The general public and all physicians are aware of the emergence of so-called superbugs, bacteria resistant to most currently available antibiotics. Many EPs consider this a scientific curiosity, trivia for the infectious disease department, and of no great consequence to their daily practice. After all, the ED has at its disposal an arsenal of broad-spectrum antibiotics designed to kill or initially contain almost all bacterial infections.
It would be naïve, however, to think that antibiotic resistance does not pertain to real-life emergency medicine practice. This month's column discusses a potentially gargantuan development in antibiotic resistance, specifically very resistant bacteria that produce enzymes that will inactivate our most powerful antibiotics. These enzymes are termed extended spectrum beta-lactamases (ESBLs).
Although ESBL-producing bacteria have been around since the late 1980s, most EPs have never heard the term or deemed it relevant to their daily ED antibiotic prescribing. I recently attended an emergency medicine conference, and only a handful of clinicians admitted ever hearing about ESBLs. Clearly, ESBL organisms now have relevance to your initial choice of antibiotics in the ED. The laboratory can often but not completely identify these horrific deadly infections a few days into the culture, but ESBL infections remain enigmatic to many front-line clinicians who do not subsequently read culture results of admitted patients.
These are scary bugs. ESBL infections make MRSA infections appear rather tame and almost benign. Previously a worrisome conundrum, MRSA infections have been sorted out, and are apparently now understood and easily treated. If you don't know about ESBL infections, it's time you incorporated this knowledge into your daily practice. These resistant organisms can be deadly, community-acquired, and quickly kill seemingly healthy individuals. ESBL organisms are no longer simply relegated to academic ramblings on ID rounds, and they are no longer just denizens of the ICU, peril to septic nursing home patients, or lethal to the immunocompromised.
The New β-Lactamases
Jacoby J, Munoz-Prince, LS
N Engl J Med
Every physician prescribes beta-lactam antibiotics on every shift and rarely considers them potentially ineffective for their therapeutic intentions. ESBL-producing bacteria, however, readily inactivate many of the once powerful and previously omniscient antibiotics intended to kill them. This complex review describes the pathophysiology of ESBL organisms, and puts laboratory and treatment options into prospective. Another excellent reference on this topic is a Pharmacotherapy article from more than 10 years ago that describes a variety of similar enzymes, but for this discussion, they will all be called extended spectrum β-Lactamases. (2001;21:920.)
Gram-negative bacteria elaborate a number of complex β-Lactamases as their major defense against antibiotics that contain a beta-lactam ring. β-Lactamases are commonly produced by bacteria from the Enterobacteriaceae family, affording these organisms high resistance to penicillins but relative low resistance to first-generation cephalosporins. ESBLs hydrolyze even the most potent third-generation cephalosporins and aztreonam via mutations, likely from overuse of broad-spectrum antibiotics. ESBL enzymes are plasmid-related, and therefore easily transferred among bacteria. Importantly, ESBLs also are resistant to aminoglycosides, fluoroquinolones, tetracycline, chloramphenicol, and sulfamethoxazole-trimethoprim. These are truly multiresistant organisms.
Many types of β-lactamase enzymes exist, the most familiar is the well known penicillinase that has been around for a long time, and has rendered penicillins rather tepid against most serious infections. Additives, such as clavulanic acid matched with pencillins, such as ampicillin-clavulanate, can inhibit some β-lactamases, but even this manipulation is now less effective. Importantly, ESBL production has evolved with the use of β-lactam antibiotics. Even Haemophilus influenzae and Neisseria gonorrhoeae are pathogens that now can produce β-lactamase.
The prevalence of ESBL-producing organisms varies greatly throughout the world. U.S. clinicians are most concerned with ESBL-producing Klebsiella pneumoniae and E. coli, but ESBL production has been identified in numerous other organisms, including Enterobacter, Pseudomonas, Proteus, Salmonella, Morganella, and Serratia species. The percentage of organisms expressing ESBL can be as high as 45 percent in Latin America, with the lowest rate found in the United States, about seven percent of strains.
Laboratory detection of ESBLs is complicated because of the heterogeneity of the enzymes and other cofounding laboratory peccadilloes. Many organisms deemed sensitive to antibiotics by standard lab testing are actually resistant in the body due to unidentified ESBLs. Simply stated, laboratories vary in their success in identifying ESBLs, but most laboratories can give the clinician an idea of the presence or absence of ESBL in a culture. Likely the technology has evolved since this 2005 article, but it is not yet perfect.
Previously, exotic ESBL infections were related to prolonged hospital stays, prior antibiotic use, nursing home patients, ICU admissions, the use of central catheters, ventilator assistance, hemodialysis, or post-surgical patients. Now many ESBL infections are community-acquired by healthy individuals. Infections with ESBL organisms prognosticate a higher mortality than expected from non-ESBL infections.
ESBL organisms are resistant to commonly used broad-spectrum antibiotics, including even late-generation cephalosporins, all pencillins, fluoroquinolones, and aminoglycosides. Fortunately, resistance to carbapenems (ertapenem, imipenem, and meropenem), and the cephamycins (cefoxitin and cefotetan) remains minimal. Resistance to the cephamycins can occur during treatment, however, limiting their ultimate use. Carbapenems remain the most effective and reliable antimicrobials against ESBL isolates.
Antibiotic restrictions, barrier protection, and hand washing regimens have decreased the incidence of outbreak infections, but it is feared that the increased use of carbapenems will lead to emergence of carbapenem-resistant bacteria, leaving no effective treatment. ESBLs resistance to carbapenems is rare so far but increasing. Unfortunately, there are no β-lactam antibiotics in end-stage development that can readily destroy organisms producing ESBLs.
Comment: ESBL-producing organisms are high on the list of potential infectious disease disasters, and bacteria-producing ESBLs are truly multiresistant. Most EPs, when faced with sepsis of unknown etiology, prescribe multiple antibiotics to cast a wide net. We use such combinations as vancomycin, a third-generation cephalosporin, or ampicillin-tazobactam (Zosyn), and a quinolone, and expect a cure. If your patient is unfortunate enough to be infected with an ESBL organism causing severe urosepsis, septic shock, or another life-threatening systemic bacterial infection, even these antibiotics in combination will be of minimal to no value.
This is not a theoretical issue or a vague concept to be pondered during a boring drug company lunch anymore. A significantly higher proportion of patients died following bacteremia caused by ESBL-producing E. coli (60%) compared with non-ESBL producing E. coli (23%) in a recent report. (J Infect 2007;55:254.) This is a rather amazing revelation. Unfortunately, E. coli and Klebsiella are the cause of some of the most frequent gram-negative infections, especially in the ICU. Mortality rates are close to 30 percent with bacteremia from these organisms. My hospital has a 10–30 percent incidence of ESBL-producing E. coli and Klebsiella organisms, a rather scary thought.
Someone with pyelonephritis may not be cured even with high doses of cephalosporins or quinolones, and even the once-powerful aztreonam has little effect against ESBL organisms. Emergency physicians must essentially guess about the best antibiotic combination in their sickest patients, but definitely should consider adding a carbapenem to the initial regimen while cultures are ferreting out the specifics. I am sure that many have never prescribed meropenem, imipenem, or ertapenem in the ED, even for suspected E. coli or Klebsiella infections, but it may be time to rethink their empiric use in the critically ill patient.
No randomized controlled trials of specific therapies have been done for ESBL infections. But it is no secret that infections with ESBL-producing organisms have been associated with poor outcome. Community-acquired ESBL bacteria are on the rise, and because of difficulty in laboratory identification, the actual prevalence is likely underestimated, leaving us with case reports or random outbreaks.
The appropriate antibiotic choice is essential in all cases of severe sepsis, and failure to choose an antimicrobial effective against ESBL-producing organisms will be associated with a lack of clinical response and increased mortality. One classic report referenced in the literature is an outbreak of 85 patients with ESBL-producing Klebsiella pneumoniae infections. (Clin Infect Dis 2004;39:31.) Treatment with imipenem, ertapenem, or meropenem has produced the best outcomes with ESBL infections. Although resistance has been reported for even these antibiotics, they still remain the drugs of choice.
Even more concerning, resistance may develop during initially effective therapy. ESBL infections are most commonly seen in urinary or biliary tract infections, and usually are the initial antibiotics chosen in the ED. Relying solely on the once-invincible cefotaxime, ceftazidime, ceftriaxone, or cefepime is likely to result in failure if an ESBL infection is extant. Perhaps higher doses and higher MICs would be effective, and there is some leeway for urinary tract infection where antibiotics concentrate. Standard doses of levofloxacin, ceftriaxone, or cefepime for severe urosepsis, however, will likely not suffice if the infection is due to E. coli-producing ESBL.
The microbiology laboratory at my hospital reports the presence or absence of ESBL for cultures positive for E. coli or Klebsiella. But even some local university hospitals still do not test for ESBLs. The lab is rather proficient at making this call, but results usually take two days, and infected individuals circling drain with severe sepsis simply don't have two days to make do with ineffective antibiotics. One might be guided by prior records confirming ESBL production in a previous infection because patients tend to be carriers of these organisms, just like they are with vancomycin-resistant enterococcus (VRE). Your hospital antibiogram may be helpful and guide your therapy. One problem, however, is that antibiograms do not always define their results with respect to severe sepsis and bacteremia so a rather innocuous ESBL may be found in a decubitus ulcer or other contaminated specimen, rather than a more clinically important blood or urine culture.
Laboratories face daunting challenges when charged with finding and reporting ESBLs. Due to MIC perturbations and other complex testing reasons, some ESBL isolates can appear falsely susceptible to a third-generation cephalosporin in vitro. Most labs assess for ESBL activity by evaluating resistance only to ceftazidime, but traditional susceptibility testing using the breakpoint to the MIC of this antibiotic may not always be an accurate reflection of ESBL activity. That may be acceptable because of the high MICs produced in the urine, but if the infection is in the blood, laboratory shortcomings are problematic.
Complicating matters further, not all cephalosporins are used in sensitivity testing. If a lab tests only for sensitivity to ceftriaxone and cefotaxime, poor indicators of ESBL, the clinician can be flummoxed even by a pristinely performed sensitivity test. It's very confusing to me, but apparently a lab test that indicates sensitivity to ceftriaxone does not mean that the bug is sensitive to ceftriaxone in vitro. A recent proficiency testing project found that 58 percent of laboratories failed to detect and report ESBL isolates correctly. It is truly discouraging that sensitivity testing is not 100 percent.
The message is clear for emergency physicians. It is prudent to treat critically ill patients with infections of the biliary tract or urinary tract with a carbapenem, especially those in septic shock where Klebsiella and E. coli are common culprits. When was the last time you prescribed meropenem or ertapenem for a patient's pyelonephritis? Your next 30-year-old with severe urosepsis or cholangitis, even in the absence of immunocompromise or other extenuating circumstances, may deserve at least one dose of a carbapenem in the ED. Meropenem seems to be the most active carbapenem against ESBL-producing organisms with MICs lower than those of imipenem. Ertapenem shows promise, but it lacks Pseudomonas coverage.
Click and Connect!Access the links in EMN by reading this issue on our website or in our iPad app, both available onwww.EM-News.com.
Infectious Disease Advice on ESBLs
By David A. Talan, MD
Thank you, Jim, for bringing attention to an urgent bacterial resistance problem. Let me add a few observations to your outstanding summary.
ESBLs are being seen in ED patients, not only those with recent hospitalization and long-term care but now occasionally in otherwise healthy people, most commonly with UTI, pyelonephritis, or urosepsis. Because of new micro testing guidelines, fewer labs will identify ESBLs specifically, but rather just report in vitro resistance to cefotaxime or ceftriaxone, which doctors should consider as a proxy for ESBL-mediated resistance, particularly if there is also carbapenem susceptibility.
Stay in close contact with the micro lab and infectious disease consultants at your hospital to understand the frequency of ESBL infections and risk populations. Some places see none, some see them rarely and only in at-risk groups, and some see them more generally. Depending on your epidemiology, empirical treatment protocols may need to be adjusted.
Use a parenteral carbapenem (ertapenem, meropenem, imipenem, or doripenem) for ESBL empirical or directed treatment in the case of serious infections. Use oral fosfomycin to treat cystitis (as a single 3 g dose or every two days for extended regimens, up to 21 days, for complicated infections, such as in men).
Dr. Talanis board certified in emergency medicine and infectious disease, and is presently studying ESBL infections through his CDC ED-based network, EMERGEncy ID NET (http://emergencyidnet.org).© 2012 Lippincott Williams & Wilkins, Inc.