Within the next five years, there should be new biomarker tests for every major cancer that would complement existing tests, such as mammograms for breast cancer and PSA tests for prostate cancer, members of the US National Cancer Institute’s Early Detection Research Network told ProteoMonitor this week.
“With many of the existing tests, the number of false positives is very high, or the tests are used only for monitoring disease and not for early detection,” said David Sidransky, a professor at Johns Hopkins University and the chairman of the steering committee of the EDRN. “I think what we’re going to see in the next five years is markers that can help solve clinical problems that we’re already dealing with, similar to the human papilloma virus DNA test that triages with abnormal pap smears.”
The NCI recently awarded $9.8 million to 17 Biomarker Developmental Laboratories within the EDRN. The award marked the second round of five-year funding for components of the EDRN. About 40 percent of grantees were new to the network, which was founded in 1999.
“There is great promise in using biomarkers for early detection of cancer,” said NCI Deputy Director Anna Barker. “EDRN provides the necessary collaborative platform and the systematic approach to validation of potential biomarkers that will ultimately bring biomarker diagnostic tests for cancer into the clinic.”
Oliver John Semmes, an associate professor of molecular biology at the Eastern Virginia Medical School, is one of many investigators who have been identifying protein profiles that are different in normal versus diseased states. This popular approach is good for initial investigation, Semmes said, but he finds it essential to follow it up by identifying proteins from the most promising profile peaks.
“The belief is that profiles [with unidentified proteins] are going to be tough to get through the [US] Food and Drug Administration,” Semmes explained. “It’s unlikely that the agencies would consider approving something other than an identified protein. For that to happen there would have to be a paradigm crossing.”
A panel of six or seven antibodies is likely to be the easiest thing to get onto the market, Semmes said.
Currently, the PSA test is the standard for detecting prostate cancer. However, the test misses about 20 percent of men who have prostate cancer. In addition, it is susceptible to false positives, resulting in 75 percent of men with abnormal levels of PSA undergoing unnecessary biopsies.
“There is a large group of men that are being overtreated for prostate cancer,” said Semmes.
Semmes’ group has identified a panel of proteins that, when combined with the PSA test, showed an accuracy rate of 80 to 90 percent during initial trials. The researchers are now conducting a larger, multi-institutional study to validate those results.
On the liver cancer front, Timothy Block, an EDRN researcher from Drexel University, has identified a protein called GP73 that he says is as good as or better than alpha fetoprotein, the current serum biomarker used to test for the malignancy. The protein could be useful in detecting AFP-negative cancers and in detecting Stage I liver cancers, said Block.
Block is now developing a GP73 ELISA test and validating the protein in a study with 500 to 600 samples.
“Things should move quickly now that the protein has been identified,” said Block. “The slow step is the discovery step. We can’t control when the marker will be out in the market, but it will soon be validated enough for either development or dismissal.”
There is no shortage of biomarkers, but finding disease-specific biomarkers is tricky, said Jeffrey Marks, an EDRN investigator at Duke University Medical Center who studies breast cancer.
“Biomarkers for disease screening are going to be very difficult because they have to be very specific for specific diseases,” said Marks.
With breast cancer, mammograms used as a screening tool can reach a sensitivity of 75 percent, said Marks. However, as with the PSA test, specificity is poor, resulting in false positives that lead to unnecessary biopsies. A test based on a panel of biomarkers could help to determine whether a biopsy is necessary.
“For populations of women who have a mammographically detected abnormality, three out of four are benign [abnormalities],” said Marks. “Most of those women would benefit from an additional test that can tip the balance to determine if the abnormality definitely needs to be biopsied or definitely doesn’t need to be biopsied.”
General Population Screening
Sidransky said that additional general-population screening tests for cancer will probably not be available for another five to 10 years because the bar for screening is very high.
“In order for a screening test to be passed [by the FDA], you can have only a very small percentage of false positives,” said Sidransky. “I think what we’ll see in the near future for the major five or six cancers is a new test that will have some role [in detection and monitoring] that will set the stage for the next few years to develop a screening test.”
Semmes pointed out that cancer biomarker research is a collaborative effort, so if separate laboratories come up with different sets of biomarkers for the same cancer, the biomarkers would likely be combined into one test.
“The first thing we would do is cross-analyze the markers and combine them to see if you could get an even better group of markers,” said Semmes.
While some researchers have pointed to cancer as a disease that has made progress in treatment over the last 50 years, Sidransky said advances in screening and early detection have resulted in lower mortality.
“Cervical cancer is almost never diagnosed in an advanced stage these days,” he pointed out. “We want to continue to shift to the left hand stage, because earlier stages of cancer do better.”