Researchers at the Center Systems Biology in Stuttgart are developing multi-scale models of tumour growth with the aim of predicting how drugs spread and disperse in the tissue. The simulations can also take into account potential effects of pharmacological compounds and irradiation. The method will benefit researchers and clinicians by assisting them in their efforts to develop more efficient therapies.
The Center Systems Biology’s (CSB) activities related to research into tumour growth are embedded in the “FORSYS – Research Units of Systems Biology” initiative funded by the German Federal Ministry of Education and Research (BMBF), which also involves other institutes at the University of Stuttgart. The CSB works with partners from the Universities of Tübingen and Magdeburg, the Dr. Margarete Fischer-Bosch Institute of Clinical Pharmacology (IKP) at the Robert Bosch Hospital in Stuttgart and Bayer as the industrial partner in the project “A Systems Biological Approach to Predictive Cancer Therapy”, which is a major pillar of tumour research in Germany. The project was set up to endeavour to simulate systematically and across several scales the effect of drugs on tumour growth. “Our multi-scale simulations cover everything from the intracellular level to the vascular systems of tumours; other project partners apply systems biology approaches to a broad range of different organs,” said Prof. Dr. Dr. h. c. Matthias Reuss, Senior Director of the CSB in Stuttgart.
The researchers from Stuttgart use their simulations to address a specific point in time in tumour development. When tumours have reached a few millimetres in size, cells within the tumour are deprived of nutrients and oxygen and die. Cells on the periphery of the tumour, which are supplied with sufficient amounts of oxygen and nutrients from the surrounding tissue, continue to multiply. This leads to a temporary equilibrium situation during which virtually as many cells die as new ones are produced. “Nutrient and oxygen deficiency causes the tumour cells to produce signalling substances (e.g. growth factors) which diffuse from the tumour into nearby blood vessels where they trigger angiogenesis, i.e. the formation of new blood vessels in the tumour,” said Reuss. The tumour is thus provided with nutrients and oxygen and able to grow and spread. In addition to supplying the tumour with oxygen and nutrients, the new blood vessels are also potential targets for anti-tumour drugs. With their simulations, the researchers from Stuttgart hope to improve existing knowledge on tumour growth and dynamics and investigate the efficiency of drugs in reaching their targets inside the human body. Drug molecules, which can have many different sizes and structures, encounter numerous obstacles on their way to their target; not all of them are able to penetrate the cell membrane, and if they do, components of the extracellular matrix might cause them to deviate from their normal path.
In order to simulate all processes on the cell and tissue level, Holger Perfahl from Professor Reuss’ group has developed special software in cooperation with international experts from Oxford, Nottingham and Barcelona. The software also permits data visualisation. Perfahl had the perfect qualifications for the project – a degree in mathematics, a degree in mechanical engineering and the desire to study biological relationships in greater detail. He uses a method known as agent-based simulation (or agent-based model; ABM) for his work. ABMs are used to simulate actions and interactions of and between autonomous agents in order to assess their effects on the system as a whole. “This is similar to the behaviour of a flock of birds or a shoal of fish, which is determined by the interactions that occur between the individual animals and their neighbours,” said Reuss. Perfahl added: “These models enable us to define individual cells as well as groups of cells as agents. It depends on the issue we are dealing with and the scale on which the processes are simulated as to whether we use individual or collective entities. On the intracellular level, we simulate relevant signalling and regulatory pathways; on the tissue level, we focus mainly on the distribution of drug molecules. We have defined interfaces on individual scales which help us simulate the system as a whole.”
The researchers are not only focused on predicting the effect of individual pharmaceuticals, but are also able to simulate the effect of combination therapies involving several drugs and the combination of chemo- and radiotherapy on destroying tumours. “We are able to combine the effects of anti-angiogenic drugs and the effects of radiation doses. The simulated effects are always verified experimentally; iterative cycles involving mathematical modelling and experiments help us to increase the accuracy of the models,” said Perfahl.
A particularly innovative approach is to combine virtual calculations with 3D images of real tumours, which the researchers from Stuttgart obtain from a group of researchers led by Prof. Dr. Bernd Pichler from the Department of Radiology at the University Hospital of Tübingen. The virtual simulations will be directly integrated into the 3D images. “We can create a kind of cyberspace which enables us to examine the effect of drugs and radiotherapy even more authentically,” Reuss said. The new approach is currently been tested using mice as models. As the tumour develops, 3D images are taken at regular intervals and compared with simulations that provided the initial 3D images and computer data. “The 3D images do not yet have the resolution we would like them to have and we are currently working on optimising our models. In order to investigate extensive tumour areas, we digitalise histological sections obtained from tumour biopsies and superimpose the data over the grid of the 3D images,” said Perfahl. The combination of three things, i.e. histological sections, 3D images and computer models, is an important step on that way towards computer-based therapies.
One of the goals of the computer simulations is to help reduce undesired drug effects. Reuss: “Exact models of the mechanism of action provide us with information on drug doses that are low but effective. Relatively small drug doses would initially lead to the partial destruction of the tumour, i.e. create “holes” in the tumour tissue. These holes could then be used to administer some more drugs, which would lead to the destruction of larger parts of the tumour.” Although the researchers are still a long way off such a computer-based therapy, researchers from many different disciplines are working together to move closer to the goal. The lack of sufficient computer capacity is another major bottleneck the researchers from Stuttgart are faced with. “At present, the standard simulation of a tumour with an edge length of one millimetre requires around eight days to complete,” said Perfahl. However, the researchers now have access to the new super computer Hermit, which is located directly next door in the Stuttgart HLRS (High Performance Computing Centre). Perfahl is currently adapting the simulation software to Hermit’s parallel computer architecture with the aim of reducing computing time in the future.
The systems biologists from Stuttgart are part of a new BMBF-funded project in which they are hoping to tackle a new challenge: they are planning to take into account different tumour geometries and individual patient differences. “The new project, which is called ‘Holistic multi-scale modelling of targeted protein therapeutic action: towards predicting effective treatment of cancer’ enables us to continue our FORSYS work. From 2013 onwards, we plan to take into account the huge variability of tumours and initially specifically concentrate on the intestines,” said Reuss. As intestinal tumours do not spread evenly into all directions, they are excellent models that will enable the researchers to integrate the tumours’ different geometries and structures into their simulations.
Further information:University of Stuttgart Center Systems Biology CSBProf. Dr.-Ing. Dr. h. c. Matthias ReussHolger PerfahlNobelstr. 1570569 StuttgartTel.: +49 (0)711/ 685 - 64573E-mail: matthias.reuss(at)ibvt.uni-stuttgart.de holger.perfahl(at)ibvt.uni-stuttgart.de