American Scientist

Among the major diseases of humankind, cancer represents a special challenge. In many cases it is stealthy, capable of growing undetected in many different tissues and systems of the body for long periods of time. Certainly, clinical research in recent decades can claim some notable victories: Regular monitoring, whether by means of radiography or of lab tests for certain biomarkers, has brought a large proportion of breast and prostate cancer cases within the bounds of treatment. A vaccine now available from family physicians holds great promise for bringing down rates of human papillomavirus, some strains of which have been found to cause cervical cancer. According to the American Cancer Society, well over 14 million cancer survivors are alive in the United States today, and in the course of the next decade that figure will likely approach 19 million.

Nevertheless, the possibility of one day receiving a cancer diagnosis remains a frightening prospect for most people. Perhaps the time has come for a new approach: What if, instead of trying to confront the invader by means of surgery, chemotherapy, radiation, or other medical armaments, on every battlefront of the body, medical research were to launch a sneak attack from a different point of approach—that is, from the perspective of the alterations in genes that actually drive the growth of a tumor? Scientists at the Dana-Farber Cancer Institute at Harvard Medical School and the Broad Institute, a collaboration between the Massachusetts Institute of Technology and Harvard University, are doing just that.

The exome (the expressed portion of the genome) consists of the 1 to 2 percent of our DNA that encodes the explicit instructions for making proteins. In a paper published online in May in Nature Medicine, a research team headed by Levi Garraway and Gad Getz announced they had developed a technique for sequencing the entire exome from 16 tumor patients.

The notion of studying the genetic mutations that lead to the development of cancer is not new, of course, but in the past such studies have focused on a limited number of hotspots—that is, portions of the DNA sequence that have given evidence of being clinically relevant to the specific type of cancer under investigation.

The gold standard for preserving these sequencing panels was by fresh-freezing them, but this technique has limited the collection of samples to places where they could be kept in optimal conditions. The Nature Medicine paper is innovative both conceptually, in sequencing the entire exome instead of selected hotspots, and technologically, in demonstrating that formalin-fixed, paraffin-embedded samples can now meet the same gold standard. “This paper is intended as a proof of concept for start-to-finish, soup-to-nuts, whole-exome sequencing,” says coauthor Nikhil Wagle, a medical oncologist with Dana-Farber and associate member of the Broad Institute.

In a further innovation, the authors developed an algorithm to help them sort through the mass of data obtained from their full-on sequencing approach. The paper’s first author, Eliezer Van Allen, also of Dana-Farber and Harvard Medical School, explains, “We developed the algorithm to help us determine the data that are most likely to be useful, either for diagnostic or prognostic purposes or for predicting a patient’s response to a particular therapy. Whole-exome sequencing was an appealing idea, but it had never been tested.”

The algorithm revealed that among all the genetic alterations associated with various types of cancer, a relatively small number stood out. These alterations (point mutations, small additions or deletions, or changes in the number of gene copies) showed up repeatedly, regardless of whether the samples came from lung cancer, melanoma, or gastric adenocarcinoma. The researchers designated these as driver genes.

“In any form of cancer,” says Wagle, “there are some driver genes and many passenger genes that go along for the ride but might not initiate anything on their own.” In the approach known as precision medicine, clinicians will aim their treatment at the target genes of a patient’s cancer, and not necessarily at the parts of his or her body that show clinical signs of disease. Anatomical classification—what we refer to conversationally as the “type” of cancer—will naturally remain a key factor in choosing the appropriate treatment: Melanoma, for example, cannot be treated with techniques that are appropriate for ovarian cancer. Rather, the genomic information will be “layered on top,” says Wagle. The more information available, the better. In the future, treatments may be even further refined, essentially custom-designed on the basis of each patient’s genomic data. As Van Allen points out, a number of studies have now demonstrated genomic diversity not only across patients with the same (anatomical) type of cancer, but even within a single patient.

Precision medicine, singling out the driver genes of a cancer for aggressive treatment, “in theory, should apply to any kind of cancer,” says Wagle. In fact, he and his collaborators are so eager to add to their data that they are crowd-sourcing the collection of tumor driver genes, inviting researchers everywhere to send in their nominations. ”Of course, we do require a certain level of evidence, and we spell out those criteria on our website,” adds Wagle.

Dana-Farber and the Broad Institute are not alone in seeking ways to move beyond the kind of broad-brush approach that offered few alternatives for treatment when a patient did not respond well to the routine, usually reliable, but necessarily generic therapies. Many cancer centers around the country are delving into molecular data to fine-tune treatments on the basis of a patient’s genomic profile. The Translational Genomics Research Institute together with George Mason University, the Cancer Institute of New Jersey, Vanderbilt University, and the universities of California, Maryland, and Washington are just a few of the research institutions that have undertaken major initiatives along these lines. Craig Thompson, president of Memorial Sloan Kettering Cancer Center, voiced the hopeful attitude of many of his colleagues when he said in a recent statement, “We’re moving away from the concept of treating cancer as many different types of the same disease and toward treating each person’s cancer as its own unique disease. . . . ushering in what will truly be a new era of precision medicine.”—Sandra J. Ackerman