Synthesis: IHC for ER/PR/HER2 and the TNBC definition

Evidence grades (GRADE-adapted): A high — multiple well-conducted RCTs or systematic reviews converge. B moderate — single pivotal RCT or consistent observational evidence. C limited — single observational study, mechanistic, or expert consensus. D preclinical / hypothesis-generating.

What the definition actually says

Triple-negative breast cancer is defined operationally by three immunohistochemical assays on biopsy or surgical resection tissue:

• Estrogen receptor (ER): < 1% of invasive tumor cell nuclei stain positive on a validated IHC assay^[1]A.
• Progesterone receptor (PR): < 1% of invasive tumor cell nuclei stain positive^[1].
• HER2: IHC score 0 or 1+, or IHC 2+ with negative ERBB2 in-situ hybridization (FISH or other validated ISH)^[2]A.

Each of those thresholds is the product of explicit consensus-building over the past two decades. None is biologically obvious; each represents a clinical compromise between sensitivity (don't deprive responsive patients of effective targeted therapy), specificity (don't subject non-responsive patients to ineffective therapy and its side effects), and reproducibility (the result has to be the same across labs and pathologists). The history of how these thresholds moved is the history of how the TNBC label itself moved.

The ER 1% threshold — history and rationale

Prior to 2010 the ASCO/CAP recommendation defined ER-positivity as ≥ 10% of tumor nuclei staining positive, a threshold inherited from the early Allred-scoring era^[3]B. Tumors with 1–9% ER staining occupied an ambiguous “low-positive” zone treated by some clinicians as endocrine-responsive and by others as functionally ER-negative.

The 2010 ASCO/CAP guideline lowered the threshold to 1%, reasoning that the tamoxifen-response evidence from BIG 1-98 and other adjuvant trials extended down to very low ER expression and that the cost of false-negative classification (denying a patient potentially effective endocrine therapy) outweighed the cost of false-positive classification (subjecting a patient to typically well-tolerated adjuvant endocrine therapy that turned out not to help)^[4]A. This change had immediate implications for the TNBC label: a meaningful subset of tumors previously called triple-negative were reclassified as low-ER-positive and therefore endocrine-therapy candidates.

Subsequent empirical work substantiated the 1% threshold's clinical relevance while also exposing its biological awkwardness. Iwamoto and colleagues showed that tumors with 1–9% ER staining had ESR1 mRNA expression profiles intermediate between ER-zero and ER-high tumors, but with substantial overlap in both directions — the 1–9% group is biologically heterogeneous, with some tumors behaving like ER-low-positive and others like ER-zero^[5]B. Yi et al. retrospectively analyzed 9,639 breast cancer patients and found that the 1–9% ER subset had survival outcomes more similar to the ER-zero group than to the ER-high group, with endocrine therapy response rates in line with that observation^[6]B.

The 2020 Allison et al. ASCO/CAP update retained the 1% threshold but added a new recommendation: tumors with 1–10% ER expression should be reported with a comment noting that the data on endocrine therapy benefit in this range is limited and that the result should be interpreted in clinical context^[1]. The category is sometimes called “ER-low-positive” in pathology reports; clinically it is closer to TNBC than to ER-high disease, and many investigators argue it should be enrolled in TNBC clinical trials rather than HR+ trials. This is an active and unresolved debate.

PR — independent of ER or coupled to it?

PR expression is downstream of ER signaling, and tumors that are ER-positive are usually PR-positive as well; the ER+/PR− subset (roughly 10% of ER+ disease) is biologically and prognostically distinct from ER+/PR+ disease^[7]B. For the TNBC definition specifically, the PR–negative requirement is rarely the binding constraint — most ER–negative tumors are also PR–negative. ER–/PR+ tumors do occur but are rare (often attributed to ER assay false-negativity rather than truly autonomous PR signaling)^[1]; current ASCO/CAP guidance recommends repeat testing in that scenario.

The 1% threshold applies identically to PR per the 2020 ASCO/CAP update; the prior 10% rule was abandoned for the same reason as for ER^[1].

HER2 testing and the 2018/2023 updates

The HER2 algorithm is more involved than ER/PR because it combines two assay technologies (IHC for protein, ISH for gene amplification) with explicit decision points where one informs the other.

The 2018 ASCO/CAP focused update (Wolff et al.) clarified the algorithm in the wake of the 2013 update's introduction of the "equivocal" category, which had been operationally unworkable for many labs^[2]. The 2018 algorithm:

• IHC 0 or 1+: HER2-negative. No further testing.
• IHC 2+: reflex to ISH (FISH or other validated ISH). If ISH-positive (HER2:CEP17 ratio ≥ 2.0 OR HER2 copy number ≥ 6.0), HER2-positive; if ISH-negative, HER2-negative.
• IHC 3+: HER2-positive. Confirmatory ISH no longer required as standard.

The 2023 ASCO/CAP update (Wolff et al.) added explicit guidance for the HER2-low scoring zone (IHC 1+, or IHC 2+/ISH-negative) following the regulatory approval of trastuzumab deruxtecan for HER2-low disease^[8]A. Critically, the 2023 update did not change the analytical scoring criteria; it added clinical-decision-support language so that pathologists explicitly report IHC 1+ and IHC 2+/ISH-negative as actionable categories, no longer collapsed into “HER2-negative”.

HER2-low: a clinically actionable subset of historically TNBC tumors

Until 2022, HER2-low tumors (IHC 1+ or IHC 2+/ISH-negative) were treated identically to HER2-zero tumors (IHC 0). Both were “HER2-negative”; HER2-targeted therapy was reserved for ISH-amplified or IHC 3+ disease. DESTINY-Breast04 changed that.

Modi and colleagues randomized 557 patients with previously treated HER2-low metastatic breast cancer (either HR+ or HR−) to trastuzumab deruxtecan vs physician's choice of chemotherapy. In the HR–negative cohort — the relevant subset for the TNBC population — median progression-free survival was 8.5 months with trastuzumab deruxtecan vs 2.9 months with chemotherapy (HR 0.46) and median overall survival was 18.2 vs 8.3 months^[9]A. The FDA approved trastuzumab deruxtecan for HER2-low metastatic breast cancer (HR+ or HR−) in August 2022.

This split the historically TNBC population into two new diagnostic groups:

• HER2-low TNBC — IHC 1+ or 2+/ISH−, eligible for trastuzumab deruxtecan after prior chemotherapy in the metastatic setting.
• HER2-zero (or HER2-ultralow) TNBC — IHC 0 or 0-with-faint-incomplete-membrane staining; not currently eligible for trastuzumab deruxtecan but under investigation in DESTINY-Breast06.

Tarantino and colleagues estimate that approximately 45–55% of historically TNBC tumors fall into the HER2-low category under modern IHC reading practices^[10]B. The implication: nearly half of patients previously told they had TNBC actually have HER2-low TNBC and will be candidates for an additional line of HER2-directed therapy when their disease progresses. This is the most clinically consequential change to the TNBC IHC framework in the past five years.

Preanalytical and analytical sources of variability

ER/PR/HER2 IHC results are sensitive to factors outside the analytical assay itself. The 2020 and 2023 ASCO/CAP updates document the major sources:

• Cold ischemia time. Tissue must be placed in fixative within 1 hour of resection; longer warm-tissue time degrades ER/PR antigens and can produce false-negative or weakly-positive results^[1]. Core biopsies have shorter cold-ischemia exposure than surgical resections and are generally preferred for receptor testing.
• Fixation. 10% neutral buffered formalin for 6–72 hours is the standard. Underfixation (< 6 h) and overfixation (> 72 h) can both shift IHC results; severely overfixed tissue is the more common problem in routine practice^[1].
• Antibody clone and detection system. ER (most commonly SP1 or 6F11 clones), PR (1E2, 1A6, PgR636 clones), and HER2 (4B5, A0485, SP3 clones; HercepTest as a complete kit) all have validated alternatives but produce slightly different staining intensities at the same biological signal. Inter-laboratory comparisons consistently show non-trivial discordance, particularly at the low-end thresholds where the TNBC definition matters most^[11]B.
• Interobserver agreement. Inter-pathologist agreement is high for IHC 0 vs IHC 3+ (kappa > 0.85) but substantially lower for the IHC 0 vs IHC 1+ boundary and the IHC 1+ vs IHC 2+ boundary — the precise distinctions that now drive HER2-low eligibility^[12]B. Fernandez et al. reported that even after the HER2-low definition was operationalized, agreement on the IHC 0 vs IHC 1+ split remained around 26% across 18 pathologists scoring the same slides^[13]B.
• Sampling and intratumoral heterogeneity. A single biopsy may not capture intra-tumoral expression heterogeneity, particularly for HER2 where focal amplification has been documented in 10–15% of cases. Repeat testing on resection specimens occasionally reclassifies tumors that were equivocal on biopsy^[2].

The aggregate effect: the same physical tumor can be reproducibly classified as HER2-zero by one pathologist and HER2-low by another, especially in the IHC 0 vs 1+ range. This is the single most consequential reproducibility issue in current TNBC pathology, because the boundary now determines a meaningful treatment option.

Inter-observer agreement and the role of digital pathology / AI

Whole-slide-image digital pathology and AI-assisted scoring have advanced rapidly in this space. The motivating observations are: (1) human inter-observer agreement on HER2 0 vs 1+ scoring is poor, (2) the boundary now matters clinically, and (3) image-based classifiers can be trained on consensus-scored reference sets to produce reproducible calls.

Several groups have reported CNN-based HER2-IHC scoring models with reported per-slide accuracies of 85–90% relative to expert consensus, with the largest gains in the IHC 0 vs 1+ boundary where humans agree least^[14]C. None of these models is currently FDA-cleared as a standalone scoring device, but several have been incorporated into pathologist-assist workflows. Independent validation across diverse cohorts and staining platforms is the main gap before regulatory clearance is plausible.

Special histologies and the TNBC label

A small but important subset of breast tumors are triple-negative by IHC but biologically distinct from the typical high-grade invasive ductal NST (no special type) that dominates the TNBC clinical category^[15]B:

• Apocrine carcinoma — often AR-positive; clinically slower-growing than typical TNBC. Frequently maps to the LAR molecular subtype.
• Secretory carcinoma — defined by the ETV6–NTRK3 fusion; indolent clinical course; eligible for tropomyosin receptor kinase (TRK) inhibitor therapy (larotrectinib, entrectinib) in advanced disease.
• Adenoid cystic carcinoma — MYB–NFIB fusion-driven; very indolent natural history; treatment is closer to head-and-neck adenoid cystic disease than typical TNBC.
• Metaplastic carcinoma — spindle-cell, squamous, or matrix-producing morphology; aggressive clinical course; relatively chemo-resistant compared to typical TNBC.
• Low-grade triple-negative tumors (e.g., low-grade invasive ductal NST with TNBC IHC) — do exist and have a substantially more indolent course than typical high-grade TNBC; can be over-treated if managed identically to grade-3 disease.

These histologies together account for roughly 5–10% of clinically labelled TNBC cases. They are usually flagged on the pathology report by histologic-type designation, but the IHC-defined TNBC label can lead clinicians to default to TNBC trial enrollment and standard regimens that are not appropriate. The 2020 ASCO/CAP ER/PR update specifically notes that special histologic types may have different treatment implications and should not be assumed to follow the typical TNBC clinical pattern^[1].

The "triple-negative" label and what it isn't

It is worth restating explicitly because the implication is so often missed in clinical conversations: TNBC is a definition by what's absent on three IHC assays, not a positive biological characterization. The label collapses substantial molecular heterogeneity (see the Lehmann/Pietenpol subtypes synthesis), includes both highly aggressive grade-3 invasive ductal disease and relatively indolent special histologies, and is now operationally subdivided by HER2-low status into two clinically distinct subsets. Treatment selection should depend on the underlying biology and the HER2-low subdivision, not on the historical TNBC label alone — though in clinical practice that label remains useful as a starting heuristic for chemo-centric treatment planning.

Evidence table — defining the boundaries

Open questions and active investigation

• HER2-low scoring harmonization. The IHC 0 vs 1+ inter-observer agreement gap is the single most consequential reproducibility issue in current TNBC pathology. Digital-pathology AI assist is the most plausible near-term remediation; FDA-cleared assistive devices are likely within 2–3 years.
• HER2-ultralow. DESTINY-Breast06 examined trastuzumab deruxtecan in patients with IHC 0-with-faint-staining (“HER2-ultralow”), finding numeric benefit; whether this extends the actionable HER2-low concept further toward HER2-zero is the next regulatory question.
• 1–10% ER (ER-low) and TNBC trial enrollment. Whether ER-low (1–10%) tumors should be enrolled in TNBC trials or HR+ trials is unresolved. The biological data favor TNBC; the regulatory and clinical-trial-eligibility data are split. A formal cross-trial-eligibility recommendation from a consensus body would resolve enormous heterogeneity in current practice.
• Liquid-biopsy-based HER2 status. ctDNA-based HER2 status assessment is technically possible but has not been validated as a substitute for tissue IHC. Whether it could resolve the inter-observer-agreement problem (by removing the IHC scoring step entirely) is an open question and a major investigational direction.
• Companion diagnostic standardization for ADC-era TNBC. As more antibody-drug conjugates enter the TNBC landscape (HER3, Trop-2, others), the parallel growth in companion-diagnostic IHC assays will reproduce the HER2-low harmonization problem at each new target. The opportunity is to establish digital-pathology-assisted scoring as the default before the proliferation, rather than retrofitting it later.

For the patient-layer companion explaining the diagnostic workflow in plain language, see How TNBC is diagnosed. For the molecular-subtype frameworks that overlay this IHC-based definition, see the Lehmann/Pietenpol subtypes synthesis. For PD-L1 assay heterogeneity (a separate but conceptually parallel reproducibility story), see PD-L1 testing and assay heterogeneity.

References

Each citation links to the original publication via DOI. The same records are searchable in the evidence library by title or DOI.

1. Allison KH, Hammond MEH, Dowsett M, et al. Estrogen and Progesterone Receptor Testing in Breast Cancer: ASCO/CAP Guideline Update. J Clin Oncol. 2020;38(12):1346–1366. doi:10.1200/JCO.19.02309. ↩
2. Wolff AC, Hammond MEH, Allison KH, et al. HER2 Testing in Breast Cancer: ASCO/CAP Clinical Practice Guideline Focused Update. J Clin Oncol. 2018;36(20):2105–2122. doi:10.1200/JCO.2018.77.8738. ↩
3. Allred DC, Harvey JM, Berardo M, Clark GM. Prognostic and predictive factors in breast cancer by immunohistochemical analysis. Mod Pathol. 1998;11(2):155–168. PMID 9504686. ↩
4. Hammond MEH, Hayes DF, Dowsett M, et al. American Society of Clinical Oncology/College of American Pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Clin Oncol. 2010;28(16):2784–2795. doi:10.1200/JCO.2009.25.6529. ↩
5. Iwamoto T, Booser D, Valero V, et al. Estrogen receptor (ER) mRNA and ER-related gene expression in breast cancers that are 1% to 10% ER-positive by immunohistochemistry. J Clin Oncol. 2012;30(7):729–734. doi:10.1200/JCO.2011.36.2574. ↩
6. Yi M, Huo L, Koenig KB, et al. Which threshold for ER positivity? A retrospective study based on 9639 patients. Ann Oncol. 2014;25(5):1004–1011. doi:10.1093/annonc/mdu053. ↩
7. Bardou VJ, Arpino G, Elledge RM, Osborne CK, Clark GM. Progesterone receptor status significantly improves outcome prediction over estrogen receptor status alone for adjuvant endocrine therapy in two large breast cancer databases. J Clin Oncol. 2003;21(10):1973–1979. doi:10.1200/JCO.2003.09.099. ↩
8. Wolff AC, Somerfield MR, Dowsett M, et al. Human Epidermal Growth Factor Receptor 2 Testing in Breast Cancer: ASCO–College of American Pathologists Guideline Update. J Clin Oncol. 2023;41(22):3867–3872. doi:10.1200/JCO.22.02864. ↩
9. Modi S, Jacot W, Yamashita T, et al. Trastuzumab Deruxtecan in Previously Treated HER2-Low Advanced Breast Cancer (DESTINY-Breast04). N Engl J Med. 2022;387(1):9–20. doi:10.1056/NEJMoa2203690. ↩
10. Tarantino P, Hamilton E, Tolaney SM, et al. HER2-Low Breast Cancer: Pathological and Clinical Landscape. J Clin Oncol. 2020;38(17):1951–1962. doi:10.1200/JCO.19.02488. ↩
11. Rakha EA, Pinder SE, Bartlett JMS, et al. Updated UK Recommendations for HER2 assessment in breast cancer. J Clin Pathol. 2015;68(2):93–99. doi:10.1136/jclinpath-2014-202571. ↩
12. Layfield LJ, Frazier S, Esebua M, Schmidt RL. Interobserver reproducibility for HER2/neu immunohistochemistry: a comparison of reproducibility for the HercepTest™ and the 4B5 antibody clone. Pathol Res Pract. 2016;212(3):190–195. doi:10.1016/j.prp.2015.11.018. ↩
13. Fernandez AI, Liu M, Bellizzi A, et al. Examination of Low ERBB2 Protein Expression in Breast Cancer Tissue. JAMA Oncol. 2022;8(4):607–610. doi:10.1001/jamaoncol.2021.7239. ↩
14. Mukundan R. Analysis of image feature characteristics for automated scoring of HER2 in histology slides. J Imaging. 2019;5(3):35. doi:10.3390/jimaging5030035. ↩
15. Geyer FC, Pareja F, Weigelt B, et al. The spectrum of triple-negative breast cancer: the importance of exceptional histologies. Histopathology. 2017;71(4):527–539. doi:10.1111/his.13234. ↩

Last reviewed: 2026-05-31. Researcher-layer synthesis page. Evidence grades follow the GRADE-adapted rubric defined at the top of this page. Citations are anchored to the full bibliographic entries above; click the ↩ arrow next to any reference to return to its first citation in the prose.