Could Mad-Cow Disease Happen Here?

THE recent British epidemic of mad-cow disease, and the twenty-seven cases of fatal human disease associated with it, have led to the slaughter of 3.7 million cattle and the near destruction of Great Britain's cattle industry. Observers have suggested that the outbreak was a factor in the toppling of John Major's Tory government. Mad-cow disease continues to haunt Britain, and Europe in general, even though the European Community, having made extraordinary efforts, appears to have contained the outbreak. The latest figures show that the incidence of the disease in Britain is less than a tenth what it was at the epidemic's height, when more than a thousand new cases were being diagnosed in cattle every week. Still, the pummeling of the British beef industry continues. Last December the British government banned the sale of most cuts of beef on the bone, including ribs, T-bone steaks, and oxtails. With (as of this writing) a worldwide ban on British beef exports, and a severe decline in domestic sales, the price of British beef has fallen to its lowest level in twenty years. Cattle tainted by association with the disease are quickly disposed of.

A similar epidemic in the United States would be even more catastrophic. Britain before the outbreak had roughly 10 million cows; we have more than 100 million.

Cattle and dairy farmers are at the heart of thousands of rural economies, and earn approximately $54 billion a year through meat and milk sales; more than $100 billion in additional revenue is generated by related industries and services. No wonder, then, that when the British epidemic hit the front pages, two and a half years ago, the U.S. government reacted emphatically. The Food and Drug Administration, the Centers for Disease Control and Prevention, and the United States Department of Agriculture rallied to reassure us that there was no sign of the disease in this country. Yet most of the conditions thought to have led to the epidemic in Britain also existed here. Despite official protestations to the contrary, and despite regulatory changes recently implemented, some of them still do. Given current agricultural practices, avoiding an American outbreak of this disease may be only a matter of chance. The question is, how lucky do we feel?

As those who followed the horrifying unfolding of the British epidemic will recall, mad-cow disease is one in a category of progressive neurological disorders called transmissible spongiform encephalopathies (TSE). The fatal human nervous-system disorder Creutzfeldt-Jakob disease, or CJD, is also among these, and the panic in Britain began when a new variant of this gruesome affliction was discovered in association with mad-cow disease. Many things about TSE in general, and the relationship between mad-cow disease and CJD in particular, remain unclear. But that's hardly reassuring. The chain of reasoning that should make us worry begins with the fact that the economics of our modern meat and milk industries dictate that many farm animals get food supplements derived in part from rendered animal protein. The rendered animal protein they eat may expose them to the TSE infectious agent, which is thought to have the potential to cross the species barrier between animals, even into human beings. TSE is 100 percent fatal, and in human beings takes up to thirty years to manifest symptoms. Thus if, however unintentionally, we encourage the spread of TSE infection, a great deal of damage will be done before we have visible signs of the problem.

It makes sense, then, that if we make mistakes in our efforts to prevent a new variant of CJD in this country, they be mistakes on the side of caution—perhaps a higher degree of caution than we have so far exhibited.

IN the United States, as in Britain, most dairy and cattle farmers are small entrepreneurs struggling to maintain a fading way of life. Larry Johnston, a seventy-four-year-old dairy farmer in Corvallis, Montana, has engaged in this struggle for more than thirty years.

When we met, one overcast morning in June, Johnston was crisply dressed but tired. Over breakfast he told me that he'd spent the previous six hours helping a cow to deliver its first calf. Birthing calves was not the way he'd planned to spend his retirement years, he said, but there was no way around it. His daughter and son-in-law were on vacation, and the hired help, though reliable, could not handle a difficult delivery. "We're just grateful that they show up," he said. "You can't demand professionalism at five dollars an hour."

Johnston would like to pay more, but he can't. Rummaging through a pile of papers, he found a balance sheet and ran a finger down the columns. The farm had lost $57,684.30 in the previous ten months. "My accountants tell me I have one smart choice," he said. "That's to sell this place and retire to a beach in Hawaii."

From the look of things, Johnston and his family have done just about everything to make their 112-acre operation pay, including feeding their cattle scientifically. Each cow's ration is determined using a formula that takes into account its weight, age, and milk output. This information is coded and integrated into a computer chip embedded in a blue tag that the cow wears on a necklace. Johnston walked me down to the feedlot for a demonstration. A six-month-old calf stood in a pen, its blue tag dangling. A computer read the tag, deciphered its message, and dumped precisely the amount and kind of feed the calf had coming to it. The setup reminded me of something out of The Jetsons, the science-fiction cartoon. It also looked like an extravagance—but Johnston told me otherwise. Without this precision he'd be out of business. He said that when he became a dairyman, a good milking cow gave maybe thirty-five pounds of milk a day. Today that cow, if healthy, would immediately be sold at auction and ground into hamburger and franks, as all healthy dairy cows are eventually.

"Our top milkers give up to a hundred and thirty pounds of milk a day," Johnston said. "We feed the best alfalfa we can buy, from Townsend, Montana, or Mud Lake, Idaho, or eastern Washington State. A company representative [from Purina Mills] works out the formula depending on the quality of our hay. And we have a special high-power supplement for the high-producing cows."

Johnston had paid $69,581.57 for feed supplements in the previous ten months alone, but he figured it was money well spent. Without the supplements, he said, milk production would slow down and his balance sheet would look even worse. He has seen a lot of farms go under. From 1987 to 1996 the number of milk-cow operations in the United States dropped by 44 percent, and the number of cows by 11 percent, while milk production increased by eight percent. Breeding, hormones, and drugs have made America's dairy cows veritable milk machines—hungry machines. I asked Johnston for a look at the supplement, and he took me to a side barn, where a pile of it spilled out from under a blue tarp. It looked like rabbit food and smelled vaguely of grain and dung. I asked him what was in the stuff, and he said that he wasn't sure but he knew why I was asking. "You want to know if I'm feeding cow to my cows," he said. "Truth is, I don't know."

A call to Johnston's supplier, Bill Shine, who at the time was the local district manager for Purina Mills, determined that the feed was custom-made for Johnston's cows, of soybeans and canola, animal fats, meat-and-bone meal, and blood meal. Shine didn't seem to know whether the meat-and-bone meal came from cows, but he did say that "regulatory issues" would prevent the inclusion of beef in cattle feed in the future. Two months after my visit to Johnston the U.S. Food and Drug Administration banned the practice of feeding to cattle or sheep almost anything containing the meat, bone, or fat of mammals other than pigs and horses (exceptions made because the FDA contends that these animals don't get TSE). Some observers are skeptical of this ruling, in part because the FDA has allotted the equivalent of only seventeen full-time inspectors to 14,000 facilities involved in feed production and rendering—the process by which animal by-products, such as meat, bone, and fat, are melted down and separated. Another objection is that the feeding of cows' blood to cattle continues to be permitted. Although TSE in human beings has not been shown to be transmitted by blood, evidence of TSE has been found in the blood of experimentally infected animals.

The skeptics argue, too, that there is no way to prove that the species barrier cannot be surmounted, and the current rules allow even animals known to be infected with TSE to be fed to pigs and chickens, which can in turn be fed to us. Or they may be fed to cattle and sheep. A study published in April in the scientific journal Nature found that TSE from one species can be carried by another species without making the carrier sick. The authors concluded that "the results presented here would strongly favour a decision to stop feeding ruminant-derived products to all animal species"—in particular, "domestic animals such as poultry which are raised for human consumption."

American dairy farmers have fed their herds supplements containing fat, bone meal, and blood and meat protein for fifty years or more. Larry Chase, an associate professor of dairy-cattle nutrition at Cornell University, says that cows fed on grass, hay, alfalfa, and other forage produce just ten to fifty pounds of milk a day, and that "you can't get a reasonable amount of milk without supplement." Protein supplements can be made from almost any organic material, including corn and soybeans. They may also contain rendered animal by-products. When I called him, Chase took a minute to check the commodities market and found that animal by-products were going for about the same price as soybeans that day. Since animal products deliver more protein per ounce than vegetable products do, it's no surprise that many farmers have turned to animal-based food supplements. "The amount of rendered animal protein produced in this country is staggering," Chase said. "And from what I understand, ten to fifteen percent of it goes into cows"—an estimate that he believes holds true today.

Feeding cow to cow may sound bizarre, but from a nutritional standpoint it makes perfect sense. Richard Race, a research veterinarian at the National Institutes of Health's Rocky Mountain Laboratories, in Hamilton, Montana, explains that animal protein is more complete than most vegetable proteins, and that the best protein an animal can eat, metabolically speaking, is derived from the tissues of its own species. "Theoretically, the best protein for a dog is dog, for a cat is cat, and for a cow is cow," he says. "For that matter, the best protein for humans is human." Donald Beermann, a professor of animal science at Cornell, observes that feeding animal protein to cows significantly accelerates their rate of growth. And, he says, animal activists are wrong when they claim that cows are strict vegetarians. "It's not uncommon for a cow giving birth in the field to eat its placenta," he says. "Makes sense. Nature is not in the practice of wasting valuable nutrients."

One may well ask, then, why most mammals approach cannibalism as, at best, a last resort. Why, for example, don't competing males kill and eat one another, and why don't most mammals eat their dead? Scientists can only speculate that such behavior bodes ill for a species' long-term survival. Disease spreads more easily within than among species, and a species that eats its own has a higher risk of being killed off by disease. Cannibalism also amplifies levels of infectious organisms, concentrating and increasing their presence in the environment and spurring their chances of hurdling the species barrier and moving into another, related species. A microbe that normally infects sheep fed repeatedly to, for example, cows may mutate in order to accommodate to its new host, thereby becoming the agent of a cow disease. This is why the Food and Drug Administration has banned the feeding of most mammals to cows or sheep and other ruminants and has stipulated that feed derived from ruminants bear the warning label "Do not feed to cattle or other ruminants." Rendered products of cow and sheep carcasses will, however, continue to be manufactured and labeled as feed for hogs, chickens, and other farm animals—a practice forbidden in Europe in 1996, when it became clear that material labeled pig or chicken feed was finding its way into cows and sheep. Gloria Dunnavan, the director of the Division of Compliance at the FDA's Center for Veterinary Medicine, says that although the agency is looking for 100 percent compliance on the regulation, it is not 100 percent optimistic. "We've developed an enforcement strategy, but it's been quite a challenge," she says. "Not a lot of people in the rendering and feed industry are used to seeing an FDA presence. But the groundwork is being laid, and to date we've received a lot of cooperation from the industry."

In his 1978 book Rendering: The Invisible Industry, Frank Burnham, the editor of Render magazine, described his business as having been "almost invisible for more than 150 years quietly doing its thing on the back streets of America." Indeed, today it is very difficult for a journalist to gain entrance to a rendering plant, but a clandestinely made videotape of a sampling of plants, which was provided to me by the Government Accountability Project, a public-interest law firm in Washington, D.C., portrays a series of altogether unsavory places where dead cats and dogs, road kill, the occasional circus animal, and the diseased carcasses of farm animals are mixed into a ghastly, belching stew.

Swift and Company, the Chicago meatpacker, was the first to mass-manufacture rendered protein and fat as animal feed, at the turn of the century, but the use of such feed really took off in Britain during the Second World War, when it became difficult to import vegetable feeds from abroad. In Britain cows were fed as much as four pounds of rendered protein a day, but in the United States, where soybeans and other vegetable proteins were plentiful, the use of rendered animal protein as feed was far less common. By the mid-1970s, however, the price of vegetable protein in the United States had risen to the point that rendered animal protein became an increasingly popular cattle feed—though never as popular as it was in Britain.

Then the energy crises hit, and the rendering industry sought to process material at lower temperatures and, in England, with smaller amounts of costly and dangerous petroleum-based solvents (solvents are not used in the United States). What it came up with was a vacuum system—a neat trick that those who recall fifth-grade science class will understand immediately. Heating water in a vacuum allows it to boil at lower temperatures. The vacuum system, though it is effective for rendering, unfortunately does not destroy all bacteria or viruses or, for that matter, the infectious agent associated with mad-cow disease. A study published in The Veterinary Record in December of 1995 reported that when neurological tissue infected with mad-cow agent was put through this process, the level of infectivity showed no reduction after rendering, whereas the old-fashioned high-temperature method reduced infectivity significantly. According to sources in the rendering industry and at the USDA, low-temperature rendering without solvents could well have led to the mad-cow outbreak in Britain, and it certainly has the potential to spread infection in the United States if the disease occurs here.

"Before the 1970s there was a reduction in infectious material by at least a factor of a hundred using the standard rendering system," a scientific consultant told me. "But the new low-temperature systems did nothing to reduce infection. In England sixty to seventy percent of rendering plants switched to the low-temperature system by the early 1980s. And in the United States anywhere from one quarter to one third of all rendering plants still use the low-temperature process."

This consultant, who declined to be named, warned that low-temperature rendering combined with the American practice of putting diseased animals into the rendering cycle was a potential threat to the public health. Although a voluntary ban has been instituted on rendering the carcasses of animals condemned for signs of neurological disease, critics worry that such a thing cannot be completely effective—especially because many animals are not examined by a veterinarian before being rendered. "If there's one industry within agriculture or food production that needs regulating in this country, it's the rendering industry," the consultant said. "The number-one World Health Organization recommendation to prevent the spread of TSE was a ban on the use in the food cycle of animals that show signs of TSE disease. Yet the FDA completely missed that recommendation, and TSE-diseased animals can still be in there. If you're going to process animal materials, you need minimum processing conditions—for example, a standard of temperature that will kill all infectious agents. And I don't see them doing that at all."

IN his excellent Plagues and Peoples (1977), William H. McNeill pointed out that many of the distinctive infectious diseases of human beings got their start in animals, particularly domestic animals. Measles is probably related to canine distemper, and influenza to a disease in hogs. Clinical and experimental evidence has led many scientists to suspect that mad-cow disease has made the jump as well. The infectious agent responsible for the British mad-cow epidemic was given ample opportunity to mutate into a human pathogen. It is thought that stricken cows ground up, processed into feed, and fed to other cows created a cycle of intense infection that eventually landed on people's dinner plates; after repeated exposure to this infected meat, a group of human beings, especially susceptible for genetic or other as yet unknown reasons, developed a human form of the illness, a variant of Creutzfeldt-Jakob disease.

CJD in its natural form was first described in the 1920s by the German physicians Hans Gerhard Creutzfeldt and Alfons Jakob. Symptoms vary, but may include loss of coordination, personality changes, mania, and dementia; death inevitably follows. It is generally said that about nine people a year in 10 million contract CJD "spontaneously"—that is, by a means unknown—and one in 10 million does so through inheritance. Still more rarely CJD has been contracted through the transplantation of infected tissues, such as corneas, through contaminated surgical instruments, or through the injection of growth hormones derived from CJD-infected pituitary extracts. Though it has not been shown to be transmissible in blood outside the laboratory, dread of the disease is such that more than $100 million worth of blood products have been destroyed for fear of CJD contamination. In the United States about 250 cases occur each year. Although young people can and do acquire CJD, and the first case ever described was in a twenty-three-year-old woman, the average age of onset in the Western world is over sixty. The new variant of CJD associated with the mad-cow outbreak differs clinically from the traditional form in two significant ways, both of them cruel. The old form is quicker, mercifully. And whereas traditional CJD is a disease of the elderly, the new form seems to prey on the young: the median age of onset is twenty-eight.

All transmissible spongiform encephalopathies, including CJD, attack and destroy the nervous system of the host organism. No one knows precisely how the TSE agent infects an organism, but once it takes hold, it is unstoppable. Strains of TSE have been found in sheep, goats, elk, mule deer, white-tailed deer, mink, and cats. In sheep and goats TSE is called scrapie. Mad-cow disease is more properly called bovine spongiform encephalopathy, or BSE. In human beings CJD and two extremely rare syndromes, Gerstmann-Sträussler-Scheinker disease and fatal familial insomnia, have been identified as forms of TSE. Clarence Joseph Gibbs, who is the chief of the Laboratory of the Central Nervous System at the National Institutes of Health and an expert on TSE, says it's probable that most if not all mammalian species acquire TSE, but that this is difficult to prove, because "no one would notice a rabbit with dementia and bad balance."

Gibbs and other scientists I recently spoke with believe that all mammals that acquire TSE do so at the human rate—about one case per million. If this figure is accurate, then approximately a hundred U.S. cows come down with BSE spontaneously each year. But detecting isolated cases of BSE in American cattle herds is exceedingly difficult, in part because the cattle population is too widely dispersed for efficient surveillance. "We've had one reported case of BSE in North America—in Canada, in a cow imported from the United Kingdom," Gibbs told me. "That cow was found by a rancher who had gone out on the range to feed his herd because of a severe snowstorm. If there had been no storm, the rancher would have stayed home, the cow would have gone down, a coyote would have eaten it, and no one would have been any the wiser. Do we have BSE in the United States? The real question is, if we do, will we find it?"

The U.S. Department of Agriculture has examined the brains of only about 7,100 dead suspect cattle for signs of BSE, and about 2,000 have undergone an additional procedure required for a diagnosis of the illness. Tam Garland, a veterinarian and a research associate in the Department of Veterinary Physiology and Pharmacology at Texas A&M University, describes this number of tests as "phenomenally small." She also says that she is not surprised that no more suspect animals have been reported. "As a rancher, you're not going to haul a vet out onto the range to look at a dying animal, only to get fingered by your neighbors as the cause of plummeting beef prices," she told me.

TSE infectivity concentrates in the central nervous system, in the spinal cord and brain. And the more of the infectious material an animal is exposed to, the likelier it is to get sick. For this reason Britain has implemented a series of more and more restrictive bans from its food chain, starting with cattle brains, spinal cords, and other tissues that have been shown to contain infectious material, and now including some sheep and goat parts as well. The United States has not followed suit, and the heads and backbones of cows, pigs, and other animals continue to figure prominently in the rendering mix.

In fact, brain doesn't just enter our food chain but goes directly into the human food supply. Each year, according to the USDA's Food Safety and Inspection Service, which is charged with regulating the meat industry, about a million cow brains are sold for consumption in the United States, and others are exported. Generally these are whole brains taken from cows slaughtered in the kosher manner—that is, their throats are cut while they are still conscious. Most cows in the United States, though, are shot in the head with a stun gun before having their throats cut, a process perhaps more humane but also, at least when a pneumatic stun gun that injects air is used, more likely to contaminate other parts of the carcass with brain. This pneumatic gun can crack the cow's skull, causing the brain to leak out. And the force of the gun is such that it can blow pieces of brain into the bloodstream, where they can be carried to the animal's lungs or liver. Tam Garland has recovered chunks of brain as large as six inches across in the lungs of slaughtered cattle. Canadian researchers have spotted pieces of brain about two and a half inches wide lodged in cow livers.

There is no law against the addition of heart, liver, and other "variety meats" to hot dogs, sausages, and canned or bottled meat products, such as chili and spaghetti sauce (when present, variety meats are noted in the list of ingredients). Nor does any regulation mandate that spinal cords be removed before carcasses are processed, although some plants are said to follow this procedure nonetheless. Unless spinal cords are removed, "mechanically deboned" meats, which are found in products such as hamburger and bologna, can be contaminated with bits of central-nervous-system tissue when the spinal columns are mechanically plucked and squeezed to get off the last bits of flesh.

Will Hueston, a veterinary epidemiologist formerly with the USDA and now an associate dean at the Virginia-Maryland Regional College of Veterinary Medicine, says that the agency has not imposed a ban on spinal column and brain in the rendering process because to do so would be very expensive and basically unenforceable. As one rendering-plant employee told me, "Who would want the job of cutting out all those spinal columns and brains? You couldn't even get occasional workers to do it." But when asked why brain and spinal column are not banned from the human food supply in the United States, Hueston was circumspect, suggesting that I speak with someone at the Food Safety and Inspection Service.

Jacque Knight, a spokeswoman for the FSIS, was vague and seemed annoyed when asked whether brain or spinal cord was getting directly into sausage or other processed meats. "Since the Meat Act of 1906 we have never prohibited brain or spinal cord," she said. "It is part of the animal. However, it is not something people expect to find in meat. Therefore, as of May of last year, we have told our inspectors that if they suspect brain or spinal cord of getting into meat, they should report it." The inspectors' union expressed concern that it would be difficult for its members, already bogged down in other duties, to take on this additional task. The agency employs 7,535 inspectors in 6,200 meat and poultry slaughtering and processing plants. To date fifty-four reports have been filed and five plants have been found in violation of the rule.

TSE infectious material does not concentrate in muscle tissue, so steaks, chops, and roasts are probably fairly safe to eat. Though contamination of these cuts is possible (through, for example, splattering when the spinal cord is cut during processing), the amount of infectious material on them, if any, is likely to be low. More than one scientist told me that if I was concerned about my family's exposure to TSE agent, I would do well to steer clear of processed and reformed meats such as bologna and head cheese. Not being a head-cheese fan, I have not found it difficult to follow this advice. In addition, I avoid any ground meat that hasn't been ground in the store where I buy it. I get hamburger by asking the butcher in my supermarket to put a chuck roast through the mill. I no longer buy spaghetti sauce with meat (an old favorite), and my eight-year-old daughter has forsaken lunch meats for what she calls "vegetarian burgers," which consist of a slice of American cheese melted over two sandwich pickles on a bun. Though whether it is possible to acquire CJD from American beef is entirely unclear, making these small culinary adjustments seems well worth doing given the gravity of the disease and the government's hesitation to do all it can to avoid it.

THE Rocky Mountain Laboratories is an unassuming NIH outpost hunkered in the shadow of the magnificent snow-crested Bitterroot Mountains. Built on a residential street in Hamilton, Montana, the red-brick structure stands out like a Brooks Brothers suit on Casual Friday. The lab was set up by the state in 1928 to study Rocky Mountain spotted fever, a deadly tick-borne disease that plagued the region. The people of Hamilton were so nervous about ticks going AWOL from the lab that a moat—really, a puddle-deep trench—was dug around its perimeter. Over time the laboratory has branched out to consider a wide range of infectious diseases, and has attracted, among other notables, a world-renowned team of experts on TSE. One of these is Byron Caughey, a strapping, bearded biochemist who looks nothing like the poet whose name he shares.

For about the past decade Caughey has studied "prion" proteins, molecules of uncertain function that are found on the surface of nerve cells. The buildup of a distorted form of these proteins—called, simply enough, abnormal prion proteins—is the identifying characteristic of TSE. In a scientific paper published in 1995 Caughey and his co-author, Peter Lansbury, a protein chemist at the Harvard Medical School, likened abnormal prion proteins to ice-nine, the sinister agent that catalyzed the freezing of all water on earth in Kurt Vonnegut's classic novel Cat's Cradle. Like Vonnegut's ice crystals, abnormal prion proteins convert neighboring prion proteins to the abnormal form, contorting protein molecule after protein molecule in a demonic domino effect. The resulting prion-protein buildup kills brain cells and forms the characteristic spongelike holes, the "spongiform" encephalopathy, of TSE.

Whereas the symptoms of TSE are clear, the debate over what initiates the conversion of normal prion proteins to the virulent form—that is, over what constitutes the infectious agent in TSE—is one of the hottest and most contentious controversies in modern biology. One theory has it that the abnormal prion protein itself is the infectious agent. The British mathematician J. S. Griffith was the first to propose the idea that proteins could be infectious, in a paper published in 1967 in Nature. He suggested that proteins might self-replicate much the way crystals do, by building up around a central seed crystal. But crystals do not "grow" in the organic sense of the word. The notion that proteins, the building blocks of life, could replicate without the aid of genetic material, or the nucleic acids that make up that material, was heretical. Even the simplest microbes require nucleic acids to direct their growth. As one scientist puts it, the notion that a protein in and of itself could control life "defies basic Darwinian principles." Coming as it did from a mathematician rather than a life scientist, the crystal theory was regarded largely as a theoretical construct. In any case, Griffith never published on the subject again, and when he died, in 1972, his theory faded. It was assumed by virtually everyone at the time that TSE was caused by a virus.

Viruses are extremely basic creatures, each consisting of a single thread of genetic material enclosed in a protein sheath. Unlike bacteria, they lack the wherewithal to multiply on their own and rely on the largesse of the creatures they infect. Generally it takes a few hours or days for a virus to replicate in large numbers. The host recognizes this viral buildup as an alien onslaught and, unless weakened or old, usually marshals its immune system in order to ward off the attack. Unlike most viruses, the TSE agent does not attack within hours or days but bides its time, hiding out and multiplying in the host for years, even decades, before making its presence known. In the meantime, TSE does not elicit an immune response, and there is very little inflammation—indeed, no outward sign of infection at all until the disease has progressed beyond the body's ability to control it. For this reason some scientists assume that a TSE virus lurks cloaked in the host's own protein, disguised as a natural part of the host until it's ready to pounce.

"Whatever is causing this disease is unique; it falls out of the realm of all other disease agents," says Richard Rubenstein, the head of the molecular and biochemical neurovirology laboratory at the New York State Institute for Basic Research, in Staten Island. "We feel that it is probably a virino—a nucleic acid protected by a host protein that camouflages it so that the immune system doesn't pick it up. There is no antibody reaction to the agent at all. The body simply does not recognize it as foreign."

Detecting this particular agent is difficult, and so far neither Rubenstein nor any other scientist has been able to find it. Many have given up the hunt, concluding that the virus cannot be found because it does not exist. Among these is Stanley Prusiner, a professor of neurology at the School of Medicine of the University of California at San Francisco. It was Prusiner who, in 1982, coined the term "prion," to describe what he believed was the probable agent of scrapie: not a virus but Griffith's long-abandoned infectious protein. Eventually—most notably in a 1995 paper in Scientific American—Prusiner went so far as to assert that prions are the cause of TSE. Prusiner's is a powerful and convincing voice, and the prion theory of TSE transmission has come to dominate the field. Last year Prusiner was awarded the Nobel Prize in medicine for discovering prions.

"Some of us are reluctant to use the term 'prion' for the infectious agent, because it presupposes that the agent is protein only," Caughey told me. "But so far most of the evidence points in that direction." Caughey explained that unlike most viruses, the TSE agent is undeterred by exposure to radiation and resistant (up to a point) to destruction by heat—both indications that it lacks nucleic acid. Peter Lansbury, Caughey's frequent collaborator, says that the agent "has the chemistry of a protein." David Bolton, a scientist at the Institute for Basic Research, who got his start in Prusiner's lab at UCSF, agrees. Bolton says, "Every aspect of this disease is best explained by the protein-only model." But Bruce Chesebro, a virologist, an immunologist, a physician, and Caughey's boss at Rocky Mountain, is not so sure. Considered by many to be the "voice of reason" in what has become a rancorous debate, Chesebro believes that the abnormal prion protein is a result of TSE infection rather than the cause of it.

"Viruses are very small, and sometimes they take decades to find," he explained to me. "We haven't found one yet, but that doesn't mean it's not there." Chesebro also said that TSE agent is not uniquely invincible, as some have implied. "Most of the infectious agent dies under exposure to prolonged high heat and detergents," he said. "You can get rid of ninety-nine-point-nine percent of it simply by heating the stuff at a high temperature for thirty minutes." To buttress his point that viruses can be very hardy creatures indeed—perhaps as tough as TSE infectious agent—Chesebro reminded me that viruses recently discovered in the geysers of Yellowstone National Park live near or at the boiling point.

Laura Manuelidis, the head of neuropathology at Yale Medical School, is perhaps the harshest critic of the protein-only hypothesis. She contends that the public mistakenly believes that Prusiner (who rarely gives interviews) has proved his case, and that this has made it difficult for scientists with an opposing view to get recognized, or even heard. "The prion has taken on a mythical component—it's become magical," she says. "But to me, TSE shares all the signs of a slow viral disease. In sheep it is found in the classic places viruses go: the spleen, the lymph nodes, the central nervous system. In cows it's in the central nervous system—where many slow viruses can hide." Despite arguments to the contrary, Manuelidis contends that the prion protein itself does not infect animals, and that the true infectious agent is still unknown. "It would be a shame," she says, "if we followed the wrong trail while an infectious epidemic was incubating."

For now, the infectious agent remains a mystery, and perhaps an academic one. Whether or not abnormal prion proteins are the cause of TSE, they are certainly among its results; and by the time the protein conversion has started, the case is beyond treatment. A more immediate question, then, is how the agent made its way into so many British cows. One theory has it that several cows with spontaneous BSE were rendered, made into feed, and fed to a large number of other cows, who were themselves rendered and served. A few cycles of this could infect a large number of cows. But many experts are dubious of this theory because spontaneous BSE, if indeed it exists, is probably rare, whereas the infection in British cows became widespread. Another possibility is that BSE got its start not in cows but in sheep infected with scrapie—the ovine TSE. What is particularly unnerving about this idea is that scrapie is not at all a rare disease.

ENGLAND has struggled with scrapie since the mid-1700s. In July the British government took the highly dramatic step of requiring all sheep and goats even suspected of being infected with the disease to be slaughtered and their carcasses disposed of by the Ministry of Agriculture, Fisheries and Food.

Scrapie was long thought to be a purely genetic disease, but by the mid-1930s French researchers had shown that it is infectious. The disease got a foothold in the United States in 1947, when an outbreak, traced to an import of purebred Suffolk sheep, was reported in Michigan. In 1952, when scrapie outbreaks were reported in California and Ohio, the USDA launched the first of two eradication programs, requiring the slaughter of entire herds infected with even a single case of scrapie. Not surprisingly, sheep farmers were reluctant to report suspected cases, and scrapie, as one scientist puts it, was "driven underground."

From the mid-1950s through the mid-1970s a scant one to twenty-two flocks a year were reported infected—numbers that many thought were too small to be credible. In 1978 the USDA instituted a program of reimbursing farmers two thirds of the appraised value of the sheep sacrificed in their flocks, up to $300 per animal. But in 1983 it was decided that farmers would be required to kill (and would be reimbursed for) not their entire flocks but only infected sheep and immediate relatives. At this point reports of scrapie increased. Since scrapie is known to spread through contact, the new rule made little sense—blood relations had a slightly elevated chance of carrying the disease but were certainly not the only sheep at risk. In 1992, for a combination of scientific and budgetary reasons, the scrapie-eradication program was essentially dismantled: farmers were given six months to report sick sheep for reimbursement, and then the coffers were closed. Today a voluntary system is in place, under which farmers can apply to have their sheep certified "scrapie-free." But as of this writing less than one percent of the country's flocks have joined the program. "No one knows the exact incidence of scrapie in the United States," Richard Race, of the Rocky Mountain Laboratories, says. "Surveillance in some states is so poor, and some vets tell me they don't do anything about it."

Known strains of the scrapie agent do not seem to be infectious to human beings: at any given time over the past couple of centuries as many as a third of British sheep have been infected with the disease, but they have not, apparently, conferred it on the people who ate them. However, American scientists have found that they are able to transmit scrapie from sheep to cows by injecting infectious material into their brains. Since scrapie is an old disease and BSE a new one, it seems that if the agent got boosted over the species barrier, it did so only in the past few decades.

Linda Detwiler, the senior staff veterinarian for the USDA, coordinates BSE surveillance and used to coordinate surveillance for scrapie. When we spoke, she reassured me that scrapie is "not rampant" in the United States—that thirty to sixty-five infected flocks a year are reported. But when I asked how many cases might go unreported, she responded, "We really don't know the amount of scrapie that there is in the United States. This is mostly due to the lack of a validated test—although one is being developed—that can detect scrapie in a live animal not yet showing signs of the disease. Some sheep producers are afraid, because the lack of a live-animal test has led to programs that have either destroyed animals without scrapie or have left producers with diseased animals." There is no telling how many herders quietly kill off their scrapie-infected sheep and bury them, burn them, or send them to a willing renderer to be converted into feed for other animals.

William Hadlow, a research veterinarian who devoted his career to the study of scrapie and other TSEs, is a world expert on them. Though he retired more than a decade ago, his house, just one block from the Rocky Mountain Labs, is piled kitchen to living room with texts on animal pathology and neurology. Hadlow explained to me when I visited him that the scrapie infectious agent appears in several strains—as might be expected if the agent is a virus, but not if it's a protein. "People who accept the idea that a protein is responsible for transmission of this disease probably haven't gotten within arm's length of a sheep," he told me, adding that he had worked with Stanley Prusiner and respected him, but that he had difficulty accepting the prion theory. And it is likely, Hadlow said, that each scrapie strain is tailored to the genetic makeup of the sheep it infects.

In 1979 two of Hadlow's colleagues injected the brains of ten cows with a strain of scrapie found in American sheep and goats. Hadlow later examined the brains and wrote up the case with his colleagues. The ensuing paper was rejected by three journals before being published in the American Journal of Veterinary Research, in May of 1995. The findings were alarming. Although Hadlow and his colleagues found only scant spongiform changes in the cattle's brains, they did find disease in three of the ten cows. "In all 3 affected cattle," they wrote, "the disease was expressed clinically as increasing difficulty in rising from recumbency, stilted gait of the pelvic limbs, disorientation, and terminal recumbency during a 6-to-10 week course." They concluded, "Pathologists should keep this variability in mind when looking for microscopic evidence of a scrapie-like encephalopathy in cattle." What Hadlow had found was that American cattle infected with American scrapie developed a very different disease from the BSE diagnosed in British cattle.

"My point is that there are a number of sheep-scrapie strains extant in the United States, and that these strains may not cause the same clinical signs in American cows as the British strain of scrapie caused in British cows," Hadlow said. "The changes in the brain we found in Texas were very subtle, and they might easily be overlooked if you're not looking for them."

Indirect evidence of just such an infectious agent in cows in the United States was uncovered by Richard Marsh, who, until he died last year, was a member of the Department of Animal Health and Biomedical Sciences at the University of Wisconsin at Madison. Marsh was an expert on transmissible mink encephalopathy. In 1985 he investigated a mysterious outbreak of TME on a Wisconsin mink ranch. Marsh was told by the rancher that the minks had been fed fish, cereal, poultry, and large amounts of butchered and ground "downer cattle." Although no one is tracking how many cows "go down" in the United States each year (that is, become too lame or too ill to stand on their feet), a common estimate is 100,000. Downer cows are by law not admitted into the food supply without first being cleared by a veterinarian, but most enter the food chain through the rendering process.

Marsh found that he could pass the mink disease into healthy cows by injecting their brains with TME, and that he could pass the disease back into healthy minks by simply feeding them the brain tissue of infected cows. The permeability of the species barrier that this experiment demonstrated led Marsh to conclude that the United States may well have its own brand of BSE—one that is perhaps being disguised by the much broader problem of downer cows.

Chronic wasting disease, another TSE, was diagnosed more than a decade ago in mule deer and elk in Colorado and Wyoming, and it is said to be spreading. Whether outbreaks of the disease have anything to do with scrapie or cattle feed is uncertain. What is known is that these and other diseased animals can be recycled back into the food chain through the rendering process. If TSE in elk or minks is capable of jumping the species barrier, it is being given every opportunity to do so in the rendering vat. And although the FDA's recent feed ban will limit the amount of rendered beef and lamb that is fed to cattle and sheep, it is almost certainly increasing the amount of rendered beef that is fed to other domestic and farm animals, notably poultry and pigs. The FDA estimates that the cost of the ban to the rendering industry is about $200 million a year, because lessened demand from the cattle industry is bringing down the price for animal feed made from beef and lamb. But if the FDA is right, this can only make the feed more attractive to farmers of other animals.

Pigs are a particular worry. Although they do not seem to get a TSE spontaneously, a recent study in England showed that one of ten pigs inoculated with BSE came down with an analogous disease. Several consumer groups have voiced concern that TSE in pigs may have appeared in this country. In 1979 the USDA sponsored a study of 106 pigs that showed symptoms of central-nervous-system disease on their arrival at a packing plant in upstate New York. The brains of seventy of the pigs were examined, and one was found to have evidence of the spongelike formations and lesions associated with scrapie and transmissible mink encephalopathy. The study was shelved until 1996, when the USDA, alarmed by the British mad-cow outbreak, sent a single slide of the pig's brain to William Hadlow. Hadlow examined the slide but could find nothing conclusive. "It was a bum slide," he recalled. "There was no evidence to make any diagnosis." Four months later the USDA sent him seven additional slides. Again, the slides were not well prepared, and Hadlow could draw no solid conclusions. But, Hadlow cautioned in a report to the USDA, "should neurologic disease occur in swine exposed to [TSE], conceivably it could be expressed microscopically, as is scrapie in some sheep and as is the encephalopathy in cattle experimentally infected with scrapie agent from American sheep." In other words, if the disease exists in pigs, it could be in a form researchers don't recognize. Meanwhile, no public agency is screening for TSE in pigs.

Last summer the Yale researcher Laura Manuelidis showed that CJD can evolve into more virulent strains by being passed from human beings to mice, from there to hamsters, and finally to rats. Her paper, published in the July 4, 1997, issue of the journal Science, concluded that when the agent changed, it provoked a variant disease, and warned that public-health agencies should be on the watch for diverse signs of brain pathology in human beings—not only those symptoms known to be associated with the new variant of CJD. Manuelidis's concern is that if a new variant of CJD has appeared in the United States, we will be slow to realize it, because we are looking for the variant that appeared in Great Britain.

"Fewer than ten percent of all deaths are investigated with an autopsy, and even a smaller percentage of victims of dementia," Manuelidis said. "Alzheimer's disease was rarely diagnosed prior to 1940, but now we diagnose all sorts of people with the disease. But Alzheimer's is a heterogeneous disease with many different causes. One cause for some people could be an infectious agent. And we really have no idea how much CJD there is. The one-in-a-million figure may be an underestimate."

From the look of it, CJD, too, is a heterogeneous disease. Some victims have plaques in their brains, whereas others don't. Some suffer with evident symptoms of the disease for years before dying, some only for months. What made the new variant of the disease particularly noticeable in Great Britain was that the victims were young, and it has not gone unmentioned that other, older victims may well be overlooked, presumed to be afflicted with another form of dementia.

Jeffrey Almond, a British scientist, is a member of the Spongiform Encephalopathy Advisory Committee, a body of scientists and physicians appointed in 1989 to advise the British government on BSE and other TSEs. When we met, at an infectious-disease meeting at the Harvard School of Public Health last year, he looked haunted. He had recently received a heartrending letter from the family of a twenty-one-year-old man who had died of the new variant of CJD. This family, he said, simply pleaded to learn the underlying cause of the young man's death, but others might have reacted differently. He said to me, "If there is an epidemic of this disease—and there might be—and you lost your son or your wife to CJD, you might wish to blame someone, wouldn't you?" He said that he regrets that Britain did not make greater efforts to control its meat industry before the outbreak. "Every bit of the carcass is used—it's all valuable, isn't it?" he murmured over a lunch of vegetarian lasagna. "They cut off what they can; then they use solvents and sprays to get off the last bits to make into mince and pâté, gravies and sauces. And they put the bones—including the vertebrae—into gelatin. It's a messy business, hard to control. And as recently as 1995 we were, I regret to say, still allowing mechanical recovery of meat off the vertebrae. We finally put a stop to that—but too late." In the United States, of course, this practice is to date unchanged.

Much in the scientific picture of CJD today is hazy. What is clear is that protecting the population from the most obvious potential source of the disease—TSE from animals—would be a relatively straightforward affair. And yet regulatory agencies in the United States have not taken up the challenge. The veterinary epidemiologist Will Hueston, who is the only American member of Britain's Spongiform Encephalopathy Advisory Committee, has witnessed firsthand the terror, chaos, and agony that mad-cow disease brought to Britain. A cautious man, he is concerned lest the public overreact. When I spoke to him not long ago, he argued that the rendering industry performs an invaluable service: to some extent it sanitizes waste products that might otherwise be dumped into landfills or, worse yet, left to fester. But he warned that we can't be certain that any current processing method will sterilize rendered products. In a better world, he said, no diseased animals or central-nervous-system tissue of any kind would go into the food chain. Daniel McChesney, a microbiologist at the Animal Feed Safety Branch of the FDA, told me that such steps would become a priority if an outbreak of BSE were to occur in this country, but for now there is simply no need. "The measures that we're taking are proactive," he said. "Great Britain has the problem; we do not." When I mentioned this to Hueston, he was not impressed. "I, for one, have no interest in eating brain," he said.

Stephen Morse, the director of the Program in Emerging Diseases at the Columbia University School of Public Health, worked with Richard Marsh on his studies of TSE agent in minks. In a letter to the journal Nature in March of 1990 he said that he did not believe that what was then generally called scrapie agent would ever find its way through the food chain and into human beings. But he has changed his mind. "Many were wrong about it, including me," he told me recently. "We didn't imagine it could pass from cows into humans. But now we think it can, and it has the potential to be terrifying. Perhaps the best analogy is to the AIDS epidemic. Although it's almost certain that TSE doesn't transmit as readily as HIV, it's similar in a number of other ways. It can remain in the body for long periods without obvious symptoms, and it is fatal. But what I'm thinking about is how we regarded AIDS in the early days, before we really understood it. We underestimated the threat. Perhaps we should avoid making that mistake again."