The Viral Storm (2 page)

The interior of the Oxford University Museum.
(©
Chris Rimmer
)

As I began my work in Uganda to study chimpanzees, my professors cautioned me that any convincing proof that chimpanzees were medicating themselves with plants would require an understanding of the infectious diseases they were treating. Unless I could show that the use of the purported medicines decreased the burden of disease, my results would be speculative at best. I needed to understand what infectious diseases plagued the chimpanzees. I knew little about microbes, so I approached Andy Spielman, a professor at Harvard’s School of Public Health and one of the few people at the time focused on understanding the ecology of microbes in nature. Despite his lab full of fellows and students and his focus on North America rather than the wilds of Africa or Asia, he kindly took me under his wing. Thus began my research on what was known about the infections of chimpanzees. Once I began thinking about microbes I never looked back. And central to my studies would be the viruses.

Viruses evolve more rapidly than any organism on the planet, yet we understand less about them than any other form of life.
2
The study of viruses provides a scientist with the opportunity to discover new species and catalog them in a way reminiscent of the world of the nineteenth-century naturalist, which had so fascinated me during my time at Oxford. A scientist can productively spend an entire career looking for new species of primate and never find one, but new viruses are discovered every year. They also have exceedingly short generations, so we can watch them evolve in real time—an ideal system for someone interested in understanding the process of evolution. Perhaps best of all from the perspective of a young scientist, there was important and urgent low-hanging fruit in this discipline: some of these viruses kill us. Thus new discoveries need not only lead to an improved understanding of nature, but they can also have important and rapid applications for controlling human disease.

*   *   *

Controlling the spread of human disease was at the forefront of public health efforts in early 2004 when the news broke of Kaptan’s death from H5N1. His death was the first confirmed mortality from this virus, the so-called bird flu, in Thailand. The truth is that while they may jump to us via other animals,
all
human influenza viruses ultimately originate in birds, so the popular designation of the virus as “bird flu” can irritate scientists. Yet within a month that name would become a mainstay of news shows and a topic of discussion for people throughout the world.

The scientific name for the virus that killed Kaptan, HPAIA (H5N1), is quite descriptive for virologists. It signifies that the virus is a highly pathogenic avian influenza A-type virus and provides the particular hemagglutinin (H) and neuraminidase (N) protein variants particular to this virus strain. But its true significance is actually much more straightforward.

H5N1 is important because it kills remarkably effectively. The virus’s case fatality rate, or the percentage of infected individuals that die, is around 60 percent. For a microbe, that’s incredibly deadly. As a comparison, we can look back to the devastating 1918 influenza pandemic. While estimates for the 1918 pandemic are imperfect, it is thought around fifty million people died. That’s the equivalent of 3 percent of the entire human population at that time, an almost unimaginable catastrophe. To put this in context, more people died from the 1918 influenza pandemic than the total number of soldiers thought to have died in battle during all twentieth-century wars
combined
. More deaths caused by a simple virus, less than one hundred nanometers in diameter and with only a paltry eleven genes, than were caused by all the battles in WWI, WWII, and all of the other wars in our last war-riddled century. Despite the enormity of the 1918 plague, the highest estimates for its case fatality rate are in the range of 20 percent, and it was almost certainly much lower than that; more careful estimates suggest around 2.5 percent.
3
Recall that H5N1 had a 60 percent fatality rate, far greater than that of the influenza virus that caused the 1918 pandemic.

But while deadliness is an important, dramatic, and ongoing obsession of the media, it is only one piece of the puzzle for microbiologists. In fact, some microbes kill virtually all people they infect: a perfect 100 percent mortality rate. And yet such microbes do not necessarily represent critical threats to humanity. Viruses like rabies, which naturally infect a number of mammal species, and the herpes B virus, which naturally infects some species of Asian monkey, will kill all the people they infect.
4
Yet unless you’re someone who is exposed to rabid animals or works with Asian monkeys, these microbes do not represent a major concern for you. That’s because they don’t have the capacity to spread from person to person. In order to be catastrophic, a microbe needs both the potential to harm or kill
and
the potential to spread.

In early 2004 there was no way to know how efficiently H5N1 would spread. Since it comes from a class of viruses that often do spread, the influenza viruses, the possibility was there. And if H5N1 were to spread in the same way that the 1918 influenza virus spread, it would create a calamity unlike any seen before in human history.

*   *   *

As impressive as H5N1 is at killing, H1N1, the so-called swine flu,
5
is equally impressive in its capacity to spread. While no one knows exactly when the H1N1 pandemic began, by August 2009, less than a year after it was first recognized, the WHO announced estimates suggesting that the virus could eventually infect over two billion people, roughly a third of the entire human population. The natural drama of this would be hard to overstate. While less visually dramatic than other forms of natural disaster, the ability of this phenomenon to touch people everywhere on the planet made it a powerful force of nature. A virus that likely infected few people in early 2009 moved around the globe to inhabit a sizable percentage of the entire population of the human species in less than a year. This occurred despite the best efforts of global public health infrastructures—structures that we tend to be very proud of and feel protected by. Yet while the case fatality rate of H1N1 is estimated to be well below 1 percent, paling in comparison to H5N1, the sheer number of people it infected has made it a real global killer. One percent of two billion is a lot of lives.

To help us understand the real threat of an outbreak, we turn to a concept in epidemiology called R
0
, or the
basic reproductive number
. For any epidemic, R
0
is the the average number of subsequent infections that each new case results in (in the context of a population with no prior immunity and no control efforts). If, on average, each case of a new epidemic leads to more than one subsequent infection, the new epidemic has the potential to grow. If, on average, each case leads to less than one subsequent infection, it will peter out. The elegant concept of R
0
helps epidemiologists distinguish between epidemics likely to “go viral” and those likely to go extinct. It’s basically a measure of scalability.

Risk interpretation is not a trivial matter, for either the public or the policy makers. In the case of H1N1 or H5N1, the potential costs of not rushing to develop a vaccine or working to decrease transmission could have been global and catastrophic.

Critically, microbes are dynamic—they do not exist in stasis. If H5N1, the deadly bird flu, gains the right combination of genetic mutations it needs to spread effectively, the results will be destructive in a way that, however less visually dramatic, will make even the most deadly earthquakes seem like a walk in the park. And if H1N1, the rapidly spreading swine flu, were to increase in virulence even minimally, its potential to kill would be striking. Neither scenario is implausible. As chapter 1 will explore in detail, influenza and a range of other viruses have an incredible capacity to negotiate the environment of their human hosts. They mutate rapidly, even swapping genes among themselves, a process referred to as reassortment.

It is just such reassortment that concerned me, and other scientists, in 2009. As the H1N1 virus spread explosively around the world, there was a nontrivial chance that it would run into H5N1 in people or animals, setting the stage for a potentially cataclysmic series of events. These are the kind of events we work to uncover early before they spread. When infected simultaneously with both of these influenza viruses, a particular human or animal could become a potent mixing vessel, providing the perfect opportunity for the bugs to swap genes. How would this happen? In a sort of sexual reproduction, H5N1 and H1N1 could produce mosaic daughter viruses with some of the genes of one virus and some of the genes of the other. Such reassortment events can occur in individuals infected with multiple similar viruses. In the case of H1N1 and H5N1, if that mosaic daughter virus inherited the potential to spread from its H1N1 parent and the deadliness of its H5N1 parent, the resulting virus would be both highly transmissible and highly fatal—exactly the formula for global impact that we most fear.

*   *   *

A mad rush to respond to pandemics has been the mainstay of global public health for the last one hundred years. Now a small but vocal group of scientists and I have begun to argue that we must do better than just
respond
to pandemics by scrambling for vaccines, developing drugs, and modifying behaviors. This traditional approach has proved a failure for human immunodeficiency virus (HIV), which nearly thirty years after its discovery continues to spread, currently infecting over thirty-three million people at last count.

But what if we had been able to catch HIV before it spread? This virus was in humans for over fifty years before it spread widely. It spread for another twenty-five years before French scientists Françoise Barré-Sinoussi and Luc Montagnier, who would go on to win a much-deserved Nobel Prize for their work, finally discovered it. How different would the world be had we stopped it before it left central Africa?

The idea that we might one day predict pandemics is a new one. The first time I heard someone discuss it was around ten years ago in the Johns Hopkins office of Don Burke, a retired medical colonel and world-renowned virologist from the Walter Reed Army Institute of Research (WRAIR) who had devoted his life to more traditional approaches to control disease before he took on a professorship at Johns Hopkins University’s Bloomberg School of Public Health. Don had hired me as a postdoctoral fellow at Hopkins a few years earlier when I was finishing up my doctoral work in the rain forests of north Borneo studying the ways that mosquitoes and other blood-feeding insects helped microbes move between primate species.

Unable to track me down himself, Don had managed to find my mother in Michigan and gave her a call. My mother, whom I would speak to infrequently on trips from our forest research station, scolded me, saying that a “general” from the US military had called her. She asked what kind of trouble I had managed to get myself into. Fortunately, all Don wanted was to see if I’d help him set up a project in central Africa to understand how viruses emerge from animals into human populations.

During the years that followed, in addition to the long, slogging work of building research capacity to catch new microbes in central Africa and Asia, Don and I engaged in hours of conversation in the field and in his office in Baltimore. We made numerous beer bets on scientific problems and asked hard questions about the future of our field. I remember the day when I first heard Don suggest that the future would include not only
response
to pandemics but
prediction
. It seemed a bold but logical idea, and we quickly moved toward thinking about ways that it could actually happen. Those early conversations formed the foundation of the work my colleagues and I conduct now, setting up and running listening posts at microbial hot spots around the world aimed at catching new microbes locally before they become global pandemics.

Among the things we listen for are novel influenza viruses, like H5N1 and H1N1. Unfortunately, the world too easily becomes complacent to threats like H5N1 and H1N1. These and other threats fade quickly from the media’s attention. Most of the world doesn’t think about either of these viruses seriously. Yet neither virus has gone extinct, and the threat is perhaps as great now as it was when they were each first noticed. Both viruses continue to infect human populations. For example, in 2009, years after the media forgot about the virus, H5N1 caused at least seventy-three laboratory-confirmed cases. This is almost certainly a substantial underestimate of the number of actual cases and does not differ notably from the number of confirmed cases from previous years. H1N1 cases continue to spread as well. We have detected them even in the most remote forest areas where we “listen.”

*   *   *