BACKGROUND: Long-term survival is reported in patients with liver metastases of colorectal cancer. Recently, an increased number of reports on liver resection following neoadjuvant chemotherapy in patients with initially unresectable liver metastases has been published.

METHODS: We analysed all published or presented trials and retrospective studies that report the rate of objective response and the rate of resection of initially unresectable metastases to correlate objective response and the rate of resection of metastases.

RESULTS: In studies that enrolled patients with metastases confined to the liver, 24-54% of patients were resected following chemotherapy, compared to 1-26% of patients in trials that included non-selected patients with metastatic colorectal cancer. A strong correlation was found between response rates and the resection rate in studies with patients with isolated liver metastases (r = 0.96, P=0.002). Likewise, in studies with non-selected patients, the resection rate of metastases also was associated with the objective response rate (r = 0.74, P <0.001).

CONCLUSIONS: Patient selection and efficacy of pre-operative chemotherapy are both strong predictors for resectability of liver metastases. Resectability is a novel endpoint focusing on the curative potential of treatment compared with classical endpoints of response or progression-free survival that are important if palliation is the aim. Therefore, patients with potentially resectable liver metastases should be investigated in special trials and interdisciplinary teams.