Ask experts to predict the future of cancer screening, and you’ll get a range of answers. But all would agree that we need better ways to detect cancers early in the course of disease, and these new tools should improve on the benefits of screening while limiting the harms.
“There have been some improvements in triaging patients with new molecular approaches, but with the possible exception of spiral CT screening for lung cancer, we haven’t had any major breakthroughs in early detection” for more than two decades, noted Dr. David Sidransky, director of head and neck cancer research at the Johns Hopkins University School of Medicine.
The dearth of such advances is not for lack of trying. Developing new screening approaches and rigorously establishing their validity is challenging, however, and there are many potential stumbling blocks along the way.
“The bar for ‘proof’ that a particular screening strategy is clinically effective is very high,” noted Dr. Mark Greene, chief of the Clinical Genetics Branch in NCI’s Division of Cancer Epidemiology and Genetics (DCEG). “A screening test must be shown to reduce the death rate from the disease for which screening is being done.”
Much of the search for new screening tests focuses on biomarkers—proteins, DNA, RNA, or other molecules that can signal the presence of cancer and be detected noninvasively in blood, urine, or other readily obtained patient samples or tissues. Researchers are also developing new imaging methods that could be used for early detection, either alone or in concert with biomarkers.
Whatever the approach, “screening is moving away from detecting an advanced consequence of cancer, which is the formation of a mass [or tumor], toward detecting the very earliest changes in the cancer process,” said Dr. Larry Norton, deputy physician-in-chief for breast cancer programs at Memorial Sloan-Kettering Cancer Center.
Dr. Norton chairs the external consulting team for the Early Detection Research Network (EDRN), an initiative of NCI’s Division of Cancer Prevention that supports efforts to discover and validate new cancer biomarkers and technologies.
“Molecular detection of cancer is possible only through evidence-based strategies and implementation,” commented Dr. Sudhir Srivastava, who directs the EDRN. “It takes a village to meet the challenges of early-detection research.”
The Post-PSA Era
In the case of some cancers, researchers are developing new screening tests because the value of existing tests for those cancers has been called into question, perhaps most notably in the case of prostate specific antigen (PSA) testing for prostate cancer. (See “Benefits and Harms of PSA Screening for Prostate Cancer.”)
“The idea that one biomarker such as PSA is going to be useful for all settings has evolved. We now believe that we’ll need panels of biomarkers,” said Dr. Mark Rubin, a professor of pathology at Weill Cornell Medical College. To identify those biomarkers, researchers are using methods such as microarrays and whole-genome sequencing, which rapidly yield a wealth of information, to profile changes that occur in cancer.
Using such an approach, Dr. Rubin, Dr. Arul Chinnaiyan of the University of Michigan, and their colleagues discovered the fusion gene TMPRSS2-ERG, which is found in about half of all prostate cancers. “That fusion gene is seen only in cancer, and, in particular, only in prostate cancer,” said Dr. Rubin, whose team has developed a test to assess the levels of this fusion gene in urine samples. “Our approach now is to try to explain the other 50 percent of prostate cancers with other cancer-specific molecular events” that could eventually form a screening test based on a panel of genetic markers.
For example, Dr. Rubin co-led a recent study that identified a gene called SPOP that is mutated in about 10 percent of prostate cancers. “We can add that gene mutation to the gene fusion to improve on the test,” he explained. “This is the sort of approach we think will be useful for prostate cancer, as well as other cancers in the future.”
Applying Lessons Learned
To avoid unnecessary biopsies or treatment of prostate and other screen-detected cancers, researchers are trying to find biomarkers that better identify which cancers are likely to progress, noted Dr. Joshua LaBaer, director of the Center for Personalized Diagnostics at the Biodesign Institute at Arizona State University and co-chair of EDRN’s steering committee. Whereas some cancers detected by screening will progress and metastasize, others may never cause illness during a person’s lifetime.
“What we’ve learned from PSA is that if we’re going to come up with new screening tools, we also have to develop tools that give us a better idea of disease prognosis,” said Dr. James Brooks, a professor of urology at Stanford University.
Dr. Brooks and Dr. Sanjiv Gambhir, chair of the department of radiology at Stanford, lead a project to deploy new technologies that could form the basis for the next generation of prostate cancer screening tests.
To pave the way for tests that rely on panels of blood-based diagnostic or prognostic protein biomarkers, they are starting to test the performance of a magneto-nanosensor chip technology developed at Stanford. The sensor, which detects proteins tagged with magnetic particles, can measure the levels of up to 64 different proteins simultaneously, in very small sample volumes.
The Stanford team also hopes to adapt an imaging technology being studied in Dr. Gambhir’s lab to improve the accuracy of prostate cancer detection by transrectal ultrasound. The method uses gas “microbubbles” that are encased in a lipid shell to which specific antibodies are attached as a contrast material for ultrasound imaging. The antibodies target a receptor for vascular endothelial growth factor, which is a protein found in newly formed tumor blood vessels. The patented antibody-labeled microbubbles are awaiting Food and Drug Administration approval for human testing.
The Stanford team’s long-term goal is to combine their blood-based biomarker and imaging methods to improve early detection and prognostic assessment of prostate cancer and eventually other cancers. Combining molecular biomarkers and imaging for cancer screening “is a very powerful approach,” commented Dr. Sidransky. “We used to believe in the power of a marker to do everything,” he added. “We now know that’s not true.”
A Sense of Urgency
Researchers have long sought an effective screening strategy for ovarian cancer, and numerous candidate biomarkers for the disease have fallen short of expectations.
“Ovarian cancer is the paradigm for why we need early detection,” said Dr. Michael Birrer, a professor of medicine at Harvard Medical School. The disease can be cured by surgery if discovered early. But “75 percent of tumors are detected at the advanced stage, and those patients are hard to cure,” said Dr. Birrer.
Dr. Birrer and Dr. Steven Skates, an associate professor of medicine at Harvard, are leading a two-pronged effort to discover new biomarker candidates that may ultimately lead to a blood test for the early detection of ovarian cancer.
The first strategy will use extensive proteomic profiling of fluids from benign and malignant tissues, such as ovarian cysts, “to find candidate biomarkers that are systematically different between the two,” Dr. Skates explained.
We used to believe in the power of a marker to do everything. We now know that’s not true.
—Dr. David Sidransky
The second strategy involves genomic analyses to identify genes that are expressed differently in ovarian cancer tissue samples than they are in normal tissues that may give rise to ovarian cancer, and then bioinformatic analyses to look for genes whose protein products are also likely to be secreted into the bloodstream.
Using either the proteomic or genomic approach, or a combination of both, the researchers hope to come up with a short list of candidate biomarkers for further testing and refinement. “We may be lucky to find that some of those candidates are actually early-detection biomarkers that can be measured in blood,” Dr. Skates said.
Those biomarkers could form the basis of a blood test to screen postmenopausal women, and other women at increased risk of ovarian cancer, at regular intervals. For women who test positive on the blood test, a follow-up test, such as transvaginal ultrasound or newer imaging methods, might be used as part of an overall screening approach in the future, Drs. Birrer and Skates suggested.
Gazing into the Crystal Ball
No one can predict with certainty which types of tests will be most effective for screening for particular cancers. However, “if you want to prognosticate the future of cancer screening, my guess is that nucleotide [RNA or DNA]-based tests are going to be the most promising, at least in the short term,” Dr. Brooks said. “The power of nucleic acids is that you can amplify them to an extraordinary degree, which you can’t do with proteins,” Dr. LaBaer added.
Future DNA-based screening tests might detect methylation or other epigenetic modifications of DNA that occur specifically in cancer. “For example, we published a paper last year showing widespread and reproducible changes in DNA methylation in prostate cancer,” Dr. Brooks said.
And future screening tests may detect biomarkers in patient samples other than blood or urine. “One area where I think you’re going to see a change is in…tumors that affect the gastrointestinal tract” or other parts of the digestive system, Dr. LaBaer predicted. “You can look in stool for aberrant nucleic acids [from cells shed by tumors].” Researchers are also investigating sputum-based tests to detect lung cancer early.
“A possibility for the future is that we may stop thinking about cancers in terms of organ sites and may think more in terms of disrupted pathways or molecular variants of cancer,” Dr. LaBaer continued. In that case, “the biomarker people are going to have to work closely with the imaging people to very quickly turn a biomarker discovery into identifying where the tumor is.”
“We’re rapidly changing our concept of what cancer is,” noted Dr. Norton. “You can’t separate screening from understanding biology, from therapy, from prevention. The biggest challenge is weaving it all together [into] the big picture.”
Furthermore, he added, “we may find out that early detection is not helpful in certain situations, and that’s also important. We may not want to screen for certain cancers if we find out that prevention may be a better place to put our resources.”
“Mortality rates for some cancers have remained constant for the past 40 years, and in some of these cancers, new therapies have extended life for a few years but are not increasing the cure rates,” Dr. Skates noted. “Improved early detection for these cancers could shift that number so that more people are cured…. The payoff could be so big.”