Abstract
PURPOSE: To comprehensively investigate the implications of lymph node dissection (LND) and the prognostic impact of the number of lymph node (LN) metastases on survival in intrahepatic cholangiocarcinoma (ICC) using a large-scale study. METHODS: Patients who underwent surgical resection for ICC between 2004 and 2018 were identified from the Surveillance, Epidemiology, and End Results (SEER) registries. The Kaplan-Meier and log-rank tests were used to compare cancer-specific survival (CSS) and overall survival (OS) between different groups. Propensity score matching (PSM) and subgroup analyses were performed to balance potential confounding factors. A multivariate Cox proportional hazards regression model was used to identify prognostic factors of survival outcomes. Restricted cubic splines fitted in the Cox proportional hazard regression models were also conducted to examine associations between continuous variables and outcomes. RESULTS: In all, 1028 patients were enrolled. There were 652 (63.4%) patients undergoing LND, with lymph node metastasis (LNM) confirmed in 212 (32.5%) cases. Patients receiving LND did not show better survival outcomes than those receiving non-LND (NLND). We divided the LND group into two subgroups: patients with LNM (+) and those without LNM (-). Among these three groups, patients with LNM experienced the worst CSS and OS, while NLND patients had similar survival times to LNM (-) patients. Restricted cubic spline analysis indicated that an increased number of LNM was associated with a decreased chance of survival (p < 0.001). Patients who received LND were further categorized as having no nodal metastasis (N0), 1-2 LNM (N1), or ≥3 LNM (N2) according to the number of LNM. The Kaplan-Meier curves showed that the mortality risk of patients with N0, N1, and N2 disease (median CSS, N0 50.0 vs. N1 22.0 vs. N2 14.0 months; median OS, N0 46.0 vs. N1 21.0 vs. N2 14.0 months, all p < 0.01) increased significantly, except for patients who had <6 LNs harvested. On multivariable survival analysis, a higher nodal stage (N1 vs. N0: CSS, hazard ratio [HR] 2.135, 95% CI 1.636-2.788, p < 0.001; OS, HR 2.100, 95% CI 1.624-2.717, p < 0.001; N2 vs. N0: CSS, HR 4.027, 95% CI 2.791-5.811, p < 0.001; OS, HR 3.678, 95% CI 2.561-5.282, p < 0.001) was an independent prognostic risk factor for survival. CONCLUSIONS: Despite the lack of a clear survival benefit of LND in patients with ICC, a significant positive association between the number of LNM and poor outcomes was observed. We still suggest adequate LND by examining at least six LNs to ensure precise staging. On this basis, the recently proposed nodal classification of N0, N1, and N2 stages may also allow better prognostic stratification of ICC patients.