Gene expression analysis has identified biologically relevant subclasses of breast cancer. However, most classification schemes do not robustly cluster all HER2+ breast cancers, in part due to limitations and bias of clustering techniques used. In this article, we propose an alternative approach that first separates the HER2+ tumors using a gene amplification signal for Her2/neu amplicon genes and then applies consensus ensemble clustering separately to the HER2+ and HER2À clusters to look for further substructure. We applied this procedure to a microarray data set of 286 early-stage breast cancers treated only with surgery and radiation and identified two basal and four luminal subtypes in the HER2À tumors, as well as two novel and robust HER2+ subtypes. HER2+ subtypes had median distant metastasis-free survival of 99 months [95% confidence interval (95% CI), 83-118 months] and 33 months (95% CI, 11-54 months), respectively, and recurrence rates of 11% and 58%, respectively. The low recurrence subtype had a strong relative overexpression of lymphocyte-associated genes and was also associated with a prominent lymphocytic infiltration on histologic analysis. These data suggest that early-stage HER2+ cancers associated with lymphocytic infiltration are a biologically distinct subtype with an improved natural history. [Cancer Res 2007;67(22):10669-76]

Biomarker development for prediction of patient response to therapy is one of the goals of molecular profiling of human tissues. Due to the large number of transcripts, relatively limited number of samples, and high variability of data, identification of predictive biomarkers is a challenge for data analysis. Furthermore, many genes may be responsible for drug response differences, but often only a few are sufficient for accurate prediction. Here we present an analysis approach, the Convergent Random Forest (CRF) method, for the identification of highly predictive biomarkers. The aim is to select from genome-wide expression data a small number of non-redundant biomarkers that could be developed into a simple and robust diagnostic tool. Our method combines the Random Forest classifier and gene expression clustering to rank and select a small number of predictive genes. We evaluated the CRF approach by analyzing four different data sets. The first set contains transcript profiles of whole blood from rheumatoid arthritis patients, collected before anti-TNF treatment, and their subsequent response to the therapy. In this set, CRF identified 8 transcripts predicting response to therapy with 89% accuracy. We also applied the CRF to the analysis of three previously published expression data sets. For all sets, we have compared the CRF and recursive support vector machines (RSVM) approaches to feature selection and classification. In all cases the CRF selects much smaller number of features, five to eight genes, while achieving similar or better performance on both: training and independent testing sets of data. For both methods performance estimates using cross-validation is similar to performance on independent samples. The method has been implemented in R and is available from the authors upon request: Jadwiga.Bienkowska@biogenidec.com.

To our knowledge we report for the first time that another 5 genes (SCNN1B, SYT6, DACH1, and the tumor suppressors TFAP2A and MT1G) are hypermethylated in renal cell carcinoma. Robust computational screens and the high throughput methylation assay resulted in an enriched set of novel genes that are epigenetically altered in clear cell renal cell carcinoma. Overall the detection of hypermethylation in these highly down-regulated genes suggests that assaying for their methylation using cells from urine or blood could provide the basis for a viable diagnostic test.

Microarray gene expression profiling has been used to distinguish histological subtypes of renal cell carcinoma (RCC), and consequently to identify specific tumor markers. The analytical procedures currently in use find sets of genes whose average differential expression across the two categories differ significantly. In general each of the markers thus identified does not distinguish tumor from normal with 100% accuracy, although the group as a whole might be able to do so. For the purpose of developing a widely used economically viable diagnostic signature, however, large groups of genes are not likely to be useful. Here we use two different methods, one a support vector machine variant, and the other an exhaustive search, to reanalyze data previously generated in our Lab (Lenburg et al. 2003). We identify 158 genes, each having an expression level that is higher (lower) in every tumor sample than in any normal sample, and each having a minimum differential expression across the two categories at a significance of 0.01. The set is highly enriched in cancer related genes (p = 1.6 × 10−12), containing 43 genes previously associated with either RCC or other types of cancer. Many of the biomarkers appear to be associated with the central alterations known to be required for cancer transformation. These include the oncogenes JAZF1, AXL, ABL2; tumor suppressors RASD1, PTPRO, TFAP2A, CDKN1C; and genes involved in proteolysis or cell-adhesion such as WASF2, and PAPPA.

Background: Clustering analysis of microarray data is often criticized for giving ambiguous results because of sensitivity to data perturbation or clustering techniques used. In this paper, we describe a new method based on principal component analysis and ensemble consensus clustering that avoids these problems.

We develop a new technique to analyse microarray data which uses a combination of principal components analysis and consensus ensemble k-clustering to find robust clusters and gene markers in the data. We apply our method to a public microarray breast cancer dataset which has expression levels of genes in normal samples as well as in three pathological stages of disease; namely, atypical ductal hyperplasia or ADH, ductal carcinoma in situ or DCIS and invasive ductal carcinoma or IDC. Our method averages over clustering techniques and data perturbation to find stable, robust clusters and gene markers. We identify the clusters and their pathways with distinct subtypes of breast cancer (Luminal,Basal and Her2+). We confirm that the cancer phenotype develops early (in early hyperplasia or ADH stage) and find from our analysis that each subtype progresses from ADH to DCIS to IDC along its own specific pathway, as if each was a distinct disease.

Molecular stratification of disease based on expression levels of sets of genes can help guide therapeutic decisions if such classifications can be shown to be stable against variations in sample source and data perturbation. Classifications inferred from one set of samples in one lab should be able to consistently stratify a different set of samples in another lab. We present a method for assessing such stability and apply it to the breast cancer (BCA) datasets of Sorlie et al. 2003 and Ma et al. 2003. We find that within the now commonly accepted BCA categories identified by Sorlie et al. Luminal A and Basal are robust, but Luminal B and ERBB2+ are not. In particular, 36% of the samples identified as Luminal B and 55% identified as ERBB2+ cannot be assigned an accurate category because the classification is sensitive to data perturbation. We identify a “core cluster” of samples for each category, and from these we determine “patterns” of gene expression that distinguish the core clusters from each other. We find that the best markers for Luminal A and Basal are (ESR1, LIV1, GATA-3) and (CCNE1, LAD1, KRT5), respectively. Pathways enriched in the patterns regulate apoptosis, tissue remodeling and the immune response. We use a different dataset (Ma et al. 2003) to test the accuracy with which samples can be allocated to the four disease subtypes. We find, as expected, that the classification of samples identified as Luminal A and Basal is robust but classification into the other two subtypes is not.

We describe a new method based on principal component analysis and robust consensus ensemble clustering to identify and elucidate the subtypes of breast cancer disease. The method was applied to microarray gene expression data using micro-dissection of samples from 36 breast cancer patients with at least two of three pathological stages of disease. Controls were normal breast epithelial cells from 3 disease free patients. Our method identified an optimum set of genes and strong, stable clusters which correlated well with clinical classification into Luminal, Basal and Her2+ subtypes based on ER, PR and Her2 status. It also revealed a hierarchical portrait of disease progression through various grades and stages and identified genes and functional pathways for each stage, grade and disease subtype. We found that gene expression heterogeneity across subtypes is much greater than the heterogeneity of progression from DCIS to IDC within a subtype, suggesting that the disease subtypes are distinct disease processes. The averaging over data perturbations and clustering methods is critical in the robust identification of subtypes and gene markers for grade and progression.