Pan-resistant?? The rise of Acinetobacter

A set of papers published this month in two journals provide an unsettling glimpse into the rocketing incidence and complex epidemiology of one really scary pathogen, Acinetobacter baumanii.

In the all-star annuals of resistant bugs, A. baumanii is an underappreciated player. If people -- other than, you know, disease geeks -- recognize it, that is because it's become known in the past few years for its propensity to attack wounded veterans shipped to military hospitals from Iraq and Afghanistan, earning it the nickname "Iraqibacter." (Important note: Steve Silberman of Wired magazine took an early look at this phenomenon in 2007, in a great story that analyzed the epidemiology of Iraqibacter to show that military infection control, not the environment of Iraq, was to blame for the bug's rapid emergence.) A. baumanii is a nasty bug, causing not just wound infections but pneumonia, urinary tract infections, meningitis and bacteremia. Even more nasty, it collects resistance factors like baseball cards, and is commonly resistant to at least 4 antibiotic classes. The most resistant strains are susceptible only to the so-toxic-we-put-it-back-on-the-shelf-decades-ago antibiotic colistin.

This is a particular concern because A. baumanii is a Gram-negative bacterium -- and while the drug-development pipeline for Gram-positives such as MRSA has slowed practically to a trickle, the one for Gram-negatives has dripped itself dry. As the Infectious Diseases Society of America and Jerome Groopman of the New Yorker highlighted back in 2008, drugs for Gram-negatives are barely on the agenda for the few companies still conducting antibiotic development.

So, the first piece of bad news. In Infection Control and Hospital Epidemiology (ICHE), a team from Brooke Army Medical Center in San Antonio take a look at their incidence of resistant Ab and find it exploding. Between 2001 and 2008, the percentage of A. baumanii isolates that were resistant to at least 3 classes of drugs went from 4% to 55%; of all the isolates, 17% (127) were resistant to at least 4 drug classes, and one was resistant to, well, everything.

How does A. baumanii spread so fast? A second paper in ICHE suggests a reason: The bug seems to do a better job than other resistant pathogens of contaminating the gear and hands of health care workers. A study done at University of Maryland found that when health care workers took care of A. baumanii patients, they ended up with contaminated gowns and gloves 39% of the time, and with contaminated hands (after glove removal) 4.5% of the time. Those are higher rates than for MRSA (18.5% of encounters) or VRE (8.5%).

A review article in Clinical Infectious Diseases reminds us why we should care about this: It examines the drugs to which some strains of A. baumanii are still susceptible, and finds all of them significantly toxic to different organs (kidneys, liver, pancreas, red blood cells, ) at the doses necessary to wipe out the bug.

Which is all troubling by itself. But a paper and editorial also appearing in Clinical Infectious Diseases make the case for A. baumanii as a bigger threat than has been understood. The bug's recent epidemiology has shown a distinct split, between the highly resistant forms affecting veterans, most of them being treated in the military evacuation chain, and less-resistant forms affecting civilians in hospitals (including in the Brooks data in the paper above). The severe wounds, aggressive treatment and rapid multiple transfers of personnel in the military system inadvertently created an environment that not only put A. baumanii under great selective pressure, but also spread it with startling efficiency.

The paper, reporting data from 4 community hospitals near Detroit, shows that the civilian medical system -- that would be the one that most of us live in -- has duplicated that churning as well. Between 2003 and 2008, all A. baumanii in their network increased 25%. A. baumanii resistant to the first 2 front-line drugs went from 2% to 33% of isolates. And "pan-resistant" A. baumanii -- resistant to all 8 drugs available for it, an essentially untreatable strain -- went from nonexistent to 14% of all the isolates that network found.

The effect on the patients was dramatic, of course: The more resistant their strains were, the more likely they were to never go home from the hospital, but (if they did not die there) to be discharged instead to a nursing home, long-term acute care facility, or hospice. But the larger point is that they carried that multiply-resistant strain with them, distributing it throughout the region: Patients came to those 4 hospitals, carrying A. baumanii, from 17 different nursing homes; from the 4 hospitals, carrying A. baumanii, they were transferred out to 28 different nursing homes.

This is a smart analysis, and devastating in its implications. American hospitals do a debatable job right now of handling infection control -- but overwhelmingly, they are handling infection control as individual institutions, not as competitors in a local market, and certainly not as members of a geographic region. Yet this data demonstrates clearly that cooperation between hospitals and other healthcare institutions -- most of which don't have hospitals' infection-control budgets or personnel -- is going to be essential if we want to put the brakes on Acinetobacter before it soars in the civilian medical system in the same way it did in the military one.