TCGA Results for Endometrial Cancer...TCGA Results for Endometrial Cancer Dougg,las A. Levine , MD Head, Gynecology Research Laboratory Associate Attending Surgeon Memorial Sloan Kettering

TCGA Results for Endometrial Cancer

Douglas A. Levine, MDg ,Head, Gynecology Research Laboratory

Associate Attending SurgeonMemorial Sloan Kettering Cancer Center

@LevineMDDouglas A. Levine, MD

The face of clinical trials is changingThe face of clinical trials is changing

@LevineMDDouglas A. Levine, MD

Types of endometrial cancerTypes of endometrial cancerEndometrioid (low grade) Serous (high grade) Serous vs Endometrioid

• More solid• Less glandular• Higher grade nuclei• Higher grade nuclei• Greater N:C ratio• Loss of polarity

@LevineMDDouglas A. Levine, MD

@LevineMDDouglas A. Levine, MD

TCGA Research Network

5

BCR Pipeline – Endometrial

6

7

Sample criteria limit ‘askable’ questions

Newly diagnosed untreated invasive tumors• Newly diagnosed, untreated, invasive tumors• Snap frozen, <60min to LN2

– ~50 mg wet weight

f

10,000

• Pathology review of contributed tissue• ≥60% tumor cells, ≤20% necrosis• Matched germline DNA• Select clinical annotation• IRB approval for TCGA use and general data

sharing10 g• MTA without retention of IP

8

CANCERGENOME.NIH.GOVENDOMETRIAL RESULTS

@LevineMDDouglas A. Levine, MD

Copy number alteration clusters

10

Andrew Cherniack, Broad

Copy number alteration clusters

1q amplificationq p

24% of high-grade endometrioid tumors cluster with serous

11

Andrew Cherniack, Broadtumors cluster with serous tumors (serous-like)

Extended 1q analysisExtended 1q analysis

• Clinically relevant analysis of TCGA dataClinically relevant analysis of TCGA data• Significantly decreased PFS for 1q amp’d group

– Independent of standard histopathologyIndependent of standard histopathology– Within stage I pts

• Limited by 26 mos f/uy /• Remarkable

– Endometrioid 1q subsetEndometrioid

Serous-like

EndometrioidLow CNASerous-likeAmp1qModerate CNA

@LevineMDDouglas A. Levine, MD

Itai Pashtan, Andrew Cherniack, Broad

Mutation spectrum

High background mutation rate; ~10-20 / Mb

Low background mutation rate; ~2-3 / Mb

Very high background

20 / Mbmutation rate; ~100-500 / Mb

Cyriac Kandoth and Li Ding, WashU Niki S h lt Nil W i h ld MSKCC

Hypermutators associated with MSI and MLH1 DNA

13

Niki Schultz, Nils Weinhold, MSKCCwith MSI and MLH1 DNA promoter methylation

SMGs – Significantly Mutated Genes

14

Progression free survival

Serous poorest PFS

15

Serous – poorest PFSNo difference between MSI and MSS groupsNo events in small POLE group

Copy number spectrum

MSI, MLH1 meth, Few SCNA, high

MSS, Many SCNA, Low mutation rate

MSS, Few SCNA, Low mutation rate

mutation rate

MSS, Few SCNA, very high

Cyriac Kandoth and Li Ding, WashU Niki S h lt Nil W i h ld MSKCC

16

SCNA, very high mutation rate

Niki Schultz, Nils Weinhold, MSKCC

Integrated features

All endometrioid, TP53 mutations, few All endometrioid, PTEN mutations, few TP53 mutations

PTEN mutations, high grade tumors, serous and some endometrioid

PTEN mutations, no TP53 mutations

17

endometrioid Cyriac Kandoth and Li Ding, WashU Niki Schultz, Nils Weinhold, MSKCC

CANCERGENOME.NIH.GOVPATHWAY ANALYSES

@LevineMDDouglas A. Levine, MD

RAS/CTNNB1 pathway - MEMo

• Extensive activation of CTNNB1 through novel mechanism; different than CRC cancer• New SOX17 hotspot mutations identified

19

Niki Schultz, Nils Weinhold, MSKCC

PI3K/AKT – most active in endometrial cancer

PIK3CA

PIK3R1

20

The endometrial cancer pathwayThe endometrial cancer pathway

@LevineMDDouglas A. Levine, MD

Multiplatform molecular similarities among ovarian serous, uterine serous, basal like breast, ,

22

Andrew Cherniack, Hui Shen, Wei Zhang, Chris Benz, Peter Laird, Yuexin Liu, Christina Yau

Mutation frequencies vary across tumors

23

@LevineMDDouglas A. Levine, MDCancer Genome Atlas Research Network, et al. Nat Genet. 2013;45(10):1113-1120.

Background Mutation FrequenciesBackground Mutation Frequencies

Di t ib ti f t ti i 127Mutation frequencies, spectra,

Background Mutation FrequenciesBackground Mutation Frequencies

Distribution of mutations in 127 SMGs across Pan-Cancer cohort

and contexts across 12 cancer types

@LevineMDDouglas A. Levine, MDKandoth C, et al. Nature. 2013;502(7471):333-339.

@LevineMDDouglas A. Levine, MDCirello G, et al. Nat Genet. 2013;45(10):1127-1133.

CommentsComments

• Genomic landscape (from TCGA or other sources) willGenomic landscape (from TCGA or other sources) will form backbone for clinical trial design of targeted agents with defined mechanism

S l ti t tifi ti– Selection vs. stratification– Intelligent trials (with biospecimens)

• Expected alterations do not result in response to p ptargeted agents with defined mechanisms of action– Endometrial phase II trials of mTORi

• Properly designed clinical trials will allow us to learnProperly designed clinical trials will allow us to learn about mechanisms of response and resistance

@LevineMDDouglas A. Levine, MD