Impact of neoadjuvant radiation on postoperative outcomes of patients with pancreatic cancer undergoing pancreatoduodenectomy

Presenter: Samer AlMasri MD | University of Pittsburgh Medical Center Background: Neoadjuvant chemotherapy (NC) and chemoradiation (NCRT) are increasingly used to downstage pancreatic adenocarcinoma (PDA), particularly in head tumors requiring pancreatoduodenectomy (PD). While NCRT may increase margin negative rates and sterilize regional lymph nodes, its impact on postoperative morbidity remains poorly characterized. This study aimed to compare the impact of NCRT and NC on the postoperative morbidity of patients with PDA undergoing PD. Methods: A single-center, retrospective analysis of consecutive patients with PDA treated with NC or NCRT followed by PD between 2009 and 2019 was performed. All complications within the first post-operative year were recorded. Logistic regression analysis was performed to identify predictors of morbidity. Kaplan-Meier estimates and Cox proportional analysis were performed with overall survival (OS) and complication-free survival as the primary outcome. Results: Of 354 patients, 271 received NC and 83 received NCRT [60% received SBRT and 40% received other RT (EBRT/ IMRT/ 3D-conformal therapy)]. Mean age was 66.4 years and 50.6% were females. Patients who received NCRT were more likely to be females, had higher rates of locally advanced disease, arterial involvement and received higher number of NC cycles (all p<0.05). Compared to NC, patients who received NCRT had higher rates of open surgeries and vascular resections, longer operative times, higher operative blood loss and rates of ICU admissions, longer hospital stay, higher 90-day mortality, serious complications (Clavien grade >3), lower rates of receipt of adjuvant therapy, and higher rate of complications within 1 year (all p<0.05). Patients who received NCRT had significantly higher rates of GI bleed (p=0.012) and portomesenteric vessel thrombosis (p=0.0001) (Table). The most common cause of GI bleed in patients who received NCRT was a pseudoaneurysm (66.67%). On multivariate analysis, NCRT remained associated with longer (>10 days) post-operative hospital stay (OR=2.38; p=0.008), higher rate of serious complications (OR=1.95; p=0.04) and less receipt of adjuvant therapy (OR=0.41; p=0.011). On subgroup analysis of various RT modalities, SBRT was not a significant predictor of serious complications (OR=1.23; p=0.65) on multivariate analysis. Patients who received NCRT had significantly shorter complication free survival within the first post-operative year compared to those who received NC, on Kaplan-Meier (p=0.018) and Cox- regression analysis (HR=1.5; p=0.049). Median OS of patients who received NCRT was not significantly different from those who received NC (28.7 vs. 31.4 months; p=0.962). Conclusion: NCRT is associated with increased morbidity following PD, which may be explained by a higher proportion of locally advanced disease and vascular resections in this cohort. In this analysis, neoadjuvant SBRT was not associated with serious complications and may offer a safer alternative to NCRT, warranting its prospective analysis.