Ampicillin-resistant enterococcal species in an acute-care hospital

Antimicrobial resistance of enterococci in Turkey

Susceptibility to ampicillin, penicillin, vancomycin and teicoplanin, high-level resistance to aminoglycosides (gentamicin and streptomycin) and beta-lactamase production were investigated among 264 consecutive clinical enterococcal isolates in Turkey. Disc diffusion test was used to detect resistance to ampicillin, penicillin, vancomycin and teicoplanin. High-level resistance to aminoglycosides was determined both by standard agar screening and by disc diffusion methods. The values of minimum inhibitory concentration (MIC) of each isolate for ampicillin, vancomycin and teicoplanin were determined by the microbroth dilution technique. The isolates were found to consist of Enterococcus faecalis (78%), Enterococcus faecium (9%) and Enterococcus spp. (12%). In all strains, the penicillin and ampicillin resistance ratios were 27% and 26%, respectively. Enterococcus faecalis was more susceptible to penicillin and ampicillin than the other strains. None of the strains were resistant to glycopeptides. High-level aminoglycoside resistance was found in 16% E. faecalis and 88% E. faecium for gentamicin, and 35% and 44%, respectively, for streptomycin. There were no differences between the two methods used to determine the aminoglycoside resistance rates in the enterococcal isolates. No beta-lactamase-producing isolates were detected in either species. In conclusion, to determine the resistance of enterococci to the penicillin group of drugs by the disc diffusion method, both penicillin and ampicillin discs should be evaluated. In serious enterococcal infections, before starting combined therapy, high-level aminoglycoside resistance should be investigated.

In vitro assessment of urinary isolates of ampicillin-resistant enterococci

OBJECTIVE

Susceptibility and minimum inhibitory concentration (MIC) studies of ampicillin-resistant enterococci (ARE) were performed with vancomycin, ciprofloxacin, and trovafloxacin. Ampicillin MICs were determined to make comparisons with achievable urinary concentrations of ampicillin.

DESIGN

From July 1998 to April 1999, all enterococci isolated from urinary specimens were tested for susceptibility to ampicillin by disk diffusion. For all ARE, vancomycin, ciprofloxacin, and trovafloxacin susceptibilities were determined by use of either disk diffusion or the E-test. Ampicillin MICs were determined for these isolates by liquid agar microdilution testing. ARE were identified to the species level on the basis of biochemical reactions.

SETTING

The study was performed at a university-affiliated tertiary care hospital.

OUTCOME MEASURES

In vitro susceptibility studies and MIC determinations were performed in accordance with the National Committee for Clinical Laboratory Standards.

RESULTS

A total of 310 urine samples were culture positive for enterococcus. Thirty (9.7%) unduplicated isolates were resistant to ampicillin. Of these, nine ARE isolates (30%) were also vancomycin resistant, whereas only 2 ampicillin-susceptible isolates were vancomycin resistant (p < 0.05). All ARE were resistant to ciprofloxacin, and 29 (96.7%) were resistant to trovafloxacin. Nine (30%), 18 (60%), and 3 (10%) isolates had an ampicillin MIC of 128, 256, and 512 micrograms/mL, respectively. Ampicillin MICs did not differ significantly between vancomycin-susceptible and -resistant isolates (p = 0.963). Twenty-seven isolates (90%) were identified as Enterococcus faecium; the other 3 were either Enterococcus avium or Enterococcus raffinosus.

CONCLUSIONS

Ampicillin resistance is associated with resistance to vancomycin. Most ARE are resistant to fluoroquinolone antibiotics such as ciprofloxacin and trovafloxacin. Ampicillin MICs for ARE found in these urinary specimens were all within 1 dilution of 256 micrograms/mL, a concentration achievable in the urine with higher doses of oral amoxicillin or intravenous ampicillin. Additional studies are needed to assess the clinical implications of these data.

Clinical prevalence, antimicrobial susceptibility, and geographic resistance patterns of enterococci: results from the SENTRY Antimicrobial Surveillance Program, 1997-1999

As part of the SENTRY Antimicrobial Resistance Surveillance Program, a total of 4998 strains of enterococci isolated from 1997 to 1999 were processed. The occurrence of enterococcal infections by species and site of infection was analyzed, as were the occurrence of vancomycin-resistant enterococci (VRE) and their resistance phenotypes and genotypes. Trends in antimicrobial susceptibility to a variety of agents (including experimental compounds) were also reported. Enterococci accounted for >9% of isolates from all bloodstream infections (BSIs) in North America. Ampicillin was active against strains from Latin America and Europe but not against those from the United States and Canada. US isolates were considerably more resistant to vancomycin (17% resistant strains in 1999) than were those from patients in the rest of the world. The highest proportion of VRE was observed among BSI isolates (81.7%). Quinupristin-dalfopristin, chloramphenicol, and doxycycline were the most active agents tested against VRE. The results of this study confirm the worldwide trend in increasing occurrence of enterococci and the emerging pattern of antimicrobial resistance among such isolates.

[Problems with gram-positive bacteria: resistance in staphylococci, enterococci, and pneumococci to antimicrobial drugs]

The resistance in staphylococci, enterococci, and pneumococci is reviewed. The author also recalls the first cases, and presents an overview of the distribution of cases in the world, the genetic and molecular mechanisms of resistance, the importance in Brazil and therapeutic alternatives. The factors that contribute to the dissemination of these problem bacteria and the measures for their control are emphasized.

Activity of HMR 3647 compared to those of six compounds against 235 strains of Enterococcus faecalis

Agar dilution was used to test the activities of HMR 3647, erythromycin A, azithromycin, clarithromycin, roxithromycin, clindamycin, and quinupristin-dalfopristin against 235 strains of Enterococcus faecalis. HMR 3647 was the most active compound (MICs at which 50 and 90% of the isolates are inhibited [MIC50 and MIC90, respectively] of 0.06 and 4.0 microg/ml, respectively). The MIC50 and MIC90 (with the MIC50 given first and the MIC90 given second; both in micrograms per milliliter) for other compounds were as follows: 4.0 and >32.0 for erythromycin A, 16.0 and >32.0 for azithromycin, 2.0 and >32 for clarithromycin, 32.0 and >32.0 for roxithromycin, 32.0 and >32.0 for clindamycin, and 8.0 and 16.0 for quinupristin-dalfopristin. All compounds were only bacteriostatic.

Activities and time-kill studies of selected penicillins, beta-lactamase inhibitor combinations, and glycopeptides against Enterococcus faecalis

The activities of piperacillin, piperacillin-tazobactam, ticarcillin, ticarcillin-clavulanate, ampicillin, ampicillin-sulbactam, vancomycin, and teicoplanin were tested against 212 Enterococcus faecalis strains (9 beta-lactamase producers) by standard agar dilution MIC testing (10[4] CFU/spot). The MICs at which 50 and 90% of the isolates were inhibited (MIC50s and MIC90s, respectively) were as follows (microg/ml): piperacillin, 4 and 8; piperacillin-tazobactam, 4 and 8; ticarcillin, 64 and 128; ticarcillin-clavulanate, 64 and 128; ampicillin, 2 and 2; ampicillin-sulbactam, 1 and 2; vancomycin, 1 and 4; and teicoplanin, 0.5 and 1. Agar dilution MIC testing of the nine beta-lactamase-positive strains with an inoculum of 10(6) CFU/spot revealed higher beta-lactam MICs (piperacillin, 64 to >256 microg/ml; ticarcillin, 128 to >256 microg/ml; and ampicillin, 16 to 128 microg/ml); however, MICs with the addition of inhibitors were similar to those obtained with the lower inoculum. Time-kill studies of 15 strains showed that piperacillin-tazobactam was bactericidal (99.9% killing) for 14 strains after 24 h at four times the MIC, with 90% killing of all 15 strains at two times the MIC. After 12 and 6 h, 90% killing of 14 and 13 strains, respectively, was found at two times the MIC. Ampicillin gave 99.9% killing of 14 beta-lactamase-negative strains after 24 h at eight times the MIC, with 90% killing of all 15 strains at two times the MIC. After 12 and 6 h, 90% killing of 14 and 13 strains, respectively, was found at two times the MIC. Killing by ticarcillin-clavulanate was slower than that observed for piperacillin-tazobactam, relative to the MIC. For the one beta-lactamase-producing strain tested by time-kill analysis with a higher inoculum, addition of the three inhibitors (including sulbactam) to each of the beta-lactams resulted in bactericidal activity at 24 h at two times the MIC. For an enzyme-negative strain, addition of inhibitors did not influence kinetics. Kinetics of vancomycin and teicoplanin were significantly slower than those of the beta-lactams, with bactericidal activity against 6 strains after 24 h at eight times the MIC, with 90% killing of 12 and 14 strains, respectively, at four times the MIC. Slower-kill kinetics by both glycopeptides were observed at earlier periods.

Reevaluation of contemporary laboratory methods for detection of antimicrobial resistance among enterococci

OBJECTIVE: To evaluate broth microdilution, disk diffusion, Etest and Vitek Systems for susceptibility testing of enterococci. METHODS: Susceptibility testing of a panel of 149 enterococci (99 vancomycin-resistant) strains, using the study methods, was performed and the results compared. RESULTS: For vancomycin susceptibility testing, categorical agreement of disk diffusion, Etest and Vitek with the reference broth microdilution test was > 95%. For aminoglycoside and ampicillin testing, categorical agreement between Etest and Vitek was 98 to 100%. CONCLUSIONS: Disk diffusion, Etest and Vitek have acceptable performance for detection of vancomycin resistance of Van A and Van B phenotypes among enterococci.

Comparison of conventional dosing versus continuous-infusion vancomycin therapy for patients with suspected or documented gram-positive infections

Ten patients were treated with conventional dosing (CD) and continuous-infusion (CI) vancomycin therapy in this prospective, randomized, crossover study. Patients were randomized to receive either CD or CI therapy for 2 consecutive days and then crossed over to receive the opposite regimen for 2 days. CD therapy consisted of 1 g of vancomycin every 12 h. CI therapy consisted of a 500-mg loading dose followed by 2 g infused over 24 h. Ten serum samples were obtained on the second day of each therapy for pharmacokinetic and pharmacodynamic analyses. Two clinical isolates of Staphylococcus aureus, one methicillin sensitive (MSSA 1199) and one methicillin resistant (MRSA 494), were chosen for pharmacodynamic evaluation of both regimens. The patient demographics (means +/- standard deviations [SD]) were as follows: sex, six males, four females; age, 36 +/- 11 years; and serum creatinine, 0.72 +/- 0.18 mg/dl. Mean pharmacokinetic parameters +/- SD for CD therapy were as follows: elimination rate constant, 0.16 +/- 0.07 h-1; half-life, 5.6 +/- 3.5 h; volume of distribution, 33.7 +/- 25 liters, 0.5 +/- 0.2 liters/kg; maximum concentration in serum, 53.4 +/- 19.3 micrograms/ml; and minimum concentration, 8.4 +/- 5.9 micrograms/ml. The steady-state concentration for CI was 20.2 +/- 11.1 micrograms/ml. Overall, both regimens resulted in the MIC being exceeded 100% of the time. The mean CD trough serum bactericidal titer (SBT) was 1:8, and the average CI SBTs were 1:16 for both isolates. Even though there was no statistically significant difference between CD trough and CI SBTs, the CI SBTs remained > 1:8 for 100% of the time versus 60% of the time for CD therapy. During CI therapy, 20 and 40% of the patients maintained SBTs of > 1:32 throughout the dosing interval for MSSA 1199 and MRSA 494, respectively. During CD therapy, however, only 10% of patients maintained SBTs of > 1:32 during the entire dosing interval for both isolates. The mean areas under the bactericidal titer-time curve (AUBC24s) +/- SD for MSSA 1199 were 528 +/- 263 for CD therapy and 547 +/- 390 for CI therapy. The mean AUBC24s +/- SD against MRSA 494 were 531 +/- 247 for CD and 548 +/- 293 for CI therapy. Similar to the AUBC24, the mean area under the concentration-time curve for a 24-h dosing interval divided by the MIC (AUC/MIC24) ratios +/- SD were 550.0 +/- 265.7 for CD and 552.6 +/- 373.4 for CI therapy, respectively. No statistically significant differences were found between any of the pharmacodynamic parameters for CD and CI therapy. In addition, no adverse effects with either CD or CI therapy were observed during the study. We conclude that CI and CD vancomycin therapy demonstrated equivalent pharmacodynamic activities. Although CI therapy was more likely to result in SBTs that remained above 1:8 for the entire regimen, the clinical impact of this result is unknown. Serum drug concentration variability was observed with both treatment regimens but to a lesser extent with CI administration. CI administration of vancomycin should be further evaluated to determine the clinical utility of this method of administration.

Species identification and antibiotic susceptibility of enterococci isolated from clinical specimens of hospitalized patients

Over a 4-month period, a total of 315 enterococci were isolated from various clinical specimens of hospitalized patients. By applying an array of biochemical tests, all strains were accurately identified to the species level, and their susceptibilities to clinically relevant antibiotics were determined by a standardized agar dilution technique. E. faecalis and E. faecium accounted for 87.3% and 9.2% of isolates, respectively. E. avium (1%), E. gallinarum (1%), E. durans (0.6%), E. hirae (0.6%), and E. casseliflavus (0.3%) isolates were also identified. Eleven strains of E. faecium and 1 E. hirae isolate were resistant to ampicillin, but none of the isolates produced beta-lactamase. Twenty-three E. faecium and 3 E. faecalis strains as well as 1 E. hirae isolate revealed imipenem resistance. A total of 25.4% enterococci (60 E. faecalis and 19 E. faecium isolates, 1 E. hirae strain) were erythromycin-resistant. Twelve strains (11 E. faecium and 1 E. avium) exhibited ciprofloxacin resistance. High-level resistance to streptomycin was found in 58 (21.1%) E. faecalis, 9 (31%) E. faecium, and both E. hirae strains, whereas high-level gentamicin resistance (HLGR) was exclusively seen in the species E. faecalis (11.6% of isolates belonging to this species). A simple agar screening test containing 500 micrograms of gentamicin per ml proved to be highly reliable for detection of HLGR. The structural gene coding for HLGR was specifically amplified by the polymerase chain reaction in all isolates showing this resistance trait. Moreover, the gene was specifically detected by a nonradioactively labelled oligonucleotide probe in colony blot hybridization assays, indicating the potential application of these molecular approaches as a diagnostic tool.

An analysis of 110 serious enterococcal infections. Epidemiology, antibiotic susceptibility, and outcome

A prospective, observational study of 110 patients with serious infections due to Enterococcus spp. in 6 university and community teaching hospitals in Connecticut was conducted to define the epidemiology of community and nosocomial serious enterococcal infections and to determine risk factors, including antibiotic resistances, that contribute to outcome. Serious community and nosocomial enterococcal infections involved a variety of sites, and antibiotic resistance was common. Types of infection by major organ system were cardiovascular, 54% (catheter-related bacteremia 28%, primary bacteremia 18%, endocarditis 6%, septic thrombophlebitis 1%); intra-abdominal, 13% (including cholangitis, 6%); renal, 13%; skin and soft tissue, 5%; bone and joint, 4%; pleuropulmonary, 4%; central nervous system, 3%; deep surgical wound, 3%; and endometritis, 2%. Sixty-one percent of infections were nosocomial; 48% of these occurred in the intensive care unit. Enterococcus faecium was responsible for 20% of all infections. Antibiotic resistances among the infections included high-level gentamicin resistance (26%), ampicillin resistance (10%), and vancomycin resistance (8%). Clinical cure was achieved in 64% of patients; 6.8% of patients relapsed, 6.8% had recurrence of the infection with a different pathogen, and overall mortality was 23%. Ampicillin resistance and a high acute physiology and chronic health evaluation (APACHE) II score were highly predictive of lack of cure.

Emerging multiply resistant enterococci among clinical isolates. II. Validation of the etest to recognize glycopeptide-resistant strains

Accurate, quantitative susceptibility testing of the enterococci for glycopeptide (vancomycin and teicoplanin) activity has become very important to guide chemotherapy as a result of emerging resistance. Reference dilution tests using broth and agar are generally difficult to perform, or in some commercial forms have demonstrated interpretive error. An alternative method, the Etest has promise as a system with qualitative and quantitative accuracy. Nearly 2000 enterococci from 97 laboratories in the United States were tested using broth microdilution, disk diffusion, and Etest methods. The Etest quantitative accuracy (+/- 1 log2 dilution) compared to the broth microdilution minimum inhibitory concentration was 90.1% (teicoplanin) to 94.1% (vancomycin). The qualitative interpretive accuracy of the Etest ranged from 98.7% for vancomycin to 99.9% for teicoplanin (no false-susceptible errors). All three tests were in remarkable agreement, with < or = 0.8% maximal discord by interpretive category. The Etest appears to be an excellent alternative to reference and standardized methods for producing quantitative activity measurements of glycopeptide susceptibility or resistance.

Emerging multiply resistant enterococci among clinical isolates. I. Prevalence data from 97 medical center surveillance study in the United States. Enterococcus Study Group

To assess the evolving problem of therapeutic drug resistances among enterococci, we organized a comprehensive national (United States) surveillance trial using 99 recruited microbiology laboratories in 48 of the 49 contiguous states or districts. All but two sites completed the protocol that generated information from nearly 2000 enterococci, usually isolated from blood cultures. All strains were speciated by the same method (API 20S) and were susceptibility tested by three methods (broth microdilution, disk diffusion, and Etest) against ampicillin, penicillin, vancomycin, teicoplanin, gentamicin, and streptomycin. Strains resistant to a glycopeptide or penicillin, or possessing high-level aminoglycoside resistance were referred to the monitor's laboratory for validation and additional susceptibility testing against other alternative antimicrobial agents. The most common species were Enterococcus faecalis and Enterococcus faecium. However, antimicrobial resistance occurred most often among the E. faecium isolates. Twenty-three percent of participant centers (22 sites) reported 87 vancomycin-resistant isolates, which accounts for 4.4% of the isolates evaluated. A recent audit (March 1994) of the laboratories not reporting vancomycin resistance during the study interval (October-December 1992) revealed that 61% of sites have now recognized these strains, a threefold increase in 12-15 months. Teicoplanin remained active against 28% (Van B phenotype) of vancomycin-resistant enterococci (10 E. faecalis, 13 E. faecium, and one Enterococcus spp.). Ampicillin-resistant beta-lactamase-positive strains were found only at one medical center (two strains, 0.2% of referred or validated strains); however, ampicillin-resistant strains represented 12% of all enterococcal, but nearly 60% of E. faecium strains.(ABSTRACT TRUNCATED AT 250 WORDS)

Risk factors for acquiring ampicillin-resistant enterococci and clinical outcomes at a Canadian tertiary-care hospital

The number of ampicillin-resistant enterococci (ARE) was noted to be increased at our teaching hospital. To determine the risk factors for acquiring this organism and to compare clinical outcomes, over a 5-month period 38 patients infected or colonized with ARE were compared with 76 patients, infected or colonized with ampicillin-susceptible enterococci (ASE). Risk factors included nosocomial acquisition, duration of hospitalization, admission to a medical service, prior antimicrobial therapy, and combination therapy for at least 7 days. The mortality rate of patients infected or colonized with ARE was higher than that of patients infected or colonized with ASE (34 versus 14%; P = 0.03), but most deaths did not appear to be related to enterococcal infection. Over a 2-year period, 16 patients with ARE bacteremia were also compared with 23 patients with ASE bacteremia. The risk factors associated with ARE bacteremia also included nosocomial acquisition, duration of hospitalization, and prior antimicrobial therapy. The mortality of patients with ARE bacteremia was also higher than that of patients with ASE bacteremia (81 versus 30%; P = 0.003), with most deaths being due to the underlying disease or a complication of it. Typing of ARE isolates by pulsed-field gel electrophoresis showed that two genotypes predominated in our institution. A prolonged hospital stay, exposure to multiple antimicrobial agents, and perhaps nosocomial transmission are important factors in acquiring ARE. The presence of ARE may also be a marker for poor outcome.

Characterization of enterococci isolated from human and nonhuman sources in Brazil

A total of 330 enterococci strains isolated from several human (272 strains) and animal (27) clinical specimens and environmental sources (31) in Brazil were identified to species level. Major human sources included urine (48.5%), blood (15.8%), and wounds (9.5%). Human isolates were identified as Enterococcus faecalis (90.0%), E. faecium (6.9%), E. gallinarum (1.1%), E. durans (0.8%), E. casseliflavus (0.4%), E. raffinosus (0.4%), and E. mundtii (0.4%). Strains isolated from animals were composed of E. hirae (40.7%), E. faecalis (33.3%), E. faecium (18.5%), and E. casseliflavus (7.5%). Among environmental isolates, 42.0% were E. faecalis, 35.4% E. faecium, 13.0% E. hirae, 6.4% E. casseliflavus, and 3.2% E. durans. Antimicrobial susceptibility tests were performed for 200 strains. Overall, high-level resistance (HLR) to aminoglycosides was found in 66 (33.0%) of the strains tested. HLR to gentamicin was detected in 11.5% of the strains, whereas 19.0% of the strains showed HLR to streptomycin and 26.0% showed HLR to kanamycin. Five (22.7%) of the E. faecium strains were resistant to ampicillin [minimum inhibitory concentration (MIC) > 32 micrograms/ml]. Vancomycin MIC50 and MIC90 were 2 and 4 micrograms/ml, respectively; only eight strains (identified as E. casseliflavus or E. gallinarum) had MIC of 8 micrograms/ml. No beta-lactamase activity was detected by the nitrocefin test.

High-level penicillin resistance and penicillin-gentamicin synergy in Enterococcus faecium

Thirty-seven Enterococcus faecium strains with different levels of penicillin susceptibility were studied in time-kill experiments with a fixed concentration (5 micrograms/ml) of gentamicin combined with different penicillin concentrations (6 to 600 micrograms/ml). Synergy was defined as a relative decrease in counts of greater than 2 log10 CFU per milliliter after 24 h of incubation when the combination of the antibiotics was compared with its most active component alone. The minimal synergistic penicillin concentrations found were 6 micrograms/ml for 16 of 16 strains for which penicillin MICs were < or = 25 micrograms/ml, 20 to 100 micrograms/ml for 14 of 17 strains for which penicillin MICs were 50 to 200 micrograms/ml, and 200 to 500 micrograms/ml for 4 of 4 strains for which MICs penicillin were > 200 micrograms/ml. Penicillin-gentamicin synergy was observed even in high-level penicillin-resistant E. faecium strains at penicillin concentrations close to one-half the penicillin MIC. The possibility of treating infections caused by high-level penicillin-resistant E. faecium strains with penicillin-gentamicin combinations in particular cases may depend on the penicillin levels attainable in vivo.

Enterococcus, an emerging pathogen

OBJECTIVE

To review the bacterial genus Enterococcus with respect to its epidemiology, specific infections in humans, mechanisms of resistance and tolerance, and antimicrobial treatment.

DATA SOURCES

A MEDLINE search of English-language journal articles published from 1977 to 1992 was completed. Articles published prior to 1977 were identified through Index Medicus and from references appearing in the bibliographies of other journal articles. Information also was acquired from abstracts, personal communication with infectious disease specialists with active research in the area of enterococcal infection, and conference proceedings.

STUDY SELECTION

In vitro data; animal models of enterococcal infection; case reports; and case-controlled, cohort, and randomized controlled trials in humans were evaluated for relevant information.

DATA EXTRACTION

Studies were evaluated by their methodologic strength (e.g., randomized controlled trial), reporting of clinically relevant outcomes (e.g., clinical response to antimicrobial therapy), statistical analyses, and accountability of all patients who entered the study.

DATA SYNTHESIS

The incidence of enterococcal infections has increased in recent years and enterococci are now the second most frequently reported nosocomial pathogens. Enterococcus faecalis is the pathogen responsible for most enterococcal infections seen today; it has been implicated as an important cause of endocarditis, bacteremia, urinary tract infections, and intraabdominal infections.

CONCLUSIONS

Enterococcal infection is of particular concern clinically because of its resistance to several antibiotics. Controlled comparative clinical trials of antimicrobial therapy in humans are lacking for several enterococcal infections. Therefore, the recommendations for antimicrobial therapy presented in this review are guidelines that reflect our current understanding of antibiotics used for enterococcal infection.

Molecular characterization of highly gentamicin-resistant Enterococcus faecalis isolates lacking high-level streptomycin resistance

Antimicrobial susceptibilities and DNA contents were analyzed for six clinical isolates of Enterococcus faecalis that had high-level resistance to gentamicin (MIC > 2,000 micrograms/ml) but not streptomycin and were obtained from patients in diverse geographic areas. Contour-clamped homogeneous electric field electrophoresis of genomic DNA showed all isolates to be different strains. Gentamicin resistance was transferred from four isolates to plasmid-free enterococcal recipients in filter matings. Restriction enzyme analysis of transconjugants showed distinct gentamicin resistance plasmids. A probe specific for the gentamicin resistance determinant hybridized to the plasmids of four isolates and to the chromosomes of two isolates. These findings suggest that clonal dissemination is not responsible for the spread of these resistant strains, that resistance determinants occur on different plasmids as well as on the chromosome of E. faecalis, and that the genetic determinants of resistance are related.

In vitro activity of nine fluoroquinolone antibiotics against 200 strains of enterococci

The in vitro activity of nine fluoroquinolones against 200 recent clinical isolates of enterococci was evaluated. The relative activity was WIN57273 > sparfloxacin > ciprofloxacin, temafloxacin > ofloxacin > fleroxacin, lomefloxacin, norfloxacin > enoxacin. Susceptibility to gentamicin or streptomycin did not influence the activity of the fluoroquinolones.

Molecular typing of ampicillin-resistant, non-beta-lactamase-producing Enterococcus faecium isolates from diverse geographic areas

Molecular typing methods were compared by using 66 ampicillin-resistant, non-beta-lactamase-producing Enterococcus faecium clinical isolates from diverse geographic areas. Whole-plasmid analysis, restriction enzyme analysis of plasmid DNA with EcoRI and HindIII, and contour-clamped homogeneous electric field electrophoresis with digestion by SmaI and ApaI were performed on all isolates. Whole-plasmid analysis identified 47 different groups. Restriction enzyme analysis of plasmid DNA identified 50 groups when EcoRI was used and 51 groups when HindIII was used. Results with EcoRI and HindIII differed in 9 of 66 isolates. Grouping results with whole-plasmid analysis differed from results of restriction enzyme analysis of plasmid DNA (combining EcoRI and HindIII) in 20 of 66 isolates. Contour-clamped homogeneous electric field electrophoresis identified 46 groups when SmaI was used and 44 groups when ApaI was used. Results with SmaI and ApaI differed in 3 of 66 isolates. Grouping results with contour-clamped homogeneous electric field electrophoresis (combining SmaI and ApaI) differed from results of restriction enzyme analysis of plasmid DNA (combining EcoRI and HindIII) in 17 of 66 isolates. The combined use of whole-plasmid analysis, restriction enzyme analysis of plasmid DNA with two enzymes, and contour-clamped homogeneous electric field electrophoresis with two restriction enzymes should be considered when E. faecium is typed for epidemiologic investigation.

Pharmacodynamic interaction between RP 59500 and gram-positive bacteria infecting fibrin clots

The fibrin clot penetration and in vivo bactericidal activity of RP 59500, a new semisynthetic streptogramin, for two Staphylococcus aureus strains (one methicillin resistant and the other methicillin susceptible), two Staphylococcus epidermidis strains (one methicillin resistant and the other methicillin susceptible), and one Enterococcus faecalis strain were evaluated. The clots, inserted subcutaneously, were infected with a mean of 10(8) CFU of the pathogen per g. For each strain, groups of four rabbits received a single intravenous injection of 50 mg of RP 59500 per kg of body weight over 30 min. The mean peak level of RP 59500 in serum in the infected rabbits was 61.9 +/- 6.3 micrograms/ml. The drug was detectable in serum at a level of 0.8 micrograms/ml up to 4 h after administration. The mean peak fibrin clot drug level at 1 h was 3.3 +/- 0.1 micrograms/g. At 6 h, the level in clots was 1.2 +/- 0.1 micrograms/g. The mean half-life in serum in infected rabbits was 0.34 +/- 0.01 h, while in clots the drug exhibited a longer half-life of 3.8 +/- 0.4 h. In vivo, this new streptogramin sterilized the clots infected with the two S. aureus strains studied in less than 1 h and induced a marked reduction in colony counts of the two S. epidermidis strains studied for up to 24 h. The activity of the streptogramin against E. faecalis was limited. These results suggest that RP 59500 should be further evaluated for the treatment of infection with methicillin-resistant staphylococci.

Comparison of daptomycin, vancomycin, and ampicillin-gentamicin for treatment of experimental endocarditis caused by penicillin-resistant enterococci

Infections with enterococci that are resistant to multiple antibiotics are an emerging clinical problem. We evaluated the antibiotic treatment of experimental enterococcal endocarditis caused by two strains with different mechanisms of penicillin resistance. Enterococcus faecalis HH-22 is resistant to aminoglycosides and penicillin on the basis of plasmid-mediated modifying enzymes; Enterococcus raffinosus SF-195 is susceptible to aminoglycosides but is resistant to penicillin on the basis of low-affinity penicillin-binding proteins. Animals infected with strain HH-22 received 5 days of treatment with the following: no treatment; daptomycin (20 mg/kg of body weight twice daily [b.i.d.], intramuscularly [i.m.]), vancomycin (20 mg/kg b.i.d., intravenously), or ampicillin (100 mg/kg three times daily, i.m.) plus gentamicin (2.5 mg/kg b.i.d. i.m.). Although vancomycin was superior to ampicillin-gentamicin (P less than 0.01), daptomycin was significantly better than all other treatment regimens (P less than 0.01) in reducing intravegetation enterococcal densities, although no vegetations were rendered culture negative by this agent. Animals infected with strain SF-195 received 5 days of no therapy, ampicillin, ampicillin-gentamicin, vancomycin, or daptomycin (all at the dosage regimens described above). Daptomycin, vancomycin, and ampicillin-gentamicin each lowered intravegetation enterococcal densities significantly better than did ampicillin monotherapy or no treatment (P less than 0.01); moreover, these three treatment regimens rendered significantly more vegetations culture negative than did ampicillin monotherapy or no treatment (P less than 0.05). Serum daptomycin levels remained above the MICs and MBCs for both enterococcal strains throughout the 12-h dosing interval used in the study. Daptomycin and vancomycin were both active in vivo in these models of experimental enterococcal endocarditis caused by penicillin-resistant strains, irrespective of the mechanism of resistance. This activity correlated with the unique cell wall sites of action of these agents (binding to lipoteichoic acid and pentapeptide precursor, respectively) compared with the sites of action of beta-lactams (penicillin-binding proteins). Beta-Lactamase production by strain HH-22 precluded in vivo efficacy with ampicillin-gentamicin combinations. In contrast, this combination was active in vivo against strain SF-195, which exhibited intermediate-level penicillin resistance (MIC, 32 micrograms/ml), likely reflecting the ability of high-dose ampicillin to achieve enough binding to low-affinity penicillin-binding proteins to cause augmented aminoglycoside uptake.

Emergence and nosocomial transmission of ampicillin-resistant enterococci

Between 1986 and 1988, the incidence of ampicillin-resistant enterococci increased sevenfold at a university-affiliated hospital. Forty-three patients acquired nosocomial infections with ampicillin-resistant enterococci, most of which were also resistant to mezlocillin, piperacillin, and imipenem. An analysis of plasmid and chromosomal DNAs of isolates revealed that the increase was due to an epidemic of 19 nosocomial infections that yielded closely related strains of Enterococcus faecium and to a significant increase in the incidence of nonepidemic, largely unrelated strains of ampicillin-resistant enterococci. The nonepidemic strains were identified as E. faecium, E. raffinosus, E. durans, and E. gallinarum. A logistic regression analysis revealed that patients with nonepidemic resistant strains were 16 times more likely than controls to have received preceding therapy with imipenem. In our institution, the increase in the incidence of ampicillin-resistant enterococci appears to be due to the selection of various strains of resistant enterococci by the use of imipenem and to the nosocomial transmission of E. faecium and E. raffinosus.

Antibiotic-resistant enterococci

Enterococci have emerged as an important cause of nosocomial infection. Successful antibiotic treatment of serious enterococcal infection usually depends on the synergistic bactericidal effect achieved by the combination of a cell wall-active agent, such as ampicillin or a glycopeptide, and an aminoglycoside. However, the prevalence of enterococci resistant to one or more of these antibiotics is increasing, and has resulted in serious therapeutic difficulties. The mechanisms of antibiotic resistance, and the epidemiology, laboratory diagnosis and management of infection with antibiotic-resistant enterococci are discussed.

Overproduction of a penicillin-binding protein is not the only mechanism of penicillin resistance in Enterococcus faecium

In 1989 and 1990, a large number of ampicillin-resistant strains of Enterococcus faecium were isolated from infected patients treated at intensive care units in Berlin, Germany. Twenty-five clinical isolates, including five different biotypes as classified by acid production from various sugars and a wide range of susceptibilities to ampicillin (MICs between 0.5 and 128 micrograms/ml), were selected for a detailed analysis of penicillin-binding proteins (PBPs). All strains contained a slowly reacting PBP with low penicillin affinity known to be present in enterococci. Overproduction of this PBP relative to susceptible isolates was noted, especially in all strains for which the MIC of ampicillin was 8 micrograms/ml, to a lesser degree in the more resistant strains, but not at all in the three highly resistant isolates for which the MIC was 128 micrograms/ml. In these three strains, this PBP appears to have a reduced affinity for beta-lactams. The results suggest that overproduction of PBP 6 correlates only with intermediate resistance levels and that higher resistance is mediated by yet another, still unknown mechanism, probably including reduction of beta-lactam affinity in one or more PBPs.

Ampicillin-resistant Enterococcus raffinosus in an acute-care hospital: case-control study and antimicrobial susceptibilities

A prospective study identified 9 (32%) of 28 ampicillin-resistant (MIC greater than or equal to 16 micrograms/ml) enterococcus isolates as Enterococcus raffinosus. A case-control study found no significant differences with respect to underlying diseases, catheterization, or surgery between patients with ampicillin-resistant E. raffinosus and those with ampicillin-susceptible Enterococcus spp. Prior treatment with antibiotics and prolonged hospitalization were more frequent among patients with ampicillin-resistant E. raffinosus. Patients with the same strain (determined by plasmid analysis) were frequently hospitalized concurrently.

Increasing resistance to beta-lactam antibiotics among clinical isolates of Enterococcus faecium: a 22-year review at one institution

To identify any change in the antibiotic resistance of Enterococcus faecium, we examined the antibiotic susceptibilities of clinical strains (n = 84) isolated at one institution during the 22 years since 1968. A significant increase in resistance to penicillin was observed during the study period: the MICs of penicillin for 50 and 90% of isolates tested were 16 and 64 micrograms/ml, respectively, from 1969 to 1988 (n = 48; geometric mean MIC, 14 micrograms/ml) , whereas they were 256 and 512 micrograms/ml, respectively, from 1989 to 1990 (n = 36; geometric mean MIC, 123 micrograms/ml) (P less than 0.001). A comparable increase in resistance to ampicillin was also noted (P less than 0.001). No strains produced detectable beta-lactamase. In contrast, susceptibilities to vancomycin, teicoplanin, and ciprofloxacin remained stable. High-level resistance to gentamicin was observed in none of 48 isolates from 1969 to 1988, but was present in 22 of 36 strains (61%) from 1989 to 1990 (P less than 0.001) and was significantly associated with resistance (MIC, greater than or equal to 128 micrograms/ml) to penicillin (P less than 0.001). To assess the potential evolution of antibiotic resistance in this species, clinical isolates (n = 24) were compared with strains isolated in 1968 from a human population in the Solomon Islands that was never exposed to antibiotics. Solomon Island isolates were significantly more susceptible than all clinical strains to penicillin, ampicillin, and vancomycin (P less than 0.001 for each), but they exhibited no differences in susceptibility to teicoplanin or ciprofloxacin. The penicillin-binding affinity of penicillin-binding protein 5 (PBP 5) in penicillin-resistant clinical strains (MIC, 512 micrograms/ml) was notably lower than that in strains with more typical susceptibilities, suggesting an alteration in this PBP as a possible mechanism for increased penicillin resistance. Solomon Island strains most susceptible to penicillin demonstrated a prominent PBP 5* and the absence of PBP 5. These changes in the antibiotic resistance of E. faecium emphasize the importance of identifying this species in patients with serious enterococcal infections and the necessity of assessing its susceptibility to both beta-lactams and aminoglycosides if effective therapy is to be identified.

Species identification and antibiotic susceptibility testing of enterococci isolated from hospitalized patients

A total of 236 enterococci from hospitalized patients were identified to the species level, and their susceptibilities to 11 antibiotics were determined. Overall, 195 (82.6%) and 38 (16.1%) isolates were identified as Enterococcus faecalis and E. faecium, respectively, but the species distribution as determined from blood culture isolates differed markedly. A total of 27 (63.2%) E. faecium isolates, but no E. faecalis strains, were ampicillin resistant (MIC, greater than 8 micrograms/ml). High-level gentamicin resistance (MIC, greater than 500 micrograms/ml) was found in 8.2% of E. faecalis isolates but was not seen in other species.

Comparison of Enterococcus raffinosus with Enterococcus avium on the basis of penicillin susceptibility, penicillin-binding protein analysis, and high-level aminoglycoside resistance

We reidentified our laboratories' collections of 57 enterococcal isolates previously classified as Enterococcus avium by the API Rapid Strep identification system (Analytab Products, Plainview, N.Y.) with the identification criteria recommended by Facklam and Collins (R. R. Facklam and M. D. Collins, J. Clin. Microbiol. 27: 731-734, 1989). Thirty isolates were identified as true E. avium, 25 isolates were identified as E. raffinosus, and 2 isolates were identified as E. pseudoavium. E. raffinosus could be differentiated from E. avium on the basis of penicillin susceptibility, as follows: MIC for 50% of E. raffinosus isolates tested (MIC50), 32 micrograms/ml; MIC90, 64 micrograms/ml (range, 4 to 64 micrograms/ml); E. avium MIC50, 1 microgram/ml; MIC90, 2 micrograms/ml (range, 0.5 to 2 micrograms/ml). No strains produced detectable beta-lactamase. Penicillin-binding protein (PBP) analysis of all E. raffinosus isolates demonstrated the unique pattern reported previously (M. D. Collins, R. R. Facklam, J. A. E. Farrow, and R. Williamson, FEMS Microbiol. Lett. 57:283-288, 1989); however, a number of newly identified PBPs were noted. Of 25 isolates, 13 had an additional PBP of 77 kDa (designated PBP 6*), while all isolates possessed a 52-kDa PBP (PBP 7) and a 46-kDa PBP (PBP 8). The presence or absence of PBP 6* did not correlate with penicillin susceptibility; however, PBP 7 demonstrated many features suggestive of low penicillin-binding affinity and may represent a possible mechanism for the relative resistance of this species to penicillin, although this hypothesis remains speculative since attempts to develop a penicillin-hypersusceptible E. raffinosus mutant were unsuccessful. E. raffinosus isolates were significantly more likely to exhibit high-level resistance to kanamycin than E. avium strains were (P < 0.001; chi-square); however, no strains demonstrated high-level resistance to gentamicin. No trend toward increasing penicillin resistance was noted among this collection of E. avium and E. raffinosus isolates collected over the past 35 and 14 years, respectively. Relative resistance to penicillin may be a helpful differentiating feature between E. avium and E. raffinosus when assessment of raffinose metabolism is not possible or is indeterminant.

In vitro activity of RP 59500, a semisynthetic injectable pristinamycin, against staphylococci, streptococci, and enterococci

The in vitro activity of RP 59500, a semisynthetic pristinamycin, was compared with the activities of vancomycin, oxacillin, ampicillin, gentamicin, ciprofloxacin, and rifampin against five Staphylococcus species, five Streptococcus species, and four Enterococcus species. For staphylococci, MICs were 0.13 to 1 microgram/ml and the MICs for 90% of the strains tested (MIC90s) were 0.13 to 0.5 microgram/ml; there were no differences between oxacillin-susceptible and -resistant strains. For streptococci, MICs were 0.03 to 4 micrograms/ml and MIC90s were 0.25 to 2 micrograms/ml; viridans group streptococci were the least susceptible streptococci. For enterococci, MICs were 0.25 to 32 micrograms/ml and MIC90s were 2 to 4 micrograms/ml; Enterococcus faecalis was the least susceptible. Vancomycin was the only comparative drug with consistent activity against all species of gram-positive cocci. With RP 59500, raising the inoculum 100-fold, lowering the pH of cation-adjusted Mueller-Hinton broth to 5.5, or omitting cation supplementation had little effect on MICs, but 50% serum increased MICs 2 to 4 dilution steps. The differences between MBCs and MICs were greater for staphylococci and enterococci than for streptococci. Time-kill studies with 24 strains indicated that RP 59500 concentrations 2-, 4-, and 16-fold greater than the MICs usually killed bacteria of each species at similar rates; reductions in CFU per milliliter were less than those observed with oxacillin or vancomycin against staphylococci and less than those observed with ampicillin against enterococci. RP 59500 antagonized the bactericidal activities of oxacillin and gentamicin against Staphylococcus aureus ATCC 29213 and that of ampicillin against E. faecalis ATCC 29212. Against the latter, combination with gentamicin was indifferent. RP 59500 has a broad spectrum of in vitro activity against gram-positive cocci; combining it with other drugs is not advantageous.