Rationales for Man-made H5N1 Experiments Evaporating?

Posted on by Laurie Garrett

With each passing day the rationales in favor of the man-made H5N1 experiments seem to further evaporate. The still-unpublished ferret experiments conducted last year by Ron Fouchier of Erasmus University in Rotterdam and Yoshi Kawaoka of the University of Wisconsin in Madison continue to evoke vigorous debate and controversy. Now that the U.S. National Institutes of Health have asked the National Scientific Advisory Board on Biosecurity (NSABB) to reconvene and again review the work the timetable for public airing of the methods and data is pushed further into the future.

However, the voluntary 60-day moratorium on such research, self-imposed by flu virologists, is theoretically coming to an end. In the February 17 WHO meeting the 21 scientists present agreed to extend the voluntary moratorium, but no clear end date for the hiatus was determined. Will the research soon resume? If so, the public and their governments have a right to ask, “Why?”

In his comments to the American Microbiology Society Fouchier implied that every aspect of his work had been distorted publicly by “misperceptions” created by “the media.” The blame-the-media strategy has most commonly been deployed by politicians caught with their hands in the public coffers, or on somebody else’s (ahem) bodies – it seems rather unseemly that the entire H5N1 debate has descended to this tawdry, and useless, level. One can sympathize with Fouchier and Kawaoka’s essential plight: They are scientists who simply did what scientists do, and need to reach final outcome: Publication. But of course nothing about these experiments, nor their ultimate publication, is now likely to proceed according to “normal science.” My goodness, the Dutch government is now threatening to block further dissemination of information about Fouchier’s work outside the European Union under terms of export control regulations.

To decide what the “new normal” of this science should be it is necessary to address the public health rationales forwarded in defense of conducting the experiments in the first place, and the recently stated reasons for insisting that the experiments are safe, and the work can be fully published without public risk. (I will not address the question of whether censoring parts of the work make any practical sense at this juncture: That is a separate issue about which I have previously written.)

Three principle rationales in favor of deliberately mutating H5N1 to create a mammal-to-mammal transmissible form of the virus have been forwarded by Fouchier, Kawaoka and their supporters. I believe all three are groundless, based on currently public information, or so weak as to crumble when weighed against risks of environmental release of the microbes (unless full disclosure reveals that the man-made H5N1 mutants are noncontagious).

Public Health Rational Number One:  The H5N1 bird flu has circulated in Asia since the mid-1990s, and across a wider swath of the world since 2005 without causing more than a handful of human-to-human transmissions, and no apparent tertiary spread among people. Virologists have identified some three to five nucleotide mutations that seem to make a difference between a form of H5N1 that spread like wildfire among birds, versus a form that can transmit between mammals. The experiments were necessary to determine what it would take for wild H5N1 to pose a serious pandemic threat to humanity.

According to Israeli government and OIE reports H5N1 is now spreading in that country from cat-to-cat. The disease watchdog ProMED has offered translation of the Hebrew media coverage. Three infected, ailing cats were identified by government officials in Moshav Shalva on March 14th. The felines were roaming among infected commercial turkeys, and apparently acquired the virus from the birds.  Officials found stray cats in a nearby village, not in direct proximity to the turkeys that were sick. A mass capture and killing of stray cats is underway, and residents have been instructed to keep pet felines indoors for their own protection. The Ministry of Health is tacking strict precautions to prevent cat-to-human transmission of the virus, including warning that nobody touch ailing or dead animals.

This is not the first report of feline H5N1 infection. A tiger and two leopards in Bangkok, fed infected chickens, died in 2003 of H5N1. Infection spread from tiger-to-tiger in 2004 in Thailand. During the second spread of H5N1 into Europe in 2007 cats died of the disease in Germany and the Netherlands, and pet owners were warned to keep their animals indoors. Though it was assumed the felines acquired H5N1 from contact with ailing birds or avian carcasses, the possibility of cat-to-cat transmission could not then be ruled out.

Feral cats capture and eat birds, so it has always been logical to assume that they would be among the first mammals naturally exposed to the avian virus. Indeed, Fouchier and collaborator Albert Osterhaus so-warned in 2003, noting, “As the H5N1 pandemic evolved, we found the H5N1 virus in domestic cats, tigers, civets, and very recently Chinese pikas, a species closely related to rabbits.”

Since 2003 H5N1 has naturally or experimentally infected at least the following mammals: Homo sapiens, cow, dog, Owston's palm civet, domestic cat/feral cat, cynomolgus macaques, stone marten, stone (beech) marten, ferret, New Zealand white rabbit, leopard, tiger, rat (Rattus norvegicus), pig (both Sus domesticusand Sus scrofa), red fox, raccoon and house mouse (Mus musculus).

If the considerable resources devoted to these ferret experiments – including the scientific expertise of hundreds of vigorously debating virologists and epidemiologists – were now turned to the Israeli feline outbreak the questions of mammalian transmission might possibly be answered – without necessitating deliberate synthesis of a lethal virus.

Public Health Rational Number Two:  Mankind will be protected from a dangerous pandemic by advance stockpiling of vaccines directed against such man-made H5N1 mammalian transmissible strains.

The notion that the global pharmaceutical industry would be interested in mass production of so-called pre-pandemic H5N1 vaccine was first introduced at the Pacific Summit of 2007 by then-VP of the Gates Foundation, Tachi Yamada.  In the Yamada scheme an advance market purchasing plan costing governments, the Gates Foundation and other unidentified donors would put up about $12 billion. The pharmaceutical industry would respond by manufacturing 7 billion doses of vaccine directed against a to-be-selected strain of circulating bird virus. The possible clout of the vaccine in the face of what would undoubtedly be a genetically different pandemic strain would be enhanced with recent innovations in adjuvants, giving far greater immune system power to responses to vaccination. Said stockpiles of vaccine would be stored in secure, hygienic, refrigerated locations around the world until WHO formally decreed that pandemic H5N1 had struck. At that magic moment, millions of still-unidentified healthcare workers would deploy worldwide to vaccinate 7 billion people.

Despite the tantalizing (though admittedly theoretical) $12 billion, the pharmaceutical industry did the math and walked away from the scheme. Industry scientists were skeptical that an eventual pandemic strain would bear enough antigenic similarity to currently circulating H5N1 to generate cross-reactive immunity. In other words, they weren’t convinced the vaccine would work. Moreover, nobody can predict when such a vaccine would be needed, and pharmaceutical companies realistically assumed they would end up stuck with billions of doses of vaccine, forcing construction – at their expense – of appropriate repositories, staffed by security and hygienic monitoring personnel. Because production on the scale of 7 billion doses – a scale that is 2,000 times larger than anything the industry has ever managed to produce worldwide in a single year -- would require 100% dedication of flu vaccine facilities, industry also felt that routine seasonal flu production would be damaged, or brought to a full halt, by the pre-pandemic effort, cutting into immediate public health efforts and profits. It made no sense. And despite significant investment by the U.S. government, manufacturers have signaled readiness to make on cue, but not to massively stockpile.

There have been no significant changes in the vaccine production situation since Yamada’s proposal. Though new start-ups for production have emerged in China, India and Indonesia, their contribution to global supply would be trivial. Moreover, routine vaccine campaigns around the world are already feeling the crunch from thepaucity of skilled personnel capable of executing all aspects of vaccine supply, distribution and administration. Worse, discussion of pre-pandemic vaccine has spawned an array of conspiracy talk, claiming the H5N1 is a false threat created to reap pharmaceutical profits. The opposite is accurate: industry comes to the pre-pandemic vaccine table very reluctantly, offering gestures but showing no interest in creating billions of doses.

This public health rationale can be dispensed with: It bears no significance in the real world of global pharmaceutical production.

Public Health Rational Number Three:  Knowing which genetic mutations to watch for will allow global surveillance efforts to hone in, sounding the alarm when those specific ones emerge.

There are two significant problems undermining this assertion. First, only supreme hubris would allow public health and epidemiology experts to conclude that a given set of five base pair mutations experimentally shown to cause mammalian transmission represents the only constellation of genetic tricks that H5N1 could play on humanity. Narrowing surveillance to a given genetic search could well leave mankind building nuclear weapons and countermeasures, only to be shattered by hijackers using commercial jets as missiles.

Worse, the notion of genetically based surveillance assumes that anybody, in any of the currently endemic regions of H5N1 spread, deploys real time genetic surveillance for flu. They do not. Genetic analysis of influenza strains is generally conducted in a haphazard manner, reflecting the interests of local research virologists and access to serological and strain samples. Such access has been highly limited and irregular. Yes, there are diagnostic kits in existence, but they have not been commercially manufactured for deployment in poor countries, and are not in serious use. There is no concrete, practical relationship at this time between the genetic virology work on H5N1 and public health disease control efforts.

Making a real time, pragmatic bridge between genetic virology and outbreak control decisions, especially in poor countries, would require construction of many laboratory, diagnostic, and veterinary capabilities, coupled with policy changes and financing, none of which currently exist. While having such capacity in place may constitute a sound aspirational goal, it is disingenuous in the extreme to argue that anything found in a laboratory in 2011 can realistically form the basis of real time H5N1 control measures in 2012, 2013, 2014, 2015…..and so on.

 When these public health considerations, in their originally stated form, were weighed by the NSABB against risks of release of the man-made viruses, alarm rose. The NSABB did classic risk/benefit analysis to decide, as requested by the NIH, whether or not the papers should be published. Since the NSABB censorship decision was handed down in December proponents of full publication have offered changes to the safety equations, as follows:

Safety Rational Number One: Ferrets are not actually a terrific model for human transmission of viruses. What happens between ferrets is not predictive of human risk and therefore ought not cause concern.

If, indeed, the ferret is a lousy experimental model, why were the experiments done on ferrets? If the ferret offers no predictive value for human pandemic potential, why should American taxpayers fund any flu research using these animals, ever again?

According to scientists that use ferrets as the primary substitute for humans in influenza work, including the flu team at the US Centers for Disease Control, there are only two options that better predict what will happen when any given genetic strain of H5N1 infects a human being.  Researchers can deliberately infect, and kill, chimpanzees. That option is illegal in most of the world, and immoral in the view of many bioethicists, animal conservationists and primate researchers. Use of monkeys, such as rhesus macaques, faces similar constraints, coupled with exponential increases in costs of experimentation. The second option – using human beings as research subjects – violates countless legal and ethical standards.

So, we come back to the ferret: It is either a useful research model, or it is not. If it is, then the NSABB was right to be worried about the man-made viruses. If it is not, American taxpayers have every right to demand that the NIH cease using their precious dollars in a recession to fund bogus research.

Safety Rational Number Two: Using a statistical technique called metanalysis, pooling divergent data collected by varied scientists, based on different techniques and serology standards, Dr. Peter Palese argues that H5N1 is not a particularly dangerous virus. According to Palese’s team, whose lab collaborates with Fouchier’s, tens of thousands of people have been harmlessly infected with the virus since 1997, with perhaps as few as 1 percent of them becoming ill. Palese argues that H5N1 is no more dangerous to human beings than seasonal flu, so release of a highly contagious form of the virus would pose no more danger to humanity than perhaps the 2009 swine flu epidemic did. “Meta-analysis shows that 1 to 2% of more than 12,500 study participants from 20 studies had seroevidence for prior H5N1 infection,” the paper states.

Epidemiologist Mike Osterholm of the University of Minnesota has presented a detailed critique of the metanalysis Palese performed. (WHO says 58.9% of people identified as having been infected with H5N1 since 2003 have died as a result, but this data is mostly based on studies of people that have taken ill.) Most of the pandemic risk analysis for H5N1 laid out by University of Hong Kong scientists in 2007 stands up today, including the biology of mammal-to-mammal transmission and rates of asymptomatic infection in people.

A more immediate and persuasive response would be close serology examination of any of the 20+ outbreaks of H5N1 now active worldwide. Pick one in Egypt, Indonesia, Hong Kong, Vietnam, Cambodia, India, Nepal, Bangladesh – there are many options. Identify an infected region, marketplace, person or flock of chickens, and conduct standard ring surveys using a well standardized serology assay. Test hundreds, maybe thousands, of people located in ever-wider circles outward from direct exposure to the infection source. See how many people turn up antibody-positive for infection without displaying symptoms.

Guess what: The experiment has just been done, in Bangladesh. (We await Palese’s response.) The US CDC and its counterpart in Bangladesh surveyed 89 farms that had H5N1 outbreaks between late 2007 and mid-2009, and three live animal markets in Dhaka where chicken outbreaks had occurred in 2008. More than 400 poultry workers from those farms and the market were tested: ZERO asymptomatic infections were found. (Citation for skeptics: Nasreen S, Khan SU, Azziz-Baumgartner E, et al. Board 183. Sero-prevalence of H5N1 antibodies among poultry workers in Bangladesh, 2008-2009. ICEID program and abstract book, p 116 (Presented Mar 13, 2012)

Having been in the 2008 H5N1 outbreak in Bangladesh and personally seen the conditions in the Dhaka live animal markets and on the farms I find this amazing. As these pictures I shot in the Dhaka market shows, the ecology of intimacy between human workers and live, angry birds is horribly close.

If the Bangladeshi study remains inadequate to convince skeptics, the experiment can be replicated now – there are plenty of outbreaks to choose from. Real observational data is always superior to statistically derived speculation.

Safety Rational Number Three: The experimental methods and results presented by Fouchier in November 2011 to the NSABB, and in February 2012 to a select experts’ meeting at WHO apparently varied significantly enough to compel the NIH’s Dr. Anthony Fauci to request NSABB take a second look. A key point that may, or may not (depending on who you are talking to), have been presented to the NSABB concerned “airborne transmission” of H5N1 between the animals. Fouchier told the WHO gathering that casual cage-to-cage spread from an infected ferret to other animals via sneezing did not cause sickness in the secondary infections. Serious sickness and infection only occurred when he literally stuffed the virus down the animals’ throats, in intratracheal passage. Therefore, Fouchier now says, it is extremely unlikely dangerous infection could be caused by this synthetic virus, were it to escape its laboratory confines. Fouchier blamed “misperceptions spread by the media” for any impression that he had created a dangerous microbe.

Assuming that on the face of it every word of the above is true, and that the august scientists on the NSABB failed somehow to appreciate the difference between “airborne transmission” and intratracheal passage, one wonders why the viral stuffing down the animals’ throats was performed. The answer: experimenters wanted to determine whether the modified virus was capable of locking onto cellular receptors lining mammals’ throats. Wild H5N1 apparently cannot do so, which is why it cannot be passed from human-to-human via coughing. Human infections result in deep lung viral cultivation, which is rarely passed to another human being. The intratracheal demonstration was not, then trivial.

Fauci insists that there were key differences between what Fouchier told the NSABB about these experimental issues, and what he said in Geneva. We cannot, therefore, judge much further until details are clear.

Blaming the media for getting this wrong is, however, inexcusable. As recently as February Fouchier told ABC News on camera that his synthesized virus had killed all of his exposed animals. It was only after the story aired, and I personally questioned the statement, that he qualified the comment with the word “intratracheal”. There is no way ABC News could have been expected to question Fouchier’s original assertion.

NEXT BLOG:  Is al-Qaeda Using Anthrax (again)?

BIRD FLU PIC_20120206095458.jpg