|Home||Books Journals Flying Publisher Guides|
PDF, 225 pages, 2.8 MB
Publish this book under your own name
- Green links: Free full-text articles -
Influenza pandemics resemble major natural disasters: we know there will be another one,
but we ignore both time and magnitude. In most other aspects, they are different. Earthquakes in Tokyo or San
Francisco last from seconds to a couple of minutes - pandemics spread around the world in successive waves
over months or a couple of years. And quite different are the consequences: an
influenza pandemic may be a thousand times more deadly than even the deadliest tsunami. As unpredictable as influenza pandemics are, as unpredictable is the virus itself. We know
nothing about the pathogenic potential of the next pandemic strain. The next pandemic may be relatively
benign, as it was in 1968 and 1957, or truly malignant, as was the 1918 episode. We don't know if the next
pandemic will be caused by the current bête noire, H5N1, or by another influenza strain. We ignore how
the next pandemic will evolve over time, how rapidly it will spread around the world, and in how many waves.
We don't know which age groups are at the highest risk of severe outcomes. We have no idea whether the next
pandemic will kill 2, 20, or 200 million people. Not surprisingly, healthcare professionals are becoming sensitised to the risk of a new
pandemic. The ongoing outbreak of H5N1 influenza among birds with occasional transmission to human beings is
of major concern because of intriguing parallels between the H5N1 virus and the 1918 influenza strain. Should
H5N1 acquire the capability of easy human-to-human transmissibility, even the most conservative scenario
anticipates up to several 100 million outpatient visits, more than 25 million hospital admissions and several
million deaths globally (
Influenza pandemics resemble major natural disasters: we know there will be another one, but we ignore both time and magnitude. In most other aspects, they are different. Earthquakes in Tokyo or San Francisco last from seconds to a couple of minutes - pandemics spread around the world in successive waves over months or a couple of years. And quite different are the consequences: an influenza pandemic may be a thousand times more deadly than even the deadliest tsunami.
As unpredictable as influenza pandemics are, as unpredictable is the virus itself. We know nothing about the pathogenic potential of the next pandemic strain. The next pandemic may be relatively benign, as it was in 1968 and 1957, or truly malignant, as was the 1918 episode. We don't know if the next pandemic will be caused by the current bête noire, H5N1, or by another influenza strain. We ignore how the next pandemic will evolve over time, how rapidly it will spread around the world, and in how many waves. We don't know which age groups are at the highest risk of severe outcomes. We have no idea whether the next pandemic will kill 2, 20, or 200 million people.
Not surprisingly, healthcare professionals are becoming sensitised to the risk of a new pandemic. The ongoing outbreak of H5N1 influenza among birds with occasional transmission to human beings is of major concern because of intriguing parallels between the H5N1 virus and the 1918 influenza strain. Should H5N1 acquire the capability of easy human-to-human transmissibility, even the most conservative scenario anticipates up to several 100 million outpatient visits, more than 25 million hospital admissions and several million deaths globally (WHO Checklist 2005).
It is wise to imagine and plan for the worst when facing an unknown threat. As the threat is global, strategies must be global - a tricky task when our planet is divided into more than two hundred nations. Dealing with nations and their leaders is like dealing with children in a kindergarten. In this difficult context, the WHO is performing an astonishing job.
In the following paragraphs, we shall take a look at the various facets of the war on influenza: the global and individual impact of the disease, the virus itself, and the individual and global management of what may one day turn out to be one of the most challenging healthcare crises in medical history. The most important thing to remember when talking about pandemic influenza is that its severe form has little in common with seasonal influenza. Pandemic influenza is not business-as-usual influenza. Bear this in mind. You wouldn't call a tiger a cat.
Epidemics and Pandemics
Influenza is a serious respiratory illness which can be debilitating and cause complications that lead to hospitalisation and death, especially in the elderly. Every year, the global burden of influenzaepidemics is believed to be 3-5 million cases of severe illness and 300,000-500,000 deaths. The risk of serious illness and death is highest among persons aged > 65 years, children aged < 2 years, and persons who have medical conditions that place them at increased risk of developing complications from influenza (CDC 2005).
New epidemic influenza A strains arise every 1 to 2 years by the introduction of selected point mutations within two surface glycoproteins: haemagglutinin (HA) and neuraminidase (NA). The new variants are able to elude human host defences and there is therefore no lasting immunity against the virus, neither after natural infection nor after vaccination, as is the case with smallpox, yellow fever, polio, and measles. These permanent and usually small changes in the antigenicity of influenza A viruses are termed "antigenic drift" and are the basis for the regular occurrence of influenza epidemics (Figure 1). In addition, there is now evidence that multiple lineages of the same virus subtype can co-circulate, persist, and reassort in epidemiologically significant ways(Holmes 2005).
Figure 1. Antigenic drift. Courtesy: National Institute of Allergy and Infectious Disease
In contrast to epidemics,pandemics are rare events that occur every 10 to 50 years. They have been documented since the 16th century (WHO 2005b), and in the last 400 years, at least 31 pandemics have been recorded (Lazzari 2004). During the twentieth century, three influenza pandemics occurred (table 1). Their mortality impact ranged from devastating to moderate or mild (Simonson 2004). The 1918 pandemic was caused by a H1N1 virus of apparently avian origin (Reid 1999), whereas the subsequent pandemic strains - H2N2 in 1957 and H3N2 in 1968 - were reassortant viruses containing genes from avian viruses: three in 1957 (haemagglutinin, neuraminidase, and the RNA polymerase PB1) and two (haemagglutinin and PB1) in 1968 (Kawaoka 1989). These major changes in the antigenicity of an influenza virus are called "antigenic shift" (Figure 2).
* H = haemagglutinin; N = neuraminidase
Figure 2. Antigenic shift. Courtesy: National Institute of Allergy and Infectious Disease
Influenza pandemics circulate around the globe in successive waves, and there is no way to prevent the spread of a new pandemic influenza virus. The new viral strain will eventually reach everywhere, and will infect practically every human being within a period of a few years. Seasonal excess mortality rates due to pneumonia and influenza may remain elevated for many years, as was shown in the A(H3N2)-dominated seasons in the decade after 1968, in persons aged 45-64 years in the United States (Simonsen 2004).
One hallmark of pandemic influenza is a mortality shift towards younger age groups. Half of influenza-related deaths during the 1968 pandemic, and large proportions of influenza-related deaths during the 1957 and the 1918 pandemics, occurred among persons < 65 years old (Simonson 1998).
The first influenza pandemic of the 20th century spread more or less simultaneously in 3 distinct waves during a 12-month period in 1918-1919, across Europe, Asia, and North America (Barry 2004, Taubenberger 2006). It was the worst pandemic in history, killing more people than World War I, and it is generally assumed that at least 50 million people died (Johnson 2002). The first wave, which started during the spring of 1918, was highly contagious but not particularly deadly. Only the second wave, beginning in September, spread the deadly form of the pandemic.
Figure 3. Emergency hospital during influenza epidemic, Camp Funston, Kansas. Images from the 1918 Influenza Epidemic. Image copyright by National Museum of Health & Medicine, Washington, D.C.http://InfluenzaReport.com/link.php?id=19
The virus of 1918 was extremely virulent and caused many deaths through secondary bacterial pneumonia. The primary viral pneumonia could kill previously healthy young individuals within 2 days. The clinical course of severe cases was so unfamiliar that investigators doubted it was influenza (WHO 2005b). Symptoms in 1918 were so unusual that, initially, it was misdiagnosed as dengue fever, cholera, or typhoid (Barry 2004).
In less severe cases, most patients experienced typical influenza with a 3- to 5-day fever followed by complete recovery (Kilbourne 2006). In contrast to subsequent pandemics, most deaths during the 1918 pandemic were among young and healthy persons aged 15 to 35 years old, and 99 % of deaths occurred in people younger than 65 years.
The recovery of the genomic RNA of the 1918 virus from archived formalin-fixed lung autopsy material and from frozen, unfixed lung tissue from an influenza victim who was buried in permafrost in November 1918 (Taubenberger 1997) has enabled the complete coding of the sequences of all eight viral RNA segments of the 1918 H1N1 virus (Taubenberger 2005). According to this investigation, the 1918 virus was not a reassortant virus (like those of the 1957 and 1968 pandemics), but more likely an entirely avian-like virus that adapted to humans.
The 1957 pandemic was caused by H2N2, a clinically milder virus than the one responsible for the 1918 pandemic. Outbreaks were frequently explosive, but the death toll was much lower. Mortality showed a more characteristic pattern, similar to that seen in seasonal epidemics, with most excess deaths confined to infants and the elderly (WHO 2005b). Patients with chronic underlying disease and pregnant women were particularly at risk of developing pulmonary complications (Louria 1957). The global excess mortality of the 1957 pandemic has been estimated at 1-2 million deaths.
The 1968 pandemic, was also a mild pandemic. The mortality impact was not even particularly severe compared to the severe epidemic in 1967-1968 (the last H2N2 epidemic), as well as two severe H3N2 epidemics in 1975-1976, and in 1980-1981 (Simonsen 2004). The death toll has been estimated to have been around 1 million, and in the United States, nearly 50 percent of all influenza-related deaths occurred in the younger population under 65 years of age. Sero-archaeological studies showed that most individuals aged 77 years or older, had H3 antibodies before they were exposed to the new pandemic virus (Dowdle 1999) and that pre-existing anti-H3 antibodies might have protected the elderly (> 77 years old) during the 1968 H3N2 pandemic.
Since 1968, there has been only one episode - in 1976- when the start of a new pandemic was falsely anticipated (Dowdle 1997, Gaydos 2006, Kilbourne 2006).
Major pandemics have occurred throughout history at an average of every 30 years and there is a general consensus that there will be another influenza pandemic. It is impossible to predict which influenza strain will be the next pandemic virus. One possible candidate is the avian H5N1 strain which has become endemic in wild waterfowl and in domestic poultry in many parts of Southeast Asia, and is recently spreading across Asia into Europe and Africa. Recent research has shown that just ten amino acid changes in the polymerase proteins differentiate the 1918 influenza virus sequences from that of avian viruses, and that a number of the same changes have been found in recently circulating, highly pathogenic H5N1 viruses(Taubenberger 2005).
At present, H5N1 avian influenza remains largely a disease of birds. The species barrier is significant: despite the infection of tens of millions of poultry over large geographical areas for more than two years, fewer than 200 human cases have been confirmed by a laboratory (WHO 200601). Human cases, first documented in 1997 (Yuen 1998), coincided with outbreaks of highly pathogenic H5N1 avian influenza in poultry. Very limited human-to-human transmission of the H5N1 strain was documented in healthcare workers and family members with contact (Katz 1999, Buxton Bridges 2000). Although H5 antibodies were detected in these groups, indicating infection with the virus, no cases of severe disease occurred.
There are little data to show to what extent asymptomatic infection or mild clinical disease occur following infection with highly pathogenic avian H5N1 strains. If asymptomatic infections were frequent, the 55 % fatality rate of severe human H5N1 disease reported as of 21 March 2006 (WHO 20060321) would of course be less alarming. However, these episodes may be the exception, at least in some settings. In a study conducted in a Cambodian village with H5N1 outbreaks in poultry and 4 fatal human cases, testing of blood samples from 351 villagers found no additional infections, although many villagers had had significant exposure to infected poultry (ProMED 20060322.0893 and Buchy, personal communication).
Until now, the disease has predominantly affected children and young adults. Of 116 patients for whom demographical data had been published on the WHO Website from December 2003 until 9 February 2006, 50 % were 16 years old or younger, 75 % were younger than 30 years, and 90 % were younger than 40 years old (Promed 20060211.0463). The reason for this age distribution (exposure risk, disease reporting bias, intrinsic host issues, etc.) is unclear. Likewise, it is not known whether, and to what extent, genetic composition plays a role in the susceptibility and resistance to infection with H5N1 influenza virus (Promed 20060216.0512).
The next pandemic is expected to cause clinical disease in 2 billion people. Best-case scenarios, modelled on the mild pandemic of 1968, anticipate between 2 million and 7.4 million cases (WHO 2005b). However, if we translate the death toll associated with the 1918 influenza virus to the current population, there could be 180 million to 360 million deaths globally (Osterholm 2005).
The fate of an individual during an influenza outbreak, be it epidemic or pandemic, is variable. It is estimated that about half of those infected have no clinical symptoms or signs. Among the others, clinical presentation varies from afebrile respiratory symptoms mimicking the common cold, to febrile illnesses ranging in severity from mild to debilitating (Hoffmann 2006a), and may cause disorders affecting the lung, heart, brain, liver, kidneys, and muscles (Nicholson 2003).
The clinical course is influenced by the patient's age, the degree of pre-existing immunity, properties of the virus, smoking, co-morbidities, immunosuppression, and pregnancy (Nicholson 2003). Death mostly occurs as a consequence of primary viral pneumonia or of secondary respiratory bacterial infections, especially in patients with underlying pulmonary or cardiopulmonary diseases. The very young and the elderly usually have the highest risk of developing serious complications; however, during pandemics, there is a mortality shift towards younger age groups (Simonson 1998).
In humans, replication of influenza subtypes seems to be limited to the respiratory epithelial cells. Once the virus enters a cell, it causes complex cytopathic effects, predominantly in the columnar epithelial cells, by shutting down the synthesis of host proteins. The loss of critical host cell proteins leads to cell death by necrosis (Yuen 2005). There are numerous individual factors associated with protection against or increasing the risk of a fatal outcome caused by a given influenza strain (Behrens and Stoll 2006), and genetic factors that affect host susceptibility are likely to play a role. Specific immunity against certain viral epitopes or some degree of cross-immunity may explain why people > 65 years were less affected by the 1918 pandemic. It is unknown whether similar mechanisms play a role in the curious age distribution of cases in the current outbreak of avian H5N1 influenza (ProMED 20060211.0463).
The unusual severity of H5N1 infection in humans was initially ascribed to multiple basic amino acids adjacent to the cleavage site, a feature characteristic of highly pathogenic avian influenza A viruses(Subbarao 1998). The presence of these basic amino acids renders the protein susceptible to proteases from many different types of tissues and allows extrapulmonary dissemination due to broadened tissue tropism (Yuen 2005). Another explanation may be that interferons are pivotal in preventing viral spread outside the respiratory tract and that H5N1 interferes with this innate defence against viral infection. It has been shown that the non-structural (NS) gene of highly pathogenic H5N1 viruses confers resistance to the antiviral effects of interferons and tumour necrosis factor alpha (Seo 2002). H5N1 viruses seem to induce higher gene transcription of pro-inflammatory cytokines than do H3N2 or H1N1 viruses, and are potent inducers of pro-inflammatory cytokines in macrophages, the most notable being TNF alpha (Cheung 2002). These mechanisms might ultimately lead to a cytokine storm and death (Peiris 2004).
In interpandemic influenza epidemics, recovery from interpandemic influenza is usually uneventful. In severe cases of human H5N1 influenza, however, mortality has so far been considerable (WHO 20060213). Dyspnoea, ARDS and multi-organ failure has been a dominant clinical feature in fatal cases (Hoffmann 2006a), with a median time from onset of symptoms to death of 9 days (n=76) (http://www.influenzareport.com/links.php?id=16).
Infectious diseases are the result of a conflict of interest between macroorganisms and microorganisms. We are not alone on earth.
Requirements for Success
To become a pandemic strain, an influenza virus must comply with a series of requirements. It has to
Ideally, it has to be more pathogenic than other competing influenza strains. In the current situation, the potential pandemic virus would have to compete with the already circulating H3N2 and H1N1 strains.
The prerequisite for success is good adaptation: adaptation to human cells; the capability to take over the production machinery of the host cell to produce new offspring; as well as making the individual cough and sneeze to spread the offspring viruses. The clue to success is virulence (Noah 2005, Obenauer 2006, Salomon 2006) - and novelty: if the virus is a true newcomer, most living human beings will have little or no protection at all. The new virus will have unlimited access to virtually every human being and will find a feeding ground of > 6.5 billion human beings. This is one of the biggest biomasses in the world.
The passing of powers from one reigning influenza subtype to a new one is called "antigenic shift" because the antigenic characteristics of the new virus need to shift dramatically to elude the immune system of virtually the entire mankind. Antigenic shift is a major change in the influenza A viruses resulting in new haemagglutinin and/or new neuraminidase proteins. This change may occur by: 1) reassortment of the segmented genome of two parent viruses, or 2) gradual mutation of an animal virus. For reassortment to take place, both the new pandemic candidate virus, normally of avian origin, and an already circulating human virus, i.e., H3N2 or H1N1, need to infect the same human host cell. Inside the cell, genes from both viruses are reassembled in an entirely new virus (they don't actually have sex, but for didactic purposes, this image might work quite nicely). That's what happened in 1957 and 1968 (Figure 2).
Reassortment may not be the best route for a candidate pandemic virus. Recent evidence with recombinant viruses containing genes from the 1918 pandemic virus shows that viruses expressing one or more 1918 virus genes were less virulent than the constellation of all eight genes together (Tumpey 2005). The 1918 virus was particular indeed: it appears that it was not the result of a reassortment of two existing viruses, but an entirely avian-like virus that gradually adapted to humans in stepwise mutations (Taubenberger 2005). It is obviously tempting to speculate that the emergence of a completely new human-adapted avian influenza virus in 1918 (n=1) could be deadlier than the introduction of reassortant viruses in 1957 and 1968 (n=2), but such speculation is not scientific. Interestingly - and worryingly -, some amino acid changes in the 1918 virus that distinguish it from standard avian sequences are also seen in the highly pathogenic avian influenza virus strains of H5N1, suggesting that these changes may facilitate virus replication in human cells and increase pathogenicity (Taubenberger 2005).
Influenza A and B viruses are enveloped viruses with a segmented genome made of eight single-stranded negative RNA segments of 890 to 2,341 nucleotides each (Gürtler 2006). They are spherical or filamentous in structure, ranging from 80 to 120 nm in diameter (Figure 4 and 5). When sliced transversely, influenza virions resemble a symmetrical pepperoni pizza, with a circular slice of pepperoni in the middle and seven other slices evenly distributed around it (Noda 2006). On the basis of the antigenicity of the surface glycoproteins, haemagglutinin (HA) and neuraminidase (NA), inﬂuenza A viruses are further divided into sixteen H (H1-H16[Fouchier 2005]) and nine N (N1-N9) subtypes. HA is the major antigen for neutralising antibodies, and is involved in the binding of the virus to host cell receptors. NA is concerned with the release of progeny virions from the cell surface. Currently, only viruses of the H1N1 and H3N2 subtypes are circulating among humans.
Figure 4. Colourised transmission electron micrograph of Avian influenza A H5N1 viruses (seen in gold) grown in MDCK cells (seen in green). Courtesy of CDC/ Cynthia Goldsmith, Jacqueline Katz, and Sharif R. Zaki, Public Health Image Library,http://phil.cdc.gov/Phil/home.asp
Figure 5. This negative-stained transmission electron micrograph (TEM) depicts the ultrastructural details of a number virions. Courtesy of CDC/ Dr. F. A. Murphy, Public Health Image Library,http://phil.cdc.gov/Phil/home.asp
Natural Reservoir + Survival
Influenza A viruses occur in a large variety of species, mainly birds, notably aquatic ones, in which infection is largely intestinal, waterborne, and asymptomatic. The domestic duck in Southeast Asia is the principal host of influenza A viruses and also has a central role in the generation and maintenance of the H5N1 virus (Li 2004). In Thailand, there was a strong association between the H5N1 virus and the abundance of free-grazing ducks and, to a lesser extent, native chickens and cocks, as well as wetlands, and humans. Wetlands that are used for double-crop rice production, where free-grazing ducks feed year round in rice paddies, appear to be a critical factor in HPAI persistence and spread (Gilbert 2006).
Highly pathogenic avian viruses can survive in the environment for long periods, especially in low temperatures (i.e., in manure-contaminated water). In water, the virus can survive for up to four days at 22°C, and more than 30 days at 0°C. In frozen material, the virus probably survives indefinitely. Recent studies indicate that the H5N1 viruses isolated in 2004 have become more stable, surviving at 37°C for 6 days - isolates from the 1997 outbreak survived just 2 days (WHO 20041029). The virus is killed by heat (56°C for 3 hours or 60°C for 30 minutes) and common disinfectants, such as formalin and iodine compounds.
Influenza is primarily transmitted from person to person via droplets (> 5 µm in diameter) from the nose and throat of an infected person who is coughing and sneezing (Figure 6). Particles do not remain suspended in the air, and close contact (up to 3-6 feet) is required for transmission. Transmission may also occur through direct skin-to-skin contact or indirect contact with respiratory secretions (touching contaminated surfaces then touching the eyes, nose or mouth). Individuals may spread influenza virus from up to two days before to approximately 5 days after onset of symptoms. Children can spread the virus for 10 days or longer.
Figure 6. An unimpeded sneeze sends two to five thousand bacteria-filled droplets into the air. Image copyright by Prof. Andrew Davidhazy, Rochester Institute of Technology. Used with permission. (http://www.rit.edu/~andpph)
As influenza viruses are normally highly species specific, they only rarely spill over to cause infection in other species. This is due to differences in the use of cellular receptors. Avian influenza viruses bind to cell-surface glycoproteins containing sialyl-galactosyl residues linked by a 2-3-linkage, whereas human viruses bind to receptors that contain terminal 2-6-linked sialyl-galactosyl moieties. For an avian virus to be easily transmitted between humans, it is fundamental that it acquires the ability to bind cells that display the 2-6 receptors so that it can enter the cell and replicate in them. While single amino acid substitutions can significantly alter receptor specificity of avian H5N1 viruses (Gambaryan 2006), it is presently unknown which specific mutations are needed to make the H5N1 virus easily and sustainably transmissible among humans, but potential routes whereby H5N1 might mutate and acquire human specificity do exist (Stevens 2006).
Since 1959, human infections with avian influenza viruses have only rarely occurred. Of the hundreds of strains of avian influenza A viruses, only four are known to have caused human infection: H5N1, H7N3, H7N7, and H9N2 (WHO 200601). Apart from H5N1, human infection generally resulted in mild symptoms and rarely in severe illness (Du Ry van Beest Holle 2003, Koopmans 2004). For the H5N1 virus, close contact with dead or sick birds (i.e., slaughtering, plucking, butchering and preparation) or exposure to chicken faeces on playgrounds seem to be the principal source of human infection (WHO 200601).
H5N1: Making Progress
At the moment, H5N1 infection in humans is relatively rare, although there must have been widespread exposure to the virus through infected poultry. This in an indicator that the species barrier to the acquisition of this avian virus is still quite high for H5N1 - despite having been in circulation for nearly 10 years. However, over the past years, H5N1 strains seem to have become more pathogenic and to have expanded their range of action:
Try not to get the bugs, and if you get them, try to treat them. In influenza management, this one-line medical wisdom theoretically translates as: 1) three prophylaxis defence lines (exposure prophylaxis, vaccination, prophylactic use of antiviral drugs); and 2) one treatment defence line (antiviral drugs). Due to the very nature of influenza infection - infected individuals may be infectious for as long as 24-48 hours before the onset of symptoms - exposure prophylaxis is virtually impossible during an ongoing epidemic or pandemic, especially in our highly mobile and densely populated world.
Basic personal hygiene measures, invented more than a century ago, are still the cornerstones of prophylaxis. Physicians should encourage regular hand-washing among family members of patients. In general, people should be discouraged to touch their eyes nose or mouth. Minimise the impact of sneezes and coughs by all possible means (WHO 2006a).
Vaccination against influenza viruses is the second cornerstone in preventing influenza. Vaccination in the northern hemisphere is recommended to start in October. Recommendations regarding the composition of the vaccine are issued yearly on the basis of detailed investigations of circulating strains. Vaccination against the prevalent wild-type influenza virus is recommended for all individuals in high-risk groups, including those aged 65 years or older (CDC 2005), and those with chronic illness, particularly diabetes, chronic respiratory and cardiac disease, and persons immunocompromised from disease or concomitant therapy. In addition, it is generally recommended that all healthcare personnel be vaccinated annually against influenza (CDC 2006b). The rate of influenza vaccination depends on a number of variables, including explicit physician recommendation and media coverage (Ma 2006).
In healthy primed adults, the efficacy after one dose may be as high as 80-100 %, while in unprimed adults (those receiving their first influenza immunisation), efficacy is in this range after two doses. With some underlying conditions (i.e., HIV infection, malignancies, renal transplantation), efficacy is lower (Korsman 2006); however, protection ultimately depends on who is vaccinated and on the match between the vaccine and the circulating virus (Wong 2005).
The evidence of efficacy and effectiveness of influenza vaccines in individuals aged 65 years or older has recently been reviewed. Well matched vaccines prevented hospital admission, pneumonia, respiratory diseases, cardiac disease, and death. The effectiveness is better in people living in homes for the elderly than in elderly individuals living in the community (Jefferson 2005). Inactivated vaccine reduces exacerbations in patients with chronic obstructive pulmonary disease (Poole 2006). Influenza vaccines are efficacious in children older than two years but little evidence is available for children under two (Smith 2006). Nasal spray of live vaccines seemed to be better at preventing influenza illness than inactivated vaccines.
In selected populations, antiviral drugs may be a useful option in those not covered or inadequately protected by vaccination. It should be emphasised, though, that the prophylactic use of available antiviral drugs is by no means a substitute for the yearly vaccination recommended by national health services.
Candidates for short-term prophylactic use of antiviral drugs are high-risk patients who are vaccinated only after an epidemic has already begun, as well as unvaccinated high-risk contacts of an individual with influenza. In some cases, prophylaxis could be indicated when a current epidemic is caused by a strain which is not represented in the vaccine. For more details, seeHoffmann 2006b.
Of the two available drug classes, the adamantanes (amantadine, rimantadine) recently came under pressure when the global prevalence of adamantane-resistant influenza viruses was found to have significantly increased from 0.4 % in 1994-1995 to 12.3 % in 2003-2004 (Bright 2005). It is believed that the elevated incidence of resistance in China is due to increased use of over-the-counter amantadine after the emergence of severe acute respiratory syndrome (SARS) (Hayden 2006). In the United States, 109/120 (91 %) influenza A (H3N2) viruses isolated in the 2005-06 season until January 12, 2006, contained an amino acid change at position 31 of the M2 protein, which confers resistance to amantadine and rimantadine (CDC 2006, Bright 2006). On the basis of these results, the CDC issued an interim recommendation that neither amantadine nor rimantadine be used for the treatment or prophylaxis of influenza A in the United States for the remainder of the 2005-06 influenza season. During this period, oseltamivir or zanamivir should be selected if an antiviral medication is used for the treatment and prophylaxis of influenza.
In uncomplicated cases, bed rest with adequate hydration is the treatment of choice for most adolescents and young adult patients (Hoffmann 2006b). Antibiotic treatment should be reserved for the treatment of secondary bacterial pneumonia.
The older drugs,rimantadine and amantadine, are only effective against influenza A virus (CDC 2005). However, there is little data available on elderly people; the drugs have more side effects; and in the 2005/2006 season, the CDC discouraged the use of these drugs (see previous section). If rimantadine and amantadine are used, it is important to reduce the emergence of antiviral drug-resistant viruses. Amantadine or rimantadine treatment should therefore be discontinued as soon as possible, typically after 3-5 days of treatment, or within 24-48 hours after the disappearance of signs and symptoms (CDC 2005).
The newer neuraminidase inhibitors are licensed for treatment of patients aged 1 year and older (oseltamivir) or 7 years and older (zanamivir). They are indicated in patients with uncomplicated acute illness who have been symptomatic for no more than 2 days. The recommended duration of treatment for both drugs is 5 days.
The problem with a new pandemic influenza strain is that there is no hiding place on earth. Virtually any single human being will eventually become infected with the new virus, be it the beggar from Paris or the President of a wealthy western country. If you don't get the virus during the first wave of the pandemic, you'll probably get it during the second. And if you don't get it during the second wave, you will get it during one of the future epidemics. If a novel pandemic influenza strain takes over as the driver of influenza disease in humans, everyone needs to mount a protective antibody response against the virus - simply because the virus is bound to stay with us for many years. Antibodies will provide some protection against the new influenza strain, but to develop antibodies you have to either be infected or vaccinated.
For the vast majority of the 6.5 billion living human beings, there will be no vaccine available any time soon after the arrival of a new pandemic influenza virus. Once a new virus has been shown to be effectively transmitted among humans, it will take approximately 6 months to start the production of the corresponding vaccine. Thereafter, vaccine supplies will be exquisitely inadequate, and years will be needed to produce enough vaccine for 6.5 billion people. In addition, production capacities are concentrated in Australia, Canada, France, Germany, Italy, Japan, the Netherlands, the United Kingdom, and the United States, and vaccine distribution can be expected to be controlled by the producing nations (Fedson 2005). We can all imagine who will be served first.
It is therefore reasonable to assume that the vast majority of people living today will have no access to either vaccine or antiviral drugs for many, many months. With no vaccine available or vaccine arriving too late, individuals might wish to work out strategies to deal with a pandemic situation. To confront or to avoid - that will be the question many people will ask themselves.
Simply confronting a new pandemic virus and hoping for a happy outcome, leaves the problem of timing. Indeed, there is conflicting evidence about the most adequate moment for getting infected:
A commonly observed phenomenon in infectious diseases is that pathogens become less virulent as they evolve in a human population. This would favour the second option, i.e., of avoiding a new influenza virus for as long as possible. An additional advantage of this choice is that several months after the start of the pandemic, the initial chaos the health systems will inevitably face during a major outbreak, will have at least partially resolved.
The most extreme option of avoiding influenza would be to flee to remote areas of the globe - a mountain village in Corsica, the Libyan Desert, or American Samoa (Barry 2004). That might work but it might not. If the direct and unprotected confrontation with the new virus becomes inevitable, some protection is still possible: face masks (but: will masks be available everywhere? and for how long?) and social distancing (don't go to meetings, stay at home as much as possible) - but what if you are working as a cashier in a crowded Paris supermarket; as a metro driver in London's tube; as a clerk in Berlin's central post office?. Where will you get money from if you don't go to work for several months? Can you retire from the world? Can you retire from life?
We don't know whether the next pandemic influenza strain will be susceptible to the currently available antiviral drugs. If it is caused by a H5N1 virus, the neuraminidase inhibitorsoseltamivir and zanamivir may be critical in the planning for a pandemic (Moscona 2005). Again, most people on earth will not have access to these drugs. They are in short supply and production capacities cannot easily be built up. Even in countries which have stockpiled oseltamivir, distribution of a drug that is in short supply will pose considerable ethical problems for treatment. In some countries with pronounced wealth disparities (i.e., some African and Latin American countries; the U.S.), social unrest can be anticipated.
Experience in treating H5N1 disease in humans is limited and the clinical reports published to date include only a few patients (Yuen 1998, Chan 2002, Hien 2004, Chotpitayasunondh 2005, WHO 2005, de Jong 2005). In particular, the optimal dose and duration of oseltamivir treatment is uncertain in H5N1 disease, and the following preliminary recommendations have been proposed (WHO 2005):
Although oseltamivir is generally well tolerated, gastrointestinal side effects in particular may increase with higher doses, particularly above 300 mg/day (WHO 2006d). For more details, check Hoffmann 2006b.
The management of an influenza outbreak is well-defined for epidemics, and less well-defined for pandemics.
The cornerstone of medical intervention ininterpandemic years is vaccination (see summary at CDC 2005). As influenza viruses mutate constantly, vaccine formulations need to be re-examined annually. Vaccine production is a well-established procedure: throughout the year, influenza surveillance centres in 82 countries around the world watch circulating strains of influenza and observe the trends. The WHO then determines the strains that are most likely to resemble the strains in circulation during the next year's winter seasons, and vaccine producers start vaccine production. The decision on the composition of the next "cocktail" is made each year in February for the following northern hemisphere winter (WHO 2006b) and in September for the following southern hemisphere winter (for more details, see Korsman 2006 and the figure at http://influenzareport.com/link.php?id=15). Predicting the evolutionary changes of the viral haemaglutinin is not easy and not always successful. In years when the anticipated strain does not match the real world strain, protection from influenza vaccine may be as low as 30 %.
- See alsoReyes-Terán 2006 and WHO 2006c -
Seriousinfluenza pandemics are rare and unpredictable events. Managing unedited situations requires some appreciation of the magnitude of the problems that lie ahead. The impact on human health may be highly variable and is expressed in the number of
It is generally assumed that during the first year of the next pandemic 2 billion people will become infected with the new virus and that half of them will have symptoms. Less accurate are the estimates of the number of people that will require hospitalisation and the death toll. During the 1957 and 1968 pandemics, the excess mortality has been estimated at around one million deaths each. In contrast, 50 million individuals are thought to have died from the 1918 influenza pandemic. Excess mortality during the last influenza pandemics varied from 26 to 2,777 per 100,000 population (Table 2). Adjusted for today's world population, these figures would translate into 1.7 million to 180 million deaths.
According to data fromhttp://www.census.gov/ipc/www/world.html +
In countries such as France, Spain and Germany, the yearly mortality from all causes is around 900 deaths per 100,000 population. A devastating pandemic might therefore, in the course of only a few months, cause three times as many deaths as would normally occur in an entire year. Indeed, social and economic disruption would occur in all countries to varying extents. In a world of extensive mass media coverage of catastrophic events, the resulting atmosphere would probably come close to war-time scenarios. In contrast, a mild pandemic similar to the 1968 episode would go nearly unnoticed and without considerable impact on national healthcare systems and on the global economy.
The concern that the world might be in for a revival of the 1918 scenario is based on the observation that the currently spreading H5N1 virus shares disturbing characteristics with the virus of the 1918 pandemic (Taubenberger 2005). However, if H5N1 is to be the candidate virus for the next devastating influenza pandemic, why has it not yet acquired the ability to spread easily between humans? Over the past years, H5N1 has had both the time and opportunities to mutate into a pandemic strain. Why hasn't it? And if it hasn't in nearly 10 years, why should it do so in the future? It is true that of the 16 influenza H subtypes, only three (H1, H2 and H3) are known to have caused human pandemics (1918, 1957, 1968, and probably 1889 [Dowdle 2006]), and it has even been hypothesised that H5 viruses are inherently incapable of transmitting efficiently from human to human. Shall we one day discover that H5 viruses are not good for human pandemics, because not all possible subtypes can reassort to form functional human pandemic strains? We don't know.
Apart from stepwise mutations that transform an avian influenza virus into a human influenza virus, reassortment is the second way in which new pandemic viruses are generated. The two pandemics that were triggered by this phenomenon occurred in 1957 and in 1968. Both were relatively mild and fundamentally different from what happened in 1918. There is some preliminary experimental evidence that reassortants of the 1918 virus might be less virulent than the co-ordinated expression of all eight 1918 virus genes (Tumpey 2005). Does that mean that pandemics resulting from reassortment events of a human and an avian virus are milder than pandemics caused by a virus which slowly accumulates mutations in order to "migrate" from water fowl hosts to human hosts? We don't know.
The revival of the 1918 catastrophe might also never happen. But the 1918 influenza pandemic did occur, and good planning means being prepared for the worst. As it is impossible to predict whether the next pandemic will result in ~20 or ~2,000 deaths per 100,000 people, the international community should prepare for the 2,000 figure. The three defence lines are containment, drugs, and vaccines.
Containment and elimination of an emergent pandemic influenza strain at the point of origin has been estimated to be possible by a combination of antiviral prophylaxis and social distance measures (Ferguson 2005, Longini 2005). To this purpose, the WHO has recently started creating an international stockpile of 3 million courses of antiviral drugs to be dispatched to the area of an emerging influenza pandemic (WHO 20000824).
If the pandemic cannot be contained early on during an outbreak, rapid intervention might at least delay international spread and gain precious time. Key criteria for the success of this strategy have been developed (Ferguson 2005). However, the optimal strategy for the use of stockpiled antiviral drugs is not known, because stopping a nascent influenza pandemic at its source has never before been attempted.
Once a pandemic is under way - and vaccines have not yet become available - national responses depend on the availability of antiviral drugs. As demand for the drug will exceed supply, stockpiling of antiviral drugs, either in the form of capsules or the bulk active pharmaceutical ingredient, has been considered a viable option by some governments.
The debate over which drugs should be stockpiled is not over. Until now, mainly oseltamivir has been used to constitute stockpiles of neuraminidase inhibitors. After the recent isolation of oseltamivir-resistant isolates in serious H5N1 infection, other antiviral agents to which oseltamivir-resistant influenza viruses remain susceptible, should be included in treatment arsenals for influenza A (H5N1) virus infections (de Jong 2005) - in other words: zanamivir.
The value of adamantanes for stockpiling is less clear. H5N1 isolates obtained from patients in China in 2003 and in one lineage of avian and human H5N1 viruses in Thailand, Vietnam, and Cambodia were resistant to adamantanes (Hayden 2006). However, isolates tested from strains circulating recently in Indonesia, China, Mongolia, Russia, and Turkey appear to be sensitive to amantadine (Hayden 2005).
With regard to the economical impact, there is some evidence that even stockpiling of the costly neuraminidase inhibitors might be cost-beneficial for treatment of patients and, if backed by adequate stocks, for short-term postexposure prophylaxis of close contacts (Balicer 2005). When comparing strategies for stockpiling these drugs to treat and prevent influenza in Singapore, the treatment-only strategy had optimal economic benefits: stockpiles of antiviral agents for 40 % of the population would save an estimated 418 lives and $414 million, at a cost of $52.6 million per shelf-life cycle of the stockpile. Prophylaxis was economically beneficial in high-risk subpopulations, which account for 78 % of deaths, and in pandemics in which the death rate was > 0.6 %. Prophylaxis for pandemics with a 5 % case-fatality rate would save 50,000 lives and $81 billion (Lee 2006).
Once a pandemic starts, countries without stockpiles of antiviral drugs will probably be unable to buy new stocks. In this context it has been suggested that governments provide compulsory licensing provisions, permitting generic manufacturers to start producing antivirals locally under domestic patent laws or to import them from generic producers at affordable prices (Lokuge 2006). In Europe, some governments are trying to build up stocks of the neuraminidase inhibitor oseltamivir for 25 % of the population. The number of treatment doses required to achieve this degree of "coverage" are based on the daily standard treatment course of 75 mg bid for 5 days. However, if doses twice as high, prescribed over a period twice as long (WHO 2005, WHO 2006d) should turn out to be required in a substantial number of patients, a stockpile planned for 25 % of a population might melt away more rapidly than expected.
For detailed information about drug treatment of influenza, seeHoffmann 2006b.
In an ideal world, we would have 6.5 billion vaccine doses the day after the pandemic starts; in addition, we would have 6.5 billion syringes to inject the vaccine; and finally, we would have an unlimited number of health personnel to administer the vaccine.
We don't live in an ideal world. At present, the world has a production capacity of about 300 million trivalent influenza vaccines per year, most of which is produced in nine countries (Fedson 2005). 300 million trivalent influenza doses translate into 900 million univalent doses, enough to vaccinate 450 million people with an initial vaccination and a booster dose - if the H5N1 vaccine is sufficiently immunogenic...
Influenza vaccines are currently prepared in fertilised chicken eggs, a process which was developed over 50 years ago (Osterholm 2005). New technologies may one day be able to develop vaccines more (Palese 2006). A dream vaccine would provide broad-spectrum protection against all influenza A subtypes (Neirynck 1999, Fiers 2004, De Filette 2006), but these vaccines are experimental and years away from industrial production.
When drug and vaccine supplies are limited, healthcare authorities have to decide who gains access to the drugs and vaccines. Who should receive short-supply vaccines and antivirals first: young people or the elderly (Simonsen 2004)? If the standard used to measure effectiveness of medical intervention was "numbers of deaths prevented," then perhaps the elderly should be given priority - assuming they can produce an adequate antibody response to the pandemic vaccine. But if the concern is to minimise the years-of-life-lost, then the vaccine may be better used in young and middle-aged adults (Simonsen 2004).
The Australian Government has acknowledged that, in the event of a pandemic, its own stockpile of antivirals will be limited and reserved for those on a confidential rationing list (Lokuge 2006). Who are they? Physicians, fire fighters, police forces - or politicians and other VIPs? Experts urge that a framework for determining priority groups be developed prior to the start of a pandemic and that such a scheme should be agreed on beforehand and be flexible enough to adapt to the likely level of disaster at hand (Simonson 2004).
The good news from epidemiological research is that past pandemics gave warning signs. In the spring of 1918, a pandemic wave occurred 6 months before the second deadly autumn wave (Olson 2005). The Asian H2N2 influenza virus was characterised by early summer, 1957, but significant mortality in the United States did not occur until October - and in 1968, the pandemic wave of mortality in Europe peaked a full year after the pandemic strain first arrived (Simonson 2004).
Epidemiological studies of the 20th century pandemics offer some insight into what can be expected when the next influenza pandemic occurs (Simonson 2004):
The next pandemic will come, but we do not know when. We do not know how severe it will be. Will it be mild like the last two pandemics of 1968 and 1957, when the new pandemic strain resulted from the reassortment of the pre-existing human strains and an avian influenza strain? Or will it be as catastrophic as the 1918 pandemic?
Only the future will tell. Let's be prepared!