On the first morning – unshaved, wearing ugg boots, tracksuit pants and a T-shirt – I drove my wife to work, returned home to a leisurely read of the paper over breakfast, had a long shower and arrived at my desk just minutes after my self-imposed deadline. I quickly scanned the top stories on the ABC News website and one item caught my eye.
‘Experts say the swine flu outbreak currently stoking fears of a global epidemic poses a greater risk to Australia, with the onset of winter bringing the peak of the flu cycle.’ I checked the BBC site: ‘Schools in the capital city of Mexico are closed due to an outbreak of a disease called swine flu.’
Now this was strange. We had been expecting the next influenza pandemic to be caused by a mutation in the avian influenza virus, the H5N1 strain that mainly affects birds, not the H1N1 virus associated with pigs. I rang my wife, who works in infection control, and asked her what was going on.
‘I can’t talk,’ she said. ‘I have to go into a meeting with the chief health officer in a minute. There’s a bit of a flap on in here.’
There were five infectious diseases specialists in our unit at the hospital at the time. Two were on duty, one was overseas working on a project, one was in the Kimberley on holidays, out of phone range and I was … ‘Oh no, you can’t do this to me,’ I said out loud. I put down the phone and tried to concentrate on writing, without success. I turned on the radio and listened to the 10am bulletin. The top story was swine flu.
* * * * *
There had been three influenza pandemics in the preceding 100 years: 1918, 1957 and 1968. The 1918 Spanish flu pandemic was the greatest natural disaster of the 20th century. Estimates vary, but a minimum of 20 million – and possibly up to 50 million – people died, at least double the ten million killed on the battlefields of World War I. In a world without antibiotics, the secondary bacterial infections that followed the epidemic were often deadly. And there were no intensive care units (ICUs) to provide the care needed to allow patients time for their own immune system to win the fight against the pathogen. Around a third of the global population was infected with Spanish flu and more than 2.5 per cent of humanity died.
The other pandemics were, by comparison, not in the same league: the Asian flu of 1957 is estimated to have caused around two million deaths worldwide, while the Hong Kong flu of 1968 is thought to have killed at most one million people.
I became intimately acquainted with Hong Kong flu as an eight-year-old when it nearly killed my father. I remember my father lying in the single bed in the spare room, the bedclothes pulled up over him as he shook his way through what I now know to be a rigor. When my mother had picked him up from the train station the previous night he had been completely well, but during the three minute drive home he started complaining of chills and had started to shiver. By the time he had put his case down in the bedroom he could hardly talk because of the shaking.
The Hong Kong flu was on that evening’s television news and on the front page of the day’s Sun newspaper, so my mother had already made the diagnosis by the time our GP arrived the next morning. I remember her concern and the hushed kitchen conversations with friends suggesting that he needed to go to hospital. He coughed and spluttered for weeks after the worst was over and he always remembered the illness as one of the worst he had experienced.
For 40 years following the 1968 outbreak it was all quiet on the influenza front. There was a false alarm for a swine flu pandemic in 1975, and virologists remained on alert, knowing the next pandemic was not a matter of if but when.
* * * * *
There are three types of influenza virus: A, B and C. Pandemics are always caused by influenza A. Infection with influenza B can be severe but is rarely fatal and infection with C is uncommon and usually mild. All three types have two main proteins on their surface, haemagglutinin (H) and neuraminidase (N), which are responsible for the confusing nomenclature. There are 16 H and 9 N subtypes but only H1, H2, H3, N1 and N2 commonly cause human disease. The Spanish flu and the 2009 swine flu were H1N1, the Asian flu was H2N2 and the Hong Kong flu was H3N2. Bird (avian) influenza is H5N1.
To understand why pandemics occur we need to know a little about the way the influenza virus reproduces and how our immune system responds to it. Viruses commandeer the proteinmaking machinery of the body’s cells to generate tens of thousands of copies of themselves. Every time a flu virus replicates (or copies) itself there is the potential for an error (mutation) to occur, resulting over time in significant changes to the structure of the virus and altering the way the host’s immune system ‘sees’ it.
When your body first encounters a virus it takes days or weeks for the immune system to produce specific antibodies against it. If the same virus is encountered in the future, your body’s reaction time is much faster because the antibodies can destroy the virus before it has time to invade your cells, replicate and cause disease. If a virus has mutated, however, the antibodies are unable to recognise it, giving the virus time to invade, replicate and cause disease all over again.
Some viruses mutate very little during replication. The measles virus that we are exposed to in childhood will be almost identical to the one that we may be re-exposed to decades later, so immunity is life-long. Influenza A, however, is notable for its sloppy proofreading, and at least one mutation occurs with every copy. Most of these mutations are minor, but enough of them can accumulate, by successive passages through infected human hosts, at the end of the flu season to make the structure of the virus quite different from what it was at the beginning.
This is known as antigenic drift and is a gradual process that allows the human immune system to almost keep pace with viral evolution; you might be susceptible to next year’s strain, but you are likely to have at least partial immunity.
Antigenic drift won’t lead to a pandemic: for that you need a sudden and profound change in the viral structure. This produces a virus strain that no one in the population has previously encountered and developed immunity to – a seismic phenomenon called antigenic shift. A shift requires more than just a few typos while transcribing the viral genetic code; it requires viruses to have ‘sex’ with each other.
Higher-order organisms, such as us, combine half their genetic material with half that of a member of the opposite sex to increase the genetic diversity of their offspring. Lower-order organisms mainly reproduce by making identical copies of themselves, but occasionally they display the traits of sexual reproduction. Sex between consenting viruses is more accurately, if less evocatively, known as reassortment.
A human influenza strain can reassort with a bird or pig strain and the resulting hybrid may have properties that produce a very severe disease in humans. The frequently cited doomsday scenario is a reassortment of the bird flu strain H5N1 with a human influenza strain, allowing bird flu to be transmitted from human to human. The reason infectious disease specialists worry so much about such an event is that the world’s insatiable appetite for poultry has produced a chicken population explosion, particularly in Asia. With so many birds living in close proximity to humans, viral reassortments are occurring all the time and the emergence of a deadly flu strain is considered inevitable.
* * * * *
All epidemics start quietly. Their early stages are usually undetectable by the routine surveillance mechanisms health departments have in place. By the time you know something is going on, the infectious agent is well established in the population. The swine flu strain responsible for the 2009 pandemic was first detected in Mexico in April of that year, in four-year-old Edgar Hernandez. This little boy recovered completely, but the viral strain would have been circulating in the community for weeks or even months before.
Within a few days of Edgar’s highly publicised diagnosis dozens of other people across Mexico were reporting fevers, muscle aches and coughs. As I read the news bulletins each morning over the next few weeks it became clear to me that an antigenic shift had occurred. By the time the virus causing their symptoms was i
dentified as swine flu (H1N1) in the US Centers for Disease Control and Prevention laboratory in Atlanta, Georgia, thousands more had been infected.
While the new strain was very similar in composition to the 1918 Spanish flu, it wasn’t immediately clear if it possessed the same virulence as its 20th century predecessor. But the World Health Organization (WHO) had been preparing for this event for more than a decade and an elaborate influenza pandemic plan had been formulated. By the end of April 2009, cases of swine flu had been reported in the US, Europe, India, Pakistan, Bangladesh and New Zealand. On 11 June Margaret Chan, the WHO’s Director-General, told a press conference in Geneva: ‘We are all in this together, and we will all get through this, together.’ She had just declared the beginning of the first influenza pandemic of the 21st century.
* * * * *
Influenza is an autumn and winter disease, so although the virus was already circulating in the northern hemisphere in early 2009 it was not expected to take off there until later in the year. The European and US public health communities were, therefore, closely watching the behaviour of the epidemic when it reached Australia and New Zealand during our winter, mid year. As had been predicted, attempts to contain the virus proved futile in a world interconnected by air travel, although many jurisdictions, including some in Australia, attempted to create a cordon sanitaire (a quarantine line), reasoning that slowing the tempo of the epidemic by even a small degree would buy some time to get local plans in place.
One of the planks of the pandemic plan is the use of antiviral prophylaxis – taking steps, such as the administration of drugs to people exposed to the virus, to reduce the spread of the illness and break the chain of transmission. The pharmaceutical antiviral agents that work against influenza, zanamavir (an Australian discovery with the trade name of Relenza) and oseltamivir (known commercially as Tamiflu), don’t have anything approaching the potency that antibiotics have against severe bacterial infections. These anti-flu drugs inhibit the neuraminidase enzyme of the virus (the N of H1N1) and, as they have few side effects on human cells, they are quite safe.
The original studies that supported the licensing of these drugs showed that patients with influenza who received either agent were, on average, free of symptoms about one day earlier than those who received a placebo – a real, but only modest, benefit. Their effectiveness against serious, life-threatening influenza has never been tested by a randomised clinical trial. Each country has established a national stockpile of the drugs and these are dispensed to those most in need during a pandemic: initially, healthcare workers on the frontline of healthcare delivery.
The problem with their use as a prophylaxis is that they only work while you are taking them: if you stop the drug and are then re-exposed to influenza you can still become infected. Antiviral prophylaxis is a little akin to a World War I soldier donning a gas mask in the trenches when the mustard gas has been released: you are protected only as long as you have the mask on. To be able to safely discard the mask, to continue the metaphor, you need a ceasefire, and this can only be achieved with vaccination.
* * * * *
Flu can manifest in many ways, with symptoms ranging from trivial and short-lived muscle aches and headaches to severe muscle pain, fever, chills and uncontrollable shaking, leading to life-threatening pneumonia, organ failure and death. Every year thousands of people around the world die from what is known as seasonal influenza, with the most affected being the frail and elderly and people with pre-existing serious chronic illness. During the 1918 pandemic the sequence of completely well, to moribund, to dead could occur within 24 hours, and the disease targeted the younger and otherwise well members of the population.
We soon discovered that the 2009 swine flu was a mild disease in most people. There were exceptions: a fortnight after contracting swine flu two doctors at my hospital developed a neurological condition called Guillain-Barré, which damages the peripheral nerves and produces muscle weakness. One continued to work until he found himself unable to get out of the chair in his consulting room; the other woke to find himself permanently unable to move many of his facial muscles.
Our hospital was busy, but because most people sick with the flu could be managed at home, there was no need for me to rush back to work and help man the pumps. A smaller number of critically ill patients required the skills of the intensive care doctors, not my humble services. My wife gave me daily updates and, while subtly mentioning the important role she was playing, made it clear that I would just be in the way. At least my dog was talking to me again.
By July it was apparent that the death rate overall was between 0.001 and 0.03 per cent of those infected, much lower than the 0.1 per cent of the 1968 pandemic. The attack rate turned out to be relatively low too, with antibody testing performed after the epidemic showing that 10–20 per cent of the population had been infected.
While some of us were busy reassuring the population that this was no 1918 Spanish flu, the intensive care doctors had witnessed something new and frightening: an epidemic of viral pneumonia that affected pregnant women, children and the middle-aged. In Australia, a total of 722 patients with H1N1 were admitted to an intensive care unit and 103 of these died, including seven pregnant women and seven children. Many more would have died if not for the heroic measures instituted in some cases – the sickest patients were treated with extra-corporeal membrane oxygenation (ECMO), the equivalent of putting someone on cardiac bypass for open-heart surgery, but instead of stopping after three hours it was continued for days or even weeks until the patient’s lungs had recovered.
After the 2009 flu epidemic was over, I attended a conference where the director of a major urban ICU said that during the epidemic his unit had reached the upper limit of its ability to cope with critically ill patients. If the number of cases had been any greater, he believed, the unit would not have been able to provide adequate care for all who needed it. If that had occurred, many older people with complications of the flu who would otherwise have received intensive care would have been denied it to allow younger people, pregnant women and other high-risk patients access to treatment.
As it was, the absence of disease in the elderly was striking. It was explained by data published in the New England Journal of Medicine revealing that they had gained protection through exposure to a similar H1N1 strain in the past. All good doctors read the medical literature and learn from the published work of others, but the biased view that one gains from personal experience can be very difficult to shake. Many ICU doctors were perplexed – and some were angered – by the casual approach to the epidemic displayed by some of their non-ICU colleagues, who, lacking direct exposure to the sickest patients, and despite the published evidence, promulgated the belief that swine flu was no different from seasonal influenza.
* * * * *
Criticism of the public health and medical response to the pandemic began to circulate almost as soon as the virus itself was identified. In Australia, the harshest comments were aimed at the vaccine policy. As we have seen, antigenic drift and shift allow flu viruses to evade the human immune system and new vaccines that anticipate these changes must be manufactured annually. There is no universal flu vaccine that will provide years of protection as the vaccines for polio, measles, and hepatitis B do, for example.
Every year an expert WHO panel decides what antigens the new seasonal influenza vaccine should contain. (It is probably not unfair to say that their deliberations are based on a mixture of experience, science and necromancy.) The viruses chosen for inclusion must be inoculated into fertilised hens’ eggs as part of a laborious and complicated vaccine manufacturing process. New techniques for vaccine production are on the horizon, but the current methods are essentially the same as they were 50 years ago. In a pandemic setting, most manufacturers seek government underwriting, as there are many commercial risks associated with the rapid roll-out of sufficient vaccine to cover an entire population. It takes four to five months to get a vaccine read
y for distribution, by which time the epidemic may have already burnt itself out or been shown to cause only a mild illness that wouldn’t justify mass vaccination.
Inevitably, there will be side effects associated with the vaccine. When tens of millions of doses are administered, even vanishingly rare adverse reactions will occur, and indemnity against claims for harm are likely to be required. Within weeks of the beginning of the 2009 pandemic, the Australian government contracted pharmaceutical manufacturer CSL to produce 21 million vaccine doses – enough for the entire Australian population. The four month lead time meant the decision had to be made when the true virulence of the virus was still unclear. It was not even known if one or two doses of vaccine would be required to produce adequate protection. To meet the unprecedented production demands, CSL had opted for multi-dose vials instead of the single-dose syringes that were usually supplied. The use of multi-dose vials has the potential for cross-contamination of blood-borne viruses and the decision was condemned by many infectious diseases experts. CSL argued that the risk of cross-infection was outweighed by the imperative of having enough vaccine ready in time. By the time it was available, in September 2009, the low death rate had become apparent and the subsequent uptake of vaccine was modest, with less than a quarter of the population opting to receive it. Millions of doses were discarded.
Australia’s federal government was loudly attacked in some medical quarters for wasting tens of millions of dollars on the vaccine. And state governments were criticised for their attempts at quarantine in the early weeks of the epidemic. The WHO was accused of having over-reacted by invoking its pandemic plan and of being conflicted because of the presence of pharmaceutical and vaccine producers on some of its expert panels.
There is a well-known medical saying that things always look clearer through the retrospectoscope: what is self-evident in hindsight is hidden to us in earlier days. It is impossible to predict the course of an epidemic with any degree of precision during its earliest stages. The complexities of human behaviour, the unpredictability of the immunological response of the population and our rudimentary understanding of the basic biology of the influenza virus mean an epidemic is only truly understandable after it has occurred. If the 2009 pandemic had turned out to have the same death rate as the Hong Kong flu, Australia could have expected up to 4000 deaths; if it had mirrored the 1918 pandemic this number would have been ten or even 20 times higher.
The Best Australian Science Writing 2012 Page 11
