Adjuvant Olaparib Benefit Persists at 6.1 Years, Including Overall Survival, in Germline BRCA-Associated High-Risk HER2-Negative Early Breast Cancer: OlympiA Update

Why This Study Matters

Germline BRCA1 and BRCA2 pathogenic variants impair homologous recombination repair, which makes tumour cells vulnerable to PARP inhibition. Olaparib is an oral PARP inhibitor, and OlympiA tested whether one year of it after standard (neo)adjuvant chemotherapy improves outcomes in high-risk early breast cancer. Earlier analyses had met the pre-specified boundaries for IDFS and DDFS (interim analysis 1, median follow-up 2.5 years) and for OS (interim analysis 2, median follow-up 3.5 years). This report asks whether the benefit lasts, and whether longer-term toxicity, especially myelodysplastic syndrome and acute myeloid leukaemia (MDS/AML), has emerged.

Study Design

OlympiA is a phase III, randomised, double-blinded, placebo-controlled trial. Participants received 1 year of oral olaparib 300 mg twice daily or matching placebo. Randomisation was stratified by hormone receptor status, neoadjuvant versus adjuvant chemotherapy, and platinum therapy for the current breast cancer. The primary endpoint was invasive disease-free survival (IDFS); secondary endpoints included distant disease-free survival (DDFS) and overall survival (OS). Efficacy analyses used the intention-to-treat population (stratified Cox model, Kaplan-Meier estimates). Because significance boundaries were already crossed at earlier analyses, this third analysis (IA3) is descriptive, with no further significance testing.

Patient Population

Eligible patients had high-risk, HER2-negative primary breast cancer with a germline BRCA1/2 pathogenic or likely pathogenic variant, and had received at least six cycles of taxane- and/or anthracycline-containing chemotherapy. Triple-negative patients qualified with residual disease after neoadjuvant chemotherapy or node-positive (or node-negative with tumour >2 cm) disease after adjuvant chemotherapy; hormone receptor-positive patients needed a CPS+EG score of 3 or more after neoadjuvant chemotherapy, or 4 or more positive nodes at surgery.

Source: Table 1 of the publication. Percentages may not total 100 because of rounding.

Primary Endpoint Results

With 2.6 additional years of follow-up since the second analysis, the IDFS benefit was maintained: 6-year IDFS was 79.6% with olaparib versus 70.3% with placebo (difference 9.4%; 95% CI 5.1% to 12.7%), HR 0.65 (95% CI 0.53 to 0.78). All first IDFS event frequencies, including local, regional, distant and central nervous system recurrences, remained lower in the olaparib group. For comparison, the first interim analysis reported IDFS HR 0.58 (99.5% CI 0.41 to 0.82) at a median follow-up of 2.5 years, as summarised in the publication.

Secondary Endpoints

Risk of MDS/AML remains an important consideration in the adjuvant treatment of early BC, especially where chemotherapy has been used.
— Garber et al., Discussion, Ann Oncol 2026 (BC = breast cancer; MDS/AML = myelodysplastic syndrome/acute myeloid leukaemia)

Subgroup Analyses

The authors report that subgroup analyses of IDFS across the stratification groups (neoadjuvant versus adjuvant chemotherapy, prior platinum, hormone receptor status) and gBRCA1 versus gBRCA2 were "generally consistent" with the overall result without evidence of heterogeneity, and that DDFS and OS subgroup analyses were similarly consistent. The subgroup hazard ratios are displayed in Figure 2 and Supplementary Appendix Tables S5 to S7 of the paper and are not reproduced in the main text.

Among 1,176 patients without prior bilateral mastectomy, the competing-risks HR for contralateral invasive breast cancer was 0.83 (95% CI 0.52 to 1.32). Fewer BRCA-associated new invasive breast cancers were reported after randomisation with olaparib than placebo (34 versus 42).

Safety Profile

Exposure data and on-treatment adverse events are unchanged since the second analysis, because all patients had completed the protocol-specified course by then. Adverse events of special interest (AESIs) were not more frequent with olaparib.

Source: Table 2 of the publication. Pancreatic cancers were not lower with olaparib (3 versus 1), per the Discussion.

Interpretation and Broader Context

The authors conclude that, at 6.1 years median follow-up, one year of adjuvant olaparib continues to show clinically meaningful improvements in IDFS, DDFS and OS, with acceptable toxicity and no increased MDS/AML risk. They caution that continued surveillance is needed, since resistance to PARP inhibitors has been observed in the metastatic setting, and that the efficacy of adding olaparib to pembrolizumab in early breast cancer is not established and lacks regulatory approval. They also note that changes in other standards of care (for example capecitabine and neoadjuvant pembrolizumab in triple-negative disease) have influenced olaparib use. Follow-up is planned until 2029.

Limitations

This third interim analysis is descriptive: the pre-specified significance boundaries were already crossed at earlier analyses, so no further significance testing was performed. The trial excluded HER2-positive disease, so the results should not be extended beyond the studied population. Subgroup hazard ratios are reported only in a figure and supplementary tables, and the authors describe them as generally consistent rather than testing them formally in the main text. The authors also note that efficacy of olaparib combined with pembrolizumab in early breast cancer is not established, that changes in other standards of care (capecitabine, neoadjuvant pembrolizumab in triple-negative disease) have influenced olaparib use, and that follow-up continues until 2029, so long-term toxicity such as MDS/AML remains under surveillance. This summary could not establish outcomes in those settings.