Palliative surgery in patients with unresectable colorectal liver metastases: a propensity score matching analysis.

Background and ObjectivesThe current study was primarily intended to determine the best surgical treatment for patients with unresectable liver metastatic colorectal cancer (CRC). In addition, we assessed whether the improvement in survival resulting from palliative resection (PR) of the primary tumor was a function of the extent of liver metastasis. MethodsThe demographics, tumor characteristics, and survival outcomes of 261 patients who underwent palliative surgery for unresectable liver metastatic CRC were analyzed. A propensity-score model was used to compare the group of patients receiving PR and non-resection (NR). ResultsThere were 195 PR patients and 66 NR. The median survival of PR and NR patients was 21 months and 10 months, respectively (P<0.001). In a Cox multivariate analysis of 51 propensity-score matched pairs, PR resulted in longer survival than NR (Hazard Ratio for NR 1.481; 95% confidence interval: 1.003-2.185; P=0.048). The extent of liver metastasis only led to better survival of PR than NR patients among patients with limited liver metastasis not among those with extensive liver metastasis (P=0.001). ConclusionsPR appears to result in better survival than NR when the patient's overall condition permits an aggressive approach, especially in patients with limited liver metastases. J. Surg. Oncol. 2014 109:239-244. (c) 2013 Wiley Periodicals, Inc.