Epigenomics said this week that it has transferred its prostate cancer classification assay to the Affymetrix GeneChip platform as it prepares to seek FDA approval for the test in the first quarter of 2007.
The tissue-based test, which measures the methylation status of the PITX2 gene to gauge the likelihood of post-prostatectomy relapse, could be the first DNA methylation-based diagnostic to win FDA approval, according to company officials.
“This would definitely be the first DNA methylation test for diagnostic use,” Achim Plum, vice president of corporate communications at Epigenomics, told Pharmacogenomics Reporter this week. “We are not aware of any IVD-approved … methylation test out there.”
The relative novelty of the platform could be an advantage for the company, which is seeking de novo 510(k) status for the test — a classification intended for devices that are not similar to any previously approved tests. The de novo 510(k) process generally enables firms to bring their products to market more quickly than if they were to seek pre-market approval.
Plum said that the company is taking no chances, however. Even though the FDA “indicated that 510(k) might be the right way to go” in its pre-filing discussions, Epigenomics also plans on filing an Investigational Device Exemption, which is not required for 510(k) approval.
“We decided to do it to have a better discussion basis with the FDA, to have closer contact, and I think that’s very good if you have a novel type of assay,” he said.
Epigenomics, which is headquartered in Berlin and has an office in Seattle, plans to begin clinical trials for the test next year, and Plum said that it is on track to reach its previously stated goal of bringing its first test to market in 2008.
The company is sticking to its product-development timeline despite several setbacks over the last year.
Roche Diagnostics, which had been partnering with Epigenomics to develop the prostate cancer test as well as another tissue-based test for breast cancer, withdrew from those projects in February [PGx Reporter 02-08-06].
Although Epigenomics is still collaborating with Roche on the development of three blood-based tests — for colorectal, prostate, and breast cancer — the diagnostic giant’s withdrawal from the other programs had a significant impact on Epigenomics’ bottom line.
For the first nine months of 2006, the company’s revenues fell 60 percent year over year to €2.3 million ($3.1 million) from €5.7 million. In a statement, the company attributed the decrease to “significantly higher R&D contributions by Epigenomics’ development partner Roche Diagnostics in the prior year period.”
In October, Epigenomics restructured in order to focus on later-stage product development. Several early-stage research projects were “scaled down, delayed, or halted altogether,” and the company announced plans to cut its Berlin-based staff by 34 positions to 78 by the end of the first quarter of 2007. After the restructuring, the company will employ 118 staffers at its Berlin and Seattle locations.
The company said when it announced the layoffs that its tissue test development would not be affected by the restructuring, and it appears that it has made good on that promise so far. The transition of the PITX2 test from prototype real-time PCR assays to the GeneChip platform was completed on schedule, six months after Epigenomics and Affymetrix signed a strategic alliance to co-develop diagnostics on the Affy platform.
Epigenomics said that it has conduced a concordance study that showed 97.8 percent agreement between the RT-PCR assay and the GeneChip-based test.
Plum said that the GeneChip platform offers several advantages over RT-PCR: First, the company will not need to purchase a pricey license to Roche’s RT-PCR technology for diagnostic use. Secondly, Epigenomics expects to benefit from Affy’s large installed base, as well as its customer base for other diagnostic tests based on the GeneChip, such as Roche’s AmpliChip.
Moreover, Plum said, the GeneChip platform gives Epigenomics the ability to add additional markers to its portfolio in the future — a benefit for customers because the cost of reagents and other consumables won’t scale linearly with the number of markers as it would with a series of RT-PCR-based tests.
The FDA approved Affy’s GeneChip platform for genetic testing when Roche sought approval for AmpliChip, but the system will have to be validated again as part of Epigenomics’ submissions for methylation testing, Plum said.
FDA approval of the test would be an important validation for Epigenomics’ approach. DNA methylation studies are quickly gaining visibility within the research community, but methylation biomarkers are still a rarity in the clinic.
The field got a boost in October with the publication in Nature Genetics of the first results of the Human Epigenome Project — a public-private consortium that includes Epigenomics, the Wellcome Trust Sanger Institute, and the French Centre National de Génotypage. The goal of that project was to examine differences in DNA methylation among 12 types of tissue samples from 43 human subjects [PGx Reporter 11-08-06].
In a statement at the time, Stephan Beck, project leader at the Wellcome Trust Sanger Institute and a co-author on the paper, said that the results demonstrated that DNA methylation is “stable, specific, and essentially binary (that is, on or off) — all key hallmarks of informative clinical markers.”
Despite these promising results, the FDA has yet to put its stamp of approval on the approach. Plum acknowledged that its DNA methylation test will be “new data for the FDA — for everybody out there,” but added that he doesn’t foresee “any particular hurdle” in getting the test approved, “other than that it’s novel.”
In the Clinic
Epigenomics sees strong demand among physicians for better molecular diagnostics for prostate cancer, the most common cancer in American and European men. Currently, around 470,000 patients are diagnosed with prostate cancer in the US and Europe annually, and about 115,000 men die from the disease each year.
“One way of using this marker could be to stratify the patients into low- and high-risk groups and do the trial only in the high-risk group. This way, you could shorten the trial, do it on a smaller number of patients, and probably show a better effect.”
Prostatectomies are performed in around 40 percent of patients, but the disease recurs in about 15 percent of these patients.
According to Epigenomics, current prognostic tests based on clinical and histopathological parameters — such as tumor size, Gleason score, and pre-surgery PSA levels — are still ineffective at helping physicians determine which patients are more likely to relapse.
The company views its PITX2 test as an additional tool to help identify patients at increased risk. The test is able to separate patients with a high likelihood of cancer recurrence from those with a low likelihood of cancer recurrence — “even in patients diagnosed with Gleason grade 7, for which reliable prognosis based on conventional parameters is particularly difficult,” the company said in a statement.
While the test should be useful for identifying patients with a poor prognosis that are currently under-treated, Plum stressed that any treatment decisions made using the test are strictly up to the discretion of the physician.
“In the end, the conclusions that the oncologists would draw are up to them because we don’t recommend specific treatment if you’re in a bad prognosis group,” he said. “That’s something that the medical community has to discuss.”
In practice, Epigenomics expects oncologists to use the PITX2 test in combination with existing methods “to get the best possible information on prognosis,” Plum said.
In addition, he noted, the marker could be useful in developing new adjuvant treatments for prostate cancer. “With a 15-percent relapse rate it’s very hard to show the benefit of a treatment. The statistics are very bad here, so you need incredibly large trials and you need to wait for a long time,” he said. “So one way of using this marker could be to stratify the patients into low- and high-risk groups and do the trial only in the high-risk group. This way, you could shorten the trial, do it on a smaller number of patients, and probably show a better effect.”