Raajit Rampal M.D. Ph.D.Leukemia Service

Memorial Sloan-Kettering Cancer Center

Genetics and targeted therapy of MPNs

Scientific Questions/Advances

• Are mutations which activate JAK2 a hallmark of all MPN patients?

• What do mutations which occur in concert with JAK2/MPL mutations do?

• What have we learned about JAK2 inhibitors?

• What novel therapies are of potential benefit for MPN patients?

JAK2V617F mutations in myeloid malignancies

Prevalence of JAK2V617F :- PV: 95%- ET: 45-50%- MF: 45-50%- CMML: 5%- MDS: RARS-T (>50%)- AML: 2%

James et al. Nature 2005Levine et al. Cancer Cell 2005Baxter et al. Lancet 2005Kralovics et al. NEJM 2005Zhao et al. JBC 2005

JAK2V617F negative MPN

• JAK2V617F-negative PV- JAK2 exon 12 mutations- loss of function mutations in LNK, negative regulator of

JAK2 (Oh et al Blood 2010)

• JAK2V617F-negative ET/PMF- MPL mutations in 10%- LNK mutations in <5%

• Somatic mutations had not been identified in 30-40% of MPN patients

Identification of CALR Mutations in JAK2 wildtype MPN

Nangalia et al NEJM 2013Klampfl et al NEJM 2013

• CALR Mutations exclusive of JAK2/MPL mutations

• Seen exclusively in ET/MF

• Small set (<10%) of ET/MF patients JAK2/MPL/CALR-negative->unknown mutant disease allele

Gene expression – can we measure gene expression and learn something about pathogenesis of MPN?

• Determine if there is a common genetic signature associated with MPN or with JAK2V617F mutations

• Identify genes which segregate with clinical phenotype

• Identify candidate genes in JAK2/MPL/CALR-negative MPN

Ben Ebert/Todd Golub

Gene Expression Profiling of MPN Samples distinguishes MPN patients from controls but does not distinguish based on disease or JAK2 status

HomozygousHeterozygousJAK2 negativeControl

PVPMFET

CONTROLJAK2 shRNA

Is there a JAK2 signature in heterozygous/WT MPN patients?

JAK2 shRNA in HEL cells to generate JAK2 signature

Similar data with JAK inhibitor

PV MMM ETCONTROL

MMM MMMETHOMOZYGOUS WT HETEROZYGOUS

JAK2 shRNA signature in MPN and control samples

Seen in all MPN patients, not in controls

JAK2 is activated in all MPN patients regardless of specific mutation

JAK2 Signature Enriched in CALR-mutant MPN Patients

FDR qvalue= 0.012

NormalCALR mutant

0.70.60.50.40.30.20.10.0

Enric

hmen

t Sco

re (E

S)

JAK2 shRNA down expression signature Homozygous vs Normal UP signature

FDR qvalue= 0.321P-value= 0.035

NormalCALR mutant

0.6

0.5

0.4

0.3

0.2

0.1

0.0

Enric

hmen

t Sco

re (E

S)

• Consistent with activated JAK2-signaling in CALR-mutant patients

Gene Expression Signature of CALR Mutation

• CALR Signature is Enriched in JAK2-mutant MPN patients

• Consistent with JAK2 and CALR mutations having same role in MPN pathogenesis – all leads to JAK2 activation

Cooperating Mutations in MPN Patients

• Recent studies have identified somatic disease alleles which occur in concert with JAK2/MPL mutations- TET2 loss of function mutations in 10% of MPN patients- ASXL1 mutations in 8-10% of MPN patients- IDH1/2 mutations in 3-5% of MPN patients- EZH2 mutations in 10-15% of patients

• Same mutations are seen in MDS and AML patients->they do not explain the PV/ET/MF concondrum

• In some cases (TET2, IDH1/2) these mutations occur most commonly at progression to AML

• Limited functional data suggest these mutations affect the epigenetic state of MPN cells->affect the way DNA is packaged and which parts of it are used in MPN cells

Leukemic Transformation of MPN

• Patients with PV, ET, and PMF are at high risk for transformation to AML- associated with a dismal prognosis

• Genetic/Epigenetic events which contribute to leukemic transformation are not known

• Approximately 50% of JAK2+ MPN patients transform to a JAK2-negative MPN*

*Campbell et al. Blood 2006 Theocarides et al. Blood 2007

Mutational Studies in post-MPN AML

Denotes frequently mutated in de novo AML

No mutations found in c-KIT, EZH2, or WT-1

Splicesome mutations

Model of MPN Pathogenesis

JAK2 rs10974944Other alleles

EZH2, TET2, ASXL1

AML

TET2, IKZF, IDH1/2Other Alleles

• Mutations which activate JAK2 are the most common lesion->best therapeutic target

• Possible other mutations affect response to JAK inhibitors

JAK Inhibitors in Current Clinical Development

Agent Company Activity Status

Ruxolitinib (INCB18424) Novartis/Incyte JAK1/JAK2 FDA approved, Phase II (PV)

TG101348/SAR302503 Sanofi Aventis JAK2, FLT3 Phase III

CYT387 Cytopia JAK1/JAK2/TYK2 Phase II

Pacritinib (SB1518) SBio JAK2, FLT3 Phase III

Lestaurtinib (CEP701) Cephalon JAK2/Flt3 Phase I

BMS-911453 Bristol-Myers Squibb JAK2 Phase I

NS-018 Nippon-Shinyaku JAK2/Src Phase I

AZD1480 Astra Zeneca JAK1/JAK2 Phase I

LY2784544 Eli Lily JAK2V617F Phase I

Tofacitinib (CP-690550) Pfizer JAK3 FDA approved (RA)

• Although structurally divergent, these inhibitors are all Type I (ATP competitive) JAK Inhibitors

Courtesy of Priya Koppikar

Resistance to JAK inhibitors in the treatment of MPNs

Phase II/III clinical trials with INCB18424 and other JAK inhibitors• Improvement in splenomegaly, constitutional symptoms, reduced progression to leukemia, survival benefit.• No decrease in allele burden in the majority of MPN patients • May be due, at least in part to presence of other (disease-initiating) alleles

JAK2-driven preclinical models argue other factors contribute• JAK2V617F knockin model: Disease initiating cells are resistant to JAK2 inhibitor treatment (Mullally et al. 2010)

MPN mutant clone persists in the presence of chronic JAK2 kinase inhibition

We have not identified second-site mutations in patients treated with INCB18424

The lack of an initial response argues for inherent insensitivity to JAK inhibitors: persistence Reactivation of JAK2/JAK-STAT signaling Activation of a redundant/collateral pathway

STAT

RASPI3K

AKTSTATSTAT

P

P

P

P

P

P

P

Nucleus

JAK2

MAPK

JAK2

JAK-STAT signaling in naïve MPN cells

RASPI3K

AKTSTAT

Nucleus

JAK2

MAPK

JAK2

JAK inhibitors block homodimeric JAK2 activation and downstream signaling in naïve cells

JAK-STAT Signaling is Intact in Patients Treated with Ruxolitinib

Koppikar et al. Nature 2012

STAT

RASPI3K

AKTSTATSTAT

P

P

P

P

P

P

P

Nucleus

JAK2

MAPK

JAK1/ TYK2

JAK inhibitors cannot inhibit heterodimeric JAK2 activation and downstream signaling in persistent cells

pJAK2 Associates with JAK1 and TYK2 in MPN Patients treated with Ruxolitinib, but Not in Inhibitor Naïve Patients

Koppikar et al. Nature 2012

• Do persistent cells remain JAK2 dependent?• Can persistence be targeted with agents which degrade

JAK2 or block JAK2 transactivation?

• Loss of JAK2 inhibits growth and signaling in persistent cells

Scr shJ2 81 shJ2 1 Scr shJ2 81 shJ2 1

SET-2 SET-2 INPer

pJAK2

JAK2

pSTAT3

pSTAT5

Tubulin

STAT3

STAT5

Persistent Cells are JAK2 dependent

Can we improve our ability to target JAK2

• Can we develop better therapies which improve the therapeutic window and target the malignant cell?

- additional therapies- alternate dosing strategies for JAK2

inhibitors

• Can novel compounds be used in a synergistic manner to inhibit JAK2 dependent proliferation?• HSP90 inhibition

Sachie Marubyashi, Priya Koppikar

PU-H71(HSP90 Inhibitor

JAK2 is an HSP90 Client and is Degraded by the Purine Scaffold HSP90 Inhibitor PU-H71

Marubayashi et al. JCI, 201025

Options: Click-Drag image to reposition | View smaller image | Expand to full image width | Reset image

JAK2V617F

JAK2V617FPU-H71 nM

pJAK2

JAK2

pSTAT3

STAT3

pSTAT5

STAT5

pMAPK

MAPK

HSP90

Raf1

Actin

• Similar results with W515L cells: JAK2 is an HSP90 client

PU-H71 Demonstrates Efficacy in vivo in JAK2V617F and MPLW515L Transplant Models

Survival

Spleen weights

HSP90 Inhibition Degrades JAK2/Inhibits JAK-STAT signaling in 1° MPN Samples

AUY922 (MSKCC IRB 12-076)

• Drug: AUY922 (HSP90 inhibitor)

• Study Design: Phase II single-arm

• Endpoint: Efficacy of AUY922 assessed by IWG criteria

• Patient Population/Eligibility: -PMF -Post-PV/ET MF-Refractory PV/ET (HU, phlebotomy, anagrelide)

p=0.02p=0.03

p=0.05

p=0.057p=0.057

p=0.05

29

Combination Treatment with JAK and HSP90 Inhibitor Shows Improved Efficacy Compared to Either Therapy Alone

Combination Treatment with JAK and HSP90 Inhibitor Shows Improved Efficacy Compared to Either Therapy

Alone

pJAK2

JAK2

pSTAT3

pMAPK

Actin

STAT3

MAPK

Veh INC30 PU+INC30INC90

Vehicle INC90 INC30 with PU

Reduced Reticulin Fibrosis with Combination Therapy

PU-H71 + Ruxolitinib Clinical Trial

• Drugs: PU-H71, Ruxolitinib

• Study Design: Phase I

• Endpoint: Efficacy assessed by IWG criteria

• Patient Population/Eligibility: -PMF on Ruxolitinib with persistent disease -Post-PV/ET MF on Ruxolitinib with persistent disease

Summary

• Mutations which activate JAK-STAT signaling are seen in almost all MPN patients->but there are additional genetic lesions seen in MPN patients which contribute to stem cell survival

• Additional novel therapeutic approaches targeted at JAK2 and at other oncogenic signaling pathways might offer benefit alone or in conjunction with JAK2 inhibitors

• Genetic studies of myeloid malignancies will likely identify novel mutations with pathogenetic and therapeutic relevance

Acknowledgements

• NHLBI, NCI, HHMI, LLS, Starr Cancer Consortium, Geoffrey Beene Foundation, Gabrielle’s Angel Foundation, MPN Foundation

MSKCC• Gabriela Chiosis• Nick Socci• Marty Tallman• Omar Abdel-Wahab

MDACC• Serge Verstovsek• Miloslav Beran• Taghi Mansouri

Harvard/Broad• Gary Gilliland• Ben Ebert• Todd Golub• Ann Mullally

Cornell• Dick Silver• Ari Melnick• Gail Roboz

Mayo• Reuben Mesa

Levine Lab• Priya Koppikar• Neha Bhagwat• Outi Kilpivaara• Jay Patel• Franck Rappaport• Alan Shih• Olga Guryanova• Lindsay Saunders• Ria Kleppe• Todd Hricik• Sophie McKenney

Northwestern• John Crispino• Jon Licht