Tracing specificity of immune landscape remodeling associated with distinct anticancer treatments

Summary

Immune cells within the tumor microenvironment impact cancer progression, resistance, response to treatments. Despite remarkable outcomes for some cancer patients, immunotherapies remain unsatisfactory for others. Here, we designed an experimental setting using the Alb-R26 ^Met “inside-out” mouse model, faithfully recapitulating molecular features of liver cancer patients, to explore the effects of distinct anticancer targeted therapies on the tumor immune landscape. Using two treatments in clinical trials for different cancer types, Decitabine and MEK+BCL-XL blockage, we show their capability to trigger tumor regression in Alb-R26 ^Met mice and to superimpose distinct profiles of immune cell types and immune-checkpoints, impacting immunotherapy response. A machine learning approach processing tumor imaging and immune profile data identified a putative signature predicting tumor treatment response in mice and patients. Outcomes exemplify how the tumor immune microenvironment is differentially reshaped by distinct anticancer agents and highlight the importance of measuring its modulation during treatment to optimize oncotherapy and immunotherapy combinations.

Graphical abstract

Highlights

• •

  Alb-R26 ^Met liver cancer model recapitulates patient molecular and immune profiles

• •

  The distinct action of anticancer drugs on tumor immunity highlights its plasticity

• •

  Distinct anticancer agents differentially reshape the tumor immune landscape

• •

  Distinct treatments partially flatten immune heterogeneity of pre-treated tumors

Abstract

Microenvironment; Immune response; Cancer; Machine learning