Beyond HER2 CAR T cells: consider how far you have fallen ……

Nabil Ahmed, MD Associate Professor

Baylor College of Medicine Houston, Texas

Disclosure Research agreement with Cellmedica … NKT cells

Consult for Adaptimmune … tgTCR T cells

ACEA Bioscience … pro bono

Concepts 1. Empiric preclinical to clinical

development of HER2CAR T-cells 2. Rational development of novel CARs

1. Target Finding 2. Exodomain Design 3. Endodomain Fitting

3. Realistic in preclinical testing 4. Pragmatic pipeline of CAR

development

Human Epidermal growth factor Receptor 2

Ozaki et al. Neurosc Res ‘98; Zhang el al. Clin Can Res ‘07; Koka et al. AJCO ‘03

• Multiple malignancies • Amplified: breast and ovarian • Non-amplified: brain, melanoma & sarcomas

• Aggressive tumor cell behavior • ↑ in stem cells of GBM & sarcoma • Associated with ↓ outcome • Protein ↓↓↓ postnatally

Wels et al. Biotech ’92; Ahmed et al. Can Res 2007 and Mol Ther 2009

FRP5 W

infried Wels

The HER2 CAR

Stev

e Got

tsch

alk

ζ

V L V

H

CD28

Herceptin

FRP5

Autologous PDC’s

CD133-PE

SSC

T cell : Tumor

0

20

40

60

40:1 20:1 10:1 5:1 0

10

20

30

40:1 20:1 10:1 5:1

% C

r rel

ease

0

20

40

60

40:1 20:1 10:1 5:1

CD133 (+) x HER2 CAR T cells CD133 (–) x HER2 CAR T cells CD133 (+) x NT T cells CD133 (–) x NT T cells

Patient 1 Patient 5 Patient 2

HER2 CAR

ζ

V L V H

CD28

Wels et al. Biotech ’92; Ahmed and Salsman et al. Clin Can Res 2009

BR

AIN

O

STEO

Day 6

LM7

LUN

G M

ET

Kleinermann; Ahmed (unpublished)

Autologous PDX’s

HER2 CAR T cells day 6 day 9 day 15

NT T cells

HER

2.C

D28

.ζ T

cel

ls

Can Res 2007; Clin Can Res 2009; Mol Ther 2009

MB

/GB

M/E

PN

OST

EO

SAR

C M

ET

Vita Brawley

Fatal Adverse Event • Patient: 39 yr; F; colon cancer metastatic to the lungs

and liver

• Lymphodepletion: CTX and FLUD

• Autologous CAR T cells: – CAR based on Trastuzumab (Herceptin®)

– CD28+4-1BB+CD3.ζ – DOSE: 1x1010

• Adverse Event: severe dyspnea; pulmonary edema; respiratory failure; died day+5

• Cytokine storm

RAC webcast, 12.2009; Morgan et al. Mol Ther 2010

A different risk profile

• DOSE: 1,000 fold less 1x108

• Trial design: mCRM escalation

• Exodomain: FRP5

• CAR: CD28+CD3.ζ (no 41BB)

• No lympho-depletion Herceptin

FRP5

Specific Cytokine Release

0

1000

2000

3000

4000

T cells

HER

2-CA

R

NT

pg/m

L

HER-ve Tumor

Normal tissue

NT T cells

HER2+ Tumor

Normal tissue

GM-CSF

IFN−γ

IL-2

TNF-α

IL-10

IL-4

IL-5

HER2-CAR T cells

HER2 CAR Trials: evolution

HER2 specific

+ durable

+ tumor resistant + stoppable …

T cellsT cells

CTLCTL

CTLCTL

HEROS

HERT-GBM

HER-TGFß

CTLCTL

2008 2009 2010 2011 2012 2013 2014 2015 2016

HEROS: complete • 43/55 (~80%) tumors HER2+ • Generated 28 cell lines; infused 19 • No SAE • Clinically

2 PR; 4 SD (142->600+ days); 13 PD

T cells

HER2 CAR

PREINFUSION POSTINFUSION

DL6

Ahmed et al. JCO 2015

Months from diagnosis

Ove

rall

surv

ival

pro

babi

lity

0 10 20 30 40 50 60

Age <18 >=18

0

0.2

0.4

0.6

0.8

1

OS=24.8 mon

Progressive GBM No SAE & No CRS Attrition 40% 7 SD 2.5 to >30 mon 1 PR 8 PD

HERT-GBM 17 patients treated

HERT-GBM in review

T cells home but don’t expand

Ahmed et al. JCO 2015

T cells home but don’t expand

HEROS 2.0 is currently recruiting

HEROS 2.0

Flu only (3 pts) Flu + CTX (2 pts)

Lymphodepletion CART cell Expansion

0 2 0 4 0 6 00

5 0 0

1 0 0 0

1 5 0 0

2 0 0 0

P o s t In fu s io n

AL

C

F lu + 1 x 1 0 8 /m 2

0 1 0 2 0 3 0 4 0 5 01

1 0

1 0 0

1 0 0 0

1 0 0 0 0

1 0 0 0 0 0

P o s t In fu s io n

co

pie

s p

er

mL

F lu + 1 x 1 0 8 /m 2

0 10 20 30 40 501

10

100

1000

10000

100000

Post Infusionco

pies

per

mL

Flu/Cy + 1x108/m2

Days Days Days

Toxicity: • 3 patients: CRS grade 1 (fever + CRP 8.9) • 1 patient: CRS grade 2 (fever + CRP 7.2)

– hypotension (low dose dopamine for ~ 36 hours) – transient oxygen requirement

Responses (5 patients): • 2 patient: PD • 2 patients: SD 8-12 months • 1 patient: CR

HEROS 2.0 is currently recruiting

HEROS 2.0

Why do tumors relapse? day 6 day 9 day 15

NT T cells

HER

2.C

D28

.ζ T

cel

ls

Can Res 2007; Clin Can Res 2009; Mol Ther 2009

MB

/GB

M

OST

EO

LUN

G M

ET

Survival Time0 10 20 30 40 50 60

Sur

viva

l Pro

babi

lity

0.0

0.2

0.4

0.6

0.8

1.0Autologous

HER2.CD28.ζ T cells (n=8)

p<0.001

Tumor only (n=8)

NT T cells (n=8)

Survival Time0 10 20 30 40 50 60

Sur

viva

l Pro

babi

lity

0.0

0.2

0.4

0.6

0.8

1.0Autologous

HER2.CD28.ζ T cells (n=8)

p<0.001

Tumor only (n=8)

NT T cells (n=8)

1. Antigenic Heterogeneity 2. Limited T cell Persistence 3. Inadequate T cell Activation 4. Inhibition @ Tumor Site

Hegde et al. Mol Ther 2013; Bielamowicz et al. Frontiers in Oncology 2014

The Great Escape 3 7 14

HER

2

IL13Rα2

Day 1

HER2 CAR T cell

HER2 CAR

IL13Rα2 CAR

IL13Rα2 HER2

Meena Hegde

~ 98%

~ 93% ~ 70%

IL13Rα2

Hegde et al. Mol Ther 2013

20 GBM (surgical excision) Flowcytometry on ≥100,000 cells IFC on 5+ cuts; 6+ hpf; 200+ cells/hpf

GBM: intratumoral heterogeneity

15 SAMPLES; MULTIPLE PIECES POOLED; 100,000-200,000 FLOW EVENTS

Bielamowicz

GBM: intertumoral heterogeneity

Hypothesis

Hegde et al. Mol Ther 2013

Possible advantages: Circumvent tumor escape

- Heterogenous expression of target - Down regulation of target - Antigen loss variants

Improve T cell activation

+ +

Tumor Profile Chris Man

ROR1

MUC1

STEAP1

SFRP1

NCAM1

L1CAM

IL11RA

FOLH1

CLDN1

CD248

BCHE

ALK

ADAM12

ROR2

PDPN

FCER2

SFRP1

ROR1

NCAM1

KDR

IL13RA2

CLDN1

BCHE

ALK

ADAM12

ES MB

SFRP1

ROR2

PDPN

NCAM1

MUC16

L1CAM

KDR

IL13RA2

EPHA2

CSPG4

CD248

CA9

BCHE

ANTXR1

ADAM12

OS

250 OS SETS

MINIMUM UNIVERSAL

PROFILE

+

Tending to 100%

Bielamowicz, Joseph and Ahmed

Kevin B

Analysis P Value H vs. H or I <.0001 H vs. H or E 0.0001 I vs. H or I 0.0001 I vs. I or E 0.0001 E vs. H or E 0.0001 E vs. I or E <.0001 H vs. H or I or E <.0001 I vs. H or I or E <.0001 E vs. H or I or E <.0001 H or I vs. H or I or E 0.0001 H or E vs. H or I or E <.0001 I or E vs. H or I or E 0.0001

0 3 0 6 00

2 5

5 0

7 5

1 0 0

6 0 7 0 8 0 9 0 1 0 0

T u m o r c e lls in p a tie n ts (% o f to ta l)

Pa

tien

ts (

%)

H

I

E

H o r I

H o r E

I o r E

H o r I o r E

Routine

2000

4000

6000

00

10000AccI

p

5' UTR

U.CARInsert 4.6 kb

HER2

IL13Rα2

EphA2

UCAR T cells

Glioblastoma Ependymoma

Medulloblastoma

Autologous GBM PDXs 1:20

Kevin B

Bielamowicz, Joseph and Ahmed

CAR T cell

Navai and Ahmed. UK Biochemical Society Proceedings. 2016

Smart CARs The Boolean Logic in Action

Multi-CAR T cell

OR

Gated-CAR T cell (iCAR)

NOT

Conditional CAR T cell

AND

On/Off CAR T cell

(GoCART) PRN

~

Concept CARs

scFv1

Target 2

Target 1

CAR

in silico Design: the virtual world

T cell

Baker, National Center for Macromolecular Imaging

Zaka

ria G

rada

Meena Hegde

scFv2

scFv1

Target 2

Target 1

TanCAR

in silico Design: the virtual world

T cell

Baker, National Center for Macromolecular Imaging

Zaka

ria G

rada

Meena Hegde

scFv2

scFv1

Target 2

TanCAR

in silico Design: the virtual world

T cell

Baker, National Center for Macromolecular Imaging

Matthew Baker

Target 1

Imaging Toolbox: exodomain

CAR Exodomain - Ligand Aggregates at IS - CAR Aggregates at IS - Synapse Properties

- F-Actin - LFA-1

Malini Mukherjee

Immune Synapse

bi valency

Mukherjee & Orange. Center for Human Immunobiology

U373 TanCAR

T cell

TanCAR T cell

TanCAR T cell

2.8µm

Confocal STED

Phase Confocal

PLA

<1500A° <400A°

FRET

<80A° <3000A°

Malini Mukherjee

CAR Signaling - pLck - pZap70

CD45 CD4/CD8.

Imaris© Rendition

pLck

CD19.28ζ CAR GFP T cells

Imaging Toolbox: endodomain

Death Pathway - Fas - Perforin

F-Actin, Perforin, Pericentrin

Target

Effector Effector

Fas receptor Fas Ligand Target

CD19.4bb.ζ

Target

CD19 CD28.ζ

CD19 CD28.ζ

CD19. 4bb.ζ

Immunological Synapse

Imaging Toolbox: killing machinery

TanCAR behavior

Hegde et al. JCI 2016

Endodomain“fitting”

Truncated

NT

Zeta 28.Z

BB.Z

OX.Z

28.OX.Z

BB.OX.Z

OX.BB.Z

28.BB.Z

Kristen Fousek

1:20

160hr

0

20

40

60

40:1 20:1 10:1 5:1

U373-GBM

... & now NON adherent cells Daudi

NT (-ve)

Tumor Only (-ve)

Triton (+ve)

CAR T cell products

Time (Hours)

1:30

Hegde et al. JCI 2016

of Surrealism & Realism

phot

ons/

cm2 /

sec/

sr

days

1.E+04

1.E+05

1.E+06

1.E+07

1.E+08

0 8 17 31 63

NT T cells

Pooled T cells

biCAR T cells

TanCAR T cells

100

**

**

**

*

*

**

**

1:3

Preclinical CAR Assembly Line Ta

rget

(s)

Time

Effort

Value

Sim

ulat

ion Fu

ncti

on

1. Empiric preclinical to clinical development of HER2CAR T-cells

2. Rational development of novel CARs

3. Realistic in preclinical testing 4. Unresolved: modeling toxicity

Look how far you have fallen THEN repent! Revelation 2:5

SILICO > VITRO > VIVO

Acknowledgments THE PATIENTS Center for Cell and Gene Therapy Baylor College of Medicine Stephen Gottschalk Helen Heslop Malcolm Brenner

Baylor College of Medicine Texas Children’s Hospital The Methodist

MD Anderson Cancer Center-UT (P Anderson, E Kleinerman)

The Methodist Research Institute (Y. Kew, R. Grossman, SZ Powell, D Baskin, J Zhang)

Children’s Mercy and PACT (Doug Meyers) The National Cancer Institute (B St Croix) University of Florida (B Fletcher) Georg Speyer Haus (J Koch; W Wels)

TEM8 in TNBC

Tiara Byrd ….

Tiara Byrd

**

DFS

pro

babi

lity