Abstract
BACKGROUND: The type of liver resection (anatomical resection, AR or non-anatomical resection, NAR) for colorectal liver metastases (CRLM) is subject to debate. The debate may persist because some prognostic factors, associated with aggressive tumor biological behavior, have been overlooked. OBJECTIVE: Our study aimed to investigate the characteristics of patients who would benefit more from anatomical resection for CRLM. METHODS: Seven hundred twenty-nine patients who underwent hepatic resection of CRLM were retrospectively collected from June 2012 to May 2019. Treatment effects between AR and NAR were compared in full subgroup analyses. Tumor relapse-free survival (RFS) was evaluated by a stratified log-rank test and summarized with the use of Kaplan-Meier and Cox proportional hazards methods. RESULTS: Among 729 patients, 235 (32.2%) underwent AR and 494 (67.8%) underwent NAR. We showed favorable trends in RFS for AR compared with NAR in the patients with KRAS/NRAS/BRAF mutation (interaction P <0.001) or right-sidedness (interaction P <0.05). Patients who underwent AR had a markedly improved RFS compared with NAR in the cohorts of RAS/NRAS/BRAF mutation (median RFS 23.2 vs. 11.1 months, P <0.001) or right-sidedness (median RFS 31.6 vs. 11.5 months, P <0.001); upon the multivariable analyses, AR [gene mutation: hazard ratio (HR)=0.506, 95% CI=0.371-0.690, P <0.001; right-sidedness: HR=0.426, 95% CI=0.261-0.695, P =0.001) remained prognostic independently. In contrast, patients who underwent AR had a similar RFS compared with those who underwent NAR, in the cohorts of patients with gene wild-type tumors (median RFS 20.5 vs. 21.6 months, P =0.333). or left-sidedness (median RFS 15.8 vs. 19.5 months, P =0.294). CONCLUSIONS: CRLM patients with gene mutation or right-sidedness can benefit more from AR rather than from NAR.