Impact of age on breast cancer outcomes in the neoadjuvant I-SPY2 trial.

Abstract

592 Background: Historically, young women with early-stage breast cancer (EBC) have had worse outcomes than older women. This is based largely on population-based or adjuvant studies, with limited neoadjuvant data. Methods: Patients (pts) with Stage 2-3 EBC treated with neoadjuvant chemotherapy +/- immunotherapy (Cx-IT) on the I-SPY2 trial between 2010-2022 were included (data cutoff 07/2025). Pts were stratified by age at trial screening (<40 vs ≥40 yrs), and clinical/molecular subtypes, including ImPrint, an immune signature predictive of Cx-IT response. Pathologic complete response (pCR), residual cancer burden (RCB), and event-free survival (EFS) were evaluated by age and disease subtypes using Kaplan-Meier methods and Cox proportional hazard ratios (HR). Results: Among 2117 evaluable pts, 510 (24%) were <40 yrs and 1607 (76%) were ≥ 40 yrs. Overall, 43% of tumors were HR+/HER2-, 35% HR-/HER2- (TN), 22% HER2+. 34% / 66% were ImPrint +/- (27% / 73% for HR+; 44% / 56% for HR-), with similar rates of ImPrint+/- between age groups. Rates of pCR (31% vs 35 %) and RCB 0/1 (48% vs 51%) were similar between pts <40 and ≥40. Overall, pts < 40 had worse EFS (p < 0.01) vs pts ≥ 40; however, in pts who achieved pCR, there was no difference in EFS in all biologic subgroups (HR 0.90, p=0.77). In contrast, in pts with residual disease, those <40 had worse EFS than pts ≥40 (HR 0.72, p<0.01). The worse EFS in pts<40 with residual disease was limited to pts with ImPrint- disease, in both HR+/HER2- (HR 0.64, p=0.02) and TN (HR 0.62, p=0.04) subtypes, where response rates were lower. In ImPrint+ pts, no differences in EFS by age were observed (HR 1.01, p=0.98). Similar results were seen for pts with HR+/HER2- disease when stratifying by RCB (RCB 0/I vs 2/3). There was no difference in EFS by age for pts with RCB 0/1 disease in any subtype (HR 0.66, p=0.09). For ImPrint-, pts<40 with RCB 2/3 disease had worse EFS than pts ≥40 (HR 0.61, p=0.01). For HR+/HER2-, 64% of tumors were MammaPrint (MP) High 1 vs 36% MP High 2, and 74% were Luminal. When stratifying by pCR or RCB 0/1, similar age-related EFS differences were observed in HR+/HER2- Luminal or MP High 1 tumors, as in ImPrint- tumors. Conclusions: Overall, young pts <40 yrs with high-risk EBC in the I-SPY2 trial had worse 5-yr EFS than pts ≥40; however, when taking response into account, there were no age-related EFS differences for pts who had an excellent response to neoadjuvant Cx-IT (i.e. pCR or RCB 1). The EFS difference was driven by pts with residual disease, specifically in the subset with ImPrint- disease, which is associated with less response to Cx-IT. Whereas, for pts with ImPrint+ disease, there was no difference in EFS by age, regardless of pCR or RCB. This highlights the unmet need for more effective therapies for immune-negative tumors. Our results reinforce the importance and prognostic value of neoadjuvant therapy, particularly in our youngest pts, to better risk-stratify and optimize treatment based on tumor biology and response to therapy. Clinical trial information: NCT01042379 .