Looking for One of the World's Most Dangerous Toxin Bacteria

By GINA KOLATA
The New York Times January 6, 1998

People in the United States are sickened by their food once a year, on average. But the food poisoning that suddenly appeared in White City, Ore., in December 1981 was something else.

It was terrifying, and dramatic, and it seemed to have no cause. In the end, however, that mini-epidemic turned out to be the first show of force by a new strain of bacteria that is now infesting the food supply, infecting as many as 20,000 Americans a year and killing 500.

This is a scientific detective story about those bacteria, known as E.coli O157:H7. It is a story that is continuing and that, in the end, may lead to sweeping changes in the way food is regulated in the United States.

The bacteria that puzzled scientists 16 years ago have turned up in some of the most notorious food poisoning episodes in recent history, including one involving Jack in the Box restaurants in the Northwest in 1993 where four children died after eating hamburgers and hundreds of people became ill.

The bacteria surfaced in 1996 in cases traced to contaminated apple juice and lettuce, then dominated headlines last August when 17 people in Colorado got sick after eating hamburgers from beef processed by a Hudson Foods plant in Nebraska. Hudson, based in Rogers, Ark., recalled 25 million pounds of its ground beef, the largest meat recall in history, and the plant closed its doors.

Normally, Escherichia coli bacteria live innocuously in the intestines of people and animals. But the new E.coli strain can cause diseases from standard diarrhea to kidney failure and death. The bacteria still befuddle medical doctors, who have misdiagnosed their infections as everything from appendicitis to blocked blood vessels of the colon.

Once someone is infected, there is no effective medical treatment to combat the disease, and all doctors can do is prevent dehydration, wait for the disease to run its course, and hope for the best.

"I say the 'E' in E.coli stands for evil," said Nancy Donley, a Chicago mother whose son, Alexander, died in 1993 from an E.coli O157:H7 infection.

In just four days, it "destroyed all of Alex's internal organs and liquefied parts of his brain," Mrs. Donley said. It is, she added, "a terrible, terrible bug."

The search for the killer strain of E.coli began when about a dozen people showed up one week in December 1981 in the emergency rooms of White City, a working-class city near the California border. The patients described terrifying symptoms: diarrhea attacks occurring several times an hour accompanied by abdominal cramps that women said were worse than labor pains and men said felt like a hot sheet of metal had been placed on their abdomen. At first, the patients said, they had watery stools, but within hours, their diarrhea turned into what looked like pure, bright red blood. And it lasted for days.

White City doctors tested patients' stools for a long list of bacteria and parasites, but could not find a culprit. It was not even clear whether the disease was caused by an infectious organism, like a bacterium or a virus, or whether it was due to a toxic chemical. Finally, on a cold damp day in February 1982, frustrated officials of the Oregon public health department called in Dr. Lee Riley, who had recently begun work as a medical detective for the Centers for Disease Control and Prevention in Atlanta, and who is now a professor of infectious disease and epidemiology at the University of California at Berkeley.

Riley immediately flew to White City. "We figured that whatever we were dealing with had to be completely new," he said.

The first question Riley tackled was, What did the patients have in common that their friends, family members and neighbors did not? One clue was that they had eaten at a local McDonald's restaurant.

Still, the association might have been purely coincidental. And if the source was McDonald's food, it was not certain what item was causing the illness. The hamburgers? Unlikely, Riley reasoned. The meat patties were cooked at such a high temperature, he said, that "we couldn't believe that that kind of cooking would allow a microorganism to survive."

While Riley dissected the Oregon cases, the disease struck again, this time in Michigan in May 1982. Once again, the patients had just eaten at a McDonald's restaurant.

At the same time, working on the mystery disease at the disease control centers, Dr. Joy Wells, a microbiologist, noticed something suspicious. Nine of the 12 patients whose stools were obtained at the start of the illness contained a type of E.coli designated O157:H7, a label based on the proteins found on the surface of that strain. When the stool of healthy people was tested for comparison, the strain was absent. But the medical investigators were leery.

"We just didn't think E.coli could do something like that," Riley said. It was possible that E.coli O157:H7 was a perfectly harmless bacterium that just happened, by chance, to be in those patients' stools.

Finally, in July 1982, the investigators found what they needed. A plant in Ohio that had supplied meat to the McDonald's in Traverse City, Mich., had happened, as part of its quality-control program, to have saved ground beef from the suspected lot. Lurking in that meat was the same strain, E.coli O157:H7.

It turned out, Riley said, to be the smoking gun.

But if it was a common strain, it could well have been a coincidence that the E.coli were in the patients' stools and in the hamburger. So Dr. Wells began searching a database of more than 3,000 E.coli samples that had been stored for years at the disease control centers. Only one was E.coli O157:H7. It had been isolated in 1975 from the stool of a 50-year-old California woman who was a naval officer in Oakland at the time.

"We tried to track down this patient," Riley said, "but we couldn't find her."

He did, however, get her medical records from the Navy. They showed that the woman had had a bloody diarrhea that was identical to the diarrhea that struck the patients in Oregon and Michigan.

Until then, said Dr. Patricia Griffin, who is a director of epidemiology research at the disease control centers, the case was shaky. "You could say, 'Maybe that's just a nice bacteria that happens to be in food,"' she said. But, she added, "finding that woman in 1975 who had had bloody diarrhea, and who had the same E.coli strain in her stool, that's when you start to get chills." The researchers now knew that there was a new strain of E.coli that could get into food and make people seriously ill.

Yet a fundamental question remained. How, or why, did these bacteria make people sick?

A crucial clue came from a seemingly unrelated investigation that was going on at the same time. Dr. Mohamed Karmali, who is the head of microbiology at the Hospital for Sick Children in Toronto, had been tormented over a mysterious illness. In 1980, in the space of 10 days, 14 children were admitted to his hospital, all with the same symptoms. Their illness began with diarrhea. About a week later, their kidneys failed. Two children died and others were left with severe, permanent damage to their kidneys and bowels.

Like Riley, Karmali and his colleagues asked what these children had in common. "There weren't too many leads," Karmali said. "We were looking at all sorts of poisons and chemicals." One hint, significant mainly in retrospect, was that the children had visited a farm market on the outskirts of Toronto and had drunk unpasteurized apple juice.

"I was working day and night," looking for infectious organisms that could have caused the strange illness, Karmali recalled. "Then we had sort of a break," he said. An 18-month-old boy died, and the autopsy showed that the boy had E.coli bacteria in his bowel that were making a potent, unidentified toxin.

Karmali began looking for the bacteria in the feces of children admitted later with the kidney disease, called hemolytic uremic syndrome, but could not find them, probably because the children's active diarrhea had already ended.

So Karmali searched for the toxin instead. "To our surprise, we found the toxin in the stool," he said.

Laboratory tests showed that the toxin killed monkey kidney cells and a type of human cell grown for research. That sounded familiar to Dr. Alison O'Brien, a microbiologist at the Uniformed Services University of the Health Sciences in Bethesda, Md., who heard about the research at a scientific meeting. She worked with a different bacterium, Shigella, that exudes a toxin with exactly the same properties. It could be, she reasoned, that the E.coli had picked up the gene that causes bacteria to make Shigella toxin. If so, the new E.coli was armed with "a nasty little toxin," Dr. O'Brien said -- indeed, probably the third most dangerous bacterial toxin known, coming right after tetanus and botulism.

Dr. O'Brien examined the rogue E.coli, and sure enough, discovered that it had somehow taken on the Shigella toxin gene. Moreover, in many of the E.coli, that gene was slightly altered in a way that made the bacterium produce toxin even more deadly than the original toxin made by Shigella.

No one knows exactly how the Shigella toxin genes jumped to E.coli, but Dr. O'Brien has an educated guess. Viruses that infect bacteria can sometimes pick up a gene from one bacterium and carry it to another.

But if that was the seminal event, Dr. O'Brien said, it probably did not occur in the United States, where Shigella bacteria do not have the dangerous toxin gene. Shigella in Central America do have that gene, however, and in the 1970s, that area was hit with a pandemic of Shigella dysentery. As the Shigella mixed with harmless E.coli in people's intestines, or as it mixed with harmless E.coli that inhabited animal manure, a virus may have carried Shigella toxin genes to E.coli. The result would have been a strain that had never been seen before: the toxin-armed E.coli O157:H7.

However the toxin does its work -- and scientists still do not know for sure -- the result is that it can injure cells that line blood vessels, plugging them with blood clots. When this happens, the first symptom is bloody diarrhea. But a small proportion of people, especially young children and the elderly, develop hemolytic uremic syndrome, the actual destruction of the kidneys that occurs when blood vessels in these organs are destroyed. The syndrome can lead to kidney failure or can be fatal. Infections with E.coli O157:H7 are now the leading cause of kidney failure in children, the CDC says, with at least 1,000 children a year developing kidney failure from these infections, and 3 to 5 percent dying.

Older people also tend to develop another complication, thrombotic thrombocytopenic purpura, a sort of leakage of the blood vessels that feed nerve cells. The result is an encephalitis-like disease, with psychosis, comas or seizures.

The bacteria are surprisingly tough and virulent. For example, said Dr. Marguerite Neill, chief of the division of infectious diseases at Brown University School of Medicine, most bacteria do not produce disease unless a person is exposed to millions of them. But as few as 10 or so E.coli O157:H7 can produce illness -- far too few to see or smell.

While most bacteria, including botulism bacteria, cannot live in acidic environments, E.coli O157:H7 is undaunted, able to grow in foods like unpasteurized apple cider and commercial mayonnaise.

As researchers studied the organism and its effects, E.coli O157:H7 epidemics kept ticking away, usually going unreported because they were sporadic and few doctors thought to test for E.coli. But some outbreaks were hard to miss. In September 1984, 34 out of 101 people living in a Nebraska nursing home became ill with E.coli infections. They had eaten contaminated hamburgers.

In October 1988, 32 junior high school students from a school in Minneapolis developed bloody diarrhea from eating contaminated hamburgers in the school cafeteria.

Nonetheless, said Dr. James Marsden, a professor of meat science at Kansas State University, "other than among scientists, I don't think it really sank in what a serious issue this was until the Jack in the Box outbreak," in 1993, when four children died and hundreds of people became ill.

Ridding meat of the new E.coli strain became "the biggest challenge ever to face the industry," Marsden said.

Public health officials also changed their tack. The Centers for Disease Control made a video about E.coli O157:H7 and sent it to public laboratories, suggesting they look for the bacteria. Suddenly, the E.coli were everywhere.

"We used to recognize outbreaks every couple of years," said Dr. Griffin, the epidemiologist for the Center for Disease Control. "Now we see 25 to 30 a year." And the list of foods that caused outbreaks keeps growing. "We look at that list and say, 'Wow,"' Dr. Griffin said.

But how, scientists asked, were the E.coli entering the food chain? Because the first outbreaks seemed to be associated with ground beef, scientists looked into cattle and other farm animals.

Initially, the scientists came up empty handed. In 1982, the Department of Agriculture could not find E.coli O157:H7 in any animals in the country, indicating that the infected cows that contaminated the McDonald's hamburgers must have been rarities. But now, said Dr. Michael Doyle, the director of the University of Georgia's Center for Food Safety and Quality Enhancement in Griffin, Ga., as many as 1.5 to 5 percent of dairy cows carry the organism. Dairy cows and other cattle seem to be the Typhoid Marys of the epidemics, carrying the bacteria harmlessly in their feces. From there, the bacteria enter the food supply. In modern meat plants, a single contaminated carcass can be ground up with scores of other cows to produce hamburgers.

E.coli O157:H7 has also gotten into apple cider, possibly because farmers fertilized their crops with cow manure, Doyle said. It has also gotten into lettuce and alfalfa sprouts.

Other animals can carry E.coli as well. Doyle said he helped investigate a case in Oregon where 11 family members got sick after eating venison jerky. The deer had been infected and the E.coli, he said, "survived the brine process."

He added, "There are reports of kids who went to petting zoos on farms and got sick from petting goats."

The outbreaks have pushed the government into revamping regulation of the food industry. Meat plants are starting to replace the decades-old "sniff and poke" system of inspection with high-tech microbial tests. The Clinton administration is seeking greater power to levy fines and order recalls.

The Food and Drug Administration recently approved irradiation for meat, a method that can kill E.coli. Non-irradiated hamburger, safety experts say, should be cooked until the inside reaches 155 degrees. They also advise thoroughly washing produce and avoiding raw milk and unpasteurized apple juice.

"I really credit the USDA and the FDA for new initiatives," said Dr. Michael Osterholm, an epidemiologist at the Department of Public Health in Minnesota who in summer 1996 traced an outbreak that sickened scores of children in a Twin Cities day-care center to a cow carcass one child's family had in their freezer. "But if you gave me all the money in the world, I'm not sure I could keep these bacteria out of the food supply."

That, however, is no excuse for complacency from regulators or the food industry, Dr. Griffin said.

"The bottom line should be, is it reasonable that if a consumer undercooks a hamburger that their 3-year-old child dies?" she asked. "Even if only a small number of children die each year, 50 or 100, I think it galls consumers and that's understandable. They are asking, Can't we have better control of how our food is produced?"