Get access

Perioperative chemotherapy with bevacizumab and liver resection for colorectal cancer liver metastasis


  • Presented at the 9th Annual Meeting of the American Hepato-Pancreato-Biliary Association, 12–15 March 2009, Miami, FL, USA.

Dr Prosanto Chaudhury, Department of Surgery, McGill University Health Center, 687 Pine Avenue West, Montreal, QC H3A 1A1, Canada. Tel: + 1 514 934 1934 (ext 31951). Fax: + 1 514 843 1503. E-mail:


Background:  Surgery remains the only curative option for patients with colorectal cancer liver metastases (CRLM). Perioperative chemotherapeutic strategies have become increasingly popular in the treatment of CRLM. Although the role of bevacizumab (Bev) in this setting remains unclear, its widespread use has raised concerns about the use of Bev as part of perioperative chemotherapy.

Methods:  We retrospectively reviewed all patients who received Bev and underwent liver resection between July 2004 and July 2008 at the McGill University Health Center. Chemotherapy-related toxicity, response to chemotherapy, surgical morbidity and mortality, liver function and survival data were assessed.

Results:  A total of 35 patients were identified. Of these, 26 (74.3%) patients received oxaliplatin-based cytotoxic chemotherapy, six (17.1%) received irinotecan-based therapy and the remainder received both agents. A total of 17 patients (48.6%) underwent portal vein embolization prior to resection and 12 (34.3%) underwent staged resection for extensive bilobar disease. A median of six cycles of preoperative Bev were administered. Nine patients (25.7%) experienced grade 3 or higher chemotherapy-related toxicities. Four events were deemed to be related to Bev. The overall response rate was 65.7% (complete and partial response). One patient progressed on therapy, but this did not prevent R0 resection. The incidence of postoperative morbidity was 42.3%. A total of 21.7% of complications were Clavien grade 3 or higher. There were no perioperative mortalities. There were no cases of severe sinusoidal injury or steatohepatitis. The Kaplan–Meier estimate of 4-year survival was 52.5%.

Conclusions:  These data confirm the safety of chemotherapy regimens which include Bev in the perioperative setting and demonstrate that such perioperative chemotherapy in patients with CRLM does not adversely affect patient outcome. There was no increase in perioperative morbidity compared with published rates. The addition of Bev to standard chemotherapy may improve response rates, which may, in turn, impact favourably on patient survival.