[903] Bevacizumab (BV) maintenance therapy significantly delays disease progression (PD) or death compared with placebo (PL) in the AVADO trial (BV + docetaxel [D] vs D + PL in 1st-line HER2-negative locally recurrent [LR] or metastatic breast cancer [mBC]).
Fumoleau P, Greil R, Rayson D, Müller V, Barni S, Aleknavicius E, Tellez E, Wilson C, Miles DW Centre R
gional de Lutte Contre le Cancer, Dijon, France; Medical University of Salzburg, Salzburg, Austria; Dalhousie University, Halifax, NS, Canada; University Medical Center Hamburg-Eppendorf, Hamburg, Germany; Azienda Ospedaliera Treviglio-Caravaggio, Treviglio, Italy; Vilnius University, Vilnius, Lithuania; Instituto de Seguridad y Servicios Sociales de los Trabajadores del Estado de Puebla, Puebla, Mexico; Addenbrookes Hospital, Cambridge, United Kingdom; Mount Vernon Cancer Centre, Middlesex, United Kingdom
Background: BV (Avastin
) is a monoclonal antibody to VEGF that significantly improved PFS and response rate when combined with 1st-line taxane chemotherapy (CTx) in two randomised phase III studies in patients (pts) with LR or mBC: E2100 (weekly paclitaxel 
BV) and double-blind, PL-controlled study AVADO (D two different BV doses). Preclinical data suggest that BV inhibits tumour neovascularisation and can limit tumour vascular regrowth. Both above trial protocols recommended BV use until PD, even after discontinuation of concomitant CTx. In the NO16966 trial in metastatic colorectal cancer, PFS was longer for pts continuing BV following cessation of CTx than for those where BV and CTx were stopped at the same time. We report results of an exploratory analysis of time to PD or death after cessation of D in pts in AVADO.
Methods: AVADO compared BV 7.5 or 15mg/kg + D 100mg/m2 with PL + D in 736 pts with inoperable LR or mBC, ECOG PS 0
1 and adequate LVEF. D and BV/PL were administered q3w, with D given for up to 9 cycles and BV/PL continued until PD or unacceptable toxicity.
Results: The table presents median time from discontinuation of D until PD or death in different subgroups in the AVADO study. Approximately 70% of pts in each arm discontinuing D went on to receive at least one dose of BV or PL before PD.
| D + PL | D + BV 7.5mg/kg | D + BV 15mg/kg | |
| ITT population, n | 241 | 248 | 247 |
| Unstratified HR for PFS (vs D + PL) | 0.79 | 0.72 | |
| Stratified HR for PFS* (vs D + PL) | 0.69 | 0.61 | |
| Pts discontinuing D, n | 228 | 235 | 240 |
| HR for time from last D to PD (vs D + PL) | 0.75 | 0.71 | |
| p value | 0.01 | <0.01 | |
| Median time from last D to PD/death, months | 2.6 | 4.2 | 4.4 |
Pts receiving  1 dose of BV/PL after stopping D, n | 156 | 166 | 177 |
| HR for time from last D to PD (vs D + PL) | 0.71 | 0.70 | |
| p value | 0.02 | 0.01 | |
| Median time from last D to PD/death, months | 3.4 | 5.1 | 4.8 |
Conclusions: Continuation of single-agent BV at either dose after discontinuation of D significantly delayed PD or death compared with single-agent PL. These results support the hypothesis that BV may inhibit tumour neovascularisation and progression following CTx and suggests that clinical benefit may be maximised by treatment with BV until progression.
Sunday, December 14, 2008 7:00 AM
Poster Discussion Session: Angiogenesis (7:00 AM-9:00 AM)Terms of Service.