353OEFFICACY AND SAFETY IN TANIA, A RANDOMISED PHASE III TRIAL OF CONTINUED OR REINTRODUCED BEVACIZUMAB (BEV) AFTER 1ST-LINE BEV FOR HER2-NEGATIVE LOCALLY RECURRENT/METASTATIC BREAST CANCER (LR/MBC)

ABSTRACT Aim: Combining BEV with 1st- or 2nd-line chemotherapy (CT) improves progression-free survival (PFS) in HER2-negative LR/mBC. The open-label randomised phase III TANIA trial evaluated further BEV in BEV-pretreated LR/mBC. Methods: Patients (pts) whose HER2-negative LR/mBC had progressed during/after ≥12 weeks of 1st-line BEV + CT were randomised 1:1 to 2nd-line single-agent CT either alone or with BEV (15 mg/kg q3w or 10 mg/kg q2w). Stratification factors were: hormone receptor status; time to 1st-line progression (PD; 1.5 × upper normal limit). 2nd-line therapy was continued until PD, unacceptable toxicity or consent withdrawal. At PD, pts in the CT arm received 3rd-line CT without BEV (no crossover); pts initially randomised to CT + BEV received 3rd-line CT + BEV. The primary endpoint was PFS from randomisation to 2nd PD/death. Additional endpoints included 2nd-line PFS in prespecified subgroups, 2nd- and 3rd-line PFS (randomisation to 3rd PD/death), 2nd-line objective response rate, overall survival (OS), safety, QoL and biomarkers. Sample size was calculated based on a log-rank test assuming median PFS of 7 → 9.3 mo and a HR of 0.75. PFS events were required in 384 of 488 pts for 80% power at 2-sided a = 0.05. Results: From Jan 2011 to Apr 2013, 494 pts were enrolled (247 CT; 247 CT + BEV). Baseline characteristics, CT vs CT + BEV: median age 54 vs 56 y; triple negative 23% vs 20%; disease-free interval ≤12 mo 10% vs 7%. The most frequently chosen 2nd-line CT was capecitabine (59% vs 61%). CT (N = 247) CT + BEV (N = 247) Median follow-up, mo 15.9 16.1 2nd-line PFS (primary endpoint) Events, n (%) 203 (82) 204 (83) Median 2nd-line PFS, mo 4.2 6.3 Stratified HR (95% CI) 0.75 (0.61–0.93) p = 0.0068a Best objective response rate, % (95% CI) 16.8 (11.7–22.9) 20.9 (15.2–27.5) Difference (95% CI) 4.1 (–4.2 to 12.4) p = 0.3457b 2nd-line grade ≥3 adverse events of special interest, %c (N = 238) (N = 245) Hypertension 7.1 13.5 Proteinuria 0.4 6.9 Venous thromboembolic event 2.1 3.3 Febrile neutropenia 1.7 3.3 Congestive heart failure 0.4 2.0 Bleeding 1.7 0.4 Arterial thromboembolic event 1.3 0 Wound-healing complication 0 0.8 Gastrointestinal perforation 0 0.4 Fistula/abscess 0 0 a2-sided stratified log-rank test. b2-sided chi-squared test with Schouten correction. cBEV basket terms, MedDRA v16.1. Data cut-off: 20 Dec 2013. Conclusions: The primary objective was met, showing statistically significantly improved PFS with BEV after PD on 1st-line BEV-containing therapy. 2nd-line safety results were as expected from previous BEV trials in LR/mBC. Final OS, 2nd and 3rd-line PFS and 3rd-line safety results are due in 2015. Disclosure: G. von Minckwitz: is a paid member on Roche's advisory boards, and has received speaker honoraria and conducted Roche-sponsored research; F. Puglisi: has received honoraria from Roche; J. Cortes: has received honoraria from Roche, Celgene, Novartis and Eisai and has acted as a consultant for Roche and Celgene; E. Vrdoljak: is a member of Roche advisory boards and has received corporate sponsored research; N. Marschner:is a member of Roche advisory boards and has received corporate sponsored research; C. Zielinski: has received speaker honoraria Roche, Eli Lilly and Pfizer; G. Romieu: is a member of a Roche Advisory Board; M. De Laurentiis: has received honoraria for talks, seminars and Advisory Board participation from Roche; C. Veyret: is an expert member of Roche Boards; S. De Ducla: is an employee of, and holds stock in Roche; U. Freudensprung: is an employee of F. Hoffmann-La Roche Ltd S. Srock: SS is an employee of, and holds stock in Roche; J. Gligorov: has acted as a Consultant/Advisor for, and has received honoraria and research funding from Roche. All other authors have declared no conflicts of interest.