ConclusionHere we present Brightplex, an integrative solution developed by HalioDx for sequential chromogenic multiplex IHCtogether with whole slide digital pathology analysis.This tool allows to:→ Evaluate multiple biomarkers specific to tumor infiltrating myeloid cells and spatial relationship between cells→ Maximize data harvesting per sample (tissue section)

Myeloid-derived suppressor cells (MDSC) assessment using an automated sequential chromogenic multiplex assay (Brightplex)Anna Martirosyan, Assil Benchaaben, Aurélie Collignon, Emilie Bonzom, Matthieu Duval, Alboukadel Kassambara,Felipe Guimaraes, Emmanuel Prestat, Christophe Haond, Jacques Fieschi | HalioDx, Marseille, France.

SITC 2018

MethodSTEP 1: Sequential Multiplex IHC Workflow

STEP 2: Digital Pathology

Module 1 • Virtual slides alignment• Color deconvolution for each biomarker• Creation of peudo-color image

Module 2HALO™

Indica Labs

• Detection of positive cells for each biomarker• Identification of regions of interest (ROIs: core tumor,

invasive margin, tumor parenchyma/ stroma)

Module 3 • Calculation of tumor-infiltrating myeloid cell densities• Map reconstruction to locate cells of interest within tumor

A) Unique combination of biomarkers to detect different MDSC populations on a single tissue section

CD11b+ CD15+ LOX1+ HLA-DR- CD14-

C) Assessing the abundance of major myeloid cellpopulations within NSCLC microenvironment

CD11b+ CD14+ S100A9+ HLA-DR- CD15-

Tissue Segmentation:

Invasive MarginTumor Parenchyma

Tumor Stroma

D) Myeloid cells distribution within NSCLC landscape

CD11b CD15 CD14 LOX1 S100A9 HLA-DR

Positive celldetection

ROI Myeloid Cell Populations Density (cells/mm2) in NSCLC Sample

TIM M-MDSC PMN-MDSC Granulocytes Neutrophils MonocytesTumor 270 2,2 43 130 68 33

Invasive Margin 190 2,8 31 82 29 40

Parenchyma 250 2,1 40 120 63 26

Stroma 440 3,5 64 210 100 68

MDSC and other tumor infiltrating myeloid cell types were analyzed within differentregions of interest.

*TIM=Tumor-infiltrating myeloid cells (CD11b+)*Monocytes=CD11b+CD14+CD15-HLA-DRhigh

*Granulocytes=CD11b+CD15+CD14-

*Neutrophils=CD11b+CD15+CD14-LOX1-

Cell Quantification:

Monocytic MDSC (M-MDSC) and polymorphonuclear MDSC (PMN-MDSC) are two major subpopulations of MDSC described by their phenotypical, morphological and functionnal characteristics. Both subsets are found in many cancers. PMN-MDSC generally represent more than 80% of all MDSC.

Core Tumor

Future DirectionsThe detection and quantification of MDSC, exhausted T cells (CD3, CD8, PD1, LAG3, TIM3 Brightplex panel) and other tests from the Cancer Immunogram could be key in Lung cancer to:→ Predict patient response to immunotherapy→ Analyze the mechanisms of anticancer drug action→ Evaluate the immunosuppressive landscape of NSCLC tumors

PMN-MDSC:CD11b+CD15+LOX1+HLA-DR-CD14-cells

M-MDSC:CD11b+CD14+S100A9+HLA-DR-CD15-cells

Make educated decisions and design new treatment strategies

MDSC are a heterogeneous population of cells:Characterized by their myeloid origin and immature stateEndowed with highly suppressive machineryHamper both innate and adaptive immune responsesM-MDSC are suspected to be more immunosuppressive than PMN-MDSC

Like all immature cells, phenotyping MDSC requires a highly complex combination of biomarkers.Brightplex, a multiplex technology initially developed to characterize T cell exhaustion, wassuccessfuly applied to the detection and quantification of MDSC on a single section of NSCLCFFPE tissue.REFERENCES: Condamine T et al, Sci Immunol, 2016, Bronte V et al, Nature Com, 2016, Gabrilovich D.I., Cancer Immunol Res, 2017, Lin Y et al, Cancer Immunol Immunother, 2018, Tcyganov E et al, Current Opinion Immunol, 2018, Fleming V et al, Frontiers in Immunol, 2018.

BackgroundThe detection and quantification of MDSC populations is part of the Cancer Immunogram. Deciphering the complexity of tumor immune contexture is requiredto understand, predict and overcome resistance to anti-PD(L)1 immunotherapy of cancer.

In the context of NSCLC, MDSC accumulation in the tumor environment isassociated with unfavorable prognosis and correlates with

The progression-free survival,The response to chemotherapy,The metastatic burden in patients.

Brig

htpl

exIH

CCe

llde

tect

ion

Cell

dete

ctio

n

→ Reconstructed map showing MDSC spatial distribution within tumor context.

→ One color corresponds to one phenotype:

TIM (Tumor-Infiltrating Myeloid cells)M-MDSCPMN-MDSCGranulocytesNeutrophilsMonocytes

TIM: Tumor-Infiltrating Myeloid Cells

HLA-DR+cells

Monocytes

Granulocytes

B) Hierarchical gating based on staining intensity to identify populations of Tumor-Infiltrating Myeloid Cells

Brig

htpl

exIH

C

M-MDSC

Neutrophils PMN-MDSC

This work was supported by French National Research Agency (ANR) in the context of the PIONeeR project

One FFPE slideDeparaffinization

Antigen retrieval

Image processingand DP analysis

Repeat6 times

Antibodiesstripping

Slide scanning

Staining

Coverslipremoving

AECdestaining

Chromogenicrevelation

Mounting

LOX1 CD11bCD15CD14S100A9HLA-DR

s

CD11b CD15 CD14 LOX1 S100A9 HLA-DR

Virtual slide with

immunostaining

Pseudo colorImage