatectomy specimens are imperative. Methods: The prevalence and clinicopathologic features associated with PDAC invasion into the duodenum/ampulla were evaluated for 341 treatment-naïve patients and 329 neoadjuvant-treated patients between 2006 to 2016. These findings were separately analyzed using a cohort of 161 neoadjuvant-treated patients that were obtained between 2017 to 2019. Kaplan-Meier survival estimates and Cox proportion hazard regression were performed to identify predictors of overall survival (OS). Results: In comparison to treatment-naïve patients, neoadjuvant-treated patients had a lower prevalence of duodenal/ampullary involvement (44% vs. 57%, p = 0.002). A cross-comparison between treatment cohorts for patients without carcinomatous spread to the duodenum/ampulla revealed neoadjuvant-treated patients harbored smaller tumors (2.4 cm vs. 2.9 cm, p < 0.001) and had a lower prevalence of both perineural invasion (70% vs. 88%, p < 0.001) and lymphovascular invasion (60% vs. 75%, p = 0.005) as compared to treatment-naïve patients. A similar relationship was not observed between treatment groups for patients with duodenal/ampullary invasion and suggests duodenal/ampullary invasion may be an indicator of neoadjuvant treatment resistance. In the setting of neoadjuvant therapy, the detection of duodenal/ampullary involvement in posttherapy specimens correlated with a decreased 3-year OS rate (22% vs. 35%, p = 0.001), while no association between duodenal/ampullary invasion and OS was seen for treatment-naïve patients. A shorter OS for duodenal/ampullary invasion was also identified using a separate validation cohort of 161 neoadjuvant-treated patients (35% vs. 46%, p = 0.010). Among a total of 490 patients receiving neoadjuvant therapy, tumor infiltration into the duodenum/ampulla was a negative prognostic factor for poor OS (p = 0.042) by multivariate analysis (Table). Conclusion: The identification of duodenal/ampullary invasion in neoadjuvant-treated patients correlates with decreased patient OS and may reflect resistance to neoadjuvant therapy. Additionally, tumor infiltration into the duodenum/ampulla is a prognostic indicator for poor patient OS and is independent of other known prognostic parameters, such as the AJCC 8th edition T-staging system. Thus, the inclusion of duodenal/ampullary invasion in future editions of the AJCC T-staging system for neoadjuvant-treated patients may be of value.