Determinants of Long-Term Outcome in Patients Undergoing Simultaneous Resection of Synchronous Colorectal Liver Metastases

Background: It remains unclear which patients can benefit from simultaneous resection of synchronous colorectal liver metastases (SCRLMs). This study aimed to examine the prognostic value of patient- and tumor-related factors in predicting long-term outcomes of patients undergoing simultaneous resection of SCRLMs and to help patients select a suitable therapeutic regimen and proper surveillance. Methods: Clinicopathological and outcome data of 154 consecutive SCRLM patients who underwent simultaneous resection between July 2003 and July 2013 were collected from our prospectively established SCRLM data and analyzed with univariate and multivariate methods, and the prognostic index (PI) was formulated based on the regression coefficients (β) of the Cox model. The patients were classified into high- and low-risk groups according to the PI value; the cut-off point was the third quartile. Results: The 5-year overall survival rate was 46%, and the 5-year disease-free survival rate was 35%. Five factors were found to be independent predictors of poor overall survival (OS) by multivariate analysis: positive lymph node status, vascular invasion, BRAF mutation, the distribution of bilobar liver metastases (LMs) and non-R0 resection of LMs. Compared to low PI (≤5.978), high PI (>5.978) was highly predictive of shorter OS. Three factors were found to be independent predictors of poor disease-free survival (DFS) by multivariate analysis: tumor deposits, BRAF mutation and bilobar LM distribution. We also determined the PI for DFS. Compared to low PI (≤2.945), high PI (>2.945) was highly predictive of shorter DFS. Conclusions: Simultaneous resection of SCRLM may lead to various long-term outcomes. Patients with low PI have longer OS and DFS, while those with high PI have shorter OS and DFS. Thus, patients with high PI may receive more aggressive treatment and intensive surveillance, This model needs further validation.