Molecular Subtypes of Breast Cancer

This comes from gene expression profiling (the Perou/Sørlie work) and is now approximated clinically using IHC surrogates (ER, PR, HER2, Ki-67):

Luminal A (~40%): ER+, PR+ (usually high), HER2−, low Ki-67. The most indolent subtype. Driven primarily by oestrogen signalling. Good prognosis, responds well to endocrine therapy (tamoxifen, aromatase inhibitors). Many won't benefit from chemotherapy — genomic assays like Oncotype DX help stratify this.

Luminal B (~20%): ER+, PR variable (often low), HER2− or HER2+, high Ki-67. More proliferative than Luminal A, worse prognosis. More likely to benefit from chemotherapy in addition to endocrine therapy. If HER2+, gets trastuzumab as well.

HER2-enriched (~15%): ER−, PR−, HER2+. Driven by ERBB2 amplification on chromosome 17q, leading to constitutive activation of RAS-MAPK and PI3K-AKT signalling. Historically aggressive, but prognosis has been transformed by HER2-targeted therapies — trastuzumab, pertuzumab, TDM-1, trastuzumab deruxtecan (an antibody-drug conjugate that's been a game-changer even in HER2-low disease).

Triple-negative / Basal-like (~15%): ER−, PR−, HER2−. Express basal/myoepithelial markers (CK5/6, EGFR). Strong association with BRCA1 mutations — defective homologous recombination leads to genomic instability and reliance on alternative DNA repair pathways. This is therapeutically exploitable with PARP inhibitors (olaparib, talazoparib) which cause synthetic lethality — if you knock out PARP-mediated base excision repair in a cell that already can't do homologous recombination, the cell dies. Immune checkpoint inhibitors (pembrolizumab) also have a role here given higher tumour mutational burden and PD-L1 expression. Worst prognosis of the subtypes, tends to metastasise viscerally (lung, brain) rather than to bone.