Open access worldwide to the new database, based on genome studies, is expected to help researchers accelerate development of new drugs and better match patients with therapies, NCI said in a statement on Monday.
"Most anti-cancer drugs that are used today are used based on their empirical activity," Dr. Yves Pommier, chief of the NCI's Laboratory of Molecular Pharmacology, said in an interview. "For most of them, we know there is a target, but they have not been connected with any genomics."
Most cancer treatments involve a lot of guess work because doctors have no way to determine how a particular patient is likely to respond to many commonly used drugs or chemotherapy, or which cancers will develop resistance.
To create the database, the NCI team sequenced 60 human cancer cell lines, generating an extensive list of cancer-specific variations for different parts of the body.
The results were published in Cancer Research, a journal of the American Association for Cancer Research.
Researchers could mine the data, for instance, to determine whether the chemotherapy drug Cisplatin is associated with specific genetic mutations, Pommier said.
"Only about half of women with ovarian cancer respond to it," he said, noting that pharmaceutical companies would have little incentive to determine if an existing cancer drug should only be used in a subset of patients.
Many recently approved cancer drugs are targeted treatments, designed to block specific pathways that cancer cells use to grow and reproduce. Before the drugs are administered, patients are tested for the specific genetic mutations that would make the drug more likely to be beneficial to them.
Melanoma drug Zelboraf, sold by Roche Holding AG, is designed to work by targeting a specific genetic mutation found in about half of all melanomas. Pfizer Inc's cancer drug Xalkori, which targets a mutation in the ALK gene, works in about 4 percent of lung cancer patients.