PerkinElmer and Predictive Diagnostics have discovered a panel of blood biomarkers that can distinguish Alzheimer's patients from controls, the companies said this week.
Researchers discovered the six-biomarker panel by using technology made by Vivascience for eluting biomarkers off high-concentration carrier proteins, PerkinElmer's prOTOF 2000 MALDI mass spectrometers, and Predictive Diagnostic's Biomarker Amplification Filter [BAMF] software.
The biomarker panel may lead to a clinical test that can be used for diagnosing Alzheimer's at early stages. Other companies, such as Ciphergen, are also working on developing such a test. PerkinElmer and Predictive's test would be unique in that it would be based on a pattern of mass spectra, rather than a set of identified proteins.
According to Mary Lopez, business leader for analytical proteomics at PerkinElmer, the study involved 302 serum samples: 62 from patients with Alzheimer's, 33 samples from patients with minimum cognitive impairment, and 207 samples from controls. Lopez, who presented the study at the American Association for Clinical Chemistry's Oak Ridge conference in Baltimore last week, said she plans to submit the study to Clinical Chemistry soon.
The team identified an Alzheimer's pattern profile using Predictive Diagnostic's BAMF software, which filters through mass spectra of diseased and control samples and finds disparate features between them. This pattern profile, or fingerprint, involves six protein biomarkers found in blood that have not been identified or characterized.
When samples were blindly processed and run through PerkinElmer's mass spectrometer, researchers found that the pattern profile could identify Alzheimer's patients and patients with minimum cognitive impairment with 94 percent sensitivity and 89 percent specificity.
"The sensitivity and specificity of the preliminary results were quite remarkable for a disease such as Alzheimer's that is so heterogeneous in its origins," said Lopez. "To my knowledge, this is the only blood-based type of initial test that has been demonstrated for this type of a study set."
Lopez said that PerkinElmer would not necessarily be developing the biomarkers into diagnostic tests for the clinic, especially since the participants of the Alzheimer's study had requested that no genes or proteins found in the study be patented.
Samples for the Alzheimer's study were collected through the Rush Alzheimer's Disease Center from a cohort of monks, priests, and nuns beginning around 2002, Lopez explained (see PM 4/21/2003). The religious cohort is unique in that their medical history is very well documented. Members of the cohort had undergone annual cognitive, psychological, and physical evaluations.
The religious participants of the study specifically asked that none of the information gained from the study be made proprietary, Lopez said. PerkinElmer is honoring that request by openly publishing all the data from that study. However, the participants' request does not mean that PerkinElmer can not use the study results to develop a proprietary diagnostic test, Lopez said.
"Everybody wants a diagnostic test for Alzheimer's, and PerkinElmer is not unaware of that," said Lopez. "But I don't want to imply that PerkinElmer is developing this as a diagnostic test. I don't want to say no to that, either. We already play in the diagnostic field, so it is perfectly aligned into our corporate field."
However, Guy della Cioppa, senior vice president of business development at Predictive Diagnostics, seemed certain that the next steps in advancing the biomarkers would involve his company developing a diagnostic for the clinic.
"The next steps would be to validate this test and to get it into the clinic. We are certainly looking to do that," he said.
Della Cioppa said his company is ready to battle the competition in developing a diagnostic, including Ciphergen, which is currently working with collaborators to develop a SELDI test for Alzheimer's (see PM 11/5/2004.) "You always have competition," he said. "The key is which technology actually works to classify blinded samples. Whoever can get there first, that will be the technology that is accepted and gets into the clinic. We certainly think we have the technology to do that."
Della Cioppa said Predictive Diagnostics has already talked with the US Food and Drug Administration twice about its BAMF technology, though not specifically as it relates to the present Alzheimer's study. Predictive believes that the FDA would be ready to approve a test based on BAMF technology, as long as the test is able to reproducibly classify blinded samples with sensitivity and specificity.
"At the end of the day, [the FDA] is data driven," said della Cioppa. "If you can show the data, I think the FDA will be quite pleased."
The next step after the preliminary study is to do a larger validation study, which will probably involve a different cohort of patients, said Lopez. PerkinElmer is currently evaluating what steps to take next in developing the study, she said.
Though there is currently no cure for Alzheimer's, a test that could diagnose early-stage disease would be valuable in helping to develop a drug therapy, Lopez said.
"Having a blood test will allow you to clearly screen many more patients during clinical trials, and to develop a treatment more quickly," she said. "Right now there's no real way to monitor the efficacy of drugs at an early stage."