Last week, researchers reported the results of the first attempt to identify the genetic alterations involved in two common cancers using the tools developed to sequence the human genome.

They identified 189 mutated genes in tumors from patients with breast or colon cancers. The vast majority of genes had not been linked to cancer previously.

The team, led by researchers at the Johns Hopkins Kimmel Cancer Center, analyzed 13,000 genes in 11 colon tumors and 11 breast tumors. Each tumor had 90 genetic mutations on average, of which 11 to 17 mutations may have been critical in causing cancer.

Different sets of genes were mutated in colon tumors and breast tumors; the types of mutations were also different. Patients with the same cancer also tended to have different types of mutations, the researchers reported online September 7 in Science.

"These results may explain the clinical diversity that physicians have been talking about for a long time," says co-leader Dr. Victor Velculescu of Johns Hopkins. "The genetic diversity of cancer may underlie the clinical diversity."

But the genetic diversity may not be as great as it now appears, he adds, because some of the mutated genes may participate in the same cellular processes, or pathways.

If this can be confirmed in follow-up studies, the hope is that drugs could be designed to intervene at a point in the pathway that may benefit patients with various related mutations.

Drs. Bert Vogelstein and Kenneth Kinzler of Johns Hopkins co-led the study, which cost about $5 million and was funded in part by NCI. In recent years, their laboratories have screened families of genes linked to cancer, and the new study expands the scope.

"We realized from our earlier work that to find cancer genes you really have to look for them in an unbiased way," says Dr. Velculescu. A genome-wide screen makes no assumptions about which genes are important and often yields surprises.

Mutations that were identified in the initial set of 22 tumors were subsequently assessed in 48 additional tumors. Mutations that occurred frequently were scrutinized for possible roles in cancer.

The 13,000 genes are the best-studied of the roughly 20,000 genes in the human genome. As with all genome projects, the results represent a beginning rather than an end.

"To make the findings most useful, we need to integrate them with other types of data, such as epigenetic changes and chromosomal anomalies," says co-author Dr. Will Parsons of Johns Hopkins.

The results, he adds, "show how much remains to be learned about cancer and also the importance of doing these types of studies for different cancers."

The sequences of cancer genomes are ultimately obtainable, the researchers conclude. The methods, they note, will improve as more sequencing data are accumulated through The Cancer Genome Atlas (TCGA) Pilot Project and other efforts now under way.

Leaders of TCGA said last week in a statement that the Science paper was "proof of principle" of the potential benefits of large-scale mutational screens for cancer.

"The strategy works," says Dr. Sanford Markowitz of Case Western Reserve University, a leader of the Science study and longtime collaborator with the Hopkins group. "But I wouldn’t want anyone to think we have finished with breast cancer and colon cancer.

"There is still very important work to been done on these diseases and on the cancer genome," he says.

By Edward R. Winstead