With an Alzheimer's disease detection test it plans to market as a research-use-only assay by the end of the year, DiaGenic joins a number of molecular diagnostics firms hoping to capitalize on the advantages of blood-based testing.
Diagenic's blood-based gene-expression profile test, which will be validated on Applied Biosystems' TaqMan platform, is an attempt to circumvent a problem that prevents conventional diagnostics from detecting Alzheimer's: a doctor cannot directly sample a patient's brain tissue. Along with competitors such as United Kingdom-based DxS, San Diego-based Sequenom, Toronto-based Chondrogene, and researchers at Purdue University in West Lafayette, Ind., the firm is laying the groundwork for a diagnostic approach that may be useful in disease areas where tissue is not available for sampling.
DiaGenic this week said it had validated its gene-expression profile in a sample of 125 newly diagnosed Alzheimer's patients and 205 archival Alzheimer's blood samples with an accuracy of 87 percent.
"The clinical validation â€" if everything works as we hope â€" should be started early next year," Anders LÃ¶nneborg, DiaGenic CEO, told Pharmacogenomics Reporter this week. The company will submit its test to the US Food and Drug Administration for clearance "within a couple or three years," he said.
"I think that what is different about molecular diagnostic approaches is, up until now, we just haven't had the ability to develop tests that were effective in blood."
There is no treatment tied explicitly to the results of DiaGenic's test, but early detection might help clinicians decide whether to treat patients with drugs that slow Alzheimer's symptoms. Such drugs include Pfizer's and Eisai's Aricept, which is an acetylcholine esterase inhibitor, and Namenda, an NMDA-receptor antagonist marketed in the United States by Forest Laboratories. But there is not yet any reason to think that the test can identify patients who will respond well to an Alzheimer's drug, said LÃ¶nneborg.
There is no test to directly diagnose Alzheimer's disease in a living person. Instead, clinicians use behavioral tests to infer that a patient is suffering from the brain-wasting condition. Definitive identification of Alzheimer's is conducted using samples of brain tissue once a patient has died.
DiaGenic is still working on establishing the specific genes that will be validated with TaqMan, and the company is still evaluating a range of platforms for the final product, said LÃ¶nneborg. The company "has not yet discussed" offering the test as an analyte-specific reagent in reference laboratories, but "that might be a possibility," he added.
There are about 12 million people with Alzheimer's disease worldwide. The company estimates that that number will rise to 22 million to 35 million by 2025.
In Alzheimer's, "You can't take a biopsy from the brain while [patients] are still alive, so you have to find other ways of doing it," said LÃ¶nneborg. "Also, for other diseases, like breast cancer, you must know that you have a [growth] before you can take a biopsy â€" if you don't know, how can you take a sample?"
The next step for DiaGenic is to apply its gene-expression test to an earlier stage of Alzheimer's disease. LÃ¶nneborg said that the firm has already begun a study looking at mild cognitive impairment patients, of whom "about 40, 50, perhaps 60 percent will eventually develop into Alzheimer's," he said. "Earlier, there is an even higher benefit of treatments."
DiaGenic is also working on a blood-based gene-expression profile for breast cancer detection. In April, the company presented data at the American Association for Cancer Research meeting in Washington, DC, indicating that its 58-gene profile correctly identified breast cancer 75 percent of the time in a sample composed of 64 patients and 76 healthy subjects.
In the Blood
Tissue-based diagnostics are much more invasive and expensive than blood-based tests, but they have traditionally wielded certain advantages, said Wayne Marshall, Chondrogene CEO. "I think that what is different about molecular diagnostic approaches is, up until now, we just haven't had the ability to develop tests that were effective in blood, because the expression-level changes that we see in blood â€" compared to tissue â€" are much smaller," said Marshall. "We see blood as a ubiquitous medium for being able to obtain biologic information."
Blood-based diagnostics are only just starting to come into broader use with the FDA's encouragement of biomarker use in drug development, said Marshall.
But full confidence in the blood-based approach is not universal. Recent promising research using DxS' blood-based EGFR assay marks "the beginning of a larger process, and we don't know how important blood-based testing is going to be," Stephen Little, DxS' CEO, said last week. Research by investigators from the National Cancer Research Center in Tokyo "makes it very clear that it might be practical to use DNA isolated from serum, and the next year or so will tell whether that converts in this first report to something that's really clinically useful," he said.
Eyeing PGx for Blood-Based Tests
So far, DxS is the only company to pursue a classically pharmacogenomic application for its blood-based test. The company plans to launch a research-use-only test for epidermal growth factor receptor mutations that may help identify patients who will respond to EGFR inhibitors, specifically AstraZeneca's Iressa and Genentech's and OSI Pharmaceuticals' Tarceva.
Chondrogene calls its approach to blood-based gene expression profiles the Sentinel Principle. The only pharmacogenomic application of the technique so far has been a pilot project conducted to find gene-expression profiles related to rheumatoid arthritis-patient response to TNF-alpha inhibitors, Marshall said.
The company's focus is early cancer detection, and its lead product is a diagnostic for colon cancer and precancerous polyps. "We would expect to go into regulatory trials in 2007," Marshall said. "We're currently exploring whether or not it would require a [premarket approval], or whether we may be able to go by way of the 510(k) process." He said he did not know whether the company may take an ASR approach with its product.
In Sequenom's case, the difficult-to-sample tissue is fetal, and the company in October said it plans to produce a blood-based test for RhD incompatibility between fetus and mother, followed by a blood-based test for another disorder, such as thalessemia, cystic fibrosis, and aneuploidy-related disorders. The company did not reply before press time to a call requesting comment for this article.
The company's RhD assay fits in the broadest definition of the field of pharmacogenomics; should a test prove reliable in finding RhD-positive fetal DNA in a RhD-negative mother, she would be treated with Rhogam to suppress an immune response to fetal cells. Currently, all RhD-negative mothers in the United States are treated with the drug.
Sequenom CEO Harry Stylli said in October that the company will provide homebrew information on the tests to CLIA labs for their own in-house testing within approximately two years, and use the labs' experiences to accrue clinical data to support an application for US Food and Drug Administration 510(k) clearance "in maybe five years, plus or minus a year or two." Sequenom plans to use this route for all of its tests on fetal DNA, applying for FDA premarket approval for those tests without a predecessor, he said.
Another effort involving blood-based molecular testing is under development at Purdue University. There, James Morre and colleagues are working on a method to detect an alternate form of tNOX messenger RNA, which is produced by metastatic cancer cells, in blood using RT-PCR.
The researchers presented data April 3 at the AACR meeting in Washington, DC, showing that the technique had successfully detected cancer in 32 patients, but not 22 healthy subjects.
The test has the potential to become a screen for post-operative and other breast cancer patients, allowing clinicians to possibly delay or circumvent chemotherapy, Morre said in a statement. He was unavailable for comment this week.
â€" Chris Womack ([email protected])