AVAGAST: A randomized, double-blind, placebo-controlled, phase III study of first-line capecitabine and cisplatin plus bevacizumab or placebo in patients with advanced gastric cancer (AGC).

LBA4007 Background: The median survival for patients (pts) with AGC in most phase III studies is less than 1 year. The addition of bevacizumab (bev) to chemotherapy (chemo) is supported by a strong preclinical rationale and by phase II evaluation. AVAGAST is the first randomized study to compare the efficacy and safety of bev + chemo vs placebo + chemo. Methods: Pts with inoperable, locally advanced or metastatic stomach/gastroesophageal junction adenocarcinoma with no prior therapy were randomized 1:1 to capecitabine (cape, or 5-FU) + cisplatin (cis) and either bev (7.5 mg/kg iv) or placebo q3w. Stratification variables: geographical region, fluoropyrimidine treatment, disease status. Cis was given for 6 cycles; bev/placebo + cape/5-FU were given until disease progression or unmanageable toxicity. Primary objectives: compare progression-free survival (PFS), overall response rate (ORR), and safety (as overseen by an independent DSMB). Results: From Sep 2007 to Dec 2008, 774 pts were enrolled. Treatment arms were balanced. Approx 95% of pts were metastatic. Two-thirds of pts were male, 49% of pts were from Asia/Pacific, 32% from Europe and 19% from the Americas. Median OS was 10.1 months with chemo + placebo and 12.1 months with chemo + bev in the intent-to-treat population (HR 0.87; p=0.1002). Median OS according to geographical region was 6.8 vs. 11.5 months (HR 0.63) in the Americas, 8.6 vs. 11.1 months (HR 0.85) in Europe and 12.1 vs. 13.9 months in Asia-Pacific (HR 0.97). Secondary endpoints and AEs of special interest for Bev are summarized below ( Table ). Conclusions: While the primary endpoint was not met (median OS HR 0.87; p=0.1002), there was a significant improvement in PFS and ORR and an acceptable safety profile for bev + chemo in patients with AGC. [Table: see text] [Table: see text]