A gene expression-based classifier for HER2-low breast cancer.

Area of Science:

• Oncology
• Genomics
• Biomarker Discovery

Background:

• HER2-low breast cancer is crucial for antibody-drug conjugate (ADC) therapy selection.
• Immunohistochemistry (IHC) scoring for HER2-low status (1+ or 2+ without amplification) faces inter-observer variability in clinical practice.

Purpose of the Study:

• To develop and validate a gene expression profiling-based signature for objective identification of HER2-low breast cancer.
• To establish an observer-independent method for HER2-low tumor classification.

Main Methods:

• A discovery approach using gene expression profiling on institutional datasets (INT1, INT2) identified differentially expressed genes (DEGs) across HER2 IHC categories.
• Principal Component Analysis generated a 20-gene HER2-low signature.
• The signature's performance was validated in independent datasets (INT3, TCGA, GSE81538) using statistical tests and receiver operating characteristic (ROC) analysis.

Main Results:

• A 20-gene HER2-low signature was identified, enriched in lipid metabolism, steroid metabolism, peptidase regulation, and immune response pathways.
• The signature exhibited a bell-shaped distribution across IHC categories, accurately distinguishing HER2-low (1+, 2+) from HER2 0 and 3+ tumors (p < 0.001).
• The signature achieved an AUC of 0.74 in distinguishing HER2-low tumors, outperforming individual ERBB2 mRNA analysis (AUC 0.52).

Conclusions:

• The developed 20-gene signature provides a robust, observer-independent method for classifying HER2-low breast cancer.
• This gene expression-based approach shows promise for improving patient selection for HER2-targeted therapies, including ADCs.
• The signature effectively differentiates HER2 0 from HER2-low expressing tumors, particularly those with IHC 1+ status.