br The availability of more sensitive breast lineage markers
The availability of more sensitive breast lineage markers that can reliably identify metastatic breast carcinoma that is tri-ple negative for ER, PR, and HER2 would be of high clinical utility. Labeling of TNBC by SOX10 and androgen receptor (AR) has been recently described in primary TNBC. However, there are few publications that have examined their perfor-mance in MBC. Furthermore, no studies have directly com-pared the sensitivity of SOX10 with the standard breast
lineage marker GATA3, specifically in the most clinically challenging context of working up metastatic TNBC. An as-sessment of the stability and concordance of these lineage markers between the primary and matched metastatic lesion is also lacking.
In this investigation, we evaluated the diagnostic utility of a panel of SOX10, GATA3, and AR in MBC that is negative for ER, PR, and HER2 and compared the Aztreonam of these markers with matched PBC.
2. Materials and methods
After institutional review board approval, we conducted a retrospective search of the pathology database at our institu-tion to identify MBC diagnosed from 2013 to 2017, which lacked expression of ER, PR, and HER2, and for which we had an available in-house PBC specimen. Outside consult cases were excluded. Clinicopathological data from the elec-tronic medical record and pathology reports were collected. For PBC, this included specimen type (core needle biopsy and/or excision specimen), patient age at the time of PBC, his-tologic type, grade (1, 2, or 3), tumor size (cm), lymph node stage, ER (SP1; Spring, Pleasanton, CA) and PR (636; Dako, Santa Clara, CA) status (positive or negative, percent staining), HER2 immunohistochemistry (IHC; scores 0, 1+, 2+, and 3+), and fluorescence in situ hybridization results (positive, nega-tive, equivocal, indeterminate, HER2/centromere enumeration probe for chromosome 17 [CEP17] ratio, HER2/cell, and CEP17/cell). For MBC, this included site of metastases; pa-tient age at the time of metastatic biopsy; ER, PR, and HER2 results; and any other immunohistochemical results.
IHC was performed at The Ohio State University Wexner Medical Center. In our institution, all MBC specimens are tested for ER, PR, and HER2. Those cases that are HER2 equivocal by IHC are reflexed to HER2 fluorescence in situ hybridization (PathVysion; Abbot, Irving, TX) in accordance with the 2013 American Society of Clinical Oncology/College of American Pathologists HER2 guideline recommendations . The frequency of IHC expression of SOX10 (Biocare, Pacheco, CA; clone BC34, dilution 1:200, antibody incuba-tion 15 minutes, retrieval high pH for 20 minutes), GATA3 (Biocare; clone L50-823, dilution 0.319333, antibody incuba-tion 15 minutes, retrieval high pH for 20 minutes), and AR (Dako, Santa Clara, CA; clone AR441, dilution 1:200, anti-body incubation 15 minutes, retrieval high pH for 30 minutes) was assessed in both MBC and PBC specimens. Staining was performed on whole-slide sections from core needle biopsy and/or resection specimens. All 3 antibodies were stained using the Leica Bond III system, and deparaffinization and
SOX10 is a useful marker of triple-negative breast cancer 223
epitope retrieval were performed using the Leica Dewax and ER2 solution (high pH), respectively. All 3 antibodies used the Leica Bond polymer refine detection kit. Slides were coun-terstained with hematoxylin and dehydrated before cover-slipped. Any degree of nuclear labeling was scored and recorded as a percentage (0-100%). Nuclear staining was scored for all 3 markers; ≥1% nuclear positivity was consid-ered a positive result. In those specimens in which GATA3, SOX10, or AR was performed at the time of clinical review, the original IHC stains were re-reviewed and scored. In speci-mens in which IHC stains GATA3, SOX10, or AR were not performed at the time of clinical review, additional IHC stains were performed on archived material where available.