Background Tumor-infiltrating immune cells play an important role against cancer and are critical to controlling tumor growth and spread. Immunotherapy drugs such as immune checkpoint inhibitors, despite inducing long-term responses in many cancer types, remain ineffective for a majority of patients. A better understanding of the tumor microenvironment (TME), and the characterization of populations of immune cells, their activation status, spatial distribution, and relationship, along with cytokine signaling, may help to better stratify patients and explain mechanisms of resistance to immunotherapy.

Veracyte’s Brightplex^® is a chromogenic multiplex immunohistochemistry (IHC) technology that leverages digital pathology enhanced by artificial intelligence. It allows the detection of up to eight biomarkers on a single FFPE slide to identify cellular subpopulations. However, to characterize the exact role of an immune cell it is often necessary to determine if it expresses soluble proteins which cannot be detected by IHC, such as cytokines or activation factors. In that case, the detection of the corresponding RNA transcripts by in situ hybridization (ISH) can be used as a substitute for protein detection.

Methods Here, we propose a multiplex technology automated on the Leica Bond RX platform which combines ISH and IHC staining on a single FFPE tissue section. In brief, that tissue section is sequentially stained to detect biomarkers of interest either with antibodies for proteins or nucleic probes for transcripts. Each round of staining is followed by the digitization of the slide. Whole slide images are fused to create a virtual multi-channel image where biomarkers are detected by digital pathology and combined to identify immune cell populations. The spatial distribution and cell-to-cell interactions within a slide or between multiple adjacent slides are assessed by combining multiplex ISH/IHC panels.

Results This technology allowed us to quantify different sub-populations of tumor-infiltrating-lymphocytes (activated, cytotoxic and exhausted), interferon-γ producing cells, and tumor-associated-macrophages expressing the chemokine CXCL9. We investigated the spatial distribution of these immune cells within the TME. We also assessed the cell-to-cell proximity between these populations to assess their interactions.

Conclusions Integrated into an Immunogram, an analytics platform that integrates multi-omics datasets from Veracyte Biopharma Atlas, this new tool could be a powerful solution to decipher the tumor landscape and predict response to immunotherapy and patient outcome.