Challenges and Opportunities for Immunoprofiling Using a Spatial High-Plex Technology: The NanoString GeoMx® Digital Spatial Profiler

Abstract

Characterization of the tumor microenvironment through immunoprofiling has become an essential resource for the understanding of the complex immune cell interactions and the assessment of biomarkers for prognosis and prediction of immunotherapy response; however, these studies are often limited by tissue heterogeneity and sample size. The nanoString GeoMx^® Digital Spatial Profiler (DSP) is a platform that allows high-plex profiling at the protein and RNA level, providing spatial and temporal assessment of tumors in frozen or formalin-fixed paraffin-embedded limited tissue sample. Recently, high-impact studies have shown the feasibility of using this technology to identify biomarkers in different settings, including predictive biomarkers for immunotherapy in different tumor types. These studies showed that compared to other multiplex and high-plex platforms, the DSP can interrogate a higher number of biomarkers with higher throughput; however, it does not provide single-cell resolution, including co-expression of biomarker or spatial information at the single-cell level. In this review, we will describe the technical overview of the platform, present current evidence of the advantages and limitations of the applications of this technology, and provide important considerations for the experimental design for translational immune-oncology research using this tissue-based high-plex profiling approach.

Keywords: biomarkers; digital spatial profiling; immune-oncology; pathology; tumor microenvironment.

Figure 1

DSP assay workflow. Schematic picture showing the workflow of profiling using the Geomx DSP.

Figure 2

Microphotographs showing immunofluorescence biomarkers used to visualize different elements of tumor tissues. (A) Surgical resected uveal melanoma sample stained with S100B/Pmell7 (melanoma cells), CD45 (immune cells), and CD3 (T-cells). (B) Surgical resected colon sample with a peripheral nerve identified by β3-Tubulin and Neurofilament. (C) Surgical resected breast carcinoma stained with PanCK (tumor), CD45 (Immune cells), and CD3 (T-cells) (D) Surgical resected lung carcinoma stained with PanCK (adenocarcinoma cells), CD20 (B-cells) and CD3 (T-cells).

Figure 3

Microphotographs showing a section of a biopsy with invasive basaloid rectal carcinoma tissue. (A), Multiplex immunofluorescence performed with DSP assay highlight epithelial cells with PanCK (green) and nuclei with Sytol3(Blue), Carcinoma nests showed low levels of panCK expression while superficial squamous epithelium shows strong panCK expression, (B) Marked up image showing ROI selection in tumor cells and tumor stroma, ROI strategy to identify carcinoma from tumor stroma is only possible with non-segmented polygons.

Figure 4

Micro photographs showing artifacts in immunofluorescence DSP slide from a non-small cell lung carcinoma tumor sample (A) ROI was drawn with a polygon ROI, avoiding elastic fibers (red arrows) that emit non-specific fluorescence signal (yellow), (B) Area on the right (red arrows) show numerous red blood cells emitting non-specific fluorescence signal (yellow). Both, elastic fibers and red blood cells, interfered with segmentation for B-cells. (C) Tumor area on the left shows fibrosis with non-specific fluorescent signals (white arrow). (D) An area out of focus (white arrow) is observed on the left side of the image.

Figure 5

Microphotograph of immunofluorescence sections of surgical resected non-small cell carcinoma (A) Squamous cell carcinoma; (B) adenocarcinoma) using DSP assay with PanCK (tumor), CD3 (T-cells), CD20 (B-cells) and Syto13 (nuclei) as visualization markers. (B, D) illustrates different segmentation strategies for tumor, stroma and T cells. In B the segmentation was performed in tumor (cyan mark up) and stroma (yellow mark up) segments based in panCK expression, with this strategy, tumor segments include tumor cells and Intra-epithelial immune cells (white arrows), and stroma segments include all tissue elements among tumor segments (B), In (C), segmentation was performed in tumor-(cyan mark up), B-cell (yellow mark up) and T-cells (red mark up) segments based in cell biomarker profile, intra-tumor T-cells (white arrows) are included in the T-cell compartment (D).

Figure 6

Heatmap of an initial dataset obtained with DSP assay to illustrate the visualization of data as quality control tool. The DSP counts of one region of interest indicated with a black arrow show no or very low DSP counts from all targets of the DSP protein panel, including the housekeeper proteins: GAPDH, Histone3 and S6. The DSP quality control report showed a positive control normalization tag in this specific region of interest.