Early Recurrence After Resection of Locally Advanced Pancreatic Cancer Following Induction Therapy: An International Multicenter Study.

Abstract

To establish an evidence-based cutoff and predictors for early recurrence in patients with resected locally advanced pancreatic cancer (LAPC). It is unclear how many and which patients develop early recurrence after LAPC resection. Surgery in these patients is probably of little benefit. We analyzed all consecutive patients undergoing resection of LAPC after induction chemotherapy who were included in prospective databases in The Netherlands (2015-2019) and the Johns Hopkins Hospital (2016-2018). The optimal definition for "early recurrence" was determined by the post-recurrence survival (PRS). Patients were compared for overall survival (OS). Predictors for early recurrence were evaluated using logistic regression analysis. Overall, 168 patients were included. After a median follow-up of 28 months, recurrence was observed in 118 patients (70.2%). The optimal cutoff for recurrence-free survival to differentiate between early (n=52) and late recurrence (n=66) was 6 months ( P <0.001). OS was 8.4 months [95% confidence interval (CI): 7.3-9.6] in the early recurrence group (n=52) versus 31.1 months (95% CI: 25.7-36.4) in the late/no recurrence group (n=116) ( P <0.001). A preoperative predictor for early recurrence was postinduction therapy carbohydrate antigen (CA) 19-9≥100U/mL [odds ratio (OR)=4.15, 95% CI: 1.75-9.84, P =0.001]. Postoperative predictors were poor tumor differentiation (OR=4.67, 95% CI: 1.83-11.90, P =0.001) and no adjuvant chemotherapy (OR=6.04, 95% CI: 2.43-16.55, P <0.001). Early recurrence was observed in one third of patients after LAPC resection and was associated with poor survival. Patients with post-induction therapy CA 19-9 ≥100U/mL, poor tumor differentiation and no adjuvant therapy were especially at risk. This information is valuable for patient counseling before and after resection of LAPC.