Selection criteria for minimally invasive resection of intrahepatic cholangiocarcinoma-a word of caution: a propensity score matched analysis using the national cancer database.

Abstract

While minimally invasive liver resection (MILR) vs. open approach (OLR) has been shown to be safe, the perioperative and oncologic safety for intrahepatic cholangiocarcinoma (ICC) specifically, necessitating often complex hepatectomy and extended lymphadenectomy, remains ill-defined. The National Cancer Database was queried for patients with ICC undergoing liver resection from 2010 to 2016. After 1:1 Propensity Score Matching (PSM), Kruskal-Wallis and χ2 tests were applied to compare short-term outcomes. Kaplan-Meier survival analyses and Cox multivariable regression were performed. 988 patients met inclusion criteria: 140 (14.2%) MILR and 848 (85.8%) OLR resulting in 115 patients MILR and OLR after 1:1 PSM with c-index of 0.733. MILR had lower unplanned 30-day readmission [OR 0.075, P = 0.014] and positive margin rates [OR 0.361, P = 0.011] and shorter hospital length of stay (LOS) [OR 0.941, P = 0.026], but worse lymph node yield [1.52 vs 2.07, P = 0.001]. No difference was found for 30/90-day mortality. Moreover, multivariate analysis revealed that MILR was associated with poorer overall survival compared to OLR [HR 2.454, P = 0.001]. Subgroup analysis revealed that survival differences from approach were dependent on major hepatectomy, tumor size > 4cm, or negative margins. MILR vs. OLR is associated with worse lymphadenectomy and survival in patients with ICC greater than 4cm requiring major hepatectomy. Hence, MILR major hepatectomy for ICC should only be approached selectively and if surgeons are able to perform an appropriate lymphadenectomy.