Half of all men and a third of all women in the United States will develop cancer in their lifetimes. Each year about 1.5 million adults in the U.S. receive a cancer diagnosis, and 600,000 adults die from their disease.1 Worldwide, annual estimates of incidence include 13 million new cancer cases diagnosed in 2008 with 7.6 million deaths. By 2030, the global burden is expected to surpass 21 million new cancer cases and 13 million deaths.2 The impact of cancer on global healthcare costs was recently estimated to be $895 billion annually, greater than any other cause of death, including cardiovascular disease.3 Clearly, cancer as a disease has staggering costs in terms of lives affected and economic impact.
Cancer is a genetic disease, arising from mistakes or alterations in DNA. Although some cancers have strong associations with inherited alleles, the majority of cancer arises from acquired DNA damage to tissues and organs, and is not heritable. This type of acquired DNA damage is referred to as somatic mutation.
What are the somatic mutations that lead to cancer? This question has important implications for diagnosis, monitoring, and treatment. Although four decades of cancer biology research has led to a functional understanding of many key genes involved in cancer, a genome-wide view of the somatic mutations prevalent in cancer is only now emerging from large-scale genome sequencing studies enabled by advances in next generation sequencing (NGS) technologies.4,5
In these studies, tumor-associated somatic mutations are identified by comparing the sequence of the tumor to that of normal tissue from the same patient, in order to exclude inherited sequence polymorphisms. In many cancer types, the tumor genome may contain tens and even hundreds of somatic alterations. These include single-nucleotide variants as well as short insertions and deletions—together, the alterations typically referred to as mutations. However, the tumor genome also contains changes in gene copy number and chromosomal translocations, as well as aneuploidy, and changes in global methylation status. Considering the myriad of alterations in the tumor genome, how can scientists determine the key aberrations leading to cancer?
One straightforward, heuristic approach to identifying relevant cancer genes from somatic mutation profiles involves characterizing the pattern of recurrent alterations across a large population. For the purposes of this article, the discussion will focus on mutations, but similar approaches apply to other types of aberrations. As pointed out by Vogelstein, the patterns of mutations in well-studied cancer genes are highly characteristic and nonrandom.6 For example, genes that promote cancer growth and proliferation often contain “hotspot” missense mutations at specific locations in the gene, and relatively few mutations that inactivate the gene.
Examples of hotspot mutations in key cancer genes include BRAF V600E, EGFR L858R, and PIK3CA H1047R. Notably, these mutant cancer genes are the targets of highly specific cancer drugs either approved or in development. Interestingly, hotspots have been observed in many other genes functionally implicated in cancer (e.g., ERBB2, KIT, KRAS) as well as genes not previously known to have a role in cancer, such as the RNA splicing factors U2AF1 and SF3B1. Therefore, the heuristic approach identifies not only genes that were known from functional studies to be key cancer genes but also new candidate cancer genes, thus expanding the opportunities for cancer drug development.
In contrast to the hotspot cancer genes, genes whose normal function is to suppress inappropriate cell growth, such as the tumor suppressors and growth regulators TP53, PTEN, and RB1, have a different pattern of mutations. The mutations in these tumor suppressor genes tend to be deleterious in nature and include nonsense or frame-shift mutations that lead to premature protein truncation. Deleterious mutations have been observed to be enriched in genes that were not previously appreciated to have a role in suppressing cancer, such as the chromatin regulators ARID1A, PBRM1, and SETD2, thus broadening our understanding of the cellular pathways deregulated during oncogenesis. Although tumor suppressor genes are not typically good drug targets, their inactivation may influence the activity of other targeted therapies.
In addition to research involving heuristic approaches to mine somatic mutation data for relevant cancer genes, other researchers have developed increasingly sophisticated statistical approaches to identify significantly mutated genes in cancer. These approaches, including MuSiC and MutSig, correct the gene specific mutation rate with the background mutation rate.7,8 Interestingly, the background mutation rate can differ based on the type of mutation (transition or transversion), the sequence context, gene size, gene expression level, and whether the gene is replicated early or late during mitosis. These methods are in continuous development, and as yet there is no widely adopted standard. However, consideration of several aspects of mutational heterogeneity improves the specificity in identification of significantly mutated genes while reducing the nomination of frequently mutated genes that likely have no contribution to oncogenesis.
Full exome somatic mutation profiles are currently available for several thousand tumors, derived from The Cancer Genome Atlas program and an increasing number of independent academic and international efforts. A general conclusion from these studies is that each cancer type has several genes that are frequently aberrant and many more that are infrequently altered. Interesting similarities have been noted in the somatic mutation profiles of different cancer types that are relevant for treatment strategies.9 Furthermore, somatic mutations that predict response to targeted therapies are observed at low prevalence across many cancer types. Therefore, a “pan-cancer” analysis is relevant to understanding the somatic mutation profile of cancer.10
In summary, a comprehensive somatic mutation profile of the tumor genome from many cancer types is now emerging. With this information in hand, pharmaceutical companies are designing and developing the next generation of breakthrough cancer therapies targeted to key somatic alterations. There are many targeted therapy success stories that have radically improved outcomes for previously hard-to-treat cancer indications, but many more are needed. Diagnostics companies are developing the next generation of tests that will reveal the relevant somatic alterations in a cancer patient’s tumor. In this area sequential single gene testing has been the norm, but the complexity of the somatic mutation profile of most tumors will require multiplex NGS testing platforms to simultaneously measure all relevant somatic alterations given limited available tumor tissue.
The last unmet need in this cancer treatment ecosystem is an information system that comprehensively identifies the drugs, clinical evidence, and clinical trials relevant to an individual patient’s somatic mutation profile. The demand for this information system is fueled by the unexpectedly rapid deployment of NGS testing platforms to clinical laboratories. This “actionability” knowledge base will prove as challenging and as important as the advances that have been made to accelerate genome sequencing over the last 15 years, since the sequencing of the first human genome. Given the tremendous toll cancer exacts on our families, our society, and our healthcare system, progress cannot come too soon.
Seth Sadis, Paul Williams, and Nick Khazanov are R&D scientists with Life Technologies Corporation. They support the Oncomine Platform, an oncology bioinformatics platform that includes Oncomine NGS Power Tools and Compendia Bioscience services.
References
- American Cancer Society. http://www.cancer.org/cancer/cancerbasics/what-is-cancer. Accessed December 31, 2013.
- American Cancer Society. Global Cancer. Facts and Figures. 2nd Edition. http://www.cancer.org/acs/groups/content/@epidemiologysurveilance/documents/document/acspc-027766.pdf. Accessed December 31, 2013.
- American Cancer Society. The global economic cost of cancer. http://www.cancer.org/acs/groups/content/@internationalaffairs/documents/document/acspc-026203.pdf. Accessed December 31, 2013.
- Vogelstein B, Papadopoulos N, Velculescu VE, et al. Cancer genome landscapes. Science. 2013;339(6127):1546-1558.
- Chmielecki J, Meyerson M. DNA sequencing of cancer: what have we learned? Annu Rev Med. 2013 Nov 20. [Epub ahead of print].
- Shyr D, Liu Q. Next generation sequencing in cancer research and clinical application. Biol Proced Online. 2013;15(1):4.doi: 10.1186/1480-9222-15-4.
- Dees ND, Zhang Q, Kandoth C, et al. MuSiC: identifying mutational significance in cancer genomes. Genome Res. 2012; 22(8):1589-1598.
- Lawrence MS, Stojanov P, Polak P, et al. Mutational heterogeneity in cancer and the search for new cancer-associated genes. Nature. 2013; 499(7457):214-218.
- The Cancer Genome Atlas Network. Comprehensive molecular portraits of human breast tumours. Nature. 2012; 490, 61–70.
- Kandoth C, McLellan MD, Vandin F, et al. Mutational landscape and significance across 12 major cancer types. Nature. 2013; 502(7471):333-339.
