No evidence of immunosurveillance in mutation-hotspot driven clonal haematopoiesis

Abstract

The theory of immunosurveillance posits that T-cells can selectively eliminate clones harbouring non-self antigens generated by somatic mutations. There is considerable evidence supporting the role of immune surveillance in cancer. Whether immunosurveillance imposes a negative selective pressure on pre-cancerous clones, however, is not well established. Here, we studied the association between MHC-variant binding and risk of clonal haematopoiesis (CH), a pre-cancer state in the blood driven by expansions of mutant haematopoietic stem cells (HSCs). We predicted MHC binding affinity towards 40 known CH hotspot variants in 380,000 UK Biobank participants, and examined the relationship between predicted binding to each variant and risk of its expansion in the blood. Despite being well powered to detect subtle differences in selective pressure, we did not find associations between predicted MHC binding and CH prevalence for any of the hotspot variants. In individuals in whom we identified CH, there was no relationship between predicted binding affinity to the variant and size of the clone. Overall, we do not find evidence for the MHC genotype to be a factor that affects which somatic variants expand in CH, suggesting a limited role for immunosurveillance in shaping the genetic diversity of the blood.