Digital Spatial Profiling

In the era of immuno-oncology, it’s critical to unravel mechanisms of action underlying sensitivity or resistance to current immunotherapies in order to i) identify predictive biomarkers of response and then ii) discover novel actionable therapeutic targets. However, cancer biology, particularly anti-tumor immunity, is complex and is characterized by a continuous interaction between tumor cells and their microenvironment – both characterized by specific features, which occurs heterogeneously within the space.

It therefore becomes primordial to benefit from a technological platform that allows to deal with the tumor biology complexity/heterogeneity and get specific investigations in compartments of interest (e.g. tumor vs stroma). Thanks to the GeoMx platform (Nanostring Inc.), which combines the spatial resolution of immunohistofluorescence with the high-plex gene/protein expression analysis, the Digital Spatial Profiling (DSP) revolution has started.

Starting from specimens as simple as FFPE tissue samples, this novel technology offers through selected panels the simultaneous analysis of up to 96 mRNAs or 10’s-100’s proteins in several regions of interest (ROIs) divided in two compartments.
In addition to these panels, the comprehensive Cancer Transcriptome Atlas (CTA) - covering 1,800 genes and more than 50 cancer-related pathways -, as well as the Whole Transcriptome Atlas (WTA) - covering 18,000 protein-coding human genes - are now running!


Illustrative data

Spatial profiling of FFPE tissue section from human lung cancer
FFPE tissue section from human lung cancer was stained for standard fluorescent morphology markers (i.e. PanCK in red, CD45 in green) and nuclei counterstained with Sytox as to allow tissue segmentation. Concomitantly with the morphology markers, oligo-tagged ISH probes (RNA) targeting 1,800 genes from the CTA were applied. ROIs were then selected and segmented into stroma and tumor areas thanks to the morphology markers. Oligos were cleaved and released from selected areas (Areas of Illumination, AOIs) via UV exposure, and then collected via a microcapillary tube and processed for library preparation and next generation sequencing. 

FFPE tissue section from human lung cancer was stained for standard fluorescent morphology ...

GeoMx analysis of the determinants of the ICI response
Tissues were incubated with the GeoMx® Immune Pathways Panel probes and stained with the morphology markers CD45 and PanCytokeratin (PanCK). Regions of interest were drawn in mixed tumor/stroma areas and were then segmented in Tumor (PanCK) and Stroma (PanCK-) compartments. Expression of 78 protein-coding genes related to tumor microenvironment profiling were finally analyzed on the Nanostring nCounter platform.
 

Tissues were incubated with the GeoMx® Immune Pathways Panel probes and stained with the ...

Unsupervised clustering and differential gene expression analysis
Heatmap view of the unsupervised clustering of 15 ROIs (divided in 15 AOIs in stroma and 15 AOIs in tumor)  according to the 1,800 genes expression profile. In the heatmap, the Y-axis represents the 1,800 genes and the X-axis represents the AOIs. The expression levels from low to high are represented as a color gradient from blue to red, respectively.Volcano plot representation of differential gene expression between stroma and tumor areas collected from the different ROIs. Vertical dotted lines represent ± Log2(FC>2) while the horizontal dotted line indicates a p-value of 0.05. Highlighted genes depicted on the volcano plot are the significantly enriched ones in the stroma (left part, green)  and tumor compartment (right part, orange).  

Heatmap view of the unsupervised clustering of 15 ROIs (divided in 15 AOIs in stroma and 15 ...


Key features and advantages of digital spatial profiling

  • Non-invasive approach preserving sample integrity
  • Highly multiplexed RNA or protein detection - up to 1800 targets in a run from a single FFPE section
  • Whole slide imaging with up to 4 fluorescent channels to capture tissue morphology and select regions of interest (ROIs) for high-plex profiling
  • Possible multianalyte analysis for deep sample characterization
  • Digital quantitation with integrated data normalization and analysis