A neoantigen fitness model predicts tumour response to checkpoint blockade immunotherapy

Area of Science:

• Immunology
• Computational Biology
• Oncology

Background:

• Checkpoint blockade immunotherapies harness the host immune system to eliminate tumor cells.
• Clinical efficacy of these therapies correlates with T-cell recognition of tumor-specific neoantigens.
• Understanding neoantigen interactions is key to predicting and improving immunotherapy outcomes.

Purpose of the Study:

• To develop a novel tumor fitness model for predicting response to cancer immunotherapy.
• To identify key factors influencing neoantigen fitness and T-cell recognition.
• To explore the evolutionary similarities between tumors and pathogens.

Main Methods:

• Developed a computational model integrating major histocompatibility complex (MHC) binding affinity and sequence similarity to assess neoantigen fitness.
• Evaluated tumor fitness based on the weighted effect of dominant neoantigens across tumor subclones.
• Validated the model's predictive power using patient data from melanoma and lung cancer immunotherapy trials.

Main Results:

• The developed immune fitness model accurately predicts patient survival in anti-CTLA-4 and anti-PD-1 treated cohorts.
• Identified specific neoantigen characteristics that determine their fitness and impact on immunotherapy response.
• Demonstrated that tumor evolution shares similarities with the evolutionary dynamics of rapidly evolving pathogens.

Conclusions:

• The neoantigen fitness model provides a powerful tool for predicting immunotherapy efficacy.
• Low-fitness neoantigens represent potential targets for novel immunotherapy development.
• This work deepens our understanding of tumor-immune interactions and evolutionary principles in cancer.