The genetic mapping of the tumors yielded other discoveries. For example, the glioblastomas' genomes revealed a defect in a gene labeled IDH1. It turned up in 12 percent of all the tumors, but in nearly half of those from patients younger than 35. Glioblastoma patients with this IDH1 defect also seemed to survive longer than patients without that mutation.
"We never would have identified it without the genome-wide analysis," said Dr. Victor Velculescu, a Hopkins professor and another co-author of the study in Science.
"In the near future, it's likely that information regarding the IDH1 alteration will be useful in disease management," he said. Patients with the defect "could be isolated and treated with different therapies."
The researchers also said many of the genetic defects could soon be detectable in genetic material released into the blood stream by dying tumor cells.
"It will be possible soon to detect many of them from patient samples - in their blood - even when tumors are relatively early," Vogelstein said. "More tumors would be curable if they can be caught early."
With more sensitive technology in the near future, he said, "it could allow blood tests to be a reality. I don't think that's any longer science fiction."
The discoveries also reveal just how complex tumor development really is. They suggest that genetic mutations - some perhaps inherited, and some acquired over decades of random errors during cell division or from environmental assaults, such as smoking - accumulate until they finally join forces in a way that touches off tumor development.
That suggests why people with genetic vulnerabilities or unhealthy habits don't always develop cancer, or develop it late in life. It may also point to opportunities for doctors to screen patients for precursor genetic defects, and tackle the cancer before it becomes lethal.
"What these studies will do is change the way people think about controlling cancer in view of this complexity," Vogelstein said. "We hope that studies like these that examine the genome in depth will stimulate people to think - 'Is this the best way to control cancer?' Or perhaps to try to think of ways ... not as dramatic as curing the patient, but to get them before [the cancer starts]. Every cancer detected early is potentially a life saved."
The federally funded Genome Atlas group reported in Nature that they worked with more donated brain tumors (206), but sequenced for fewer genes (601) than the Hopkins group.
Their sequencing turned up three significant tumor cell mutations that had not been reported before. Patients with these defects could receive individually tailored therapies. The study also revealed three biological pathways that appeared to be disrupted in more than three-quarters of the glioblastomas studied. These, too, may offer targets for future therapies and drugs aimed at all tumor types sharing the same pathway.
"This represents another major step towards our major ultimate goal of using information about the human genome to improve human health," said Dr. Alan E. Guttmacher, acting director of the National Human Genome Research Institute.
