At the annual meeting of the American Society for Clinical Oncology in Chicago this week, Memorial Sloan-Kettering Cancer Center's Douglas Levine chaired a session on the clinical implications of findings from The Cancer Genome Atlas. He gave the oncologists an overview of TCGA and what kinds of tumor samples it solicits — primary, newly diagnosed, and untreated cancer samples, accompanied by germline DNA as a control. Once the samples go through quality control, they're sent out for sequencing and genomic characterization, Levine said, before the data is sent to a repository to be accessed by various research groups for analysis. The better the technology gets, the more analyses are run, Levine added. Currently, the samples are assayed using many different platforms, including SNP or CNV testing, methylation testing, analysis of microRNAs and messenger RNAs, mutation analysis, and, in some cases, whole-exome and whole-genome sequencing.
TCGA data has already shown that even though some of the same genes are implicated in different cancers, there are different mechanisms of pathway activation and different point mutations, requiring different treatment strategies, Levine said.
But clinicians will likely find that the best use of this data is in combining it with known clinical variables. "You can take clinical variables and overlay them with genomic signatures. ... You really can affect the prognosis of different groups," Levine said. "We have the clinical data available and some good genomic signatures, and at the end of the day, if we can combine them," we can better treat patients. He added that the "genomic landscape from TCGA and other sources will form the backbone for clinical trial design in the future."
Paul Spellman of Oregon Health and Science University added that with sequencing becoming less expensive, the goal is to do whole-genome sequencing of all samples by 2014 as well as methylation profiles of entire cancer genomes.