Towards a consensus definition of immune exclusion in cancer.

Background: The immune cell topography of solid tumors has been increasingly recognized as an important predictive factor for progression of disease and response to immunotherapy. The distribution pattern of immune cells in solid tumors is commonly classified into three categories - namely, "Immune inflamed", "Immune desert" and "Immune excluded" - which, to some degree, connect immune cell presence and positioning within the tumor microenvironment to anti-tumor activity. Materials and methods: In this review, we look at the ways immune exclusion has been defined in published literature and identify opportunities to develop consistent, quantifiable definitions, which in turn, will allow better determination of the underlying mechanisms that span cancer types and, ultimately, aid in the development of treatments to target these mechanisms. Results: The definitions of tumor immune phenotypes, especially immune exclusion, have largely been conceptual. The existing literature lacks in consistency when it comes to practically defining immune exclusion, and there is no consensus on a definition. Majority of the definitions use somewhat arbitrary cut-offs in an attempt to place each tumor into a distinct phenotypic category. Tumor heterogeneity is often not accounted for, which limits the practical application of a definition. Conclusions: We have identified two key issues in existing definitions of immune exclusion, establishing clinically relevant cut-offs within the spectrum of immune cell infiltration as well as tumor heterogeneity. We propose an approach to overcome these limitations, by reporting the degree of immune cell infiltration, tying cut-offs to clinically meaningful outcome measures, maximizing the number of regions of a tumor that are analyzed and reporting the degree of heterogeneity. This will allow for a consensus practical definition for operationalizing this categorization into clinical trial and signal-seeking endpoints.