Even a small mouse has anger.
—Chukchi proverb
In May of 1993, nineteen-year-old Merrill Bahe was in a car with family, heading into Gallup, New Mexico, when he suddenly had trouble breathing. His family stopped at a convenience store to start CPR and call for help, and an ambulance came and took him to the Gallup Indian Medical Center. By the time he got there, his lungs were so full of fluid that the doctors there were unable to revive him. And yet, other than a fever and some mild flu-like symptoms, Merrill had been an otherwise healthy young man, a track star on the Navajo reservation.
This kind of unexplained, sudden death of a seemingly healthy young adult is dramatic enough, but the reason Merrill had been going into town was to attend the funeral of his fiancée, twenty-one-year-old Florena Woody. She had died under similar conditions five days earlier at the nearby Crown-point Medical Center.
George Tempest, an officer in the US Public Health Service, as well as chief of medicine at the Indian Medical Center, quickly got on the phone with other doctors who tended to the Navajo Nation. He discovered that within the Four Corners region, where the boundaries of New Mexico, Arizona, Colorado, and Utah intersect, five people had died under similar circumstances during the previous six months.
In keeping with a law that required all unexplained deaths to be reported to a central registry, the local investigator for the New Mexico Office of the Medical Investigator was also alerted. For his own part, this investigator had been perplexed by a similar case of a young thirty-year-old Navajo woman who arrived at the same emergency room complaining of flu-like symptoms and sudden, severe shortness of breath, followed by rapid death soon after arrival. At autopsy, her lungs were almost twice the normal weight for a woman her age. During the postmortem, the medical examiner ladled out liters of fluid.
Public health officials thought this might be the plague, which is endemic to the region, but tests ruled it out. This was something distinctly new and different, yet within a short time a dozen more people had contracted the mysterious illness, most of them young Navajos in New Mexico. News outlets began reporting on the unexplained deaths, often using the pejorative term “Navajo flu.”
That spring, I was completing my second year of the Epidemic Intelligence Service program. As advertised, it had been an amazing experience that allowed me to see a greater variety of outbreaks than most doctors see in a lifetime, ranging from vertigo linked to a gut virus in Thermopolis, Wyoming, to viral meningitis (a brain inflammation) in wealthy parents of young children in a day care in California, to measles on a college campus in New Jersey. My wife had just given birth to triplets, and I had to make critical decisions about my next job. Plan A was to take an infectious disease fellowship, but that prospect dimmed as I thought about supporting a family as a clinical trainee. But more than that, I had fallen in love with public health, the people I worked with, and the work that we did.
There was a moratorium on hiring people at CDC, but when I approached my division director, Dr. Brian Mahy, he said, “Don’t worry about it. We’ll get you a job.”
I said, “Well that’s very kind of you, but there’s this hiring freeze.”
He said, “Don’t worry about it. That’s my problem.”
I never asked another question and took it on blind trust. On July 1, when my stint at the EIS ended, I simply showed up for work with CDC’s Special Pathogens Branch.
One of the reasons there was space for me in Special Pathogens was this strange outbreak that seemed to be expanding across the 27,000-square-mile region that’s home to the Hopi, Ute, Zuni, and Navajo Nation reservations. Adding to the mystery, reports of other suspected cases were coming in from across the United States.
The first clue to solving the puzzle came when Tom Ksiazek, who had been recruited from the US Army Medical Research Institute of Infectious Diseases to the Special Pathogens Branch, ran the patient samples against a battery of infectious agents. He found an unexpected immune response against certain Old World viruses known to be transmitted by rats.
These so-called hantaviruses, named for the Hantaan River in South Korea, were first identified during the Korean War, when American soldiers were getting infected with something called Korean hemorrhagic fever. A collective term for infections from any of a number of related viruses in Eurasia and Africa, this illness was commonly known as hemorrhagic fever with renal syndrome (HFRS), but the illness we were seeing in the Four Corners area did not involve the kidneys.
While one of the infected rodent reservoirs or natural host (and vector for transmission), the common rat, had spread worldwide via shipping lanes, it rarely if ever caused disease. Even so, our lab people began using genetic methods to probe the blood of various rodents throughout the New World.
More definitive lab evidence trickled in from genetic studies of this new virus’s genetic material in the blood of infected persons, and this allowed CDC to develop more sophisticated diagnostic tests. It turned out that what we were seeing was a rare New World virus, closely related to the Old World hantaviruses, that had never been associated with disease.
So forty years after its first appearance in Korea, this appeared to be a pulmonary version, hantavirus pulmonary syndrome (HPS). But what was the specific agent? Where did it come from? And what was the rodent host?
We launched multiple lines of investigation, but this was before CDC had established incident management systems. We had no task force structure and no emergency operation center for coordinating large outbreaks, so it was more chaotic than it should have been. Even so, we kept the focus on the patients.
During the Korean War, when our soldiers starting coming down with Korean hemorrhagic fever, researchers identified the antiviral drug ribavirin as a fully effective treatment.
So our first line of action in the Four Corners region was to put together a ribavirin trial in patients to see whether or not this drug was a therapeutic option. That task was entrusted to a team led by Dr. Louisa Chapman.
Our second objective was to more sharply define the clinical illness. That meant going to all the hospitals where these potential cases came from and observing firsthand the clinical spectrum, which is the only way you can know what you’re seeing when you see it and thus identify new cases. The next step was to talk to survivors, and to the family members of those who didn’t survive, visiting their homes and trying to identify the risks, while also trying to understand the course of the illness.
Our goal in all of this was to keep new cases from happening. Which necessarily involved finding the virus’s natural host or reservoir, and determining the risk factors and behaviors of the virus, the rodents, and the infected people that accounted for infection. It also demanded that we create diagnostic tests and figure out the laboratory tools we were going to need to identify the disease.
The parallel phase of our job for the mammologist was to try to figure out where the reservoir of the disease was located. For hantavirus, this required trapping and studying rodents all across the southwestern United States to try to nail down everything we could about the disease vector. (In the case of hantaviruses the terms “vector” and “reservoir” are often used interchangeably because the rodents serve as both: the natural host and the direct source of human infection. In other diseases such as West Nile, the natural reservoirs may be birds, but mosquitoes act as the vector.)
Once we’d assembled all these pieces of information, the task was to link them together to create a final prevention strategy that would work in the specific communities affected.
That was an incredible amount of work in and of itself, but then we also needed to review patients’ charts and get details on every individual, right down to their temperature, respiratory rate, and heart rate. What was the history of the current illness? What was the previous history? What was their travel history? What were their hobbies? What were potential exposures? What medications were they on? What about their surgical history? And then all of their lab values: sodium, potassium, chloride, liver function tests, platelets, white count, red cells. And all this updated on a daily basis.
Out of this welter of information you try to put together a picture of the disease so that you can be better at identifying additional cases, and then deduce the effectiveness of various treatments to try to arrive at the best way to manage future cases.
In early June, I joined the response team to try to establish a case definition for the initial ribavirin study in order to determine whom we should treat. This is the difficult work of making those clinical decisions, because you don’t have a diagnostic test yet, and often you want to treat sooner than later, so even if you have a diagnostic test, you might not want to wait for the results to come back before you start a regimen for your patients.
Our first condition for possible treatment was that you had to be from the Four Corners region, and that you had to have fever. You could not be immune compromised, and you had to have pulmonary infiltration in both lungs. For comparison, we looked at the cases of hemorrhagic fever with renal syndrome that had occurred in Korea forty years earlier.
Since we had gotten this early signal that it was hantavirus, we thought that even though we had pulmonary involvement, maybe there were renal cases as well that we just hadn’t picked up on.
Traditionally, you name a virus after the place where it is first identified. Needless to say, the local residents weren’t very happy with calling this deadly microbe “Four Corners virus.” Not just because the bad publicity would cause vacationers to drive their RVs elsewhere, but because of the more general stigma. Already, a group of Navajo children visiting the state capital had been discriminated against because they happened to come from the Four Corners area.
So then we suggested “El Muerto virus,” after the Canyon del Muerto that was close to the site of capture of the rodent from which the virus was isolated. Unfortunately, this was near the site of a pivotal battle between the Navajo and Kit Carson’s US Army. When people understood the history of what had actually happened there, we decided that maybe this wasn’t such a good name either. Then there was the internal bickering among the scientists about who was the first to identify the virus, who had the right to name it, and so on. After this experience, there has been a gradual trend toward not naming diseases after places but to be descriptive about the nature of the clinical illness.
In the end, the decision was made to call it Sin Nombre, which is Spanish for “no name.” You’d never have gotten away with calling it No Name in English, but somehow the Spanish gave it a little flourish, and the name stuck. And nobody caught the irony that a virus that was killing Indians was called No Name for a people who have had no voice.
After setting up case definitions in order to begin treating patients, I began working on a clinical database. I also worked with CDC Drug Service to figure out how to enroll patients, and I worked with the Food and Drug Administration to get this done as quickly as possible.
Setting up a clinical database meant sorting cases into three groups. The first were those residing in the Four Corners area; the second, people residing outside the area who had traveled to the Four Corners in the previous six weeks; and the third group, those who had neither resided in nor traveled to this region. We did not give drugs at this time to the third group but focused on soliciting samples and reevaluating lab results in light of any newly available data on rodents and their distribution in that locality.
As the database grew, we began revising the case definition. To accurately identify cases based on a clinical description to quick-start a drug, you need to figure out exactly who you’re looking for, and this often means excluding cases that are not typical. Our case definition excluded people who might have had a hospital-acquired pneumonia, people who had been sick for more than two weeks, and people who were immunocompromised. This would help us zoom in on the people we should be looking for, the ones who might benefit from ribavirin therapy. This became much simpler as diagnostic tests were distributed by CDC to all the state health departments.
Media attention had prompted clinicians elsewhere to rethink signs and symptoms in light of this new disease. Unlike those of us in the center of the storm, they didn’t limit themselves to searching for cases within the Four Corners area.
It’s routine for clinicians to identify cases of established diseases. But for novel diseases, you’re dependent on the astute clinician who says, “You know what? This patient’s symptoms sound like what I just read about in the newspaper.”
One such case occurred in Lufkin, Texas, where a patient had died at the end of June with acute respiratory distress. Records indicated that she had tested positive on our rudimentary test with Seoul virus, one of the Old World hantaviruses. She had an antibody count of 1:1,600, which is high, but a negative immunofluorescent assay, a test that lights up infected cells with your tagged blood if you have antibodies from a current or previous infection.
I flew into Houston at seven in the evening and picked up my rental car, then began the two-hour drive into the Piney Woods area of East Texas.
In Lufkin, a town of about 35,000, built around forestry, oil field equipment, and poultry, I met the state epidemiologist at seven thirty the next morning. I talked to the parents of the victim. I talked to her husband. I talked to the daughter, a nurse at the same hospital where the patient had been treated, and tried to solicit as much of the health history as I could. I needed to know about her habits, hobbies, and especially travel, trying to figure out how she might have gotten infected and by what sort of rodent. She wasn’t a gardener. Apparently she was only a so-so housekeeper, but reportedly there had been no mice in the house for over a year. However, there were seeds and squirrels out back, as well as rats.
The next task was to review reams of records, looking for any other potential cases in the nearby area. I looked through the charts of anyone who might have had a similar disease, hoping to locate blood or autopsy samples that might have been preserved, in order to retroactively investigate these patients.
The case in Lufkin seemed like it made sense as soon as we tentatively identified the rodent reservoir for this disease: Peromyscus maniculatus, the white-footed deer mouse. This little beast was widely distributed throughout the Southwest—in Texas, Nevada, and California. But soon we started seeing cases as far away as Louisiana and Florida. Which led us to wonder if there were other pathological virus and rodent species.
Eventually, from these cases we found outside the distribution of the white-footed deer mouse, we discovered that there were numerous other rodent species in the same family that had their own hantaviruses.
Having followed the media attention, a doctor in Dade County, Florida, suspected hantavirus as the cause of illness in one of his patients. He sent us a sample, and he turned out to be right. But this wasn’t the Sin Nombre virus we’d seen in the Southwest; it was a completely different strain, called Black Creek virus.
Our biggest concern with this Florida case was that the infected individual happened to be staying at a treatment center with thirty other people. So next I flew to Dade County and met with the administrator for District 2, the environmental administrator for the Dade County Public Health Unit, the deputy district administrator for health, and the public information officer. My job in those days involved managing the hantavirus disease-monitoring databases during the weekday, and spending the weekend investigating unusual cases. This work load was typical of the scientists in Special Pathogens—my mentor’s wife actually took to visiting him at CDC on weekends—but the intensity took a toll on my family, given that I was leaving my wife alone with six-month-old triplets.
As usual, a big part of my job in the field was to determine the risk of exposure, so I needed to see who else might have gotten sick, while also making sure that investigators were on the lookout for a novel type of rodent. And we quickly found one.
The culprit for Black Creek virus wasn’t the usual deer mouse, but a cotton rat. So our task was no longer a question of tracking down and isolating one rodent with one virus. This was a scenario involving multiple rodent species, each with its own, somewhat related hantavirus.
As we delved further, what we discovered was that these North American variations on the disease were not even new.
In 1978, a young man from Preston, Idaho, population 3,170, had died so abruptly and inexplicably that the attending physician contacted CDC. That must have been a slow day in Atlanta, because they sent both a CDC officer and an EIS officer, Dr. Rick Goodman, to investigate. After they completed their work, though, they were none the wiser about what had actually happened.
But then fifteen years later—again, amid the media swirl—the doctor who’d initially called CDC about the Preston case said to himself, You know what, I think this guy had hantavirus. So he contacted us again.
Fortunately, the paraffin blocks from the victim’s autopsy had been preserved, and they were transferred to Dr. Sherif Zaki, CDC’s chief pathologist, who looked at them under the microscope, applied a serum containing the right type of antibody/immune marker, examined the results, and concluded, yep, this patient died of hantavirus pulmonary syndrome way back during the Carter administration.
I had an opportunity to talk to the individual’s dad, mother, brother, and sister, as well as his wife. They all came to Preston to chat with me, and their memories of the event—it was the death of a loved one, after all—were very fresh and vivid.
“Oh yeah, middle of summer. Temperature hitting a hundred and four that day, I remember, and he started getting fatigued and tired, his neck was sore. First he was taken to a chiropractor. He was sleeping all day. He wasn’t eating. Cold sweats and fever.”
They remembered the doctor who’d initially diagnosed him with flu. But then he didn’t get better. He developed a cough and went to a different hospital, where they gave him a shot and some suppositories. He had trouble breathing.
Four days into it, his feet and legs were cold, and his cough was getting worse. He had difficulty breathing, so maybe he had pneumonia. By the time he went back to the county hospital, he looked blue-gray and he was gasping for breath. They found out that his lungs were filled with fluid, and shortly thereafter he died.
It turns out that the victim had been a welder in Logan, Utah, who’d lived in a prefab home with crawl space. He’d hunted elk and deer, but he also spent a lot of time going after mice in and around, and underneath, his house. According to his family, he not only shot them but stomped on them, sometimes even picking them up and throwing them against his truck to smash them.
What this case showed us is that, just because a disease hasn’t been identified, doesn’t mean it isn’t there. It’s just that it usually takes a cluster before something shows on the radar, because it takes a reasonable amount of diagnostic evidence to tie cases together and recognize that you have something new.
If we’d never had the hantavirus test from Korea, it would have taken much longer to understand the scope of the outbreak in 1993, which took place not just in the Four Corners region but well beyond it.
The confluence of factors that allowed us to identify it were, first of all, enough typical cases that people were concerned about, and a diagnostic methodology that allowed us to figure out exactly what it was. The moment we identified it, it became very clear that this disease very likely had been in the United States for thousands of years, because these viruses likely had coevolved with a variety of rodent species and had cross-species infections. The likely ancestral rodent host for virus for both the New and Old Worlds was a shrew or a mole.
Zoologists talk about how the search image affects perception. For example, a certain kind of bird will be extraordinarily adept at spotting a certain kind of bug that it feeds on, but we humans won’t normally see the bug. However, if we are conditioned to look for the specific pattern of that bug on a specific kind of leaf, we become much better at being able to see it.
Before the outbreak in the Four Corners, there was no search image for hantavirus pulmonary syndrome in the United States. Physicians saw random cases with seemingly random symptoms that didn’t add up to any larger pattern. This is not surprising since we have only so many ways to be ill and die. These cases had been occurring for decades, but there was no vocabulary and there was no scientific framework. Until somebody could imagine it, it remained invisible. Sometimes you need a theory before you can find the facts, because otherwise the facts are random signals, needles scattered in a haystack.
Before Four Corners, we also never would have considered that rodents carrying hantaviruses might be distributed throughout the New World. And without that awareness, we never would have observed that hantaviruses can, in fact, be spread from person to person.
In 1996, we sent a wonderful epidemiologist, Rachel Wells, down to Brioche, Argentina—a lovely Patagonian ski resort town from what I’ve been told—to do an investigation of what appeared to be a new hantavirus. When she came back she said, “Okay, based on my findings I think this may be transmitted person to person, because a lot of these people are closely linked.”
My message back to her was, “Based on our extensive experience of forty years, that isn’t the case with these viruses. These people may appear to share a link in transmission simply because they all live in proximity to one other, and thus have the same set of rodents in their homes.”
We told her to go back and look at her data, which she did, and due to some of the unique relationships among these individuals, she was still convinced that person-to-person transmission was taking place.
Then, out of the blue, in December 1996, we were told that two cases of hantavirus pulmonary syndrome had been diagnosed in Buenos Aires at the Malbran Institute. That made no sense, because we’d been studying the distribution of the infected rodent species in the district of Bariloche.
The story became even more interesting when we were told of an infected physician, Monica, who’d traveled to Buenos Aires from Bariloche after her husband became ill. She was accompanied by her good friend, Marina, also a physician. Marina in turn became ill and died at the end of the month, while a second physician who tended to Monica, who’d never set foot outside Buenos Aires, also became ill. Both of these cases occurred twenty-seven and twenty-eight days after first contact.
Even with this information staring me in the face, including some history of arterial blood exposure of the first attending physician friend, I still wrote a note recorded in my green logbook: “Could this not have been a needle stick involving the patient?”
Working with our colleagues in Argentina, we did a bunch of studies to see if anybody else was infected. We drew blood at all the clinics where these individuals had been, and at various other hospitals, identified clinicians who might well have been infected, and asked them to suggest patients who might be at risk. At that point, we knew to look for relatives who had nursed these patients in the hospital, as well as sexual partners—which was one of the things that Rachel had figured out: a number of these people had been intimate with those who had become infected in Bariloche. Subsequently we determined that this type of hantavirus, called Andes hantavirus, could indeed be transmitted among people by large particle droplets in the air.
This is a great example of a new researcher teaching an experienced virologist and epidemiologist (namely me) new tricks. Even though none of the other strains of the virus seemed to be associated with person-to-person transmission, this one was. The case reminded us how you have to be careful about your assumptions whenever you’re dealing with emerging infectious diseases, because there is always room for surprises.
The Sin Nombre virus, the infection from the Four Corners, has to date never shown a human-to-human transmission. Which is a gigantic difference. Same virus, same rodents, or very closely related, but a gigantic difference. This is why, when discussing a disease that people are worried about and asking, “Well, could this go airborne?” we try very carefully to explain that this is not what usually happens. Rabies, which has persisted for hundreds of thousands of years, has never become airborne. Diseases usually have a typical way of transmitting themselves, and they stick to it.
However, given the fact that there is continuous gene shuffling every time a virus goes through its cell cycle and replicates, you can never rule out something novel—and perhaps deadly.
The classical investigative model is to examine personal factors, pathogen factors, and environmental factors, then determine how they come together to define what we should say to the public. After all, the individual doesn’t know until after the fact that there’s an outbreak of something deadly, whether it involves rodents or mosquitoes or birds. So it’s essential to get the information out there as soon as possible, but also to be clear on the interplay of factors that affect the disease’s transmission.
At first you have to ask, Is this the right carrier? It’s not fair to pick on chipmunks if the culprits are gray squirrels. Once you establish the vector, it’s a question of whether that rodent might spray you, or if you could contract the disease by sweeping up some dust contaminated with rodent urine, and thus inhaling the pathogen. (We always recommend to anyone who needs to sweep up rodent droppings that they first spray it down with Lysol.)
Then there’s also the host factor. How much of the infectious agent is entering your body? What is your innate susceptibility to this pathogen?
All these variables come into play in determining whether or not an individual becomes sick. Sometimes it’s simply the luck of the draw. Some smokers die of old age. Other people who never smoked a day in their life die of lung cancer. Sometimes the media is ahead of the medical community and drives the public health response. That’s what was going on when a reporter from Quepasa, a science and tech magazine in Chile, contacted me, asking, “Help me understand how person-to-person transmission occurs with hantaviruses.”
He was concerned about the city of Coyhaique, in Aisén province in the south of Chile. A case of what appeared to be hantavirus had occurred in a desert area with no woods and no rodents, yet the patient died. “This city has lots of rodents,” he said, “but health care workers aren’t taking any precautions. What are you at the CDC doing?”
Chile is obviously a sovereign nation, and CDC doesn’t have the responsibility for protecting the whole world; that’s the mandate of the World Health Organization (WHO). But CDC is often called on to help with outbreaks outside the United States.
A couple of days later, I spoke with Dr. Jeanette Vega from the Pan American Health Organization in Chile. She served in the Chilean Ministry of Health, and later as well at WHO as managing health director of the Rockefeller Foundation.
She led us to the case of a male individual who’d had onset of disease on September 5, 1997, and died six days later. What had attracted media attention was that his sister, mother-in-law, and brother-in-law had also died of hantavirus pulmonary syndrome. It was this cluster that got everybody involved.
Dr. Roberto Belmar had already been set up as the chief of the Hantavirus Commission in Chile, trying to determine how to respond to this outbreak. We had a series of conversations back and forth with our colleague, Dr. Elsa Salguero, and others in Argentina, trying to find useful comparisons. A week later, we were again on the phone with Dr. Vega in a conference call with many other practitioners.
They had determined that the first cluster had occurred in an area called Lagos Verde, with five deaths in three months. But then they had a second cluster in Coyhaique, with four deaths within four days!
When you look at the first and second clusters so concentrated in time, you have to assume that the most likely reason all these people got infected was exposure to some common source. When you look at the first cluster, though, where onset was spread out over many months, there’s far greater likelihood that this was person-to-person transmission from one case to the next. It was this possibility that really had us intrigued.
Once again, the moment you identify the cluster—the search image—all of a sudden cases start coming out of the woodwork. We began to hear of suspected cases from all over the country, and we began to worry.
As of that time, 1997, we had been in the hantavirus business for four years, and we knew that some previous surveys had been done in this community, with rodent studies that followed up on the possibility of person-to-person transmission.
When we got there, we helped establish the National Surveillance System, then tried to compile everything that could be known about the clinical illness, focused on the Aisén province where this was going on. We wanted to describe the epidemiologic condition, who lived, who died, and then maybe do a couple of studies—some cohort studies among the family members, among the health care workers to see who was and who was not infected, and maybe some community survey work in the rural area to see whether maybe this disease was just a lot more prevalent there than elsewhere.
We met at Christmastime with the acting minister of health, the head of the health system from that province, the head of environmental health in Chile, and the Epidemiologic Surveillance Unit, as well as WHO representative Dr. Piña, who was there in Chile.
We learned that there were 14,622,354 inhabitants, 84 percent rural, living in thirteen regions, with twenty-nine health service units. At this point, they thought there were twenty-six cases, sixteen of them male, ranging in age from one year eleven months to forty-six years. Twelve of these cases happened to be in the province of Aisén. People were dying within one to eight days from the onset of symptoms.
We headed down to Coyhaique and started going to the villages, sometimes on horseback, trapping rodents, trying to get a better sense of what was going on.
We were there for two or three weeks, keeping track of which rodents were going into what box, checking the databases to make sure we’d aligned the right rodent type to the right place. We learned very quickly that there were a lot of rodents: over 40 percent of the traps were coming back positive. We were also investigating tons and tons of suspect cases to try to help the local health authorities put together a system for determining who needed to be tested and who didn’t.
Aisén was 515,000 square kilometers and contained five provinces, with a population of 84,000, or 1.7 inhabitants per kilometer square, which was pretty rural. Forty-two thousand of those people resided in Coyhaique.
Lagos Verde, where our target family lived, had fifty homes and approximately one hundred inhabitants. We wanted to determine what percentage of the population was infected with hantavirus, and of that, how many people actually got the disease; and of that number, how many got severe disease and died.
Along the way, we were able to connect the everyday patterns of life in that region of the world to the spread of the disease. For example, I saw data that indicated one of the clusters, where those infected were all next to each other, was a family that had opened up a house before the rich owners came to live there, and that’s probably how all those people got infected at one time.
In four of the twelve cases in that area we got a history that looked more like hemorrhagic disease, which affects multiple organs, damages the blood vessels, and affects the body’s ability to self-regulate. The patients also seemed to have more kidney disease. The doctors already were trying to be very proactive in what they were thinking about treatment. If patients had low blood pressure, doctors gave them plasma from someone who had survived the disease as a way to help them respond and get better. In all ten to eleven of the cases where they could, they actually gave patients plasma therapy, an experimental technique that would reappear later when we were confronting Ebola.
In late spring and early summer we got calls about another cluster of cases in Brazil near the capital, São Paulo, suggesting person-to-person transmission, and I flew down on June 10, 1998.
The government wanted assistance not just from CDC, but also from the Pan American Health Organization, to try to understand what was going on and to try to improve their public health measures.
There were twenty-four health regions in São Paulo, overseen by the Adolfo Lutz Institute of the Brazilian Ministry of Health. They already had a surveillance system for dengue, a viral disease spread by mosquitoes within the genus Aedes, principally A. aegypti, as well as for leptospirosis, caused by a bacteria carried by a variety of animals, especially rodents. Now they were developing a separate surveillance system for hantavirus pulmonary syndrome.
They had a handful of cases in 1996, and then they had this cluster two years later, in roughly the same area, which is why we were called in.
Clinicians may see tens of thousands of patients during their career, but there will be one or two cases strange enough that they call the health department for assistance. This is especially important in developing or low-to middle-income countries where there’s not necessarily great diagnostic support, and autopsies are rare. You really are depending completely on clinical acumen, and intuition, to discover the anomalies. Once there’s a big outbreak everybody realizes it, but the initial ones—you need to find them and put prevention measures in place if you want to decrease the size of the outbreak before it gets out of hand.
Very often, it’s when a prominent member of the community gets sick that doctors take note of special cases.
In the São Paulo region, it was a fifty-one-year-old woman named Palusta who triggered concern. She got sick and was dead six days later, which was quick for an otherwise healthy, relatively young woman. Her doctors initially suspected dengue, but when her blood sample was sent off for diagnostic testing it came back negative. But the blood was positive for hantavirus.
Palusta owned a fifteen-hectare farm with grazing land for cattle. She went there about twice a week to oversee operations, and three times a week she went to visit the home of a woman who’d once worked for her, but had died in the local hospital while being treated for pulmonary heart disease.
Another subsequent case was one of Palusta’s employees on the large plantation she owned that produced sugarcane for alcohol. He had stomach pain and so much chest pain that his physicians thought he’d had a heart attack. His electrocardiogram was normal, but his blood pressure was so low that it was difficult to put in an IV. He was throwing up coffee-ground vomitus, which often suggests bleeding inside the stomach. He had very little shortness of breath, but he also died in shock. So in this case the symptoms pointed toward dengue.
Palusta’s clinical history was that she had a fever on a Saturday, with headache, muscle aches, and pain behind her eyes, but no real shortness of breath. Four days later, when she returned to her doctor, he said, “You’ve got some protein and some white cells in your urine,” and diagnosed her with dengue. She also had bruising all over her body, which contributed to the diagnosis of dengue.
The next day, Thursday, she was feeling better, but later that afternoon she deteriorated. On Friday she was back at the doctor’s office in shock, with significant shortness of breath and lots of fluid in her lungs. She died within three hours.
In trying to figure out how one individual got sick, you couple the individual clinical investigation with the epidemiologic investigation. Then you align that with the entomology, or in this case the mammology, and go out and see which animals are there and which are potentially infected. Then you put these three pieces together to think about what is the right set of prevention strategies to try to keep other people from getting sick. This approach is a “One Health” approach to emerging infections. Not just to focus on humans but also the implicated animals and the environment to identify potentially novel prevention strategies. This is the pattern we follow when we prevent rabies in humans by vaccinating dogs.
Given this cluster of cases, CDC was invited to come and investigate. So we went out to see this farm of Palusta’s, which was a brick building surrounded on three sides by fields. There was a concrete floor, as well as an indoor bathroom, but some of the roof tiles were missing. There were also a large shed for grain, a wood pile twenty feet from the door, and the sugarcane blocks called rapadura that were made on the property. I also noted a vegetable garden, chickens, turkeys, guinea hens, pigs, cattle, bananas, oranges, beans, corn, and coffee being produced. Eight people lived there, and all tested negative for hantavirus. Palusta also had another five-hectare farm where there was a grain silo. All seven household members found there also tested negative.
When Palusta first got sick and was diagnosed, the Ministry of Health sent a trapper who found two types of rodents—Rattus rattus, the common rat, and Mus musculus, the common house mouse—which usually aren’t infected with hantavirus. It seemed highly unlikely that these were the culprits.
So we put down traps and captured twenty-eight rodents on our first night, and had forty captures the second. They turned out to be Calomys, also known as vesper mice.
In part, what we were seeing was the tremendous diversity of the family of rodents called sigmondontinae, which invaded the New World about 5 million years ago, then radiated out into nearly four hundred different species. What probably happened is that an ancestral virus infested a sigmondontinae rodent, and as the rodents evolved over time, the viruses co-evolved and diversified with cross-species infections as well.
After we assessed the farm, we went to Palusta’s primary residence, which was a beautiful house in the center of town, surrounded by paved streets in a gated community with an electronic front gate, five miles from her farm. The kitchen had granite butcher-block countertops with Italian tiles up to the ceiling, and there were polished terrazzo floors throughout.
Only three months after her death, we began trapping in this area: 130 of our 160 traps had captures in them, and all were along the edge of a plowed cornfield where the brush had been allowed to grow. I had put the traps out myself the night before, one by one, and went with my team personally to pick them up. Early the next morning—before the rodents had time to bake in the sun in the metal traps, which is never good—I took them to the processing site, occasionally putting on protective gear and doing the dissections myself. The team taught me how to do orbital bleeds: you take a little capillary tube, put it right through the top of the mouse’s eye, and break all the little blood vessels so that the blood drips out. We then euthanized them, harvested their organs, and placed the organs in liquid nitrogen.
A lot of these diseases had similar signs and symptoms, and we followed up with dozens of potential cases, trying to pin down the diagnosis so we could have better guidance in what to do for these patients.
We spent most of our time looking out at these large sugarcane farms and thinking about what might have changed the ecological balance to give rise to a new species of rodents, which just happened to be carrying a deadly disease.
Eventually we focused on those rough borders left around the corn and the sugarcane, which turned out to be a rodent paradise. The government had been trying to reduce the burning of cane fields because of environmental concerns—smoke equals air pollution. In 1998, three-quarters of the fields were not burned.
Which is where the law of unintended consequences came in. The government had tried to do something good, and it wound up causing a spike in a deadly disease.
This is why we always need to think long and hard about the impact of any man-made changes. Sometimes the danger is not the primordial mysteries of nature. Sometimes the danger is us.