An immunologic portrait of cancer

Abstract

The advent of high-throughput technology challenges the traditional histopathological classification of cancer, and proposes new taxonomies derived from global transcriptional patterns. Although most of these molecular re-classifications did not endure the test of time, they provided bulk of new information that can reframe our understanding of human cancer biology. Here, we focus on an immunologic interpretation of cancer that segregates oncogenic processes independent from their tissue derivation into at least two categories of which one bears the footprints of immune activation. Several observations describe a cancer phenotype where the expression of interferon stimulated genes and immune effector mechanisms reflect patterns commonly observed during the inflammatory response against pathogens, which leads to elimination of infected cells. As these signatures are observed in growing cancers, they are not sufficient to entirely clear the organism of neoplastic cells but they sustain, as in chronic infections, a self-perpetuating inflammatory process. Yet, several studies determined an association between this inflammatory status and a favorable natural history of the disease or a better responsiveness to cancer immune therapy. Moreover, these signatures overlap with those observed during immune-mediated cancer rejection and, more broadly, immune-mediated tissue-specific destruction in other immune pathologies. Thus, a discussion concerning this cancer phenotype is warranted as it remains unknown why it occurs in immune competent hosts. It also remains uncertain whether a genetically determined response of the host to its own cancer, the genetic makeup of the neoplastic process or a combination of both drives the inflammatory process. Here we reflect on commonalities and discrepancies among studies and on the genetic or somatic conditions that may cause this schism in cancer behavior.

Figure 1

Transcriptional profiling of 114 melanoma metastases; A) self organizing heat map displaying selected IRF-1 related transcripts together with the expression of MITF and melanoma differentiation antigens. Each column represents a melanoma metastasis. The dendrogram indicates the degree of similarity among genes (rows) or melanoma samples (columns) using Pearson's correlation coefficient. B) Matrix similarity based on Pearson correlation for the same genes. C) Self organizing map of IRF transcripts with related genes that have been observed to be expressed in tumors displaying better prognosis [23,24,45] and expressed during immune-mediated tissue-specific destruction [9,10]. Highlighted is the IFN-γ/STAT-1/IRF-1/IL-15 cluster associated with a Th1 type of immune response. The yellow dashed lines underline the close relationship between the expression of IFN-γ and IRF-1 in the tumor microenvironment.

Figure 2

Functional gene network analysis based on 349 genes derived from a Student's t test comparing relapse free and relapse patients affected by breast cancer [24]. Ingenuity pathway analysis of canonical pathways over-expressed in relapse free breast cancer patients compared to patients who did not relapse identified 46 pathways significantly affected. Immune function genes involved in interferon signaling associated with allograft rejection, autoimmune reaction, antigen presentation, B cell development and NK Cell Signaling ranked at the top and were up-regulated in lesions of breast cancer patients who endured a relapse-free survival.