Performance Of Neoadjuvant Therapy In Locally Advanced Adenocarcinoma Of The Pancreas: A Single Large Volume Center Experience.

e16745 Background: Pancreatic adenocarcinoma (PC) is a disease with low 5-year survival rates even in patients who undergo surgical resection: ~10% overall and 20-35% in those with negative lymph nodes and R0 operation. Adjuvant chemotherapy (CT) has demonstrated benefit; however, not all patient are candidates for multi-agent CT post-op. Early use of systemic therapy might improve outcomes. As a leading referral center for pancreatic resection in a tri-state area, we present a single institution real-world experience of neoadjuvant CT (neo-CT) in PC patients principally with degrees of vascular abutment at presentation. Methods: We identified 90 patients scheduled for pancreatectomy from 2015 to 2019 who received neo-CT. Three had an aborted procedure due to occult metastatic disease. Of those who underwent resection (n = 87), most were borderline or locally advanced (n = 79) with a few deemed resectable (n = 5) or metastatic (n = 3) prior to CT. 80 patients (92%) had R0 resection, though final pathology showed R1 resection in 5 and distant metastases in 2 patients. 72% of patients (n = 63) received 5-FU-based neo-CT, while 28% (n = 24) received gemcitabine-based CT. 36% (n = 31) received neoadjuvant radiation (RT) (with or without concurrent CT) in addition to CT, and 43% (n = 37) had adjuvant therapy, either CT, RT, or a combination. 4 patients received additional therapies on clinical trial. 41 patients (47%) completed >6 months of peri-operative CT, with 23 of those (26%) completing 6 months prior to surgery. We evaluated pathologic regression scores and compared those with very good response (score 0-1) post-CT vs. partial (score 2) or no response (score 3). Results: Median progression-free survival (PFS) from start of treatment in our cohort was 1.9 years (95% CI 1.4-NR) while median overall survival (OS) was 3.0 years (95% CI 2.4-NR). Duration of neo-CT did not affect PFS or OS. Logrank tests of OS curves of the 3 pathologic response groups indicated a statistically-significant difference (p = 0.004), with longer OS in patients with score 0-1 and shorter OS in patients with score 3. Conclusions: This analysis involves patients treated in both community and academic settings. We observed improved survival among those with best pathologic responses. The majority of cases with lower residual tumor burden received FOLFIRINOX. We are analyzing molecular data in this cohort and will conduct adjusted analyses for clinical characteristics. The prospective selection of multiagent CT in PC may be influenced by molecular profile and early response assessment.