Portal de Periódicos da CAPES

Infectious Disease: Can We Avert a Lethal Flu Pandemic?

2005; Elsevier BV; Volume: 15; Issue: 22 Linguagem: Inglês

10.1016/j.cub.2005.10.063

1879-0445

Robert M. May,

SARS-CoV-2 and COVID-19 Research

If avian flu becomes directly transmissible among humans, could we prevent a pandemic by using prophylactic antivirals? Possibly, if the virus is not too transmissible, and we react fast and efficiently. If avian flu becomes directly transmissible among humans, could we prevent a pandemic by using prophylactic antivirals? Possibly, if the virus is not too transmissible, and we react fast and efficiently. Since 1997, a highly pathogenic avian influenza virus (AIV), of the subtype H5N1, has emerged in poultry in Southeast Asia. Concern is rising, for several reasons. In May 2005, the virus was found in migratory waterfowl in western China [1Liu J. Xiao H. Lei F. Zhu Q. Qin K. Zhang X.W. Zhang X.L. Zhao D. Wang G. Feng Y. et al.Highly pathogenic H5N1 influenza virus infection in migratory birds.Science. 2005; 309: 1206Crossref PubMed Scopus (621) Google Scholar] and most recently AIV is reported among birds in Turkey and Romania. To date, essentially all the few, but increasing, human cases appear to have come from human-to-bird contact, causing around 60 deaths from something like 120 known infections [2WHO. (2005). Cumulative number of confirmed human cases of Avian Influenza. http://www.who.int/csr/disease/avian_influenza/table_2005_10_10/en/16/10/2005.Google Scholar]. Past flu pandemics are believed to have arisen when AIV has reassorted with a human influenza virus, within a dually infected human patient, resulting in a new variant capable of direct transmission from human to human. So, what is our chance of containing such a H5N1 or similar flu epidemic, if and when it appears? This question is addressed in two separate and recent numerical studies [3Longini Jr., I.M. Nizam A. Xu S. Ungchusak K. Hanshaoworakul W. Cummings D.A. Halloran M.E. Containing pandemic influenza at the source.Science. 2005; 309: 1083-1087Crossref PubMed Scopus (921) Google Scholar, 4Ferguson N.M. Cummings D.A. Cauchemez S. Fraser C. Riley S. Meeyai A. Iamsirithaworn S. Burke D.S. Strategies for containing an emerging influenza pandemic in Southeast Asia.Nature. 2005; 437: 209-214Crossref PubMed Scopus (1343) Google Scholar]. Both assume that AIV jumps to humans in Southeast Asia, and use evidence from experience with previous flu epidemics to assess the parameters in complex and detailed models of the spread of infection originating in Thailand, and the possibilities of averting a pandemic. Central to any such study is the infection's basic reproductive number, R0, which quantifies its transmissibility; R0 is defined as the average number of secondary cases generated by a typical primary case in an entirely susceptible population [5Anderson R.M. May R.M. Infectious Diseases of Humans: Dynamics and Control. Oxford University Press, Oxford1991Google Scholar]. Epidemics can arise if R0 exceeds one, and not otherwise. Control strategies aim to reduce R0 below one, by effectively removing a proportion 1–(1/R0) of the susceptible population. Longini et al. [3Longini Jr., I.M. Nizam A. Xu S. Ungchusak K. Hanshaoworakul W. Cummings D.A. Halloran M.E. Containing pandemic influenza at the source.Science. 2005; 309: 1083-1087Crossref PubMed Scopus (921) Google Scholar] estimate R0 as around 1.4 — based on about 33% of the population being infected in past Asian pandemics — but their simulations explore R0-values from 1.1 to 2.4. Ferguson et al. [4Ferguson N.M. Cummings D.A. Cauchemez S. Fraser C. Riley S. Meeyai A. Iamsirithaworn S. Burke D.S. Strategies for containing an emerging influenza pandemic in Southeast Asia.Nature. 2005; 437: 209-214Crossref PubMed Scopus (1343) Google Scholar] estimate R0 at 1.8 — based on re-analysis of earlier data, suggesting the average 'generation interval' between an individual becoming infected and infecting a contact is around 2.6 days, rather than the previously estimated 4 — but they also explore a range 1<R0<2. Longini et al. [3Longini Jr., I.M. Nizam A. Xu S. Ungchusak K. Hanshaoworakul W. Cummings D.A. Halloran M.E. Containing pandemic influenza at the source.Science. 2005; 309: 1083-1087Crossref PubMed Scopus (921) Google Scholar] assume a model population of 500,000 people, with age-structure and patterns of movement within and between 36 geographical regions, characteristic of rural Thailand (as revealed in its 2000 census). Ferguson et al.[4Ferguson N.M. Cummings D.A. Cauchemez S. Fraser C. Riley S. Meeyai A. Iamsirithaworn S. Burke D.S. Strategies for containing an emerging influenza pandemic in Southeast Asia.Nature. 2005; 437: 209-214Crossref PubMed Scopus (1343) Google Scholar], using massive parallel computational capacity, work with spatially detailed simulations of the 85 million people in Thailand and in a 100kilometre-wide zone outside its borders, explicitly including households, schools and workplaces. Figure 1illustrates how an outbreak of flu might spread in Thailand, if there is no intervention, roughly 90 days after the first case. It comes from Ferguson et al.'s [4Ferguson N.M. Cummings D.A. Cauchemez S. Fraser C. Riley S. Meeyai A. Iamsirithaworn S. Burke D.S. Strategies for containing an emerging influenza pandemic in Southeast Asia.Nature. 2005; 437: 209-214Crossref PubMed Scopus (1343) Google Scholar] simulations, seeded with a single rural individual. For the first 30 days, the epidemic tends to be spatially confined. As 'sparks' are increasingly shed into other regions, numbers increase exponentially and infectious individuals spread over larger distances [6May R.M. Gupta S. McLean A.R. Infectious disease dynamics: what characterizes a successful invader?.Philos. Trans. R. Soc. Lond. B Biol. Sci. 2001; 356: 901-910Crossref PubMed Scopus (107) Google Scholar]. Between 60 and 90 days the epidemic changes from being mainly local to being country-wide. Any control strategy needs to be implemented effectively before this time; after this, logistic constraints make success unlikely. If R0 is 1.8, the unchecked epidemic infects roughly two-thirds of the modelled population of 85 million. Methods available for stopping a flu epidemic come under three headings. First, 'targeted antiviral prophylaxis' (TAP). Antiviral drugs, of which the most effective currently is oseltamivir (Trade Name 'Tamiflu', which inhibits the action of neuraminidase, the N of H5N1), work by significantly reducing transmission. Although they can also help an individual to fight off infection, for population-level prophylaxis they need to be administered in advance of apparent infection. Ferguson et al. [4Ferguson N.M. Cummings D.A. Cauchemez S. Fraser C. Riley S. Meeyai A. Iamsirithaworn S. Burke D.S. Strategies for containing an emerging influenza pandemic in Southeast Asia.Nature. 2005; 437: 209-214Crossref PubMed Scopus (1343) Google Scholar] estimate that blanketing an entire country or region with Tamiflu should be able to eliminate a pandemic virus with an R0 of even 3.6 or greater. But such action is logistically difficult, if not impossible. Targeted strategies are therefore needed. Both Ferguson et al. [4Ferguson N.M. Cummings D.A. Cauchemez S. Fraser C. Riley S. Meeyai A. Iamsirithaworn S. Burke D.S. Strategies for containing an emerging influenza pandemic in Southeast Asia.Nature. 2005; 437: 209-214Crossref PubMed Scopus (1343) Google Scholar] and Longini et al. [3Longini Jr., I.M. Nizam A. Xu S. Ungchusak K. Hanshaoworakul W. Cummings D.A. Halloran M.E. Containing pandemic influenza at the source.Science. 2005; 309: 1083-1087Crossref PubMed Scopus (921) Google Scholar] look at forms of social targeting as the most straightforward approach. This involves prophylaxing individuals in the same household, school or workplace as newly diagnosed symptomatic cases. And doing so very promptly. Indeed, given that 'social targeting' may be too slow to be effective, both sets of authors assume somewhat wider targeting to local neighbourhoods. Longini et al. [3Longini Jr., I.M. Nizam A. Xu S. Ungchusak K. Hanshaoworakul W. Cummings D.A. Halloran M.E. Containing pandemic influenza at the source.Science. 2005; 309: 1083-1087Crossref PubMed Scopus (921) Google Scholar] call this Geographical targeted antiviral prophylaxis (GTAP). Second, vaccination would be excellent if we had a vaccine. Although a human influenza H5N1 vaccine is currently being tested, and may be available in time, the problem is — as is the case for normal seasonal flu vaccines — we cannot be certain how effective the vaccine under development may be against the strain that eventually appears. Third, we have quarantine, or other ways of effectively reducing contact rates within the population (Ferguson et al. [4Ferguson N.M. Cummings D.A. Cauchemez S. Fraser C. Riley S. Meeyai A. Iamsirithaworn S. Burke D.S. Strategies for containing an emerging influenza pandemic in Southeast Asia.Nature. 2005; 437: 209-214Crossref PubMed Scopus (1343) Google Scholar] call this "social distance measures"). Exploring various combinations of these actions, Longini et al. [3Longini Jr., I.M. Nizam A. Xu S. Ungchusak K. Hanshaoworakul W. Cummings D.A. Halloran M.E. Containing pandemic influenza at the source.Science. 2005; 309: 1083-1087Crossref PubMed Scopus (921) Google Scholar] concluded that a prepared response with GTAP would have a high probability of containing the epidemic, provided R0 was below 1.6. Such containment is achieved, in their simulations, with an antiviral stock pile of around 100,000 to 1 million courses of treatment. If effective pre-vaccination occurred — assuming only 30% efficiency, and only halving transmissibility — then TAP could be effective against strains with R0 as high as 2.1. And if one could use combinations of TAP, pre-vaccination and quarantine, then success could be achieved even against strains with R0 as high as 2.4. Ferguson et al.'s [4Ferguson N.M. Cummings D.A. Cauchemez S. Fraser C. Riley S. Meeyai A. Iamsirithaworn S. Burke D.S. Strategies for containing an emerging influenza pandemic in Southeast Asia.Nature. 2005; 437: 209-214Crossref PubMed Scopus (1343) Google Scholar] studies of control were focussed mainly on targeted antiviral prophylaxis. This was partly because they saw logistic difficulties in "social distance measures", and partly because they believed some such actions could have unintended adverse consequences as a result of people dispersing. Broadly, their more computationally ambitious model — remember, they included the actual 85 million people in and around Thailand, versus Longini et al.'s [3Longini Jr., I.M. Nizam A. Xu S. Ungchusak K. Hanshaoworakul W. Cummings D.A. Halloran M.E. Containing pandemic influenza at the source.Science. 2005; 309: 1083-1087Crossref PubMed Scopus (921) Google Scholar] 500,000, rather tidily distributed — concluded that a combination of GTAP and social distance measures could contain an outbreak in Thailand and avoid a pandemic, provided the new virus' R0 is below 1.8. But their corresponding calculations required a stockpile of 3 million courses of antiviral drugs. And for larger values of R0, even this would seem insufficient. In the light of these conclusions, a glance backward to 1918 is in order. The number killed in that pandemic is estimated at 20 to 50 million. But the global population then was less than 2 billion, with only one quarter urban. And the relatively smaller number of people crossing oceans did so in ships. Today's threatened pandemic looms over a more crowded world of 6.5 billion, half urban, constantly and rapidly moving around. We know much more, but our circumstances are inherently more difficult. In the UK, the government has announced plans to stockpile enough courses of Tamiflu to cover one quarter of the UK population by the end of 2006. In a world of good surveillance, efficient delivery of TAP, and no panic, this would be adequate provided H5N1 waits until the UK is ready. In the real world of panicking people, the UK plans could prove more problematic. Meanwhile, the USA is committing $25 million to boosting surveillance in Asia. An official at the National Immunization Program of the US Centers for Disease Control has juxtaposed this sum against the $800 million spent by the USA on anthrax vaccines, "against a pathogen that has killed only a handful of Americans and whose bioterrorist potential is unproven" [7Butler D. Drugs could head off a flu pandemic – but only if we respond fast enough.Nature. 2005; 436: 614-615Crossref PubMed Scopus (7) Google Scholar]. We live in interesting times.