Maybe this time: What we’ve learned in the decade since SARS


Illness fascinates us. It’s in our books, our movies, our board games, our whispered hallway gossip. It stars in an endless rotation TV medical procedurals, grizzled surgeons shouting back and forth about the fate of their patient’s life.

An episode of such a show might start like this: On an early summer day in Saudi Arabia, a middle-aged man checks into the hospital. He’s developed a fever and a cough, and is having trouble breathing. He isn’t on any long-term medications. He doesn’t smoke. The doctors guess infection, admit him to the ICU, and prescribe a polysyllabic combination of treatments. Unfortunately, they don’t work. Eleven days later, he dies.

A few months later, a different man from another country falls ill with something suspiciously similar. His admission to the hospital stirs a flurry of discussion on professional mailing lists and science news blogs. The disease is soon identified and linked to a famous, relatively recent epidemic that killed hundreds. The World Health Organization (WHO), the international hub for public health information, issues an official alert notifying hospitals to be on the lookout.

Now let’s say that these aren’t really the imagined scenes on our television. Say this is all something unfolding right now.

This new disease is still unnamed. Six people have been infected, five of whom are in Saudi Arabia, two of whom are related. We aren’t sure yet how it is transmitted, how many more patients are out there, or how best to treat it, but we do know that it’s in the same virus family as SARS.

SARS—fully named Severe Airborne Respiratory Syndrome, Coronavirus SARS coronavirus to virology researchers, the phenomenon described by one New York Times journalist as “an epidemic of illness and fear”—first appeared in China on November 16, 2002. It began with one man with what seemed to be atypical pneumonia, who checked into a hospital in China and promptly died. By the time the WHO issued their first public notification about his sickness in February 2003, SARS had already spread to over three hundred people in southern Asia, killing four. The then-unidentified virus snaked its way through hotel lobbies and airport security checkpoints, carried in the lungs of unwitting carriers later nicknamed “superspreaders” in the scientific literature. Total damage: 8,098 cases in 27 countries. 774 dead.

For an epidemic, SARS was relatively mild. The 1918 Spanish flu, for example, killed several million people, a whopping (and mostly young and otherwise healthy) three percent of the world population. And while SARS has essentially been contained and eradicated, other familiar diseases continue to take lives around the world today.

“The SARS coronavirus in the end killed a couple of hundred people. That’s it,” says virologist Marc Van Ranst. Van Ranst has researched viruses like SARS at the Rega Institute of Medical Research, and he oversaw Belgium’s national response to the swine flu epidemic in 2009. “Thousands of people every day die because of AIDS, malaria, tuberculosis. These are really the bad bugs… and they don’t get on the front page every day.”

But despite all this, SARS was still deeply scary to us, and not just because some people love a good panic. SARS arrived in a wave of uncertainty, announced amid headlines of unsolved anthrax attacks and the newly hatched war on terror. It had incubated in a farm animal halfway across the world that most of us hadn’t ever heard of. It infected its way past national borders with apparently uncontrolled abandon.

We didn’t know how to treat it. It didn’t feel like we knew anything. As an early report from the Center for Disease Control put it, “The severity of the illness might be highly variable, ranging from mild illness to death.”

SARS invited the world to take a critical look at the way that we handle emerging epidemics. The debate began even before WHO officially pronounced the outbreak under control in July of 2003. People questioned the amount of time it had taken for news of the illness to break outside of China; the manner in which governments and international bodies had responded to the crisis; the way that information was communicated throughout the outbreak; and the uneven treatment process at different hospitals.

Now, almost exactly a decade after the first SARS patient, this new and similar illness in the Middle East has begun to capture our attention. The WHO recommended that hospitals quarantine their infected patients, and they have asked doctors to test all cases of severe pneumonia—even those involving people who haven’t been to Saudia Arabia or Qatar—for signs of the virus. This virus may not be the next epidemic, but it all begs the question of whether something like SARS could happen to us again.

University of Hong Kong virologist Leo Poon, one of the first scientists to identify the genetic code of the SARS virus, definitely thinks so.

“If we look at the evolutionary history of coronaviruses, there are many interspecies transmission events from one animal to another, and from animals to humans as well,” Poon says. “This suggest that there might be another event of SARS in the future.”

We have had ten years to think about the lessons SARS can teach us, about surveillance, about disease identification, and about coronavirus treatment. If history now repeats itself, will we be ready for it?

In order to tackle an epidemic, we have to know that it’s happening. This can be a surprisingly difficult trick.

At first, the Chinese government tried to contain the news of the nascent SARS outbreak to its borders, but information leaked through the a mailing list of epidemiology professionals. And though the first WHO report on SARS didn’t arrive until three months after the start of the outbreak, one computer application called the Global Public Health Intelligence Network (GPHIN) had picked up the first mention just eleven days after the start of the spread.

GPHIN was started by the Canadian government in 1997 to track several hundred internet sites for mentions of specific communicable diseases, and it was one of the first attempts at this kind of massive public surveillance. The system had several important drawbacks. It only handled a few languages. There was a limit to the number of users who could access the site at any given time. It wasn’t capable of weeding out garbage and misinformation from the vast onslaught of internet data.

But others realized the potential of such a program. If the WHO had started fighting SARS earlier, they could have prevented hundreds or even thousands of people from being infected. Soon, epidemiologists such as Google’s Larry Brilliant began calling for an even more powerful electronic surveillance system than GPHIN—one that could effectively, independently, and transparently use early detection to help stop the spread of deadly viruses.

Boston Children Hospital’s HealthMap is one modern actualization of Brilliant’s imagined system.

“HealthMap grew out of an interest to support the traditional hierarchical infrastructure,” explains Sumiko Mekaru, a surveillance epidemiologist at HealthMap. “We’re bringing organization to the chaos of online sources, and we’re bringing a way to access the public and let them participate and take some ownership of their own public health in a  way that hasn’t been done much before.”

Every hour, HealthMap crawls news feeds in ten different languages across the internet, looking for mentions of familiar infectious diseases. It aggregates this information in an online map and daily email to healthcare professionals. The organization has also spawned several online portals and iPhone applications that allow individuals to self-report local outbreaks.

HealthMap is a fast way to track diseases we know to worry about. However, unknown diseases like new coronavirus outbreak still pose a significant challenge.

“The major diseases are a little easier for our system to detect,” says Mekaru. “It’s a little more challenging when you have the undiagnosed or mysterious illnesses show up, because even having the right vocabulary or the right search string can be challenging.”

One quick fix is to search for more general reports that mention possibly alarming symptoms like respiratory illness or rash. This has helped HealthMap uncover smaller outbreaks, such as a parasite spread at a rock concert about a year ago.

In the future, the gap might be filled by more advanced surveillance techniques. Several groups, HealthMap among them, are working on programs that can digest Twitter information to find mentions of outbreaks, and hope to find a good method to extract the signal from the (very loud) noise. Other projects want to develop useful ways of tracking indirect indicators, like the amount of flu medicine being sold at local pharmacies.

“There are a lot of different methods you can use to detect disease outbreaks,” says Mekaru. “I don’t think it’s realistic for any single entity to employ all of them, but I think it’s important that the public health community and the surveillance community has as many people as possible using this data and sharing it freely.”

At the end of the day, more data—and more ways to sift through it, and more ways to share it—are always better. The WHO has attempted to promote freer sharing of information by revising their International Health Regulations, but ultimately has no power to mandate governments to act in a particular way. And word of Saudi Arabia’s new virus spread in much the same way as the first news SARS did: through email chains between virologists in the know.

While the rest of us are gaping at our computer feed, scientists have to figure out what the heck the mystery illness actually is.

The diagnostic process begins with independent patients, as their doctors try to match the symptoms against diseases they already know. With SARS, it first looked like pneumonia. Different patients had inconsistent complaints, but all cases had a fever above 100.4˚F and some kind of difficulty breathing. That pair of symptoms could easily fall under the umbrella of dozens of different diseases, and one by one, hospitals ruled out anthrax, hemmorhagic fever, avian influenza, pulmonary plague, Weil’s disease, Legionnaire’s disease, typhus, or an unusual strain of chlamydia .

The mystery finally began to clear up in March of 2003, after the WHO organized a collaborative effort of research laboratories from around the world to sequence the new disease. Sequencing is the process of peering inside a virus, which is a vessel for long strands of information written in the four-base language of either RNA or DNA. This information allows scientists compare the new virus against familiar ones, which in turn can help doctors find an effective treatment for their patients.

A particular technique called next generation sequencing, which was in its infancy at the beginning of the SARS outbreak, is one of the best ways to do this. Next generation sequencing reveals a passage of a particular strand by breaking it into many small pieces, and then comparing those pieces against a genome that we already know. And within in a month, scientists had successfully pinned SARS to a a virus they named SARS-CoV, a never-before-seen strain of the spiky-looking coronavirus family. a spiky-looking bug dubbed SARs-CoV. SARS-CoV was a only the third of its kind of be found in humans.

The processing of sequencing SARS was fast—for 2003. Now, next generation techniques are standard practice in new epidemics, and the process has gotten much, much better.

“If you really put your mind to it, it takes days to characterize a novel virus,” says virologist Marc Van Ranst. Van Ranst led a coronavirus research team at the Rega Institute of Medical Research, and later oversaw Belgium’s national response to the 2009 swine flu. “It would have taken months ten years ago. If you go back twenty years ago, it took years.”

Higher throughput—more data produced in a single round of sequencing—is one of the main reasons behind the increased speed. Today’s sequencing turns out one thousand times more data than it did five years ago. (Compare that to computer hardware’s famous Moore’s Law, which considers doubling a chip’s transistors every 18 months an impressive pace.) That means that we can look at larger volumes of genomic data, and we can test the virus against more estimations at once.

Sequencing is also faster because of advances in polymerase chain reaction primers. Primers are small snippets of DNA or RNA that are required to copy the unknown virus’s own genetic code. Back in 2003, scientists needed to use a different primer for every single guess as to what the virus might be. Now, we have consensus primers, which can be used to rule out an entire virus family.

Improvements in next generation sequencing have pushed it to the front lines of all manner of recent outbreaks, from yellow fever in rural Uganda to E.coli in European produce. In the case of the new coronavirus, doctors turned to sequencing after ruling out several other theories like influenza virus, and officially announced their discovery at the end of the summer.

If we don’t know the possible origin of the virus, or if it’s significantly different from the ones we know, sequencing once again becomes a time-consuming procedure. Some scientists say that the procedure needs to be more user-friendly, allowing doctors who might not be particularly accustomed to the procedure to sequence a new virus at their local laboratory. And there isn’t a push to work quickly or immediate cause for concern, as there wasn’t initially with the first death in Saudi Arabia in June, then we just aren’t going to know about a new disease no matter how fast the process is.

At the rate medical technology is developing, Van Ranst wouldn’t be surprised to see diagnostic machines go much further in his lifetime. He thinks that doctors in the future will, like the medical officers on Star Trek, be able to point a handheld device at their patient and receive an accurate, real-time diagnosis.

“Gene Roddenberry wasn’t that far off,” he says. “We’re not there yet, but give me ten years.”

But despite our successful efforts to sequence the virus’s genome in 2003, we still had no idea how to treat SARS. There had never been much research done on human coronaviruses. Despite the many related illnesses in animals like mice, cows, chickens, wales, and cats, we had only encountered two coronavirus strains that actually affected humans, and both kinds merely resulted in a common cold.

Without a familiar treatment plan to turn to, doctors often elected to try everything they had.

“We starting to treat individuals with a cocktail of drugs with the hope that one of them was going to do something,” explains Donald Low, chief microbiologist at Mount Sinai Hospital. Low was a major spokesperson and strategist during the Toronto outbreak, which started after an elderly woman picked up SARS on a trip to Hong Kong. “We had cases coming out of everywhere, and we were just throwing the kitchen sink at them.”

Mount Sinai’s so-called medical cocktail included ribavirin, an antiviral drug that has been around since the 1970s. Although the way that ribavirin actually operates in the body is not well understood, it has been found effective against a number of DNA and RNA viruses, including hepatitis C and hemmorhagic fevers like Lassa virus and Ebola. Some thought there was a chance it could be effective for SARS patients too.

But ribavirin was unsuccessful. In many cases, it actually made people sicker. About half of the hospital’s patients experienced a significant drop in their magnesium and calcium levels. Even more of them developed hemolytic anemia, a condition which causes red blood cells to break down prematurely. As a result, at least nineteen Toronto patients ultimately needed blood transfusions.

Though Mount Sinai stopped using ribavirin when they recognized the problem, it was not the only example of SARS treatment gone awry. Communication between the different hospitals was haphazard. Doctors in Hong Kong, who had yet to hear about ribavirin’s adverse effects in Canada, continued to use the drug with their SARS patients. Meanwhile, experimentation with corticosteroids in Chinese hospitals had mixed results, causing some patients to experience fungal infections and even bone death. A study later released in the American Journal of Medicine would find that SARS patients who had received either ribovirin or corticosteroid treatments were slightly more likely to die.

Ultimately, the only sure treatment we had for SARS was time. Isolate the patient, give him IV fluids and respiratory support, and wait and see.

In retrospect, Low says it was unethical to include drugs that scientists didn’t fully understand in their SARS treatment. He wishes that, instead of jumping to ribavirin, the hospital had performed a quick, local study with twenty or so patients. Such a trial, would have taken only three or four weeks, and it would have quickly become apparent the ribavirin was not the right way to go.

But in the heat of an epidemic, controlled research is not at the front of most doctors’ minds.  And there’s no organization, not even the WHO, that can require individual hospitals across countries to react in such a coordinated way, though some researchers called for the establishment of a special contingency fund for quick disease research in the wake of the 2003 outbreak.

“You lose sight of the problem. People are desperate. They’re seeing people die, they’re not doing anything for them, and they’ve got to do something,” says Low. “Those are the kind of things we have to get our heads around if we’re faced with the next outbreak. We could be back in that very same situation again, because there’s no central coordinating body that would be responsible for doing that.”

Of course, this wouldn’t be a problem if we had spent our time developing a new coronavirus treatment. But we haven’t.

Part of the problem is simply unavoidable: without more SARS cases to treat, can’t know for sure which experimental treatments will work in humans. Theories can be tested up and down with lab primates or Petri dishes, but clinical trials with real people are the only way to be scientifically certain that a medicine works the way we think it will.

But those kinds of studies aren’t really going on either. Some labs, like the one at Mount Sinai, became fearful of keeping their SARS samples after a few accidents at laboratories in Singapore, Taiwan, and Beijing caused researchers and their family members to become sick. It’s not hard for such an accident to happen. Someone forgets to wear the right safety gear, or mistakenly reaches back for the wrong container. They breathe in, shake their colleague’s hand, take public transportation home, kiss their spouse hello. The infection spreads again.

That’s why, in early 2004, the doctors at Mount Sinai made the decision to nuke their twelve remaining SARS samples in an autoclave machine. And that’s why, in October of this year, the United States government increased the biosafety level of SARS, restricting samples to the most highly regulated laboratories.

Furthermore, coronavirus research is limited by a simple bottleneck in funding. Industry interest is low, because there’s little to no money to be made. And though outside funding initially surged after SARS appeared in 2003, those sources has waned too as time went on and other, more immediate problems caught our attention.

Perhaps, if we’re hit with another epidemic, interest in the research will wane again.

“We really should learn from the history,” says Poon. “One day, we might run out of our luck.”

This piece was written in November 2012 for the MIT Science Writing Program.

Aviva Hope Rutkin is a science and technology reporter in the Boston area. She currently writes for the MIT Technology Review. Follow her @realavivahr.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Basic HTML is allowed. Your email address will not be published.

Subscribe to this comment feed via RSS

%d bloggers like this: