Skip to main content
main-content

01.12.2017 | Research article | Ausgabe 1/2017 Open Access

Breast Cancer Research 1/2017

Integrative clustering reveals a novel split in the luminal A subtype of breast cancer with impact on outcome

Zeitschrift:
Breast Cancer Research > Ausgabe 1/2017
Autoren:
Miriam Ragle Aure, Valeria Vitelli, Sandra Jernström, Surendra Kumar, Marit Krohn, Eldri U. Due, Tonje Husby Haukaas, Suvi-Katri Leivonen, Hans Kristian Moen Vollan, Torben Lüders, Einar Rødland, Charles J. Vaske, Wei Zhao, Elen K. Møller, Silje Nord, Guro F. Giskeødegård, Tone Frost Bathen, Carlos Caldas, Trine Tramm, Jan Alsner, Jens Overgaard, Jürgen Geisler, Ida R. K. Bukholm, Bjørn Naume, Ellen Schlichting, Torill Sauer, Gordon B. Mills, Rolf Kåresen, Gunhild M. Mælandsmo, Ole Christian Lingjærde, Arnoldo Frigessi, Vessela N. Kristensen, Anne-Lise Børresen-Dale, Kristine K. Sahlberg, OSBREAC
Wichtige Hinweise

Electronic supplementary material

The online version of this article (doi:10.​1186/​s13058-017-0812-y) contains supplementary material, which is available to authorized users.

Abstract

Background

Breast cancer is a heterogeneous disease at the clinical and molecular level. In this study we integrate classifications extracted from five different molecular levels in order to identify integrated subtypes.

Methods

Tumor tissue from 425 patients with primary breast cancer from the Oslo2 study was cut and blended, and divided into fractions for DNA, RNA and protein isolation and metabolomics, allowing the acquisition of representative and comparable molecular data. Patients were stratified into groups based on their tumor characteristics from five different molecular levels, using various clustering methods. Finally, all previously identified and newly determined subgroups were combined in a multilevel classification using a “cluster-of-clusters” approach with consensus clustering.

Results

Based on DNA copy number data, tumors were categorized into three groups according to the complex arm aberration index. mRNA expression profiles divided tumors into five molecular subgroups according to PAM50 subtyping, and clustering based on microRNA expression revealed four subgroups. Reverse-phase protein array data divided tumors into five subgroups. Hierarchical clustering of tumor metabolic profiles revealed three clusters. Combining DNA copy number and mRNA expression classified tumors into seven clusters based on pathway activity levels, and tumors were classified into ten subtypes using integrative clustering. The final consensus clustering that incorporated all aforementioned subtypes revealed six major groups. Five corresponded well with the mRNA subtypes, while a sixth group resulted from a split of the luminal A subtype; these tumors belonged to distinct microRNA clusters. Gain-of-function studies using MCF-7 cells showed that microRNAs differentially expressed between the luminal A clusters were important for cancer cell survival. These microRNAs were used to validate the split in luminal A tumors in four independent breast cancer cohorts. In two cohorts the microRNAs divided tumors into subgroups with significantly different outcomes, and in another a trend was observed.

Conclusions

The six integrated subtypes identified confirm the heterogeneity of breast cancer and show that finer subdivisions of subtypes are evident. Increasing knowledge of the heterogeneity of the luminal A subtype may add pivotal information to guide therapeutic choices, evidently bringing us closer to improved treatment for this largest subgroup of breast cancer.
Zusatzmaterial
Additional file 1: a Log2-transformed and median-centered RPPA data. b Molecular classifications of the 425 Oslo2 tumors. c Pearson correlation values and corresponding p-values calculated for correlation between each molecular subtype level and each COCA cluster by giving 0/1 numerical values to the binary categorical variables. d PAM50 distribution in total number and percentage in each of the six COCA clusters. e P-values for association between each of the six COCA clusters and clinical/molecular classification of the tumors (chi-squared association test). The p-values are Bonferroni-corrected for multiple testing. f miRNAs significantly differentially expressed (Benjamini-Hochberg adjusted p-value <0.01 and log2 |fold-change| >1) between luminal A samples in COCA cluster 1 and COCA cluster 4. g Annotation of the 1808 genes that were correlated with the 71 miRNAs differentially expressed between luminal A tumors in COCA cluster 1 versus COCA cluster 4 (absolute Spearman correlation >0.4). h Pathways enriched among the 850 genes upregulated in luminal A tumors in COCA cluster 4. i miRNAs differentially expressed between luminal A tumors in COCA cluster 1 vs COCA cluster 4 and predicted target genes that were among the proteins differentially expressed between luminal A tumors in COCA cluster 1 vs COCA cluster 4. j Comparison of the composition of RPPA-defined subtypes in luminal A tumors separated on miRNA expression. The RPPA subtype data are taken from [ 9].(XLSX 871 kb)
Additional file 2: a Supplementary methods. b Summary of clinicopathological properties of the 425 primary breast tumors in the Oslo2 cohort. (PDF 137 kb)
13058_2017_812_MOESM2_ESM.pdf
Additional file 3: Four miRNA patient clusters (1‒4) derived from clustering the expression of 421 miRNAs using Pearson correlation and complete linkage. The PART algorithm was used to identify clusters [23]. (PDF 136 kb)
13058_2017_812_MOESM3_ESM.pdf
Additional file 4: P-values and log2 fold-change resulting from testing miRNA differential expression between one miRNA cluster versus all other clusters using Wilcoxon rank-sum tests. P-values are corrected for multiple testing using Benjamini-Hochberg false discovery rate correction. (XLSX 85 kb)
13058_2017_812_MOESM4_ESM.xlsx
Additional file 5: Oslo2 tumors sorted according to membership of each of seven PARADIGM clusters (columns) with heatmap representation of the top 253 most deregulated pathway entities (IPLs) across the clusters (rows; filtering out IPLs with activity -0.25 > x < 0.25). IPL name details can be seen by zooming in. (PDF 2274 kb)
13058_2017_812_MOESM5_ESM.pdf
Additional file 6: P-values and statistics from analysis of variance identifying the top 500 pathway entities defining the seven PARADIGM clusters. (XLSX 39 kb)
13058_2017_812_MOESM6_ESM.xlsx
Additional file 7: Clinical and molecular distribution in the six COCA clusters. (PDF 7 kb)
13058_2017_812_MOESM7_ESM.pdf
Additional file 9: Pathway enrichment map of genes correlated with miRNAs differentially expressed between luminal A tumors in COCA cluster 1 and COCA cluster 4 and upregulated in luminal A tumors in COCA cluster 4. A blue line connects any two pathways when there are more than five genes in common between them (exact number indicated). Ingenuity Pathway Analysis (IPA) was used to identify enriched pathways among the upregulated genes. (PDF 58 kb)
13058_2017_812_MOESM9_ESM.pdf
Additional file 10: Six proteins differentially expressed between luminal A samples in COCA cluster 1 versus COCA cluster 4. (PDF 771 kb)
13058_2017_812_MOESM10_ESM.pdf
Literatur
Über diesen Artikel

Weitere Artikel der Ausgabe 1/2017

Breast Cancer Research 1/2017 Zur Ausgabe

Neu im Fachgebiet Onkologie

Mail Icon II Newsletter

Bestellen Sie unseren kostenlosen Newsletter Update Onkologie und bleiben Sie gut informiert – ganz bequem per eMail.

Bildnachweise