Mathematical modelling in oncology: A heterogeneous subject

Research output: PhD ThesisPhd-Thesis - Research and graduation internal

Abstract

Mathematics can be applied in many different fields, oncology being no exception. Mathematical models require data to be constructed, tested and validated, which makes their applications largely driven by opportunity. Furthermore, different hypotheses require different mathematical models. As such, many different kinds of mathematical models can be used depending on the available data and proposed research question. In this work, three distinct aspects of mathematical oncology are studied as a result.
In the first part, intra-tumour heterogeneity (ITH) is studied. High ITH is well known to correlate to worse patient prognosis in many cancer types. However, ITH does not have a single definition, as it can be measured through the genome, proteome and the tumour microenvironment. In this work, a measure for chromosomal copy-number ITH (CNH) is constructed. CNH is shown to outperform other ITH-based biomarkers for predicting patient survival across cancer types. Furthermore, a subset of breast cancer patients with low CNH are identified that could benefit from less intensive treatment.
In the second part, mathematical models for tissue dynamics are established. In particular, a mathematical model combined with experimental data shows that the healthy endocrine pancreas does not contain stem cells. A different mathematical model applied to colorectal cancer xenografts shows that colorectal tumour stem cells are defined by their environment.
In the third part, mathematical models, in particular machine learning models, are used to predict complications and identify risk factors for colorectal cancer resections by exploiting a large clinical audit containing 60,000+ patients.
Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • University of Amsterdam
Supervisors/Advisors
  • Vermeulen, Louis, Supervisor
  • Miedema, Daniël M., Co-supervisor
Award date21 Nov 2022
Print ISBNs9789464196221
Publication statusPublished - 2022

Cite this