Identifying Spatial Expression Patterns in the Tumor Microenvironment
In this webinar, Dr. Rognoni describes the development of an automated workflow for quantitative analysis of multispectral immunofluorescence whole-tissue scans. This technique, developed at Definiens, allows the identification of spatial patterns for immunoprofiling, overcoming the limitation of small regions of interests and providing significant amount of data on the entire tumor region.
Advancement in cancer immunotherapy is associated with unraveling the complexities of immune suppressive mechanisms across different cancers. Quantification on multispectral multiplex-immunofluorescence (mIF) images allows detection of several biomarkers in a single section. In addition, new evidence using mIF techniques suggests that spatial analysis reveals novel insights in the tumor microenvironment. However, multispectral imaging is tile based due to long scanning periods, which leads to insufficient data acquisition for significant spatial analysis.
In this webinar, we present an automated workflow to study the spatial patterns of infiltrating cells in the tumor microenvironment based on multispectral mIF whole slide scans. This was used to study the relationship between tumor proliferation and immune-response in non-small cell lung cancer (NSCLC) resections.
Formalin fixed, paraffin embedded NSCLC resection samples were stained by MedImmune with a custom-developed 7-plex mIF panel (CD68, CD8, Ki67, PD1, PD-L1, pancytokeratins-CK & DAPI) using the Opal method (PerkinElmer) in combination with an automated staining paltform to ensure high throughput and consistent sample processing. Tiled scans were acquired with a Vectra Polaris (PerkinElmer) multispectral imaging system.
We developed a workflow which composes the tiled unmixed multispectral data to a whole-slide image and optimizes the layers for screen display and automated image analysis. Furthermore, images were shared on Definiens collaboration platform along with a chromogenic-IHC pseudocolor of the IF CK/DAPI signals and co-registered H&E section for pathologist annotations. These annotations were used in defining tumor center and invasive margin.
The image analysis includes single-cell detection on the complete slide along with classification of subpopulations based on multi-marker positivity of individual cells. Part of the analysis is a high-quality tumor stroma separation based on the CK signal. The single-cell readouts were used to construct spatial biomarker-expression patterns.
About the speaker:
Dr. Rognoni joined Definiens in 2015 and is currently leading the development of multiplex IHC image analysis (bright field and multispectral immunofluorescence) used in service projects. Before working in the field of digital pathology, he studied physics at the FU Berlin and LMU Munich and specialized in biophysics during his Ph.D. and postdoctoral research at TU Munich.