Friday, December 23, 2016 Written by Ellen Jo Baron, Ph.D., D(ABMM), Prof. Emerita, Stanford University Director of Medical Affairs, Cepheid

Influenza today: Still a hot topic on many fronts

It seems like just yesterday that influenza was on everyone’s minds, but it has been almost a year since we featured influenza in an edition of On Demand. The novel influenza A H1N1 strain that caught everyone by surprise in 2009 seems to have become a very minor component of the circulating strains now, except in the Middle East and India, where it still predominates. This season, influenza B is prominent in the Americas and Africa, and influenza A H3N2 comprises half the reported cases in Europe and Australia, and more than 75% of strains in ChinaA. The World Health Organization reported that the most common strain worldwide, A(H3N2), is again targeting the traditional at-risk age groups of >60 years and <2 years. In temperate South America, numbers of influenza cases began to increase in May, peaking around July with the largest numbers seen for A(H3N2), but showing a second peak in August with more influenza B strains and untyped influenza A strains Figure 1. There was a sharp drop-off in September, heralded by the disappearance of almost all influenza A reports. Africa peaked in July, but numbers remained high into September, with influenza B assuming the majority of cases. The United States is just beginning its influenza season, so the epidemiology is not known yet Figure 2. Overall, influenza activity is lower than historical levels. This may be a result of increasing numbers of Americans receiving their annual flu vaccine. However, a new virus, variant H3N2 (H3N2v), associated with pigs and originally discovered in 2011, has cropped up this year in some human outbreaks periodically since July, 2012.1,B Although the strains of influenza circulating globally now are genetically slightly different from those in the current vaccine, the CDC feels that there will be significant cross-reactive protection, so they recommend that the vaccines not be changed. The common viruses causing disease are generally susceptible to both neuraminidase inhibitor antiviral agents oseltamivir (Tamiflu®) and zanamivir (Relenza®). This correlates with the seeming disappearance of the previously circulating H1N1 strain (called “seasonal”) that was resistant to oseltamivir. Of course, occasional resistance can arise and patients who fail to improve after a week of therapy should be evaluated for drug resistant strains, as well as for other complications. Zanamivir is not recommended for patients <7 years old or for those with underlying respiratory disease. The CDC treatment guidelines state that antiviral therapy should not be delayed while waiting for diagnostic test results when clinical indications suggest influenza and antiviral treatment is indicated.C

The current influenza activity is interesting to public health authorities and of paramount importance to individual patients and their caregivers and those close to them, but the topic of most discourse in the literature this year is the ethics of performing certain types of research involving influenza. The debate over this activity rose to the highest levels of academia and the national science community. Some readers may remember back in 1996 when reports started trickling out of China about a new strain of avian influenza that had a surprisingly high mortality rate in chickens, H5N1. In 1997 in Hong Kong there were 18 cases of the same virus infecting people who were in contact with sick poultry; and most unusual for influenza, one-third of the patients died. No one was particularly alarmed because all patients had contact with poultry and it was not thought that such viruses could transmit an infection from human to human, and there were no more cases reported for several years. But in 2003, the virus resurfaced in Hong Kong.

After that, cases began to spring up elsewhere and in other species of animals. A zoo in Thailand fed fresh chicken carcasses to two leopards and two tigers, all four of which died of fulminant disease in a short period of time. Poultry in both Korea and Vietnam were dying and influenza A H5N1 was identified as the culprit. Human cases resulting in many deaths in Vietnam began and continued, caused by the same virus. By 2004, the virus was widely disseminated throughout Southeast Asia. In early 2004, 9 million poultry were culled in China to stem the epidemic. A study published in January, 2005 reported the first well-documented case of human to human transmission, in which a young Thai girl passed the infection to her mother. Sadly, the mother died.

In 2005, a huge die-off of wild birds in a large lake populated by numerous migratory species (Qinghai Lake) was determined to be due to a new, more lethal variant of H5N1. Japan and Philippines also reported disease in poultry and illness seemed to have spread to migratory birds as far away as Russia, Mongolia, and to poultry, pigs, and then humans in Indonesia. Throughout the next few years, cases in humans or animals were reported throughout the world, moving to Europe, India, and other areas. It appeared that children and young adult patients were more susceptible to infection than the elderly and very young patients, in whom influenza is typically more common. Research was initiated to explore the pathogenesis of the virus, now known as Highly Pathogenic Avian Influenza (HPAI) and one study showed that human disease was diverted at least partly because the virus preferentially adhered to epithelial cells deep in the lungs rather than those cell types that it is most likely to encounter in the upper respiratory tract. By June, 2006, there had been 205 laboratory-confirmed human cases reported to the World Health Organization. By spring of this year (2012), several hundred human cases had been reported worldwide. WHO publishes a timeline of events that is updated often.D A major human outbreak has not occurred, but many people are concerned, and some scientists are studying this virus with new technological and molecular tools.

Two specific groups of researchers, one collaboration between the University of Wisconsin and several Japanese institutions and the other a collaboration between the National Institutes of Health, the Erasmus Medical Center in Netherlands and the University of Cambridge in UK, endeavored to determine if the current widely circulating strains of HPAI could acquire the genetic determinants that would allow the virus to infect humans more easily, including factors that would facilitate binding to upper respiratory tract epithelial cells. Such ability would surely contribute to the possibility of a great pandemic of lethal influenza, resembling that of 1918. The virus of the 1918 pandemic was similar to this virus, in that it preferentially affected relatively healthy children and young adults. In fact, a number of genetic characteristics of the HPAI H5N1 resembled those of the 1918 strain.2 Those scientists who were studying the characteristics that would allow inter-species transmission by the airborne route had chosen ferrets as the animal model for their research. Ferrets are often used for this type of respiratory virus research because they can develop a respiratory infection similar to human influenza; in fact, ferrets sneeze just like people do. The experiments were carried out with utmost care under rigorous scientific protocols. Mutations that developed in the viral strains grown in the laboratory were carefully controlled and characterized. Finally, the two independent groups of scientists were able to modify A/H5N1 enough to allow ferret-to-ferret respiratory transmission. Does this mean that those genetic changes could occur naturally in nature? Does it mean that even if such changes were to occur, that they would behave the same way in humans? Nobody knows the answers, but the prospect is disconcerting. Both manuscripts were submitted for publication at the same time to Nature and Science, among the most prestigious and widely read scientific journals in the world. These manuscripts were to generate a huge amount of public and private dialogue and controversy.3

After 9/11 and the anthrax distribution in the U.S., many scientists and government officials felt that the information necessary to weaponize anthrax spores could have been found in legitimate scientific literature (much like the information on how to build a bomb is available on the internet), and they wanted to censor any future publications that could be seen as helping potential bioterrorists achieve their goals. This type of research is called Dual-Use Research of Concern (DURC), meaning that both peaceful and military applications are possible based on the results. The academic and government communities agreed that bioterrorism was a threat sufficient to warrant serious concern and put an official oversight function in place. The government created an advisory board, the National Science Advisory Board for Biosecurity (NSABB), peopled by eminent and respected scientists, whose job was to review publications of research findings that could lead to potentially dangerous information becoming available and being exploited for the wrong purposes. The Board’s focus was to decide how to protect the public from harm while not inhibiting the advancement of science.

Clearly, the two submitted manuscripts describing the genetic manipulations of the HPAI virus that allowed it to infect ferrets by the airborne route merited review by this board. After much deliberation, the NSABB decided that the full protocols should be published and available to the general public. This decision provoked an outcry from some scientists who felt that potential security risks outweighed the free dissemination of research findings. The supporters of total publication felt that partial publication (holding back key experimental details) would have a detrimental effect on other scientists contemplating similar research. They felt that the information presented could potentially lead to better control, and perhaps a better vaccine. The final upshot of this highly vocal and visible rift within the influenza research community and the broader scientific community was the publication, in full, of both manuscripts.2,4 Amidst many calls for better decision-making in the future, the safeguards and oversight activities have been modified.5,6,E Still, at this time, the HPAI has not moved into the general population. Could our currently available rapid influenza tests detect H5N1 if it suddenly swept into the human population? At least one group studied two lateral flow antigen assays with samples from infected poultry.7 Rectal/genital (i.e., cloacal) swabs were less sensitive for detection of HPAI H5N1 than feathers and neither sample was very sensitive compared with a molecular assay. Cloacal samples are appropriate, since infected birds shed the virus via the gastrointestinal tract. However, this result does not bode well for detection of the virus in human samples.

The inevitable emergence of new strain variants fueled by genetic reassortments among existing strains has important implications for diagnostic tests.

Can the rapid antigen tests developed for circulating strains at the time of test creation detect the new variants? Clearly they have challenges, as illustrated by the very public failure of existing rapid antigen tests to detect the influenza A 2009 novel strain8 and the H3N2v strain.9

Both the rapid antigen tests and the antibodies used in the direct fluorescent antibody stains used by those laboratories performing DFA assays were developed before avian influenza was considered to be a threat. Genetic tests have the advantage, as they can be tailored with the addition and subtraction of specific primers and probes to match the circulating strains.

How do the current diagnostic tests fare with the influenza strains currently in circulation? The general laboratory population was surprised to learn that the novel A H1N1 2009 was not detected very well by the commonly used rapid antigen tests. Current recommendations from the CDC about which diagnostic tests to use are based on that experience. First of all, patients should be tested as early in their disease as possible. Test results are less reliable at 4 days after onset of illness. The CDC recommendations state that rapid influenza diagnostic tests (RIDT) are not very sensitive, averaging sensitivities from 40-70% with a range of 10-80%.F A negative test should be confirmed with a more sensitive test, such as reverse transcriptase PCR or viral culture. These tests are more sensitive, but slower. Although a truly rapid test (15-20 minutes) that could unequivocally rule out influenza would be ideal, the state of our technology still falls short of that goal. Whether the patient presents at the height of influenza season or in the middle of the summer, major decisions will be made based on the test results. Should the patient be admitted into a respiratory isolation room? Should the patient be sent home? Should the patient be given antiviral medication or antibiotics? An accurate test result in a timely fashion is important to that individual patient, his or her family, and the institution. Fortunately, there are options available for the laboratory, although all have pros and cons.

Some laboratories continue to use rapid antigen tests, usually immunochromatographic lateral flow format. A group from the Stanford University Medical School laboratory of Dr. Benjamin Pinsky, led by first author Dr. Mike Dimaio (Figure 3), has recently published an evaluation of two of the most popular such assays, BinaxNOW and BD Directigen, along with Cepheid Xpert Flu for their sensitivity and specificity with today’s circulating virus strains.10 The Stanford laboratory, with its highly skilled and experienced virology staff, routinely used direct fluorescent antibody (DFA) testing and its own laboratory-validated real-time reverse-transcriptase PCR assays for influenza A (based on the CDC published protocol) and the primary mutation that confers oseltamivir resistance. In 2008, the laboratory had performed a comparison of the rapid direct antigen tests that it had been using in previous years and found the sensitivity and specificity to be so low (unpublished data) that it had discontinued use of those tests even before the novel H1N1 virus outbreak began in the spring of 2009, instead ramping up the number of times per day that DFA testing was performed. But physicians wanted a more rapid turnaround time. The group used 200 previously submitted samples (frozen) including 84 from children

Novak-Weekley and colleagues (Figure 4) from five institutions in the U.S. and Australia provided another evaluation of Cepheid’s new Xpert Flu test.11 Both nasopharyngeal swabs and nasal aspirates/washes, prospective and previously frozen, were included. More than 1500 samples were included, divided almost equally between fresh and frozen. The ProFlu+ molecular assay, viral cultures, and sequencing of viruses were all used as reference tests. Compared to the ProFlu+ molecular assay, frozen nasal wash samples tested by Xpert Flu had sensitivities of 98-100% for all influenza A and B viruses. For nasopharyngeal swabs, there were somewhat lower sensitivities (93.8%) for influenza B containing samples, although the results for influenza A were comparable to those seen with washings. Specificities were 99-100% for both aspirates and swabs. The prospective samples had 100% sensitivities for all samples except for nasal washings, for which slightly lower sensitivities were seen with seasonal influenza A, not including the 2009 H1N1 strains. The PCR assay performed better than culture, after discrepant analysis by sequencing. Novak-Weekley and colleagues noted that the ease of use, the rapid turnaround time, and the random access nature of the assay, which helps diminish the chance of cross-contamination, all serve to make the Xpert Flu an attractive test for laboratories that lack molecular expertise or that require rapid turnaround time and do not wish to wait to accumulate a batch of samples for testing.11

The recently newsworthy influenza A H3N2v, another pig to human influenza virus, was probably not included in either of the Xpert Flu publications described above. This virus has become a risk especially for patients who tend pigs for prolonged periods, and who frequent county fairs. H3N2v is likely not well detected by any current FDA-cleared tests. In fact, a recent study shows that a number of variant viruses were not detected by current rapid antigen tests at all.12 So far, these viruses have all been detected by the current GeneXpert assay, including the H5N1 highly pathogenic strain, which is reported as "influenza A." Cepheid scientists are working now on the next generation influenza assay, which embodies our commitment to continuously improve our assays, and in the case of influenza, to try to keep up with this wily virus as it evolves.

The flu season has started, and laboratories have several choices of tests to employ. Knowing the pros and cons of the options can help microbiologists and physicians choose the best tests and testing algorithms for their needs. The case report in the next section of this edition of On Demand describes one laboratory’s approach.


1. Lindstrom, S., et al., Human infections with novel reassortant influenza A(H3N2)v viruses, United States, 2011. Emerg Infect Dis. 18(5): p. 834-7.
2. Imai, M., et al., Experimental adaptation of an influenza H5 HA confers respiratory droplet transmission to a reassortant H5 HA/H1N1 virus in ferrets. Nature. 486(7403): p. 420-8.
3. Osterholm, M.T. and D.A. Relman, Creating a mammalian-transmissible A/H5N1 influenza virus: social contracts, prudence, and alternative perspectives. J Infect Dis. 205(11): p. 1636-8.
4. Herfst, S., et al., Airborne transmission of influenza A/H5N1 virus between ferrets. Science. 336(6088): p. 1534-41.
5. Fauci, A.S. and F.S. Collins, Benefits and risks of influenza research: lessons learned. Science. 336(6088): p. 1522-3.
6. Lipsitch, M., et al., Evolution, safety, and highly pathogenic influenza viruses. Science. 336(6088): p. 1529-31.
7. Slomka, M.J., et al., Evaluation of lateral flow devices for identification of infected poultry by testing swab and feather specimens during H5N1 highly pathogenic avian influenza outbreaks in Vietnam. Influenza Other Respi Viruses. 6(5): p. 318-27.
8. Control, C.f.D., Performance of rapid influenza diagnostic tests during two school outbreaks of 2009 pandemic influenza A (H1N1) virus infection - Connecticut, 2009. MMWR Morb Mortal Wkly Rep, 2009. 58(37): p. 1029-32.
9. Centers for Disease Control and Prevention. Evaluation of Rapid Influenza Diagnostic Tests for Influenza A (H3N2)v Virus and Updated Case Count - United States, 2012. Morbidity and Mortality Weekly Report, 2012. 61: p. 1-3.
10. Dimaio, M.A., et al., Comparison of Xpert Flu rapid nucleic acid testing with rapid antigen testing for the diagnosis of influenza A and B. J Virol Methods. 186(1-2): p. 137-140.
11. Novak-Weekley, S.M., et al., Evaluation of the Cepheid Xpert Flu Assay for rapid identification and differentiation of influenza A, influenza A 2009 H1N1, and influenza B viruses. J Clin Microbiol. 50(5): p. 1704-1710.
12. Balish, A., et al., Analytical detection of influenza A(H3N2)v and other A variant viruses from the USA by rapid influenza diagnostic tests. Influenza Other Respi Viruses.