BACKGROUND: Recurrence-free survival (RFS) may not be a surrogate for overall survival (OS) in patients with resectable colorectal liver metastases (CLM). We investigated whether a new composite tool-time to surgical failure (TSF)-is a suitable endpoint. METHODS: The medical records of consecutive patients who underwent curative resection for CLM at our center over a 17-year period were reviewed. Patients with liver-limited tumors (n = 371) who had not received previous treatment for metastasis were eligible for analysis. TSF was defined as the time until unresectable relapse or death. The correlations between TSF and OS, and between RFS and OS, were assessed for all the eligible patients. RESULTS: The median OS, TSF, and RFS were 5.7, 2.7, and 0.7 years, respectively, and the 5-year OS, TSF, and RFS rates were 52.6, 39.8, and 23.7 %, respectively, for all patients. The rates of first, second, and third relapse were 75.5, 77.6, and 70.8 %, respectively, and repeat resections were performed in 54.3 % (first relapses), 40.7 % (second relapses), and 47.1 % (third relapses) of patients. The concordance proportions of TSF and RFS for OS events were 0.83 and 0.65, respectively. The correlation between TSF and OS was stronger than that between RFS and OS in terms of the predicted probabilities. CONCLUSIONS: The correlation between TSF and OS was stronger than that between RFS and OS after curative hepatic resection. TSF could be a suitable endpoint for CLM overall management.
BACKGROUND: Recurrence-free survival (RFS) may not be a surrogate for overall survival (OS) in patients with resectable colorectal liver metastases (CLM). We investigated whether a new composite tool-time to surgical failure (TSF)-is a suitable endpoint. METHODS: The medical records of consecutive patients who underwent curative resection for CLM at our center over a 17-year period were reviewed. Patients with liver-limited tumors (n = 371) who had not received previous treatment for metastasis were eligible for analysis. TSF was defined as the time until unresectable relapse or death. The correlations between TSF and OS, and between RFS and OS, were assessed for all the eligible patients. RESULTS: The median OS, TSF, and RFS were 5.7, 2.7, and 0.7 years, respectively, and the 5-year OS, TSF, and RFS rates were 52.6, 39.8, and 23.7 %, respectively, for all patients. The rates of first, second, and third relapse were 75.5, 77.6, and 70.8 %, respectively, and repeat resections were performed in 54.3 % (first relapses), 40.7 % (second relapses), and 47.1 % (third relapses) of patients. The concordance proportions of TSF and RFS for OS events were 0.83 and 0.65, respectively. The correlation between TSF and OS was stronger than that between RFS and OS in terms of the predicted probabilities. CONCLUSIONS: The correlation between TSF and OS was stronger than that between RFS and OS after curative hepatic resection. TSF could be a suitable endpoint for CLM overall management.