Why is tracing the source of cyclosporiasis outbreaks so difficult?

The foodborne illness cyclosporiasis has exploded in the United States this summer: More than 6700 cases have been confirmed and more than 11,500 potential ones have been reported to the Centers for Disease Control and Prevention (CDC) by 45 states between 1 May and 28 July. (The true toll is likely much higher because many people don’t seek medical attention.) Figuring out where all those infections came from has been a struggle. Caused by the tiny parasite Cyclospora cayetanensis, the illness commonly manifests as watery, “explosive” diarrhea. Dehydration and electrolyte imbalance can also lead to kidney damage that requires hospitalization. People catch cyclosporiasis by consuming food or liquids contaminated with the parasite through contact with human feces, which can happen anywhere along the food supply chain. The cases seen so far are not part of a single national outbreak. Instead, at least six different clusters are under investigation. The source of the contamination has been identified for the biggest one—an outbreak affecting at least 1947 people across nine states that a Food and Drug Administration (FDA) investigation has linked to shredded iceberg lettuce sourced from Mexico by Taylor Farms. (That outbreak has resulted in 98 hospitalizations; no deaths have been reported.) So far, genetic evidence to back up that finding is lacking, however, and the source of the other five clusters is still unknown. Here’s how health officials are trying to find the origins of cyclosporiasis outbreaks—and why that is often so difficult. How do researchers get to the bottom of this kind of outbreak? Foodborne disease outbreaks are investigated using three types of data that, in the U.S., are collected by local, state, and federal health agencies, including CDC and FDA. The first type is epidemiologic data, obtained by asking patients what they ate, where and when, and looking for common factors. With Cyclospora, that’s often more difficult than with bacterial infections because the incubation period tends to be longer—about 1 week on average—which makes it harder for people to remember details of their diet, says Lucy Robertson, a parasitologist at the Norwegian University of Life Sciences. Using the information from patients, investigators then examine the food supply chain to find out where contaminated products might have come from. This type of detective work led FDA to conclude that the origin of the biggest multistate outbreak was iceberg lettuce from Taylor Farms, sourced and processed in Mexico, that was served in multiple Taco Bell franchises but also sold in stores. (Taylor Farms initiated a recall of the products on 17 July.) Finally, testing samples from patients’ stool, food, and the environment can provide microbiological and genetic data to nail down the links between cases and their sources. Why is nailing down the link between cases hard when it comes to Cyclospora? Compared with bacterial food spoilers such as Salmonella, Escherichia coli, or Campylobacter, which each cause many more cases annually, Cyclospora, a protozoan parasite, hasn’t been “on the radar” as much, says Kellogg Schwab, a public health scientist at Johns Hopkins University. PulseNet, a national network that uses DNA data from pathogens to connect cases across states, only tracks bacteria-related foodborne illnesses. And unlike bacteria, Cyclospora only reproduces in humans’ small intestine; it can’t be grown in the lab. That gives researchers considerably less material to analyze. “All we can work with is the parasites that are in the person’s stool,” says Joel Barratt, an associate professor at Emory University’s School of Medicine who led the Cyclospora laboratory team of CDC’s Division of Parasitic Diseases and Malaria until his resignation last year. What’s more, whole-genome sequencing, which makes it easy to compare organisms, would be too expensive for Cyclospora because its genome is “massive,” Barratt says: some 44 million base pairs, about 10 times larger than those of Salmonella or E. coli. Instead, CDC relies on tests that target eight specific genetic markers. And finally, bacteria reproduce asexually in the human gut, creating nearly perfect clones that make it easy to identify genetic clusters. Cyclospora reproduces sexually, a process in which two parasites scramble their genes. That means even samples from the same source “are not going to be genetically identical,” Barratt says. And how about finding the potential source of contamination in food? That, too, is “tricky,” Robertson says. C. cayetanensis is related to many other parasites that can find their way into food, so false positive tests are not uncommon. FDA may have encountered this problem itself. On 19 July, the agency reported that a positive test of a lettuce batch from Taylor Farms, collected at the Mexican border and analyzed a few days earlier, was actually a false positive. The test FDA used has a specificity of almost 99%, meaning the risk of a false positive is 1%. The agency has not explained how the false positive occurred and did not respond to multiple requests for information from Science. (FDA has emphasized that the error doesn’t change its conclusions about the source of the outbreak, which is supported by “overwhelming epidemiological data.”) Negative results can also be hard to interpret. On 24 July, Mexico’s health authorities said a joint investigation by the U.S. and Mexico had “not detected the parasite” in more than 10 samples of lettuce and water collected from Taylor Farms’s facilities during an inspection. But that doesn’t mean C. cayetanensis wasn’t present at those sites, because the parasite isn’t evenly distributed across produce. When taking random samples from a processing plant for testing, “Which crates do you choose?” Robertson asks. If you’re testing water, “How much salad must you wash to get the parasites off?” And the modern food chain is extremely complex, Schwab points out. At any given plant, “tens of thousands of heads of lettuce come in a day,” he says. Pieces from one contaminated head might end up in many different bags that may be shipped to different regions for consumption. “The sheer volume of produce precludes testing everything,” he says. How have budget cuts and layoffs affected these tracing efforts? A lack of funding has historically slowed down research on parasitic diseases, Barratt says. Although the current eight-marker genetic test works pretty well, for example, his team has worked on a test with more than 50 markers that could make the analysis of clusters more accurate. But “that research hit a wall because there was no funding,” he says. Moves by President Donald Trump’s second administration have made the situation worse. Last year, contracts were not renewed for several members of Barratt’s team at CDC, which was in charge of monitoring and analyzing cyclosporiasis cases, and staff dropped from 11 people to three, he says. Even though others have probably joined to respond to the outbreak, he says the experience of “highly trained, specialized scientists that were ready to go” is hard to replace. Learning the ins and outs of Cyclospora-specific protocols can take weeks to months, even for scientists already specialized in bacterial foodborne diseases, Barratt says. Frank Yiannas, FDA’s former deputy commissioner for food policy and response, told Politico on 24 July that the government’s management of the outbreaks has been “catastrophic,” and called for an independent review. Teresa Estrada-Garcia, a biologist with Mexico’s Center for Research and Advanced Studies of the National Polytechnic Institute who recently co-authored a World Health Organization report on foodborne diarrheal diseases, says the U.S. used to have an “excellent” epidemiological surveillance system. Given how it has suffered under the current administration, she says, a situation like the current one was “overdue.”