Dataset Information

Correlating nuclear morphometric patterns with estrogen receptor status in breast cancer pathologic specimens.

ABSTRACT: In this pilot study, we introduce a machine learning framework to identify relationships between cancer tissue morphology and hormone receptor pathway activation in breast cancer pathology hematoxylin and eosin (H&E)-stained samples. As a proof-of-concept, we focus on predicting clinical estrogen receptor (ER) status-defined as greater than one percent of cells positive for estrogen receptor by immunohistochemistry staining-from spatial arrangement of nuclear features. Our learning pipeline segments nuclei from H&E images, extracts their position, shape and orientation descriptors, and then passes them to a deep neural network to predict ER status. After training on 57 tissue cores of invasive ductal carcinoma (IDC), our pipeline predicted ER status in an independent test set of patient samples (AUC ROC?=?0.72, 95%CI?=?0.55-0.89, n?=?56). This proof of concept shows that machine-derived descriptors of morphologic histology patterns can be correlated to signaling pathway status. Unlike other deep learning approaches to pathology, our system uses deep neural networks to learn spatial relationships between pre-defined biological features, which improves the interpretability of the system and sheds light on the features the neural network uses to predict ER status. Future studies will correlate morphometry to quantitative measures of estrogen receptor status and, ultimately response to hormonal therapy.

SUBMITTER: Rawat RR

PROVIDER: S-EPMC6123433 | biostudies-literature | 2018

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Correlating nuclear morphometric patterns with estrogen receptor status in breast cancer pathologic specimens.

Rawat Rishi R RR Ruderman Daniel D Macklin Paul P Rimm David L DL Agus David B DB

NPJ breast cancer 20180904

In this pilot study, we introduce a machine learning framework to identify relationships between cancer tissue morphology and hormone receptor pathway activation in breast cancer pathology hematoxylin and eosin (H&E)-stained samples. As a proof-of-concept, we focus on predicting clinical estrogen receptor (ER) status-defined as greater than one percent of cells positive for estrogen receptor by immunohistochemistry staining-from spatial arrangement of nuclear features. Our learning pipeline segm ...[more]

PMID: 30211313

Similar Datasets

Project description:BackgroundEstrogens are known to regulate the proliferation of breast cancer cells and to modify their phenotypic properties. Identification of estrogen-regulated genes in human breast tumors is an essential step toward understanding the molecular mechanisms of estrogen action in cancer. To this end we generated and compared the Serial Analysis of Gene Expression (SAGE) profiles of 26 human breast carcinomas based on their estrogen receptor alpha (ER) status. Thus, producing a breast cancer SAGE database of almost 2.5 million tags, representing over 50,000 transcripts.ResultsWe identified 520 transcripts differentially expressed between ERalpha-positive (+) and ERalpha-negative (-) primary breast tumors (Fold change >or= 2; p < 0.05). Furthermore, we identified 220 high-affinity Estrogen Responsive Elements (EREs) distributed on the promoter regions of 163 out of the 473 up-modulated genes in ERalpha (+) breast tumors. In brief, we observed predominantly up-regulation of cell growth related genes, DNA binding and transcription factor activity related genes based on Gene Ontology (GO) biological functional annotation. GO terms over-representation analysis showed a statistically significant enrichment of various transcript families including: metal ion binding related transcripts (p = 0.011), calcium ion binding related transcripts (p = 0.033) and steroid hormone receptor activity related transcripts (p = 0.031). SAGE data associated with ERalpha status was compared with reported information from breast cancer DNA microarrays studies. A significant proportion of ERalpha associated gene expression changes was validated by this cross-platform comparison. However, our SAGE study also identified novel sets of genes as highly expressed in ERalpha (+) invasive breast tumors not previously reported. These observations were further validated in an independent set of human breast tumors by means of real time RT-PCR.ConclusionThe integration of the breast cancer comparative transcriptome analysis based on ERalpha status coupled to the genome-wide identification of high-affinity EREs and GO over-representation analysis, provide useful information for validation and discovery of signaling networks related to estrogen response in this malignancy.

Project description:BackgroundMTO1 and MRPL41 are nuclear-encoded mitochondrial genes encoding a mitochondrial tRNA-modifying enzyme and a mitochondrial ribosomal protein, respectively. Although both genes have been known to have potential roles in cancer, little is known about their molecular regulatory mechanism, particularly from an epigenetic approach. In this study, we aimed to address their epigenetic regulation through the estrogen receptor (ER) in breast cancer.MethodsDigital differential display (DDD) was conducted to identify mammary gland-specific gene candidates including MTO1 and MRPL41. Promoter CpG methylation and expression in breast cancer cell lines and tissues were examined by methylation-specific PCR and real time RT-PCR. Effect of estradiol (E2), tamoxifen, and trichostatin A (TSA) on gene expression was examined in ER + and ER- breast cancer cell lines. Chromatin immunoprecipitation and luciferase reporter assay were performed to identify binding and influencing of the ER to the promoters.ResultsExamination of both cancer tissues and cell lines revealed that the two genes showed an opposite expression pattern according to ER status; higher expression of MTO1 and MRPL41 in ER- and ER+ cancer types, respectively, and their expression levels were inversely correlated with promoter methylation. Tamoxifen, E2, and TSA upregulated MTO1 expression only in ER+ cells with no significant changes in ER- cells. However, these chemicals upregulated MRPL41 expression only in ER- cells without significant changes in ER+ cells, except for tamoxifen that induced downregulation. Chromatin immunoprecipitation and luciferase reporter assay identified binding and influencing of the ER to the promoters and the binding profiles were differentially regulated in ER+ and ER- cells.ConclusionsThese results indicate that different epigenetic status including promoter methylation and different responses through the ER are involved in the differential expression of MTO1 and MRPL41 in breast cancer.

Dataset Information

Correlating nuclear morphometric patterns with estrogen receptor status in breast cancer pathologic specimens.

Publications

Correlating nuclear morphometric patterns with estrogen receptor status in breast cancer pathologic specimens.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets