WHEN her two sisters died from an aggressive form of breast cancer, Gail Walters was anxious not to meet the same fate. She scheduled surgery to have both of her breasts removed. Walters (not her real name) had been identified as being in a "BRCA2 family". In other words, she was likely to carry a mutation of the BRCA2 gene that would make her susceptible to breastcancer.
As luck would have it, a few weeks before her mastectomy, Mike Stratton at the Institute of Cancer Research in Sutton, Surrey, managed to sequenceBRCA2. This meant that Walters could be screened - and quickly - for specific variants of the gene. Although both her sisters had the harmful variant, Walters did not, so her risk of developing breast cancer was no higher than average and thesurgery was cancelled.
Translating research into action that helps patients doesn't always happen so quickly or dramatically, but Walters's tale shows how cancer research has shifted gear. Cancer is no longer seen as a uniform disease. Instead, researchers are focusing on the molecular abnormalities within a particular patient's cancer cells, often by exploiting the vast amount of informationgenerated by the Human Genome Project.
"Before the HGP we had no idea where the genes were," says Stratton's colleague, Colin Cooper, who leads the section that made the discovery. "Now all we have to do is look it up on the internet." Out of the 30,000 genes in a single cell, five or six go wrong to cause cancer, so the job of a cancer researcher is no longer one of gene identification, but geneselection, he says.
It doesn't end at the selection process, of course. You also need to identify important interactions. Identifying protein interactions helped Jason Carroll at the Cancer Research UK Cambridge Research Institute to discover how breast cancers become resistant to the drug tamoxifen and to identify patients who are unlikely to respond to the treatment. "When you use the power of thehuman genome we can learn a lot about why treatments work and why they fail," he says.
Looking at the interaction between proteins and the genome is a relatively new approach to cancer treatment. "It turns out that the human genome is much more complex than we thought," says Carroll. "We knew what genes were involved, but we didn't know where the switches were that turn the genes on and off."Until recently, he explains, scientists had been looking for these switches at "promoter regions" biologically upstream of the genes involved in tumour growth. "Everyone assumed that was where the control switches were. Now people are seeing that these genes can be switched on and off from a great distance. In fact, we've found them in places that used to be called 'junk DNA'," Carroll adds.
Thisknowledge has led to a host of potential new cancer treatments - and with them, lots of new drugs to test. And while you can have a good idea about which drugs a patient may respond to through their genetic profile, checking out the reality relies on developments in another area - imaging.
"If you have good imaging techniques, you can know very quickly if your treatment is having a successfulpatient response," says Kevin Brindle, a biochemist at the University of Cambridge.
Traditionally, doctors and researchers assessed the success of a treatment by looking at CAT or MRI scans, which allow you visualise the internal structure of the body to see if tumours had shrunk. Now we're seeing more use of functional imaging techniques such as fMRI, says Brindle. Tumours use up a great deal ofglucose, and fMRI is good at detecting the rate of glucose uptake through increased blood flow at a particular site. A change in this activity is often obvious way before you see anatomical changes in the tissue, Brindle says.
These insights into human genomics are also changing the skills needed from a typical cancer researcher. The rapid pace of change has left many feeling ill-equipped to deal...