TRPS1 and GATA3 Expression in Invasive Breast Carcinoma With Apocrine Differentiation

A total of 63 female patients with primary IBC, 3 patients with primary cutaneous apocrine carcinoma, and 3 patients with primary cutaneous eccrine carcinoma were included in this study. A total of 52 primary IBCs with apocrine differentiation and high AR expression (>90%), including 41 TNBCs and 11 HER2⁺/ER⁻ IBCs, were collected from 2012 to 2020 at The University of Texas MD Anderson Cancer Center (Houston) and The University of Texas Southwestern Medical Center (Dallas). IBCs with apocrine differentiation were defined by histologic criteria, including distinct cell margins, abundant acidophilic cytoplasm with eosinophilic granules, and vesicular nuclei with prominent nucleoli (Figure 1). Distant metastasis was observed in 2 TNBCs with apocrine differentiation, and TRPS1 and GATA3 expression were tested as well. In addition, 11 primary TNBCs without differentiation but with high AR expression (>90%) were included for comparison. Clinical and pathologic information was collected, including patient age, sex, histologic diagnosis, grade, tumor size, stage, and ER, PR, and HER2 status. This study was approved by the Institutional Review Board at The University of Texas MD Anderson Cancer Center.

Formalin-fixed, paraffin-embedded tumor sections were obtained and stained with antibodies for AR (RM7, RevMAb Biosciences, San Francisco, California), TRPS1 (PA5-84874, Invitrogen/Thermo Fisher Scientific, Waltham, Massachusetts), and GATA3 (polyclonal L50-823, Cell Marque, Rocklin, California) by immunohistochemistry. Immunohistochemical staining was performed in a Leica BOND-MAX autostainer system (Leica Biosystems GmbH, Nussloch, Germany) following standard automated protocols. Nuclear staining was considered positive staining. Immunoreactivity was quantified by multiplying a variable representing the percentage of positive cells (0, <1%; 1, 1%–10%; 2, 11%–50%; and 3, 51%–100%) by another variable representing the staining intensity (0, negative; 1, weak; 2, moderate; and 3, strong). Final immunoreactivity scores were graded as negative (0 or 1), low positive (2), intermediate positive (3 or 4), or high positive (6 or 9).

The differences in immunochemical expression of TRPS1 and GATA3 were analyzed using the Fisher exact test for categoric variables and the Wilcoxon rank sum test for continuous variables. All tests were 2 sided, and P values less than .05 were considered statistically significant.

Clinicopathologic features of 63 IBCs with and without apocrine differentiation are summarized in Supplemental Table 1 (see supplemental digital content at https://meridian.allenpress.com/aplm in the February 2024 table of contents). A total of 52 cases exhibited apocrine differentiation, including 41 TN and 11 HER2⁺/ER⁻ cases, and the other 11 cases showed no apocrine differentiation. In all cases, AR expression was high (strong and >90%; Figure 1, B).

Benign breast epithelial cells are usually positive for TRPS1 and GATA3, whereas epithelial cells with apocrine metaplasia are negative for TRPS1 and positive for GATA3, as exemplified in Figure 2.

TRPS1 expression was significantly different from GATA3 expression in IBCs with apocrine differentiation (P < .001; Table). In the 41 TNBCs with apocrine differentiation, TRPS1 was negative in 36 cases (88%), low positive in 4 (10%), and intermediate positive in 1 (2%), whereas GATA3 was high or intermediate positive in all cases (100%). In the 11 HER2⁺/ER⁻ cases with apocrine differentiation, TRPS1 was negative in 9 cases (82%), whereas GATA3 was high or intermediate positive in all cases (100%). Between the 2 groups (TNBC versus HER2⁺/ER⁻) of IBCs with apocrine differentiation, there was no significant difference in TRPS1 expression (predominantly negative) or in GATA3 expression (all positive).

Morphologic apocrine differentiation and negative TRPS1 expression by immunohistochemistry were also identified in distant metastasis of 2 TN IBCs with apocrine differentiation. Histologic slides of lymph node metastasis in 1 HER2⁺ and 9 of 13 TN IBCs with apocrine differentiation were available for review, and all showed apocrine differentiation morphologically.

In the 11 TNBCs without apocrine differentiation, TRPS1 was positive in all cases (100%), with low positive expression in 1 (9%) and intermediate or high positive expression in 10 (91%). Similarly, GATA3 was high or intermediate positive in all 11 cases (100%; Figure 3).

Similarly to the breast, the apocrine glands in the normal dermis showed negative TRPS1 expression and positive GATA3 expression, whereas the eccrine glands in the dermis showed positive TRPS1 expression and negative GATA3 expression (Figure 4). In primary cutaneous apocrine carcinoma, TRPS1 was negative in all 3 tested cases, whereas GATA3 was moderate or high positive in all 3 cases. In contrast, in primary cutaneous eccrine carcinoma, TRPS1 was positive in all 3 tested cases, whereas GATA3 was negative in 2 cases and low positive in 1 case (Figure 5).

TNBC is classified into 6 tumor-specific subtypes by gene expression profiling, including basal-like 1, basal-like 2, immunomodulatory, mesenchymal, mesenchymal stem–like, and luminal AR subtypes.^16,17  Most TNBCs are basal-type IBC, but TNBC with strong AR, with or without apocrine morphology, overlaps with or belongs to the luminal AR type.^16,17  In comparison with nonapocrine TNBC, TNBC with apocrine differentiation usually has strong AR expression and confers better overall survival.^18,19  Genetically, TNBC with apocrine differentiation has a lower level of genetic instability, harbors TP53 mutations less frequently, and displays a high mutation frequency in PIK3CA and other PI3K signaling pathway–related genes.^17,18  Consistent with its high AR expression, TNBC with apocrine differentiation usually belongs to the luminal AR subtype, so it is not surprising that this type of tumor is positive for luminal marker GATA3.

TRPS1 is a breast lineage biomarker and is highly expressed in both ER⁺ luminal and TN/basal types of breast carcinoma, as shown in previous studies.^1,10  On the contrary, GATA3 involves breast luminal differentiation and is a luminal cell marker, mainly expressed in luminal A and luminal B types of IBC, but not in TN/basal-like IBC.^20,21  Therefore, the sensitivity of TRPS1 in basal-like IBC or TNBC is better than that of GATA3, because more than 90% of TNBCs (IBC–no special type [NST]) are positive for TRPS1, and almost all GATA3-negative TNBCs (IBC-NST) are positive for TRPS1.¹¹  However, the present study showed TNBC with apocrine differentiation was predominantly negative for TRPS1 expression and positive for GATA3. To our knowledge, this is the only special type of TNBC identified so far that has negative TRPS1 but positive GATA3.

Similarly to TNBC with apocrine differentiation, HER2⁺ IBC with apocrine differentiation exhibited negative TRPS1 expression in 82% of cases, whereas GATA3 expression was positive in 100%. Therefore, when TRPS1 is used as a diagnostic tool to confirm the breast origin of metastatic tumors, negative TRPS1 cannot exclude breast origin for tumors with apocrine features, although negative TRPS1 is an extremely rare event for IBC-NST. A panel of TRPS1 and GATA3 immunostains can be helpful when the tissue origin of such tumors is clinically relevant.

The present study also demonstrated that TNBC with strong AR expression but without apocrine differentiation was predominantly positive for TRPS1 and GATA3, suggesting that high AR expression is not associated with negative TRPS1. Consistent with this finding, early data revealed that TRPS1 was expressed in all male IBCs regardless of the extent of AR expression.²²  AR is expressed in 70% to 90% of ER⁺ IBCs, 30% to 60% of HER2⁺ IBCs, and 10% to 53% of TNBCs with variable intensity.^23–25  Based on our study, apocrine differentiation is the major factor that is associated with negative TRPS1 expression, rather than AR expression.

Differentiating primary cutaneous adnexal carcinoma in the axilla from a metastasis from the breast is a diagnostic challenge. Approximately 25% of patients with breast cancer may develop cutaneous metastasis.^26,27  Histologically and immunohistochemically, primary cutaneous apocrine carcinoma mirrors breast apocrine carcinoma, and primary cutaneous eccrine carcinoma resembles IBC without apocrine differentiation.²⁸  In many situations, a wide immunohistochemical panel can be orientating but not conclusive, because of the similar morphologic features and immunohistochemical profiles.^28–30  Primary cutaneous adnexal carcinoma shows positive immunoreactivity for p63 and CK 5/6, whereas p63 and CK 5/6 are also commonly expressed in TNBC.³¹  The commonly used breast cancer markers GATA3, mammaglobin, ER, and PR were expressed in primary cutaneous adnexal carcinoma to varying degrees as well.²⁹  Breast tissue consists of the parenchyma and stroma, originating from ectodermal and mesodermal elements, respectively, and the mammary gland is a highly evolved modified epidermal appendage derived from the sweat gland embryonically.³²  Consequently, mammary carcinoma has similar morphologic and histochemical profiles to those of cutaneous adnexal carcinoma. In the current study, we showed TRPS1 has negative expression in the skin apocrine duct and positive expression in the skin eccrine duct, which is the opposite of GATA3. In addition, TRPS1 is negative in primary cutaneous apocrine carcinoma, the same as in breast carcinoma with apocrine differentiation, and positive in primary cutaneous eccrine carcinoma, the same as in breast carcinoma without apocrine differentiation. Hence, TRPS1 is unlikely to help differentiating primary cutaneous adnexal carcinoma in the axilla from a metastasis from the breast.

More study is needed to further investigate the underlying cause of loss of TRPS1 expression in IBC with apocrine differentiation. Recently, overexpression of specific apocrine biomarkers 15-hydroxyprostaglandin dehydrogenase (15-PGDH) and hydroxymethylglutaryl coenzyme A reductase (HMG-CoA reductase) has been identified in early and advanced apocrine lesions, and 15-PGDH expression is related to androgen.^33,34  Studies in prostate cancer suggest that androgen withdrawal resulted in an increase in TRPS1 protein in prostate cancer cells.³⁵  It is possible that at the early stage of tumorigenesis in IBC with apocrine differentiation, the androgen pathway is activated and TRPS1 expression is shut down, and subsequently 15-PGDH and HMG-CoA reductase overexpression induces apocrine differentiation. IBC with apocrine differentiation overlaps with or belongs to luminal AR subtype of breast cancer at the molecular level. The luminal AR subtype is characterized by luminal gene expression with high AR. The luminal cell marker GATA3 is preserved during tumorigenesis.

In summary, most ER⁻/PR⁻/HER2⁺ IBCs with apocrine differentiation are TRPS1 negative and GATA3 positive by immunohistochemistry, regardless of HER2 status. Therefore, it is important to be aware that in this type of tumor, TRPS1 negativity does not exclude breast origin. A panel of TRPS1 and GATA3 immunostains can be helpful when the tissue origin of such a tumor is clinically relevant. In addition, TRPS1 expression is similar between breast carcinoma and primary cutaneous adnexal carcinoma, and therefore is not a differentiating marker in that regard.

We would like to thank Sarah Bronson, BA, of the Research Medical Library at The University of Texas MD Anderson Cancer Center for her assistance in editing this document.