Spillover

by David Quammen

  By then they knew that the pathogen was a virus. That knowledge derived from isolations performed quickly on clinical samples shipped to overseas laboratories, including the CDC. (Johnson, before flying to Zaire, had led the CDC isolation effort himself.) They knew that this virus was similar to Marburg virus, another lethal agent, identified nine years before; the electron micrographs showed that it was equally filamentous and twisty, like an anguished tapeworm. But the lab tests also revealed Ebola virus as distinct enough from Marburg virus to constitute something new. Eventually these two wormy viruses, Ebola and Marburg, would be classified within a new family, Filoviridae: the filoviruses.

  Johnson’s group knew also that the new agent, Ebola virus, must reside in some living animal—something other than humans—where it could exist less disruptively and maintain a continuous presence. But the question of its reservoir was less urgent than other concerns, such as how to break the chain of person-to-person transmission, how to keep patients alive, how to end the outbreak. “Only limited ecological investigations were made,” the team reported later, and the results of those investigations were all negative. No sign of Ebola virus appeared anywhere except in humans. But the negative data are interesting in retrospect, at least as a record of where these early researchers looked. They pureed 818 bedbugs collected from Ebola-affected villages, finding no evidence of the virus in any. They considered mosquitoes. Nothing. They drew blood from ten pigs and one cow—all of which proved Ebola-free. They caught 123 rodents, including 69 mice, 30 rats, and 8 squirrels, not one of which was a viral carrier. They read the entrails of six monkeys, two duikers, and seven bats. These animals also were clean.

  The International Commission members were chastened by what they had seen. “No more dramatic or potentially explosive epidemic of a new acute viral disease has occurred in the world in the past 30 years,” their report warned. The case fatality rate of 88 percent, they noted, was higher than any on record, apart from the rate for rabies (almost 100 percent among patients not treated before they show symptoms). The Commission made six urgent recommendations to Zairian officialdom, among which were health measures at the local level and nationwide surveillance. But the identification of Ebola’s reservoir wasn’t mentioned. That was a scientific matter, slightly more abstract than the action items offered to President Mobutu’s government. It would have to wait.

  The wait has continued.

  Three years after Yambuku, Karl Johnson and several other members of the Commission were still wondering about the reservoir question. They decided to try again. Lacking funds to mount an expedition devoted solely to finding Ebola’s hideout, they hitched their effort to an ongoing research program on monkeypox in Zaire, coordinated by the World Health Organization. Monkeypox is a severe affliction, though not so dramatic as Ebola virus disease, and also caused by a virus that lurks in a reservoir host or hosts, at that time still unidentified. So it seemed natural and economical to do a combined search, using two sets of analytical tools to screen a single harvest of specimens. Again the field team collected animals from villages and surrounding forest in Bumba Zone, as well as in other areas of northern Zaire and southeastern Cameroon. This time their trapping and hunting efforts, plus the bounties they paid for creatures delivered alive by villagers, yielded more than fifteen hundred animals representing 117 species. There were monkeys, rats, mice, bats, mongooses, squirrels, pangolins, shrews, porcupines, duikers, birds, tortoises, and snakes. Blood was taken from each, and then snips of liver, kidney, and spleen. All these samples, deep-frozen in individual vials, were shipped back to the CDC for analysis. Could any live virus be grown from the sampled tissues? Could any Ebola antibodies be detected in blood serum? The bottom line, reported with candor by Johnson and coauthors in the pages of The Journal of Infectious Diseases, was negatory: “No evidence of Ebola virus infection was found.”

  One factor making the hunt for Ebola’s reservoir especially difficult, especially hard to focus, is the transitory nature of the disease within human populations. It disappears entirely for years at a time. This is a mercy for public health but a constraint for science. Viral ecologists can look for Ebola anywhere, in any creature of any species, in any African forest, but those are big haystacks and the viral needle is small. The most promising search targets, in space and in time, are wherever and whenever people are dying of Ebola virus disease. And for a long interlude, no one was dying of that illness—no one whose death came to the attention of medical authorities, anyway.

  After the Yambuku outbreak of 1976, and then two small episodes in Zaire and Sudan between 1977 and 1979, ebolaviruses barely showed themselves anywhere in Africa for fifteen years. There may have been some scattered cases during the early 1980s, retrospectively suspected, but there was no confirmed outbreak that evoked emergency response; and in each of those minor instances the chain of infection seemed to have burned itself out. Burning out is a concept with special relevance to such highly lethal and moderately contagious pathogens. It means that a few people died, a few more got infected, a fraction of those also died but others recovered, and the pathogen didn’t continue to propagate. The incident expired on its own before shock troops from WHO, the CDC, and other centers of expertise had to be mustered. Then, after an interval, it returned—with the outbreaks at Mayibout 2 and elsewhere in Gabon, and even more alarmingly at a place called Kikwit.

  Kikwit, in Zaire, lay about three hundred miles east of Kinshasa. It differed from Yambuku, and Mayibout 2, and the timber camp outside Booué in one crucial way: It was a city of two hundred thousand people. It contained several hospitals. It was connected to the wider world in a way that those other outbreak sites weren’t. But like them it was surrounded by forest.

  The first identified case in the Kikwit outbreak was a forty-two-year-old man who worked in or near that forest and probably, to some small extent, disturbed it. He farmed several patches of cleared land, planting corn and cassava, and made charcoal from timber, all at a spot five miles southeast of the city. How did he get his wood supply, how did he clear daylight for his gardens? Presumably by cutting trees. This man fell sick on January 6, 1995, and died of a hemorrhagic fever a week later.

  By that time he had directly infected at least three members of his family, all fatally, and launched the infection into his wider circle of social contacts, ten more of whom died within coming weeks. Some of those contacts evidently carried the virus into the city’s maternity hospital, where it infected a laboratory technician, and from there into Kikwit General Hospital. The technician, while being treated at Kikwit General, infected several doctors and nurses who did surgery on him (suspecting a gut perforation related to typhoid, they cut open his abdomen), as well as two Italian nuns who helped with his care. The technician died, the nuns died, and local officials hypothesized that this was epidemic dysentery, a misdiagnosis that allowed the virus to spread further among patients and staff at other hospitals in the Kikwit region.

  Not everyone accepted the dysentery hypothesis. One doctor at the Ministry of Health thought it looked instead like a viral hemorrhagic fever, which suggested Ebola. That good guess was confirmed quickly from blood specimens received by the CDC, in Atlanta, on May 9. Yes indeed: It was Ebola virus. By the end of the outbreak, in August, 245 people had died, including 60 hospital staff members. Performing abdominal surgery on Ebola patients, when you thought they were suffering from something else (such as gastrointestinal bleeding from ulcers), was risky work.

  Meanwhile, another international team came out to search for the reservoir, converging on Kikwit in early June. This group consisted of people from the CDC, from a Zairian university, from the US Army Medical Research Institute of Infectious Diseases (USAMRIID, formerly a bioweapons lab but now committed to disease research and biodefense) in Maryland, and one fellow from the Danish Pest Infestation Laboratory, who presumably knew a lot about rodents. They began work at the site to which the spillover seemed traceable—that is, at the charcoal pit an
d crop fields of the unlucky forty-two-year-old man, the first victim, southeast of the city. From that site and others, over the following three months, they trapped and netted thousands of animals. Mostly those were small mammals and birds, plus a few reptiles and amphibians. All the traps were set within forest or savanna areas outside the city limits. Within Kikwit itself, the team netted bats at a Sacred Heart mission. They killed each captured animal, drew blood, and dissected out the spleen (in some cases other organs too, such as a liver or a kidney), which went into frozen storage. They also took blood from some dogs, cows, and pet monkeys. The total yield included 3,066 blood samples and 2,730 spleens, all shipped back to the CDC for analysis. The blood samples, after having been irradiated to kill any virus, were tested for Ebola virus antibodies, using the best available molecular method of the time. The spleens were transferred to a biosafety level 4 (BSL-4) laboratory, a new sort of facility since Karl Johnson’s early work (and of which he was one of the pioneering designers), with multiple seals, negative air pressure, elaborate filters, and lab personnel working in spacesuits—a containment zone in which Ebola virus could be handled without risk (theoretically) of accidental release. No one knew whether any of these Zairian spleens contained the virus but each had to be treated as though it did. From the spleen material, minced finely and added to cell cultures, the lab people tried to grow the virus.

  None grew. The cell cultures remained blithely unspotted by viral blooms. And the antibody tests yielded no positive hits either. Once again, Ebola virus had spilled over, caused havoc, and then disappeared without showing itself anywhere but in the sick and dying human victims. It was Zorro, it was the Swamp Fox, it was Jack the Ripper—dangerous, invisible, gone.

  This three-month, big-team effort at Kikwit shouldn’t be considered a total failure; even negative results from a well-designed study tend to reduce the universe of possibilities. But it was another hard try ending in frustration. Maybe the Kikwit team had gotten there too late, five months after the charcoal maker fell ill. Maybe the shift from wet season to dry season had caused the reservoir, whatever it is, to migrate or hide or decrease in abundance. Maybe the virus itself had declined to a minimal population, a tenuous remnant, undetectable even within its reservoir during the off season. The Kikwit team couldn’t say. The most notable aspect of their eventual report, apart from its long list of animals that didn’t contain Ebola virus, was its clear statement of three key assumptions that had guided their search.

  First, they suspected (based on earlier studies) that the reservoir is a mammal. Second, they noted that Ebola virus disease outbreaks in Africa had always been linked to forests. (Even the urban epidemic at Kikwit had begun with that charcoal-maker out amid the woods.) It seemed safe to assume, therefore, that the reservoir is a forest creature. Third, they noted also that Ebola outbreaks had been sporadic in time—with years sometimes passing between one episode and the next. Those gaps implied that infection of humans from the reservoir is a rare occurrence. Rarity of spillover in turn suggested two possibilities: that either the reservoir itself is a rare animal or that it’s an animal only rarely in contact with people.

  Beyond that, the Kikwit team couldn’t say. They published their paper in 1999 (among a whole series of reports on Ebola, in a special supplement of the Journal of Infectious Diseases), authoritatively documenting a negative conclusion. After twenty-three years, the reservoir still hadn’t been found.

  12

  “We need to know where it is,” Trish Reed had said. She was alluding to the two unanswered questions about Ebola virus and its location in space. The first question is ecological: In what living creature does it hide? That’s the matter of reservoir. The second question is geographical: What’s its distribution across the African landscape? The second may be impossible to answer until the reservoir is identified and its distribution traced. In the meantime, the only data reflecting Ebola virus’s whereabouts are the plotted points of human outbreaks on a map.

  Let’s glance across that map. In 1976 Ebola virus made its debut, as I’ve mentioned, with the dramatic events in Yambuku and the slightly smaller crisis in southwestern Sudan, which was nonetheless large enough to account for 151 deaths. The Sudanese outbreak centered at a town near the Zairian border, five hundred miles northeast of Yambuku. It began among employees of a cotton factory, in the rafters of which roosted bats and on the floor of which skittered rats. The lethality was lower than in Zaire, “only” 53 percent, and laboratory analysis revealed that the Sudanese virus was genetically distinct enough from the virus in Zaire to be classified in a separate species. That species later became known, in careful taxonomic parlance, as Sudan ebolavirus. The official common name is simply Sudan virus, which lacks the frisson of the word “Ebola” but nonetheless denotes a dangerous, blazing killer. The version Karl Johnson found at Yambuku, originally and still called Ebola virus, belongs to the species Zaire ebolavirus. This may seem confusing, but the accurate, up-to-date labels are important for keeping things straight. Eventually there would be five recognized species.

  In 1977 a young girl died of hemorrhagic fever at a mission hospital in a village called Tandala, in northwestern Zaire. A blood sample taken after her death and sent unrefrigerated to the CDC yielded Ebola virus, not in cell cultures but only after inoculating live guinea pigs and then finding the virus replicating in their organs. (These were early days still in the modern field campaign against emerging viruses, and methodology was being extemporized to compensate for difficulties, such as keeping live virus frozen under rough field conditions in the tropics.) Karl Johnson again was part of the laboratory team; this seemed a logical extension of his work on the first outbreak, just a year earlier and two hundred miles east. But the nine-year-old girl, dead in Tandala, was an isolated case. Her family and friends remained uninfected. There was not even a hypothesis as to how she got sick. The later published report, with Johnson again as coauthor, only noted suggestively, in describing the girl’s native area: “Contact with nature is intimate, with villages located in clearings of the dense rain forest or along the rivers of the savannah.” Had she touched a dead chimpanzee, breathed rodent urine in a dusty shed, or pressed her lips to the wrong forest flower?

  Two years later Sudan virus also resurfaced, infecting a worker at the same cotton factory where it had originally emerged. The worker was hospitalized, upon which he infected another patient, and by the time the virus finished ricocheting through that hospital, twenty-two people were dead. The case fatality rate was again high (65 percent), though lower than for Ebola virus. Sudan virus seemed to be not quite so lethal.

  Then another decade passed before filoviruses made their next appearance, in another shape, in an unexpected place: Reston, Virginia.

  You know about this if you’ve read The Hot Zone, Richard Preston’s account of a 1989 outbreak of an Ebola-like virus among captive Asian monkeys at a lab-animal quarantine facility in suburban Reston, just across the Potomac from Washington, DC. Filovirus experts express mixed opinions about Preston’s book, but there’s no question that it did more than any journal article or newspaper story to make ebolaviruses infamous and terrifying to the general public. It also led to “a shower of funding,” one expert told me, for virologists “who before didn’t see a dime for their work on these exotic agents!” If this virus could massacre primates in their cages within a nondescript building in a Virginia office park, couldn’t it go anywhere and kill anyone?

  The facility in question was known as the Reston Primate Quarantine Unit and owned by a company called Hazelton Research Products, which was a division of Corning. The unfortunate monkeys were long-tailed macaques (Macaca fascicularis), an animal much used in medical research. They had arrived in an air shipment from the Philippines. Evidently they brought their filovirus with them, a lethal stowaway, like smallpox virus making its way through the crew of a sailing ship. Two macaques were dead on arrival, which wasn’t unusual after such a stressful journey; but over th
e following weeks, within the building, many more died, which was unusual. Eventually the situation triggered alarm and the infective agent was recognized as an ebolavirus—some sort of ebolavirus, as yet unspecified. A team from USAMRIID came in, like a SWAT team in hazmat suits, to kill all the remaining macaques. Then they sealed the Reston Primate Quarantine Unit and sterilized it with formaldehyde gas. You can read Preston for the chilling details. There was great anxiety among the experts because this ebolavirus seemed to be traveling from monkey to monkey in airborne droplets; a leak from the building might therefore send it wafting out into Washington-area traffic. Was it lethal to humans as well as to macaques? Several staff members of the Quarantine Unit eventually tested positive for antibodies but—sigh of relief—those people showed no symptoms. Laboratory work revealed that the virus was similar to Ebola virus yet, like Sudan virus, different enough to be classified in a new species. It came to be known as Reston virus.

  Notwithstanding that name, Reston virus seems to be native to the Philippines, not to suburban Virginia. Subsequent investigation of monkey-export houses near Manila, on the island of Luzon, found a sizable die-off of animals, most of which were infected with Reston virus, plus twelve people with antibodies to the virus. But none of the dozen Filipinos got sick. So the good news about Reston virus, derived both from the 1989 US scare and from retrospective research on Luzon, is that it doesn’t seem to cause illness in humans, only in monkeys. The bad news is that no one understands why.

  Apart from Reston virus, ebolaviruses in the wild remain an African phenomenon. But the next emergence, in November 1992, added yet another point to the African map. Chimpanzees began dying at a forest refuge in Côte d’Ivoire, West Africa. The refuge, Taï National Park, lying near Côte d’Ivoire’s border with Liberia, encompassed one of the last remaining areas of primary rainforest in that part of Africa. It harbored a rich diversity of animals, including several thousand chimpanzees.