Purpose: Colorectal liver metastases (CLM) have significant molecular heterogeneity, which contributes to the risk of recurrence following surgery. Most of the traditional scores intended to predict recurrence is based on clinicopathological variables and it is unclear whether incorporating molecular biomarkers might improve our assessment of the risk of recurrence. Our aim was to determine if molecular biomarkers might be associated with the risk of recurrence after surgery of CLM.
Patients and methods: A total of 121 patients diagnosed with colorectal cancer (CRC) with resected liver metastases were included. The role of several clinicopathological variables to predict patient's outcome after resection of liver metastases was analyzed. Eighteen genes related to CRC pathogenesis were also included in the analyses. Univariate and multivariate stepwise Cox regression analyses were performed to identify factors associated with recurrence and the risk of death.
Results: Eight prognostic factors for progression-free survival and nine factors for overall survival were identified in the univariate analyses. After adjusting for other risk factors, only the expression of two molecular factors was associated with the risk of recurrence: TS (HR 0.631, 95 % CI 0.422-0.944) and SMAD4 (HR 1.680, 95 % CI 1.047-2.695). None of the variables was significantly associated with the risk of death in the multivariate analyses.
Conclusions: The prognostic significance of most traditional clinicopathological variables might be insufficient to define patients at risk for recurrence after liver metastases resection. Molecular biomarkers might improve the identification of patients with higher risk of recurrence.