One finding in particular by the Johns Hopkins group shows the value of the genome-wide approach. Victor Velculescu, who led the Hopkins glioblastoma study, and his colleagues discovered that a mutation in one gene differentiates one subset of glioblastomas from another in a disease that researchers had always believed was quite homogeneous. The gene, called IDH1, had never before been implicated in any cancer. But the IDH1 mutation occurred in 12 percent of glioblastoma patients, and those people were, on average, 20 years younger and survived significantly longer than patients without the mutation. This finding--perhaps the most instantly clinically relevant piece of the three studies released today--is one that the scientists hope could soon be used to help physicians better predict their patients' survival. The finding could also help clinicians determine if existing therapies might be more effective on this brand of glioblastoma and ultimately help create treatments directed specifically at the IDH1 pathway.

Cancer researchers welcome the flood of data gleaned from both approaches. "I'm just glad the information is in the till," says Paul Mischel, a neuropathologist at the University of California, Los Angeles, who specializes in glioblastoma therapy development and application. "These studies provide the first really well-delineated set of road maps." Chin and Velculescu hope that sequencing costs will soon drop low enough to allow them to combine the two techniques, sequencing large numbers of genes in many tumors.

The studies have also revealed to scientists looking to treat these diseases just how difficult their challenge really is. "For the first time, these are giving you the complete picture of these two cancer types," Velculescu says. "This is important, because if we ever want to cure cancer, we have to know what's wrong with it. And unfortunately, what appears to be wrong with most cancers is more complicated than we may have anticipated."