Multiscale modelling of intestinal crypt organization and carcinogenesis

TL;DR

This paper reviews multiscale mathematical models of intestinal crypts and colorectal carcinogenesis, highlighting their predictions, challenges, and potential for reducing computational costs in understanding cancer development.

Contribution

It provides a comprehensive overview of mathematical approaches to multiscale crypt modelling, emphasizing validated predictions and discussing methods to simplify complex models.

Findings

Validated model predictions of crypt behaviour

Identification of key multiscale modelling challenges

Discussion of coarse-grained model derivations

Abstract

Colorectal cancers are the third most common type of cancer. They originate from intestinal crypts, glands that descend from the intestinal lumen into the underlying connective tissue. Normal crypts are thought to exist in a dynamic equilibrium where the rate of cell production at the base of a crypt is matched by that of loss at the top. Understanding how genetic alterations accumulate and proceed to disrupt this dynamic equilibrium is fundamental to understanding the origins of colorectal cancer. Colorectal cancer emerges from the interaction of biological processes that span several spatial scales, from mutations that cause inappropriate intracellular responses to changes at the cell/tissue level, such as uncontrolled proliferation and altered motility and adhesion. Multiscale mathematical modelling can provide insight into the spatiotemporal organisation of such a complex, highly regulated and dynamic system. Moreover, the aforementioned challenges are inherent to the multiscale modelling of biological tissue more generally. In this review we describe the mathematical approaches that have been applied to investigate multiscale aspects of crypt behaviour, highlighting a number of model predictions that have since been validated experimentally. We also discuss some of the key mathematical and computational challenges associated with the multiscale modelling approach. We conclude by discussing recent efforts to derive coarse-grained descriptions of such models, which may offer one way of reducing the computational cost of simulation by leveraging well-established tools of mathematical analysis to address key problems in multiscale modelling.