Genomic Characterization of Invasive Lobular Breast Carcinoma
Michael L. Gatza, Ph.D.
TCGA Breast Cancer AWG
Invasive Breast Carcinoma
Invasive Ductal Carcinoma (IDC) 50-80% Ductal
Mixed IDC.ILC 4-5%
webmd.com Lobular
2 Invasive Lobular Carcinoma (ILC) 10-15% Summary of Data Freeze
Pathology centrally re-reviewed (Andy Beck, Harvard)
817
490
127
88
112
3 Identification of differentially expressed genes
LumA LumA Ductal Lobular Ductal Lobular Normal
ATM network Immune signaling (multiple) MAPK signaling
N=663 genes 2 Class 2 Class FDR=0SAM
MYC targets (multiple) E-cadherin stabilization CDH1 Low High 4 mRNA Expression Mike Gatza, UNC Development of Integrated MAF
DNA Exome sequencing
UNCeqR (mRNAseq / DNAseq) Integrated MAF
ABRA (CDH1, TP53, GATA3, PTEN, RB1)
Integrated MAF Integrated MAF
DNA-based MAF DNA-based MAF
Matt Wilkerson (UNC), Lisle Mose (UNC) Giovanni5 Ciriello (MSKCC), Cyriac Kandoth (MSKCC) Mike McLellan (Wash U) Comparison of significant alterations: IDC vs. ILC
6 Giovanni Ciriello, MSKCC Identifying IDC LumA and ILC LumA-specific alterations
IDC ILC GATA3 (p=0.0002)
Low High Protein Expression
7 Giovanni Ciriello, MSKCC PARADIGM analysis identifies IDC and ILC- associated signaling pathways
Blue: ILC DOWN Red: ILC UP
CDH1 p53/DNA Damage Response IDC ILC mRNA MYC
IDC ILC Protein
Low High mRNA Expression
Low High Protein Expression
XBP1
Immune Related
8 Christina Yau, Buck Institute Development of mRNA-based ILC classes
1 2 3
Centroid (90genes) Centroid classifier
ConsensusClusterPlus to ID 3 ILC classes Identified samples with positive sil. width ClaNC developed centroid predictor
TCGA ILC LumA (n=106) TCGA ILC LumA (n=89) TCGA ILC LumA (n=89)
9 Mike Gatza, UNC 2 Class SAM identifies differentially expressed genes in ILC classes
Class1 C2 C3
N= 722 genes EGFR MET GLI1 FGF17
WNT6 AREG KIT
KRT 14, 15, 17, 32, 81 KRK 1, 6-8 CLDN 8,10,11, 19 PTCH2 988 genes (FDR=0) 988 genes TP63 VIT ID4
N= 268 genes Immune-related genes: >100
Low High LCK mRNA Expression IFNG 10 Mike Gatza, UNC ILC class mRNA / miRNA expression patterns correspond with IDC and Adjacent Normal
ILC IDC Norm
FDR<0.05) SAMseq (
988 genes (FDR=0) miRNA
LowILC IDC High Normal mRNA Expression Class 1 61 49 94 Class 2 39 167 0 Low High miRNA Expression Class 3 27 274 0 11 Mike Gatza, UNC P<0.0001 Reanne Bowlby, BC Cancer Agency ILC Class1 corresponds with RPPA Reactive Subtype
ILC Class (n=127) Class1 Class2 Class3 Reactive P<0.0001 Non-reactive RPPA Subtype (n=70) Missing data
Annexin1 Caveolin1 Collagen IV RPPA Myh11 RMB15
Low High Protein Expression
Mike Gatza, UNC 12 Gordon Mills, MDACC ILC Class1 tumors exhibit altered PDGFR/ STAT3 and FoxM1 signaling
FOXM1 sub-network
PDGFR
PARAGIGM FoxM1
pSRC Y527 pSTAT3 Y705
RPPA FoxM1
Low High Protein Expression
Christina Yau, Buck Institute 13 Mike Gatza, UNC ILC class 2 defined by high immune signaling and proliferation
Class1 Class2 Class3 B-cell BCR B-cell (CS) CD8 CD8 (CS) T-cell (CS) LCK T-cell (TNBC) NK T-cell (Teschendo) MΦ TH1 (CS) MΦ CSF1 TCR TCGA ILC (n=127) Low High Signature Score
P=1.39e-10 ProliferationScore (PAM50) 14 Class 1 Class 2 Class 3 Mike Gatza, UNC PARADIGM analysis identifies IFNG and FOXM1 as key pathways in ILC class 2 tumors
15 Christina Yau, Buck Institute Summary
• Developed unique integrated MAF utilizing both DNA exome and mRNA sequencing
• ILC vs. IDC – FOXA1, CDH1 mutations associated with ILC – GATA3 mutation associated with IDC – Altered signaling: CDH1, Myc, p53/DNA damage, immune signaling – Identified differentially expressed miRNA and methylation
• ILC classes – Class 1 associated with Reactive subtype – Class 2 immune component and highly proliferative
16 TCGA Breast Cancer Analysis Working Group
Baylor College of Medicine The University of Texas University of California Chad Creighton MD Anderson Cancer Center Santa Cruz Xiaosong Wang Roel Verhaak Buck Institute Rehan Akbani Chris Benz British Columbia Cancer Nancy Shih David Haussler Agency Gordon Mills Christina Yau Andy Chu Sam Ng Elizabeth Chun Memorial Sloan-Kettering University of North Ted Goldstein Andy Mungall Cancer Center Carolina Kyle Ellrott Gordon Robertson Giovanni Ciriello Chuck Perou (co-chair) Charlie Vaske Dominik Stoll Niki Schultz Katie Hoadley Josh Stuart Ethan Cerami Mike Gatza Broad Institute Jing Zhu Arthur Goldberg Joel Parker Andrew Cherniack Caitlin Byrne University of California, Xiaping He Greater Poland Cancer Center Anders Jacobsen San Francisco Michael Iglesia Maciej Wiznerowicz Tari King Fred Waldman Grace Silva
Chris Sander Wei Zhao Harvard Medical School University of Southern Terrence Wu Nationwide Children’s California The Genome Institute Yonghong Xiao Hospital Peter Laird at Washington University Jay Bowen Institute for Systems Biology Swapna Mahurkar Matthew Ellis (co-chair) Julie Gastier-Foster Sheila Reynolds Simeen Malik Li Ding Ilya Shmulevich National Cancer Institute Dan Weisenberger Lucinda Fulton Chunhua Yan Daniel Koboldt Lawrence Berkeley National John Demchok Windber Research Elaine Mardis Laboratory Laura Dillon Institute Paul Spellman Margi Sheth Hai Hu Mayo Clinic Peter Good Richard Mural Jim Ingle Jacqueline Palchik 17 Heidi Sofia Kenna Shaw