No matter what they tried, the staff at Queen Elizabeth Hospital Birmingham in England could not quell an outbreak of a multidrug-resistant pathogen. After an incredible 40 weeks infection control specialists worried that if they could not get on top of the situation, Acinetobacter baumannii could become a permanent hospital resident lying in wait for any patient, the way multidrug-resistant Clostridium difficile does in many hospitals.
Another 40 weeks passed before the staff broke the stranglehold, declaring the hospital free of a pathogen considered among the most difficult to control. A. baumannii can lead to such infectious complications as bacteremia, pneumonia and meningitis. Bringing it to heel required bird-dogging the bacteria's movements from patient to patient, room to room and even to specific pieces of equipment while noting the tiny changes in the bacterium’s genome. The same technique is now catching on in hospitals worldwide for a variety of multidrug-resistant bugs.
Details about how the techniques finally cracked the Queen Elizabeth case, published recently in the journal Genome Medicine, suggest how theU.S. Centers for Disease Control and Prevention will employ whole genome sequencing to stop the spread of multidrug-resistant infections and even food-borne bacterial incidents occurring across wide regions.
The Queen Elizabeth outbreak, which began in 2011, at first confounded infection control. "We would feel we were more or less on top of it and then there would be some further ongoing transmission," says Beryl Oppenheim, who leads the hospital's infection control efforts. "In some cases it was obvious what the route of transmission was: the patients were in close proximity on a ward. But in other cases it wasn't clear-cut. It means you're trying to focus your attention on everything."
Oppenheim sent a sample of the bug to the U.K.’s national reference library, which characterized the bacteria and assigned it a number. This one was numbered 27, meaning it was the 27th time Queen Elizabeth Hospital had sent a new strain of A. baumannii to the laboratory. To solve No. 27, Oppenheim turned to Mark Pallen, a medical microbiologist at the University of Warwick Medical School. Pallen knew about Acinetobacter. "We had kind of broken the ground in that area," he says.
Whole genome sequencing, the ability to rapidly and inexpensively read the genetic code of an organism right down to the last A, T, G or C (for adenine, thymine, guanine and cytosine—the nucleotides that make up the genetic ribbon), had allowed Pallen's group to find genetic alterations as small as a single nucleotide. By sequencing the bacteria of each patient, Pallen could map these changes onto a sort of bacterial family tree. That family tree allowed him to trace just how the bacteria had jumped around the hospital by looking at how closely related each bacterial sample was to the next.
An earlier hospital study by Pallen had looked back at a previous outbreak at another hospital. The Queen Elizabeth case was a chance to track the culprit in real time. First, Pallen looked back at the bacteria's course through the hospital before his team was brought in. Bacterial samples “came from patients who were on the same ward at the same time. Although we don't know exactly how it was transmitted, we felt kind of comfortable that we knew that's how the outbreak happened," Pallen says. Then things started looking a bit odd. “As the outbreak progressed,” he explains, “it became clear there were patients popping outside the ward where it started."
Although hospital staff followed infection-control protocols, Oppenheim says the nature of Acinetobacter makes it easy to miss. "It isn't easy to screen people and be certain who was carrying it and who wasn't. There isn't just one test," she says. Swabbing a bit of skin for bacteria and getting a bacteria-free sample didn't guarantee that another patch of skin wasn't home to Acinetobacter. As Pallen's whole-genome sequencing results came in, transmission patterns emerged. Unlikely kinships existed among bacteria infecting patients who had no contact with one another. Where was the mixing zone in which this transmission occurred?
Pairing the hospital's electronic medical records with the sequence data provided a clue, Oppenheim says. The records system stored not just medication history and laboratory results, it also catalogued which rooms patients had occupied, the operating theaters in which they were treated and what special equipment was used on them. An informatics team took the patients records (with the names removed), overlaid Pallen's whole-genome sequence information and looked for where infected patients had crossed paths. The answer soon emerged: It was an operating theater often used to debride wounds of burn victims. Tests in the operating theater confirmed it
Of course the operating theater was routinely cleaned but Acinetobacter “has an amazing propensity for sticking to surfaces," Oppenheim says. It can form what's called a biofilm on just about anything. Biofilms, once thought to be only possible in wet environments (think: dental plaque), are bacterial communities encased in their excreted polymers. Some reports say biofilms are 100 to 1,500 times more resistant to biocides than bacteria not in a biofilm. Researchers think biofilms may play a role in many hospital-acquired infections. For instance, the risk of acquiring an infection in a hospital rises 73 percent if a patient sleeps in a bed previously occupied by someone with Acinetobacter, methicillin-resistant staphylococcus aureus (MRSA) or C. difficile—all of which can form biofilms.
The hospital staff shut down the operating room and subjected it to methodical cleaning, extending into ventilation systems. Several weeks passed without a new infection. But Oppenheim wasn't ready to declare victory. The outbreak was generally among burn, accident and plastic surgery patients. "These are very, very long-stay patients, so you can never relax completely," she says. Her caution proved necessary. Before long, another outbreak emerged. This time, Pallen's whole-genome sequencing and informatics analysis pinpointed the transmissions to a special bed used by burn patients. From there, the bacteria again colonized the operating theater. This time the staff enhanced cleaning of both the bed and the operating theater. Several weeks later the hospital was finally able to declare the outbreak resolved. It had lasted 80 weeks—about one and a half years.
Oppenheim believes informatics and whole-genome sequencing will be essential to the control of hospital infectious outbreaks everywhere. "I think this is really going to be the future. This will give infection prevention teams extra tools to work with."
Pallen and Oppenheim were part of another effort that showed the efficacy of the method, this time finding the source of a Pseudomonas aeruginosa outbreak when the Queen Elizabeth burn ward was new. The source proved to be a biofilm in the plumbing on a valve that mixes hot and cold water to regulate water temperature and prevent scalding. "And of course that is the exact temperature that many bacteria like, providing them with an ideal place to grow these biofilms," Oppenheim says.
Successes such as these have helped make whole-genome sequencing a central component in a new strategy to combat antibiotic resistant bacteria in the U.S. In September the White House Council of Advisors on Science and Technology issued a report that called for whole genome sequencing to detect where infections arise and how they spread. To scale up such an approach, the report called for the establishment of a reference collection of genome sequences from antibiotic-resistant bacteria, the establishment of a national laboratory network for pathogen surveillance, funding to enhance computational methods and the maintenance of a publically available genomic information database by the National Institutes of Health. The report was released in conjunction with the White House’s National Strategy for Combating Antibiotic-Resistant Bacteria, including an executive order directing federal agencies to increase efforts to combat antibiotic resistance, and a $20-million prize for the development of a rapid test to identify resistant infections.
The ability to track pathogens with sensitive genome sequencing is playing a growing role in infection control on the broadest scale, says Stephan Monroe, actingdirector of the CDC’s Office of Advanced Molecular Detection. "Whole genome sequencing is sort of the ultimate in genetic discrimination," he says. Previous tools could not detect the nucleotide changes that revealed pathogen family trees. "Now we can link things together that may have occurred in different parts of the country," he adds, a big help in tracking food-borne illnesses, for example.
Indeed, one of the CDC's goals, in collaboration with the U.S. Food and Drug Administration, Monroe says, is to create a panel of well-characterized bacterial strains for which the resistance patterns to many antibiotics is known—and to perform whole genome sequencing to create a reference set for studying resistance or developing new antibiotics. The CDC is working on this with state public health departments and has already made some significant progress, particularly around listeria," he notes. "We know there are already a number of state public health departments with the capacity to do whole genome sequencing, and we're working with those states," he says. "It's likely that this technology will end up in major hospitals as well."