Background
Tumor immune contexture (TIC) provides key information to predict the clinical outcome or to understand the mechanism of action of a drug, particularly in the context of immunotherapy. Quantitative approaches assessing cell density of several cell types are commonly used to evaluate the complexity of TIC. Although informative, this only partially reveals all the information present in the analyzed image. To get the missing valuable information, we developed a spatial analysis pipeline to qualitatively assess the pattern of infiltration and interaction of different cell subsets.
Methods
Based on our Brightplex® technology, FFPE clinical samples were sequentially stained with different biomarkers to identify several immune cell types and regions of interest. This includes the use of RNA in situ hybridization technology to assess the expression of cytokines such as IL-10 in regulatory T cells or macrophages. Importantly, we were able not only to analyze multiple cell subsets present on one slide (intra-slide) but also to combine with other cell subsets on a consecutive slide (inter-slide). The presented case study was realized on Urothelial Bladder cancer from PEMBIB clinical trial (NCT02856425).
Results
In this case study, T-cell infiltration exhibited an excluded pattern at baseline. Treatment did not impact this pattern, however, we observed that Natural Killer (NK) and CD8 T cells get closer from each other and also the development of interacting cell networks involving these two cell types after treatment. Such structures have been described in the literature as key for the response to Immune Checkpoint Inhibitors. This underlies the importance of performing this type of analysis.
Conclusions
The pipeline provides insightful information on pattern of T-cell infiltration in the tumor, T-cell interaction with both tumor cells or NK cells. In the presented case study, we were able to identify NK-CD8 T-cell networks, especially in post-treatment. Thus, spatial analysis allows the identification of new key specific features that may be predictive of the clinical outcome or unravels the mechanism of action of a drug. This type of analysis can be performed as well at a cohort level to assess the spatial evolution of TIC.