Cancer Signatures Promise Better Detection, Staging, Treatment

BETHESDA, Maryland—As researchers probe the complex nature of individual cancer cells, unique molecular patterns, or signatures, have emerged. Several drugs based on early findings in the field have already earned US Food and Drug Administration approval. A goal set by the National Cancer Institute (NCI) is to "catalog distinguishing molecular signatures of cancer cells to develop new diagnostic and therapeutic approaches and predict response."

Dr. Robert L. Strausberg, PhD, director of NCI’s Cancer Genomics Office, discussed the Institute’s extraordinary opportunity in signature research with ONI Washington bureau chief Patrick Young.

ONI: Would you define what NCI means by the signatures of cancer cells?

Dr. Strausberg: Traditionally, we have defined cancer by body location and various features that can be seen under the microscope. When we talk about signatures, we are talking about understanding the molecular events that distinguish one cancer cell from other cancer cells and from normal cells, and using changes in the genetic blueprint to identify a unique signature that tells us how that cancer cell is different. That allows us, for example, to distinguish different types of breast tumors.

ONI: Can you give me a specific example of a cancer signature?

Dr. Strausberg: I will give you several. Gleevec [imatinib mesylate], whose development for CML [chronic myelogenous leukemia] predates NCI’s extraordinary opportunity, was based on a chromosomal aberration—the 9;22 translocation, or Philadelphia chromosome, which is seen in almost all cases of CML. This is what we would call a molecular signature.

Going from that signature to actually having an intervention was a long process. First, researchers had to determine if the translocation was involved in the genesis of the cancer and also its maintenance. Then they could develop a drug that targeted the protein of that genetic region to treat the cancer.

Herceptin [trastuzumab] is another example. It was noticed that in about 30% of breast cancers, a particular gene, HER2/neu, is amplified and there is more of its protein in the cancer cell. So the antibody that forms the basis of Herceptin is targeted against a gene product, that is, a signature.

The type of signature that has gotten a lot of attention recently is based on the analysis of the cellular expression of thousands of genes at the same time. That means we can look globally at the genes and their expression in cancer cells and normal cells, and, from that, try to uncover signatures at the expression level. Of course, this suggests that we could not only identify a particular cancer but also find various routes toward intervention.

This line of research has been happening for a number of years, but there are new technologies that allow us to do this much more efficiently. By looking at all of the changes in the genome at the same time, we should be able to find the most informative signatures that could be used for detection and intervention.

ONI: How would signatures help in screening and prevention?

Dr. Strausberg: One of the great challenges is simply to know early on that a cancer is developing. We would like to be able to find changes in cells in the early stages of cancer and then use that information, which you can refer to as a signature, to develop relatively noninvasive tests to identify a cancer early. If you knew of a signature in colon or in lung cancer, you could then try to develop a test to detect that change specifically in body fluids at an earlier stage than you could have in the past, and certainly in a less invasive manner.

This research may prove useful in testing preventive strategies. Prevention studies are challenging to do because you have to follow patients for many years to discern the early onset of cancer. If we had signatures that we could use to understand what cancers are developing, then we could obtain much earlier results on the effectiveness of various prevention strategies.

ONI: Does the recently reported potential blood test for ovarian cancer rely on a signature?

Dr. Strausberg: Yes. That is exactly the sort of test to which I am referring. The basis there is the possibility of looking at serum proteins that are associated with the early development of ovarian cancer. There is another report of looking at changes in mitochondrial DNA in lung cancer patients. The idea is that you could look at the sputum and identify these changes that are specifically associated with lung cancer.

ONI: How do you expect signatures to influence staging and prognosis.

Dr. Strausberg: Actually, much of the demand for this information comes from pathologists, who have a difficult time staging tumors and knowing exactly what a grade means. There is a strong demand for molecular information, not only to confirm staging but also to suggest other ways that we could stage these tumors.

Researchers have found a signature difference in lymphoma that is not only useful in staging but also to differentiate between who has a good prognosis and who does not. The signatures occur in patients who are under the same treatment strategy, and it is very clear that these patients fall into groups, and that outcome is related to the signature.

ONI: How is NCI stimulating signature research?

Dr. Strausberg: There are a variety of programs. The Cancer Genome Anatomy Project is cataloging cancer genes and normal genes. There are programs, such as the Director’s Challenge, that take the cancer genes that are identified and use technologies such as microarrays to look at many, many cancer samples to try to discern how they are related to each other. We can actually correlate differences in patient outcome, and there has been some very nice work done in those areas, particularly in B-cell lymphoma.

ONI: What problems lie ahead?

Dr. Strausberg: The big challenge is going from a signature of cancer to an intervention. The challenge is to interface the molecular information that we are generating about cancer with therapeutic information at the more applied level, such as drug development, and to build informatics systems that allow us to move relatively seamlessly between basic research and survivorship. A lot of development is required, but I think that over the next decade, we will move from these signatures to effective interventions.

Another challenge is to consider not only the tumor itself but also the microenvironment in which it is living. Tumors require a blood supply, for example. We need a more complete picture of the tumor cells themselves and of the environment that supports them.

We need to look more closely at particular cell types, the ones directly involved in tumors and the ones that support the tumors. If we can target genes in the tumor cells and in cells that are required for the tumor to live and grow, we will wind up with more effective therapies than we have today. There is also much discussion about genes vs the environment. Signatures reflect what is going on in a cell, and they offer a very practical approach to studying the interactions of cells and the environment.

ONI: When would you expect early clinical trials of more sophisticated signatures, say those that involve the changing ratios of many different proteins in a cell?

Dr. Strausberg: That will start very soon, I think. The signature approach has demonstrated effectively that we can distinguish significant differences among cancers that were previously classified as one disease. And it is clear that when you look at the genes, there is a molecular basis for these differences. That type of technology and information will be part of clinical trials and will inform those clinical trials very soon.