Short-term risk of recurrence in patients (pts) with HR+/HER2− early breast cancer (EBC) treated with endocrine therapy (ET) in randomized clinical trials (RCTs): A meta-analysis.

541 Background: ET ± chemotherapy (CT) has been standard of care for HR+/HER2− EBC for ≈20 yrs, yet pts continue to experience disease recurrences up to decades after completing adjuvant (adj) treatment (tx). There is little information on early recurrences (≤5 yrs) in the more recent tx landscape, which includes optimized local tx and improved systemic tx. This meta-analysis of ET arms in RCTs in pts with HR+/HER2− EBC seeks to fill this gap by evaluating 3- and 5-yr risk of recurrence (RoR). Methods: A systematic literature review identified RCTs that met study criteria: phase 3 trials that included pts with HR+/HER2− EBC who received 5 yrs of adj ET (aromatase inhibitors ± ovarian suppression or tamoxifen) ± CT with invasive disease-free survival (iDFS) or disease-free survival (DFS) reported at 3 and/or 5 yrs post tx randomization. Neoadj or extended ET trials were excluded. Survival curves were digitized for iDFS estimates. Meta-analysis of proportions, using a random-effects model with double arcsine transformation, was used to evaluate pooled 3-yr RoR (100 − iDFS/DFS rate) for all pts, by lymph node (LN) status (N0: 0, N1: 1-3, N+ any positive LN), and 5-yr RoR where available. The 3-yr RoR across ET arms in a subset of only adj cyclin-dependent kinase 4/6 inhibitor (CDK4/6i) trials was evaluated given the recency of these RCTs. Results: RoR estimates were obtained from 14 trials (across 23 distinct ET-only arms), including 4 adj CDK4/6i trials that met study criteria; studies included pts with stage I-III disease ± nodal involvement. Pooled 3- and 5-yr RoR for the overall HR+/HER2− EBC pt population regardless of LN/stage were 7% and 12%, respectively. In trials with nodal status subgroups, pooled 3-yr RoR was 5% for N0, 6% for N1, and 13% for N+ pts. In a sensitivity analysis excluding cohorts with low to medium risk scores based on genomic testing, pooled 3-yr RoR estimates for N0 and N1 subgroups were 7% and 10%, respectively. Meta-analysis across ET arms of CDK4/6i trials revealed 3-yr RoR of 15% (all stage I-III) and 10% (N0). Conclusions: Despite the potential for late recurrence of HR+/HER2− EBC, short-term RoR is considerable and accumulates over time for these pts treated with ET ± CT in the adj setting. These findings highlight the unmet need for tolerable tx escalation among at-risk pts, including node-positive and select N0 pts with high-risk features.