IMODI is an operational consortium to continuously deliver new predictive models in regards to specific clinical needs and diversity, All results are available for new therapeutic and diagnostic candidate selection, Ex-vivo assay predicts in-vivo cisplatin and docetaxel sensitivity in a lung PDX model (other drugs and models are under investigation), Chemotherapeutic agents could impact the microbiota composition and microbiome analysis could learn about drug toxicity and tumor response, 2nd generation of PDX model with human microenvironment would help the selection of clinical drug candidates (impact of the humanization on the pharmacological drug profile is pending).

F. Le Vacon2, D. Guenot22, L. Arnould12, A. Bruno10, L. Calvet9, M. Colombel16, J. Corre17, O. Cuvillier11, B. Eckel20, O. Degoul3, A.Gonzalez-Jouhanneaud8, J. Iovanna19, M. Kuras1, Ch. Lautrette6, B. Malavaud13, Ph. Merle18, F. Meyer-Losic7, F. Praz15, O. Rosmorduc23, J.E. Sarry17, S. Tabone14,21, Ph. Vaglio4, L.Ysebaert13, O. Duchamp5

1 Ariana Pharmaceuticals, Paris; 2 BIOFORTIS MERIEUX NUTRISCIENCES, Saint-Herblain; 3 CTI-BIOTECH, Meyzieu; 4 Modul-Bio, Marseille; 5 Oncodesign, Dijon ; 6 OncoMedics, Limoges; 7 Ipsen Innovation, Les Ulis; 8 Pierre Fabre Research Institut, St-Julien-en-Genevois; 9 Sanofi, Vitry-sur-Seine; 10 Servier Research Institut, Suresnes; 11 CNRS U5059, Toulouse; 12 Centre Georges François Leclerc, Dijon; 13 Toulouse Hospital; 14 Centre Léon Bérard, Lyon; 15 INSERM U938, Paris; 16 INSERM U1033, Lyon; 17 INSERM U1037, Toulouse; 18 INSERM U1052, Lyon; 19 INSERM U1068, Marseille; 20 INSERM U1111, Lyon; 21 Synergie Lyon Cancer, Lyon; 22 Strasbourg University; 23 Pitié-Salpetrière Hospital, Paris.

This work was supported by a grant from and the French Government

What about IMODI…

IMODI Initiative: a Novel Holistic and Integrative Approach with Patient-Derived Tumor Models # A38

Tumor model development: 9 cancer pathologies 10-30 models / cancer 200 in-vivo models (SC & OT) 50 in-vitro cancer cell lines and

CAF

PDX model samples: Tumor Blood Stools 200 models

Tumor characterization Histology Copy number & polymorphisms (Cytoscan , Affymetrix) Gene Seq: Exome or panel of 100 genes (Illumina, MySeq) RNA expression: Illumina RNAseq or U133-Affymetrix transcriptome In-vivo & ex-vivo Pharmacology (sensitivity to 4 standards of care) Gut microbiota characterization Microbiota gene metasequencing (Microbiota identity card)

DATA MINING

Identification of biomarkers

Patient samples: Tumor Blood Stools Clinical history 800 patients

Example of a well characterized NSCLC PDX collection Highly conserved phenotype and genotype Histological PDX profile are in concordance with those

observed in the patient’s tumor Major molecular subtypes are represented in the

NSCLC collection EGFR-mutated models are under development

PDX dissociation and culture lead to expansion of both mouse fibroblast and human epithelial cells

Culture at high level of fibroblast confluency allowed the development of epithelial cell organized in colonies

Once a critical cell number is reached, the epithelial clusters are subcultured and develop into a mouse fibroblast-free culture of epithelial cell currently being characterized

IM-LUN-060-PDX-INV (Basal Medium P0 to P3) 250µm

The national IMODI (Innovative MODels initiative) consortium including 25 partners (pharmas, SMEs, academic research labs and clinical centers) aims at developing more predictive tools for better selection of new effective treatments to combat 9 cancer pathologies. These developments include:

Collection of in-vivo PDX models, Collection of in-vitro derived cell lines, 2D & 3D ex-vivo assays, In-vivo humanized models (immune system, liver and tumor stroma), Characterization of tumor histology, gene mutation, gene expression, pharmacological responses, gut microbiota, Biobanks of tumors, blood, serum and stools, Central data base, Data mining,

Results on NSCLC lung cancer model developments, characterization and data analysis are presented as an example of the IMODI holistic and integrative approach.

L U N -N IC -0 0 1 4

T im e p o s t tu m o u r in d u c tio n (d a y s )

Tu

mo

r v

olu

me

(m

m3

)

2 0 4 0 6 0 8 0 1 0 0 1 2 00

5 0 0

1 0 0 0

1 5 0 0

2 0 0 0

V e h ic le

C is p la tin e 3 m g /k g

L U N -N IC -0 0 1 4

T im e p o s t tu m o u r in d u c tio n (d a y s )

Tu

mo

r v

olu

me

(m

m3

)

2 0 4 0 6 0 8 0 1 0 0 1 2 00

5 0 0

1 0 0 0

1 5 0 0

2 0 0 0

V e h ic le

D o c e ta x e l 9 m g /k g

L U N -N IC -0 0 6 0

T im e p o s t tu m o u r in d u c tio n (d a y s )

Tu

mo

r v

olu

me

(m

m3

)

2 0 4 0 6 0 8 0 1 0 00

5 0 0

1 0 0 0

1 5 0 0

2 0 0 0

V e h ic le

C is p la tin e 3 m g /k g

L U N -N IC -0 0 6 0

T im e p o s t tu m o u r in d u c tio n (d a y s )

Tu

mo

r v

olu

me

(m

m3

)

2 0 4 0 6 0 8 0 1 0 00

5 0 0

1 0 0 0

1 5 0 0

2 0 0 0

V e h ic le

D o c e ta x e l 9 m g /k g

Chemosensitivity evaluation in 2D PDX-derived primary cultures of cells extracted from LUN-NIC-0060 PDX tumors in defined medium after 5 days of culture by ATP level measurement (metabolic activity) and calcein & ethidium test (cell mortality, data not shown).

Cisplatin is more effective than docetaxel on the LUN-NIC-0060 model as confirmed by the in-vivo data

What about IMODI

Histology and Genomic Characterization

In-vivo Pharmacological Response to Standards of Care

Significant efficacy of cisplatin and gemcitabine on the LUN-NIC-0060 epidermoid model Marginal activity of gefitinib on the LUN-NIC-0014 acinar

adenocarcinoma model (EGFR wt, KRAS wt, BRAF wt, ALK wt, ROS1 wt)

LUN-NIC-0014 PDX response to cisplatin and docetaxel correlates with patient outcome (non responsive to cisplatin + docetaxel)

Conclusion and perspectives

INTRODUCTION

General process

RESULTS

Gut Microbiota Analyses

In-vitro PDX-Derived Cell Line Establishment

Tumor Microenvironment Humanization

Ex-vivo Pharmacology Assay

Nb PDX collection under development

CD34+ humanization of mice did not modify the PDX tumor growth

Tumor xenografting increases Treg circulating cells

Human Treg cells infiltrate the LUN-NIC-0084 tumors (cytometry)

Treatment

T0 Graft

T4 End of

experiment

T3 1 week

post-treatment

T2 End of

treatment

T1 Before

treatment

ESTABLISHMENT OF LUNG PDX MODELS

TREATMENT & STOOL SAMPLE COLLECTION METASEQUENCING ANALYSE OF MICROBIOTA

COMPOSITION

Characterization of the microbiota composition and variation during chemotherapy

• 5 groups:

• Experimental timeline:

• Targeting variable regions of the 16S rRNA gene (V3 & V4 conserved regions)

- Vehicule - Cisplatin - Pemetrexed - Gefitinib - Docetaxel

Mesure of Weight & Tumor Volume

pemetrexed T2 Vehicule T2

20% 0.5% Escherichia-Shigella

T1 T2 T3 T4

V Ci P G D V Ci P G D V Ci P G D V Ci P G D Pr

opor

tion

of re

ads

assi

gned

to

Phy

lum

(%

)

MICROBIOTA COMPOSITION

Phylum: Proteobacteria Family: Enterobacteriaceae Genus: Escherichia-Shigella

after pemetrexed treatment

Healthy LUN-NIC-0084

KRAS

EGFR

ALK

HER2

BRAF

PIK3CA AKT1

MAP2K1 NRAS

ROS1

KIF5B

RET

Others

PDX gene mutation diversity

3

5

8

1

5

2

2

1

1

2