H5N1 Mammalian Transmissibility Research - U.S. Senate ...
www.hsgac.senate.gov/.../2012-04-26-inglesby-testimony-biol...
File Format: PDF/Adobe Acrobat - Quick ViewApr 26, 2012 – Will engineering novel H5N1 mammalian-transmissible viruses help us improve surveillance for avian flu? Everyone in the flu community can ...
http://www.hsgac.senate.gov/download/2012-04-26-inglesby-testimony-biological-security
Mammalian-Transmissible H5N1 Influenza: Facts and Perspective
+ Author Affiliations
- Address correspondence to Michael T. Osterholm, mto@umn.edu.
Abstract
Two recently submitted (but as yet
unpublished) studies describe success in creating mutant isolates of
H5N1 influenza A virus
that can be transmitted via the respiratory
route between ferrets; concern has been raised regarding human-to-human
transmissibility
of these or similar laboratory-generated
influenza viruses. Furthermore, the potential release of methods used in
these studies
has engendered a great deal of controversy
around publishing potential dual-use data and also has served as a
catalyst for
debates around the true case-fatality rate of
H5N1 influenza and the capability of influenza vaccines and antivirals
to impact
any future unintentional or intentional release
of H5N1 virus. In this report, we review available seroepidemiology data
for
H5N1 infection and discuss how case-finding
strategies may influence the overall case-fatality rate reported by the
WHO. We
also provide information supporting the position
that if an H5N1 influenza pandemic occurred, available medical
countermeasures
would have limited impact on the associated
morbidity and mortality.
Perspective
Life science is currently at a critical crossroads. Two recently submitted manuscripts to Science and Nature
raise serious questions regarding research censorship and a grave
concern for global biosecurity, biosafety, and public health.
These two studies describe success in creating
mutant isolates of H5N1 influenza A virus that can be transmitted via
the respiratory
route between ferrets. The National Science
Advisory Board for Biosecurity (NSABB) was asked by the U.S. government
to assess
the dual-use research implications of these
manuscripts. The NSABB review concluded that a significant potential for
harm
existed in fully publishing the methods and
results; thus, the NSABB recommended that details of the work not be
fully communicated
in an open forum (1, 2). However, the World Health Organization (WHO) has stated a preference from a public health perspective for full disclosure
of the information in these two studies (3).
The recommendation from the NSABB has resulted in a flurry of commentaries regarding both the merits and flaws of this decision
(4–15). There have been calls for “solid science and not speculation” and for decisions to be based on “sound scientific principles”
(12, 14).
Some of these same commentaries have provided selective data, not the
entire body of published literature, to address specific
issues related to the NSABB decision (12, 14).
The issues are (i) the determination of the human case-fatality rate
for H5N1 infection and its implication for a future
possible pandemic and (ii) the capability of our
current influenza vaccines to impact any future unintentional or
intentional
release of H5N1 virus.
H5N1 VIRUS INFECTION CASE-FATALITY RATE
The WHO reports that 345 of the confirmed 584 cases of human H5N1 infection have died; the cumulative case-fatality rate is
59% and ranges from 30 to 80% depending on the country (16). One group of investigators recently concluded that “the case-fatality rate that is offered by the WHO, and that is driving
this controversy, is likely orders of magnitude too high” (14). Their conclusion was based on 10 published studies of H5N1 seroepidemiology in potentially exposed persons. The WHO has
established criteria for the serologic identification of H5N1 in humans (17). Using literature search strategies similar to those of a previous review of H5N1 seroepidemiology (18), we identified 24 studies published to date that evaluate the seroepidemiology of H5N1infection in humans (19–42).
This analysis includes three follow-up studies related to the 1997
outbreak of H5N1 influenza in Hong Kong; therefore,
we excluded those three studies from further
analysis since current H5N1 viruses are not similar to the strain that
caused
that outbreak (43). This is consistent with the findings and actions of the WHO: the 1997 cases are not currently included in the case count
for H5N1 infections, and the 1997 isolate is not recommended for inclusion in current H5N1 vaccines (16, 44).
The remaining 21 studies were conducted and published after 2004; 13
meet the WHO criteria for serologic confirmation (titer
of ≥1:80) (19, 24, 26, 27, 33–41). Three of the
13 studies reported serologic evidence of H5N1 infection (24, 39, 41). In total, 26 (0.47%) of 5,487 participants in these studies were seropositive for H5N1.
All 13 studies that used the WHO
serologic screening criteria were conducted within 4 months of the
occurrence of human H5N1
cases or within 6 months of H5N1 poultry
outbreaks in the area from which participants were enrolled. The timing
of the participant
surveys maximized the possibility that these
studies would detect recent H5N1 infections had they occurred (45). Most of the individuals tested were exposed to sick poultry and/or a symptomatic human case of H5N1 infection.
One seroepidemiology study of H5N1 infection in Thai villagers conducted by Khuntirat et al. (29) has been cited as documentation of an increased rate of H5N1 subclinical infections(12, 14).
In that study, approximately 6% and 3.5% of participants were reported
to have elevated antibody levels to one of two H5N1
viruses, respectively. Serologic samples were
obtained from villagers more than 2 years after sporadic H5N1 outbreaks
were
reported in poultry and one confirmed and two
possible human H5N1 cases occurred in the area. This study was not
included
in the 13 studies detailed above because
serologic results did not meet the WHO criteria for serologic
confirmation; the investigators
used a low threshold antibody titer (≥1:10) as
evidence of previous infection. This study has been criticized, since
using
such a low threshold could lead to
overestimation of the true seroprevalence because of the increased
likelihood of false-positive
results (46, 47). In addition, the study was not initiated until 2 years after H5N1 infections in humans or poultry had been reported in
the area.
Of the 13 studies that used the WHO criteria, five reported the range of serologic titers detected (26, 27, 33, 34, 37).
These studies were conducted within 4 months of the occurrence of human
cases and within 6 months of poultry outbreaks.
The five studies, which included 2,629
participants, found that no participants had evidence of previous
infection based on
the WHO criteria for serologic confirmation and
only 13 (0.49%) had neutralization titers between 1:10 and 1:40. One
study
documented that most participants had detectable
H5N1 titers below 1:80; however, a reevaluation of the dilutions used
in
the study led the authors to conclude that these
serologic titers did not represent detectable antibody (33;
J. Katz, personal communication). The results of these five studies,
which involved sampling of participants more contemporary
to evidence of circulating H5N1 viruses than the
study by Khuntirat et al., demonstrated at least a 10-fold-lower
prevalence
of intermediate serologic results (i.e.,
neutralization titers between 1:10 and 1:40) than to those found in the
latter study.
We believe that these data support the concern
regarding false-positive results in the Khuntirat et al. study (29).
Some researchers have stated that,
because of the specificity of the WHO case definition, milder or
asymptomatic H5N1 cases
have been missed by traditional case-based
surveillance and therefore a small fraction of the total number of
infected cases
has been accounted for under the WHO
surveillance system (12, 14).
When population-based seroepidemiology studies are used to supplement
clinical-based surveillance, a more complete picture
of the epidemiology of that infection is
generated than that by use of clinical case-based surveillance alone.
Mild or asymptomatic
cases can be detected by serologic evidence of
prior infection even if such cases are missed by traditional
surveillance at
the time of their infection. Another case
detection strategy is to follow exposed persons over time to identify
any influenza-like
illnesses in such groups. Exposed persons can
include persons with known exposure to confirmed or suspected human H5N1
cases,
persons with occupational or household exposure
to sick birds, or persons living in the same locations as human and/or
avian
cases. Periods of observation of exposed persons
by health authorities have lasted for up to 6 months. To date, this
type
of targeted surveillance has not uncovered
additional cases of mild influenza-like illness caused by H5N1 infection
(48).
As with any surveillance system, case ascertainment for H5N1 infections
certainly has not captured 100% of the cases. However,
all of the data presented above suggest that the
number of mild infections that have been missed is likely relatively
small.
Up to this point, we have discussed
the likelihood of missing mild cases of H5N1 infection; however, it is
also important
to consider the potential to miss fatal cases of
H5N1 infection. Investigators have shown that current case-based
surveillance
for H5N1 infection in countries with ongoing
avian H5N1 transmission does not always document fatal cases, with such
cases
being missed either because the diagnosis was
not considered or virologic confirmation was lacking (46, 47, 49).
A more complete ascertainment of fatal cases of H5N1 infection would
increase the current H5N1 case-fatality rate. While
this is also likely an infrequent occurrence,
the phenomenon of missing cases of fatal infectious illnesses has been
documented
in other situations. For example, one would
expect that fatal cases of human rabies would rarely escape detection;
however,
instances of previously fatal human rabies cases
have been uncovered only after recipients of their donated organs
subsequently
developed rabies and died (50, 51).
The available seroepidemiologic data for human H5N1 infection support the current WHO-reported case-fatality rates of 30%
to 80% (16). While some have suggested that concern about such high case-fatality estimates was a major factor in the NSABB decision,
such estimates were only one of a number of factors considered by the NSABB (12).
In fact, if H5N1 virus does become a pandemic virus, the virulence (as
measured by the case-fatality rate) could decrease
10- to 20-fold from what is currently documented
and the virus would still generate a more severe pandemic than the 1918
pandemic,
where the overall case-fatality rate was
probably about 2%. Given the global population and the current dynamics
of population
movement around the world, an H5N1 pandemic,
even with a relatively low case-fatality rate, would be a truly
catastrophic
event.
ROLE OF VACCINES AND ANTIVIRAL AGENTS IN MITIGATING AN H5N1 INFLUENZA PANDEMIC
The primary public health response to
an influenza pandemic is a pandemic vaccine. Secondary to the pandemic
vaccine is the
use of antivirals. If an H5N1 strain, regardless
of its origin, becomes readily transmissible between humans and begins
to
spread in the population, it likely will result
in an influenza pandemic. The proposal that viable vaccines and
available
antivirals will make a substantial difference in
the global morbidity and mortality associated with the pandemic is not
supported
by data from the previous three pandemics. The
time required to manufacture both egg-based and cell culture-based
influenza
vaccines has resulted in “too little, too late”
vaccine responses for the 1957, 1968, and 2009 pandemics on a worldwide
scale.
For example, by 28 October 2009, only 16.8 million doses of pandemic 2009 A(H1N1)pdm09 vaccine had been shipped by the U.S.
federal government to states (52).
An ample supply of the vaccine was not available until after the second
wave had subsided in early October; by that time,
demand for the vaccine had dropped dramatically.
The Centers for Disease Control and Prevention (CDC) estimated that the
2009 A(H1N1)pdm09
pandemic vaccine prevented only 200 to 520
deaths in the United States because of delay in availability (53).
Mammalian cell-based pandemic vaccines
were licensed for use in the European Union in 2009 and were used there
during the
pandemic response. As in the United States, both
the egg-produced and the cell culture-produced influenza vaccines
arrived
too late and in too little quantity to have a
significant impact on the pandemic in the European Union. According to
the date
of marketing authorization in Europe, a
mammalian cell-based vaccine was available only after three adjuvanted
egg-based influenza
vaccines were already in distribution (54).
The European experience with the availability of a pandemic cell-based
vaccine did not demonstrate a measurable improvement
in vaccine production speed nor was it
sufficient to alter the overall public health impact of the pandemic in
Europe.
Given the need to distribute pandemic
vaccines globally and the fact that not all countries have the financial
assets to purchase
sufficient quantities of vaccine for their
populations during a pandemic, the WHO coordinates a program for
donation and distribution
of pandemic influenza vaccines. As of 10
November 2010, the last WHO update for pandemic vaccine distribution,
only 78 million
doses of A(H1N1)pdm09 vaccine had been
distributed to 77 countries (55). All of these vaccine doses were distributed well after the second wave of the pandemic, months after developed countries
had started their vaccine campaigns.
From a historical perspective,
influenza vaccine also arrived in quantities too small and too late to
have a significant public
health impact in the United States for both the
1957 and 1968 pandemics. Given the experience of the three previous
pandemics,
unless newer and more effective influenza
vaccine technologies are developed that facilitate substantially faster
production
and generate far greater numbers of doses in
much shorter time frames, it is unlikely that influenza vaccine will
have a significant
public health impact during the next pandemic.
The technology behind our current influenza vaccines is simply not
sufficient
to address the complex challenges associated
with an influenza pandemic in the 21st century.
Currently there are limited H5N1
influenza vaccine stockpiles around the globe, including one in the
United States. These
vaccines are not designed to protect the general
population but are to be targeted to a small subset of critical-asset
individuals,
until a pandemic-specific vaccine can be
produced. Furthermore, it is unclear as to how effective the currently
stockpiled
H5N1 vaccines would be against an emergent
pandemic strain, given the diversity of existing H5N1 clades.
Antivirals are the only other
pharmaceutical intervention available for influenza. As with influenza
vaccines during a pandemic,
the global capacity for antiviral manufacturing
falls far short of global needs. While there are global and national
stockpiles
of antivirals devoted to pandemic response,
during the 2009–2010 pandemic, there were significant disparities
regarding use
and availability of antivirals around the world (56).
Current antiviral stockpiles and antiviral manufacturing capacity
support the conclusion that should another influenza
pandemic occur in the foreseeable future, the
impact of antiviral use on human morbidity and mortality will be no
better than
was documented in the 2009 pandemic.
SUMMARY
In summary, we believe that the debate
about the case-fatality rate of H5N1 influenza in humans and the
suggested important
role of currently available antivirals and
vaccines in mitigating an H5N1 pandemic are without merit. Furthermore,
we do not
believe that continued focus on these issues
helps to address how best to manage research involving influenza
viruses, such
as H5N1, that are transmissible between mammals
and have the potential to be highly virulent in humans. Future
discussions
specific to the current controversy need to
resolve critical questions such as how we safely conduct H5N1 virus
transmission
studies in mammals, how we share critical
methods and results with those who have a need to know, and how we
ensure that laboratory-generated
H5N1 viruses do not escape controlled
environments. Resolution of these issues with regard to H5N1 influenza
viruses has the
potential to serve as a template for similar
situations involved existing or emergent pathogens. It is our belief
that the
current controversy provides a valuable
opportunity for scientists and public policy experts to work together in
creating
this road map for the future.
Notes
Dr. Osterholm is a member of the National Science Advisory Board for Biosecurity. His views expressed here do not represent
the official policy or scientific conclusions of the NSABB.
Footnotes
-
Citation Osterholm MT, Kelley NS. 2012. Mammalian-transmissible H5N1 influenza: facts and perspective. mBio 3(2):e00045-12. doi:10.1128/mBio.00045-12.
- Copyright © 2012 Osterholm et al.
This is an open-access article
distributed under the terms of the Creative Commons
Attribution-Noncommercial-Share Alike 3.0
Unported License, which permits unrestricted
noncommercial use, distribution, and reproduction in any medium,
provided the
original author and source are credited.
No comments:
Post a Comment