The cancers to be studied first in the pilot phase of The Cancer Genome Atlas (TCGA) project will be brain, lung, and ovarian, officials from NCI and the National Human Genome Research Institute (NHGRI) announced on September 13.

The Institutes launched the pilot phase of TCGA in December 2005 to assess the feasibility of systematically identifying the significant genomic changes involved in cancer using contemporary genomic analysis technologies, including resequencing of selected genes.

The selection of brain, lung, and ovarian cancers, which collectively account for more than 210,000 cases annually in the U. S., followed a rigorous assessment of tumor collections that began over a year ago.

"The three cancer types were selected based on specific scientific, technical, and bioethical criteria," said NCI Deputy Director Dr. Anna Barker. "The biospecimen collections representing brain, lung, and ovary met or exceeded these criteria and represent high-mortality diseases that present significant challenges to the cancer research community."

For example, each collection had to exceed a certain number of samples to achieve statistical power; contain a defined percentage of tumor cells; and have matched normal tissue, outcomes data on patients, and appropriate informed patient consent, to name a few of the criteria.

"We were looking for biospecimen sets of the highest possible quality so that we could have confidence in the results," said Dr. Carolyn Compton, director of NCI's Office of Biorepositories and Biospecimen Research.

Three institutions are donating the biospecimens for the TCGA Pilot Project.

The lung cancer samples are from the Lung Cancer Tissue Bank of the Cancer and Leukemia Group B clinical trials group, housed at Brigham and Women's Hospital; the brain cancer samples (glioblastoma multiforme grade 4) are from the University of Texas M.D. Anderson Cancer Center; and the ovarian cancer samples are from the Gynecologic Oncology Group tissue bank at the Children's Hospital of Ohio State University.

The tissue samples will be stored and processed at the TCGA Biospecimen Core Resource (BCR) facility. The International Genomics Consortium, in collaboration with the Translational Genomics Research Institute, of Phoenix, Ariz., will establish and manage the BCR.

When fully operational, the TCGA Pilot Project will have two groups of centers doing different types of genomic analyses on tissue samples provided by the BCR.

The Cancer Genome Characterization Centers, funded by NCI, will identify changes such as chromosomal rearrangements and the presence of extra copies of genes. Genes of interest will then be resequenced at Genome Sequencing Centers, funded by NHGRI.

The information generated by the centers will be integrated and managed by a Principal Bioinformatics Resource (PBR) and made available to researchers over the Internet. Data are expected to be accessible through this TCGA database by mid-2007.

"The variety of data will provide a view of the cancer problem that we have never had before," noted NHGRI Director Dr. Francis Collins.

During the pilot project, tumor samples will be assessed for alterations in DNA expression, gene regulation (including epigenetic changes that alter gene activity but not DNA sequence), gene copy numbers, and other changes. These data, along with patient outcomes and other clinical data, will be accessed through specific portals that are currently being defined.

"The revolutionary aspect of TCGA is the integration of all of these different types of genomic, biologic, and clinical data," said Dr. Barker. "TCGA promises to open up a whole world of cancer genomics that will reach far beyond these three cancers."