Triple-negative breast cancer represents a clinically important breast-cancer phenotype characterized by absent estrogen receptor (ER) and progesterone receptor (PR) expression and absence of HER2 overexpression or amplification. It accounts for approximately 10%-20% of invasive breast carcinomas and is enriched among younger patients and carriers of germline BRCA1 pathogenic variants. At the population level, TNBC has a higher early recurrence rate, a greater propensity for visceral and central nervous system metastasis, and a shorter interval between recurrence and death than hormone receptor-positive breast cancer [1].

The label “triple negative,” however, describes what the tumor lacks rather than a single biological entity. TNBC includes basal-like, mesenchymal, luminal androgen receptor-like, immune-rich, metaplastic, medullary-pattern, apocrine, and other phenotypes. The practical consequence for pathology is that two tumors meeting the same ER/PR/HER2 definition can have markedly different morphology, immune contexture, proliferative activity, chemosensitivity, and long-term outcome.

Histopathology provides several potential prognostic dimensions that are inexpensive and reproducible. Tumor size and nodal stage remain dominant conventional variables, but lymphovascular invasion, fibrotic focus, residual disease after neoadjuvant therapy, histologic subtype, tumor necrosis, pushing versus infiltrative borders, and the host lymphocytic response may provide additional information. Among these, stromal tumor-infiltrating lymphocytes (sTILs) have accumulated the most robust evidence as a favorable prognostic biomarker in early TNBC [4-6].

Immunohistochemistry can further refine the phenotype. Ki-67 provides a measure of proliferative activity; AR identifies a subset with luminal/apocrine differentiation; CK5/6 and EGFR are commonly used to approximate a basal-like phenotype; p53 overexpression osssr null-pattern staining is frequent in TNBC; and PD-L1 marks a therapeutically relevant immune axis. Yet the prognostic literature is difficult to translate directly into practice because antibodies, scoring compartments, cutoffs, fixation, tissue-microarray methods, treatment exposure, and endpoints differ between studies.

The prognostic value of treatment response also deserves specific consideration. In patients receiving neoadjuvant systemic therapy, pCR and RCB are post-treatment histopathological variables that strongly stratify recurrence and survival risk. Consequently, any review of pathology-associated prognosis in TNBC should consider both pretreatment tumor biology and post-treatment residual disease.

This systematic review therefore aimed to synthesize the evidence linking histopathological and immunohistochemical features with prognosis in TNBC, identify markers with the strongest and most reproducible prognostic value, characterize areas of disagreement, and define practical priorities for pathology reporting and future validation.

Review Design and Reporting Framework

The review was structured according to PRISMA 2020 principles. The review question was: among patients with histologically confirmed TNBC, which routinely assessable histopathological or immunohistochemical features are associated with recurrence or survival outcomes? The review was not prospectively registered.

Eligibility Criteria

Eligible studies included adult patients with invasive TNBC, reported a defined histopathological or immunohistochemical feature, and provided a time-to-event or clinically interpretable prognostic outcome. Accepted outcomes included disease-free survival (DFS), invasive DFS, recurrence-free survival (RFS), event-free survival (EFS), distant DFS/recurrence-free interval, breast-cancer-specific survival (BCSS), and overall survival (OS). Cohort studies, prospective-retrospective biomarker analyses of clinical trials, and population-based studies were eligible. Mixed breast-cancer cohorts were included only when TNBC-specific results were separable.

Studies were excluded when they focused exclusively on genomic, transcriptomic, proteomic, radiomic, or circulating biomarkers without a histopathological or immunohistochemical correlate; did not report survival or recurrence; included only metastatic tissue without primary-tumor context; were case reports or very small descriptive series; or represented duplicate populations without additional prognostic information. Reviews and meta-analyses were used for contextual interpretation but were not counted as primary included studies.

Information Sources and Search Strategy

MEDLINE/PubMed, Embase, Scopus, Web of Science, and Cochrane CENTRAL were searched from January 2000 to 30 June 2026. Reference lists of relevant systematic reviews and landmark biomarker studies were also examined.

A representative search strategy combined terms for TNBC with pathology and prognostic markers: ("triple negative breast cancer" OR TNBC) AND (histopathology OR histologic OR morphology OR "tumor infiltrating lymphocytes" OR TILs OR immunohistochemistry OR Ki-67 OR androgen receptor OR EGFR OR CK5/6 OR p53 OR PD-L1 OR "lymphovascular invasion" OR "residual cancer burden") AND (prognosis OR survival OR recurrence OR outcome). Database-specific indexing terms were added where appropriate.

Study Selection and Data Extraction

Titles and abstracts were screened for relevance, followed by full-text eligibility assessment. Extracted variables included study design, country or setting, sample size, treatment context, specimen type, histopathological feature, immunohistochemical antibody or scoring method when available, cutoff, survival endpoint, effect estimate, and whether the marker remained significant in multivariable analysis.

Risk-of-Bias Assessment

Observational studies were assessed using domains adapted from ROBINS-I and the Newcastle-Ottawa framework, with particular emphasis on cohort selection, completeness of follow-up, blinding or reproducibility of pathology scoring, prespecification of cutoffs, adjustment for stage and treatment, assay heterogeneity, and selective reporting. Prospective-retrospective biomarker analyses embedded within randomized trials were considered at lower risk when pathology scoring was prespecified and outcomes were mature.

Data Synthesis

A new pooled meta-analysis was not performed because the included literature varied markedly in treatment setting, assay platform, biomarker cutoff, scoring compartment, endpoint definition, and covariate adjustment. Evidence was therefore synthesized narratively, with greater weight given to large cohorts, trial-based biomarker analyses, long-term follow-up, continuous-marker analyses, and multivariable models.

Review Design and Reporting Framework

The review was structured according to PRISMA 2020 principles. The review question was: among patients with histologically confirmed TNBC, which routinely assessable histopathological or immunohistochemical features are associated with recurrence or survival outcomes? The review was not prospectively registered.

Eligibility Criteria

Eligible studies included adult patients with invasive TNBC, reported a defined histopathological or immunohistochemical feature, and provided a time-to-event or clinically interpretable prognostic outcome. Accepted outcomes included disease-free survival (DFS), invasive DFS, recurrence-free survival (RFS), event-free survival (EFS), distant DFS/recurrence-free interval, breast-cancer-specific survival (BCSS), and overall survival (OS). Cohort studies, prospective-retrospective biomarker analyses of clinical trials, and population-based studies were eligible. Mixed breast-cancer cohorts were included only when TNBC-specific results were separable.

Studies were excluded when they focused exclusively on genomic, transcriptomic, proteomic, radiomic, or circulating biomarkers without a histopathological or immunohistochemical correlate; did not report survival or recurrence; included only metastatic tissue without primary-tumor context; were case reports or very small descriptive series; or represented duplicate populations without additional prognostic information. Reviews and meta-analyses were used for contextual interpretation but were not counted as primary included studies.

Information Sources and Search Strategy

MEDLINE/PubMed, Embase, Scopus, Web of Science, and Cochrane CENTRAL were searched from January 2000 to 30 June 2026. Reference lists of relevant systematic reviews and landmark biomarker studies were also examined.

A representative search strategy combined terms for TNBC with pathology and prognostic markers: ("triple negative breast cancer" OR TNBC) AND (histopathology OR histologic OR morphology OR "tumor infiltrating lymphocytes" OR TILs OR immunohistochemistry OR Ki-67 OR androgen receptor OR EGFR OR CK5/6 OR p53 OR PD-L1 OR "lymphovascular invasion" OR "residual cancer burden") AND (prognosis OR survival OR recurrence OR outcome). Database-specific indexing terms were added where appropriate.

Study Selection and Data Extraction

Titles and abstracts were screened for relevance, followed by full-text eligibility assessment. Extracted variables included study design, country or setting, sample size, treatment context, specimen type, histopathological feature, immunohistochemical antibody or scoring method when available, cutoff, survival endpoint, effect estimate, and whether the marker remained significant in multivariable analysis.

Risk-of-Bias Assessment

Observational studies were assessed using domains adapted from ROBINS-I and the Newcastle-Ottawa framework, with particular emphasis on cohort selection, completeness of follow-up, blinding or reproducibility of pathology scoring, prespecification of cutoffs, adjustment for stage and treatment, assay heterogeneity, and selective reporting. Prospective-retrospective biomarker analyses embedded within randomized trials were considered at lower risk when pathology scoring was prespecified and outcomes were mature.

Data Synthesis

A new pooled meta-analysis was not performed because the included literature varied markedly in treatment setting, assay platform, biomarker cutoff, scoring compartment, endpoint definition, and covariate adjustment. Evidence was therefore synthesized narratively, with greater weight given to large cohorts, trial-based biomarker analyses, long-term follow-up, continuous-marker analyses, and multivariable models.

The principal finding of this review is that prognosis in TNBC can be meaningfully refined using information already present on routine H&E slides and widely available immunohistochemistry. However, the evidentiary strength is highly uneven. sTILs, LVI, nodal status, tumor size, and post-neoadjuvant residual disease show the most consistent associations. By contrast, many individual IHC markers produce cohort-dependent results because they reflect overlapping biological programs rather than single deterministic pathways.

TILs are particularly compelling because they combine biological plausibility, low cost, treatment-independent prognostic value, and increasing standardization. The International Immuno-Oncology Biomarker Working Group has provided a reproducible framework for scoring stromal TILs on H&E sections [18]. The strongest data support reporting TILs as a continuous percentage, with optional clinically useful categories, rather than relying on a universal binary cutoff.

LVI deserves greater attention in TNBC pathology reports. The recent chemotherapy-naive long-term cohort is important because it shows that LVI adds adverse prognostic information independently of sTILs. A TIL-rich tumor should therefore not automatically be regarded as low risk when LVI is present. Similarly, fibrotic focus may capture stromal biology associated with aggressive behavior, although external validation is needed before it is treated as a routine decision marker.

For patients treated neoadjuvantly, response assessment supersedes many pretreatment markers. pCR and RCB integrate chemosensitivity with residual tumor volume and nodal burden, providing strong post-treatment risk stratification. The marked survival difference between pCR/RCB-I and RCB-II/III supports meticulous standardized examination of the tumor bed and lymph nodes.

Ki-67 is appealing because it is inexpensive and familiar, but its clinical implementation remains limited by cut-point instability. Laboratories using Ki-67 should prioritize standardized preanalytics, scoring procedures, and preferably continuous reporting. A single universal threshold derived from heterogeneous retrospective studies should not be assumed to be prognostically interchangeable across populations.

AR illustrates the difference between a biologically meaningful phenotype and a fully validated prognostic biomarker. AR-positive TNBC often shows apocrine morphology and lower proliferation and may represent the luminal androgen receptor subtype. Yet improved survival in some cohorts may partly reflect those accompanying features. Recent analyses showing loss of significance after adjustment suggest that AR should not replace stage, TILs, or other established variables.

PD-L1 is even more context dependent. Its clinical role is primarily predictive in the setting of immunotherapy, while prognostic analyses in untreated or conventionally treated TNBC are difficult to interpret because PD-L1 positivity often functions as a surrogate for immune activation. Clone-specific scoring systems such as SP142 immune-cell score and 22C3 combined positive score are not interchangeable for prognostic purposes.

The evolving phenotype of residual disease after neoadjuvant therapy also has practical implications. Repeating selected IHC markers in residual tumor may capture therapy-driven selection and identify biologically different post-treatment disease. This concept is especially relevant as adjuvant treatment decisions increasingly depend on residual risk.

A clinically useful pathology-centered prognostic framework for TNBC should therefore be hierarchical. First, establish conventional anatomic and histopathologic risk: tumor size, nodal status, LVI, histologic type and grade. Second, quantify immune context using sTILs. Third, in neoadjuvant cases, report pCR and RCB. Fourth, use IHC markers such as Ki-67, AR, EGFR/CK5/6, and PD-L1 to refine phenotype when the information is clinically relevant, recognizing that most remain supportive rather than independently decisive.

Clinical and Pathology Reporting Implications

Strengths

This review integrates conventional histopathology, immune morphology, treatment-response pathology, and clinically accessible immunohistochemistry rather than focusing on a single biomarker class. It emphasizes multivariable and long-term evidence, distinguishes prognostic from predictive interpretation, and highlights where routine H&E features perform as well as or better than more complex IHC panels.

Limitations

The evidence base is dominated by retrospective or prospective-retrospective studies, and many biomarkers were evaluated using tissue microarrays rather than whole sections. Definitions of TNBC changed over time, particularly ER/PR thresholds and HER2 criteria. Treatment regimens also evolved substantially from historical anthracycline cohorts to contemporary chemoimmunotherapy.

Assays and cutoffs were heterogeneous. Ki-67 thresholds varied widely; AR positivity ranged from low nuclear-percentage cutoffs to higher thresholds; PD-L1 studies used different clones and scoring compartments; and basal-like status was approximated with non-identical EGFR/CK panels. These differences limit direct comparison and preclude reliable universal cutoffs.

Finally, this review did not perform a new quantitative meta-analysis. Therefore, the synthesis emphasizes consistency and biological plausibility rather than generating a single pooled effect for each marker.

Future Research

Future prognostic studies should use standardized TNBC definitions, central or digitally assisted pathology review, whole-slide sTIL assessment, prespecified biomarker cutoffs, and time-to-event endpoints defined according to STEEP criteria. Biomarker analyses should report both continuous and categorical effects and adjust for treatment, stage, nodal status, LVI, and pCR/RCB where relevant.

Prospective trials should test whether combining sTILs with LVI, RCB, and a limited IHC panel can safely identify patients for treatment escalation or de-escalation. Digital pathology may improve reproducibility of TILs, Ki-67, necrosis, and spatial immune features, but algorithms should be benchmarked against standardized human scoring before clinical adoption.

In residual disease, paired pre- and post-treatment profiling is especially important because therapy can change phenotype. The prognostic value of dynamic changes in AR, Ki-67, basal markers, PD-L1, and immune infiltrates should be evaluated prospectively alongside modern adjuvant therapies.

Histopathological assessment remains central to prognostic stratification in TNBC. Stromal TILs are the most consistently favorable pretreatment histopathologic marker, whereas LVI, nodal involvement, larger tumor size, fibrotic focus, and substantial residual disease after neoadjuvant therapy identify higher-risk disease.

Among immunohistochemical markers, Ki-67, AR, EGFR/CK5/6, PD-L1, and p53 provide useful biological context, but their independent prognostic value is less consistent than that of TILs, LVI, and treatment response. AR often marks a more indolent apocrine/luminal phenotype, while basal markers frequently identify aggressive disease; however, neither should be interpreted independently of clinicopathological context. PD-L1 is particularly assay- and compartment-dependent.

The most clinically useful approach is therefore integrated rather than marker-specific: combine conventional morphology and stage with standardized sTIL assessment and, after neoadjuvant therapy, pCR/RCB. Selected IHC markers can then refine phenotype and support risk interpretation when locally validated. Greater standardization and prospective validation are required before individual IHC biomarkers can be used to guide prognostic decisions independently.

1. Dent R, Trudeau M, Pritchard KI, et al. Triple-negative breast cancer: clinical features and patterns of recurrence. Clin Cancer Res. 2007;13(15 Pt 1):4429-4434. doi:10.1158/1078-0432.CCR-06-3045.
2. Rakha EA, El-Sayed ME, Green AR, Lee AHS, Robertson JF, Ellis IO. Prognostic markers in triple-negative breast cancer. Cancer. 2007;109(1):25-32. doi:10.1002/cncr.22381.
3. Loibl S, Müller BM, von Minckwitz G, et al. Androgen receptor expression in primary breast cancer and its predictive and prognostic value in patients treated with neoadjuvant chemotherapy. Breast Cancer Res Treat. 2011;130(2):477-487. doi:10.1007/s10549-011-1715-8.
4. Loi S, Sirtaine N, Piette F, et al. Prognostic and predictive value of tumor-infiltrating lymphocytes in a phase III randomized adjuvant breast cancer trial: BIG 02-98. J Clin Oncol. 2013;31(7):860-867. doi:10.1200/JCO.2011.41.0902.
5. Adams S, Gray RJ, Demaria S, et al. Prognostic value of tumor-infiltrating lymphocytes in triple-negative breast cancers from two phase III randomized adjuvant breast cancer trials: ECOG 2197 and ECOG 1199. J Clin Oncol. 2014;32(27):2959-2966. doi:10.1200/JCO.2013.55.0491.
6. Dieci MV, Criscitiello C, Goubar A, et al. Prognostic and predictive value of tumor-infiltrating lymphocytes in two phase III randomized adjuvant breast cancer trials. Ann Oncol. 2015;26(8):1698-1704. doi:10.1093/annonc/mdv239.
7. Yue Y, Astvatsaturyan K, Cui X, Zhang X, Fraass B, Bose S. Stratification of prognosis of triple-negative breast cancer patients using combinatorial biomarkers. PLoS One. 2016;11(3):e0149661. doi:10.1371/journal.pone.0149661.
8. Salgado R, Denkert C, Demaria S, et al. The evaluation of tumor-infiltrating lymphocytes in breast cancer: recommendations by an International TILs Working Group 2014. Ann Oncol. 2015;26(2):259-271. doi:10.1093/annonc/mdu450.
9. Sharma P, López-Tarruella S, García-Sáenz JA, et al. Pathological response and survival in triple-negative breast cancer following neoadjuvant carboplatin plus docetaxel. Clin Cancer Res. 2018;24(23):5820-5829. doi:10.1158/1078-0432.CCR-18-0585.
10. Cabezas-Quintario MA, et al. Androgen receptor as prognostic marker in triple-negative breast cancer patients. J Cancer Sci Ther. 2018;10:170. doi:10.4172/2472-4971.1000170.
11. Schmolze D, Behrendt CE, Lee PP. The prognostic value of PD-L1 expression in triple-negative breast cancer: a cohort study and systematic literature review. Pathol Lab Med Open J. 2019;2:1-10. doi:10.17140/PLMOJ-2-107.
12. Wang X, Liu Y. PD-L1 expression in tumor infiltrated lymphocytes predicts survival in triple-negative breast cancer. Pathol Res Pract. 2020;216(3):152802. doi:10.1016/j.prp.2019.152802.
13. Huang M, O'Shaughnessy J, Zhao J, et al. Association of pathologic complete response with long-term survival outcomes in triple-negative breast cancer: a meta-analysis. Cancer Res. 2020;80(24):5427-5434. doi:10.1158/0008-5472.CAN-20-1792.
14. Chen B, Zhang Z, Wang C, Lv X. Prognostic relevance of tumor-infiltrating lymphocytes in residual tumor tissue from patients with triple-negative breast cancer following neoadjuvant chemotherapy: a systematic review and meta-analysis. Oncol Lett. 2023;26:410. doi:10.3892/ol.2023.14028.
15. ElFeky AM, Saied EM, Shawky H. Prognostic value of basal markers (EGFR and cytokeratin 5/6) expression in triple-negative invasive breast cancer. J Pak Med Assoc. 2023. doi:10.47391/JPMA.EGY-S4-33.
16. Wu L, Chen M, Lin Y, et al. Prognostic value of immunohistochemistry-based subtyping before and after neoadjuvant chemotherapy in patients with triple-negative breast cancer. Am J Surg Pathol. 2024;48(1):27-35. doi:10.1097/PAS.0000000000002139.
17. de Boo LW, Jóźwiak K, ter Hoeve ND, et al. Prognostic value of histopathologic traits independent of stromal tumor-infiltrating lymphocyte levels in chemotherapy-naïve patients with triple-negative breast cancer. ESMO Open. 2024;9(3):102923. doi:10.1016/j.esmoop.2024.102923.
18. Pastena P, Perera H, Martinino A, et al. Unraveling biomarker signatures in triple-negative breast cancer: a systematic review for targeted approaches. Int J Mol Sci. 2024;25(5):2559. doi:10.3390/ijms25052559.
19. Sigurjonsdottir G, De Marchi T, Ehinger A, et al. Evaluation of alternative prognostic thresholds for SP142 and 22C3 immunohistochemical PD-L1 expression in triple-negative breast cancer: results from a population-based cohort. Breast Cancer Res Treat. 2025;210(2):271-284. doi:10.1007/s10549-024-07561-x.
20. Singh AR, Kumar P, Choubey S. Histopathological and immunohistochemical correlation of triple-negative breast cancer with clinical outcomes. Int J Med Biomed Stud. 2025;9(1). doi:10.32553/ijmbs.v9i1.3065.
21. Tran LDN, Nguyen KK, Pham TM, et al. Predictive and prognostic value of tumor-infiltrating lymphocytes in triple-negative breast cancer: a real-world study from Vietnam. Ther Adv Med Oncol. 2025. doi:10.1177/17588359251403410.
22. Pal M, Chakrawal A, Singh B, et al. Androgen receptor as a prognostic biomarker in breast cancer: a systematic review and meta-analysis. Crit Rev Oncol Hematol. 2026;105112. doi:10.1016/j.critrevonc.2025.105112.