Minimally Invasive Liver Resection for Early-Stage Hepatocellular Carcinoma: Inconsistent Outcomes from Matched or Weighted Cohorts

Background The aim of the current study was to re-evaluate the role of minimally invasive liver resection (MILR) among patients with early-stage (stage I or II) hepatocellular carcinoma (HCC) undergoing partial hepatectomy. Methods A retrospective analysis of the National Cancer Database (NCDB) was conducted to identify patients with early-stage HCC who underwent partial hepatectomy in the USA from 2010 to 2013. Overall survival (OS) was compared in three cohorts: crude; stabilized inverse probability of treatment propensity score weighting (IPTW); and propensity score matching (PSM). Results Among 4027 patients included in the study, only 11.7%, (n = 473) underwent MILR. In the stabilized IPTW cohort, patients who underwent MILR versus open resection were more likely to have tumors greater than 3 cm (63.9%, n = 285 vs. 51.4%, n = 228, p < 0.001) and poorly/undifferentiated tumors (21.5%, n = 96 vs. 12.9%, n = 57, p < 0.001). Within the crude cohort, a 5-year OS was superior among patients in the open surgical group (67.8%) compared with patients who underwent MILR (56.6%) (p < 0.001). After classic PSM analysis, the 5-year OS of patients undergoing MILR and open surgery were noted to be comparable (57.3% vs 63.8%, p = 0.17; HR 1.16, 95% CI 0.92-1.45). In contrast, after applying IPTW, the 5-year OS of patients who underwent MILR (55.5%) was worse compared with patients who had an open resection (67.5%) (HR 1.46, 95% CI 1.15-1.84; p < 0.001). Conclusions Long-term outcomes of patients undergoing MILR were comparable with patients who had open surgery when assessed by standard PSM. The use of IPTW resulted in more unbalanced groups leading to residual confounding and bias.