Understanding rates of attrition and barriers to surgical resection in patients undergoing neoadjuvant therapy for localized pancreatic cancer.

701 Background: Treatment paradigms are shifting toward neoadjuvant therapy (NAT) for resectable (R) and borderline resectable (BR) pancreatic cancer (PC) based on the rationale of treating occult metastatic disease, potentially downstaging patients (pts) to resectability if anatomically unresectable and, most importantly, selection of biology and pts most likely to benefit from potentially morbid surgery. While the benefits of NAT for PC have been well described, what remains underreported are the true attrition rates and underlying reasons for failure to proceed to surgical resection among pts receiving NAT. Herein, we present true intent-to-treat (ITT) attrition data of pts undergoing NAT for localized PC (LPC). Methods: Pts with LPC and who underwent ≥ 1 cycle NAT from January 2013-2022 were included. Clinicopathologic data were compared between those who did and did not proceed to pancreatectomy. Multivariable models identified factors associated with odds of attrition. Overall survival (OS) was estimated using the Kaplan-Meier method. Results: 427 LPC pts received NAT, including 57 (13%) with R, 133 (31%) BR, and 237 (56%) locally advanced (LA) disease. 64% had head tumors with median tumor size of 3.3 cm (IQR 2.7-4.2) and median CA 19-9 of 267 U/mL (IQR: 58-766). 60% of pts received FOLFIRINOX and 28% Gemcitabine-Abraxane for a median of 4.2 months (IQR 2.5-5.7). During NAT, 22% switched regimens and 30% received radiation. Following NAT, 207 (48%) pts underwent attempted resection, and 182 pts (43%) had successful pancreatectomy. The rate of attrition among all LPC pts was 52%, of which 20% had disease progression, 18% remained inoperable, 11% had physiologic decompensation, 2% were lost to follow-up, and 1% declined surgery. Rates of attrition were: 23% for R, 44% for BR, and 73% for LA (p<0.01). Higher stage was associated with increased odds of attrition (OR 9.7, 95% CI 4.8 – 19.5 for LA and OR 3.0, 95% CI 1.4 – 6.1 for BR compared to R). The most common reasons for attrition were: metastatic progression (5% in R, 18% in BR, 26% in LA), the primary tumor was deemed inoperable (2% in R, 12% BR, 32% LA), or physiologic decompensation on NAT (7% R, 11% BR, or 12% LA). Pts undergoing pancreatectomy had a median OS of 30 vs 16 months for those not (p<0.01). ITT median OS inversely correlated with stage – R: 30, BR: 21 and LA: 18 months (p<0.01). There was no difference in OS for those who received NAT and resection – R: 31, BR: 27, and LA: 34 months (p=0.66). Conclusions: At a tertiary center with 87% BR and LA pts, 43% of pts proceeded to pancreatectomy following NAT including 27% with LA tumors. Attrition of pts remains a significant issue even in the setting of R, highlighting the importance of NAT in pt selection. Pts who receive NAT and pancreatectomy experience similar OS, regardless of stage, further supporting the NAT approach in choosing patients most likely to benefit from resection.