Mathematical model for delayed responses in immune checkpoint blockades

TL;DR

This paper presents a mathematical model using ODEs to simulate delayed immune responses in checkpoint blockade therapy, highlighting the delicate parameter conditions needed for such delays and their implications for tumor immune evasion.

Contribution

The study introduces a novel ODE-based model that captures delayed immune responses in checkpoint blockade therapy, emphasizing the importance of precise parameter calibration.

Findings

Delayed responses occur in a narrow parameter space.

Short-lived anti-tumour T cell storms can break the delay.

Tumors may survive in hostile immune environments.

Abstract

We introduce a set of ordinary differential equations (ODE) that qualitatively reproduces delayed responses observed in immune checkpoint blockade therapy (e.g. anti-CTLA-4 Ipilimumab). This type of immunotherapy has been at the forefront of novel and promising cancer treatments over the past decade and was recognised by the 2018 Nobel Prize in Medicine. Our model describes the competition between effector T cells and non-effector T cells in a tumour. By calibrating a small subset of parameters that control immune checkpoint expression along with the patient's immune-system cancer readiness, our model is able to simulate either a complete absence of patient response to treatment, a quick anti-tumour T cell response (within days) or a delayed response (within months). Notably, the parameter space that generates a delayed response is thin and must be carefully calibrated, reflecting the observation that a small subset of patients experience such reactions to checkpoint blockade therapies. Finally, simulations predict that the anti-tumour T cell storm that breaks the delay is very short-lived compared to the length of time the cancer is able to stay suppressed. This suggests the tumour may subsist off an environment hostile to effector T cells; however, these cells are -- at rare times -- able to break through the tumour immunosuppressive defences to neutralise the tumour for a prolonged period. Our simulations aim to qualitatively describe the delayed response phenomenon without making precise fits to particular datasets, which are limited. It is our hope that our foundational model will stimulate further interest within the immunology modelling field.