Andrew Tutt, Judy Ellen Garber, Bella Kaufman, Giuseppe Viale, Debora Fumagalli, Priya Rastogi, Richard D. Gelber, Evandro de Azambuja, Anitra Fielding, Judith Balmana Gelpi, Karen A. Gelmon, Nigel Baker, Amal Arahmani, Elzbieta Senkus-Konefka, Eleanor Mc Fadden, Vassiliki Karantza, Sunil R. Lakhani, Greg Yothers, Christine Campbell, Charles E. Geyer; Breast Cancer Now, Toby Robins

Background: PARP inhibitors (PARPi) target cancers with homologous recombination repair defects by synthetic lethality. The PARPi olaparib (OL) is licensed for metastatic HER2-negative breast cancer with BRCA1/2 germline mutation (gBRCAm). Despite (neo)adjuvant chemotherapy ([N]ACT), recurrence rates in patients (pts) with gBRCAm early breast cancer (EBC) can be high. Novel adjuvant treatments are needed.

Methods: OlympiA (NCT02032823), a randomized, double-blind, phase III study, enrolled pts with gBRCAm and HER2-negative (TNBC or hormone-receptor+ [HR+]) high-risk EBC after primary local treatment and ACT/NACT. Eligible pts with TNBC had ≥pT2 or ≥pN1 disease prior to ACT or non-pCR after NACT; those with HR+ BC had ≥4 positive nodes prior to ACT or non-pCR and CPS&EG score ≥3 after NACT. Pts were randomized 1:1 to 1 year of continuous oral OL (300 mg BID) or placebo (PL). Endocrine therapy and bisphosphonates were allowed. The primary endpoint was invasive disease-free survival (IDFS) in the ITT population. Secondary endpoints included distant DFS (DDFS), overall survival (OS) and safety. Safety analysis included adverse events of special interest (AESI) (myelodysplastic syndrome/ acute myeloid leukemia, new primary malignancy, pneumonitis). Per protocol IDMC interim analysis (IA) review was triggered at 165 IDFS events in the first 900 pts, with superiority boundaries based on a hierarchical multiple testing procedure: P < 0.005 for IDFS, followed by P < 0.005 for DDFS and p<0.01 for OS.

Results:1836 pts were enrolled between 06/14–05/19; 49.9% had ACT, 50.1% NACT. Baseline demographics and tumor characteristics were balanced between arms. 82.2% had TNBC; 26.5% received a platinum agent. The IDMC recommended data unblinding as IA showed a significant benefit of OL vs PL for IDFS (hazard ratio [HR] 0.58; 99.5% CI 0.41, 0.82; P < 0.0001) at 2.5 yrs median follow-up. IDFS events occurred in 106/921 and 178/915 pts assigned to OL and PL, respectively. 3-yr IDFS was 85.9% vs 77.1% (diff. 8.8%; 95% CI 4.5%, 13.0%). DDFS was significantly improved with OL (HR 0.57; 99.5% CI 0.39, 0.83; P< 0.0001); 3-yr DDFS was 87.5% vs 80.4% (diff. 7.1%; 95% CI 3.0%, 11.1%). OS was greater for OL than PL but was not statistically significant at IA (HR 0.68; 99.0% CI 0.44, 1.05; P = 0.024); 3-yr OS% 92.0% vs 88.3% (diff. 3.7%; 95% CI 0.3%, 7.1%). Median intended OL exposure was 94.8%. AEs were consistent with the label. G3+ AEs in >1% of OL pts were; anemia (8.7%), neutropenia (4.8%), leukopenia (3.0%), fatigue (1.8%), and lymphocytopenia (1.2%). SAEs and AESI were not increased by OL, SAE 8.7% vs 8.4% and AESI 3.3% vs 5.1%, OL vs PL respectively.

Conclusions: Adjuvant OL following ACT or NACT significantly improved IDFS and DDFS with acceptable toxicity in pts with gBRCAm and high-risk HER2-negative EBC.

Clinical trial information: NCT02032823