Building Prognostic Models for Breast Cancer Patients Using Clinical Variables and Gene Expression Signatures
ABSTRACT: Our findings indicate that the integration of expression signatures and clinicopathological factors can better determine the individual risk of recurrence for newly diagnosed patients with lymph-node negative ER-positive breast cancer. Models incorporating other variables yet to be discovered will be needed to obtain robust prognostic models for ER-negative and HER2-positive breast cancer patients. A large data set was created by combining five different publicly available microarray datasets of node-negative breast cancer patients treated with local therapy only. The microarray gene expression data was combined using the batch effect adjustment by the Distance Weighted Discrimination method.
Project description:3 fresh frozen Breast tumor from the University of North Carolina at Chapel Hill (UNC) were obtained from the UNC Tissue Procurement Facility under an IRB approved protocol. Each experimental sample was ssayed versus a common reference sample that was a modified version of the Stratagene Human Universal Reference (which was further augmented with a 1/10 amount of MCF7 mRNA and 1/10 amount of ME16C mRNA) Keywords = Agilent microarray Keywords = EGFR Keywords = keratin Keywords: parallel sample
Project description:Recently, long oligonucleotide (60-70mer) microarrays for two-color experiments have been developed and are gaining widespread use. In addition, when there is limited availability of mRNA from tissue sources, RNA amplification can and is being used to produce sufficient quantities of cRNA for microarray hybridization. Taking advantage of the selective degradation of RNA under alkaline conditions, we have developed a method to “strip” glass-based oligonucleotide microarrays that use fluorescent RNA in the hybridization, while leaving the DNA oligonucleotide probes intact and usable for a second experiment. Replicate microarray experiments conducted using stripped arrays showed high reproducibility, however, we found that arrays could only be stripped and reused once without compromising data quality. The intraclass correlation (ICC) between a virgin array and a stripped array hybridized with the same sample showed a range of 0.90-0.98, which is comparable to the ICC of two virgin arrays hybridized with the same sample. Using this method, once stripped oligonucleotide microarrays are usable, reliable, and should help to reduce costs. Keywords = Agilent microarray Keywords = Stripped arrays Keywords = Replicate reproducibility Keywords: other
Project description:Five molecular subtypes (Luminal A/B, HER2-enriched, Basal-like, and Claudin-low) with clinical implications have been identified. In this report, we evaluated molecular and phenotypic relationships of a large in vitro panel of human breast cancer cell lines (BCCLs), human mammary fibroblasts (HMFs) and human mammary epithelial cells (HMECs) with (1) breast tumors, (2) normal breast cell-enriched subpopulations and (3) human embryonic stem cells (hESCs) and bone marrow-derived mesenchymal stem cells (hMSC). First, by integrating genomic data of 337 breast samples with 93 cell lines we were able to identify all the intrinsic tumor subtypes in vitro, except for the Luminal A. Secondly, we observed that cell lines recapitulate the differentiation hierarchy observed in the mammary gland, with Claudin-low BCCLs and HMFs cells showing a stromal phenotype, HMECs showing a mammary stem cell/bipotent progenitor phenotype, Basal-like cells showing a luminal progenitor phenotype, and Luminal B cells showing a luminal phenotype. Thirdly, we identified Basal-like and highly migratory Claudin-low subpopulations of cells within a subset of triple-negative BCCLs (SUM149PT, HCC1143 and HCC38). Interestingly, both subpopulations within SUM149PT where found to have Tumor Initiating Cell (TIC) features, but the Basal-like subpopulation grew faster than the Claudin-low subpopulation. Finally, Claudin-low BCCLs were found to resemble the phenotype of hMSCs, whereas hESCs cells were found to have an epithelial phenotype without basal and luminal differentiation. The results presented here should help improve our understanding of the cell line model system through the appropriate pairing of cell lines with relevant in vivo tumor and normal cell counterparts. reference x sample
Project description:Introduction: In breast cancers, the basal-like subtype has high levels of genomic instability relative to other breast cancer subtypes with many basal-like-specific regions of aberration. There is evidence that this genomic instability extends to smaller scale genomic aberrations as well, as shown by a previously described micro-event in the PTEN gene in the Basal-like SUM149 breast cancer cell line. Methods: We sought to identify if small regions of genomic change exist by using a high density, gene centric Comparative Genomic Hybridizations (CGH) array on both cell lines and primary tumors. A custom Agilent tiling array for CGH (244,000 probes, 200bp tiling resolution) was created to identify small regions of genomic change and was focused on previously identified basal-like-specific, and general cancer genes. Tumor genomic DNA from 94 patients and 2 breast cancer cell lines was labeled and hybridized to these arrays. Aberrations were called using SWITCHdna and the smallest 25% of SWITCHdna-defined genomic segments being called micro-aberrations (<64 contiguous probes, ~ <15kb). Results: Our data showed that primary tumor breast cancer genomes frequently contained areas of small-scale copy number gains and losses, termed micro-aberrations, which are undetectable using lower-density genome-wide platforms. The basal-like subtype exhibited the highest incidence of these events. These micro-aberrations sometimes altered expression of the involved gene as suggested by data from microarray and mRNA-seq studies. We confirmed the presence of the PTEN micro-amplification in SUM149 and by mRNA-seq showed that this resulted in loss of expression of all exons downstream of this event. Micro-aberrations disproportionately affected the 5’ regions of the affected genes, including the promoter region, and a high frequency of micro-aberrations was associated with poor survival outcomes. Conclusion: Using a high probe density, gene-centric aCGH microarray, we present evidence of small-scale genomic aberrations that contribute to gene inactivation, and thus, genomic instability and tumor formation through a mechanism not detected using conventional copy number analyses. reference x sample
Project description:Breast cancer subtypes identified in genomic studies have different underlying genetic defects. Mutations in the tumor suppressor p53 occur more frequently in estrogen receptor (ER) negative, basal-like and HER2-amplified tumors than in luminal, ER positive tumors. Thus, because p53 mutation status is tightly linked to other characteristics of prognostic importance, it is difficult to identify p53's independent prognostic effects. The relation between p53 status and subtype can be better studied by combining data from primary tumors with data from isogenic cell line pairs (with and without p53 function). In this study, the p53-dependent gene expression signatures of four cell lines (MCF-7, ZR-75-1, and two immortalized human mammary epithelial cell lines) were identified by comparing p53-RNAi transduced cell lines to their parent cell lines. Cell lines were treated with vehicle only or doxorubicin to identify p53 responses in both non-induced and induced states. Each cell line displayed unique patterns of gene expression, but cell type specific trends were evident. A common gene expression signature associated with p53 loss across all four cell lines was identified. This signature showed overlap with the signature of p53 loss in primary breast tumors and predicted relapse-free survival and overall survival in independent test data sets. Experiment Overall Design: We analyzed 48 arrays performed using 48 polyA RNA samples. RNAs were collected from cell lines treated with an IC50 dose of doxorubicin hydrochloride or with a feeding control. Each cell line had its own reference which represented the second sample on the dual channel array. These untreated RNAs were prepared by pooling four harvests of that cell line at 60-80% confluence and 48h after feeding
Project description:Results When compared with each other, primary tumors and regional metastases showed statistically indistinguishable gene expression patterns. Supervised analyses comparing patients with distant metastases versus primary tumors or regional metastases showed that the distant metastases were distinct and distinguished by the lack of expression of fibroblast/mesenchymal genes, and by the high expression of a 13-gene profile (that is, the ‘vascular endothelial growth factor (VEGF) profile’) that included VEGF, ANGPTL4, ADM and the monocarboxylic acid transporter SLC16A3. At least 8 out of 13 of these genes contained HIF1α binding sites, many are known to be HIF1α-regulated, and expression of the VEGF profile correlated with HIF1α IHC positivity. The VEGF profile also showed prognostic significance on tests of sets of patients with breast and lung cancer and glioblastomas, and was an independent predictor of outcomes in primary breast cancers when tested in models that contained other prognostic gene expression profiles and clinical variables. Conclusions These data identify a compact in vivo hypoxia signature that tends to be present in distant metastasis samples, and which portends a poor outcome in multiple tumor types. Microarrays and immunohistochemistry were used to analyze primary breast tumors, regional (lymph node) metastases, and distant metastases in order to identify biological features associated with distant metastases.
Project description:Purpose: The biological subtypes of breast cancer designated as Luminal A, Luminal B, HER2+/ER-, and Basal-like are clinically important for prognosis and planning treatment strategies. Recognizing that there is a continuum in both the spectrum of breast cancer disease and the risk of survival, we sought to develop a clinical test for the biological subtypes using a supervised risk classier.Methods: Microarray and real-time quantitative RT-PCR (qRT-PCR) data from 189 samples, procured as fresh-frozen and formalin-fixed, paraffin-embedded tissues, were used to statistically select prototypical samples and genes for the biological subtypes of breast cancer. Predictions for biological subtype and risk of recurrence were determined for different stages of disease, treatments, and across analytical platforms. Results: The biological subtype predictions on a large combined microarray test set showed prognostic significance across all patients (1244 subjects; p<0.0001), on node negative patients with no adjuvant systemic therapy (738 subjects; p<0.0001), and on patients treated with endocrine therapy (404 subjects; p=0.001). Analysis of a neoadjuvant chemotherapy study revealed a high pathologic complete response (pCR) rate in HER2+/ER- and Basal-like patients. The subtype and risk predications were also highly significant when using the qRT-PCR assay from archived FFPE breast cancers. Conclusion: Our risk predictor based on distance to biological subtype centroids provides a continuous risk score that applies to all stages of breast cancer given current therapies. The assay can be performed using archived breast tissues and a real-time qRT-PCR assay, thus facilitating application to retrospective cohorts and clinical samples. Keywords: reference x sample Comparison of reference samples against treatment
Project description:Metaplastic breast cancers (MBC) are aggressive, chemoresistant tumors characterized by lineage plasticity. To advance understanding of their pathogenesis and relatedness to other breast cancer subtypes, 28 MBCs were compared with common breast cancers using comparative genomic hybridization, transcriptional profiling, and reverse-phase protein arrays and by sequencing for common breast cancer mutations. MBCs showed unique DNA copy number aberrations compared with common breast cancers. PIK3CA mutations were detected in 9 of 19 MBCs (47.4%) versus 80 of 232 hormone receptor-positive cancers (34.5%; P = 0.32), 17 of 75 HER-2-positive samples (22.7%; P = 0.04), 20 of 240 basal-like cancers (8.3%; P < 0.0001), and 0 of 14 claudin-low tumors (P = 0.004). Of 7 phosphatidylinositol 3-kinase/AKT pathway phosphorylation sites, 6 were more highly phosphorylated in MBCs than in other breast tumor subtypes. The majority of MBCs displayed mRNA profiles different from those of the most common, including basal-like cancers. By transcriptional profiling, MBCs and the recently identified claudin-low breast cancer subset constitute related receptor-negative subgroups characterized by low expression of GATA3-regulated genes and of genes responsible for cell-cell adhesion with enrichment for markers linked to stem cell function and epithelial-to-mesenchymal transition (EMT). In contrast to other breast cancers, claudin-low tumors and most MBCs showed a significant similarity to a "tumorigenic" signature defined using CD44(+)/CD24(-) breast tumor-initiating stem cell-like cells. MBCs and claudin-low tumors are thus enriched in EMT and stem cell-like features, and may arise from an earlier, more chemoresistant breast epithelial precursor than basal-like or luminal cancers. PIK3CA mutations, EMT, and stem cell-like characteristics likely contribute to the poor outcomes of MBC and suggest novel therapeutic targets. Comparison of reference samples against treatment
Project description:Breast cancer is no longer viewed as a homogenous disease, but rather a compilation of several distinct subtypes as defined by microarray or other large scale genomic analyses. Based on prior reports, we hypothesized that younger women’s breast tumors would be enriched for more aggressive subtypes (i.e. Basal-like) and higher grade, and that age-specific gene expression differences may be highly dependent on subtype classification and/or grade. Using two independent datasets, our current analysis shows that breast tumors arising in women aged = 45 years are enriched for the Basal-like subtype (and higher grade) while those aged = 65 years are enriched for Luminal tumors. Moreover, when evaluating gene expression differences between age-defined groups, sizable gene lists were identified which diminished to few, if any, age-specific genes when statistically correcting for significant clinical factors (i.e. subtype, grade, etc). Keywords: reference x sample 344 breast tumor samples hybridized with Stratagene common reference and profiled on Agilent microarrays.
Project description:The ability to predict metastatic potential is of clinical and biological importance. Numerous metastasis/relapse predictors exist for breast cancer patients; however, what is less well established is whether predicting metastasis to specific organs sites is feasible. In this study we sought to determine: 1) the degree to which gene signatures vary across tumors and their metastases, 2) if genomic intrinsic subtypes associate with particular organs of relapse, and 3) if other genomic signatures can predict spread to specific organs. Using a gene expression microarray data set of >1000 breast tumors and metastases, we observed that >90% of 298 gene signatures were similarly expressed between matched pairs of breast tumors and metastases; those most altered were reflective of cell types including fibroblasts and immune cells. Significant associations were identified between tumor subtypes and organ of first relapse. Among these, HER2-enriched tumors were significantly associated with liver, and Basal-like and Claudin-low tumors with brain and lung. Correspondingly, previously published brain and lung metastasis signatures, along with embryonic stem cell and tumor initiating cell signatures, were also associated with Basal-like and Claudin-low subtypes. These signatures strongly correlated with low Differentiation Scores (DS) and, to a lesser extent, high proliferation. Interestingly, within Basal-like and Claudin-low tumors, low DS further predicted for brain and lung metastases. In total, intrinsic subtype and DS provide clinically useful information that identifies the distant organ sites that should be most closely monitored for signs of disease recurrence. 414 samples profiled on Agilent microarrays.