New test for flu

New test may enable more labs to subtype flu viruses

Robert Roos News Editor

Aug 29, 2006 (CIDRAP News) – Government-funded researchers say they have developed a test that may enable more laboratories to fully identify influenza viruses than is possible with existing tests.

The FluChip test involves a microarray of genetic material—hundreds or thousands of short sequences of known DNA or RNA deposited on a microscope slide, according to a news release from the Centers for Disease Control and Prevention (CDC). The sequences are used to "capture" matching sequences from viruses in samples gathered from patients.

In a study reported in the Journal of Clinical Microbiology, the test determined the type and subtype of more than 70% of 72 flu viruses, including H5N1 avian flu, in less than 12 hours. The study was led by Kathy L. Rowlen, PhD, of the University of Colorado, with assistance from the CDC and funding from the National Institute of Allergy and Infectious Diseases (NIAID).

The test doesn't require the same level of biological security as other tests used to gather detailed information on flu viruses, including the subtype and clade, or family, according to the CDC. That means more labs can use the test.

"Because the sample is inactivated and you're not amplifying the virus, you don't have to use a BSL-3 [biosafety level 3] lab to do the diagnosis," Dr. Nancy Cox of the CDC, a coauthor of the study, told CIDRAP News. The test can be used in BSL-2 labs, a category that includes many clinical labs, said Cox, who is director of the CDC's influenza division.

Rapid tests now used in clinical practice can identify flu viruses in less than 30 minutes, but, at best, they can only detect a virus and determine whether it is type A or type B; they can't identify the subtype, such as H3N2. More sophisticated tests, such as polymerase chain reaction, immunofluorescence, and viral culture, provide more information but are more difficult to conduct and take from several hours to several days, according to the CDC.

The developers of the FluChip test hope to reduce the time required to subtype a flu virus to less than an hour, the news release said. Their journal article says the methods most widely used to subtype flu viruses typically take about 4 days.

The FluChip used in the study contained 55 sequences of RNA representing a variety of type A and type B flu viruses, including H5N1 avian flu and two of the most common seasonal flu subtypes, H3N2 and H1N1. The process for selecting these sequences is described in a companion article in the August issue of the Journal of Virology. The article explains a new method for combing through large amounts of viral genetic information to find sequences that are most helpful for distinguishing between different types and subtypes.

"Our goal was to develop an efficient method for mining large databases to identify regions of the flu genome that are largely the same from strain to strain, as well as strain-specific sequences," Rowlen said in the CDC news release.

To test the FluChip, Rowlen and colleagues used it to analyze 72 unlabeled flu virus samples provided by the CDC, including strains from humans, birds, horses, and pigs. The researchers used RT-PCR to amplify some of the viral genes and then chemically fragmented them. The FluChip then was exposed to a solution containing the gene fragments.

Bits of viral RNA or DNA, when they come in contact with a microarray such as FluChip, will bind to, or be captured by, matching sequences on the microarray, according to the CDC. Researchers then can identify the pathogen by analyzing the pattern of captured targets.

Rowlen and colleagues ran two rounds of tests on the sample viruses. In the first round, they correctly identified the virus type and subtype (type A or B and the hemagglutinin [H] and neuraminidase [N] number) of 66% of the samples. They identified the correct type and the correct H or N number for another 17% of the viruses, and identified only the correct type for 12%.

In the second round, the test correctly typed and subtyped 79% of the samples. The average results for the two rounds were as follows: correct type and subtype, 72% of samples; correct type and partly correct subtype, 13%; correct type only, 10%; false-negatives, 4%; and false-positives, 1%.

About a third of the faulty results occurred because the viruses supplied by the CDC were not covered by the RNA sequences on the FluChip, the article says. Many of the other incorrect results occurred because the attempt to amplify the viral RNA failed, leaving too little to detect.

"We were surprised and pleased at how well the chip performed in these early tests," Rowlen commented in the news release.

"This paper is a proof-of-concept paper that this approach will be quite useful," Cox told CIDRAP News.

"Depending on the design of your capture sequences [on the microarray], you can get more or less information about the viruses," she added. For example, she said, the test could be designed to identify particular strains of the H5N1 avian flu virus, such as subgroups within clade 2. "It’s a very versatile approach to obtaining information about the viruses."

Cox said the test will have to be commercialized to be brought into wide use, and there are private companies interested in it, though she declined to name any. She estimated it will take at least 2 years of development to make the test ready for commercial use.

"If it can be made inexpensively, it would be an ideal tool to use for regular seasonal flu as well as for strains such as H5N1 which are novel to humans," she said. "The aim is to make it competitive with, if not less expensive than, the commercial tests, which cost between about $14 and $20 per test. The aim is to make this less expensive but provide more information."

Townsend MB, Dawson ED, Mehlmann M, et al. Experimental evaluation of the FluChip diagnostic microarray for influenza virus surveillance. J Clin Micro 2006;44(8):2863-71 [Abstract]

Mehlmann M, Dawson ED, Townsend MB, et al. Robust sequence selection method used to develop the FluChip diagnostic microarray for influenza virus. J Clin Micro 2006;44(8):2857-62 [Abstract]