2 Clarke Drive
Suite 100
Cranbury, NJ 08512
© 2024 MJH Life Sciences™ and OncLive - Clinical Oncology News, Cancer Expert Insights. All rights reserved.
Gad Getz, PhD, MSc, focuses on cancer genome analysis, particularly cataloging all of the genomic events that occur during the development of cancer, in his laboratory at the Broad Institute at MIT and Harvard.
Gad Getz, PhD, MSc
Gad Getz, PhD, MSc, focuses on cancer genome analysis, particularly cataloging all of the genomic events that occur during the development of cancer, in his laboratory at the Broad Institute at MIT and Harvard. He also is a co-principal investigator of The Cancer Genome Atlas Project and a co-leader of the International Cancer Genome Consortium.
The Broad Institute team has published a number of studies in recent years that describe new genes and pathways involved in different tumor types, including a recent Nature article that has expanded the knowledge of genes involved in cancer by 25% (2013;499[7457]:214-218).
Michael S. Lawrence, PhD, a member of the team, is the lead author of the Nature article. In this interview, Getz and Lawrence discuss the laboratory’s latest findings.
What is the significance of the Nature study?
In order for us to really attack cancer on all fronts, we first have to learn all the genes and pathways that drive it. Creating this full list of cancer drivers is a fundamental process of scientific discovery that will happen only once in the history of science, and it’s happening right now. We view it as analogous to what chemists did 200 years ago when they first understood and wrote down the periodic table of elements. It was a seismic shift in the history of science, and it propelled whole new fields of understanding.
Once we have uncovered the whole list of cancer’s “tricks” for cheating the system, then we will be in a good position to solve the cancers existing in patients’ bodies today. However, the news is even better than this: Once we can solve the cancers of today, we can solve the cancers of tomorrow. The reason for why there won’t be an everlasting arms race between the disease and the doctor, as there is with viruses like HIV and SARS, or bacteria like MRSA, which continue to evolve and pass down ever newer and more sophisticated descendants to our children, is that cancers die with their hosts. If we can solve the cancers of today, we will solve cancer for all time. This is something that will happen just one time in human history.
How do your new findings fit into our current understanding of the role and frequency of genetic mutations in cancer?
Our findings resoundingly confirm the established model of carcinogenesis elaborated by Robert A. Weinberg, PhD, Bert Vogelstein, MD, and others: Over years of cycles of copying and division, all cells accumulate random mutations, and occasionally one of these mutations causes a cell to grow and divide too aggressively. These precancerous cells start to snowball from there, accumulating several driver mutations. As Weinberg famously said, the cells are like a car: To get a cancer cell, you have to jam down the gas pedal, cut the brakes, and get access to unlimited fuel. Eventually the process of natural selection leads to the emergence of clonal populations of full-blown cancer cells.
Some quiet tumor types, such as leukemias and childhood cancers, have background mutation frequencies that can be 1000-fold lower than the densely mutated genomes of lung cancer and melanoma. Different genes also have vastly different levels of background mutation, underscoring the crucial importance of properly accounting for these multiple levels of heterogeneity in the system.
A few cancer driver genes are mutated in large fractions of patient populations, for example, TP53, KRAS, NRAS, BRAF, PIK3CA, which can be mutated in 20% of patients, or even (in some tumor types) up to almost 100% of cases. We’ve found most of these high-frequency genes, because they’re very easy to find—the signal jumps right out of even very small data sets. But really, the vast majority of cancer driver genes play a role in only smaller fractions of patient populations.
These are genes like ELF3, STAG2, ARHGAP35, which are mutated at levels like 2% to 5%. We’re still actively discovering many genes like these and are projected to discover more as samples sizes increase.
What are the potential therapeutic/clinical implications of the types of new genes identified and the findings relating to the frequency of gene mutations in cancer?
The new genes we identified fall into groups bearing the classic hallmarks of cancer: genes that drive growth signaling, regulate the cell cycle and apoptosis, monitor DNA damage, etc. A few genes fell into potentially novel cancer hallmarks, for example, posttranslational protein modification. The work has huge clinical and therapeutic implications. As an example, consider multiple myeloma (MM), a cancer of the bone marrow. MM is fairly typical of tumor types we’ve studied, in that there are a few genes (eg, NRAS, KRAS, TP53) mutated in large fractions (ie, 10%-20%) of the patients, and then a long tail of lower-frequency genes occurring in 5% or fewer of the patients. When we look way down the list, we find the gene BRAF, which is mutated in 2% of patients with MM. BRAF mutations are typically associated with melanoma, where they affect over 60% of patients.
In the clinic, patients with melanoma whose tumors have BRAF V600E mutations are treated with targeted therapeutics: small-molecule BRAF inhibitors such as vemurafenib and dabrafenib. Patients with MM carrying these same BRAF V600E mutations, while very rare, are very important to identify, because the patients might also benefit from these drugs.
Do you think that current technology will ultimately permit us to develop a complete catalog of cancer genes in the near future?
Yes, we can definitely achieve this goal. Our calculations predict that if we can sequence 100,000 cancer patients, we’ll be able to learn the complete catalog of cancer genes down to the level of genes occurring in only 2% of patients across 50 tumor types (~2000 patients per tumor type). With the continuing decrease in sequencing costs, this is an entirely reasonable goal to have, and with >600,000 people dying of cancer each year in the United States alone, the urgency of the challenge cannot be overstated. Getting down to the 1% level will take even more patients, of course, but in all likelihood, there will come a day, probably sooner rather than later, when one million people have contributed DNA sequencing data to cancer research.