555P - Efficacy and safety of bevacizumab in metastatic colorectal cancer (mCRC): first-line analysis of pooled data from randomized controlled trials (RCTs)

Date 01 October 2012
Event ESMO Congress 2012
Session Poster presentation III
Topics Anticancer agents
Colon and Rectal Cancer
Biological therapy
Presenter Fairooz F. Kabbinavar
Authors F.F. Kabbinavar1, H. Hurwitz2, N. Tebbutt3, B.J. Giantonio4, Z. Guan5, L. Mitchell6, D. Waterkamp7, J. Tabernero8
  • 1Thoracic Oncology And Kidney Cancer Programs, David Geffen School of Medicine at UCLA, 90024 - Los Angeles/US
  • 2Duke University Medical Center, 27710 - Durham/US
  • 3Austin Health, University of Melbourne, Melbourne/AU
  • 4The Abramson Cancer Center, University of Pennsylvania, 19104 - Philadelphia/US
  • 5Cancer Center, Sun Yatsen University, guangzhou/CN
  • 6Global Medical Affairs Biometrics, F. Hoffmann-La Roche Ltd, Basel/CH
  • 7Global Medical Affairs, F. Hoffmann-La Roche Ltd, Basel/CH
  • 8Medical Oncology / Gastrointestinal Tumors Group, Vall d'Hebron University Hospital / VHIO, 08035 - Barcelona/ES



Bevacizumab (BV) with chemotherapy (CT) is a standard treatment for mCRC. This analysis pooled individual patient data from the clinical databases of six RCTs (phase 2 or 3) of BV to further define clinical outcomes with first-line treatment, including within subgroups.


Patient data were pooled from first-line (AVF2107, NO16966, ARTIST, AVF2192, AVF0780, AGITG MAX) mCRC trials of BV. All analyses were based on the intent-to-treat population. Overall and progression-free survival estimates (OS, PFS) were calculated by Kaplan-Meier methods. To assess differences in time to response variables by treatment arm (CT vs BV + CT), stratified random (overall) and fixed (subgroup comparisons) models were used to estimate pooled hazard ratios (HRs) and 95% confidence intervals (CIs), with each study included as a stratum.

Overall: BV + CT vs CT
HR (95% CI) P value
OS 0.81 (0.70 − 0.93) .0034
PFS 0.58 (0.46 − 0.73) <.0001
Subgroup comparisons: BV + CT vs CT
HR (95% CI) for OS HR (95% CI) for PFS
Monotherapy (n = 751) 0.86 (0.72–1.02) 0.56 (0.48–0.67)
Doublet therapy (n = 2427) 0.84 (0.77–0.92) 0.73 (0.67–0.80)
Patients with liver metastases only (n = 1095) 0.87 (0.76–1.00) 0.67 (0.59–0.77)
Patients with extensive disease (n = 1049) 0.79 (0.69–0.90) 0.67 (0.59–0.77)
KRAS wildtype patients (n = 364) 0.70 (0.54–0.91) 0.57 (0.45–0.72)
KRAS mutant patients (n = 166) 0.85 (0.60–1.22) 0.54 (0.38–0.76)


Of the 3178 pooled first-line patients (CT [n = 1481]; BV + CT [n = 1697]), 58.5% were male, 40.1% were ≥65 years, and 44.9% had an ECOG performance status ≥1. OS and PFS were statistically significantly increased in BV-treated patients vs control patients. The BV-associated adverse event profile from this pooled analysis identified no new safety signals.


The addition of bevacizumab to first-line CT resulted in statistically significant improvements in OS and PFS for mCRC patients in the overall analysis, with PFS benefit extending across subgroups defined by CT intensity, site of metastatic disease, and KRAS status.


F.F. Kabbinavar: Dr Kabbinavar has received honoraria from Genentech/Roche.

H.I. Hurwitz: Dr Hurwitz has served in a consultant/advisory role for Genentech, Roche, Sanofi, Amgen, and BMS. He has received honoraria from Roche. Dr Hurwitz has received research funding from Roche, Genentech, BMS, Sanofi, and Pfizer.

N.C. Tebbutt: Dr Tebbutt has received research funding from industry.

B.J. Giantonio: Dr Giantonio has received honoraria from Genentech/Roche.

L. Mitchell: Dr. Mitchell is an employee of F. Hoffmann-La Roche Ltd.

D. Waterkamp: Dr Waterkamp is an employee of F. Hoffmann-La Roche Ltd.

J. Tabernero: Dr Tabernero has served in a consultant/advisory role for Genentech/Roche.

All other authors have declared no conflicts of interest.