Pathogenesis and management of CMML

Raphaël Itzykson, Hôpital Saint-Louis, Paris

International Conference of the Korean Society of Hematology March 29th 2018

대한혈액학회 Korean Society of Hematology

COI disclosure

Name of author ： Raphael Itzykson

I currently have, or I have had in the past two years, an affiliation or financial

interest with business corporation(s):

(1) Consulting fees, patent royalties, licensing fees ： No

(2) Research fundings： Yes, Janssen, Novartis

(3) Others No

Clinical presentation of CMML

Monocytosis Hyperleukocytosis Tumor symptoms

Dysplasia Anemia Thrombocytopenia

Myeloproliferation

Myelodysplasia

Granulomonocytic Hyperplasia

Somatic mutations in CMML

Whole genome 475 mutations

Exome

15 mutations

Recurrent Oncogenes 2 mutations

Nat Commun. 2016;7:10767. J Clin Oncol. 2013;31(19):2428-36.

Three families of recurrent mutations in CMML

RAS

CBL

JAK2

Signaling

EZH2

TET2

Epigenetics

N

NH2

N O

ASXL1

IDH

DNMT3A

A Exon 2

Splicing

ZRSR2

U2AF1

SRSF2

SF3B1

At least one of those in 95% of patients None is specific of CMML

~90% ~75% ~60%

TET2

IDH2

IDH1

CBL

NRAS

KRAS

JAK2

FLT3

KIT

SRSF2

U2AF35

SF3B1

ZRSR2

J Clin Oncol. 2013;31(19):2428-36.

HSC

MPP

CMP

MEP GMP

Two mechanisms for the granulomonocytic hyperplasia in CMML

Platelets RBC PN Monocytes

GM-CSF Hypersensitivity?

JAK2

Grb2

Sos Ras

Ras GTP Nf1

Raf

MEK

ERK

Cbl

AP1

GM-CSF

GM-CSF hypersensitivity in MDS/MPN

Shc

Shp2

PI3K

Akt

JMML : ~100%

Mutations

control

JMML

GM-CSF

JAK2

Grb2

Sos Ras

Ras GTP Nf1

Raf

MEK

ERK

Cbl

AP1

CMML: ~60 %

JMML: ~100%

Mutations

GM-CSF

GM-CSF hypersensitivity in MDS/MPN

Shc

Shp2

PI3K

Akt

control

JMML

GM-CSF

GM-CSF hypersensitivity in MDS/MPN

restricted to GM-CSF restricted to CMML with

Signaling mutation

(RAS, CBL, JAK2)

Blood. 2013;121(25):5068-77.

HSC

MPP

CMP

Erythro Granulo Mono

A model for the pathogenesis of CMML

GM-CSF Hypersensitivity

Enhanced Self-renewal

Epigenetic hits (TET2) Splice hit (SRSF2)

Differentiation bias

MD-CMML

Signaling mutations

MP-CMML

1. Persistent PB monocytosis (≥1 x109/L and ≥ 10% of WBC)

• No impact of BM monocyte %

2. Not meeting criteria for BCR-ABL CML, PMF, PV, or ET

3. If eosinophilia: No evidence of PDGFRA, PDGFRB, or FGFR1 rearrangement or PCM1-JAK2

4. <20% myeloblasts or monoblasts in PB or BM

• Including promonocytes

5. Evidence of dysplasia in one or more lineages

• If lacking: acquired, clonal cytogenetic or genetic abnormality

6. or persistent monocytosis > 3 months, with exclusion of all other causes

• CMML-0: <2% PB blasts and <5% BM blasts

• CMML-1: 2-4% PB blasts and 5-9% BM blasts

• CMML-2: 5–19% PB blasts and 10–19% BM blasts

Arber, Blood 2016

WHO-2016 criteria for CMML

Molecular biology as diagnostic tool?

RAS

CBL

JAK2

Signaling

EZH2

TET2

Epigenetics

N

NH2

N O

ASXL1

IDH

DNMT3A

A Exon 2

Splicing

ZRSR2

U2AF1

SRSF2

SF3B1

• No single specific mutation • Preferential combo: TET2/SRSF2 • CHIP genes:

• TET2, DNMT3A, ASXL1 • One mutation • Low VAF (<20%)

Itzykson JCO 2013, Busque Nat Genet 2011, Genovese NEJM 2014, Jaiswal NEJM 2014

Age-matched controls CMML Reactive monocytosis

Accumulation of ‘classical’ monocytes (MO1) is a key feature of CMML

CD14

CD

16

Flow cytometry as diagnostic tool

Selimoglu-Buet et al. Blood 2015

* P < 0.05 ** P < 0.01 *** P < 0.001

OS

AMLFS

fav unfav

*

*** ***

*

*

* **

*

*

0.1 1.0 10.0

U2AF1

SF3B1

IDH2

ZRSR2

KRAS

JAK2

CBL

NRAS

RUNX1

ASXL1

SRSF2

TET2

HR (95% CI)

Itzykson, JCO 2013

Prognostic Impact of Gene Mutations

120 3624 48 60

0%

20

%4

0%

10

0%

60

%8

0%

Months

Cu

mu

lativ

e P

rob

ab

ility

of S

urv

iva

l

0 6 12 18 24 30 36 42 48 54 60

0.0

0.2

0.4

0.6

0.8

1.0

187 167 134 106 84 65 50 33 19 11 5 jco$asxl1=0

125 95 68 44 29 17 10 8 6 4 1 jco$asxl1=1

187

125

134

68

84

29

50

10

19

6

5

1

N à risque Mois

Su

rvie

Glo

ba

le (

%)

ASXL1 sauvage

ASXL1 muté

Overall Survival

ASXL1 wildtype

ASXL1 mutated

Ove

rall

Surv

ival

(%

)

Months

48 months

18 months

‘Next-generation’ Prognostic scores in CMML

Score CPSS GFM Mayo CPPS-mol

Clinical Features

Blasts WBC

RBC-TD Cytogenetics

Age WBC Hb

Platelets

Monocytes IMC Hb

Platelets

Blasts WBC

RBC-TD Cytogenetics

Molecular Features

No ASXL1 No ASXL1 NRAS

RUNX1 SETBP1

Risk groups 4 3 3 4

mOS (mths) 5 - 72 14-60 10-32 17 - 70

Validation Yes Yes Yes Yes

Reference Such Blood 2012

Itzykson JCO 2013

Patnaik Leukemia 2013

Elena Blood 2016

Management of CMML

Eltrombopag in CMML with thrombocytopenia

• Prospective multicentric GFM Phase II trial (n=19/30)

• Lower-risk CMML-0 with platelets < 50 000/mm3

• IWG 2006 Response rate: 63%

• Median response duration: 8 mois

• RUNX1 mutations do not impair response achievement

Itzykson et al ASH 2017

Wattel Blood 1996

Hydroxyurea in proliferative CMML

• HY versus VP16 in ‘advanced’ MP-CMML (N=105)

• Overall Response Rate: 60% (CR: 20%)

Ruxolitinib in CMML

Padron et al ASH 2017

• US Phase 1/2 trial

• Few hematological responses captured by IWG 2006

• Spleen and general symptoms improvements

• Prolonged survival compared to historical control?

JAK2

Grb2

Sos Ras

Ras GTP Nf1

Raf

MEK

ERK

Cbl

AP1

CMML

JMML

Mutations

GM-CSF

Targeted therapy in proliferative CMML

Shc

Shp2

PI3K

Akt

JAK2

Grb2

Sos Ras

Ras GTP Nf1

Raf

MEK

Cbl

GM-CSF

Targeted therapy in proliferative CMML

Shc

Shp2

PI3K

Akt

Targets Lyubynska Science Transl Med 2012, Padron Blood 2013, Goodwin, Blood 2014, Kong, J Clin Invest 2014

ERK

Allogeneic SCT in CMML

NRM 35% Relapse 35%

Park et al Eur J Hematol 2013

Time from randomisation (months)

Pro

po

rtio

n o

f p

ati

en

ts s

urv

ivin

g

CCR (n=179)

Azacitidine (n=179)

0 5 10 15 20 25 30 35 40

0

0.2

0.4

0.6

0.8

1.0

Lancet Oncol 2009;10:223–32

CMML: n=11

CMML: n=5

AZA is licensed in CMML-2 with WBC < 12 G/L

• « meta-analysis » of 17 studies

• Overall Response Rate: 50%

• Complete Response Rate: 25%

• Regression of myeloproliferative features (poorly captured)

• MP-CMML retains adverse prognosis

• No difference between azacitidine and decitabine • PSM models Alfonso, Am J Hematol 2017 ; Duchmann, submitted

Hypomethylating agents in CMML

Retrospective, 174 patients, EU+US

Duchmann et al, ASH 2017

Genetic biomarkers in CMML treated with HMAs

Specific methylation signatures predict decitabine response in chronic myelomonocytic leukemia

Meldi J Clin Invest 2015

Epigenetic biomarkers in CMML treated with HMAs

DACOTA Trial

A Randomized Phase III study of Decitabine

with or without Hydroxyurea versus Hydroxyurea

in patients with advanced proliferative Chronic Myelomonocytic Leukemia

EMSCO FISM

Decitabine 20mg/m2/d x5d q.28d

± HY during the first 3 cycles

CMML

WBC > 13 G/L

≥ 2 criteria:

Marrow blasts ≥5 %

Abnormal K (except –Y)

ANC > 16 G/l

Hb < 10 g/dL

Platelets < 100 G/L

Splenomegaly > 5 cm

Or Extramedullary localization

HY

Primary Endpoint: Event-free Survival

- Death

- Transformation to AML

- Progression of myeloproliferation

N=168

N=168

Management of CMML

MDS/MPN in the GFM

• Academic and Industry Trials

• Patients registry

• Academic EU trials through EMSCO

• Academic Int’l trials through MDS/MPN IWG

• Cell bank and research focusing on CMML • INSERM U1170, Pr. E. Solary