Pan-Cancer Initiative Finds Patterns of Drivers

The first set of results from a large collaboration comparing genomic data across 12 tumor types, published in 18 papers across several journals this fall, provides numerous early demonstrations of how such analyses may aid in identifying new patterns of drivers for cancer.

Studies in the Pan-Cancer Initiative drew on data from about 5,000 tumors gathered by The Cancer Genome Atlas (TCGA). Launched as an informal TCGA collaboration in October 2012, the Pan-Cancer program has grown to more than 250 researchers at dozens of institutions.

Project goals were to put together coherent TCGA data sets across tumor types and platforms and to study, among other questions, the extent to which different types of tumors are driven by common abnormalities across tissue types and the extent to which they are driven by factors specific to their tissue of origin.

The researchers are sifting through the pooled data sets, which include protein and phosphoprotein abundance, DNA methylation, copy number, single-nucleotide and structural variants by whole-exome sequencing, microRNA sequencing, and RNA sequencing and gene expression data.

Among early Pan-Cancer analyses, one published in Nature examined point mutations and small insertions/deletions and identified 127 significantly mutated genes across the 12 tumor types (Nature 2013;502:333–9). Most tumor types have two to six mutations, suggesting a relatively small number of driver mutations. Additionally, “mutations in transcriptional factors/regulators show tissue specificity, whereas histone modifiers are often mutated across several cancer types,” the study notes.

A study in Nature Genetics that looked at somatic copy number alterations (SCNA) discovered 140 regions containing significantly recurrent focal SCNAs across all tumor lineages (Nature Gen 2013;45:1134–40). Among these 140 regions, 102 have no known oncogene or tumor suppressor gene targets and 50 have significantly mutated genes. Amplified regions without known oncogenes were enriched for genes involved in epigenetic regulation.

“Analysis of the molecular aberrations and their functional roles across tumor types will teach us how to extend therapies effective in one cancer type to others with a similar genomic profile,” the Pan-Cancer researchers noted in a Nature Genetics commentary introducing the initiative (Nature Gen 2013;45:1113–20).

The next round of Pan-Cancer analyses will draw on an expanded TCGA data set covering 10,000 tumors from 25 tumor types, including whole-genome sequencing of tumors from 1,000 patients and their corresponding normal genomes, says Joshua Stuart, PhD, professor of biomolecular engineering at the University of California, Santa Cruz, and senior author of the commentary. TCGA sample collection for this expanded round will be completed in about a year, if all goes as planned.

Stuart and his Pan-Cancer colleagues also are working with the International Cancer Genome Consortium to pool their data for a total of 2,000 tumors analyzed with whole-genome sequencing. These full sequences may reveal a greater role for phenomena such as viral integration and mobile elements in the genome, says Stuart.

For more news on cancer research, visit Cancer Discovery online at http://CDnews.aacrjournals.org.