Improving patient outcome by integrating the generic with the personal

Cancer is one of the top two healthcare challenges: 1 in 3 people will have cancer in their lifetime. To translate that to the ideas factory that generated this proposal, about 30 people attended the ideas factory in various capacities, therefore about 10 of them will have cancer during their lifetime.
While many advances have been made in cancer treatment, there are still ways in which therapy can be improved. It is, for example, usually treated by combinations of surgery, radiotherapy, and/or chemotherapy, but the precise interaction of these treatments and the ways in which different people react to them is poorly understood. This makes it virtually impossible to prescribe the best therapeutic strategy for an individual patient.
The challenge addressed in this project is to build a framework in which to view disease through a personalised lens of predictive modelling, in order to improve future combination therapy planning. We propose to do this through an unprecedented multidisciplinary project: mathematics-led, but drawing on our expertise in biology, physics, and computer science. Our project reflects the structure of life through a stratified, multi-scale description which deals with the important parts, e.g. the organ containing the tumour, in great detail, whilst describing the remainder of the whole in a more chunked way, able to efficiently capture the essence of the necessary detail. Our longer term goal is for our modelling framework to be generic, and adaptable to a range of diseases and combined therapies. In this project, a generally adaptable framework and the associated interconnected mathematical and computational models and methods will be created. Having been validated by biological experiments, these models will be refined and populated with data to provide clinically useful predictions for our exemplar, combined chemo/radiotherapy of glioma (a type of brain tumour).
Each component of the project draws on specific expertise provided by the investigators:
- three-dimensional, spatially-resolved mathematical models of drug delivery and tumour growth, coupling mass transport with cell response, and simulated using fast computational algorithms, provides a detailed, patient-specific, representation of response to therapy;
- radiation interaction modelling, with associated algorithms to speed-up accurate radiation therapy planning, provides details of the influence of radiotherapy;
- experimental cell biology work, delivers data with which to validate the models;
- mathematical modelling and supporting synergy experimentation integrates whole-body effects with disease- and person-specific models;
- process algebra modelling of signalling inside and between cells, bystander effects, and metastasis, provides models of cell response.
These will all have scientific outputs, but where the project really reaps the benefits of multi-disciplinarity is at the interfaces of these work packages, and through the combination of our joint approaches to problems.
Through this project we will lay the foundations for our 20-year goal of a generic framework for combined therapies by addressing a specific and important example: combined radiation/drug therapies for glioma.

The project is very amenable to becoming an outreach vehicle capable of demonstrating the public benefit of mathematics in a visual way. This will be exploited through a variety of social media (e.g. animations showing the spatio-temporal variation of drug and radiation delivery at the local and patient scales, delivered via YouTube) and more traditional forms of engagement (e.g. web presence, presentations to local schools and at Science festivals).

Both the nature of the problem tackled, and the generalisable mathematical framework used to approach the problem, give substantial routes to impact.

Academic:
The proposed framework is general and straightforwardly extensible to other cancers, e.g. the mathematical and computational models will be easily generalised to other cancers (the mathematical framework remains unchanged, whatever the cancer). At all stages we will build the models to be modular and as generic as possible, so that new case studies can be introduced simply. The new algorithms developed will benefit academics whose research requires rapid methods for removing statistical noise or the inversion of linear systems of equations (present in many scientific and engineering problems) and should ultimately become embedded in scientific and engineering software toolkits.
The biological experimentation protocols developed can be transferred into other systems with minimal experimentation to obtain patient-specific rate data. The models and software developed will be made available to others, e.g. through the Leeds software repository, the GEANT 4 collaboration, and through the project website. Long term, the increased capability of desktop computing will put the modelling developed in this project on the desktop of practising clinicians, for use in real-time consideration of patient options.
For other (non-cancer) diseases, more work will be required to develop appropriate component models (the part equivalent to the tumour model here); however, the overarching framework, with its novel approach to combination therapy, will be applicable in other disease domains with appropriate re-parameterisation, e.g. infectious diseases are often treated with multiple anti-microbials.

Economic and Societal:
A recent Oxford University report (Nov 2012) calculated that cancer costs the UK economy approx. £15bn per year, in terms not just of healthcare but economic losses due to time off work (for patients and their carers), and early deaths.
Increasingly, the emphasis in cancer care is on combination therapies, despite there being no tools for rationally designing trials (e.g. when should the drug be administered relative to the radiation?), let alone for the proper planning of combination therapies in a manner akin to clinical radiotherapy planning. Engaging with clinicians, we see our exemplar acting as a beacon for the transformation to a unified, optimised, synergistic planning of patient therapies. Furthermore, given the high mortality of glioma, clinicians are willing to engage in new trials schedules.

Our goal is to model full therapy delivery in a unified way, with both dosing and timing being optimised for each individual patient, thus maximising the effectiveness of the intervention synergy achieved by the use of multiple therapeutics.

For patients, we offer:
- the ability to stratify and eventually personalise intervention;
- better outcomes from combined therapy;
- steps towards clinical trials.

For clinicians, we offer:
- a therapy planning tool;
- the chance to influence its development through a steering committee of stakeholders.

For the general (healthy) public, we offer:
- opportunities to raise awareness of potential of mathematics and technology in healthcare (through our website, YouTube, science festivals, public lectures, schools visits);
- an important target disease, to captivate their imagination.

Our unified approach will allow us to roll out the framework to other cancers and diseases with comparative ease, and apply to disease-specific charities for further project funding.