Treatments for prostate cancer, including surgery, radiotherapy and androgen deprivation therapy, can lead to variable and often temporary results, mainly due to the heterogeneity of cancer cells. The non-cancerous portion of the tumour microenvironment, such as immune and stromal cells, is key for tumour development, and being a genetically stable target holds the potential for both diagnosis and treatment. Improving our knowledge on the molecular interactions between cancer and non-cancer cells could unlock such potential.
An observational study was designed to identify driver molecular interactions among cancer, stromal and infiltrating immune cells. Such cell types were enriched from biopsies of patients across pathological stages, and their transcriptome was profiled. A novel statistical model allowed the identification of unexpected microenvironmental molecular drivers of several hallmarks in prostate cancer. For example, dramatic transcriptional changes in t cells and fibroblasts allow a sustained hormonal and lipid unbalance in the tumour core; while cancer cells suppress key microenvironmental modulators, of which some novel in prostate cancer.
Furthermore, an innovative computational tool will be presented, which enables large-scale studies of tissue composition and immune cell infiltration on public and novel RNA sequencing data sets. Such tool allowed drawing a landscape of the associations between tumour relapse and abundance of infiltrating cells, for the whole TCGA data base.