BackgroundThe role of paraaortic lymph node (PALN) sampling for pancreatic cancer (PC) remains controversial. This study aimed to assess the yield of PALN sampling and its impact on long-term outcomes. MethodsPatients treated from 2005 to 2018 were included. Since 2015, patients with borderline resectable or unresectable (BR/UR) tumours received standardized neoadjuvant therapy (NAT), whereas patients with resectable tumours underwent upfront resection. ResultsAmong 771 patients with PALN excision, metastasis was confirmed in 76 patients (10%) and 59 proceeded to planned resection. PALN metastasis was associated with early recurrence and shorter survival after upfront resection (1-year recurrence-free rate: 23% vs 57% for resectable tumours, P < 0.001, 9% vs 52% for BR/UR tumours, P = 0.006; median overall survival (OS): 19 vs 31 months for resectable tumours, P < 0.001, 17 vs 23 months for BR/UR tumours, P = 0.057). In contrast, they were both similar between patients with/without PALN metastasis following NAT for BR tumours (1-year recurrence-free rate, 50% vs 58%, P = 0.88; median OS, 28 vs 35 months, P = 0.45, respectively). ConclusionsOccult metastasis to PALNs is not uncommon in patients with potentially resectable PC. Its prognostic impact depends on the treatment strategy, and the intraoperative decision for resection should be individualized.