Combined resection and RFA in colorectal liver metastases: stratification of long-term outcomes

Background: Combined hepatic resection and radiofrequency ablation (resection-RFA) is a widely accepted multidisciplinary treatment for unresectable colorectal cancer liver metastases. Worse prognosis after resection-RFA is correlated to tumor morphology, although unfavorable morphology is inherent to this patient cohort. This study aimed to select patients who may or may not benefit from resection-RFA with the aid of tumor biology. Method: Data from 485 patients who underwent curative hepatectomy with or without concurrent RFA were retrospectively collected and analyzed. Clinicopathologic characteristics, predictors of overall survival (OS), and OS of patients stratified by tumor biology in resection-RFA were analyzed. Results: Combined resection-RFA was performed in 86 patients (17.7%) and a standalone resection in 399 patients. Baseline patients' characteristics of the resection-RFA group were significantly different in terms of median number of tumors (5 versus 2) and bilobar distribution (84.9% versus 29.1%) from those of the resection-only group. Multivariate analysis identified four independent predictors of decreased OS in the resection-RFA group. Three were related to tumor biology: primary tumor nodal metastases (hazard ratio [ HR], 2.32; 95% confidence interval (95% CI), 1.16-4.64], Kirsten rat sarcoma viral oncogene homologmutation (HR, 2.64; 95% CI, 1.36-5.14), and preoperative high carcinoembryonic antigen (HR, 2.33; 95% CI, 1.13-4.81), and one related to tumor morphology-ablated lesions >= 3 (HR, 2.05; 95% CI, 1.41-3.80; P = 0.023). To examine the prognostic influence of tumor biology, the resection-RFA group was stratified into two groups by number of predictors related to tumor biology (lowrisk: 0-1 risk factors; n = 56 and high risk: 2-3 risk factors; n = 30). Median OS of the low risk, high risk, and resection-alone groups were 61.8, 20.7, and 75.3 mo, respectively. The 5-y OS rate was similar between the lowrisk and resection-alone group (52.7% versus 58.7%, respectively; P = 0.323). Conclusions: Patients with low-risk tumors undergoing a combined resection-RFA approach had roughly comparable OS to those who only underwent resection, irrespective of advanced tumor morphology. Combined resection-RFA procedures might be of value to these patients. (C) 2016 Elsevier Inc. All rights reserved.