By Ben Butkus
Clinical researchers from the University of California, San Francisco, have published a research paper demonstrating that Cepheid's Xpert Flu A panel test performs as well as or better than several more established methods for detecting 2009 H1N1 influenza, but still falls a bit short when compared to the gold standard method.
Nevertheless, Cepheid's test, which runs on its automated GeneXpert molecular testing platform, is easier to use than all other available assays and represents "a great improvement" over testing methods currently employed by a majority of clinical labs, the study's lead author said this week.
In the study, published online this week in the Journal of Clinical Microbiology, the UCSF researchers used Xpert Flu A and the US Centers for Disease Control real-time reverse transcriptase PCR assay to analyze 105 frozen upper respiratory specimens that had been previously tested for the presence of 2009 H1N1 flu using direct fluorescence antibody detection and Luminex's xTag Respiratory Viral Panel.
Of the 105 specimens tested, 26 had been previously found positive for influenza A by direct fluorescence antibody; while 13 were DFA negative or indeterminate, but positive by Luminex RVP testing for influenza A matrix gene and negative for seasonal H1 and H3 genes.
In addition, 49 samples had been previously found to be DFA negative and were not tested with Luminex RVP; and 17 samples were both DFA and RVP negative. These results led the researchers who conducted the previous study to conclude that Luminex RVP testing has superior sensitivity to DFA for detecting 2009 H1N1 flu. The results of that study were published last year in the Journal of Clinical Virology.
In the most recent study, the UCSF researchers retested samples with invalid or discordant results from the prior study using Xpert Flu A and the CDC RT-PCR assay, which detects the influenza A matrix gene, swine influenza A matrix gene, and 2009 novel H1 hemagluttinin gene. The CDC assay, which also tests for specimen integrity by PCR for the RNAse P gene, is widely considered the gold standard for H1N1 testing.
All samples having detectable 2009 novel H1 gene were also positive for matrix gene using Xpert Flu A; and all samples with any positive result on Xpert Flu A were confirmed as positive using the CDC assay.
Using the CDC test as the gold standard, the GeneXpert assay detected influenza A matrix gene in 40 of 44 positives, for 91 percent sensitivity; and detected 2009 novel H1 gene in 31 of 44 positives, for 70 percent sensitivity.
Overall, Xpert FluA was able to detect influenza in all DFA-positive samples and some DFA-negative samples, but "showed limited ability to detect 2009 novel H1N1 influenza in DFA-negative, [Luminex] RVP-positive samples," the researchers wrote in their paper.
In an e-mail this week to PCR Insider, corresponding author Steve Miller of the UCSF Department of Laboratory Medicine noted that the Cepheid test "did perform well in our trial;" but has some "distinct advantages as well as drawbacks" when compared to other testing methods.
"The performance was superior to DFA for sensitivity and the specificity was excellent," Miller said. "Use of the Cepheid assay would have identified a significant number of patients in our institution as having influenza A, as our routine testing was done by DFA. When you keep in mind that many clinical labs are currently using antigen assays, which have poor sensitivity for influenza detection, the Cepheid assay is quite a large improvement."
However, Miller also noted that his group's study "shows some limitations in sensitivity for the Cepheid assay when compared to the Luminex RVP test. I would consider a gold standard PCR test to detect [greater than] 95 percent of true positives," and as such, the CDC assay can be considered the gold standard.
Miller also pointed out that the version of the Cepheid assay used in the study was approved through the US Food and Drug Administration's Emergency Use Authorization, which is no longer applicable (PCR Insider, 1/7/2010).
Future versions of the Xpert Flu A test "may have improved sensitivity and will have to go through full FDA approval before the product is available to clinical labs," Miller said.
Cepheid could not be reached for comment. However, the company has said that it is developing an expanded influenza panel that will test for influenza A subtypes H1, H3, and 2009 H1N1, and influenza B. The company has also said that it plans to submit a 510(k) application for that test this year. The current status of that filing is unknown.
Regardless of whether Cepheid improves the performance of its flu test, the fact that it runs on the GeneXpert system gives it a leg up on the competition, according to Miller.
"The GeneXpert system distinguishes itself from all other commercially available, FDA-approved PCR assays by its ease of use," Miller said. "The sample is added to the test cartridge, along with reagents … and the result is determined automatically by the system.
"The cartridge itself is never opened, which reduces risk of contamination by PCR amplicons," Miller added. "Lab technologists do not need to be familiar with complicated PCR techniques, so it can be run evenings and weekends. Importantly, individual tests are run rather than batches, reducing turnaround time. It is not an exaggeration to say that our technologists love the [GeneXpert] test system."
Other researchers contributing to the JCM paper are affiliated with the San Francisco Department of Public Health, which "only tests immunocompromised and critically ill patients" using Luminex RVP testing. In those patient populations, other viruses besides influenza are of concern, Miller added. As such, "even a test without perfect sensitivity is very useful, since most cases are detected quickly, and these particular patient populations are worked up further during the course of their hospitalization."
In general, however, Miller said that his group sees Cepheid's test as a "large improvement" over methods such as DFA testing, and that it can be implemented in most laboratories without PCR facilities and expertise.
"It does not obviate the need for reference tests and well-thought-out clinical algorithms, but is a great improvement over tests performed in many labs today," he concluded.