Biomarker research paves the way to precision cancer medicine.
Her doctor calls her his “miracle child,” even though she’s over 60, the mother of two adult sons and a grandmother of three. But Vera Lynn Coe, or “just Lynn” as she prefers, has survived advanced breast cancer for nearly a dozen years—a medical milestone few reach.
The Cocoa, Fla., resident says she has always maintained a positive attitude, even when refusing to limit her options to palliative care after learning her breast cancer had spread to her liver. Instead, she chose to undergo chemotherapy while she continued working. After surgery, she took Herceptin (trastuzumab), a drug that targets HER2, a protein in her cancer cells that fuels rapid growth. Within a few months, her cancer went into remission.
“It locks the gates on my tumor cells,” Coe says of Herceptin, describing how it prevents tumors that produce too much of the HER2 receptor to gain a survival edge. Because she tested positive for this molecular biomarker, her physician knew immediately that she had a particularly aggressive cancer and that her cancer would likely respond well to this targeted therapy.
The biomarker that guided Coe’s treatment is one of several dozen cancer-associated biomarkers considered “actionable” by clinicians. Some biomarkers, such as HER2, are predictive, meaning they can be used to predict how a patient will respond to treatment. Others are prognostic, meaning they can be used to estimate how a disease will progress. Still others are diagnostic—they can be used to confirm a diagnosis even in its earliest stages. Broadly defined, biomarkers refer to an array of biological molecules in the body, which take many forms—genes, their protein products and the organic chemicals inside cells—that can be objectively measured and evaluated, in the blood, urine, other bodily fluids or tissues, to distinguish normal from abnormal processes and indicate the presence or severity of disease. Biomarkers can be inherited, such as certain genetic mutations, or acquired, such as damage to DNA that occurs from certain exposures.
In cancer, these biological substances are playing an increasingly important role in helping doctors “personalize” medicine, tailoring treatment to each individual’s genetic fingerprint. Molecular mutations inside tumor tissues can help predict responses to certain therapies, as in Coe’s case, or flag differences between aggressive and indolent disease, allowing some patients to avoid or delay chemotherapy. With the aid of other tests, biomarkers can also confirm the early presence of disease or monitor treatment for a risk of recurrence. In blood tests, biomarkers can reveal an inherited vulnerability to certain cancers, including breast and ovarian cancer and a familial type of colon cancer.
As a screening tool in an asymptomatic general population, the bar is high for finding biomarkers unique enough to detect cancer at its earliest, most curable stages while at the same time resulting in the fewest possible false alarms, making it an intensive research goal.
Cancers are Darwinian. The cancer may get better for short periods of time, but then it returns.
Since mapping the entire human genome more than a decade ago, scientists have generated thousands of candidate biomarkers, using advances in technologies to pinpoint genetic errors that may signal disease. Although international teams are sifting through this dense molecular landscape for clues to many illnesses, the early focus has been largely on cancers and how they alter normal cellular processes to promote explosive and uncontrolled cell growth.
In no cancer is that process clearer than in chronic myeloid leukemia (CML), a “disease defined by its biomarker,” the Philadelphia chromosome, according to Mikkael A. Sekeres, director of the leukemia program at the Cleveland Clinic in Cleveland and chairman of the Food and Drug Administration’s (FDA) oncologic drugs advisory committee. Individuals with chronic phase CML carry this genetic abnormality in their leukemia cells, and this “fusion gene” encodes a protein called BCR-ABL, which tells leukemia cells to “grow, grow, grow,” he says, crowding out healthy blood cells and damaging bone marrow.
Named for the city in which it was discovered, the Philadelphia chromosome “is the first genetic abnormality linked to a cancer,” Sekeres says. With the approval in 2001 of the targeted therapy Gleevec (imatinib), which directly inhibits the activity of the BCR-ABL protein, it became “the first real success we’ve had in biomarker monitoring and molecular therapy,” he adds.
Richard Levine, president and founder of Space Coast Cancer Center in Brevard County, Fla., and Coe’s oncologist, describes the success seen in treating CML as the standard bearer for treating other cancers. “You can have patients in complete remission,” he says, “but they do have to take it [the targeted drug] forever, or the cancer recurs. It’s a new concept—the idea of turning cancer into a chronic disease,” similar to hypertension or diabetes, which can be contained for many years.
“Of course, not everybody responds to targeted therapy, and there is relapse,” Levine adds, saying cancer is a complex disease and molecular differences can be seen even within the same cancer.
Biomarkers have a relatively short history in medicine, starting in the mid-19th century when an English scientist identified the first potential cancer marker in the urine of a patient with multiple myeloma. A century later, however, the discovery of DNA’s double helix laid the groundwork for unraveling the genetic code, leading to a burst of biomedical activity. The first protein biomarker targeted by a cancer drug—the estrogen receptor in breast cancer—emerged in 1977, and the prostate-specific antigen (PSA) became the first biomarker for early detection in the early 1980s.
Biomarkers that hold the greatest promise for patients are those seen in the cancerous process itself, such as the Philadelphia chromosome. However, many candidate biomarkers are also found in normal cells or cellular functions, such as cell death. Finding biomarkers elevated only in cancer has proven challenging. Moreover, once a key molecular pathway is shut down, tumors employ “work-arounds,” allowing them to bypass that pathway and continue to grow.
“Cancers are Darwinian,” says Sekeres, referring to the 19th-century English naturalist Charles Darwin, who theorized that nature evolves by selecting for those most adaptable to environmental forces. For example, even though researchers have identified genetic abnormalities in 75 percent of patients with types of leukemia other than CML, drug-targeting efforts have failed. “The cancer may get better for short periods of time, but then it returns,” he says.
Right now, I think we're in the early days of molecular discovery. We're just nibbling at the edges. But there's still a lot more we need to learn about cancer.
Another hurdle: the sheer number of biomarkers discovered over the past decade through “omic” technologies, such as genomics, proteomics and metabolomics. These terms refer to different functional groups of proteins that give us windows into the cancer cell’s behavior. Despite the massive volume of molecular information generated through these large-scale efforts, comparatively few biomarkers have been validated for clinical usefulness, affordability or whether they can be independently reproduced or replicated, says Sudhir Srivastava, chief of the cancer biomarkers research group in the division of cancer prevention at the National Cancer Institute (NCI). “We still want discovery done,” he says, “but driven by clinical usefulness.”
In 2000, the National Institutes of Health formed the Early Detection Research Network, a consortium of more than 300 investigators and 40 private or academic institutions, for just that purpose—to move the discovery and development of biomarkers from a disorganized, piecemeal effort into a collaborative, highly structured endeavor.
The multitiered process, establishing both a scientific standard and a road map, takes a biomarker from discovery and analysis to large validation studies in tumor specimens and patients, and ultimately, to the FDA. The consortium focuses on early detection markers in populations at high risk for developing cancer, Srivastava says, with the hope of identifying the earliest possible molecular shifts in cells as they march toward malignancy.
So far, just five network-supported biomarkers have moved forward to clinical trials, illustrating the sustained, long-term effort required to reach the clinic. Once a biomarker makes it to the FDA, the agency treats it as neither a device nor a drug, says Susan Laine, an FDA spokeswoman, but as a companion test for “pharmaceutical and diagnostic partners who are developing products that detect and target the molecular drivers of cancer.”
Of the five consortium-validated biomarkers, all but one can be measured in blood. Two of those five seek to refine screening for PSA.
Because the PSA cannot distinguish between aggressive and indolent cancers, it is a somewhat problematic biomarker, Srivastava says, “yet it is so widely used around the world, we wanted to increase its performance.” One new biomarker, proPSA, may reduce the number of unnecessary initial biopsies during screening, he says, as aggressive prostate cancers preferentially make this PSA by-product. The other, prostate cancer gene 3, or PCA3, measurable in urine, would be used to see whether men at risk for the disease require repeat biopsies.
From his perspective, James D. Brooks, chief of urologic oncology at Stanford University Medical Center in Palo Alto, Calif., sees PSA screening as a cautionary tale about how biomarkers can fail and lead to overtreatment.
“We have to be careful,” Brooks says. “We learned a lot from PSA screening about detecting stuff that might not be life-threatening,” yet can lead to invasive and unnecessary procedures. Also, PSA testing “falls apart in patients with metastatic disease,” he says, because biomarker expression can change during the course of disease.
Still, Brooks says he’s both an optimist and a realist about the pace of biomarker research today and its potential for improving patient care in the future. To reduce the risk of overtreatment, for example, he says, investigators could couple a cancer-detection biomarker with one that delivers prognostic information about how rapidly a cancer may grow.
“Right now, I think we’re in the early days of molecular discovery,” identifying certain genetic mutations that can bring “stunning results to small groups of patients—5 percent here, 2 percent there,” he says. “We’re just nibbling at the edges. But, there’s still a lot more we need to learn about cancer.”
Coe takes Herceptin every three weeks and will do so for the rest of her life until her cancer progresses, and then she will probably move on to another targeted agent based on her biomarker information. Other than losing her hair from chemotherapy, which is often given with Herceptin, she’s suffered few ill effects. “I’m alive, I keep my spirits up, and I keep wearing my wig,” she says with characteristic grit.