TRPS1 in Breast Carcinoma: A Narrative Review

1. Introduction

Breast cancer is the most common malignancy in women, and 20–30% of patients develop metastatic disease. The fore, when a carcinoma is found in the lung, liver, bone, brain, pleura or lymph nodes of a female patient, a mammary origin must always be considered in the differential diagnosis, even in the absence of a documented history.

Immunohistochemistry is the main tool for solving this problem. Over the last decade, GATA3 has been the most widely used breast-lineage marker because of its nuclear staining and its sensitivity, which is higher than that of GCDFP-15 and mammaglobin [1,2,3]. However, GATA3 has two well-documented limitations. First, its sensitivity decreases in triple-negative breast carcinomas (TNBC) and in poorly differentiated and metaplastic carcinomas [2,3]. Second, it is expressed in other tumors, such as urothelial carcinoma, salivary gland tumors, parathyroid carcinoma and some gynecologic and cutaneous tumors [1]. GCDFP-15 and mammaglobin, which show cytoplasmic staining, are even less sensitive in TNBC [2,3].

In 2020, Ai et al. analyzed gene expression data from The Cancer Genome Atlas (TCGA) across 31 solid tumor types and identified TRPS1 as the gene with the most breast-selective expression, with high and homogeneous levels in all four intrinsic subtypes, including the basal-like subtype [4]. Immunohistochemical validation in 479 breast carcinomas and 1234 tumors from other organs established it as a highly sensitive marker, especially for TNBC [4]. Since then, dozens of studies have confirmed its sensitivity but qualified its specificity [5,6].

The aim of this narrative review is to summarize the evidence available up to October 2026 on TRPS1 biology, its diagnostic performance across breast carcinoma subtypes, its expression in non-mammary neoplasms and practical recommendations for its use in immunohistochemical panels.

2. Methods

A narrative review of the literature was performed. PubMed, PubMed Central and Google Scholar, as well as journal websites, were searched using the terms “TRPS1”, “trichorhinophalangeal syndrome type 1”, “breast carcinoma”, “triple-negative”, “metaplastic”, “immunohistochemistry”, “GATA3”, “cytology” and “metastatic”, combined with Boolean operators. Original articles, case series, tissue microarray (TMA) studies and reviews published between 2000 and October 2026, in English or with an English abstract, were included. Priority was given to immunohistochemical and cytological studies on human tissue published between 2021 and 2026, and to key studies on the biology of the gene. Conference abstracts without complete data and studies performed exclusively in animal models without diagnostic correlation were excluded. As this is a narrative review, no systematic protocol (PRISMA) was applied and no meta-analysis was performed. Percentages are reported as presented in each original study.

A generative artificial intelligence tool (Claude Opus 5.5, Anthropic) was used to assist with drafting, translation into English and verification of bibliographic data. All content was reviewed and verified by the authors, who take full responsibility for the manuscript.

3. Molecular Biology of Trps1

3.1. Gene and Protein

TRPS1 is located on the long arm of chromosome 8 (8q23–q24). Its germline mutations cause trichorhinophalangeal syndrome type I, an autosomal dominant disorder characterized by sparse hair, a bulbous nose and skeletal abnormalities of the phalanges [7]. The protein is an atypical GATA-family transcription factor: it has a single GATA-type zinc finger and additional Ikaros-like zinc fingers, and it acts mainly as a transcriptional repressor [7]. It participates in the development of cartilage, bone, kidney and hair follicles [4].

3.2. Role in the Mammary Gland and in Carcinoma

In the normal breast, TRPS1 is expressed in the nuclei of luminal epithelial cells of ducts and lobules but not in myoepithelial cells [4]. Normal luminal epithelium therefore serves as a positive internal control.

Radvanyi et al. described overexpression of the gene in breast cancer in 2005 [8]. Chen et al. validated it by quantitative immunohistochemistry in human tissue and described associations with clinical outcome [9]. Copy-number gain of 8q, which is frequent in breast cancer, contributes to this overexpression.

Functional studies have shown that TRPS1 is a lineage-specific transcriptional dependency of breast cancer: breast cancer cell lines require TRPS1 to survive, unlike cell lines derived from other tumors [10]. In murine models, it acts as a regulator of mammary epithelial growth and differentiation, with context-dependent effects [11]. In luminal cells, it modulates chromatin accessibility and the binding of the estrogen receptor to its target genes [12]. It has also been linked to angiogenesis, through VEGFA, and to epithelial–mesenchymal transition [5,13].

This lineage dependency explains why TRPS1 is retained in tumors that lose other features of luminal differentiation, such as TNBC and metaplastic carcinoma. Database analyses have also shown that TRPS1 and GATA3 are independent biomarkers in breast cancer, with different expression patterns and prognostic value [14].

4. Immunohistochemistry: Antibodies, Technique and Interpretation

4.1. Available Antibodies

The initial study by Ai et al. used a rabbit polyclonal antibody (PA5-84874, Invitrogen/Thermo Fisher) with heat-induced antigen retrieval at pH 6.0 on an automated platform [4]. Several commercial monoclonal clones are now available, including MSVA-512R, EP392, EPR16171, 8D11, 1D6, RM518 and OTI3B2 [5]. Clones MSVA-512R and EP392 provide the most intense and homogeneous nuclear staining in luminal and HER2-positive carcinomas, whereas 8D11 shows greater variability in intensity in TNBC and in tumors with apocrine differentiation [5].

4.2. Technical Variables

Protocol conditions modify the performance of the marker. Rammal et al. compared two protocols on the same TMAs: with high-pH antigen retrieval, sensitivity was 91% and specificity 66%; with low-pH retrieval, these values were 87% and 74%, respectively [15]. In addition, TRPS1 is expressed at low intensity in a wide variety of normal tissues, so antibody titration is critical: high concentrations increase weak positivity in non-mammary tumors [16].

In bone specimens, the type and duration of decalcification affect nuclear immunoreactivity; in the literature reviewed, the best results were obtained with 10% nitric acid for 6 h [5].

4.3. Interpretation

Only nuclear staining is considered positive; membranous or cytoplasmic staining, described in isolated cases, should be regarded as negative [4]. Adjacent normal mammary luminal epithelium and, in the skin, cutaneous adnexa serve as positive internal controls [4,17]. Three systems are used in the literature to grade TRPS1 expression. They are not equivalent, so the report should state which one was used.

A. Nuclear intensity. Intensity is graded on four levels: 0 (negative), 1+ (weak), 2+ (moderate) and 3+ (strong), using the internal control as reference. In its simplest form, the predominant intensity is reported together with the percentage of positive cells. Lennartz et al. combined intensity with extent into four categories, which they applied to more than 19,000 tumors and three different antibodies [16]:

• Negative: no stained tumor cells.
• Weak: weak (1+) staining in ≤70% of tumor cells, or moderate (2+) staining in ≤30%.
• Moderate: weak (1+) staining in >70% of cells, moderate (2+) staining in 31–70%, or strong (3+) staining in ≤30%.
• Strong: moderate (2+) staining in >70% of cells, or strong (3+) staining in >30%.

In other words, the more intense the staining and the more cells it involves, the higher the category. This system is quick and easy to apply in daily practice, but its categories are broad and it depends on visual calibration of intensity.

B. H-score (histochemical score). The percentage of tumor cells stained at each intensity is estimated, and each percentage is multiplied by the value of its intensity: 1 for weak, 2 for moderate and 3 for strong. The sum of the three products is the H-score, which ranges from 0 (no stained cells) to 300 (all cells with strong staining) [18]. For example, a tumor with 50% strongly stained cells, 30% moderately stained cells, 10% weakly stained cells and 10% unstained cells has an H-score of (3 × 50) + (2 × 30) + (1 × 10) = 220. Its advantage is that it reflects intratumoral heterogeneity of intensity. In cell blocks from metastases (clone EPR16171), Donthi et al. considered an H-score ≥10 as positive: the mean was 247 in metastatic breast carcinomas, compared with 45 in Müllerian, 8 in pulmonary and 6 in gastrointestinal metastases, with a sensitivity of 96% and a specificity of 73% [19]. In the skin, squamous cell carcinoma showed a median H-score of 200, basal cell carcinoma of 5 and endocrine mucin-producing sweat gland carcinoma of 300 [20]. The H-score is more laborious, shows greater interobserver variability and lacks a consensus diagnostic cut-off.

C. Semiquantitative immunoreactive score (IRS). This score derives from the system proposed by Remmele and Stegner for the estrogen receptor (0–12) [21]. Ai et al. proposed a 0–9 version for TRPS1, which is the one most widely used in the literature on this marker [4]: the percentage of stained cells is scored 0 (<1%), 1 (1–10%), 2 (11–50%) or 3 (>50%), and this score is multiplied by the intensity score: 0 (negative), 1 (weak), 2 (moderate) or 3 (strong). The result can only be 0, 1, 2, 3, 4, 6 or 9, and it is classified as negative (0 or 1), low (2), intermediate (3 or 4) or high (6 or 9) [4]. For example, a tumor with 60% moderately stained cells scores 3 × 2 = 6, i.e., a high score. When intensity is heterogeneous, some authors use the average intensity [22] and others the highest. There is also a 0–12 variant in which the percentage of cells is scored from 0 to 4; with this variant, 18 of 19 cutaneous metastases of breast carcinoma reached the maximum value [17].

In practice, any of the three systems is valid if applied consistently; the essential point is to record the percentage of positive cells and the intensity separately, so that the result can be compared with the literature and between laboratories.

The most important practical point is that both intensity and extent matter. Breast carcinomas usually show diffuse and strong staining, whereas many positive non-mammary neoplasms show weak or focal staining [15,16]. Weak and focal positivity therefore has a much lower predictive value for mammary origin than diffuse and intense staining.

5. Sensitivity According to Breast Carcinoma Subtype

TRPS1 is positive in approximately 90% of breast carcinomas overall, and its advantage over GATA3 is concentrated in estrogen receptor (ER)-negative tumors. Table 1 summarizes the main comparative studies.

5.1. Luminal and Her2-Positive Carcinomas

In ER-positive carcinomas, TRPS1 and GATA3 have virtually identical sensitivities, close to 95–98% [4]. In HER2-positive carcinomas, both are around 87–88% [4]. In these groups, the contribution of TRPS1 is smaller, although it may identify isolated high-grade GATA3-negative cases [4].

5.2. Non-Metaplastic Triple-Negative Carcinoma

This is the setting in which TRPS1 is most useful. In the series by Ai et al., TRPS1 was positive in 86% of non-metaplastic TNBC, compared with 51% for GATA3, and 69 cases were GATA3-negative but TRPS1-positive with intermediate or high intensity [4]. Using whole sections, Yoon et al. reported positivity in 99% of triple-negative carcinomas of no special type (NST), compared with 63% for GATA3 and 74% for SOX10 [23]. Kong et al. found a significant inverse correlation between TRPS1 and androgen receptor expression in 969 TNBC [25], which anticipates the behavior of apocrine carcinoma (see Section 5.4). In TNBC expressing cytokeratin 5 (basal phenotype), TRPS1 also retains its value as a marker of breast origin [27].

5.3. Metaplastic Carcinoma

GATA3 is positive in only 21–50% of metaplastic carcinomas, whereas TRPS1 is positive in 86–98% [4,23,24]. Positivity is maintained in the spindle cell, squamous, chondroid and giant cell components [4]. This finding is especially useful when a spindle cell breast neoplasm is encountered in a core needle biopsy, where the differential diagnosis includes phyllodes tumor and sarcomas (see Section 6.4).

5.4. Special Types with Low Expression

Some histological types are characteristically negative or show low expression:

• Apocrine carcinoma: among TNBC with apocrine differentiation, TRPS1 was positive in only 12%, whereas GATA3 was positive in 100% [28]. In the series by Kong et al., which included apocrine carcinomas with and without hormone receptor expression, positivity was 63.7% [25].
• Triple-negative neuroendocrine carcinoma: negative in 7 of 7 cases, whereas ER-positive neuroendocrine tumors were positive in 97.4% [25].
• Acinic cell carcinoma and classic cribriform adenoid cystic carcinoma: usually negative for both TRPS1 and GATA3 [23]. In contrast, solid-basaloid adenoid cystic carcinoma and secretory carcinoma are usually positive [23,29].
• Mucinous, tubular and lobular carcinomas: almost constant expression, in 99–100% of cases [25].

Overall, 5–10% of TNBC are TRPS1-negative, most of them belonging to these special types [5]. When TNBC are classified by immunohistochemical subtype, TRPS1 is the most frequently expressed breast-lineage marker [30].

6. Specificity and Differential Diagnosis

The specificity of TRPS1 proved to be lower than suggested by the initial study. Ai et al. had found absent or minimal expression in urothelial carcinoma, adenocarcinomas of the lung, pancreas, colon and stomach, renal cell carcinoma, thyroid carcinoma and melanoma [4]. However, even in that study, 25% of pulmonary squamous cell carcinomas, 24% of salivary duct carcinomas and 14% of ovarian serous carcinomas were positive, mostly with low intensity [4].

The largest study, by Lennartz et al., evaluated 19,201 tumors from 152 entities with clone MSVA-512R [16]. Positivity ranged from 51% to 100% across breast carcinoma types and reached up to 100% in some soft tissue tumors, 46% in salivary gland tumors, 35% in squamous cell carcinomas and 40% in gynecologic cancers [16]. However, TRPS1/GATA3 co-expression was observed in only 29 of 9835 tumors (0.3%) from 134 non-mammary, non-salivary entities [16]. Series from routine practice confirm a low overall specificity of 58–74% [15,26]. Table 2 summarizes TRPS1 expression in the non-mammary neoplasms most relevant to the differential diagnosis.

6.1. Urothelial Carcinoma

This is the setting in which TRPS1 adds the most specificity relative to GATA3, because urothelial carcinoma is almost always GATA3-positive and almost always TRPS1-negative [4,16]. However, Bachert et al. found positivity in up to 27.4% of muscle-invasive urothelial carcinomas and 31.9% of prostatic adenocarcinomas, although only a minority showed intermediate or high expression [22].

6.2. Gynecologic Tumors

Metastatic high-grade serous carcinoma to the breast or axilla is a classic source of error, because it shares papillary morphology and ER expression with breast carcinoma. TRPS1 was expressed in 14% of ovarian serous carcinomas in the series by Ai et al. [4] and in up to 71% of endometrial carcinomas in the series by Rammal et al., although with lower median intensity [15]. The panel should include PAX8 and WT1 [26].

6.3. Skin and Cutaneous Adnexa

The skin is the site where the specificity of TRPS1 is lowest, probably because of the embryological relationship between the mammary gland and cutaneous adnexa [31]. TRPS1 is expressed in 94% of cutaneous squamous cell carcinomas, in most adnexal tumors and in all endocrine mucin-producing sweat gland carcinomas, whereas Merkel cell carcinomas and most basal cell carcinomas are negative [20]. It is also expressed in cutaneous mesenchymal neoplasms [32]. Nevertheless, it remains useful in mammary and extramammary Paget disease and in the evaluation of cutaneous metastases of breast carcinoma [31].

6.4. Spindle Cell Lesions of the Breast

In a biopsy showing a spindle cell proliferation, TRPS1 positivity supports metaplastic carcinoma over sarcomas of other sites, but it does not distinguish metaplastic carcinoma from malignant phyllodes tumor, whose stromal component is often TRPS1-positive [35]. The same applies to a relevant proportion of mammary angiosarcomas [36]. In this setting, TRPS1 should be combined with a broad panel of cytokeratins, p63 and CD34.

7. Cytology, Metastases and Other Applications

7.1. Cytology and Serous Effusions

TRPS1 performs well in cell blocks and smears. In pleural effusions with metastatic breast carcinoma, Wang et al. found it positive in 92% of cell blocks [37]. Compared with mesothelioma, TRPS1 outperformed GATA3 in identifying metastatic breast carcinoma in pleural fluid [38], although reactivity of both markers has been described in some mesotheliomas [5]; mesothelial (calretinin, WT1 and D2-40) and epithelial (claudin-4) markers should therefore be added. Results in effusions are not uniform: in the series by Bradt et al., TRPS1 was positive in 31 of 41 ductal carcinomas compared with 40 of 41 for GATA3, and both markers were positive in 15 of 15 lobular carcinomas [39]. This supports the combined use of both markers.

In cytological specimens from metastases, TRPS1 showed high sensitivity for breast carcinoma, including triple-negative carcinoma [40,41]. On cytological smears, Rohra et al. observed TRPS1 positivity in 55 of 56 invasive breast carcinomas, compared with 37 of 56 for GATA3 [42].

7.2. Distant Metastases

ER expression may be lost in metastases, which increases the need for a receptor-independent lineage marker. In 107 ER-negative or ER-low distant metastases, Zhang et al. found that TRPS1 and GATA3 had similar and complementary performance: 6 cases were positive only for TRPS1 and 8 only for GATA3, and none was positive only for GCDFP-15 [43]. In the series by Hashmi et al., TRPS1 was the only positive breast marker in three metastases (bladder, lung and bone) [26]. TRPS1 is currently regarded as one of the most promising markers for metastatic breast carcinoma [6].

7.3. Prognostic and Predictive Value

The prognostic role of TRPS1 remains exploratory. Early quantitative immunohistochemistry and database studies described associations between TRPS1 levels and clinical outcome, with inconsistent results across cohorts [9,14]. In TNBC, its inverse relationship with androgen receptor expression suggests that TRPS1 could contribute to biological subclassification [25]. At present, TRPS1 has no prognostic or predictive indication in clinical practice, and its validated use is diagnostic.

8. Limitations, Use in Panels and Future Directions

8.1. Limitations of the Evidence

• Technical heterogeneity: studies use different clones, dilutions, platforms and positivity thresholds, which explains much of the variation in specificity between series [5,15].
• TMA design: many studies use 0.6–1.5 mm cores, which underestimate the heterogeneity of expression and do not reproduce the situation of a small biopsy [16,22].
• Selection bias: cohorts from routine practice include few non-mammary tumors, which limits the precision of the calculated specificity [26].
• Lack of standardization: three scoring systems coexist (nuclear intensity, H-score and immunoreactive score), and there is no consensus on which system to use or on the positivity threshold for diagnostic purposes (see Section 4.3).

8.2. Proposed Practical Panel

Based on current evidence, TRPS1 should be used as part of a panel guided by morphology and clinical context, never in isolation [5,26]:

• ER-positive carcinoma of uncertain origin: TRPS1 + GATA3, plus PAX8 and WT1 if there is papillary architecture or a history of gynecologic cancer.
• ER-negative or triple-negative carcinoma: TRPS1 + GATA3 + SOX10. The TRPS1/GATA3 combination maximizes sensitivity, and SOX10 adds value in basal-like and metaplastic TNBC [23,44].
• GATA3-positive carcinoma with possible urothelial origin: TRPS1 + p63/p40 + uroplakin; TRPS1 negativity favors urothelial origin [4,16].
• Lung nodule in a patient with a history of breast cancer: TRPS1 + GATA3 + TTF-1 + napsin A; TRPS1 can be positive in squamous cell carcinomas [4].
• Spindle cell breast neoplasm: TRPS1 together with broad-spectrum cytokeratins, p63 and CD34; TRPS1 does not separate metaplastic carcinoma from malignant phyllodes tumor [35].
• Skin lesion: interpret with caution; TRPS1 is useful in Paget disease but does not reliably separate a mammary metastasis from a primary adnexal tumor [20,31].

As a general rule, diffuse and strong staining supports mammary origin more strongly than weak or focal staining, and TRPS1/GATA3 co-expression is much more specific than either marker alone [16].

8.3. Future Directions

Open lines of research include standardization of the clone and positivity threshold, validation in small biopsies and cytology with multicenter cohorts, digital quantification of staining intensity and assessment of prognostic and predictive value in TNBC. The dependency of tumor cells on TRPS1 also points to it as a potential therapeutic target, although this possibility remains at the preclinical stage [10,12].

9. Conclusions

TRPS1 is an atypical GATA-family transcription factor on which breast carcinoma cells depend, which explains its preserved expression across all molecular subtypes. It is the most sensitive immunohistochemical marker of breast lineage currently available: it is positive in approximately 90% of breast carcinomas and in 86–99% of triple-negative and metaplastic carcinomas, in which it clearly outperforms GATA3. Its specificity, however, is limited, as it is expressed in squamous cell carcinomas, gynecologic tumors, salivary gland tumors, cutaneous adnexal tumors and sarcomas; its main differential strength is its almost constant negativity in urothelial carcinoma. Apocrine, triple-negative neuroendocrine, acinic cell and classic cribriform adenoid cystic carcinomas are known causes of false-negative results. TRPS1 should therefore be integrated into a panel with GATA3 and markers relevant to the differential diagnosis, interpreting the intensity and extent of staining in the morphological and clinical context.