Precision Cancer Medicineand Big Data

with a focus on children with cancer

Gilles Vassal

3

Each human being is unique

Desoxyribonucleic acid, DNA

A T

G C

4 bases

1 human genome=

3 x 109 basepairs

25,000 to 30,000 genes

DNA sequencing of the First Human GenomeStarted 199013 years, $2.7 bnCompleted on April 14, 2013

Cancer

• Interaction between Genome and Environment

– Eg, Tobacco, UV, virus,…….

– Somatic alterations

• Genetic predisposition

– Eg Breast cancer (BRCA)

– Constitutional DNA

Cancer

Normal Lung Lung cancer

Genes

Cancer

Molecular alterations

Heterogeneity

Immune system

Proteins

definition of cancer

a tumor

an organ

a pathological sample

= A definition from the XIXth century

Significantly mutated pathways in adenocarcinoma of the lung

Ding et al. Nature 455, 1069, 2008

A definition from the XXth century

Lung Cancer Molecular Subtypes

Enrichment of phase 1 with EMLA4-ALK patients

RR = 1CR + 46PR/82 (57%) with a PFS at 6 months = 72%

EML4-ALK fusion gene = 2 to 7% of NSCLC

Phase I48 y old female

NSCLC with EML4-ALK

NEJM 363, 2010

Development of oral crizotinib, ALK MET inhibitor

Specific genetic traits can predict for the success of targeted agents

Recent examples of oncogenic addiction in solid tumors leading to active inhibitors - i.e.“superstars”!

B-RAF inhibitor in melanoma

(V600E BRAF mutation) NEJM 2010

ALK inhibitor in NSCLC

(ALK translocation) NEJM 2010

Hedgehog inhibitor in BCC

(PTCH mutation) NEJM 2010

Phase I data; all drugs are registered

vemurafenib

crizotinibvismodegib

Which drug(s) for Mister X?

Challenge

To prescribe

the right drug combination

At the right time

Precision Medicine

A public health and societalissue

To Increase survival

using expansive drugs

only for patients who willbenefit

Drugs/biomarker

Tailored medicine

B

M

B

M

B

M

BMB

M

Algorithms for decision making

19

Precision cancer medecine

At least for a first step : Stratified Medicine

1. Moving away from empirism and serendipityto a biology-based therapy

2. Matching the right drug with the right cancer type

3. Defining the right response biomarker on each patient’s tumor

4. Organizing molecular follow up of patients under therapy

The Ultimate goal

individual biology-guided treatment decision

MOLECULAR SCREENING*

CGH Array & NGS

WES, RNAseq

CLINICAL DECISION

Max <21 calendar days

FRESH TUMOR

BIOPSY PATHOLOGICAL CONTROL

TREATMENT

• Monocentric

• 1200 patients – 2011-2015

Antoine Hollebecque et al., ASCO 2013; Charles Ferte et al, AACR 2014

MOSCATO 01 trial:

*From 15 genes to 25000 genes In 4 yearsPer tumor sample

0

5

10

15

20

25

30

35

40

45

50

Mut

atio

n PI

K3C

AA

mp

lific

atio

n FG

F4A

mp

lific

atio

n FG

F3M

uta

tion

KRA

SA

mp

lific

atio

n FG

FR1

Am

plif

icat

ion

PIK3

CAA

mp

lific

atio

n ER

BB

2Po

siti

ve IH

C P-

MET

Am

plif

icat

ion

MD

M2

Am

plif

icat

ion

EGFR

Tran

slo

cati

on N

OTC

H2

Am

plif

icat

ion

FGF1

9M

uta

tion

BRA

FA

mp

lific

atio

n N

OT

CH2

Mut

atio

n ER

BB2

Mut

atio

n EG

FRA

mp

lific

atio

n M

ET

Mut

atio

n PT

ENLo

ss P

TEN

Gai

n PI

K3C

AA

mp

lific

atio

n AR

Mut

atio

n N

RA

SM

uta

tion

FBXW

7M

uta

tion

AKT

1A

mp

lific

atio

n CC

ND

1Lo

ss N

OTC

H1

Tran

slo

cati

on A

LKA

mp

lific

atio

n KI

TG

ain

FGFR

1D

elet

ion

PTEN

Mut

atio

n ER

BB3

Am

plif

icat

ion

AKT2

Am

plif

icat

ion

CCN

E1A

mp

lific

atio

n FG

FR2

Am

plif

icat

ion

IGF

1RA

mp

lific

atio

n N

OT

CH1

Am

plif

icat

ion

PIK3

CBD

elet

ion

CDK

N2A

Del

etio

n CD

KN

2BM

uta

tion

FGFR

3M

uta

tion

HR

AS

Mut

atio

n ID

H1

Mut

atio

n IN

PP4B

Mut

atio

n M

ETM

uta

tion

TSC

1Lo

ss R

B1A

mp

lific

atio

n CC

NE2

Am

plif

icat

ion

CDK

4G

ain

FGF3

Mut

atio

n N

OTC

H4

Mut

atio

n TS

C2

Am

plif

icat

ion

EIF4

EBP1

Am

plif

icat

ion

PIK3

R5A

mp

lific

atio

n RI

CTO

RD

elet

ion

CHE

K1D

elet

ion

NF1

Del

etio

n TS

C1G

ain

AKT

1G

ain

AR

Mut

atio

n A

PCM

uta

tion

BRC

A2

Mut

atio

n FG

FR1

Mut

atio

n LK

B1

Mut

atio

n M

AP2K

4M

uta

tion

RB1

Mut

atio

n TP

53M

uta

tion

VH

LLo

ss C

DKN

2A

Loss

CD

KN2

BTr

ansl

oca

tion

ER

GTr

ansl

oca

tion

PTE

NTr

ansl

oca

tion

RO

S1A

mp

lific

atio

n AK

T1A

mp

lific

atio

n AL

KA

mp

lific

atio

n BR

AF

Am

plif

icat

ion

BRD

4A

mp

lific

atio

n CD

K2

Am

plif

icat

ion

CDK

6A

mp

lific

atio

n CD

K8

Am

plif

icat

ion

FGFR

Am

plif

icat

ion

FGFR

4A

mp

lific

atio

n ID

H2

Am

plif

icat

ion

IGF

1A

mp

lific

atio

n KD

RA

mp

lific

atio

n M

YCC

Am

plif

icat

ion

NO

TCH

3A

mp

lific

atio

n PD

GFR

AA

mp

lific

atio

n VE

GFA

Del

etio

n ST

K11

Del

etio

n TP

53

Gai

n FG

F4Po

siti

ve IH

C To

tal-M

ETM

uta

tion

CDKN

2A

Mut

atio

n FG

F19

Mut

atio

n FG

F3M

uta

tion

FGFR

Mut

atio

n FG

FR4

Mut

atio

n H

ER3

Mut

atio

n KI

TM

uta

tion

MTO

RM

uta

tion

PDG

FRA

Mut

atio

n ST

K11

Tran

slo

cati

on R

ET

105 actionable aberrations used to choose the targeted treatment

Fre

qu

en

cy (N

)

27 most frequent abormalitiesrepresent 78% of the actionable aberrations

BSI LCI

LUNG BIOPSYSCAPULAR BIOPSY

FI

Biopsy IrradiationLegend:

BSI

BSI: brain stereotxic irradiation; LCI:lung conformational irradiation; FI:Femoral irradiation; SC: scapula irradiation

2012

Dec

SI BSI

JanDecSeptJul Dec Oct Nov Jan

2009 2010 2013 2015

EGFR Del19 and T790M-positive

Gefitinib AZD9291

EGFR del19 and T790M wild-type

CDDPPEM

EGFR Del19 and T790M wild-type

EGFR Del19 and T790M wild-type

EGFR Del19 and T790M wild type

Figure 1. Patient I Clinical Course Including Treatment History and Relevant Imaging Studies and Tumor Biopsy Specimens

FISH HER2 positiveHER2 amplification (CGH)

No HER2 amplification (CGH)

Courtesy A Bardelli

Cancer in Young People in Europe

Each year: 35 000 new cases 15,000 <15 years and 20,000 15 -24 years

80% are disease-free at 5 years with multidisciplinary treatments 300,000 EU citizens are childhood cancer survivors

2/3 have long-term side-effcets

6,000 young people die each year

Cancer: first cause of death by disease beyond one year in EU

STILL A PUBLIC HEALTH ISSUE

Pediatric malignancies are different from

malignancies in adults

http://www.siope.eu/SIOPE_StrategicPlan2015/

By 2025:

To increase cure rate in patients with poor

prognosis malignancies

To increase quality of life (cure) in survivors

To tackle inequalities

SIOPE Strategic Plan

A long term sustainable strategy

Designed within the FP7

Created in 2003, ITCC is a consortium of 49 institutions in 12 countries

INNOVATIVE THERAPIES FOR CHILDREN WITH CANCER 2

ITCC runs a comprehensive clinical and biological early evaluation program of anticancer drugs for children and adolescents.

Each year, 4 500 patients with cancer are diagnosed in ITCC centers.

20% of them experience recurrenceof their disease and are offered to participate in ITCC trials.

78 patients included; Data from 60 with results from 65 interventions

62% solid tumors; 38% CNS tumors

Median age: 11.1 years (range, 0.8-24.3y)

56/60 patients had a molecular analysis done

32 of 56 (57%) had at least 1 actionable target

MOSCATO-01 (pediatric cohort)

Bioinformatics of Precision Cancer MedicineSpeed and storage are essential!

• Raw data for one sample and calculation = 300 Go

• One patient = Tumor + constitutional DNA

• quality control+mapping+variant calling+annotation

• DELL Cluster* (since 215) [servers + storage+network]

– 96 samples in parallel per day versus 12 per day

• Acceleration = x11

• 1000 patients = 600 To duplicated = 1,2 Po

*dedicated to the pediatric programm

M

A

T

C

H

SHARE

With ECTGs

The Innovative Therapies & PCM Programme

EU Clinico

Biological

Database

New knowledge, targets, pathways 3

Paediatric

New

Drug

Development

4

New targeted and

immune therapy

drugs

2

Trials w single agents

and combinations

A tumour molecular

profile* for patients

at relapse

Molecular Matching Trials

1

*WES, RNAseq, Immuno

MATRIX trial (Genentech/Roche)(2 drugs – atezolizumab ; cobimetinib)

MAPPYACTS (France, Spain, Denmark, Italy)

1. Generate molecular profiling for each patient

2.MATCH

4. Create

new druggable pathways

for specific pediatric drug development

INFORM (Germany)

iTHER (Netherland)

S- PED (UK)

The ITCC Precision Cancer Medicine program

Molecular Matching Trials at relapse

eSMART trialIST multi-agent from multi-company

Phase 1 & 2 ITCC Trials(sponsored by industry and ISTs)

3. Evaluate drugs and combinations

European clinico –biological database

5. New knowledge

WES

, RN

A s

eq

, met

hyl

om

eim

mu

no

ph

en

oty

pe

1000 childrenand adolescents

4 years

Projet National MAPPYACTS

ON-PURPOSE FRESH TUMOR BIOPSY/SURGERY & PATHOLOGY CONTROL

MOLECULAR PROFILING(CGH & NGS)

MOLECULARTUMOR BOARD

TREATMENTBiopsie à la

rechutePortrait Moléculaire

Tumoral (WES/RNAseq)

RCP moléculaire Traitement

RCP

moléculaire

Enfant et adolescent, en rechute, tumeur solide et leucémie

MoleculAr Profiling for Pediatric and Young Adult Cancer Treatment Stratification

300 enfants en 3 ans

PHRC 2015

Fondation ARC

Ouverture décembre 2015

PI B Geoerger

RCP

Nationale

Children and their parentsneed to travel to investigating centersto access early innovation

A multistakehoder platform(Academia, Parents, Industry, Regulatory)

Vassal, Eur J Cancer, 2015, 51, 218

Precision Cancer Medicine Programm

Thanks to the patients and their family

Thanks to: