Role of para-aortic lymph node sampling in pancreatic cancer surgery.

Abstract

715 Background: In the management of pancreatic cancer, para-aortic lymph node (PALN) metastasis is regarded as distant metastasis, and systemic treatment is recommended. However, imaging study is not perfect to detect all PALN metastasis and the management of intraoperatively discovered PALN has been controversial. We hypothesized that sampling of PALNs on exploration could allow us to avoid pancreatic resection for patients who would not benefit. In this study, we evaluated the incidence and the effect on the long-term outcomes for patients with potentially resectable pancreatic cancer. Methods: Three hundred and ninety-two patients who had PALNs sampled upon potentially resectable pancreatic cancer from 2005 through 2014 were included in the study. All patients were appropriately staged preoperatively with CT/MRI and those with suspected PALN metastasis were not considered as candidates for resection. The patients whose resections were aborted because of liver metastasis or peritoneal dissemination discovered on exploration, or those who died within 30-days after the operation were not included. Evaluated outcomes were incidence of PALN metastasis and their recurrence-free and overall survivals (RFS, OS). Results: The patients’ median age was 74 years, and 58.6% was man. 67.8% had tumors at pancreatic head. Preoperative chemotherapy was given only on 16 patients (3.2%). Among 392 patients with PALNs sampled, 53 (13.5%) patients had metastasis; Resection was completed on 40 patients and resection was aborted on the rest. Among patients who underwent pancreatic resection, median RFS and OS were 10 and 12 months for patients with PALN metastasis, compared to 17 and 26 months for those without PALN metastasis (p < 0.001 for RFS and p < 0.001 for OS). The 5-year-OS rates for patients with/without PALN metastasis were 5.9% and 25% (p < 0.001). Among 53 patients with PALN metastasis, OS were not different between the patients who underwent resection and those who did not (median 13 months vs 17 months, p = 0.06), and there were no recurrence-free survivors. Conclusions: PALN sampling and evaluation before committing to resection is useful to identify the patients who can unlikely benefit and to avoid unnecessary morbid operation.