Table 1.

Explaining differential responses to immunotherapy across patient populations

Sources of variationExamplesPractical impact
Somatic neoantigensTumor mutational load and mutations that affect antigen presentation (MHC and antigen processing)
  • Patient selection

  • Vaccines (neoantigen or shared antigens for low mutational load tumors)

Somatic driver mutationsDriver mutations that correlate with tumor immune infiltration
  • Patient selection

  • Targeted therapies

  • Drug combination selection

  • New drug target identification

Inflammatory status/phenotype of tumor
  • Checkpoint ligand expression on tumor or TIL (e.g., PD-L1, PD-L2)

  • Expression of checkpoint molecules on T effector cells in TILs vs. periphery (immunotherapy na├»ve and immunotherapy failures/nonresponders)

  • Differential expression of molecules in Treg cells in TILs vs. periphery

  • Innate immune cell infiltrate (e.g., NK, macrophage, dendritic cells)

  • Patient selection

  • Drug combination selection (e.g., PD-1 pathway blockade + anti-LAG-3, TIGIT, TIM-3, GITR, OX40, CD137, ICOS, CSF1R, KIR2DL)

  • New drug target identification

Patient immune status including the microbiomeFactors impacting immune response to cancer [age, history of infection, vaccination, and microbiome(s)]
  • Patient selection

  • New drug target identification

Germline allelic variationPolymorphisms in target and ligand(s), target pathway genes, FcgR
  • Patient selection

  • Drug combination selection

  • New drug target identification