Optimal cut-off value for the number of colorectal liver metastases: a project study for hepatic surgery of the Japanese Society of Hepato-Biliary-Pancreatic Surgery.

Background: The optimal cut-off value of the number of colorectal liver metastases (CRLM) to predict prognosis after hepatic resection remains unclear. This study was conducted to determine a suitable cut-off value.

Methods: A total of 727 hepatectomized patients with CRLM were evaluated. We proposed the following optimal cut-off values: first, a small P-value for the log-rank test with no overlapping of the 95% confidence interval (CI) for median survival time using the Kaplan-Meier method and the hazard ratio (HR) using the Cox proportional hazards model and, second, the maximum HR value for accurate separation.

Results: For disease-free survival analysis, of the three group separations, A2 (1, 2-4, and ≥5) showed a small P-value and the largest HR, whereas two group separations, B2, B3 and B4 showed similarly small P-values, but B4 (1-4, ≥5) indicated the largest HR. Regarding the overall survival analysis, of the three group separations, A2 showed the smallest P-value, whereas the two group separations, B4 showed similarly small P-values, with the largest HR.

Conclusions: Tumor number separation in patients with CRLM after hepatic resection should be performed using the A2 (1, 2-4, and ≥5) or B4 (1-4 and ≥5) classifications.