Spillover

by David Quammen

  Nannies had given birth to their kids during the usual kidding season, which can stretch from January to as late as April. Mostly those births had gone well, though on certain farms of the province, including at least one in the Herpen area, many females aborted during the last month of pregnancy. Even full-term kids seemed a bit weak and puny, with a higher mortality rate than usual. Evidently something was troubling the goats, an infection of some sort, possibly new, and veterinarians took note, trying to forestall the abortions with antibiotics. That didn’t help. The general public noticed this situation little or not at all.

  Then came a balmy spring—far warmer and drier than normal. In April, by the recollection of one resident, “there was no drop of rain.” Even before summer arrived, lands surrounding the village had gotten dusty. Breezes blew. In early May, people began to get sick.

  A local physician named Rob Besselink, with an office in Herpen, saw an odd, flulike ailment in several of his patients: high fever, severe headache, muscle aches, shortness of breath, coughing. Was it a bacterial pneumonia? “We started treating them,” Besselink said later, “and it turned out that they didn’t react as we expected them to react on the antibiotics they were given.” He discussed it with a colleague. “After that first week we said to each other, ‘There’s something strange going on,’ because I had three or four people having the same symptoms, and he had also two or three.” Within a couple weeks, the two doctors had seen about twenty patients fitting the profile, of whom almost a dozen, unresponsive to antibiotics, had to be hospitalized.

  Around the same time, in another part of Noord-Brabant, a medical microbiologist named Ineke Weers, employed at a regional laboratory, heard murmurings about a similar cluster. Despite Weers’s broad training and experience—she was an MD with a PhD in microbiology and twenty-one years of work in the diagnostics of infectious diseases—this turned out to be something new to her. An internist at one of the hospitals mentioned that doctors there had lately seen quite a few patients with an atypical, antibiotic-resistant pneumonia. Did Weers know what it might be? Had she read anything about such a syndrome? No, nothing, she answered. But she offered to call the Municipal Health Service in Den Bosch, a large city nearby, and ask whether those authorities could offer some glimmer of insight or advice. They could not; they had heard no other such reports.

  Four days later, Rob Besselink called the same office of the MHS about his situation in Herpen. Two weeks after that, another general practitioner in Noord-Brabant made a similar report to the MHS. This aggregation of puzzling cases was enough to trigger the beginnings of a response. The physicians took blood samples, some of which went to a nearby laboratory, some to a more specialized lab, where the sera were tested for antibodies. After a bit of confusion about what sort of microbe might be causing such “atypical pneumonia,” both labs eventually converged on an answer: It was Coxiella burnetii, the agent of Q fever.

  Q fever wasn’t unknown in the Netherlands but for fifty years it had been blessedly rare. Although the bacterium seemed to be endemic among livestock populations, based on occasional surveys, it had seldom caused noticeable disease in cows, in sheep, or in humans. Now the outbreak in Noord-Brabant caught the attention of the National Institute for Public Health and the Environment (commonly known by its Dutch initials, RIVM), up near Utrecht. Scientists there made an informed guess that maybe the high incidence of abortions on dairy goat farms, which had begun back in 2005, and which had been diagnostically linked to Q fever, might be a source of the human cases. Coxiella burnetii was known to be capable of airborne transmission. At this point, RIVM sent people south to the village of Herpen and surrounding areas to conduct a study. Someone had to learn what was happening downwind of the goats.

  46

  I drove down from Utrecht to Herpen myself, three years later, on a dreary day in February when the gray of the sky and the fog seemed to blend almost seamlessly, along a flat line of horizon, with the gray of the snow. Dr. Rob Besselink received me, just after working hours, in his little medical office on the village’s main street. He was a thin man, in his late forties, with a wide smile that pinched creases into his narrow face. Wearing a black sport coat, a blue paisley shirt, and faded jeans, he looked more like lead guitarist in a rock band than what you’d expect of a rural Dutch physician. Among the first things he mentioned, when I asked about the character of Herpen as a community, was the big change that had come in local farming practices within the past decade: the increase in goats.

  This change had actually started back in 1984, when the European Community established quotas on cow milk that pushed Dutch farmers away from dairy cattle. Many continued as dairymen but started milking goats. The dairy-goat trend grew stronger after 1997 and 1998, when outbreaks of classical swine fever (caused by a virus, but not zoonotic) led to mass cullings of pigs, and many pig farmers, hard hit financially and scared about a recurrence, sought an alternative line of husbandry. “So they started keeping goats, in quite some amounts,” Besselink told me. It was true in Noord-Brabant and true across the country. From a low of about 7,000 animals in 1983, the total Dutch goat population had increased to 374,000 by 2009, of which 230,000 were dairy goats. Most of those lived indoors—stabled year-round inside buildings such as the large, redbrick sheds I had seen on the outskirts of Herpen. You might think that keeping the goats within four walls and a roof should minimize chances of their releasing an infection. But circumstances in the nature of Dutch goat husbandry, as I learned from Besselink and others, conspired to bring C. burnetii out of those sheds in great quantity and launch it on the wind.

  Coxiella burnetii is an assertive bug. It not only causes abortion in goats but also concentrates massively in the placental material expelled during those abortive deliveries. A single gram of placenta from an aborting goat can contain as many as 1 billion bacterial particles. It is also excreted in milk, urine, feces, and during normal deliveries of kids carried to term. Assuming those deliveries and abortions occur within the kidding shed, how does the stuff escape? Very simply, Besselink explained: Goat feces and dirty bedding straw are shoveled up and carried outside by the farmers to fertilize their fields. From there the bacteria can waft into a nearby village as easily as the pleasant, autumnal smell of smoke from a pile of leaves.

  Two goat farms in the Herpen vicinity attracted attention. One was a sizable commercial operation with almost four thousand goats, which had suffered a storm of abortions in April. The other was a “hobby farm” with less than ten animals. When the study team came down from RIVM to look for the source of the outbreak, they visited both places, taking samples of urine, milk, manure, and straw from the stable floors; insects from a light trap; and water from drinking buckets. The hobby farm seemed to be clean. From the commercial farm, every category of sample included evidence of Coxiella burnetii except the milk, the urine, and the water. “There were a lot of Coxiella bacteria in the farm,” Besselink recalled. It was only a kilometer south of the village—virtually right next door. That farmer and his family endured some obloquy during the following year. “He has a wife, he has kids, the kids go to school here, so they were having a hard time because they were having the blame, of course, of what was going on,” Besselink said. The goat farmer hadn’t done anything illegal, merely been unlucky and maybe a little careless, but he suffered lost revenue, sapped energy, sleepless nights. A village doctor comes to know these things. The farmer’s children were stigmatized and his kids—that is, his nanny goats’ kids—were suspect also, having been born under circumstances that included a plume of virulent microbes.

  Arnout de Bruin, a molecular biologist with a background in evolutionary studies, was part of the RIVM team that went to Herpen. When I met him at the institute’s headquarters, a fenced complex in a suburb of Utrecht, he wore a light stubble of beard and a brown T-shirt reading VARSITY TEAM—NORTH DAKOTA. He was a bright young man with a dark sense of whimsy. The funny thing about his involvement with the outbreak down there,
de Bruin told me cheerily, was that it only happened because he’d been studying Q fever as a possible bioterrorist threat. (The bacterium had a history of attracting dark interest; biological warfare researchers in the United States had worked on it during the 1950s, so had the Soviets, and four decades later the Japanese cult Aum Shinrikyo seems to have considered it, before using sarin gas for their 1995 attack on the Tokyo subway.) De Bruin’s group on that project, a “biological calamities” team, had developed PCR primers for detecting Coxiella burnetii in a sample. So when the cases started piling up in Noord-Brabant, both among goats and among people, and the health authorities wanted urgently to trace the source, they asked de Bruin’s team for help. Okay, yeah, sure. He and his partners jumped at the chance for a field test of their new molecular tools. On the advice of veterinary officials, who knew of the abortion wave on the big commercial farm, they went to that place.

  “And the farmer said, ‘This is the secure area, and this is the nonsecure area, because here the goats have been standing which had aborted,’ ” de Bruin told me. “So we took all kinds of samples. Surface area swabs, water from the drinking buckets, vaginal swabs from the goats. What did we take more? Oh yeah, for instance, insects, from the insect lamp. Dust particles, hay, manure.” He laughed grimly. “We found it everywhere.”

  What sort of protection were you wearing? I asked. Masks, respirators? None, he said, laughing again, at his own foolishness and the laxity of supervisory vigilance. “But nobody got sick.” Maybe he and his colleagues were lucky. Anyway, the farmer was wrong about which parts of his property should be scrutinized. “We found it everywhere,” de Bruin repeated. “There was no secure/unsecure area because the whole farm was infested.”

  On the basis of this field sampling and the lab results, he told me, some health officials became overly eager, inclined toward concluding too much. “They said immediately, ‘Oh, that’s the source!’ And we said, ‘Well, it is a source.’ ” But no one had checked the other farms in the neighborhood, any of which might also have been leaking Coxiella burnetii into the air. You should test those too, de Bruin advised. Meanwhile his team worked on other aspects of the outbreak-response study.

  They gathered blood samples from 443 people in the Herpen area and, in 73 of those individuals, found evidence of recent infection with C. burnetii; another 38 had been infected sometime in the past. From questionnaire information, the study team matched positives against different forms of potential exposure. The most revealing result from this analysis was that direct contact with animals was not a significant risk factor for infection. Nor was drinking raw milk. Some of the cases—but only a minority, less than 40 percent—involved contact with agricultural products such as hay, straw, and manure. From these data, the team narrowed it down to “windborne transmission” as the most likely source of Q fever in the area. The high incidence of infection among goats, the cascade of abortions, the practice of fertilizing fields with manure from the kidding sheds, the nature of the bacterium itself (more on this below), the dry April weather, and the easterly winds had combined to becloud the village of Herpen with Coxiella burnetii.

  De Bruin himself, having helped gather and analyze these data, was acutely aware how well the bacterium went airborne. Later, as the epidemic continued into 2008 and 2009, he grew more wary about field sampling. “I said, ‘Hey, we’re not going anymore without protection—because we’re lab people, we’re not immune.’ ” If you’re a farmer, he said, you may have developed immunity from prior exposure to Q fever at a level that never caused overt illness. That turns out to be quite common among Dutch farmers and veterinarians—but not among molecular biologists. “So we went with masks.” Still, it’s hard to work in a mask—your breathing constrained, your glasses or goggles fogging up—and you find that you don’t want to wear such gear a minute longer than necessary. De Bruin saw more dark amusement in the absurdity of drawing a line between what was impracticable and what was safe. He recalled driving down to another major outbreak site in the south. “I came to that farm, and the only place I could park my car was in front of the stable. So I opened my car, and there was a big wind blowing through the stable.” He got out. He breathed the wind. He thought, “And now I’m going to put on my mask?” This time we both laughed.

  The outbreak continued, growing worse in 2008, worse still in 2009. By the end of that year, 3,525 human cases had been recorded since the first alerts in May 2007, most of those still in Noord-Brabant. The infection generally made itself manifest as fever, pneumonia, and in some cases hepatitis. At least twelve people died—not a high lethality compared to some of the grisly viruses, but fairly severe when you remember that this is a bacterial infection, supposedly treatable with antibiotics.

  One cluster of cases, in 2008, occurred at a psychiatric care institution in the town of Nijmegen. After three of the psychiatric patients came down with atypical pneumonia and were hospitalized, the Municipal Health Service screened patients, employees, and visitors, finding twenty-eight cases of C. burnetii infection. What was the source? A goat farm near Nijmegen had suffered a storm of abortions, and Q fever was confirmed from vaginal swabs. The bacteria could have traveled downwind from those aborted kids. But in this instance, there was also a more immediate possibility. The psychiatric institution maintained a small flock of sheep on a meadow within the premises. During that year’s lambing season, one lamb had been abandoned by its mother—and was then adopted by a patient, who took it into her bedroom and bottle-fed it six times a day. The pet lamb was also cuddled consolingly by several other patients. This seems to have been somebody’s idea of therapy, until the lamb tested positive for Q fever.

  On the day after my conversation with Arnout de Bruin, I drove north to the Central Veterinary Institute, a university-affiliated facility near the city of Lelystad, with an annex devoted partly to research on dangerous zoonotic agents. Whatever was happening in the Netherlands to account for these sequential outbreaks, it was clearly a veterinary concern as well as a matter of human health. The CVI annex, tucked among trees off a secondary road, was so discreet that I had to circle the neighborhood twice to find it. There I was welcomed by Hendrik-Jan Roest, a slim veterinary scientist in rimless glasses and a casual blue sweater, tall enough to play forward on the Dutch national basketball team, who led me back outside immediately so we could peer in the window of a BSL-3 lab where he and his technician were growing C. burnetii. Through the little window I could see incubators and a negative-airflow hood, like the fan hood above a stove, meant to suck away ambient bacteria as his technician worked at her bench. In this building, Roest told me, we work also on West Nile virus, Rift Valley fever, and foot-and-mouth disease, among other things. Rift Valley fever, I said, you have that in the Netherlands? Not yet, he said.

  Back in his office, Roest sketched a verbal portrait of Coxiella burnetii, listing the traits that make it so unusual and problematic. First of all, it’s an intracellular bacterium, meaning that it reproduces within cells of its host—as does a virus, though by dissimilar mechanisms—not out in the bloodstream or the gut, where it could be more easily targeted by immune response. Furthermore, it exists in two forms of bacterial particle, one large and one small, each with different characteristics suited to different phases of its life history. The large form replicates prolifically inside host cells and then transmogrifies to the small form, which is tougher and more stable. The small form, almost like a spore, is packaged for survival in the external environment. (The smallness of this small form may account for why Macfarlane Burnet and some others mistook it for “a filterable virus,” a microbe so tiny it passed through filters designed to scoop away ordinary bacteria.) It is resistant to desiccation, resistant to acids, resistant to high and low temperatures, and resistant to ultraviolet light. It can live in salt water for more than six months. No wonder it travels so well, not just from host to host but from place to place—even from continent to continent.

  “Does anyone know where
it came from?”

  “I think it was always there,” Roest said.

  Always where? Always everywhere? In Montana, where Herald Cox found it, and in Australia, where Macfarlane Burnet found it, and in the Netherlands, where you’re finding it now? No, not quite everywhere, he said. There is no record of Coxiella burnetii in New Zealand. So far.

  Then why had the disease just lately—since 2007—become so troublesome in Noord-Brabant? When I asked him about the increase in dairy goats, he brushed that idea aside as too simplistic and began showing me photos and charts on his computer. One image revealed a vast building, like a train depot, filled with white goats.

  “This is the way they are goat farming.”

  “Wow.”

  “They are huge, huge barns.”

  “Big barns,” I agreed.

  Another shot gave a clearer view of what he called a “deep litter shed,” the standard arrangement for housing hundreds or thousands of dairy goats. The shed had a concrete floor, recessed below ground level so that it could contain weeks’ or months’ worth of bedding straw, goat shit, and urine, a savory mulch of organic waste that grew ever deeper and, warmed by decay, offered a lovely culture medium for microbes. New straw was added regularly, as long as possible, to stiffen and mitigate the mess. “Very slowly the package of manure and straw is getting thicker and thicker,” Roest explained, “and so the level where the animals live on is coming up.” Shin-deep in their own ordure, the nannies milled around, converting their feed to milk. As the manure rose, composting gently, it harbored uncountable abundances of C. burnetii, “alive and kicking, down deep in the litter.” By the time such a shed had filled to its brim, any single infected goat could have passed its infection to many or most of the others. Then the goats were moved out, machinery came in, shoveling began, the valuable manure was transferred to crop fields and pastures—and billions more particles of the bacterium, in its small and resistant form, were launched on the breezes.