Li Joshua J X, Cheng Hiu Yu, Lee Conrad H C, Ng Joanna K M, Tsang Julia Y, Tse Gary M
Department of Pathology, Queen Mary Hospital, The University of Hong Kong, Hong Kong, Hong Kong.
Department of Anatomical and Cellular Pathology, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, Hong Kong.
Cancer Cytopathol. 2025 Jan;133(1):e22910. doi: 10.1002/cncy.22910. Epub 2024 Oct 8.
Metastatic breast cancers are frequently encountered in cytology and require immunocytochemistry (ICC). In this study, traditional and multiplex ICC (mICC) for GATA-binding protein 3 (GATA3), gross cystic disease fluid protein 15 (GCDFP15), and mammaglobin (MMG) were performed with the aim of validating mICC in cell blocks, with further single-cell expression pattern analysis to identify the single markers and combinations of markers most sensitive in subtypes of breast cancer.
GATA3, GCDFP15, and MMG were paired with OptiView 3,3'-diaminobenzidine and Ventana DISCOVERY Purple and Blue, respectively, with cyclical and serial staining. Bright-field imaging was performed with the Mantra 2 system and analyzed with the inForm Tissue Finder (Akoya Biosciences). Cell detection and phenotyping were further confirmed by two pathologists.
In the 36 cases studied, traditional ICC and mICC demonstrated good concordance (kappa coefficient, >0.5; p < .01) at three cutoffs (1%, 5%, and 50%), except for GATA3 at the 1% cutoff. Single-marker positivity outnumbered double-marker positivity and the exceedingly rare triple-marker positivity (<3%). GATA3 was the leading single marker-positive phenotype in all breast cancer subtypes, except for MMG in estrogen receptor-positive, progesterone receptor-positive, and human epidermal growth factor receptor 2-positive (ER+/PR+/HER2+) breast cancers. Limited to two markers, GATA3/MMG included the greatest number of tumor cells for luminal breast cancers (ER+/PR+/HER2+, 60.6%; ER+/PR+/HER2+, 31.4%), whereas HER2-overexpressed breast cancers (27.4%) and triple-negative breast cancers (26.4%) favored the combination of GATA3/GCDFP15.
For a single marker, GATA3 displayed the highest sensitivity. The addition of MMG for hormone receptor-positive breast cancers and GCDFP15 for hormone receptor-negative breast cancers further increased sensitivity. The low proportion of multimarker-positive cells suggested that the coexpression observed with traditional ICC is attributable to intratumoral heterogeneity, not genuine coexpression.
转移性乳腺癌在细胞学检查中经常遇到，需要进行免疫细胞化学（ICC）检测。在本研究中，我们进行了针对GATA结合蛋白3（GATA3）、乳腺囊肿病液蛋白15（GCDFP15）和乳腺珠蛋白（MMG）的传统和多重ICC（mICC）检测，目的是验证细胞块中的mICC，并进一步进行单细胞表达模式分析，以确定在乳腺癌亚型中最敏感的单一标志物和标志物组合。
GATA3、GCDFP15和MMG分别与OptiView 3,3'-二氨基联苯胺以及Ventana DISCOVERY Purple和Blue配对，进行循环和连续染色。使用Mantra 2系统进行明场成像，并使用inForm Tissue Finder（Akoya Biosciences）进行分析。细胞检测和表型分析由两名病理学家进一步确认。
在研究的36例病例中，传统ICC和mICC在三个截断值（1%、5%和50%）下显示出良好的一致性（kappa系数>0.5；p<0.01），除了GATA3在1%截断值时。单标志物阳性的数量超过双标志物阳性，三标志物阳性极为罕见（<3%）。GATA3是所有乳腺癌亚型中主要的单标志物阳性表型，但雌激素受体阳性、孕激素受体阳性和人表皮生长因子受体2阳性（ER+/PR+/HER2+）乳腺癌中的MMG除外。限于两种标志物时，GATA3/MMG在管腔型乳腺癌（ER+/PR+/HER2+，60.6%；ER+/PR+/HER2+，31.4%）中包含的肿瘤细胞数量最多，而HER2过表达乳腺癌（27.4%）和三阴性乳腺癌（26.4%）则倾向于GATA3/GCDFP15组合。
对于单一标志物，GATA3显示出最高的敏感性。对于激素受体阳性乳腺癌添加MMG以及对于激素受体阴性乳腺癌添加GCDFP15进一步提高了敏感性。多标志物阳性细胞的比例较低表明，传统ICC观察到的共表达归因于肿瘤内异质性，而非真正的共表达。
