Cancer communications and causing traffic jams
Blog posted by: Martin Christlieb
Francesca Buffa has been awarded a five-year grant by the European Research Council (ERC) to develop computer simulations to personalise cancer treatment and prevent drug resistance.
We increasingly understand which parts of a cell have gone wrong and make it divide uncontrollably, leading to cancer. We are also becoming better and better at developing drugs which match the bits that have gone wrong.
However, there are complications. The cancer in two individuals may have different mutations making each tumour grow. Since our new drugs are so minutely targeted we will need a different drug for the two patients.
A single tumour might have three mutations driving it. In some breast cancers the tumour may be driven by a combination of three mutations, snappily called PTEN, p53 and EGFR, so each patient needs a combination of up to three drugs – one for each mutation.
Finally, when we give our drugs in real clinics, cancers often become resistant and the patients relapse.
We have lots of drugs or potential drugs, but before we use them we need to prove that they work and work well. The problem is that setting up multiple clinical trials testing each combination of drugs in different patients, each carrying a different set of mutations takes thousands of patients, millions of pounds, and takes decades.
But, what if a computer could conduct the trials in the relatively cheap safety of its processors? Francesca is trying to do just that.
Inside each cell is a network of paths that the cell uses to shuttle signals. A bit like the road network allows cars to shuttle round a town. If one set of traffic lights is permanently set to green, too many cars will pass through and a traffic jam may result somewhere else. In cancer some paths are permanently on resulting in uncontrolled cell multiplication.
With a smart drug the broken always-green traffic light can be set to red stopping the flow and halting the cancer growth.
Rush-hour drivers and cancer cells will not tolerate this for long before they discover the rat-runs that get round the red-light.
We could block each rat-run, but a computer with knowledge of the roads and the traffic lights, might be able to predict which rat-runs the drivers will use and predict which other light to turn red to block the whole thing from the start.
In the cancer clinic this would mean that we pick the correct drug(s) to block the mutated communication point(s), and add a drug which blocks the rat-runs the cancer pathways will try to use to get round our block. This would prevent the tumours becoming resistant … or at least slow the process down.