Genetic complexity in MPN, MDS/MPN and MDS · 2015. 2. 20. · 2 OH CH 2 OH TET2 IDH1/2* isocitrate α-KG 2HG EZH2 PRC2 ASXL1 H3K27me3 ET PV PMF CMML 1-5% 5-10% 5-10% 40-50% ASXL1:

Genetic complexity in MPN, MDS/MPN and MDS

Nick Cross

Wessex Regional Genetics Laboratory, Salisbury

Faculty of Medicine, University of Southampton

Genetic complexity in chronic myeloid neoplasms

• Classes of mutations

• Complexity, subclonality and prognosis

• Specificity of mutations

Myeloproliferative disorders and myelodysplastic syndromes

MPN MDS MDS/MPN

Proliferation Effective haemopoiesis

Dysplasia Ineffective haemopoiesis

ET PV

PMF CML CEL CNL

MPN-U

CMML JMML aCML

MDS/MPN-U RARS-T

RCUD RCMD RARS RAEB

MDS-U

Mutation discovery: cytogenetics, DNA arrays, exome/genome sequencing

Reiter et al., Blood 1998;92:1735-42; Gelsi-Boyer et al., Br J Haematol. 2009;145:788-800; Ernst et al., Nat Genet. 2010;42:722-6

138 der(13)

13

12

11.1

11.2

11

12.212.1

12.3

14.214.3

14.1

21.3

21.121.2

22

31

32

33

34

13

24.1

24.2

24.3

23

13

12

12

22

11.21

11.111.2

21.121.221.3

23.123.223.3

21.1

21.2

21.3

22.122.2

22.3

11.2211.23

11.1

der(8)

normal 8

der(8)

normal 13

der(13)

16-L19

normal 8

der(8)

der(13)

normal 13

52-M15

Classes of somatically acquired driver mutation in MPN, MDS/MPN and MDS

SIGNALLING EPIGENETIC mRNA

SPLICING TRANSCRIPTION/

REPAIR COHESIN CYTO

TK fusions JAK2 MPL CBL NF1

NRAS KRAS CSF3R RIT1

PTPN11 FLT3 KIT

SETBP1?? CALR??

TET2 DNMT3A IDH1/2 EZH2

ASXL1 PHF6

CREBBP EP300

SF3B1 SRSF2 U2AF1 ZRSR2 LUC7L2 PPRF8

RUNX1 TP53 ETV6 BCOR CUX1

STAG2 SMC1A SMC3 RAD21

5q- -7/7q

+8 +19

i(17q) del(11q) del(12p) del(20q) inv(3q) t(3;3)

No abnormality defines specific disease entities except BCR-ABL1



REPAIR COHESIN CYTO



PTPN11 FLT3 KIT

SETBP1 CALR


ASXL1 PHF6

CREBBP EP300


RUNX1 TP53 ETV6 BCOR CUX1


5q- -7/7q

+8 +19


Drug targetable mutations?



REPAIR COHESIN CYTO



PTPN11 FLT3 KIT

SETBP1 CALR


ASXL1 PHF6

CREBBP EP300


RUNX1 P53

ETV6 BCOR CUX1


5q- -7/7q

+8 +19


Classes of somatically acquired driver mutation in MPN, MDS/MPN and MDS

MPN MDS MDS/MPN

Signalling

Epigenetic, transcription

Splicing, cohesins

Signalling abnormalities

• Activate growth factor signalling pathways

• Thought to be largely responsible for proliferative phenotypes

• Gain of function mutations in transducers of signalling – JAK2 V617F, KIT D816V, RAS mutations

– Tyrosine kinase fusion genes e.g. BCR-ABL1, FIP1L1-PDGFRA

• Loss of function mutations of negative regulators of signalling, – eg CBL, SH2B3 (LNK)

Targeted therapy for myeloid disorders with activated tyrosine kinases

Less aggressive More aggressive

MPN with FGFR1

Fusion genes

MPN with JAK2

Fusion genes

MPN with FLT3

Fusion genes

Mastocytosis with KIT D816V

MPN with PDGFR

Fusion genes

BCR-ABL1 positive

CML

MPN with JAK2 V617F

imatinib

imatinib

FLT3 inhibitors

ruxolitinib

ponatinib dovitinib midostaurin

JAK2 inhibitors

LOC113386 19q13

CNTRL 9q33

ZNF198 13q12

FGFR1OP2 12p11

CFS1 12q15

TRIM24 7q34

N=13

FGFR1 8p11

PCM1 8p21

LRRFIP1 2q37

JAK2 9p34

FLT3 13q12

SYK 9q22

LYN 8q12

SPTBN1 2p16

MYO18A 17q11

RABEP1 17p13

PDE4DIP 1q22

WDR48 3p22

BIN2 12q13

HIP1 7q11

GOLGA4 3p22

CCDC6 10q21

TRIP11 14q32

CCDC88C 14q32

GIT2 12q24

GPIAP1 11p13

PRKG2 4q21

TPM3 1q21

NIN 14q24

SPECC1 17p11

TP53BP1 15q22

NDE1 16p13

N=25

N=7

ERC1 12p13

SART3 12q23

DTD1 20p11

PDGFRB 5q33

PDGFRA 4q12

ABL 9q34

Tyrosine kinase fusions in MLN-eo and related MPN

September 2014: 60 tyrosine kinase fusion genes

ETV6 12p13

BCR 22q11

CDK5RAP2 9q33

FIP1L1 4q12

KIF5B 10p11

STRN 2p24

KANK1 9p24

CUX1 7q22

KIT 4q12

RPN1 3q21

RANBP2 2q13

ALK 2p23

CEP85L 6q22

FOXP1 3p14

NTRK3 15q25

TPR1 1q25

RET 10q11

FGFR1OP 6q27

GOLGB11 3q12

Identification of BCR-ABL1 negative, imatinib responsive patients

• All known imatinib-responsive fusions are associated with cytogenetically visible abnormalities except for FIP1L1-PDGFRA.

– Screen patients with Eos-MPN or persistent unexplained eosinophilia

for FIP1L1-PDGFRA (and BCR-ABL1).

– Only screen for other fusions if indicated by karyotype: abnormalities of 4q11-12 (PDGFRA), 5q31-33 (PDGFRB), 9q34 (ABL)

– If cytogenetics fails consider split apart FISH

– Screen suspected mastocytosis for KIT D816V; only screen for FIP1L1-PDGFRA if eosinophilia present.

Identification of BCR-ABL1 negative, imatinib responsive patients

• Most BCR-ABL1 negative imatinib responders are male (>10:1 ratio for PDGFR fusions) with eosinophilia.

• BCR-ABL1 negative imatinib responders are very rare.

JAK2 fusions: clinical responses to ruxolitinib

Schwaab et al., Ann Hematol. 2015;94(2):233-8

STAT5 phosphoflow to identify potential responders to TKI therapy?

Roberts et al., N Engl J Med. 2014;371:1005-15

0

20

40

60

80

100

120

1 2

unt

dasatinib

Sample 1 and Sample 2 response with dasatinib

• Untreated vs +imatinib, dasatinib, ruxolitinib, ponatinib (1 hour)

• Overnight fix, permeabilisation

untreated

dasatinib

untreated

dasatinib

Signalling abnormalities in MDS/MPN

• Up to 50% of CMML cases have abnormalities that activate TK/RAS signalling

• Higher frequency in proliferative (WBC >13.0 × 109/L) vs dysplastic CMML

• 10-60% of CNL/aCML cases have activating CSF3R mutations

Gene Mutated in CMML

NRAS 11%

CBL 10%

JAK2 8%

KRAS 8%

NF1 4%

FLT3 3%

PTPN11 3%

KIT <1% Gelsi-Boyer et al. Br J Haematol. 2010;151:365-75 Itzykson et al, J Clin Oncol, 2013;31:2428-36 Haferlach et al. Leukemia. 2012;26:834-9

Plon S. Genet Med. 2011;13:203-4

Maxson et al., NEJM 2013;368:11-20

Epigenetic mutations

• Epigenetics: heritable (through cell generations) changes in gene expression without changes in DNA sequence

• Modification of: – DNA (methylation of CpG) DNMT3A, TET2, IDH1/2, WT1

– histones (methylation, acetylation etc) EZH2, ASXL1

Zhou et al. Nat Rev Genet. 2011;12:7-18

Epigenetic mutations in MPN and MDS/MPN

DNMT3a

CH3 CH3

CH2OH

CH2OH

ET PV

PMF CMML

1-5% 5-10% 5-10% 2-10%

DNMT3a:

Cross, Am Soc Hematol Educ Program. 2011:208-14 Jankowska et al., Blood. 2011;118:3932-41

Grossmann et al. Leukemia. 2011;25:877-9 Itzykson R et al, J Clin Oncol. 2013;31:2428-36


DNMT3a

CH3 CH3

CH2OH

CH2OH

TET2

α-KG

ET PV

PMF CMML

1-5% 5-10% 5-10% 40-60%

TET2:




DNMT3a

CH3 CH3

CH2OH

CH2OH

TET2

IDH1/2

isocitrate

α-KG



ET PV

PMF CMML

1-5% 5-10% 5-10% 40-60%

TET2:


DNMT3a

CH3 CH3

CH2OH

CH2OH

TET2

IDH1/2*

isocitrate

α-KG

2HG

ET PV

PMF CMML

1-5% 5-10% 5-10% 1-5%

IDH1/2:




DNMT3a

CH3 CH3

CH2OH

CH2OH

TET2

IDH1/2*

isocitrate

α-KG

2HG

H3K27me3

ET PV

PMF CMML

1-5% 5-10% 5-10% 1-5%

IDH1/2:


WT1



DNMT3a

CH3 CH3

CH2OH

CH2OH

TET2

IDH1/2*

isocitrate

α-KG

2HG

PRC2

H3K27me3

ET PV

PMF CMML

1-5% 5-10% 5-10% 5-15%

EZH2:


EZH2

WT1



DNMT3a

CH3 CH3

CH2OH

CH2OH

TET2

IDH1/2*

isocitrate

α-KG

2HG

EZH2 PRC2

ASXL1

H3K27me3

ET PV

PMF CMML

1-5% 5-10% 5-10% 40-50%

ASXL1:


WT1


No clear association between mutations in epigenetic regulators

EZH2

TET2

ASXL1

Ernst et al., Nature Genetics. 2010; Grossmann et al., Leukemia 2011

TET2 mutations may be acquired before or after JAK2 V617F in MPN

Delhommeau et al. N Engl J Med. 2009;360(22):2289-301. Schaub et al. Blood. 2010;115(10):2003-7.

Several somatically mutated genes in myeloid disorders are constitutionally mutated in rare

developmental disorders

EZH2

SETBP1

ASXL1

DNMT3A

BCOR

cohesins

EP300

Weaver Syndrome

Schinzel-Giedion

Bohring-Opitz Syndrome

Overgrowth syndrome

OFCD syndrome

Cornelia de Lange

Rubinstein-Taybi syndrome

MPN/MDS/T-ALL

aCML

Myeloid malignancies

Myeloid malignancies

AML, MDS

MDS/MPN, MDS

Lymphoma, T-ALL, myeloid

MLL SET

SETBP1

Deregulation of HOXA gene expression May provide a stem cell advantage that promotes clonal expansion (landscaping mutations)

Common consequences of EZH2, ASXL1 and other mutations?

EED SUZ12

EZH2

ASXL1

Blood 2012;119:6099-108

Cancer Cell 2012;22:180-93





MDS: prevalence of mutations

>10%

>5%

>3%

13 genes 30 genes

34% of total

oncogenic mutations

Papaemmanuil et al. Blood. 2013;122(22):3616-2

Prognostic significance of mutations in specific genes

• EZH2, RUNX1, TP53, ASXL1 mutations generally associated with poor prognosis in multiple studies (MPN, MDS/MPN, MDS)

• Is the level of mutation important, ie clonal vs subclonal?


TP53 Mutation Status Divides MDS Patients with Complex Karyotypes into Distinct

Prognostic Risk Groups

Bejar et al., ASH 2014

Prognosis in MDS associated with somatic genetic complexity (111 candidate genes)


Prognosis in CMML associated with somatic genetic complexity (19 genes; 312 cases)

Itzykson et al, J Clin Oncol, 2013;31:2428-36

Prognostic significance of somatic complexity in myelofibrosis: ASXL1, EZH2, SRSF2, IDH1/2

Lasho , Guglielmelli et al. Blood 2013;122:104 ASH 2013

Which genes should be screened for in diagnostic labs. Who is going to pay? Is it worth paying for?

Is up front mutation profiling the best prognostic indicator?

• Collaboration between Jackie Boultwood, Andrea Pellagatti (Oxford) and Moritz Gerstung, Elli Papaemmanuil, Peter Campbell (Sanger Institute, Cambridge)

• 124 MDS cases: NGS mutation screen + GEP microarray data (Affymetrix GeneChip Human Plus 2.0 arrays) on bone marrow CD34+ cells from 124 MDS patients

• Built statistical models to disentangle the effect of 12 mutated genes and 4 cytogenetic alterations on gene expression, diagnostic clinical variables and outcome in patients with MDS

Gerstung et al., Nat Commun. 2015 Jan 9;6:5901

Prognostic power of gene expression, mutations and clinical parameters

• Genetics, gene expression, blood and bone marrow counts all contain information for predicting survival

• Prognostic power of expression data was greater than that of genetics, cytogenetics or IPSS score

• Gene expression data provide greatest prognostic information

• Prognostic information present in genetics and cytogenetics is mostly contained in expression and blood/bone marrow count data and does not add independent prognostic information

pro

gno

stic

po

we

r

Gerstung et al., Nat Commun. 2015 Jan 9;6:5901

Molecular MRD status provides the most powerful prognostic factor in many scenarios

e.g. NPM1 mutant AML





Somatic myeloid mutations in the general population

• n=17,182; 160 genes

• Mutations rare <40 years old but seen in 10% >70 years old

• DNMT3A, ASXL1, TET2

• Increased risk of hematologic malignancy (HR = 11.1)

• n=12,380; WES

• Mutations 1% <50 yrs; 10% >65 yrs

• DNMT3A, ASXL1, TET2

• Increased risk of hematologic malignancy (HR = 12.9)

• 42% of hematologic malignancies arose in persons who had clonality

Jaiswal et al., N Engl J Med. 2014 ;371(26):2488-98 Genovese et al., N Engl J Med. 2014;371(26):2477-87

Somatically acquired uniparental disomy (copy number neutral LOH) in healthy individuals

• Study of 108 elderly men using the Illumina 1M-Duo beadchip analysis plus 78 elderly monozygotic twins

Forsberg et al., Am J Hum Genet. 2012 ;90:217-28

• Somatic abnormalities seen in 3.4% of cases >60 years old

• Included abnormalities associated with malignancy – del(5q), del(20q), 4q aUPD

• Blood counts normal

Is the aUPD at 4q in ULSAM-697 associated with an acquired TET2 mutation?

• Sanger sequencing entire TET2 coding sequence

• Found 21bp deletion that disrupts exon 4

• Mutation found at ages 71, 82, 88 and 90 but absent in fibroblasts

Why do some JAK2 V617F positive patients develop PV and others develop ET?

• Strength of JAK2 signalling – Higher JAK2 V617F allele burden (%V617F) in PMF and PV compared to ET

– Frequent homozygous JAK2 V617F clones in PMF and PV; rare in ET

– JAK2 exon 12 mutations associated with erythroid phenotype and show stronger signalling in model systems

• Constitutional genetic differences? – MPN phenotype in retroviral transplant models dependent on mouse strain

Zaleskas et al., PLoS One. 2006;1:e18 Godfrey et al., Blood. 2012 Sep 27;120(13):2704-7.

Scott et al., N Engl J Med. 2007;356:459-68.

Constitutional genetic variation at HBSL1-MYB influences whether JAK2 V617F positive cases

develop ET or PV

• n=1112 (556 ET, 556 PV) from the UK

• Top hit (excluding 9p) = rs9399137 in HBSL1-MYB polymorphic intergenic region

Tapper et al. Nat Commun 2015 in press

Summary

• Complex interplay between somatic and inherited variants in myeloid disorders

• >40 recurrent somatically mutated genes identified that can be grouped into 5 principal functional classes

• Small subset of genes and overall somatic complexity prognostically significant [but GEP and MRD analysis may be more powerful]

• Driver mutations may be present at high level in healthy elderly individuals

Acknowledgements

Andy Chase Will Tapper Amy Jones Jo Score Catherine Bryant Thomas Ernst Will Leung Elli Papaemmanuil Peter Campbell Jyoti Nangalia Carlo Gambacorti-Passerini Tony Green Mario Cazzola Eva Hellstrom-Lindberg David Bowen Jackie Boultwood Andrea Pellagatti Andreas Reiter Juliana Schwaab

Salisbury Sanger/ICGC Mannheim

GWAS Alessandro Vannucchi Paola Guglielmelli Mario Cazzola Giovanni Barosi Robert Kralovics Heinz Gisslinger Konny Döhner Frank Stegelmann Andreas Hochhaus Katerina Zoi Heike Pahl Susanne Schnittger

David Oscier Andrew Duncombe Tony Green Anna Godfrey Claire Harrison Rosemary Gale Adam Mead Anna Schuh Andrew Jack Paul Evans Jo Ewing Mike Griffiths

Uppsala Jan Dumanski Lars Forsberg

Documents

Genetic complexity in MPN, MDS/MPN and MDS · 2015. 2. 20. · 2 OH CH 2 OH TET2 IDH1/2* isocitrate α-KG 2HG EZH2 PRC2 ASXL1 H3K27me3 ET PV PMF CMML 1-5% 5-10% 5-10% 40-50% ASXL1: