Differential prognostic impact of liver resection by the site of concurrent extrahepatic diseases in patients with colorectal cancer liver metastases: a nationwide multicenter study.

BACKGROUND

Liver resection for colorectal liver metastases (CRLM) with concurrent extrahepatic disease (EHD) has demonstrated potential benefits for long-term prognosis; however, its effectiveness remains controversial. Additionally, the prognostic impact of different EHD sites is not well elucidated. This study aimed to assess the significance of liver resection in patients with CRLM with concurrent EHD and evaluate how different EHD sites influence prognosis.

METHODS

A nationwide multicenter database was used for a retrospective analysis of patients diagnosed during two periods: 2005-2007 and 2013-2015. EHD was classified into the following five subgroups: lung, peritoneum, lymph nodes, local, and others. The inverse probability of treatment weighting (IPTW) method was applied to minimize selection bias. Kaplan-Meier survival curves and Cox proportional hazards models were used to compare the overall survival (OS) between the different treatment groups and EHD subgroups.

RESULTS

Among 3787 patients, 874 (23.1%) underwent liver resection. Following IPTW adjustment, the hepatectomy (HT) group demonstrated significantly better OS than the non-hepatectomy (non-HT) group (5-year hazard ratio, 0.322; 95% confidence interval, 0.273-0.379; P < 0.001). Analysis by the EHD site subgroup demonstrated that liver resection was associated with a better prognosis across all sites. However, the prognostic impact differed by EHD site, with peritoneal metastasis associated with poorer outcomes in both the HT (5-year OS rates, 30.1% vs. 45.0%) and non-HT (5-year OS rates, 4.4% vs. 8.6%) groups.

CONCLUSIONS

Regardless of the EHD site, liver resection was associated with a significantly better OS in patients with CRLM with concurrent EHD. The prognostic impact varies across EHD sites, underscoring the significance of considering differential prognostic risks when selecting treatment strategies.