Optimisation of locally advanced MSI-H/dMMR colorectal cancer treatment with neoadjuvant pembrolizumab using data-driven delay integro-differential equations

Abstract

with neoadjuvant pembrolizumab using data-driven delay integro-differential equations Authors: ['Gunilla Håwi' 'Peter Kim' 'Peter P. Lee'] Colorectal cancer (CRC) poses a major public health challenge due to its increasing prevalence, particularly among younger populations, with it being the third most commonly diagnosed cancer and the second leading cause of cancer-related deaths worldwide as of 2024. Microsatellite instability-high (MSI-H) CRC and deficient mismatch repair (dMMR) CRC comprise 15% of all CRC, and have shown remarkable response to immunotherapy, especially with PD-1 inhibitors. Despite this, there is still a significant need to optimise immunotherapeutic regimens to maximise clinical efficacy and patient quality of life, whilst minimising monetary costs. Mathematical models are a promising avenue for treatment optimisation, and we use delay integro-differential equations to model the interactions and dynamics of various cell types, immune checkpoints, damage-associated molecular patterns (DAMPs), and cytokines. We provide deterministic equations for the concentrations of cancer cells, DAMPs, CD4+ and CD8+ T cells, dendritic cells (DCs), cancer-associated fibroblasts (CAFs), neutrophils, macrophages, NK cells and various cytokines of pro-inflammatory (IL-2, IFN-γ, TNF, IL-12, IL-1β) and immunosuppressive (TGF-β, IL-10, IL-4, IL-6) nature. We consider these cells and their phenotypes, noting that many of these cell types are being modelled deterministically for the first time in cancer, paving the way for a deeper understanding of the complex underlying immune dynamics. We consider two compartments: the tumour site and the tumour-draining lymph node (TDLN), taking into account phenomena such as DC migration, T cell proliferation, acquired resistance to pembrolizumab, and CD8+ T cell exhaustion and reinvigoration. The parameter values and initial conditions for the model are estimated from experimental data, incorporating various pharmacokinetic, bioanalytical, and radiographic studies, as well as deconvolution of bulk RNA-sequencing data from the TCGA COADREAD database.