[49] Benefit from exemestane (EXE) as extended adjuvant therapy after 5 years of tamoxifen (TAM): intent-to-treat analysis of NSABP B-33.

Mamounas E, Jeong J-H, Wickerham L, Smith R, Geyer C, Ganz P, Land S, Hutchins L, Eisen A, Ingle J, Costantino J, Wolmark N.. NSABP Operations Biostistical Centers, Pittsburgh, PA; Altman Health Foundation, Canton, OH; The Regional Cancer Center, Erie, PA; University of California at Los Angeles, Los Angeles, CA; Arkansas Cancer Research Cntr, Little Rock, AK; Toronto Sunnybrook Regional Cancer Cntr, North York, ON, Canada; Mayo Clinic, Rochester, MN; Allegheny General Hospital, Pittsburgh, PA

Until disclosure of ATAC trial results in 12/01, 5 years of TAM had been standard adjuvant hormonal therapy in early stage breast cancer (BC). In 5/01 NSABP began a randomized, placebo-controlled, double-blind clinical trial to evaluate EXE as extended adjuvant therapy after 5 years of TAM.
Subjects and Methods:Postmenopausal pts with clinical stage T1-3 N0-1 M0 BC, who were disease-free after 5 years of TAM were eligible. Prior adjuvant or neoadjuvant chemotherapy was allowed. Initially pts were randomized to 2 years of EXE (25 mg PO daily) or placebo (PLAC); in 2002 the protocol was amended to extend the EXE/PLAC duration to 5 years. Primary aim was to test whether EXE after 5 years of TAM prolongs disease-free survival (DFS). Secondary aims were to test the effect of EXE on overall survival (OS) and relapse-free survival (RFS) and the effect of EXE and TAM withdrawal on bone and lipids. A total of 3,000 pts were needed to detect a 21.3% reduction in hazard rate with a power of 80% (two-sided 0.05-level log-rank test). In 10/03, results of NCIC MA.17 showing benefit from extended adjuvant letrozole after 5 years of TAM, necessitated closure of accrual, treatment unblinding, and offering EXE to PLAC pts.
Results:At the time of unblinding, 1598 pts had been randomized. Among 1577 eligible pts, 52% were node-negative and 49% were under the age of 60. Upon unblinding, 560 of 783 pts on EXE continued EXE; of 779 pts on PLAC, 344 switched to EXE. Despite the crossover, with median follow-up of 30 months there was a borderline statistically significant improvement in DFS in favor of the EXE arm (91% vs. 89%, RR=0.68, p=0.07) and a statistically significant improvement in RFS (96% vs. 94% RR=0.50, p=0.03) but no difference in OS (13 deaths on PLAC vs. 16 on EXE, RR=1.2, p=0.63). Toxicity was assessed up to time of unblinding. There was 1% grade 4 toxicity in each arm and 9% vs. 6% grade 3 toxicity with EXE vs. PLAC, respectively (p=0.03). Most common grade 3/4 toxicities in EXE vs. PLAC were arthralgia (1.0% vs. 0.5%), fatigue (0.9% vs. 0.5%) and bone pain (0.5% vs. 0.7%). At 6 months after unblinding there were 28 fractures in EXE and 20 in PLAC (p=0.33).
Conclusion:Despite premature closure and crossover to EXE, original EXE assignment resulted in borderline improvement in DFS and significant improvement in RFS of a similar magnitude to that seen in NCIC MA.17.

Friday, December 15, 2006 9:30 AM

General Session 3 (9:30 AM-11:15 AM)