The discovery and pre-clinical development of the first clinical stage EZH2-inhibitor, EPZ-6438 (E7438)

Kevin W. Kuntz1†, Roy M. Pollock1,2, Sarah K. Knutson1, Natalie Warholic1, Tim Wigle1, Chris Sneeringer1,3, Victoria Richon1,4, Richard Chesworth1, Margaret Porter Scott1,3, Robert A. Copeland1, Heike Keilhack1 1Epizyme Inc., 400 Technology Square, 4th floor, Cambridge MA 02139, USA 2Currently at Warp Drive Bio, 400 Technology Sq, Cambridge MA 02139, USA 3Currently at Genentech, 1 DNA Way, South San Francisco CA 94080, USA 4Currently at Sanofi, 270 Albany Street, Cambridge MA 02139, USA

† corresponding author

Mutations within the catalytic domain of the histone methyltransferase EZH2 have been identified in subsets of patients with non-Hodgkin lymphoma (NHL). These genetic alterations are hypothesized to confer a cellular survival dependency on EZH2 enzymatic activity in these cancers. Here, we disclose the discovery of EPZ-6438 (E7438), as a potent, selective and orally bioavailable small molecule inhibitor of EZH2 in preclinical models of NHL. Previously we have disclosed the properties of EPZ005687, a tool compound useful for exploring the in vitro biology of EZH2 inhibition. Multi-parametric optimization of the potency, pharmokinetics, oral bioavailability and tolerability of this series led to the discovery of the clinical compound, EPZ-6438. Modulation of the log P was required to reach the optimal balance between clearance and bioavailability while maintaining the requisite potency. EPZ-6438 selectively inhibits intracellular lysine 27 of histone H3 (H3K27) methylation in a concentration- and time-dependent manner in both EZH2 wild-type and mutant lymphoma cells. Inhibition of H3K27 trimethylation (H3K27Me3) leads to selective cell killing of human lymphoma cell lines bearing EZH2 catalytic domain point mutations. Treatment of EZH2-mutant NHL xenograft-bearing mice with EPZ-6438 caused a dose-dependent tumor growth inhibition, including complete and sustained tumor regressions with correlative diminution of H3K27Me3 levels in tumors and selected normal tissues. EPZ-6438 recently entered clinical testing as E7438 in a dose escalation phase 1 trial in relapsed or refractory malignancies.

Abstract Results

Introduction

Conclusions

www.epizyme.com

#277 P057

References 1. Knutson et al. A selective inhibitor of EZH2 blocks H3K27 methylation and kills mutant lymphoma cells, Nature Chemical Biology, 2012, 8:890-896. 2. Keilhack et al. Preclinical characterization of E7438, a potent, selective inhibitor of protein methyltransferase EZH2 with robust antitumor activity against EZH2 mutated non-Hodgkin

lymphoma xenografts in mice, Blood (ASH Annual Meeting Abstracts) Nov 2012: 120: 3712. 3. Sneeringer et al. Coordinated activities of wild-type plus mutant EZH2 drive tumor-associated hypertrimethylation of lysine 27 on histone H3 (H3K27) in human B-cell lymphomas,

PNAS, 2010, 107(49): 20980-20985. 4. Knutson et al. Durable tumor regression in genetically altered malignant rhabdoid tumors by inhibition of methyltransferase EZH2, PNAS, 2013, 110(19): 7922-7927.

• EPZ-6438 is a potent and selective small molecule inhibitor of EZH2 and EZH2 SET domain mutants.

• EPZ-6438 inhibits cellular H3K27 methylation leading to killing of lymphoma cell lines expressing EZH2 SET domain mutants.

• Antitumor activity has been observed in several EZH2 mutant lymphoma xenograft models ranging from tumor growth inhibition to durable regressions (e.g. KARPAS422) at well tolerated doses and schedules.

• EPZ-6438 (E7438) has transitioned into clinical development and results from the Phase I study are being presented in a separate oral presentation by V. Ribrag entitled “Phase 1 first-in-human study of the enhancer of zeste-homolog 2 (EZH2) histone methyl transferase inhibitor E7438 as a single agent in patients with advanced solid tumors or B cell lymphoma” in Plenary Session 5

Disclosures: Kuntz, Knutson, Warholic, Wigle, Chesworth, Keilhack: Epizyme, Inc.: Employment, Equity Ownership, Patents , Stock options Copeland: Epizyme Inc. : Employment, Equity Ownership, Patents, Stock options ; Mersana: Membership on an entity’s Board of Directors or advisory committees. Pollock: Epizyme, Inc.: Equity Ownership, Patents , Stock options ;Warp Drive Bio.: Employment Sneeringer, Porter Scott: Epizyme, Inc.: Equity Ownership, Patents , Stock options ; Genetech.: Employment Richon: Epizyme, Inc.: Equity Ownership, Patents , Stock options ; Sanofi.: Employment

Methylation

Demethylation

Change of function mutation• Non-Hodgkin lymphoma

PRC2 COMPLEX

K27(me)3K27

Loss of function due to INI1 deficiency• Synovial sarcoma, MRT, others

SWI/SNF COMPLEX

INI1

At least 3 distinct genetically defined cancers• Non-Hodgkin lymphoma, germinal center (EZH2 point mutations)

• Synovial sarcoma (SSX-SS18 fusion)

• MRT (INI1-deletion)

EZH2 Inhibition for Genetically Defined CancersEZH2 Inhibition for Genetically Defined Cancers

• EZH2 is the catalytic subunit of the multiprotein PRC2 (polycomb repressive complex 2) complex

• PRC2 catalyzes mono-, di- and tri-methylation of H3K27 • H3K27 is the only significant substrate for PRC2 • H3K27me3 is a transcriptionally repressive histone mark • Hyper-trimethylation of H3K27 is tumorigenic in a broad spectrum of human

cancers, including GC NHL

Hit Expansion

EPZ004851EZH2 IC50 2.5 mM

EPZ004759EZH2 IC50 0.5 mMSAM competitiveNucleosome non-competitiveSolubility <10 mM at pH 7

EZH2 HTS Yielded One Tractable Series

Design Tenets1.Improve solubility2. Improve biochemical potency3. Maintain good property space

EPZ005030EZH2 IC50 0.4 mM10 mM in cell Me assayMouse PKF = 41 %Cl = 34 ml/min/kg

Oral Bioavailability achieved: 1st major hurdle overcome

An HTS was performed under balanced conditions (Km for SAM and Km for Nucleosome) to identify inhibitors of EZH2, of which is EPZ004851 is an example. Additional analogs were purchased based on structural similarity to the original hit and EPZ004759 demonstrated improved potency. This compound was characterized for its mode of inhibition and was found to be SAM competitive. Aqueous solubility was identified as a key issue with EPZ004759, and therefore groups known to increase solubility were used to functionalize the template leading to EPZ005030 which showed improved solubility. EPZ005030 was assayed in mouse PK shown to have moderate clearance and bioavailability.

Optimization of the 5,6 Ring Template

Design Tenets

Improve potencyMaintain good PK

EPZ005687 was identified as a tool compound to explore EZH2 biology

EPZ005687EZH2 IC50 50 nM1.1 mM in cell methylation assayMouse PKOral F ~40%Cl 11 ml/min/kg, Vdss 0.5 L/kgT1/2 1.5 hppb 99.2 (m) 99.7 (h)

R1 = lipophilic groups preferredR2 = Me, Et & Pr preferred

In order to increase potency, substitutions in a variety of vectors were explored. Equal-potent compounds could be made with different small groups in the R2 position. Increasing the size and lipophilicity of groups in the R1 position led to significant increases in potency. Combining a cyclopentyl group in the R1 position with a change from a pyrrazolopyridine to an indazole led to EPZ005687 which showed increased cell potency while maintaining the mouse PK profile.

Development of a new sub-series

New sub-series expanded scope of SAR allowing optimization of potency and PK

EPZ00609333 nM

8

EPZ005991EZH2 IC50 200 nM

EPZ00622215 nMCell Me IC50 166 nMMouse PKF = 2%Cl = 49 ml/min/kgppb = 98%

EPZ-643814 nM94 nM

55%13 ml/min/kg93%

Disconnection of the five-membered ring led to an increase in potency as illustrated by the potency enhancement going from EPZ005991 to EPZ006093. Importantly a second substitution on the aniline allowed an additional vector where polarity could be incorporated. When the aniline was alkylated with a tetrahydropyran and an ethyl group it furnished potent compound, EPZ006222, however the PK properties were not optimal. Replacing the chloro with a benzyl-morpholine group, as in EPZ-6438, improved the solubility, decreased the clearance and improved the oral bioavailability without sacrificing potency. EPZ-6438 was selected for clinical development.

EPZ-6438 is a Specific and SAM-Competitive Inhibitor of EZH2

EPZ-6438 is potent against both wild-type and mutant EZH2

EPZ-6438 is SAM competitive

EPZ-6438 is similarly active against wild-type and SET domain mutants of EZH2. It is about 50-fold selective against EZH1. EPZ-6438 has shown no activity across a panel of protein methyltransferases up to the top concentration tested (10 or 50 µM), meaning it is greater than 1000-fold selective for EZH2. EPZ-6438 was shown to be a reversible inhibitor of EZH2 and is SAM competitive and nucleosome non-competitive.

EPZ-6438 Specifically Inhibits Cellular H3K27 Methylation in a Time and Dose-Dependent Manner

WSU-DLCL2 cells in vitroMethylation by ELISA

WSU-DLCL2 cells in vitroTime Course at 1 µM

WSU-DLCL2 cells in vitro4-Day Treatment

WSU-DLCL2 cells treated with EPZ-6438 for four days showed a concentration dependent decrease in H3K27 trimethylation. WSU-DLCL2 cells treated with 1 µM EPZ-6438 showed a time-dependent decrease of H3K27Me3 with greater than 50% reduction after 24 h and maximal decrease (within the limits of detection) at four days. Treatment of WSU-DLCL2 cells for four days with 2.7 µM EPZ-6438 showed a reduction of H3K27 mono-, di- and tri-methylation with no significant effects on other methyl marks.

EPZ-6438 Selectively Kills EZH2 Mutant Cells Despite Similar Target Inhibition in Both Mutant and WT Cells

EZH2 Y646F Mutant EZH2 Y646 WT

H3K27me3

H3

Methylation IC50 = 0.008 µMMethylation IC50 = 0.0091 µM

µM E7438

µME7438

Day Day

Treatment of WSU-DLCL2 cells (EZH2 Y646F mutant) with a range of concentrations of EPZ-6438 showed time and concentration dependent defects in proliferation. While at four days the number of cells were similar (within 10-fold), by seven days dramatic differences in cell numbers appeared. Treatment of OCI-LY19 cells (EZH2 wild-type) with a range of concentrations of EPZ-6438 showed no major proliferation defects even after 11 days of treatment. Both cell lines showed equal reduction of H3K27Me3 suggesting that the differences in phenotypic responses are due to differences in the cells’ dependency on EZH2 and not due to differences in drug exposure.

KARPAS422 EZH2 Y646N Mutant Xenografts are Sensitive to Orally Dosed EPZ-6438 in a Dose

Dependent Manner

All doses were BID in efficacy study, no significant body weight loss during study

Mice were kept alive and remain tumor free 63 days (top two dose groups) after cessation of dosing

28 day efficacy study

12

7 day PK/PD studyTarget inhibition in tumor (ELISA)

0

200

400

600

800

1000

1200

1400

1600

0 5 10 15 20 25 30

Tu

mo

r V

olu

me

(m

m3)

Day

Vehicle

80.5 mg/kg

161 mg/kg

322 mg/kg

*

**

Veh

icle

75 m

g/k

g

150 m

g/k

g

301 m

g/k

g

602 m

g/k

g

0

50

100

Trim

eth

yla

tio

n L

evel

of

H3

K2

7 (

%)

Veh

icle

150 m

g/k

g

301 m

g/k

g

602 m

g/k

g

1203 m

g/k

g

0

50

100

*

*

* *

BID QD

*

*

*

*

Mice implanted with KARPAS422 cells were randomized once tumors reached a median size of 150 mm3. The mice were treated orally with EPZ-6438 twice daily using the doses indicated. Tumor volumes were measured by calipers every 3-4 days. After 28 days, treatment was halted and the mice observed for an additional 63 days. The top two doses led to complete elimination of the tumor with no regrowth seen. In a separate study, the changes in the H3K27Me3 levels in tumors after seven days of BID PO dosing showed a dose dependent decrease. At the ~150 mg/kg dose level, complete elimination of the tumor is observed at 28 days while only ~60% decrease in H3K27Me3 is observed at 7 days.

*statistically significant difference from vehicle