German officials declare E. coli O104:H4 a sproutbreak

Via H5N1, German officials are calling it for sprouts:

Germany on Friday blamed sprouts for a bacteria outbreak that has left at least 30 dead and some 3,000 ill, and cost farmers across Europe hundreds of millions in lost sales.

“It’s the sprouts,” Reinhard Burger, the president of the Robert Koch Institute, Germany’s national disease centre, told a news conference on the outbreak of enterohaemorrhagic E. coli (EHEC) in northern Germany.

“People who ate sprouts were found to be nine times more likely to have bloody diarrhoea or other signs of EHEC infection than those who did not,” he said, citing a study of more than 100 people who fell ill after dining in restaurants.

As a result, the government lifted a warning against eating raw tomatoes, lettuce and cucumbers.

There still haven’t been any positive tests, but as I mentioned yesterday, the epi seems to strongly point to sprouts. Confirmation via bacterial isolation and typing would be ideal, but I’m not holding my breath for that to happen at this late date. Larger studies also, I’m hoping, will be done–the numbers above state that they came from ~100 people, out of approximately 3,000 sickened so far, and we still don’t know how the implicated sprouts were contaminated. Did it originate in the seeds? (If so, still from where?) Was it human-to-sprout contamination from a sick worker on the farm? (If so, where again did the worker pick it up?) Still so many unanswered questions, but at least this should let some of the other farmers’ lives get back on track.

The case of the missing smoking sprouts

Maryn McKenna has a great update today on the E. coli situation, looking at where we are as far as unanswered questions about the outbreak and the strain. It’s been a messy day; more evidence seems to point to the sprout farm, but CIDRAP also notes that another contaminated cucumber was found in the compost bin of a family sickened by the bacterium (this one had the correct serotype–O104), but it’s impossible to tell at this point whether the cucumber was the source of that bacterium or it ended up there from one of the sickened family members. Twists and turns abound in this investigation. I’ve not seen any confirmation that the remaining sprout isolates tested negative yet, either.

One thing I want to emphasize and expand upon, from the CIDRAP article:

Most of the investigation findings point back to a sprout source, and microbiological testing a month after the fact doesn’t change that, Hedberg said. “Negative micro results cannot negate positive epi results. This is an important principle that we cannot state too strongly.”

At this late date, it’s hard to say whether we’ll be able to definitively trace this back to its source–too much time may have passed for there to be any remaining contaminated source material left. This means we might not ever find the “smoking gun” (or smoking sprouts, as the case may be). With such a severe outbreak–725 cases of hemolytic uremic syndrome, over a quarter of those infected–that’s bad news if we can’t confirm the vehicle, as it may make it more difficult to find the ultimate source of this strain. However, as Hedberg notes, we do still have the epi. This was used long before we had today’s molecular typing techniques, or even before we had microbiology culture ability, for that matter. Think John Snow’s cholera investigations, where he didn’t even know about bacteria and yet was able to determine the water as the vehicle for infection. So while confirmation may not happen, it’s still looking like most lines of evidence point to the implicated farm.

Maryn also brings up a great point that what we’re seeing as far as cases may be over-estimating the actual severity of the infection. I’ve talked about this previously regarding influenza infections, particularly H5N1. Right now H5N1 has a high mortality rate–but is it artificially high, because mild or asymptomatic infections are being missed?

With O104, as with any food-borne infection, surely this is happening. Mild diarrhea or stomach cramping isn’t something people frequently go to their healthcare provider over, so inevitably cases are missed. However, it probably happens with any E. coli outbreak, yet in most others we still see HUS rates between about 2-7% of the confirmed infections, while this one is at about 26%. So it doesn’t seem (to me, at least) that missed mild infections are the whole story. Is this acting like the novel Clostridium difficile strains, which have a mutation in a regulatory gene that leads them to pump out higher levels of toxin than “regular” strains? More than just genetic analysis will be needed to investigate that–some basic microbiology will also be needed. If nothing else, this outbreak has given us much research fodder over the coming years.

E. coli O104:H4 in Europe–is it new?

Mike has has a great new post up looking at some molecular analyses of the current European outbreak strain. For anyone who hasn’t been paying close attention to what’s happening across the pond, there’s an ongoing outbreak of enterohemorrhagic E. coli (EHEC)–the type of E. coli that includes O157:H7, which has been associated with outbreaks of disease associated with food. The most infamous outbreak was the 1993 Jack-in-the-Box disaster, associated with undercooked hamburgers contaminated with the organism, but there have also been outbreaks associated with contaminated vegetables (such as the 2006 outbreak due to spinach). Infections with this bug can cause serious illness, including bloody diarrhea (due to production of a protein called the Shiga toxin) and eventually can shut down the kidneys. Permanent damage can result, and even death.

In most outbreaks, children have been the most affected group, and the outbreaks tend to be fairly small (as outbreaks go–~200 people were confirmed to be infected due to spinach in 2006, though many more mild or asymptomatic cases likely went undetected). That’s reason number 1 this European outbreak is a bit odd. Adults are the largest group affected, and of those, most have been women. It’s also a huge outbreak–at least 1600 affected and 16 deaths to date. Almost a third of those–roughly 500–have been diagnosed with hemolytic uremic syndrome (HUS), one of the most serious complications of the infection. That’s a huge number, and cases don’t seem to be slowing down, as we usually see with EHEC outbreaks.

News out yesterday also includes notice that one of the outbreak strains has been sequenced:

Meanwhile, a Chinese genomics laboratory, BGI (formerly the Beijing Genomics Institute), announced today that it has sequenced the outbreak strain and completed “a preliminary analysis that shows the current infection is an entirely new super-toxic E coli strain.” The analysis was done by BGI-Shenzen in collaboration with the University Medical Centre Hamburg-Eppendorf, the BGI statement said.

The analysis confirmed that the pathogen is an E coli O104 but said it is a new serotype, “not previously involved in any E coli outbreaks,” according to BGI. The strain is 93% similar to a strain found in the Central African Republic, but it has acquired sequences that seem similar to those involved in causing “hemorrhagic colitis” and HUS, the statement said.

The statement also said the E coli strain carries genes that confer resistance to several classes of antibiotics. Earlier reports from Europe had said the strain was resistant to multiple drugs.

A WHO official agreed that the outbreak strain is new, according to the AP report. “This is a unique strain that has never been isolated from patients before,” said Hilda Kruse, a WHO food safety expert.

Earlier this week, the CDC called the outbreak strain very rare but not brand new. In today’s AP story, Dr. Robert Tauxe, a CDC foodborne disease expert, said the strain was seen in a case in Korea in the 1990s. He said the genetic fingerprints of the current strain and the Korea one may vary slightly, but not enough to call the European strain new, according to the AP.

I believe that this is the Korean paper they’re referring to, describing a case of O104:H4 infection, but it’s not from the 1990s, at least that I can tell (published in 2006, though it may be an old case). Mike is skeptical that this is a new strain as well. The wording of the article doesn’t make sense either; O104:H4 *is* the serotype, so that obviously isn’t novel, though some elements of the bacterium could be. Reports are saying that it produces more toxin than ordinary EHEC strains, and that it’s resistant to multiple antibiotics. For these infections, the former is important; the latter, not so much, as treating EHEC infections with antibiotics actually makes the infection worse. (However, E. coli can also cause other types of infections, including meningitis and septicemia, for which antibiotics would be appropriate–so it’s not completely OK that it’s multi-resistant; it just doesn’t matter as much for the diarrhea/HUS combination).

So what’s going on? Still hard to tell. We don’t yet know the vehicle for bacterial transmission. Salad ingredients–lettuce, tomatoes, and cucumbers have been implicated in case-control studies but no one has yet found this strain on vegetables. We don’t really know if the virulence in this strain is higher than other EHEC strains, or if the higher apparent levels of HUS are due to better reporting/surveillance in Europe. (I think this unlikely–it’s a pretty large difference–but still, it needs to be examined). Basically, we’re closing in on a month into this outbreak and we still know very little, and it doesn’t seem to be slowing down at a rapid pace. And, we probably haven’t even identified all the cases to date–there have now been three diagnosed in the U.S. following travel to Germany, and likely more sporadic cases in other areas that haven’t been linked back to this outbreak yet. Stay tuned; this one’s going to be in the news for awhile as we get it all figured out.

Edited to add: see also other posts on this, especially the sequencing/novelty issues, here at phylogeo, here at bacpathgenomics, here at pathogenomics, or here at genomic.org.uk.