HER2 Scoring

The ASCO/CAP 2018 guideline defines four discrete scores based on staining intensity and completeness of membrane staining:

• Score 0 (Negative): No staining observed, or faint, incomplete membrane staining in ≤10% of invasive tumor cells.
• Score 1+ (Negative): Faint, barely perceptible incomplete membrane staining in >10% of tumor cells.
• Score 2+ (Equivocal): Weak-to-moderate complete membrane staining in >10% of cells, or strong complete staining in ≤10%. Requires reflex in situ hybridization (ISH) testing.
• Score 3+ (Positive): Intense, uniform circumferential membrane staining in >10% of invasive tumor cells. Patient is eligible for anti-HER2 therapy.

Accurate scoring depends on evaluating only the invasive component of the tumor and assessing the circumferential completeness of the dark brown diaminobenzidine (DAB) chromogen precipitate on the cell membrane. Key rules include:

• Ignore cytoplasmic staining: Only linear membrane signal is diagnostic.
• Exclude in situ carcinoma: Ductal carcinoma in situ (DCIS) with strong HER2 staining is not scored.
• Assess heterogeneity: Document if distinct areas of amplified and non-amplified tumor exist, as this may indicate genetic heterogeneity requiring ISH confirmation.
• Use internal controls: Normal breast epithelium should be negative (0), serving as a built-in negative control.

Deep learning models, often built on U-Net or Vision Transformer (ViT) architectures, automate the objective quantification of HER2 IHC. These algorithms perform semantic segmentation to identify invasive tumor regions, then classify each cell's membrane staining completeness and intensity. The output is a continuous H-score (range 0-300) calculated as: (1 × % cells 1+) + (2 × % cells 2+) + (3 × % cells 3+). This reduces inter-observer variability, which is particularly problematic for discriminating 1+ from 2+ cases, and provides a standardized digital readout for pathologist review.

The binary determination of HER2 positivity (3+ by IHC or amplified by ISH) is a predictive biomarker that unlocks a specific therapeutic class. HER2-positive patients receive monoclonal antibodies (trastuzumab, pertuzumab) and antibody-drug conjugates (trastuzumab emtansine, trastuzumab deruxtecan). Conversely, a false-positive score exposes a patient to cardiotoxic therapies with no benefit, while a false-negative score denies access to life-prolonging treatment. The recent recognition of HER2-low (IHC 1+ or 2+/ISH-negative) as a distinct therapeutic category for trastuzumab deruxtecan has made the accurate discrimination of low-level expression critically important.

Scoring fidelity is highly sensitive to the pre-analytical phase. The ASCO/CAP guidelines specify strict protocols:

• Cold ischemic time: Must be documented and kept under 1 hour.
• Fixation: 10% neutral buffered formalin for 6-72 hours. Under-fixation causes false-negative edge artifact; over-fixation masks epitopes.
• Tissue processing: Incomplete dehydration or paraffin infiltration degrades morphology.
• Validation: Every new antibody lot must be validated against a known standard on a tissue microarray (TMA) containing cell lines with defined HER2 expression levels.

The foundational laboratory technique underlying HER2 scoring. IHC uses antibodies directed against the HER2 receptor protein, visualized through an enzymatic color reaction that produces a brown precipitate on the tumor cell membrane. The intensity and completeness of this membranous staining form the basis of the 0 to 3+ scoring system. Standardization requires rigorous positive and negative control tissues on every staining run to ensure assay validity.

A reflex testing method used when IHC yields an equivocal (2+) result. FISH quantifies HER2 gene amplification by counting the ratio of HER2 gene signals to chromosome 17 centromere signals (CEP17) within tumor nuclei. A HER2/CEP17 ratio ≥ 2.0 or an average HER2 copy number ≥ 6.0 signals defines amplification. This orthogonal DNA-level assessment resolves the ambiguity of moderate protein expression.

The 2+ IHC category represents a diagnostic gray zone where protein overexpression is ambiguous. Approximately 15–20% of breast cancers fall into this category. Resolution pathways include:

• Single-probe FISH: Counts HER2 signals only
• Dual-probe FISH: Calculates HER2/CEP17 ratio
• Alternative chromosome 17 probes: Used when CEP17 abnormalities are suspected
• mRNA-based testing: Quantifies HER2 transcript levels as a tiebreaker Accurate resolution is critical to avoid denying effective therapy or administering cardiotoxic treatment unnecessarily.

An emerging clinical category encompassing tumors with IHC 1+ or IHC 2+/FISH-negative results. Historically classified as HER2-negative, these patients were ineligible for HER2-directed therapy. The DESTINY-Breast04 trial demonstrated that antibody-drug conjugates (ADCs) like trastuzumab deruxtecan provide significant survival benefit in this population. This has redefined scoring as a three-tier system: HER2-positive, HER2-low, and HER2-null (IHC 0).

Deep learning algorithms that automate membrane staining quantification from whole-slide images. These systems perform semantic segmentation of invasive tumor regions and classify individual cells based on staining completeness and intensity. Benefits include:

• Continuous scoring: Generates a quantitative H-score rather than discrete categories
• Inter-observer reduction: Eliminates variability between pathologists
• Equivocal triage: Flags borderline cases for molecular confirmation Models are trained using multiple instance learning with slide-level ASCO/CAP labels as weak supervision.