Building on Discoveries in Cancer Genomics

Building on Opportunities in Cancer Research

Scientists recognized that it would be necessary to decipher the genomes of many cancers to understand the extent of their complexity and diversity. This understanding led the NCI to launch two large research programs that have undertaken the comprehensive analysis of the DNA and RNA in approximately 10,000 tumors from more than 30 types of cancer. These programs are The Cancer Genome Atlas (TCGA) for adult tumors and the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) Initiative for childhood cancers.

TCGA is a joint project of the NCI and the National Human Genome Research Institute (NHGRI). This comprehensive program, made possible by advances in sequencing technologies beyond those used to sequence the human genome, has resulted in substantial progress in understanding the biology of cancer and has led to new approaches to cancer diagnosis and treatment.

TARGET is a TCGA-like effort in children’s cancers that is managed by the NCI. Genomic technologies are being used to search for therapeutic targets in five cancers that are common in, but not always exclusive among, children: acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML),neuroblastoma, osteosarcoma, and high-risk Wilms tumor.

Opportunities in Research - Building on Discoveries in Cancer Genomics

TCGA and TARGET have brought together several hundred investigators to work on various cancer projects, which also required the development of an efficient infrastructure for organizing the many steps involved in processing and sequencing the DNA in each tumor and storing the large amount of generated data in a readily retrievable format for analysis.

The technical resources, reagents, and personnel for these programs represent a considerable investment by the NCI and the NHGRI. Support for these large, ambitious programs was greatly facilitated by the increased funding associated with the American Recovery and Reinvestment Act, which led to faster progress in completing the analyses of the various tumors.

To build on this momentum, the extensive information developed by TCGA and TARGET is stored in databanks that can be accessed and analyzed by the entire cancer research community, even before papers arising from the data have been published.

Analysis of the tumors in TCGA and TARGET has made it possible to organize each tumor type, often for the first time, into subsets that share particular genetic and epigenetic changes. The latter type of changes, unlike genetic changes, are potentially reversible, which makes it possible to consider treatment to re-express epigenetically silenced genes in the tumor or to silence genes that have been aberrantly activated epigenetically.

The targeting of some of the identified genetic abnormalities can be evaluated by using drugs that have already been approved by the Food and Drug Administration or experimental drugs that are inclinical trials. In addition, there are efforts to develop specific inhibitors to target other identified abnormalities, once they have been validated as key drivers of the tumor types in which they are found.

Anatomical illustration of a human stomach showing some of the features associated with each of the four molecular subtypes of gastric cancer — TCGA analysis of stomach tumors identified Epstein-Barr virus (EBV) as the probable cause of a distinct subset of stomach cancers and suggested new approaches for its treatment. This illustration lists some of the features associated with each of the four molecular subtypes of gastric cancer.

In one example, TCGA investigators reported that many squamous cell lung cancers have abnormalities in enzymes called protein kinases for which experimental drugs are in development. This finding has led the NCI to develop the Lung-MAP trial, a master protocol for patients with this form of lung cancer, who will receive experimental targeted treatments as determined by the molecular abnormalities present in their tumors.

In another example, TCGA analysis of stomach tumors identified Epstein-Barr virus (EBV) as the probable cause of a distinct subset of stomach cancers and suggested new approaches for its treatment. EBV had been identified previously in some stomach cancers, but the significance of this finding remained uncertain. However, TCGA analysis indicated that a wide range of tumor suppressor genes are epigenetically silenced in EBV-positive stomach cancers and, in addition, that most of the tumors have mutations in a particular protein kinase for which experimental inhibitors are currently in clinical trials. It should therefore be possible to test whether reactivating these silenced tumor suppressor genes and/or inhibiting the mutated protein kinase can help patients with this cancer.

In acute myeloid leukemia, TCGA investigators identified at least one key mutation in every case, a finding with both short- and long-term clinical implications. In glioblastoma multiforme, an aggressive form of brain cancer, reactivation of the tumor was found to occur through epigenetic changes, an observation with potential implications for preventing reactivation.

Although it is recognized that cancers at a given organ site may have several subtypes, some important characteristics may be shared among cancers that arise at different sites. For example, TCGA researchers identified four genomic-based subtypes of endometrial cancer and, in addition, uncovered important similarities between endometrial, ovarian, and breast cancers.

Recognizing the value of comparing genomic data from diverse types of cancer, TCGA investigators developed a formal project for cross-tumor analysis called the Pan-Cancer project. This effort has brought together more than 250 collaborators from 30 institutions to analyze the same dataset. It is leading to a deeper appreciation of features common to several cancer types. Some of the results point to potential new uses for existing drugs based on shared molecular targets across cancer types.

Recognizing the value of comparing genomic data from diverse types of cancer, TCGA investigators developed a formal project for cross-tumor analysis called the Pan-Cancer project. This effort has brought together more than 250 collaborators from 30 institutions to analyze the same dataset.

A recent analysis of 12 cancers by TCGA investigators identified 11 major subtypes based on their molecular profiles. Although some types were specific to their organ site of origin, others, such as squamous cell lung, head and neck, and a subset of bladder cancers were found to have many shared molecular characteristics between organ sites. Such analyses are leading to a potentially new classification of tumors according to their molecular abnormalities, which would go beyond the traditional histologic classification according to the site of origin. A molecular classification may have therapeutic implications in addition to highlighting similar pathogenetic features between cancers of different origins. Finding that a particular targeted treatment is beneficial in one form of cancer may indicate that it could also be clinically useful in other tumor types that share similar molecular abnormalities.