Publication: ASCO® 2025

Authors: Laila Samiian, Adam Brufsky, Sahra Uygun, Isha Kapoor, Victoria Poillucci, Joyce O’Shaughnessy, William Audeh

Title:Molecular Insights into HR+/HER2+Early-Stage Breast Cancer: Neoadjuvant Therapy Responses by MammaPrint and BluePrint genomic subtypes

Background:

Clinical HER2+ (cHER2+) early breast cancer (EBC) represents 15-20% of invasive EBC and is typically treated with Neoadjuvant HER2-targeted therapy (NHT) combined with chemotherapy, regardless of ER status. NBRST and I-SPY2 trials showed varied NHT responses in cHER2+ tumors based on genomically-defined molecular subtypes, emphasizing the importance of understanding tumor biology. Genomic assays MammaPrint® (MP) and BluePrint® (BP) predict therapy response and inform treatment decisions. Here, we explored the biological pathways underlying differential NHT response in triple positive (HR+HER2+) tumors using whole transcriptome analysis (WTA).

Methods:

Patients with HR+/HER2+ early-stage breast cancer treated with NHT (N=720) were included from FLEX (NCT03053193). MP classified tumors as UltraLow (UL), Low, High 1, or High 2 Risk, while BP categorized them as Luminal A, Luminal B, HER2, or Basal. Differences in clinical characteristics and pathological complete response (pCR) rates were assessed by Chi-Square or Fisher’s exact tests and proportional Z-test, respectively. Differential gene expression analysis of WT profiles was performed between tumors with and without pCR, using limma package in R, followed by pathway enrichment analysis in Metascape.

Results:

Among 720 HR+/HER2+ EBCs, MP classified 19 (2.6%) as UL, 107 (14.9%) as Low, 385 (53.5%) as High 1, and 209 (29.0%) as High 2. BP classified 120 (16.7%) as Luminal A, 307 (42.6%) as Luminal B, 278 (38.6%) as HER2, and 15 (2.1%) as Basal. Compared to other BP subtypes (Luminal A/B), BP HER2 tumors were associated with younger age (54 vs 60, p<0.001), premenopausal status (p=0.002), higher grade (G3: 54.7%, p<0.001), and T3 tumors (10.7% vs 3-4%, p<0.001). pCR rates with NHT were higher in BP HER2 tumors compared to Luminal A/B tumors (n=41, 61.2% vs n=18, 26.5%, respectively, p<0.001). WTA of BP HER2 tumors with pCR showed 23 genes with 2-fold change (not statistically significant after correction), 20 of which were upregulated and associated with regulation of bone morphogenic protein encoding genes and increased cell-substrate/cell matrix adhesion, compared to tumors that had residual disease.

Conclusions:

These data show heterogeneity within HR+/HER2+ tumors, with approximately 60% genomically reclassified as non-HER2-type by BP. Consistent with I-SPY2, BP HER2 cancers showed higher pCR rates than Luminal A/B, suggesting that additional therapeutic strategies are needed to increase the pCR rates in these cancers. Although WTA in BP HER2 tumors with and without pCR identified DGE, the findings were not statistically significant. Future analyses of WTA in larger numbers of BP subtypes within the HR+ HER2+ EBC patients who are being enrolled on the FLEX trial may elucidate the biology of the cancers with pCR vs non-pCR.