PTU-074 Anatomic vs non-anatomic resection of colorectal liver metastases: a comparison of surgical and oncological outcomes

Introduction Non-anatomic resections (NAR) for colorectal liver metastases (CRLM) have become increasingly common in an attempt to increase resection rates and to enable parenchyma-preserving resections, though some investigators have reported NAR to be associated with higher rates of resection margin positivity and poorer long-term survival than anatomic resections (AR). The aim of this study was to compare surgical outcomes, recurrence and survival in patients treated with NAR and AR. Methods Patients undergoing hepatic resection for CRLM between January 2002 and December 2008 were identified from a prospectively maintained database. Patients were divided into two groups: those who underwent AR and those who underwent NAR. Patients who underwent simultaneous AR and NAR were excluded from analysis. Clinicopathological variables and perioperative outcomes, as well as long-term disease-free and overall survival were compared between the AR and NAR groups, using the χ 2 test, Mann–Whitney U test, logistic regression analysis, log-rank analysis and multivariable Cox regression analysis employing a backwards stepwise model. Results Over the study period 91 patients underwent AR and 77 patients underwent NAR. Perioperative (90-day) morbidity, major morbidity and mortality occurred in 25.0%, 8.3% and 1.8% of cases respectively. There was no significant difference in rates of perioperative morbidity, major morbidity or mortality according to resection type (all p>0.30). There was also no difference in the incidence of positive resection margins between resection types (p=0.413). Overall survival was 48.5% and 36.6% at 3 and 5 years respectively, with disease-free survival of 31.4% and 25.3% at 3 and 5 years respectively. Type of resection was not associated with any significant difference in overall (p=0.430) or disease-free survival (p=0.625) following metastasectomy. Conclusion AR is not superior to NAR in terms of perioperative risk, tumour clearance, time to recurrence or long-term survival following metastasectomy. NAR should remain an integral component of the surgical treatment of CRLM and may enable increasing rates of operability. Furthermore, the preservation of hepatic parenchyma by NAR may enable a larger proportion of patients subsequently developing recurrent metastatic disease to undergo repeat metastasectomy. Competing interests None declared.