Discussions about the benefit of liver resection (LRx) for non-colorectal, non-neuroendocrine metastases are controversial. This study aimed to analyze the outcome of LRx for these patients and validate a previously published prognostic risk model.

The study analyzed 150 patients who underwent LRx for non-colorectal non-neuroendocrine (NCNN) metastases. Patients' demographics, tumor characteristics, treatment options, and postoperative outcome were investigated. The Kaplan-Meier method and Cox regression models were used to assess survival and prognostic variables.

After a median follow-up period of 61 months, 39 % of the patients were alive. The 30-day mortality rate was 0.7 %. The overall, disease-free, and intrahepatic recurrence-free survival rates were respectively 42, 29, and 51 % at 5 years and 28, 23, and 47 % at 10 years. The negative prognostic factors identified in the multivariate analysis were melanoma (p = 0.04), squamous tumors (p = 0.01), and a primary tumor liver metastasis, with an interval shorter than 2 years (p = 0.02), whereas the predictive prognostic factors identified were breast cancer (p = 0.04), stromal tumors (p = 0.03), and major LRx (p = 0.04). The prognostic risk score stratified patients into low risk (0-3 points: n = 50; 5-year overall survival [OS] 58 %), medium risk (4-6 points: n = 91; 5-year OS 35 %), and high risk (≥7 points: n = 9; 5-year OS, 33 %) groups (p = 0.01).

Liver resection for patients with NCNN metastases is a safe treatment option. More than 25 % of patients can achieve a long-term survival of 10 years when the histology of the primary tumor and the surrogates for the individual biologic tumor behavior are taken into account. Exclusion of patients with NCNN liver metastases from surgical therapy is no longer justified.