Abstract P6-12-03: Meta-Analysis of Patients with Triple-Negative Breast Cancer (TNBC) from Three Randomized Trials of First-Line Bevacizumab (BV) and Chemotherapy Treatment for Metastatic Breast Cancer (MBC)

Abstract

Abstract Background: Three randomized Phase III trials (E2100, AVADO, and RIBBON-1) in patients with MBC have demonstrated that combining BV with first-line chemotherapy (taxane-, anthracycline-, or capecitabine-based) improves progression-free survival (PFS). We present a metaanalysis of patients with TNBC treated in the first-line setting. Methods: Kaplan-Meier methodology was used to estimate the median duration of PFS and overall survival (OS). Unstratified and stratified log-rank tests were used to assess the difference in PFS and OS between treatment cohorts (chemo only vs chemo+BV). For AVADO, only the 15 mg/kg BV group was included. Unstratified and stratified hazard ratios (HRs) were estimated using the Cox regression model. In the stratified analysis, disease-free interval, number of metastatic sites, visceral involvement (liver and/or lung), and the study were included as stratification factors. 1-year survival rates were compared using normal approximation method (Greenwood estimate for standard error). All p-values were two-sided and the type I error rate was set to 0.05 with no multiplicity adjustment. 95% confidence intervals (CIs) were calculated. Results: 621 patients were included in the pooled analysis: 258 in the chemo-only cohort and 363 in the chemo+BV cohort. Baseline characteristics in the two cohorts were similar. Median PFS was 5.4 months with chemo alone vs 8.1 months with chemo+BV (unstratified HR=0.65, 95% CI: 0.54-0.78, log-rank P<0.0001; stratified HR=0.68, 95% CI: 0.56-0.83, log-rank p=0.0002). PFS by baseline characteristics will be presented. The objective response rate was 23% for the chemo-only cohort (95% CI: 18-28%) and 42% for the chemo+BV cohort (95% CI: 37%-48%); P<0.0001 (unstratified and stratified analyses). The percentages of patients with progressive disease as best response were 28% and 11%, respectively (P<0.0001). OS was not significantly different between the two cohorts: in the chemo-only cohort the median OS was 17.5 months vs 18.9 months in the chemo+BV cohort (unstratified HR=0.96, 95% CI: 0.79-1.16, log-rank p=0.673; stratified HR=0.99, 95% CI: 0.81-1.21; log-rank p=0.930). The 1-year survival rates were 64.8% (95% CI: 58.9-70.7%) in the chemo-only cohort and 70.9% (95% CI: 66.2-75.6%) in the chemo+BV cohort, giving a difference in rates of 6.1% (95% CI: -1.5-13.6%; p-value for difference in rates=0.114). Grade 3-5 adverse events were observed in 46% of patients in the chemo-only cohort vs 53% of patients in the chemo+BV cohort. Grade 3 hypertension was reported in more patients in the chemo+BV arm (7.5%) than in the chemo-only cohort (1.6%). There was no grade 4 or 5 hypertension. Conclusions: The meta-analysis of patients with TNBC treated in the first-line Phase III trials of BV showed a statistically significant 35% decrease in the risk of disease progression or death (unstratified analysis) and a statistically significant 19% improvement in response rate. There was no difference in OS. The safety profile in this subgroup of patients is consistent with that observed in multiple Phase III studies across a variety of tumor types. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P6-12-03.