Computational Oncology: Exploring the origin of colon cancer

In the framework of the Mathematics in Oncology initiative, we have used computer simulations to model biological processes behind the initial steps of colon cancer development to unravel the otherwise invisible cancer formation process. The results have now been published in the Special Issue “Frontiers in quantitative cancer modeling” of the journal “Computational and Systems Oncology”.


The human colon consists of several million cells which, like other highly proliferating cells in our body, divide rapidly. During this replication process, mutations can occur that may alter the behavior of the cell. Certain types of mutations enable the mutated cell to divide more frequently than its neighboring cells without this mutation. After some time, the mutated cells overgrow the non-mutated cells in the colon because of their rapidly increasing number. The mutated cells might even take over an entire colonic crypt (also referred to as intestinal gland), which is a collection of a few thousand cells in the colon wall. Those single mutated crypts might then develop into a polyp or adenoma (precancerous tissue formations that are looked for in colonoscopy) or even a manifest cancer.

Current research suggests that specific mutated crypts are the origin of colon cancer. Therefore, analyzing these mutational processes within a crypt is key for understanding early cancer development and may have significant implications for cancer prevention. However, as the colonic crypts are too small and too numerous, scientists cannot directly observe these processes in humans.

In an interdisciplinary group of scientists, we have now built a computational model to simulate these mutational processes within a crypt on a computer. These computer simulations allow to analyze if and how fast different so-called driver mutations which play a key role in cancer formation take over a crypt. We focused in our work on the Lynch syndrome, a genetic condition which leads to an increased risk of developing colon cancer during lifetime. The research was conducted in the framework of the “Mathematics in Oncology” initiative including scientists from Heidelberg University, HITS, Leipzig University and the University Hospital Heidelberg. The paper (doi: 10.1002/cso2.1020) has now been published in the Special Issue “Frontiers in quantitative cancer modeling” of the journal Computational and Systems Oncology. Saskia Hauptfrom EMCL has also presented those results at the online SMB Meeting (13-17 June 2021), the annual meeting of the Society for Mathematical Biology (see also News item).

Cancer research: A computational approach

Building on existing approaches which are already used for simulating healthy colon tissue, we incorporated recent biomedical data and clinical observations into our model to make the computer simulations for the crypts as realistic as possible. In our paper, we quantified each mutation’s potential to take over entire crypts for different types of mutations and investigated how the mutation spread is influenced by the cell location within the crypt or the stem cell dynamics.

The now conducted simulations show that simulated driver mutations in an active stem cell almost always take over the entire crypt within a few weeks. Depending on the location of the mutated cell, there are different possibilities to spread throughout the crypt, either in a top-down or in a bottom-up process (see video). We showed in our simulations that both scenarios are theoretically possible.

Current studies indicate that there is always one active stem cell populating a crypt which is, after some time, replaced by a neighboring stem cell. In simulations, we observed the following scenario: If a mutated stem cell is replaced by a non-mutated stem cell, it is possible that the previous mutation does not take over the crypt or at least that the time of spread is prolonged. This means that stem cell exchange can restore the integrity of crypts and contribute to the elimination of specific mutations which are of particular importance for individuals with Lynch syndrome. In other words, a stem cell exchange might explain why some mutated crypts do not progress further to cancer and support the hypothesis of spontaneous regression of precancerous lesions in Lynch syndrome.

In general, the predictions of the time span required for taking over a crypt carrying certain mutations provide the basis for future studies. In particular, we will address the time required for a mutated crypt to become an endoscopically visible lesion. Further, we will analyze how long it usually takes until a mutation occurs in a cell of a crypt. This is essential to obtain estimates for the duration of the whole cancer formation process. With these findings, we want to support tailored treatment approaches for Lynch syndrome cancer patients and prevention strategies for cancer-free Lynch syndrome carriers.

Link to the paper:

HITS press release: English / German