Novel Image-Based Screening of Mammary Tumors

描述(申请人提供)：通过自我检查、临床检查和乳房X光检查发现乳腺肿瘤，并对可疑病变进行活检。随后的组织学分类在治疗决策中起着决定性的作用，但相关的恶性风险、适当的治疗和复发的风险很难确定。因此，患者的治疗是基于流行病学结果，而不是个人需求。非侵入性成像提供了一些信息，但通过常规筛查进行早期乳腺癌检测目前还不可行。然而，一旦检测到肿瘤并进行活组织检查，就迫切需要新的方法来辅助检测和分类病变的亚类。细胞表型的一个关键的表观遗传标记是核蛋白的组织，它指导和反映正常的细胞功能。基于染色质相关蛋白的重新分布反映了伴随着细胞表型变化的基因表达的变化这一假设，我们开发了图像分析方法来从三维、高分辨率、荧光免疫染色的图像中定量特定染色质相关蛋白的核分布。通过将这些方法应用于模拟正常和恶性乳腺上皮组织的培养模型，我们已经证明了核有丝分裂装置蛋白(NUMA)和赖氨酸-20甲基化的异染色质相关蛋白-4(H4-K20M)的分布是能够明确区分非肿瘤性和恶性的人乳腺上皮细胞的生物标志物。该项目的目标是量化特定核蛋白在模拟癌前和恶性组织的培养模型中的分布，以揭示癌前疾病的表观遗传学特征。我们的基于图像的方法将被扩展，以产生一种新的技术，即表型组织图，能够以细胞分辨率解析局部组织表型，并揭示组织形态和行为的细微差异。这项技术将与常规的H&E染色和组织学技术一起使用，将在各种癌前肿瘤的针芯活检上进行测试，目的是确定分级病变的亚类。结果将与初始针芯的组织病理学和后续的手术活检相关联，以期预测初始筛查漏掉的更多侵袭性表型。这项研究的未来公共卫生效益是帮助乳腺癌患者的治疗决策过程。通过更好地了解细胞内分子成分的组织以及这些成分的组织在癌变过程中如何改变，我们可以开发并为病理学家提供新的图像分析工具，以帮助和支持活检的乳腺组织的组织学分类。

DESCRIPTION (provided by applicant): Breast tumors are detected by self exam, clinical exam, and mammogram and suspicious lesions are biopsied. The ensuing histological classification plays a determining role in the treatment decision but the associated risk of malignancy, the appropriate treatment, and the risk of reoccurrence are difficult to determine. As a consequence, patient treatment is based on epidemiological findings rather than individual needs. Non- invasive imaging provides some information but early breast cancer detection by routine screening is not currently feasible. However, once a tumor has been detected and biopsied, there is urgent need for novel methods to aid in the detection and classification of sub-classes of lesions. One key epigenetic marker of cell phenotype is the organization of nuclear proteins which direct and reflect normal cell function. Based on the hypothesis that the redistribution of chromatin-related proteins reflects changes in gene expression that accompany alterations in cell phenotype, we have developed image analysis methodologies to quantify the nuclear distribution of specific chromatin-associated proteins from three-dimensional, high-resolution, fluorescently immunostained images. By applying these methods to culture models that mimic normal and malignant breast epithelial tissue, we have demonstrated that the distribution of nuclear mitotic apparatus protein (NuMA) and heterochromatin related protein histone-4 methylated on lysine-20 (H4-K20m) are biomarkers capable of clearly distinguishing non-neoplastic and malignant human mammary epithelial cells. The goal of this project is to quantify the distribution of specific nuclear proteins in culture models that mimic premalignant and malignant tissue to uncover epigenetic characteristics of premalignant disease. Our image-based methodologies will be expanded to produce a novel technique, the phenotype tissue-map, capable of resolving local tissue phenotype at cellular resolution and uncovering subtle differences in tissue morphology and behavior. The technology, which will work alongside the usual H&E staining and histological techniques, will be tested on needle-core biopsies of a variety of premalignant tumors with the aim of defining sub-classes of graded lesions. The results will be correlated with the histopathology of the initial needle-core and the follow-up surgical biopsy with the hope of predicting more aggressive phenotypes missed by the initial screen. The future public health benefit of this research is to aid the treatment decision process of breast cancer patients. By better understanding the organization of molecular components within the cell and how the organization of these components is altered during the progression to cancer, we can develop and provide pathologists with novel image analysis tools to aid and support the histological classification of biopsied breast tissue.