immunogram to decipher pd1/l1 ici resistance: a proof of … · 2019. 12. 21. · 01-01-022....

Data acquisition and distribution across patients

Patient groups display heterogeneous immune profilesTechnical workflow for immunoprofiling at baseline

Immunogram to decipher PD1/L1 ICI resistance: a proof of concept in advanced NSCLC patients of the PIONeeR ProjectFlorence Monville1, Frédéric Vély3,4, Joseph Ciccolini2,3, Florence Sabatier2,3, Stéphane Garcia2,3, Vanina Leca1, Marion Fabre2, Christelle Piperoglou3, Pernelle Outters1, Laurent Arnaud3, Laurent Vanhille1, Caroline Laugé1, Anna Martirosyan1, Aurélie Collignon1, Marie Roumieux2, Julien Mazières5, Maurice Perol6, Francoise Dignat-George2,3,7, Eric Vivier3,4,8, Fabrice Barlesi2,3, Jacques Fieschi1

1 HalioDx, Marseille/France, 2 Centre de Recherche en Cancérologie de Marseille (CRCM), Inserm U1068 Aix Marseille University, Marseille/France, 3Assistance Publique des Hôpitaux de Marseille (APHM), Marseille/France , 4 Centre d’Immunologie de Marseille-Luminy (CIML), Marseille/France, 5 CHU Toulouse/France, 6 Centre Léon Bérard, Lyon/France, 7 Vascular Research Center (VRCM), Marseille /France, 8 Innate Pharma, Marseille/France.

BackgroundIn the management of advanced Non-Small Cell LungCarcinoma (NSCLC), PD1/L1 Immune CheckpointInhibitors (ICIs) have been shown to increase OverallSurvival (OS) over standard 2nd-line chemotherapy (CT).While this long-term increase in OS is driven by about20% of patients, others display disease progressionduring the first weeks.In clinical practice, high PD-L1 expression in tumor cellsas well as high Tumor Mutational Burden provide with anenriched population of PD(L)1 inhibitors responderswithout being sufficiently precise to exclude patientsfrom treatment.

ChallengeUnderstand biological background behind resistancesto PD1/L1 ICIs through a comprehensive multiparametricbiomarkers strategy (PIONeeR biomarkers program).Identify a relevant predictive algorithm based onbiomarkers combination adaptable to clinical practice.

Reference• Blank CU. et al., Science, 2016

ConclusionThis preliminary profiling highlights immunological heterogeneityacross patients not evaluated in current clinical practice.Such analyses confronted to clinical parameters may highlightbiological mechanisms explaining resistance to anti-PD(L)1 ICIs.These profiles will be expanded to additional biomarkers andoptimized on more than 400 patients to identify reliable predictivebiomarker combinations.

Strategy450 patients prospectively collected with stage IV orrecurrent NSCLC with first line anti-PD1/L1immunotherapy.Preliminary analysis of more than 20 biomarkers on 11patients at baseline as a proof of concept ofImmunogram1 performances to identify relevantcombinations of biomarkers to predict response to ICItreatment.

PIONeeR study sampling workflow

Foreignness : TMB (1/2), TCR clonality (1/2)Immune suppression within the tumor : Treg (1/3), M-MDSC (1/3), PMN-MDSC (1/3)PD-L1+ / Exhaustion : PD-L1+ cell density (1/2), PD1/LAG3/TIM-triple positive Tcells (1/4), PD1/LAG3/TIM-double positive Tcells (1/4).Immune infiltration : Tumor Infiltrating Lymphocytes (1/2) and Tumor Infiltrating Monocytes (1/2)Circulating lymphocytes : Total CD45+ lymphocytes

This work is supported by French National Research Agency (17-RHUS-0007); a partnership of AMU, AP-HM, CNRS, Inserm, Centre Léon Bérard, Institut Paoli-Calmettes, AstraZeneca, HalioDx, Innate Pharma & ImCheck Therapeutics, sponsored by AP-HM and initiated by Marseille Immunopole.Drug supply is funded by Innate Pharma and AstraZeneca.

0102030405060708090

100

Tota

l TCR

βRe

adsC

ount

Immune infiltration

Circulatinglymphocytes

PD-L1 and immune checkpoints

Immuno-suppressive cellsRegulatory T cells (Treg)Myeloid-derived Suppressor Cells:- Monocytic (M-MDSC)- PolyMorphoNuclear (PMN-MDSC)

Foreignness

CD45

SSC

SSC

CD3 CD8

CD4

CD19

CD16/56

M-MDSC

PMN-MDSC

CD3 CD8 CD11bCD4

PDL1 PD1

0

50

100

150

200

250

300

350

0

200

400

600

800

1000

1200

1400

1600

0

50

100

150

200

250

300

350

Cell

dens

ities

(cel

ls/m

m2

0

200

400

600

800

1000

1200

1400

Cell

Dens

ities

(cel

ls/m

m2 )

T Cell clones diversity(each clone represented by one color) bar size representsTCRβ counts and TumorMutational Burden

0

500

1000

1500

2000

2500

3000

Bloo

d Ce

llco

unt (

cells

/µL)Circulating lymphocytes

Immune cells infiltration

T-cells exhaustion

Immuno-suppressive cellsM-MDSC:

CD11b+ CD14+ S100A9+ HLA-DRlow/-

PMN-MDSC:CD11b+ CD15+ S100A9+ LOX1+

HLA-DRlow/-

Treg: CD4+ FOXP3+

T-cell clonality Tumor Mutational BurdenPD-L1 < 1%, TMB > 10

0

20

40

60

80

100

PD-L1+ / Exhaustion

Immune Suppression

Foreignness CirculatingLymphocytes

TumorInfiltration

PD-L1 < 1%, TMB < 10

0

20

40

60

80

100

Immune Suppression

Foreignness CirculatingLymphocytes

TumorInfiltration

PD-L1+ / Exhaustion

S100A9CD14CD11b HLA-DR

S100A9CD15CD11b HLA-DRLOX1

01-0

1-02

2

01-0

1-02

6

01-0

1-02

7

01-0

1-03

3

01-0

1-04

0

01-0

4-00

2

01-0

4-00

6

01-0

5-00

2

02-0

1-00

1

03-0

1-00

2

03-0

2-00

1

Patients IDs

0 17 4,3 6,4 7,1 1,4 13 5 5 9,2 8,5TMBAll events Lymphocytes CD3+ CD3-

Total CD3+ T cellsCytotoxic CD3+CD8+ T cellsHelper CD3+CD4+ T cellsCD3+CD56+ T cellsActivated CD3+ HLA-DR+ T cellsCD19+ B cellsCD3-CD56+ NK cells

CD3+CD8+ T cellsCD3+CD8- T cellsGranulocytesMonocytes Myeloid

Lymphoid

Moderate: 1 checkpointSevere: 2 or 3 checkpoints

PD-L1+ cell density in the tumor

Variables were combined in five biological axes to define an Immunogram; five arbitrary scoreswere calculated:• Each test results were transformed to fit 0 – 100, where 100 = maximum value observed

across the 11 samples• Each normalized result was weighted as indicated in brackets below

Mutations per megabase relative to normal counterpart

TCRβ reads count

LAG3 TIM3

CD4brown

FOXP3pinkV2VS

Post-anti-PD(L)1 treatment samplesPre-anti-PD(L)1 treatment samples

Diag V3 V4 V5V1 V6R R R R

Longitudinal blood sampling

Blood samplePre-tttFFPE

tumor

Pre-tttBlood sample

Pre-tttStool sample

Post-tttFFPE tumor

Post-tttBlood sample

Post-tttStool sample

Tumor biopsy at diagnostic

Blood sampling prior to treatment

20 FFPE slides

DNA/RNA extractionTargeted NGS on DNA

IHC staining + Digital pathology analysisIMMUNOSCORE® CR IC – CD8, PD-L1IMMUNOSCORE® CR TCE – CD3, CD8, PD1, TIM3, LAG3IMMUNOSCORE® CR Treg – CD4, FOXP3IMMUNOSCORE® CR MDSC – CD11b, CD14, CD15, S100A9, LOX1, HLA-DR

Genomics - Targeted sequencingTMB: 7661 exonic regions, 478 genes, 1,41MbT cell clonality: TCRβ sequencing

Immunoprofiling through immunostaining & cytometryCD45, CD3, CD8, CD4, CD19, CD56

Endothelial activation through MACS & cytometry on‐ Circulating endothelial cells (CEC, CD146+)‐ Endothelial MicroParticles (EMP, CD31+CD41-)‐ Circulating Endothelial leukocytes (7AAD, CD45, CD34, KDR)

NA

NA

NA N

A

01-01-02203-02-00101-01-02601-04-00601-04-00203-01-002

PD-L1 > 1 %

0

20

40

60

80

100

PD-L1+ / Exhaustion

Immune Suppression

Foreignness CirculatingLymphocytes

TumorInfiltration

Anti-PD(L)1 respondersenriched population

Anti-PD(L)1 treatmentexcluded population

PD-L1 expression

Poster P117

NKG2A inhibitor (Monalizumab)

+ Durvalumab (N=30)CD73 inhibitor

(MEDI9447) + Durvalumab (N=30)

ATR inhibitor (AZD6738)

+ Durvalumab (N=30)

Stat3 inhibitor*

(AZD9150) + Durvalumab (N=30)

Control arm Docetaxel

monotherapy (N=30)