Clinical Updates in Infectious Diseases
Supported by an unrestricted educational grant from Rhone-Poulenc Rorer
Volume IV, Issue 4 - November 1998
Diagnosis and Treatment of Serious Antimicrobial-Resistant Staphylococcus aureus Infection
Although strains of Staphylococcus aureus resistant to penicillin have caused infections for many years, isolates resistant to methicillin, oxacillin, and other ß-lactams have become predominant-primarily in the last 20 years. Strains resistant to ß-lactams and other cell-wall-active agents fall into several categories (Table 1).
Methicillin-resistant S. aureus (MRSA) strains were first described in England in 1961, shortly after methicillin became available for clinical use. They have subsequently spread throughout the world and are an important cause of nosocomial infections in many geographic areas, including the United States. Data from the National Nosocomial Infection Surveillance System reveal MRSA accounts for up to 40% of nosocomial S. aureus infections in large hospitals and 25% to 30% of such infections in smaller hospitals.
The vast majority of MRSA infections are acquired in hospitals or long-term care facilities (LTCFs). In a few cities, MRSA has been acquired in the community by intravenous drug users. Several recent reports have suggested MRSA may occasionally be transmitted in other community settings, particularly among preschool-age children, some of whom have attended daycare centers. Further studies are needed to determine if transmission of MRSA in community settings is becoming more common or is limited to a few geographic areas.
Classification of S. aureus
MRSA is of special concern because it is resistant not only to methicillin, oxacillin, and nafcillin but also to all other ß-lactams, including cephalosporins, imipenem and meropenem, and aztreonam. All MRSA strains contain a mecA gene and regulatory sequences that encode for production of a low-affinity penicillin-binding protein (PBP-2´) not present in methicillin-susceptible strains of S. aureus.
Most strains of MRSA are multidrug resistant. Resistance to erythromycin and clindamycin are very common and many strains are resistant to gentamicin, tobramycin, and ciprofloxacin. In some geographic areas, resistance to co-trimazole and rifampin are also common. Many strains are susceptible to minocycline. Recently, strains resistant to methicillin and oxacillin but susceptible to many non-b-lactam agents such as clindamycin and gentamicin have been recovered from a few patients who appear to have acquired the organism outside of healthcare settings. Such strains have been confirmed to possess the mecA gene, and will grow on oxacillin-salt screening plates. Clinicians need to be aware that these relatively "sensitive" MRSA strains can be confused with strains that do not contain the mecA gene.
Some strains of S. aureus that over-produce ß-lactamase appear resistant to oxacillin with routine disk diffusion susceptibility tests but are mecA-negative. Such strains usually have "borderline-resistant" minimum inhibitory concentrations (MICs) to oxacillin but are susceptible to nafcillin and many non-ß-lactam antibiotics. A few strains with decreased susceptibility to beta-lactam antibiotics have been shown to possess modified PBPs (but not PBP-2´).
Detection of Resistant S. aureus in the Laboratory
Many MRSA isolates are heterogeneously resistant to ß-lactams, i.e., only 1 daughter cell out of 104 to 106 cells appears phenotypically resistant when routine antimicrobial susceptibility tests are performed. For this reason, methicillin or oxacillin resistance is not detected well by some susceptibility testing systems (Table 2). The gold standard for determining if a strain of S. aureus is MRSA or not is to test the isolate for the presence of the mecA gene using polymerase chain reaction methods. However, this technique is not yet widely available in hospitals. Other reliable methods for detecting MRSA include oxacillin-salt screening plates containing 6 µg/mL of oxacillin plus 4% NaCl, broth microdilution tests with 2% NaCl, and agar dilution tests with 2% NaCl. Disk diffusion tests detect most MRSA but may not classify some heterogeneously resistant strains correctly. The above tests should be incubated at 35°C for a full 24 hrs. Most automated susceptibility testing systems reliably detect MRSA. Occasionally, methicillin-susceptible strains of S. aureus may be misclassified as MRSA by certain automated susceptibility testing systems, which may lead to inappropriate management of affected patients. Accurate detection of resistance to most non-ß-lactam antibiotics is not a problem for most clinical laboratories. S. aureus isolates with "borderline resistance" to oxacillin seldom, if ever, yield growth on oxacillin-salt screening plates, a trait that helps differentiate them from true MRSA isolates.
Epidemiology of MRSA
The prevalence of MRSA in hospitals varies considerably from one region to another and among hospitals in the same city. In some hospitals MRSA accounts for <10% of all S. aureus isolates, whereas in other facilities, they account for up to 65% of S. aureus isolates. MRSA is also reasonably common in LTCFs, especially those located near hospitals with high prevalence rates. In LTCFs affiliated with hospitals where MRSA is endemic, from 25% to 35% of residents may be culture-positive for MRSA at any given time. Although residents are often colonized, outbreaks of MRSA infections have seldom been reported in LTCFs.
The main reservoir of MRSA in hospitals is patients colonized or infected with MRSA. Although colonized patients have no signs or symptoms of infection, they can still serve as a source from which transmission may occur. Colonized personnel and contaminated environmental surfaces can also serve as reservoirs, but are not as important as affected patients. Presumably, MRSA reservoirs in LTCFs are similar to those in hospitals.
MRSA is most frequently transmitted from 1 patient to another via personnel who have not washed their hands between patients. Healthcare personnel who have persistent nasal colonization can transmit the organism to patients, especially if they develop a concomitant viral upper respiratory infection or MRSA sinusitis. Personnel with dermatitis colonized or infected with MRSA have also been responsible for outbreaks of MRSA infection. Transmission via contaminated articles or environmental surfaces may occur but this is a less important source. Airborne transmission is uncommon but may be a problem in special units housing patients with burn injuries or extensive dermatitis. Risk factors associated with acquisition of MRSA are shown in (Table 3).
Like other strains of S. aureus, the body site most commonly colonized with MRSA is the anterior nares. Other body sites that may be colonized with MRSA include open wounds, the respiratory tract, perineum, upper extremities, umbilicus (in infants), urinary tract, and axilla. Some patients remain colonized for only a few weeks and then become culture-negative without any specific therapy. However, patients with serious underlying diseases that require repeated hospitalization may remain colonized for more than 3 years.
About 40% to 60% of hospitalized patients colonized with MRSA develop an overt infection. In LTCFs, only 5% to 15% of colonized residents eventually develop an MRSA infection. The clinical criteria used to differentiate colonization from infection are the same as for other common bacterial pathogens. The most common infections caused by MRSA include surgical site infections (28%), bacteremia (21%), and other skin and soft tissue infections including omphalitis (21%) and lower respiratory tract infections (15%). Urinary tract infections, suppurative arthritis, and osteomyelitis are less common.
MRSA infections are of special concern for several reasons. These infections are associated with prolonged hospital stays and increased hospital costs, and few therapeutic options are available to treat affected patients. For example, 1 study showed that patients with serious MRSA infections stayed in the hospital an average of 12 days longer, and had average hospital costs of $5100 greater than comparable patients with methicillin-susceptible S. aureus (MSSA) infections. Fatality rates among patients with MRSA infections are not significantly higher than those observed among patients with infection caused by MSSA.
Treatment of MRSA Infections
Vancomycin continues to be the drug of choice for treating most MRSA infections caused by multidrug resistant strains. Clindamycin, co-trimoxazole, fluoroquinolones, or minocycline may be useful when patients do not have life-threatening infections caused by strains susceptible to these agents. For serious infections caused by strains that test susceptible to rifampin, adding this agent to vancomycin or a fluoroquinolone may contribute to improved outcomes. For example, a recent study found patients with S. aureus infections related to orthopedic implants had a better response rate if treated with a ß-lactam or vancomycin plus rifampin for 2 weeks, followed by 3 to 6 months of high-dose oral ciprofloxacin plus rifampin. Rifampin should not be used alone to treat MRSA infections because S. aureus strains easily develop resistance to this agent. For individuals with MRSA sinusitis, consider using a 2-drug combination that includes agents that penetrate well into nasal secretions, such as co-trimoxazole, rifampin, or minocycline. Test the infecting strain for susceptibility to these agents before committing patients to a course of therapy. Infections caused by non-MRSA isolates with borderline resistance to oxacillin may be treated with nafcillin. Treatment with vancomycin and special infection control precautions are not necessary for such patients.
Eradicating MRSA Colonization
Patients with serious underlying diseases often remain colonized in the anterior nares for a prolonged time, often several years. Such patients occasionally develop invasive MRSA infections after remaining colonized for months. For this reason, consider eradicating MRSA nasal carriage. Eradication of MRSA carriage can also eliminate the need for colonized patients to be placed in isolation when they are readmitted to acute care hospitals. Currently, the agent of choice for eradicating MRSA nasal carriage is calcium mupirocin ointment. Application of a small amount of ointment the anterior nares 2 to 3 times a day for 5 days is often effective. Longer treatment periods are not indicated. Follow-up nasal swab cultures should be performed 1 or more weeks after decolonization therapy, since nasal colonization will not be successfully eradicated in all patients. Those with wounds, ulcers or tracheostomy sites colonized by MRSA are not as likely to respond to topical intranasal therapy.
Eradicating MRSA from decubitus ulcers or other chronic wounds is difficult and there is controversy about the value of decolonization therapy in such patients. Any attempts to eradicate MRSA from such wounds using topical mupirocin should be limited to a 10-day course. Longer courses of topical mupirocin administration to such wounds has been associated with development of mupirocin-resistant strains of MRSA. Strains with both low-level and, to a lesser extent, high-level mupirocin resistance are being seen with increased frequency in the United Kingdom, where mupirocin has been used widely for several years.
Glycopeptide-Intermediate S. aureus
Most strains of S. aureus, including MRSA, are inhibited in vitro by vancomycin concentrations ranging from 0.5 to 2 µg/mL. Accordingly, it was of particular concern that in 1996, a strain of MRSA with reduced susceptibility to vancomycin (MIC=8 µg/mL) was isolated in Japan from a patient who responded poorly to vancomycin therapy. The isolate was found to have reduced susceptibility to another glycopeptide, teicoplanin, and therefore has been referred to as a glycopeptide-intermediate S. aureus (GISA). Since mid-1997, 3 patients with serious MRSA infections caused by similar GISA strains have been identified in Michigan, New Jersey, and New York and one case was reported from France.
All 4 patients had received preceding therapy with glycopeptides. Several had received repeated courses of vancomycin for many weeks before GISA was isolated from clinical specimens. In several instances, patients had infections related to indwelling vascular or peritoneal dialysis catheters that had not been promptly removed. Such infections are likely to be seen with increasing frequency, particularly among patients who receive prolonged (and sometimes inappropriate) courses of vancomycin therapy for MRSA infections.
Clinicians need to be aware that GISA strains are not detected by routine disk diffusion susceptibility tests. Broth dilution, agar dilution, and E-test methods appear to be reliable if the laboratory incubates susceptibility tests for a full 24 hrs. The Vitek system usually classifies GISA strains as having vancomycin MICs of 4 µg/mL (which is very unusual among S. mec isolates). Alternatively, isolates can be screened for reduced susceptibility to vancomycin by inoculating the organism onto agar plates containing 6 µg/mL of vancomycin. In patients with MRSA infections that appear to respond poorly to vancomycin, repeat cultures should be obtained from infected sites, and isolates should be tested using one of the methods mentioned above. Any confirmed GISA isolates should be promptly reported to the Centers for Disease Control and Prevention.
The first patient identified with a GISA infection responded to a combination of ampicillin and sulbactam plus arbekacin, an aminoglycoside available in Japan. Two patients in the United States have been treated with various regimens. To date, optimal therapy for GISA strains has not been established. Preliminary evidence from a rabbit model of infective endocarditis suggests a combination of vancomycin plus a ß-lactam may be effective against GISA infections. Investigational antimicrobials with in vitro activity against GISA strains include quinupristin-dalfopristin, linezolid, LY333328, and everninomycin.
Because GISA strains do not respond to treatment with vancomycin and are often resistant to many other antimicrobials, the Hospital Infection Control Practices Advisory Committee has issued guidelines for preventing the spread of GISA strains. Recommended measures should be implemented immediately if a GISA isolate is recovered from any patient, whether colonized or infected.
John M. Boyce, MD
Professor of Medicine
Providence, Rhode Island
Dr. Boyce currently has a research grant from Eli Lilly to study Staphylococcus aureus isolates with subpopulations that have reduced susceptibility to vancomycin. This research may also include looking at new investigational agents being developed by Eli Lilly.
- Clin Infect Dis 1997;24:S131-S135.
- JAMA 1998;279:593-598.
- J Antimicrob Chemother 1997;40:135-146.
- MMWR 1997;46:813-815.
- J Clin Microbiol 1998;36:1020-1027.
- MMWR 1997;46:626-635.
National Foundation for Infectious Diseases / 4733 Bethesda Avenue / Suite 750 / Bethesda, MD 20814 / (301) 656-0003
Published by Postgraduate Institute for Medicine / Englewood, CO / (800) 423-3576
Copyright © 1998. All rights reserved.