Does Asymptomatic #H5N1 Influenza Infection Occur?

Center for Biosecurity of UPMC
Clinician's Biosecurity News
By Amesh A. Adalja, MD, FACP, December 2, 2011

Asymptomatic influenza virus infection during regular flu seasons is very common, but with the highly pathogenic H5N1 avian influenza virus, it is thought to occur only infrequently. Because so little asymptomatic infection has been reported, the seroprevalence of H5N1 is believed to be 0%–12%, with an associated case fatality rate of approximately 60%.1

However, results of a new study by Powell and associates suggests that the standard hemagglutination inhibition assay (HAI) used to detect H5 antibodies is not very sensitive, which means that the established seroprevalence may reflect underestimation, and case fatality rates may reflect overestimation. The researchers studied an alternative approach that quantified the amount of interferon-gamma secreted by peripheral blood mononuclear cells (PBMCs, ie, T cells) in response to H5 hemagglutinin.1 The results suggest the potential need to update our understanding of how to gauge the immune response to the H5N1 influenza virus.
Study Design

The researchers established a study cohort composed of households of potentially exposed persons in an area of Vietnam where there were cases of both human and poultry H5N1 infection. Cohort members provided blood and plasma samples. For comparison, the researchers also established a recovered patient group (N=19) and a healthy control group (N=271).

PBMCs were tested using an ELISPOT assay for interferon-gamma secretion in response to H5, H3, and H1 hemagglutinin, as well as to N1 and N2 neuraminidase, and internal proteins of H3N2. Horse erthryocyte HAI assays, a more sensitive form of HAI, were also performed.1
20% Response to ELISPOT Assay

ELISPOT testing was conducted on 747 samples; responses to H5 HA occurred in 36 (4.8%). Of those positive responses, 24 (3.2%) were specific or dominant responses to H5, as indicated by a 2-fold greater response when compared to H1 and H3 responses. An additional 111 donors responded to H5, but at a level equal to or lower than the response to H1 and H3.1

Taken together, results indicate that 20% of the exposed population responded by ELISPOT to H5N1.2 None of the 271 healthy controls exhibited any response to the H5 HA, while 16% of recovered individuals did respond. As measured by the horse HAI, H5 seroprevalence was found in 37 (5%) members of the community cohort. Of these 37, only 4 had specific H5 ELISPOT results.1,2 The discordance between the HAI and ELISPOT results indicates that measurement of both arms of the immune system (humoral and cell) is needed to capture the full breadth of the response.
Antibody Response May Underestimate Exposure

The results of this study illustrate that when comparing H5N1 seroprevalence data to T-cell assays, a discrepancy of up to 15% in response occurs. Furthermore, the lack of overlap between those with antibody positivity and ELISPOT positivity calls for the use of both methods to ascertain asymptomatic infection.

If it is the case that asymptomatic H5N1 infection occurs to a greater degree than has been thought, then one major implication of this work may be an adjustment to the high case fatality rate that has been associated with H5N1 infection. It could well be that the high case fatality rate is the result of underestimation of the true prevalence of H5N1 infection. A lower case fatality rate for H5N1 could have a significant effect on pandemic planning. Clearly, robust data about total case counts is needed to establish an accurate parameter for case fatality. Another important implication of these results may be increased understanding of H5N1 pathogenicity. If we are able to determine why some people remain asymptomatic, then we may gain new insights into host immune factors that could be exploited in vaccine strategies.
References

Powell TJ, Fox A, Peng Y, et al. Identification of H5N1-specific T-cell responses in a high-risk cohort in Vietnam indicates the existence of potential asymptomatic infections. J Infect Dis 2011; [epub ahead of print] http://jid.oxfordjournals.org/content/early/2011/11/10/infdis.jir689.short?rss=1. Accessed December 1, 2011.

Epstein S. T-cell immune responses and asymptomatic H5N1 influenza infection. J Infect Dis 2011; [epub ahead of print] http://jid.oxfordjournals.org/content/early/2011/11/10/infdis.jir692.short?rss=1. Accessed December 1, 2011.