Genome-wide association analysis identifies genetic correlates of immune infiltrates in solid tumors

• DOI: 10.1371/journal.pone.0179726
• Publication Date: 2017-07-27T13:51:28Z
• Authors: Nathan O. Siemers; James L. Holloway; Han Chang; Scott D. Chasalow; Petra B. Ross-Mac...; Charles F. Voliva; Joseph D. Szustak...
• Keywords: Immune checkpoint

Therapeutic options for the treatment of an increasing variety cancers have been expanded by introduction a new class drugs, commonly referred to as checkpoint blocking agents, that target host immune system positively modulate anti-tumor response. Although efficacy these agents has linked pre-existing level tumor infiltrate, it remains unclear why some patients exhibit deep and durable responses while others do not benefit. To examine influence genetics on state, we interrogated relationship between somatic mutation copy number alteration with infiltration levels 7 cell types across 40 cohorts in The Cancer Genome Atlas. Levels cytotoxic T, regulatory total natural killer, B cells, well monocytes M2 macrophages, were estimated using novel set transcriptional signatures designed resist interference from cellular heterogeneity tumors. Tumor mutational load estimates purity included our association models adjust biases multi-modal genomic data. Copy alterations, mutations summarized at gene level, position-specific evaluated infiltration. We observed strong loss large region chromosome 9p decreased lymphocyte melanoma, pancreatic, head/neck cancers. Mutations oncogenes PIK3CA, FGFR3, RAS/RAF family members, suppressor TP53, changes infiltration, usually restricted types. Associations specific WNT/beta-catenin pathway genetic state limited, but noted link expression WNT receptor FZD3, suggesting there are interactions pathways. Finally, two different death regulators, CASP8 DIDO1, often mutated tumors had higher infiltrates. In summary, study supports relevance questions resistance blockade therapies. It also highlights need assess genome-wide influences during exploration any hypothesized be relevant therapeutic Some links like loss, may response cancer Others, pathways, help guide combination approaches.