ABSTRACT#

P39Use of ex vivo Histoculture to identify potential predictive biomarkers for the ICOS agonist antibody, JTX-2011Heather A. Hirsch, Jason Reeves, Tong Zi, Amit Deshpande, Guang Yang, Alexander Needham, Jenny Shu, Christopher Harvey, Sriram Sathyanaryanan, Jennifer Michaelson, Emma Lees, Elizabeth Trehu, Deborah A. LawJounce Therapeutics Inc., Cambridge, MA

Summary

Presented at the Society for Immunotherapy of Cancer (SITC) Annual Meeting ● November 8-12, 2017 ● National Harbor, MD

ICOS (Inducible T cell CO-Stimulator) is a co-stimulatory molecule expressed primarily on Tlymphocytes. Clinical correlations and preclinical data suggest that ICOS plays an importantrole in the immune response to cancer. Therefore, we generated JTX-2011, an ICOS agonistantibody currently in clinical development in advanced solid tumors in the ICONIC trial. Inpreclinical studies, single agent efficacy correlates with the percentage of ICOS-expressing Tcells within the tumor. Thus, ICOS expression is being used as a biomarker to enrich forpatients in Phase 2 of the ICONIC trial. Building on our biomarker-driven strategy, we haveexplored additional potential predictive biomarkers using ex vivo tumor histoculture whichallows for in functional analysis of therapies using patient intact tumor tissue. Herein, wereport on the results of such analysis, including assessment of the induction of an IFNJ genesignature.

Tumor processing: Fresh human tumor samples were obtained post-surgery. A section ofeach tumor was cut and fixed for IHC. 300 PM slices of remaining tumor were placed in a 6-well plate. Treatments were added into the medium and plates were incubated at 37°C.Tumor slices were stored in RNAlater after incubation.

RNA extraction and QC: Tumor slices were lysed using Qiagen’s TissueLyser processor andFFPE samples were deparaffinized. RNA was extracted from FFPE and fresh tumor samples,quantified using Quibit, and QC’d using AATI’s Fragment Analyzer.

Gene expression: Gene expression for interferon-J signature and other genes of interestwere performed using Taqman qPCR probes or NanoString nCounter using the HumanImmunology V2 panel.

IHC: ICOS (Spring SP98) and PD-L1 (CST E1L3N) levels were assessed by IHC. Sampleswere considered to be PD-L1 high if they were 5% positive for PD-L1. Samples areconsidered ICOS high with an ICOS IHC score of 2 or 3.

Background

Methods

“Primed”T effector cell

ICOS↑

APC

TCR

1st signalICOS

JTX-2011

↑ IFN-γ↑↑ IFN-γ↑

Activation of T eff

T regulatory cell

NK cell

XReduction of T regs

“Non-Primed”T effector cell

TCRICOS

Fresh tumor from surgery

Post-Dose Expression AnalysisNanoString Analysis

RT-qPCR Analysis

• IFNJ signature• Unique immune molecular

signatures

Drug treatment

cultivate6-72hr

Baseline Analysis

Baseline tissue: FFPE embedded

ICOS

H&E, ICOS, PD1, CD8

NanoString

• Measure response to therapeutic• Assess PD biomarkers• Insights into MOA

Connect baseline to response to assess molecular determinates of response

Expanded assessment of histoculture platform and characterization of baseline predictive

markers

Establishment of histoculture system and association of IFNJ signature induction with

treatment

RNA signatures as potential predictive biomarkers

ICOS expression

Indu

ced

Inte

rfero

n S

igna

ture

Determination of HPV status

Integrative analysis of RNA expression, RNA

signatures and IHC Molecular subtyping

Concordance of RNA and IHC

Indu

ced

Inte

rfero

n S

igna

ture

Indu

ced

Inte

rfero

n S

igna

ture

PD-L1 highPD-L1 low

24 hour nivolumab treated NSCLC tumors

24 hourJTX-2011 treated NSCLC tumors

(A) Post-dose profiling of NSCLC and HNSCC to establish IFNJ induced signature.Threshold across treatments was set based on PHA treated positive control samples.(B) Ex vivo treatment of NSCLC tumors with nivolumab. (C) Ex vivo treatment ofNSCLC with JTX-2011. Samples compared to response of PHA at 24hrs (D) In sometumors, treatment with both JTX-2011 and nivolumab results in increased IFNJsignature induction that is greater than single agent alone. The bar chart abovecompares single agent treatment to combination treatment.

ICOS highICOS low

PD-L1 highPD-L1 low

Indu

ced

Inte

rfero

n S

igna

ture

Indu

ced

Inte

rfero

n S

igna

ture

ICOS 3ICOS 2ICOS 1ICOS 0

IFN-γ Signature Score

Indu

ced

Inte

rfero

n S

igna

ture

Indu

ced

Inte

rfero

n S

igna

ture

Histogram of inducers vs non-inducer (as measured by IFNJ signature inductions) are defined as shown inthe waterfall plots above using the cutoff for the strong inducers. Tumors are placed into 5 equal groupsbased on range of signature expression at baseline. In tumors where JTX-2011 induced IFNJ signature,baseline signature scores are higher as illustrated by right shift of the histogram suggesting ICOSexpression, ICOS signature and 10-gene IFNJ signatures are potential predictive biomarkers.

(A) Schematic of the process by which histoculture is performed. (B) Humanhistoculture is used as a functional assay to test both indication selection biomarkers aswell discovery of potential PD biomarkers

A

B

A

B

C

A B

C

• We have confirmed that treatment of NSCLC and HNSCC histoculture samples with nivolumab leads to induction of an IFNJ gene signature in a subset of the treated samples.

• We have established that JTX-2011 induces IFNJ signature both as a monotherapy and in combination with nivolumab.

• Additionally, induction correlated with baseline ICOS expression for JTX-2011-treated samples, and with baseline PD-L1 expression for nivolumab-treated samples.

• Baseline expression of ICOS mRNA, ICOS RNA gene signature, and 10-gene IFNJ signature correlated with IFNJ induction in JTX-2011-treated samples providing novel predictive biomarkers to assess in the ICONIC clinical trial.

• In conclusion, ex vivo histoculture is a robust tool that can be used to assess candidate predictive biomarkers, identify novel potential biomarkers, and interrogate post-dose responses to novel therapeutics.

Indu

ced

Inte

rfero

n S

igna

ture

HPV+ IMSHPV- IMSBasalClassical

Human tumor slices treated

with mAbs

Established system using nivolumab

Applying to additional IO Targets

Assessed JTX-2011 in histoculture system

• Induction of IFNJ-like signature was observedconsistent with data from the clinic (Ayers et al, AACR 2015)

• Identified induction of IFNJ-like signature • Identified potential additional genes for

assessment as RNA signature biomarkers

Indu

ced

Inte

rfero

n Si

gnat

ure

JTX-2011NivolumabCombination

1.2

1.0

0.8

0.6

0.4

0.2

0

-0.2

D

A B

C D

Jounce Therapeutics © 2017

Correlating potential predictive biomarkers with induction of IFNJ signature in HNSCC tumors

(A) Ex vivo histoculture analysis of anti-PD1 treatment of HNSCC tumors. Induction of IFNJ signature isused as a proxy for tumor response to anti-PD1 therapy. The black line shows strong IFNJ induction anddashed red lines shows moderate IFNJ induction (B) Ex vivo histoculture analysis of JTX-2011 treatmentof HNSCC tumors suggests ICOS high tumors are more likely to respond. (C) Ex vivo histocultureanalysis of JTX-2011 treatment of HNSCC tumors suggests greater response in IMS molecular subtype.

Nivolumab treatment

JTX-2011 treatment

JTX-2011 treatment: molecular subtype comparison

QC metrics and molecular characterization of baseline tumors:(A) HPV status was determined by NanoString analysis of HPV16 E6 mRNA inbaseline tumors samples. (B) Box plots showing concordance of IHC and relevant RNAexpression at baseline as well as the expression of these genes in HPV+ tumors. (C)Integrative analysis of IHC and mRNA data using complete linkage clustering withcorrelation distance. (D) Molecular subtypes (Keck et al, 2015) were determined viatwo dimensional clustering using representative genes from each subtype as well asHPV E6 and E7 mRNAs.

ICOS gene expression ICOS signature IFNJ signature

Num

ber o

f Tum

ors

Num

ber o

f Tum

ors

Non-Inducer Non-Inducer Non-Inducer

Inducer Inducer Inducer

Bins of increasing baseline gene expression

Bins of increasing baseline gene expression

Right shift of histogram peak suggests cutoff for favorable NPV

Right shift of histogram peak suggests cutoff for favorable NPV

Right shift of histogram peak suggests cutoff for favorable NPV