Breast cancer classification

Editor-In-Chief: C. Michael Gibson, M.S., M.D. [1] Associate Editor(s)-in-Chief: Mirdula Sharma, MBBS [2] Soroush Seifirad, M.D.[3] Nehal Eid, M.D.[4]

Breast cancer classification uses three complementary axes: histopathologic type, biomarker-defined clinical subtype, and intrinsic gene-expression subtype. These categories describe different features and are not interchangeable.^[1][2]

Breast epithelial lesions are classified as in situ or invasive. In situ categories include Ductal carcinoma in situ (DCIS) and lobular neoplasia, including Lobular carcinoma in situ (LCIS). Classic LCIS is generally regarded as a risk marker/non-obligate precursor rather than invasive carcinoma.^[1]

Invasive carcinomas are classified by morphology:

• Invasive carcinoma of no special type (IBC-NST): the most common invasive histotype; formerly termed invasive ductal carcinoma.
• Invasive lobular carcinoma (ILC): commonly shows discohesive, single-file growth. Loss of E-cadherin supports lobular differentiation but is not required for diagnosis.
• Special histologic types: include tubular, cribriform, mucinous, micropapillary, apocrine, metaplastic, adenoid cystic, secretory, neuroendocrine, tall cell carcinoma with reversed polarity, and salivary gland-type carcinomas.

A special histologic type generally requires at least 90% of the tumor to show that morphology; otherwise, report a mixed IBC-NST and special-type carcinoma. Core biopsy may not establish whole-tumor purity.^[3][4]

Invasive carcinoma grade is reported separately from histologic type. The Nottingham (Elston-Ellis) system scores tubule formation, nuclear pleomorphism, and mitotic count to assign grade 1–3.^[3] DCIS is classified by nuclear grade as low, intermediate, or high.^[4]

The 2026 WHO classification recognizes ILC with extracellular mucin and updates criteria for mucinous and neuroendocrine carcinoma, among other refinements.^[5]

ER, PR, and HER2 should be assessed on every new primary and newly metastatic invasive breast cancer using validated ASCO/CAP methodology.^[6][7][4] ER and PR are positive when at least 1% of tumor-cell nuclei stain. HER2 status is determined by validated immunohistochemistry (IHC) and in situ hybridization (ISH) criteria.^[6][7]

Frequency estimates vary by cohort. HER2-low and HER2-ultralow are cross-cutting expression categories and are not mutually exclusive with HR status.^[1][4]

An ER result of 1–10% is reported as ER-low-positive; results below 1% are ER-negative.^[6] ER-low tumors are heterogeneous, and some have biology resembling ER-negative/TNBC disease. The endocrine benefit remains uncertain: the ASCO/CAP-recommended comment notes limited data, while recent observational meta-analyses associate endocrine therapy with improved outcomes. These findings do not establish the magnitude of benefit for an individual patient.^{[6][8][9][10]}

HER2 IHC scores are 0, 1+, 2+, and 3+. IHC 2+ is equivocal and requires reflex ISH testing; amplification establishes HER2-positive status.^[7][4]

HER2-low generally denotes IHC 1+ or IHC 2+/ISH-negative disease; HER2-ultralow denotes IHC 0 with faint, incomplete membrane staining in ≤10% of tumor cells. These are expression/reporting categories within HER2-negative disease, not separate intrinsic gene-expression subtypes. Reported estimates suggest HER2-low expression occurs in roughly half of breast cancers, and about 88% of HER2-low tumors are HR-positive; estimates vary by cohort and assay.^[11][12][13]

Document the specific IHC score and staining pattern, including the distinction between HER2-null (IHC 0/absent staining) and HER2-ultralow (IHC 0+/faint incomplete staining in ≤10%). This distinction can affect biomarker eligibility for trastuzumab deruxtecan in certain metastatic settings; HER2-null does not meet the HER2-low/ultralow expression criterion for that indication.^[4][14][15]

Scoring near the IHC 0/1+ boundary has limited reproducibility in reported cohorts. One report found a change between IHC 0 and HER2-low in approximately 40% of paired primary/metastatic samples; in an expert re-review series, approximately 55% of initially IHC 0 cases were reclassified as HER2-ultralow.^[7][16]

Re-review with consideration of repeat or consultative HER2 testing is recommended when a grade 1 tumor of any histologic type, or a pure mucinous, tubular, or cribriform carcinoma, tests HER2-positive, because the result is biologically unexpected.^[4]

Gene-expression classifiers such as PAM50 define intrinsic molecular subtypes. They describe tumor biology and are distinct from the IHC-defined clinical categories.^[2][1]

IHC-based surrogate schemes may subdivide HR-positive/HER2-negative cancers into luminal A-like (lower proliferation) and luminal B-like (higher proliferation) groups. These are approximations and do not reproduce PAM50 subtyping exactly. No standardized Ki-67 cutoff separates luminal A-like from luminal B-like; 14% and 20% thresholds have both been used. Ki-67 in an intermediate range should not be used as the sole basis for assigning a subtype.^[17][18]

Clinical and intrinsic molecular subtypes are not interchangeable; a meaningful minority of tumors are discordant between IHC-based classification and gene-expression classification.^[19][20]