Impact of surgical timing on outcomes in patients with borderline-resectable and locally advanced pancreatic cancer undergoing neoadjuvant therapy.

Abstract

e16460 Background: Neoadjuvant therapy (NAT) followed by resection is standard for borderline-resectable (BR) pancreatic ductal adenocarcinoma (PDAC) and increasingly used in select locally-advanced (LA) cases with adequate tumor response. However, the clinical implications of the interval between NAT completion and surgery are not well characterized, and surgical timing varies widely in practice. We evaluated the association between surgical timing, pathological response, and survival outcomes. Methods: This retrospective study included patients (pts) with BR- and LA-PDAC who underwent NAT followed by resection at a single center. Clinicopathologic and outcome data were extracted from electronic health records. The interval from NAT completion to surgery (interval) was measured in weeks (wks); pathologic response was graded 0–3 using College of American Pathologists (CAP) tumor regression grade (TRG), and associations were evaluated using multivariable logistic regression. Recurrence-free survival (RFS) and overall survival (OS) were calculated from surgery to recurrence/death and death, respectively, and analyzed using Kaplan-Meier and Cox methods. Results: We identified 304 pts with PDAC (71% BR, 29% LA) treated between Feb 2012 and Jun 2022. Median age was 67 years (range: 31–85); 48% were female. NAT consisted of chemotherapy (CT) plus chemoradiotherapy (CRT) in 63% and CT alone in 37% (68% FOLFIRINOX, 26% gemcitabine–nab-paclitaxel), followed by resection (94% pancreaticoduodenectomy, 6% distal pancreatectomy). Median NAT-to-surgery interval was 6.4 wks (1.9–35.9). CAP TRG distribution was: 0 (4%), 1 (9%), 2 (59%), and 3 (28%). At a median follow-up of 65.3 months (mo), 73% developed recurrence and 70% died. Median RFS and OS were 15.2 (95%CI: 11.7–18.7) and 34.4 (95%CI: 28.5–39.4) mo, respectively. Improved TRG, negative margins, absence of lymphovascular space invasion (LVSI), and perineural invasion (PNI) were associated with superior RFS and OS (all p < 0.01) on univariate analysis. On multivariable analysis, longer interval increased the odds of favorable TRG in pts receiving CRT (OR 1.094 per week, 95%CI 1.028–1.164; p = 0.004) but not in those receiving CT alone (OR 1.023, 95%CI 0.896–1.166; p = 0.606). While interval was not independently associated with survival in the overall cohort, pts achieving excellent path response (TRG 0-1, 13% of CRT pts) had significantly longer intervals (10.7 vs 7.5 wks, p = 0.017), improved mRFS (38.4 vs 11.7 mo, p = 0.003), and mOS (NR vs 27.5 mo, p < 0.001). Conclusions: Longer intervals between NAT and surgery in BR- and LA-PDAC were associated with improved path response in pts receiving CRT. Path response correlated with survival outcomes. This potentially reflects a time-dependent maturation of treatment effect and warrants further prospective studies to define the optimal timing of surgery.