While researchers still need to determine which of these mutations are true drivers of metastasis and which are merely carrier mutations that don't affect the cells, they have identified some interesting candidates. For example, the patient had normal versions of a gene called CTNNA1, which has been linked to cells' ability to stick to each other. But both tumors samples had a large deletion knocking out both copies of the gene, an occurrence that was particularly common in the metastatic cells. This mutation might allow cancer cells to break free of the primary tumor and spread through the bloodstream.

In addition to studying tumor samples from the patient, researchers implanted some of her tumor tissue into a mouse with a compromised immune system. This approach, called a xenograft, is often used to study the properties of human cancers. Just as in the patient, the cancer cells quickly multiplied and spread. When the team sequenced DNA from these cells, they found it had a similar genetic profile to that of the metastatic tumor samples. "It was a big surprise to see so many similarities between the xenograft and the metastatic genome," says Elaine Mardis, codirector of the Genome Center. Both cancers originated from the same primary tumor and seemed to evolve in similar ways, despite growing in completely different environments. "This is just one case and we need to study more, but this does look like an interesting model for studying metastatic cancer," she says. If the findings are confirmed more broadly, drug developers can use this system to test new treatments on human tumor cells, knowing that the cells behave similarly in the mouse and in the human body.

Mardis, Wilson, and others aim to sequence hundreds of cancer genomes over the next year. "The capacity for sequencing instruments has been on a dramatic uptick," says Mardis. "The biggest challenge now is, how do you do multigenome analysis, for example comparing 20 to 50 genomes at the same time from a carefully defined phenotype like drug resistance?" The researchers hope that studying this volume of DNA will give broader insight into cancer genomics, letting them identify key metabolic or signaling pathways that are affected in cancer and which might be good targets for new therapeutics.