Safety of bevacizumab when given perioperatively for colorectal cancer with liver metastases: McGill University pilot study

Abstract

e15083 Background: Bevacizumab (BV) increases responses and survival rates when used with chemotherapy in metastatic colorectal cancer (CRC). Current practice is to give peri-operative chemotherapy for resectable CRC liver metasases (LM). The safety and efficacy of chemo-BV in this setting is unknown. Post-operative complications have been reported in patients (pts) who underwent surgery while receiving BV. The goals were to determine safety of perioperative BV. Methods: In a prospective pilot study, patients with resectable LM from October 2005 to 2008 that received BV perioperatively along with chemotherapy. Of a total of 60 pts, 34 had oxaliplatin-based CTX, 22 CPT-11 (FOLFIRI) and 4 IROX. All but seven pts received BV pre and postoperatively. The average age was 55 years. All patients underwent liver surgery 6–8weeks post last dose of BV. Univariate Cox regression models were used to evaluate the association of patient and tumour characteristics, therapy and postoperative complications. Results: Postoperative complications developed in a total of 24 pts (40%). 12 pts (35%) who received Oxaliplatin CTX + BV, 9 pts (40%) with CPT 11 CTX + BV and 3 pts (75%) with Oxaliplatin + CPT-11 CTX +BV. The average time from BV discontinuation to surgery was 49 days. No significant associations were identified between BV and CTX regimen or timing and postoperative complications. Wound healing complication were most frequent with 8 pts (13%), DVT diagnosed in 6 pts (10%), protracted pancytopenia 4 pts (7%), sepsis 1pt, infections 2pts, MI 1pt, acute cardiomyopathy 1pt, and billiary leak 1pt. No bowel perforations or sudden deaths were reported. These side effects apart from thromboembolic events are comparable to previously reported post-operative complications. Conclusions: Neither the use of BV with CTX nor timing of BV administration were associated with a non acceptable increase in complication rates as compared to previously published by EORTC intergroup. Our data confirm that the combination of BV with neoadjuvant chemotherapy is feasible and safe in patients CRC LM. A higher incidence of thromboembolic events was seen. To determine the optimal timing and drug combination prospective randomized trials are urgently required. [Table: see text]