Using SELDI technology from Ciphergen, researchers at Johns Hopkins University have identified three protein biomarkers that when combined with the standard CA125 biomarker detected early-stage ovarian cancer in 74 percent of women study patients who had the disease.
The new biomarkers, the latest in a wave of new proteomic biomarker discoveries in this area (See PM 06-11-04), can potentially be developed further as a diagnostic to improve detection of early-stage ovarian cancer.
“Typically, only half of early-stage ovarian cancer patients have high levels of a standard marker called CA125,” said Zhen Zhang, an associate professor of pathology at Johns Hopkins who was the first author of a study on ovarian cancer biomarkers published in the Aug. 15 issue of Cancer Research. “But combining CA125 with our new markers may improve early detection capabilities.”
In reaction to the positive news, shares of Ciphergen jumped more than 21 percent on Monday to $3.29 per share.
“If someone has a test that proves early stage ovarian cancer, there’s clearly a big market for that,” said Matt Hogan, the Chief Financial Officer at Ciphergen.
Hogan said that about five million CA125 tests are given each year. The company envisions the recently-discovered biomarkers being used as an additional tool by doctors to detect ovarian cancer in women that are at high risk of having the cancer due to family history or symptoms such as pelvic pains.
Daniel Chan, the last author of the study who is a professor and the director of the Biomarker Discovery Center at Johns Hopkins, said researchers used four different Ciphergen chips to capture proteins from serum samples of patients with and without ovarian cancer. The serum samples were fractionated into six different pH levels before being applied to the chips.
Previous researchers had identified protein profiles — patterns of proteins immobilized on chips — that could potentially identify a patient as having ovarian cancer, but Chan and his research team were not satisfied with merely having a pattern.
“A profile is a signature at a given time point — if I’m walking around I might have a protein profile, and if I go to sleep I may have a different profile, and if I have a couple of drinks, the profile may change,” Chan explained. “We don’t just stop there (at the profile). Once we identify a host of proteins, we go on to identify what those specific protein biomarkers might be.”
After selecting out the most promising protein peaks from their protein chip/mass spectrometry data, the researchers tested proteins for their ability to act as biomarkers for ovarian cancer.
Ultimately, the researchers narrowed down the test proteins to the three that were the most successful biomarkers: Apolipoprotein A1 was down-regulated in cancer; a truncated form of transthyretin was also down-regulated in cancer; and a cleavage fragment of inter-alpha-trypsin inhibitor heavy chain H4 was up-regulated in cancer.
“We are excited by these results and also to have licensed the test from Johns Hopkins in preparation for commercialization,” said Gail Page, president of Ciphergen’s diagnositic division.
Page said Ciphergen is now working on validating the results of Chan’s study in a 1500 sample trial.
Under a licensing agreement between Ciphergen and Johns Hopkins, Chan is entitled to a share of royalties received by the university from sales of ovarian cancer biomarker products. Financial terms of the deal are not being disclosed by the company, Hogan said.
Chan’s group used serum samples from five clinical centers to screen for initial potential biomarkers. Once they had honed in on the three most successful biomarkers, they tested them out on serum from a cohort of 41 healthy women and 41 women with ovarian cancer who had been treated at the Johns Hopkins Medical Institutions .
The research group combined the three new protein biomarkers with CA125, the standard biomarker that is commonly used to monitor women with ovarian cancer to see if the disease is progressing. They found that of 23 patients with early-stage ovarian cancer, the biomarker group correctly identified cancer 74 percent of the time. When CA 125 was used alone with the same sample, it correctly identified cancer 65 percent of the time. The difference in sensitivity was not statistically significant, given the sample size of patients.
Researchers then adjusted cutoff values for CA125 to below current standards. With the adjustment, the new markers combined with CA125 correctly identified cancer 83 percent of the time — an amount significantly higher than CA125 alone.
“Of course we would like it to be perfect - 100 percent,” said Chan. “But no diagnostic test is capable of doing 100 percent, and the patient may be misdiagnosed. The question is, this kind of performance is good enough for what?”
Chan said that the new biomarkers were not good enough for screening the general population for ovarian cancer, but they could be used for screening high-risk women - women that already have some signs and symptoms of ovarian cancer.
“Right now CA125 is the only biomarker in blood that is approved for (screening women with symptoms),” said Chan. “So definitely this is better than that. A combination of this blood test with ultrasound could help improve detection.”
Tim Veenstra, the director of the laboratory of proteomics and analytical technologies at Science Applications International Corporation, a research arm of the National Cancer Institute, said that Chan’s approach in using several biomarkers instead of one biomarker is a sensible one.
“If you go to a doctor sick he won’t just look to see if you’re coughing, he’ll look at other things too,” said Veenstra. “Looking at multiple biomarkers is like looking at multiple symptoms. I think they’re working on a very sound hypothesis that we can increase our cancer diagnostic sensitivity by instead of using one biomarker, using a panel of biomarkers.”
Veenstra said that the researchers need to test the biomarkers on a larger, statistically significant sample of women, and that the true test of whether or not the biomarkers are valuable will come five years from now when researchers can see if the biomarkers are being used in the clinics.
“The goal of all this is to develop a clinical assay. That’s where we’re all headed for,” said Veenstra. “If we don’t achieve that, then more or less we may have failed.”
Chan said he and his research group are looking to further validate the new biomarkers by putting them through clinical trials. If the clinical trials go well, the researchers will then seek approval from the US Food and Drug Association to use the biomarkers as a diagnostic test.
“The intention is to make the diagnostic test available to other people, to clinical labs,” said Chan.
Aside from investigating ovarian cancer, Chan’s research group has also used proteomics techniques to identify biomarkers for pancreatic cancer, prostate cancer and breast cancer.