Abstract P2-16-03: Mature results on metastatic breast cancer patients with prolonged (≥1 year) exposure to first-line bevacizumab combined with paclitaxel from a large observation study

Abstract

Abstract Background: Bevacizumab (BEV) combined with chemotherapy (CT) significantly improves progression-free survival (PFS) and response rate (RR) vs CT alone in the 1st-line treatment of HER2-negative metastatic breast cancer (mBC). BEV was continued for ≥1 year in 20% of patients (pts) in the global ATHENA safety study. We examined a similar subgroup of pts with long-term antibody treatment in a large German observation study. Meanwhile, long-term follow-up of up to 4.5 years allows to present the final results, including mature data on overall survival. Methods: Pts who had received no prior CT for their mBC received BEV-PAC per the European label. Efficacy and safety were documented in detail for up to 1 year (or until progression, death, or BEV discontinuation if earlier), with subsequent retrieval of long-term follow-up. Results: By April 2013, final data were available for 865 pts, of whom 167 (20%) were recorded to have received BEV for ≥1 year. Baseline characteristics of this subset relative to the overall population are summarized in the table. CharacteristicOverall population (n = 865)Subgroup treated with BEV for ≥1 year (n = 167)Median age, years (range)58(26-87)57(28-79)Age ≥65 years,%3230Metastatic at diagnosis,%2018Disease-free inreval <12 months,%24*14#Liver metastasis,%4235≥3 metastatic sites,%3329Triple-negative disease,%1812ECOG performance status >1,%98Prior (neo)adjuvant chemotherapy,%6661*n = 591, initially M0. #n = 113 Dose reductions of long-term BEV occurred in only 3% of cycles and (at least once) in 17% of pts. In the majority of pts treated for ≥1 year, BEV was continued as a single agent after discontinuation of CT. The overall best RR in pts treated for ≥1 year was 80% (complete response in 19%). Median PFS, based on observed events in 83% of all pts, was 18.4 months compared to 9.6 months in the overall population. Overall survival data, based on 524 deaths (61%), show a median of 21.6 months. In the subgroup with prolonged therapy, median OS amounts to 35.7 months with 40% of pts still alive after 4 years. Safety of long-term BEV is as previously reported with only a slightly increased hypertension rate (35%, grade ≥3: 11%), and no cases of gastrointestinal perforation, arterial thromboembolic event, or reversible posterior leukoencephalopathy syndrome. Conclusions: According to the mature data of this study, including those on OS, a notable proportion of pts appear to derive benefit from prolonged exposure to first-line BEV-containing therapy. Although baseline characteristics exhibited only moderate indicators for an a priori more favorable prognostic profile in the subset of pts treated for ≥1 year, of course one has to keep in mind, that time-related efficacy data are biased towards improved outcome in those able to continue BEV for ≥1 year. However, the median survival of almost 3 years suggests that in a considerable number of pts continued first-line BEV-PAC correlates to favorable outcomes with limited adverse effects. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P2-16-03.