Preoperative bevacizumab and surgery for colorectal liver metastases: a propensity score analysis.

Purpose: Bevacizumab associated with conventional chemotherapy has become standard care in the management of metastatic colorectal cancer. However, its impact on outcomes after liver resections (LRs) remains debated. The aim of this study was to evaluate the impact of neoadjuvant bevacizumab-based chemotherapy (BBC) on postoperative outcomes of LR for colorectal liver metastasis (CLM) using a validated approach.

Methods: All patients who received neoadjuvant therapy for CLMs between January 2005 and May 2011 were included. Risk factors for major complications (Clavien ≥3) were analyzed by univariate and multivariate analysis. Evaluation of BBC's impact on morbidity was conducted after a propensity score adjustment on factors identified to influence major complications (MCs).

Results: LR for CLMs after neoadjuvant chemotherapy was performed in 199 patients (127 men and 72 women). Major LR was performed on 111 patients (55.78%), and MCs occurred in 41 cases (20.6%). After multivariate analyses, major LR (OR 2.85; 95% CI 1.29-6.85; P = 0.013) and combined resections of both the primary tumor and CLMs (OR 7.12; 95% CI: 2.6-20.5; P < 0.001) were independent predictive factors for MCs. After a propensity score matching, 56 patients with a BBC regimen were compared to 112 patients without BBC. No difference in terms of biliary fistula occurrence (P = 0.94) or 90-day mortality (P = 0.66) was found. Both in the univariate and multivariate analyses, BBC was not associated with MCs (P = 0.95).

Conclusion: The present study using propensity score matching demonstrated that BBC did not impair outcomes of LR for CLM.