Disease Drop Tied to Gene Network

The incidence of one of the nation’s leading intestinal diseases has plunged to its lowest levels in 13 years, and surprised federal health officials credit a new electronic network that allows them to read and compare the genetic “fingerprints” of bacteria in food-poisoning outbreaks.

The news--to be included in a food safety report that the federal Centers for Disease Control and Prevention sends to Congress next month--could not come at a more critical time. There has been a growing sense of urgency in recent years over the dangers posed by tainted food, which federal health officials estimate sickens up to 76 million Americans and kills 5,000 annually.

In the case of shigellosis--a nasty infection that produces severe, bloody diarrhea--disease detectives for the first time were able to link a cluster of food poisoning cases in Los Angeles to similar outbreaks in Minnesota, Massachusetts and Canada.

With the discovery of virtually identical genetic “fingerprints” of Shigella sonnei, U.S. officials were able to trace the source to tainted parsley in Mexico. And since shigella is often passed through dirty water, the trail ultimately led them not only to the parsley but to the water used to irrigate it.

Discovery Leads to Education

U.S. officials helped the Mexican agriculture ministry launch a course in good agricultural practices, including the importance of using clean water to wash crops, and monitored the produce coming into the United States.

As a result, the number of incidents plummeted. According to the CDC’s preliminary estimates, cases of shigellosis dropped by more than one-fourth, to an estimated 18,000 in 1999.

Health officials were startled by the drop--each of the previous three years had seen 25,000 cases--because the highly infectious shigella is passed not only through food and water but also through person-to-person contacts, particularly in day-care settings, and is regarded as extremely difficult to control.

“It’s a surprisingly big jump,” says Dr. Robert Tauxe, who heads CDC’s food-borne diseases branch.

Since the new fingerprinting network was launched more than a year ago, health officials say, they are better equipped to conduct outbreak investigations “bigger and faster, where we had not succeeded before,” Tauxe says.

In the Los Angeles shigellosis outbreak, for example, in the summer of 1998, health officials were inclined, as they had been in other such cases, to dismiss the 16 illnesses as isolated situations, confined only to those who had eaten a few days earlier in two local restaurants.

But the genetic match, which extended beyond U.S. borders, allowed health officials “to do something we had never done before”: to identify an international outbreak with a common food source, the parsley, says Bala Swaminathan, who runs CDC’s food safety lab.

It was, he says, “a defining moment” for the program.

“This improvement in technology has probably done more good in a short period of time in reducing illnesses and saving lives than any other single advance in food safety outbreak response,” says Joe Levitt, director of the Food and Drug Administration’s food safety center.

The CDC, FDA and the U.S. Department of Agriculture--working with health departments in 32 states, Los Angeles County and New York City--run the genetic fingerprinting network, known as PulseNet.

Federal officials regard it as a valuable addition to food safety sleuthing. Until recently, these investigations relied mostly on old-fashioned shoe leather--traditional epidemiology, which is effective but time-consuming. It involved painstaking interviews of food poisoning victims to uncover what they had in common with each other--what they ate, for example, and where.

When bacterial genetic fingerprinting was used at all--the actual technology has been around for more than a decade--CDC usually used the technique to confirm what was already known about outbreaks, a process that could take months because of the workload.

In those days, fingerprinting methods were tediously slow and differed among the few states using it. And because every state had its own way of typing DNA patterns, the results were inconsistent and impossible to compare.

But a two-year effort to set standards for state labs and refine typing methods--speeding up the process from four days or longer to 24 hours--catapulted the approach from a confirmatory tool to a prime part of virtually every investigation.

The Need for Quick Analyzing

Today, states move quickly to collect and analyze samples from sick people, while the CDC and USDA analyze those from suspected foods, then compare what they have. If the results match, officials know where and how to act.

Certainly the need is pressing: Food poisonings have become more complicated and lethal than ever before with the emergence of new, and sometimes deadly, organisms. And the nature of outbreaks has changed as well: No longer confined to the traditional “church supper,” in which a lot of people get sick from sharing a single dish in a single location, food poisoning outbreaks today can cross state and international lines, making them much more difficult to track.

The ability to solve them quickly often can spell the difference between life and death, particularly since 1993, when the first of a spate of frighteningly lethal food-borne outbreaks occurred.

That year, four children died and 700 people became ill in four Western states from eating undercooked meat at Jack in the Box restaurants. The meat was contaminated with Escherichia coli 0157:H7, an especially virulent strain of bacteria. Today federal health officials estimate that had PulseNet been in place, about 500 of those cases could have been prevented.

PulseNet now can provide data “within days--rather than weeks,” Levitt says. “With this, we will be able to catch food [poisoning] outbreaks before they get out of hand.”

In addition to shigella, PulseNet is building a database on three other bacteria--including the deadly E. coli 0157:H7, listeria and salmonella--and has plans to expand to 10 within two years.

But some officials worry that uneven capabilities in many states could hamper their efforts for nationwide prevention.

“This whole system is predicated on the ability of state and local health departments to get isolates [of the bacteria] quickly and investigate,” says Dr. Michael Osterholm, former state epidemiologist in Minnesota, who now runs an Internet-based information company on infectious diseases. “In principal it’s a great idea--the problem is the need for resources.”

CDC’s Swaminathan agrees that “you have to have the states typing these bacteria as quickly as possible when they come in, then transmitting it as quickly as possible to the CDC. Unless every state has one or two people dedicated only to this function, it won’t get done.”

He complains that he does not have enough scientists for his own lab and often has to pull people away from other projects to help. Without more, “I’m not sure we’ll get the full benefit of [the technology],” he says.

Only seven of the 32 participating states are directly connected to the “main” server at CDC because of problems that developed in the software. But Swaminathan says that the software has been fixed and all participating states should be hooked up by April 30.

Federal officials hope that 40 states will get on board by the end of the year.

“Their participation often depends on the sophistication of their lab support. We’re trying to work with states to get them the infrastructure they need,” such as equipment and facilities, says Robert Buchanan, a senior science advisor in the FDA’s food safety center.

Some states with small populations and limited resources rely on support from neighboring states whose labs are better equipped and can share services, Buchanan says.

PulseNet’s fiscal 1999 budget was about $7 million, of which 70% is federal dollars and the rest is state and local government funds. President Clinton’s proposed budget for 2001 calls for an expansion of PulseNet.

In California, major health facilities treating food poisoning cases send their bacterial isolates to the PulseNet lab in Berkeley or to the Los Angeles County PulseNet lab. In the year since the program was introduced, the results have been dramatic.

“It’s really allowing us to look at clusters we wouldn’t have looked at before, and it’s made our surveillance more sensitive,” says Dr. Roshan Reporter, a medical epidemiologist at the county health department’s acute communicable diseases division.

She says that the county was able to identify about a half-dozen multi-state outbreaks that “otherwise would have been dismissed as just a few cases or a coincidence. We are more likely to pick up outbreaks we might have missed, and we’re better able to connect them to cases in other health jurisdictions.”