Is Hepatectomy Justified for BRAF Mutant Colorectal Liver Metastases? A Multi-institutional Analysis of 1497 Patients

Objective: To analyze clinical outcomes and prognostic variables of patients undergoing hepatic resection for BRAF mutant (BRAF-mut) colorectal liver metastases (CRLM). Background: Outcomes following hepatectomy for BRAF-mut CRLM have not been well studied. Methods: All patients who underwent hepatectomy for CRLM with complete resection and known BRAF status during 2001 to 2016 at 3 high-volume centers were analyzed. Results: Of 4124 patients who underwent hepatectomy for CRLM, 1497 had complete resection and known BRAF status. Thirty-five (2%) patients were BRAF-mut, with 71% of V600E mutation. Compared with BRAF wild-type (BRAF-wt), BRAF-mut patients were older, more commonly presented with higher ASA scores, synchronous, multiple and smaller CRLM, underwent more major hepatectomies, but had less extrahepatic disease. Median overall survival (OS) was 81 months for BRAF-wt and 40 months for BRAF-mut patients (P < 0.001). Median recurrence-free survival (RFS) was 22 and 10 months for BRAF-wt and BRAF-mut patients (P < 0.001). For BRAF-mut, factors associated with worse OS were node-positive primary tumor, carcinoembryonic antigen (CEA) >200 mu g/L, and clinical risk score (CRS) >= 4. Factors associated with worse RFS were node-positive primary tumor, >= 4 CRLM, and positive hepatic margin. V600E mutations were not associated with worse OS or RFS. A case-control matching analysis on prognostic clinicopathologic factors confirmed shorter OS (P < 0.001) and RFS (P < 0.001) in BRAF-mut. Conclusions: Patients with resectable BRAF-mut CRLM are rare among patients selected for surgery and more commonly present with multiple synchronous tumors. BRAF mutation is associated with worse prognosis; however, long-term survival is possible and associated with node-negative primary tumors, CEA <= 200 mu g/L and CRS < 4.