This article has been updated to more clearly reflect Celera Diagnostics’ view toward bacterial strain-typing tests.
The Translational Genomics Research Institute and Northern Arizona University have begun collaborating on a pathogen strain-typing and -detection project that may ultimately lead to a new line of PCR-based diagnostics for Celera Diagnostics.
TGen and NAU announced last week that one goal of their collaboration, funded by a five-year, $9 million NIH grant, is to identify pathogen-specific DNA signatures for diagnostics using real-time PCR technology from Applied Biosystems, an Applera company. Celera Diagnostics, a joint venture between ABI and sister company Celera Genomics, has the option to develop and commercialize the tests.
"The goal of the collaboration is to develop a diagnostic test for sepsis and community-acquired pneumonia," said an ABI spokesperson.
Any clinical diagnostics resulting from the collaboration would not be commercialized by ABI, according to a spokesperson. "Although Applied Biosystems won't specifically be developing any potential diagnostic, we have the commercial rights to do that, and Celera Diagnostics has the first right of refusal, because of our structure," she said.
"You could imagine, for example, that you could run the PCR-based assays after 6 to 12 hours of culturing and come away with a whole breadth of information that you wouldn't even get after 48 or even 72 hours, if you just continue culturing."
But there is some doubt as to whether Celera Diagnostics is interested in bacterial strain typing. “It’s not part of our core competencies,” a Celera spokesperson told Pharmacogenomics Reporter. Celera’s focal points include: strain typing and viral-load testing for hepatitis C virus, human immunodeficiency virus; detection of pathogens, such as gonorrhea, chlamydia, and human papilloma virus; and testing for genetic diseases, such as cystic fibrosis and fragile-X syndrome, according to the spokesperson.
ABI and the two research institutions are currently negotiating licensing agreements concerning information generated by the project, which will involve the design and validation of TaqMan assays. Like HIV strain typing, which was among the first PCR-based molecular diagnostic to achieve FDA clearance, the research aims to enter true pharmacogenomic territory.
Using techniques developed by Paul Keim, director of pathogen genomics at TGen and a professor of microbiology at NAU, for previous studies involving anthrax, the project will "develop a panel of [DNA] signatures for each of the pathogens on the list," Keim said. "These will certainly include species identifiers, but more importantly, we'll get down to the strain level, we'll be looking at virulence genes and we'll be looking at [antibiotic]-resistance genes." Armed with strain-specific information, clinicians should be able to tailor treatment to each patient's infection.
The project is a partnership between NAU, TGen, ABI, and Banner Health. Banner is a large non-profit healthcare provider in Arizona that will supply specimens and a clinical venue. The work will focus on sepsis pathogens, such as Staphylococcus, Enterobacter, Escherichia and Candida genera, and common causes of community-acquired pneumonia, such as Streptococcus pneumoniae, non-typable Haemophilus influenzae, Mycoplasma pneumoniae, and respiratory syncytial virus.
"This particular award was for moving into the clinical diagnostic arena, and hopefully by the end of five years of funding we'll have something that's at least moving toward FDA approval," said Keim.
Celera currently manufactures a TaqMan PCR-based HIV diagnostic that has been cleared by the FDA, as well as an HCV diagnostic that does not have FDA clearance. But it may be too early to say whether the company will get into bacterial strain typing, said a Celera spokesperson. "That is something that ABI's applied markets are getting into, and because of the relationship [between ABI and Celera], we would look at that and see what the feasibility is," he said.
The impact of sepsis is enormous, and a rapid diagnostic might have a substantial impact. Big hospitals often have one or two dozen sepsis patients, and "they just plunge in there and start treating [patients] while they're waiting for the diagnostic tests to come back," Keim said. "If it's multi-drug resistant and they haven't anticipated it, then they've lost two to three days of therapy." A molecular diagnostic using current real-time PCR technology may take "an hour or two" in a clinical lab, he said, But the project is not focused on developing a point-of-care assay, so getting samples to the laboratory may be the most time-consuming step, he said.
According to Stephen Kingsmore, director of the Santa Fe, NM-based National Center for Genome Resources, which conducts sepsis research, about 750,000 US patients contract the condition annually. Besides drug-resistant Staphylococcus aureus , sepsis can also be caused by other pathogens, including Escherichia coli, which is a common cause of urinary-tract-related sepsis.
According to the US National Institutes of Health, sepsis is the second leading cause of death among patients in non-coronary intensive care units and the tenth leading cause of death overall in the United States. The disease is also the number one cause of death in neonatal ICUs for low birth-weight children.
Sepsis accounted for a total intensive-care unit expenditure of $7.6 billion in Europe and $16.7 billion in the United States in 2000, the NIH said. Annual health care costs associated with community-acquired pneumonia are estimated to be $23 billion in the United States. "Improved detection of specific etiologic agents early in the course of infection may impact patient management costs by earlier administration of appropriate treatment," the NIH said.
Pneumonia is the sixth most common cause of death in the U.S. Annually, two to three million cases of CAP result in 10 million physician visits, 500,000 hospitalizations, and 50,000 deaths, according to the NIH. Prospective studies have failed to identify the causative agent in 40 percent to 60 percent of cases, the NIH said.
Using procedures developed in Keim's lab, "we can identify the Ames strain [of Bacillus anthracis], for example, that was used in the anthrax letters [found in the US in 2001], from SNPs at the single-molecule level," he said. "You can't do any better than that." Assays developed under the current project would eventually be useful not only with blood samples, but also sputum, pus, wound swabs, and traditional bacterial cultures, said Keim. "Sure, we'd like to be able to take a drop of blood and tell you everything, but whether or not that's going to be feasible really depends upon the pathogen, how the patient presents, and [how far] along they are," he said.
Existing methods of clinical strain typing still depend on time-worn microbiology, and while the adoption of high-throughput molecular technology is already common in epidemiology and research, it would represent a significant step forward in medicine. "You could imagine, for example, that you could run the PCR-based assays after 6 to 12 hours of culturing and come away with a whole breadth of information that you wouldn't even get after 48 or even 72 hours, if you just continue culturing," said Keim. In epidemiology, prior research on multi-locus sequence typing of certain bacterial genes has been used to define "global clonal strain patterns" that TGen and NAU will capitalize on in their own work, he said.
Many resistance genes arise from single-nucleotide changes in DNA sequence, Keim said.
BioMerieux of Marcy L'Etoile, France, has also expressed interest in developing bacterial strain-typing diagnostics. The company has a license to use Affymetrix' GeneChip technology to develop assays for bacterial strain typing, hepatitis C, respiratory infections, and central nervous system infections, but its product stable so far includes only histological tests for some of these disease areas. BioMerieux was unable to comment before deadline.
A microarray-based approach would allow the interrogation of thousands of SNPs, but the real-time PCR approach TGen and NAU are using requires a more streamlined strategy. "In anthracis, we studied [about] three and a half thousand SNPs in the entire species, and then we boiled down those into about a dozen, which we call 'canonical SNPs,' and basically capture all the information from those three and a half thousand SNPs back into the subset — that's because much of the information you get is redundant," said Keim.
— Chris Womack ([email protected])