Introduction While colonic and rectal cancers are often considered as a single disease entity there is a growing body of evidence that these are in fact separate disease processes. While a variety of factors have been identified in the literature as predictive of outcome following resection of metastatic disease from colorectal primaries it is not known if colonic liver metastases (CLM) behave differently to rectal liver metastases (RLM). The aim of this study was to determine those factors which predict long-term survival following resection of either CLM or RLM.

Methods We analysed a prospectively maintained Hepatobiliary database of 418 patients (with complete follow-up) who underwent liver resection for CRM between January 2000 and December 2010. The cohort was stratified according to the site of the primary tumour with rectal tumours being defined as those within 15 cm of the anal verge. Continous variables were compared with the Mann–Whitney U test whereas categorical variables were compared with χ² test. Survival analysis was performed with Kaplan–Meier plots and significance assessed with log rank test. Multivariate analysis was performed using a Cox-Regression model. A p value of <0.05 was considered significant.

Results 55% of patients had CLM (n=227) whereas 45% had RLM (n=191) (p=0.258). Patients with CLM were less likely to have node positive primary disease (52% vs 62%; p<0.05). Overall 5-year survival was similar for both CLM and RLM (42% vs 45%; p=0.62). Following resection of CLM multivariate analysis identified a CEA ≥200 (OR 2.39; p<0.01) and the presence of 4 or more tumours (OR 2.4; p<0.05) as independent predictors of long term survival. While there was a strong trend towards poorer 5-year overall survival in those with a resection margin <1 mm this did not reach statistical significance (p=0.383) on univariate analysis. Following resection of RLM the presence of a resection margin <1 mm was the only independent predictor of survival (OR 2.86; p<0.001).

Conclusion Overall 5-year survival following resection of both CLM and RLM is similar. Those factors which predict long-term survival following liver resection however differ and this may have implications for selecting those for intensive follow-up or who may be candidates for adjuvant therapy.

Competing interests None declared.