Abstract
ABSTRACT
Background
Estrogen receptor (ER)‐positive/HER2‐positive breast cancer demonstrates significantly lower rates of pathological complete response (pCR) to neoadjuvant HER2‐targeted therapies compared to ER‐negative/HER2‐positive disease. However, the optimal ER positivity cutoff for clinically meaningful patient stratification remains undefined.
Methods
We analyzed a retrospective cohort of 741 HER2‐positive breast cancer patients treated with neoadjuvant chemotherapy plus dual HER2 blockade (trastuzumab and pertuzumab) at Sun Yat‐sen University Cancer Center. The optimal ER cutoff for predicting pCR was determined by ROC analysis with Youden index and validated by bootstrap resampling (1000 iterations). Transcriptomic data from TCGA and SCAN‐B cohorts were analyzed to characterize biological differences between ER subgroups. Drug response was predicted using the oncoPredict algorithm, and cancer dependency was assessed using DepMap CRISPR screening data.
Results
ER ≥ 50% positivity was identified as the optimal predictive cutoff (bootstrap 95% CI: 25.0%–77.5%) and was an independent predictor of significantly lower pCR rates in multivariate analysis (OR = 0.27; 95% CI: 0.19–0.40;
p
p
Conclusions
ER ≥ 50% positivity defines a clinically and biologically distinct HER2‐positive subgroup with poor response to standard neoadjuvant therapy and predicted resistance to ADCs, consistent with lower HER2 and CD16A expression in this subgroup. The convergent transcriptomic and functional evidence for ESR1‐CCND1 axis activation provides mechanistic support for combining CDK4/6 inhibitors with endocrine and anti‐HER2 therapies to improve outcomes in this resistant subgroup.