Influenza virus has only eight genes. The molecular structure of the most important proteins they code for is known in intimate detail. The coming and going of its epidemics have been studied by statisticians continually since the 1840s. But predicting pandemics remains a fools’ game. It falls into the category of Alvin Weinberg’s ‘trans-science’ – a question of fact that can be stated in the language of science but is unanswerable by it. Weinberg’s examples focused on the impossibility of predicting the probability of extremely improbable events. There have only been three influenza pandemics in the last century: in 1918, 1957 and 1964. The uncertainty is massively amplified by evolution – the random and frequent genetic mutations and the swapping of genes between bird, pig and human viruses.
For public health planners, trying to predict when a pandemic will occur is not the only problem. Foreseeing what will happen when one starts is nearly as difficult, because the working assumption that a brand new virus will behave in the same way as its predecessors has a good chance of being wrong. One of the big issues right now is whether there will be a second wave of cases next winter, after the current epidemic in North America has run its course, and whether their mortality rate will be higher than at present. This happened in 1918. A popular current explanation is that the virus evolved during the pandemic to become more lethal. But there is no scientific evidence to support this. Only three cases from the 1918 pandemic have been confirmed in the laboratory; all died in the second wave.
Gene sequencing of viruses isolated since the beginning of the current epidemic hasn’t shown any changes, so far. The only major one has been to its name, altered by the US Centers for Disease Control and Prevention (CDC) from ‘swine flu’ to ‘novel H1N1 influenza’ (although the CDC flu specialists have coined the useful shorthand title S-OIV – swine-origin influenza virus). The change signifies its most important property, which gives it pandemic potential: it is different enough from its predecessors to make it unlikely that anyone has good protective immunity against it. The other certainty about the virus is its ancestry. Six of its genes are related to those found in North American pig influenza viruses (which themselves have a mixed ancestry from bird, human and pig viruses) and two have come from pig viruses from Europe/Asia.
This porcine origin has been a big surprise. The UK Pandemic Plan mentions pigs, but only in passing, because its working assumption is that the next pandemic will be caused by a bird virus. Indeed, the plan was prepared in response to the occurrence and spread of the highly pathogenic avian H5N1 virus in poultry in South East Asia since 2003 and the rare but lethal human infections caused by it. ‘A potentially pandemic or pandemic strain . . . is most likely to emerge in China or the Far East,’ the plan says. So Mexico was another surprise. According to the plan, ‘it seems reasonable to assume that, if the disease starts in Asia, taking about a month to build up to about a thousand cases and a further two to four weeks to spread to the UK, estimates of the mortality rate will be available by the time it reaches the UK.’ This is optimistic. More than a month after the first laboratory-confirmed cases occurred outside Mexico, it is still far from certain what the mortality rate will be across all age groups.
And under the heading ‘What We Know’, the Plan says: ‘A pandemic flu outbreak cannot be contained in the UK because of the large numbers of seed cases (i.e. people bringing in the disease in from abroad) that would be expected. The disease would be expected to spread to all major UK centres of population within one to two weeks.’ Wrong again, but pessimistic this time. After more than two weeks, the UK is still in containment mode, aggressively searching for cases and identifying their contacts, quarantining them, and closing schools.
That so many assumptions in the Pandemic Plan must now be questioned does not mean that it is bad. But just as J.K.Galbraith’s The Great Crash, 1929 is enjoying a boost in sales, I guess that Richard Neustadt and Harvey Fineberg’s The Epidemic That Never Was: Policy-Making And The Swine Flu Affair (1983) will be reread. It describes the events that followed the first recorded death of a healthy individual from swine influenza. David Lewis was an 18-year old army recruit at Fort Dix, New Jersey. He died on the night of 4 February 1976 after a forced march done against medical advice; he had developed a respiratory infection earlier in the day. The virological consensus then was that the 1918 pandemic virus came from pigs (the work of Jeffery Taubenberger at the Armed Forces Institute of Pathology in Washington, DC has shown since that it came from birds).
There was intense lobbying. On 24 March a ‘Blue Ribbon Panel’ of experts met with President Ford in the Oval Office. The decision was taken to vaccinate the whole population. The programme started in October. But its complication rate was higher than expected, and the virus never got out of Fort Dix. The vaccination programme was suspended on 16 December, and never restarted. In February 1977 Neustadt and Fineberg were commissioned by the incoming Carter administration to review the programme and look for lessons to be learned. Their report was called ‘The Swine Flu Affair: Decision-Making on a Slippery Disease’. Slipperiness comes not only from the changing character of the virus, but from ignorance about it. In 1976 experts were ‘overconfident in theories validated through but two or three pandemics . . . it is not that conclusions were inconsistent with evidence, but that the paucity of evidence belied the force with which conclusions were advanced.’
A lot of research has been done since then. But the work done on influenza in pigs and its relevance to the evolution of new human pandemic viruses has been almost completely overshadowed by worries about the potential threat from H5N1 bird flu. Pigs are susceptible to infection with human and bird flu strains. The evidence that they act as ‘mixing vessels’ in which human, bird and swine viruses evolve by swapping their genes is sound and strong, as is the evidence that people can be infected from pigs, and vice versa. Many papers have been published on the subject, including one in 2005 from Yucatan in Mexico showing that Mayans living very close to their pigs, turkeys, ducks and chickens showed evidence of infection with swine viruses. It concluded by saying that the human/animal encounters in Yucatan could lead to the generation of novel reassortant viruses. But the minute of the Department of Health Joint Committee on Vaccination and Immunisation meeting on 21 June 2006 speaks for itself:
The Advisory Committee on Dangerous Pathogens (ACDP) had noted at an earlier meeting that the hypothesis that pigs act as the mixing vessel for influenza viruses from which the new pandemic human strain arises is no longer widely held to be correct as there is little evidence that a flu strain dangerous to human health has been produced in this way.
The ACDP meeting had been convened urgently in October 2005 to consider the public health measures that would be needed if bird flu occurred in the UK.
Most of the evidence about S-OIV comes from CDC. As of yesterday there have been 896 laboratory-confirmed cases in the US, in 43 states, with two deaths. The average age of the cases is 15. So far the virus is behaving like seasonal flu because it is targeting the young, and because there are many outbreaks happening at different stages of progression. Maybe more cases than would be expected are being hospitalised. It is too early to be certain. But so far the elderly have been spared. It is not known whether this is because previous influenza has given them some immunity, because the young are more susceptible, because differences in social networks favour transmission to them, or because there is a bias in case detection thanks to the vigorous investigation of outbreaks in schools. It is still too early to judge the current hypothesis that the deaths in Mexico can be explained as the tip of the iceberg of a big S-OIV outbreak.
Whether Tamiflu, Relenza and the internet (CDC is using Widgets, Buttons, Podcasts, eCards, Twitter/Microblogs, Online Videos, Facebook, MySpace and DailyStrength to get its messages across) will help to bring the epidemic in North America – at six o’clock this morning GMT, 1112 laboratory-confirmed cases had been reported from Mexico and 214 from Canada – to a close remains to be seen. So far the rest of the world has had very few cases, and there has been no sustained transmission of the virus. As Sam Goldwyn said, however, ‘making predictions is difficult, particularly about the future.’