Johns Hopkins Magazine
Johns Hopkins Magazine Current Issue Past Issues Search Get In Touch
a Cure

In the war against cancer, molecular biomarkers hold out tantalizing promise.

By Kristi Birch
Illustrations by
Ferruccio Sardella

It doesn't look good, at least not judging from the body count: Cancer kills more than 1,500 Americans a day, according to the Centers for Disease Control and Prevention. That's one of every four deaths, or more than half a million people a year. And it gets worse. More than 30 years after President Nixon declared the War on Cancer in 1971, the American Cancer Society reported in January that despite some advances in treatment and prevention, cancer surpassed heart disease as the leading cause of death in the United States for people under the age of 85.

Tell all that to leading Johns Hopkins cancer expert Bert Vogelstein, and he'll tell you something different. Over the last 25 years, Vogelstein's seminal discoveries have helped establish cancer as a genetic disease — and earned him status as the most highly cited scientist in the world. If we're losing a war, our general doesn't seem to know it. "When I went to medical school, cancer was a black box. Now that has completely changed," Vogelstein says. "There has been a revolution."

The revolution is a whole new understanding of the molecular biology of cancer that should enable doctors to detect and remove cancers in their earliest stages, when the cancer is still curable. Specifically, they'll use a new generation of diagnostic tests that look for defective molecules.

At the heart of it all are damaged genes. Thirty years of a war, and the enemy turns out to be a few typos. That's what causes the insidious beast that is cancer — a handful of spelling mistakes in the code of DNA. The DNA molecule is a twisted ladder of two strands of acid joined by rungs of four chemical bases. These bases are represented by letters: T, A, C, and G. When cells reproduce, they must replicate the exact sequence of Ts, As, Cs, and Gs. Many genetic alterations known as mutations are typos: substitutions of one letter for another. Other mutations are caused by extra or deleted letters or even whole words or paragraphs. One mutation, and you've got a damaged gene. And once you've got enough mutations in genes that affect cell growth, cells can grow out of control, forming tumors and invading normal tissue throughout the body. That's cancer. So far, around 100 genes have been found to be associated with specific types of cancer, and the number keeps growing.

Gene mutations have two origins. The first is when people inherit mutations from their parents. Tests already exist to screen people for their predisposition to colon, kidney, and breast cancers by detecting associated gene mutations. But testing positive for an inherited mutation does not guarantee that someone will get the disease. All cancers have mutations, but not all mutations result in cancer. Women with inherited mutations in the breast cancer susceptibility genes known as BRCA1 or BRCA2, for example, have up to an 80 percent chance of developing the disease, while people with defects in the MSH2 genes have a 70 percent chance of developing colorectal cancer.

Nevertheless, heredity accounts for just 5 percent of all malignancies. The other 95 percent happen because cells sometimes make mistakes when they copy their DNA. These mistakes are exacerbated by exposure to such things as cigarette smoke, ultraviolet light, toxins, and infections.

Unraveling the biology of cancer has already yielded new cancer drugs such as Gleevec for leukemia and Iressa for lung cancer. These drugs target defective cells and try to fix their particular molecular defects. Because the drugs aim to cure people who are already sick, they also grab headlines, captivating investor interest and snaring research dollars. In fiscal 2003, the National Cancer Institute allotted over $1 billion (23 percent of its budget) for treatment research, but only $318 million (7 percent) for research on diagnosing and detecting cancer. Nonetheless, treatment often prolongs rather than saves lives, and sometimes by just a few months. That's why Vogelstein contends that the best way to win the war is not with new cancer therapies and drugs, but by pre-empting cancer through early diagnosis.

"Any cancer can be cured if it's caught early enough," Vogelstein says. "Cancer develops in a place in the body, in an organ. As long as it hasn't spread to other organs, it generally can be removed."

Existing tests such as pap smears for cervical cancer and mammograms for breast cancer are a testament to the effectiveness of early detection; they have significantly reduced deaths caused by these cancers. But even these tests aren't perfect. They depend on the skill of the lab technician and the capability of the imaging, which means some cancers go missed. They also result in many false positives, which cause unnecessary anxiety for patients and can lead to unnecessary invasive procedures, such as biopsies, which carry the risk of complications and even death.

Vogelstein: "Any cancer can be cured if it's caught early enough."
Photo by Keith Weller
This is where the new generation of diagnostic tests comes into play. Doctors will be able to detect many cancers through new tests based on molecular "biomarkers" such as gene mutations or increases in proteins that definitively indicate a certain cancer. Unlike tests that screen for a single mutation to indicate a person's predisposition to a cancer, these tests will screen for a whole series of molecular changes that indicate the presence of a cancer in the patient's body. And they will do it with certainty. By combining these tests with improved imaging techniques (such as an MRI, digital mammography, or a CT scan with high resolution) to pinpoint the smallest malignancies, physicians will be able to remove tumors before they become invasive. Because they will catch cancer before it spreads, there will be no need for chemotherapy or radiation therapy.

This vision is not as far from being realized as one might think. Just ask Ted Levy. In 1982, Levy's father died of pancreatic cancer, and his uncle died of it four years before. Pancreatic cancer, a silent killer with few symptoms, claims the lives of 96 percent of its victims — most within a year of diagnosis.

In 2001, Jane, a close friend of the now 58-year-old Internet consultant and his wife, Lenore Levy, was also diagnosed with the disease. In an effort to help Jane, Lenore went on the Internet, looking for a clinical trial or even just something to make the rest of Jane's life more comfortable. That's when she stumbled upon a link for the National Familial Pancreas Tumor Registry at Johns Hopkins Hospital. Scientists believe that familial risk (family history) accounts for 1020 percent of pancreatic cancer cases.

"Bells started going off in my head," Lenore recalls. "That was my husband. I kept shoving it at Ted and saying, 'This is you.' Then finally he said, 'Oh, my God. It is.'"

Levy sent an e-mail to the registry, and they sent him a detailed questionnaire. More than a year later, long after Jane had died, Hopkins contacted him about participating in the CAPS2 (Cancer of the Pancreas Screening) study. The study screens people with a high familial risk for pancreatic cancer through an aggressive program of advanced imaging.

In January 2003, Levy traveled from his home in Cherry Hill, New Jersey, to Hopkins for the initial screening, which included a consultation with a geneticist, blood testing, and endoscopic ultrasonography (EUS). The geneticist recommended that Levy be tested for mutations in the BRCA genes, which were originally associated with breast cancer and now are also associated with pancreatic cancer. As an Ashkenazi Jew (of Eastern European descent), Levy was much more likely to have this mutation. Gastroenterologist Marcia Canto conducted the EUS, and she found a subtle abnormality. "I was surprised," Levy says. "My attitude when I went into this study was, they want to collect data and I'll have the opportunity to get some sophisticated imaging. I didn't think it would show anything, but I had nothing to lose."

Canto ordered an ERCP (endoscopic retrograde cholangiopancreatography), in which a catheter goes into the pancreatic ducts and injects dye so that an X-ray can be taken, and a multi-detector CT scan. The lesion appeared to be chronic pancreatitis, not anything to be happy about, but not cancer, either. But at a follow-up test in January 2004, Canto concluded that the lesion was intraductal papillary mucinous neoplasm (IPMN), or pre-cancerous cells — and that it appeared to be progressing. In the meantime, another warning sign had come in: the results of Levy's genetic tests. He had a BRCA2 mutation.

Doctors recommended a Whipple resection, surgery that removes the head of the pancreas, where Levy's lesion was located, as well as a portion of the small intestine, common bile duct, and gall bladder. The Whipple procedure is major surgery and requires weeks for recovery. But IPMNs are thought inevitably to become invasive cancer — and the lesion was already at 1 centimeter. Levy remembered his uncle and father, and made his decision.

While he was in Johns Hopkins Hospital, Levy met three other men also having the Whipple. These men had actual pancreatic cancer. The same surgery Levy was choosing gives people with full-blown pancreatic cancer their only real hope: overall, a 20 percent chance of survival. "They were way worse off than me," Levy says. "I was going through it just as an elective procedure."

Six days after the surgery, Ted Levy walked out of the hospital, a little sore but cancer-free. For the time being, at least, he has stopped the disease dead in its tracks. "Early detection of that IPMN using clinical screening probably cured him," says Michael Goggins, director of the Johns Hopkins Pancreatic Cancer Early Detection Laboratory, who designed the CAPS studies with Canto. "This is really going after cancer."

As fortunate as Levy was, imaging is only one side of the new diagnostic coin. Often, doctors cannot discern from imaging alone if an abnormality is cancer. In an earlier CAPS study, subjects also underwent imaging of their pancreas. Two patients who went to surgery with subtle imaging abnormalities did not have a tumor. This is why the new molecular tests are so necessary: to know whether a patient has cancer and thus needs surgery in the first place. And it's why the CAPS2 researchers now also collect blood and pancreatic fluid from patients to find a common molecular biomarker among those who turn out to have the disease. Goggins' group has identified a number of markers that can be identified in the blood and in pancreatic fluid. The tests they are working on will help best decide who needs surgery and who does not. Molecular tests like this are in the works for many different kinds of cancers all over the scientific world, and especially at Johns Hopkins.

One of these tests is already on the market and being used in a clinical setting. PreGen-Plus", a stool-based DNA test that detects colon cancer, was released in August 2003. Developed by EXACT Sciences, it is the first cancer test available that uses biomarkers to screen for the presence of actual cancer.

Colon cancer is the fourth most common cancer for both men and women, but it's very curable when caught early. Though colonoscopies effectively screen for the cancer, they are invasive (a tube with an attachable camera and other tools is inserted through the rectum to obtain tissue samples), and many people are unwilling or unable to have them. PreGen-Plus provides an alternative.

The research for the stool-based DNA test began in 1991 with David Sidransky, now director of the Head and Neck Cancer Research Division at Hopkins, when he was a postdoc in the molecular genetics laboratory co-directed by Vogelstein and oncologist Kenneth Kinzler. They found that mutations in K-Ras, an oncogene important to the growth of colon cancer, could be detected in stool samples.

Not only was this discovery important for colon cancer, but it was also a eureka moment for cancer in general: Now scientists knew that they could detect cancer not just in tumors, but in cellular debris found in bodily fluids and secretions.

If the next generations of DNA tests become sensitive enough, they could replace colonoscopy as a screening test for patients who show no symptoms of colon cancer. PreGen-Plus is based on alterations in K-Ras and other genes found by the Kinzler-Vogelstein laboratory to contribute to colorectal cancers. Right now, the test detects mutations in close to two-thirds of patients with colorectal cancers, including patients whose cancers are detected early enough that they can be cured through surgery alone. Because it is based on mutations associated with the presence of existing colorectal cancer, the stool-based DNA test has a very low rate of false positives.

If the next generations of DNA tests become sensitive enough, they could replace colonoscopy as a screening test for patients who show no symptoms. Already, the DNA test is making a difference. "We know of some patients who had taken the test, and early-stage cancer was detected. Over half the people using the test to date had never been screened before," says EXACT Sciences' Amy Hedison. "It is absolutely saving some people's lives."

In the future, some cancer tests could even be conducted at home. That's what Sara Sukumar, professor of oncology at Hopkins School of Medicine, has in mind for breast cancer detection.

Fifty years ago, a woman discovered she had breast cancer when she felt a mass large enough to have been present in her body for years. Treatment was often too little, too late. It wasn't until the 1980s that women started getting regular mammograms — low-level X-rays that produce an image of the inner breast tissue that reveals suspicious areas. Though mammograms remain the gold standard of breast cancer detection, most lesions that mammograms find are benign, leading to undue anxiety in women when they have to go through the process of getting a biopsy.

On the path to a home test for breast cancer: Sara Sukumar
Photo by Bill Denison
"Our dream is to have a home test women could administer themselves, like a pregnancy test. If it turns blue, she goes in for further testing. If it's white, she's okay," Sukumar says. "That's where I'm headed."

Most breast cancers originate in the lining, or epithelium, of the breast ducts. The test will depend on testing fluid from breast ducts for a biomarker called hypermethylation. When chemical units called methyl groups are attracted to DNA in concentrated form, they shut off the genes to which they are attached. This becomes a problem when they shut off genes that control for cell growth. Once these genes are silenced, tumors can form. More than 50 growth suppressor genes have been found in breast cancer to be turned off because of hypermethylation, and more are being discovered.

Sukumar envisions a test whereby a woman would massage fluid out of her nipples and either send the sample to a lab to be tested for hypermethylated genes, or, ideally, put it on blotting paper at home. If the test is positive, the woman would have a mammogram to find the exact location of the cancer, and then have it removed. Disaster averted.

In addition to using biomarkers to detect cancer, scientists aim to use them to gauge the aggressiveness of the disease. Take prostate cancer, which is very deadly once it metastasizes. Since 1986 and the advent of the PSA (prostate-specific antigen) test, which screens for high levels of an antigen often associated with prostate cancer, doctors have been able to catch more cases of the cancer in its earliest stages; today, only about 6 percent of patients are diagnosed with metastatic disease, compared to 50 percent just 20 years ago.

The test is doing a good job of finding the cancer. Sometimes, too good. Some prostate cancers aren't life-threatening, which means there are men who undergo needless treatment, which carries the risk of incontinence and some degree of erectile dysfunction. If the surgery is performed at a center of excellence such as Hopkins, the risk of these side effects is low, but doctors still need to know who can benefit most from surgical treatment. For some men, the "wait and watch" approach can be the best choice.

"Prostate cancer is so much more complicated because you have to figure out what cancers need to be cured. Some prostate cancers grow slowly enough that the man doesn't need to be treated, particularly if he has other potentially life-threatening illnesses, such as heart disease," says William Isaacs, professor of urology and oncology at the James Buchanan Brady Urological Institute at the Johns Hopkins Hospital. "We want to say more than just if a guy has the disease or not."

Isaacs' lab is working with other Hopkins urologists and scientists to find better biomarkers for detecting prostate cancer with certainty and for gauging its aggressiveness. He and his colleagues have found that a gene known as DD3 is overexpressed only in malignant prostate tissue and present in the urine of men with the disease. Preliminary data suggest that the level of DD3 may be higher in men with aggressive disease. Though not yet FDA-approved, a urine test (developed by Jack Schalken of the Netherlands and launched in 2004 from Gen-Probe and DiagnoCure) screens for the DD3 urine-based marker of prostate cancer and may also gauge its aggressiveness. This is one of the first gene-based tests that could provide a prognosis as well as a diagnosis of a given cancer.

Some of these new early detection tests will be based on proteins. Daniel W. Chan, director of the Clinical Chemistry Division at the Hopkins Medical Institutions, and a team of researchers have identified a panel of three proteins that, when present in different concentrations in the blood, indicate early-stage ovarian cancer in 83 percent of cases.

Ovarian cancer kills more women than any other type of reproductive cancer. Yet if caught before it has spread outside the ovaries, the survival rate is over 90 percent. The problem is catching it early: The symptoms of ovarian cancer — abdominal pain, digestive problems, backache, fatigue — get confused with other diseases, and often don't even appear until the cancer is too advanced. The result is that only 24 percent of ovarian cancers are found early, and after the disease has spread, the survival rate drops to 25 percent.

Currently, the only test for ovarian cancer looks for a blood protein called CA125, but it detects only one-third of cases. The FDA has approved its use only to screen women who have already had the disease. Chan hopes that his new three-protein screening test will become available to all women in a few years, after development and additional testing verify his findings. His hope? "To detect stage 1 for all ovarian cancer patients."

These are just some of the tests being devised at Johns Hopkins. There are plenty of others. Oncologist Stephen Baylin is working on finding markers for lung cancer. And, 13 years after his crucial discovery in Vogelstein's lab, David Sidransky has a urine test in clinical trials for bladder cancer that looks not for mutations but for missing chromosomal material and repetitive sequences of DNA that also fingerprint the disease. This test should pave the way for new, more sensitive and specific tests for other cancers, including colon, lung, head and neck, and prostate.

In the meantime, in January — a year after he pre-empted the disease that killed his father and uncle — Ted Levy had his first annual post-surgery exam. The results were normal. Now he's busy preparing for a trip to Italy, where he and Lenore will bicycle from Florence to Venice. They say they'll always be a little nervous about Ted's yearly follow-up exams, but they feel lucky as well. "Nobody in Ted's family had a chance," says Lenore. "But I feel like if anyone has a chance, we do."

And, as the optimists point out, it took just 50 years from the time Watson and Crick determined the double helix structure of DNA until scientists completed the Human Genome Project in 2003. It identified all 20,00025,000 genes in human DNA, as well as the sequence of the 3 billion chemical base pairs that make it up. That laid the groundwork for even more discovery. In medical research, action comes relatively soon after understanding. That's why it might not be long before we can gauge our progress in the war on cancer not by the number of lives lost in the battle, but by the number of people who barely had to fight at all.

Kristi Birch writes from Baltimore.

Return to April 2005 Table of Contents

  The Johns Hopkins Magazine | 901 S. Bond St. | Suite 540 | Baltimore, MD 21231
Phone 443-287-9900 | Fax 443-287-9898 | E-mail