Genomic analysis of 33 cancer types completed

This image shows the uncontrolled growth of cells in squamous cell carcinoma, a common form of skin cancer.Markus Schober and Elaine Fuchs, The Rockefeller University

Cancer is caused by changes to DNA that affect the way cells grow and divide. There are at least 200 forms of cancer, with many subtypes. Identifying the changes in each cancer’s complete set of DNA—its genome—and understanding how these changes interact to drive the disease will lay the foundation for improving cancer prevention, early detection, and tailored treatments.

The Cancer Genome Atlas (TCGA) was launched in 2005 by NIH’s National Human Genome Research Institute (NHGRI) and National Cancer Institute (NCI) to map the key genomic changes in 33 types of cancer. The multi-institution collaboration focused not only on cancer genome sequencing, but also on different types of molecular data collection and analysis, such as investigating gene and protein expression profiles (when they are turned on or off) and associating them with clinical and imaging data. With over $300 million in total funding, the project involved more than 150 researchers at more than two dozen institutions.

The PanCancer Atlas sums up the work accomplished by TCGA in a collection of 27 papers across a suite of Cell journals. Three summary papers published on April 5, 2018, recap the core findings, and companion papers report more in-depth explorations.

The first summary paper describes a technique called molecular clustering that groups cancers based on their molecular characteristics rather than their tissue of origin. The scientists analyzed gene expression, DNA modifications, protein expression, and other data from about 10,000 tissue samples representing 33 different types of cancer. The team identified 28 distinct clusters based on molecular similarities. Although most of these clusters could be linked to tissue of origin, many contained different cancer types. The most diverse group had 25 cancer types. These findings could help guide the treatment of many cancer patients whose tumors are of unknown origin.

The second paper presents findings on oncogenesis, the processes that lead to cancer development and progression. The authors focused on three critical oncogenic processes: the DNA mutations that drive cancers; the influence of DNA alterations on gene and protein expression; and the interplay of tumors with their surroundings, particularly immune cells. The results will help in the development of new treatments for a wide range of cancers.

The final paper details genomic alterations in 10 key signaling pathways that control the stages of the cell’s life cycle, growth, and death. The researchers found that 89% of tumors had at least one significant alteration in these pathways. About 57% of tumors had at least one alteration that could be targeted with currently known drugs and 30% had multiple targetable alterations. These findings will help researchers explore treatments with more tailored approaches, such as using a combination of drugs to target multiple pathways at the same time.

“TCGA was the first project of its scale to characterize—at the molecular level—cancer across a breadth of cancer types,” says Dr. Carolyn Hutter, NHGRI team lead for TCGA. “At the project’s infancy 10 years ago, it wasn’t even possible, much less on such a scale, to do the types of characterization and analysis that were being proposed. It was a hugely ambitious project.”

“The PanCancer Atlas effort complements the over 30 tumor-specific papers that have been published by TCGA in the last decade and expands upon earlier pan-cancer work that was published in 2013,” says Dr. Jean Claude Zenklusen, director of the TCGA Program Office at NCI.