Elaine Mardis and Richard Wilson: Taking Cancer's Genetic Measure

It's a faster way to get to "personalized medicine," tailoring treatment to each person's genes.


In 2006, only three years after helping to decode the human genome, Elaine Mardis and Richard Wilson were running out of money. A grant from the National Institutes of Health to look for genetic mutations associated with the deadly blood cancer acute myeloid leukemia was winding down.

The biochemists, who together direct the genome sequencing center at Washington University School of Medicine in St. Louis, had an inspiration. They had been looking only at the parts of the genome where they would most likely find AML-related genes. They realized it would be more productive to decode the complete genomes of an AML patient's normal cells and cancer cells and compare them; the differences would reveal the mutations. The NIH said no: too complicated, too many unknowns, and too expensive.

Mardis and Wilson found a funding angel, and in just eight months, working with Washington University oncologist Timothy Ley, they had used an AML patient's normal skin cells and cancer cells in her bone marrow to identify 10 genetic mutations that had set the stage for her cancer. "These [mutated genes] are the ones that you are going to want to go after," says Mardis.

They plan to sequence 150 more cancer patients by early next year, focusing on leukemias and cancers of the breast and lung. Pharmaceutical labs can then design molecules for specific combinations of genetic mutations that switch off the cancer process. If your genes fit one of the combinations, you get the matching drug—that is, if a pharmaceutical company has developed it.

"We may find after lots and lots of tumor sequencing that every tumor is different," says Mardis. "And that really invokes this extreme personalized medicine approach." But will drug companies pay to pursue it? "It concerns me a lot," says Wilson.

James Watson, who shared a Nobel prize for modeling DNA, is a fan of their research. "The money is well spent," he says. But he has personal cause to think so. Watson learned last year from his own decoded genome that a mutation undercut the effect of his blood pressure drug. A different medication worked much better.