This story originally ran on Jan. 14.
By Tony Fong
Startup antibody biotech firm MabCure has begun a clinical trial for a monoclonal antibody-based diagnostic it has developed for ovarian cancer.
The trial, which will be done in collaboration with Ramathibodi Hospital at Mahidol University in Bangkok, Thailand, will evaluate the efficacy of the antibodies in detecting the cancer, and represents the first step in MabCure's proposed path to get a diagnostic test approved in Europe.
If successful, the company would then file for clearance from the US Food and Drug Administration, followed by commercialization in that market, Amnon Gonenne, the company's president, CEO, and co-founder, told ProteoMonitor this week.
Hasselt, Belgium-based MabCure joins a growing list of firms that are targeting ovarian cancer with proteomic-based assays. Microarray firm Arrayit last month announced its intention to file for 510(k) clearance for its OvaDx test for early detection of the cancer [See PM 01/08/10]. And in September Vermillion received FDA clearance for its OVA1 test, which is meant to indicate the likelihood of ovarian cancer prior to biopsy or exploratory surgery [See PM 09/17/09]. With the recent approval of its Chapter 11 reorganization plan, Vermillion also will be restarting its program to develop other tests for ovarian cancer.
But according to Gonenne, MabCure's technology is an improvement over biomarker-based strategies employed by most proteomics companies because rather than relying on statistical estimates of expressed proteins that may correlate with cancer, it seeks to detect antigens specific to ovarian cancer in order to diagnose the disease with a high level of certainty.
The methodology being used by the company is based on hybridoma technology that was developed by Eli Orr, the co-founder and chief scientific officer of MabCure.
Hybridoma technology is a method of producing monoclonal antibodies invented in 1975. Hybrid cell lines are formed by fusing a specific antibody-producing B cell with a myeloma cell.
But rather than develop antibodies against defined and well-known antigens, the company took the opposite approach: They generated and found antibodies against specific antigens without knowing a priori what they were.
The company took this tack, Gonenne said, because it wanted to be "pragmatic and find out that if we can do that, then later on we can do the basic science to figure out what the antigens are. And if we're unable to find those antibodies then there's no point in doing any further work."
Because MabCure's technology has not yet been patented, Gonenne declined to describe it in detail, but he said that it is directed at detecting antigens specific to various types of cancer.
In all the years of antibody-based research, no cancer-specific antigens have ever been found. Instead, the antibodies target more general antigens that are overexpressed in some subpopulations of patients with cancer. Orr hypothesized that in the early stages of cancer, antigens specific to cancer have an evolutionary rationale for hiding themselves and are masked by non-cancerous antigens.
To overcome this, Orr set about to develop a "fishing strategy [using antibodies as the lure] that would be able to uncover those few hidden antigens." Existing phage-display technology is not able to do this because it is directed mostly at native proteins, according to Gonenne. Hybridoma technology can differentiate cancer-specific proteins from non-cancerous proteins, but because it takes months to develop a sufficient amount of antibodies, it is not an efficient approach.
"If you're looking to screen … a very large population, you need to be able to [make] very large libraries" of antibodies, Gonenne said. By taking apart the classic hybridoma technology and re-engineering it, "we are able to generate thousands of antibodies in about four to six weeks."
Another reason researchers previously failed to detect cancer-specific antigens was that they were not paying enough attention to the tertiary structures of antigens, according to Gonenne.
"If you take a cancer cell and treat it chemically, about 99 percent [of the time] you create a structure that does not necessarily correlate with its native structure," he said. "If you inject [those antigens] into animals, you create antibodies, of course, against those structures, but those antibodies are not relevant to the native antigen."
[ pagebreak ]
So Orr created a series of processes to keep the structures of the antigens on the cell surface in their native 3D state.
Testing his technology first against melanoma, Orr was able to generate three "highly specific" antibodies against melanoma surface antigens that were not found on normal cells or in any other cancer.
It is this specificity that gives the MabCure assay its advantage over traditional proteomics methods, according to Gonenne.
Proteomic assays to date have not been based on the identification of specific biomarkers for cancer, but on the relative abundance of "normal" proteins and a statistical analysis of the measurements of those proteins to predict the likelihood of cancer.
But if there is an antibody that targets a marker that appears only on the surface of a disease cell and nowhere else, "then the answers you get are X and Y — it's a zero or one. It's a binary answer," he said. "Our results [do not] say a patient is likely to have the disease or probably has the disease. Our results say the patient has the disease or doesn't."
While numerous firms are searching for specific protein cancer biomarkers that definitively can tell whether a patient has cancer, progress on that front has been slow, Gonenne said.
Even existing markers currently used to diagnose cancer have severe limitations, he added, citing specifically prostate-specific antigen, which he called a "ridiculous standard" because of a high false-positive rate. According to the National Cancer Institute, most men with elevated PSA levels do not have cancer.
MabCure is developing a panel of antibodies for prostate cancer that, compared to PSA, is not prostate-specific but prostate cancer-specific and does not react with normal prostate antigens. In the next few months, the company will begin a clinical study for that panel, Gonenne said.
Meanwhile, the company also has conducted preliminary work into colorectal cancer, though development of that antibody panel, as well as a panel for melanoma, is on hold. As a year-and-a-half old company with only four employees, MabCure has to limit its focus, Gonenne said.
For the time, ovarian cancer is the company's main priority. Two years ago, MabCure took some of the antibodies being investigated in the clinical trial and tested their ability to detect disease in the blood of patients with ovarian cancer. The researchers discovered that they were able to detect disease "in each and every patient," according to Gonenne, including those patients — about half of the total sample — who had undergone chemotherapy.
For those patients, traditional methods indicated they were in clinical regression, but in fact, some eventually were later diagnosed with a recurrence of the disease, Gonenne added.
"Our antibodies were able to show that even for those patients who were shown to have no residual disease, we found each and every one of them had cancer," he said. Based on those results, the company decided to proceed to the clinical trial.
Gonenne declined to describe in detail the trial design, but said that recruitment of patients has started. The trial will look at the ability of MabCure's technology to detect antigens in both blood and urine.
The company plans to extend the clinical trial to include centers in Europe for potential approval of a diagnostic in that market. Then, it would file with the FDA to market such a test in the US, Gonenne said.
MabCure plans to follow the path that Agendia used in getting FDA clearance for its microarray-based MammaPrint genetic test for breast cancer, which was cleared in Europe in 2004 and approved in the US as an in vitro diagnostic multivariate index assay in 2007.
In that process, Agendia submitted to the FDA studies of tumor samples and clinical data from more than 300 patients at five centers in Europe that validated MammaPrint's intended use. The move obviated the need to conduct clinical trials in the US for the FDA.
By extending its clinical trial to Europe and Thailand, the company anticipates it will have enough data across a wide enough population to meet FDA requirements, Gonenne said.