INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion References

With the global spread of methicillin-resistant strains, glycopeptide antibiotics have become the mainstay of chemotherapy in both coagulase-negative staphylococci as well as Staphylococcus aureus infections worldwide. Loss of the antistaphylococcal efficacy of this family of drugs would pose a serious challenge to the treatment of methicillin-resistant S. aureus (MRSA) infections. Such a situation may emerge by the transfer of enterococcal glycopeptide resistance genes to staphylococci, which often cocolonize wound infection sites with enterococci. In fact, transfer of the vancomycin resistance from Enterococcus faecalis to an S. aureus strain has already been demonstrated under simulated laboratory conditions (11). An additional concern relates to the increasingly frequent detection of coagulase-negative staphylococcal isolates with reduced susceptibility to glycopeptide antibiotics in clinical specimens associated with a variety of clinical diseases, particularly in patients undergoing continuous ambulatory peritoneal dialysis (CAPD) (8, 13-15). Clinical failure of vancomycin therapy in an MRSA-infected patient has already been described in Japan (9). Soon after that, MRSA infections for which the vancomycin MICs showed a similar range (~8 µg/ml) were also reported in the United States (2, 3).

In this communication, we describe the characterization of five consecutive isolates of a single methicillin-resistant Staphylococcus epidermidis (MRSE) strain with reduced susceptibility to vancomycin and teicoplanin, recovered from the dialysate of a peritoneal dialysis (PD) patient who experienced recurrent episodes of peritonitis during an extended period of therapeutic and/or prophylactic use of vancomycin. Ominously, the PD catheter site (as well as several other body sites) of the patient was also colonized by a vancomycin-resistant E. faecalis strain carrying the vanA gene. Nevertheless, the MRSE isolates did not react with the vanA DNA probe but appeared to carry a distinct glycopeptide resistance mechanism. The vancomycin MICs for the five MRSE isolates, as determined by broth dilution, were in the range of 2 to 16 µg/ml. However, population analysis showed that cultures also contained bacteria for which the vancomycin MICs were as high as 25 and 50 µg/ml at significant frequencies (10⁴ to 10⁵). Enrichment for these subpopulations occurred without difficulty under laboratory conditions.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion References

Bacterial strains were grown in tryptic soy broth (TSB; Difco, Detroit, Mich.) at 37°C with aeration. Organisms were identified to the species level by the API test system (bioMerieux Vitek, Inc., Hazelwood, Mo.). A preliminary determination of MICs of vancomycin was performed by a broth microdilution method, with TSB (Difco) and incubation of samples at 37°C for 24 and 48 h. The vancomycin (and methicillin)-susceptible S. aureus strain NCTC 8325 was always used as a susceptible control. Population analysis profiles (PAPs) were done for a more accurate and quantitative evaluation of antibiotic susceptibility (19). Overnight cultures of bacteria (10⁹ CFU/ml) were plated at a series of dilutions on tryptic soy agar plates containing antibiotic-free medium or twofold dilutions of the test antibiotic within the drug concentration range of 0.75 µg up to 100 µg/ml in the cases of vancomycin and teicoplanin and up to 800 µg/ml in the case of methicillin. Plates were incubated at 37°C for 48 h, and the number of bacterial colonies was counted. Plotting colony counts against drug concentrations provides a graphic display (PAP) of the composition of the bacterial culture in terms of the homogeneity or heterogeneity of the antibiotic susceptibility phenotype. Methicillin and glycopeptide MICs for the majority and for subpopulations of cells were determined by inspection of PAPs.

Preparation of chromosomal DNA for pulsed-field gel electrophoresis (PFGE) and separation of SmaI-restricted fragments in a contour-clamped homogeneous electric field (CHEF) apparatus (CHEF-DRII; Bio-Rad, Richmond, Calif.) were carried out as described previously (7). Autolysis was induced by suspending bacteria in buffer containing Triton X-100 (6). The titer of vancomycin in the growth medium was determined with a bioassay (16). Aggregation of cells and the ultrastructure of bacteria were determined by phase-contrast microscopy and electron microscopy with a procedure previously described (18).

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion References

Clinical history. A 34-year-old female first developed renal failure of unknown etiology in 1985, and by 1988, she required hemodialysis. Two years later, because of inability to maintain a vascular access, she was placed on long-term PD. A year later, in 1991, she developed peritonitis, and over the next 6 years, she had recurrent episodes of peritonitis associated with fever, chills, nausea, vomiting, and abdominal pain. By 1994, the episodes became more frequent and severe, and S. epidermidis was first isolated from cloudy dialysate. She was placed on vancomycin (30 mg per liter) in each dialysis exchange for 10 to 60 days (5 exchanges per day) when symptoms or cloudy dialysate appeared. S. epidermidis was frequently recovered from her dialysate prior to each therapeutic trial. Because of the increasing frequency of her episodes and the need to continue peritoneal dialysis, she was placed on prophylactic vancomycin (30 mg per liter) in each dialysate exchange for 3 weeks of each month (5 exchanges per day) from January to June of 1997. Because of the identification of S. epidermidis with intermediate resistance to vancomycin (Table 1), the peritoneal dialysis catheter was removed on 16 June 1997, and she was placed on hemodialysis thrice weekly. Except for one intravenous dose associated with vascular access surgery, she did not receive vancomycin from 16 June until 10 September 1997. During this time period, she felt well and showed no clinical signs or symptoms of peritonitis, although she continued to have minimal drainage from the PD catheter site.

Properties of the bacterial isolates. Table 1 summarizes the relevant properties of bacteria recovered from a variety of body sites and the PD fluid. Cultures of peritoneal dialysate collected between 27 March and 28 May yielded pure cultures of MRSE (isolates RR4 through RR8) with reduced susceptibility to vancomycin and teicoplanin. Vancomycin-resistant enterococci, consisting of an E. faecium strain (isolate RR1) carrying the vanB gene and E. faecalis strains (RR2, RR3, and RR9 through RR11) carrying the vanA gene, were also detected at several body sites of the patient.

View larger version (46K): [in this window] [in a new window]   FIG. 1.   PFGE patterns of isolates recovered from the PD patient. Chromosomal DNA was prepared, SmaI restricted, and separated by PFGE (see Materials and Methods). The bottom panel shows results of hybridization with a probe for vanA. E. faecium strains EFSK2 (vanB) and EFSK33 (vanA) were used as controls. l.m.w., low molecular weight.

Heterogeneity of vancomycin resistance. Evaluation of the broth dilution vancomycin MICs for the S. epidermidis strains after 24 and 48 h of incubation allowed detection of modest increases in MICs with extended incubation time. A more detailed characterization of overnight cultures of the five S. epidermidis isolates by population analysis showed that they were heterogeneous with respect to susceptibility to vancomycin, teicoplanin, and methicillin (Fig. 2). While the MIC of vancomycin for the majority of cells in these cultures was between 3 and 6 µg/ml, the cultures also contained bacteria for which the MICs were more substantially increased at low but significant frequencies. For instance, cultures of isolate RR4 contained subpopulations of cells capable of growing on plates that contained 12 µg of vancomycin per ml (MIC, 25 µg/ml) at a frequency of 10⁴. Cells that formed colonies on 25 µg of vancomycin per ml (MIC, 50 µg/ml) were also present (frequency, 10⁵).

View larger version (17K): [in this window] [in a new window]   FIG. 2.   Phenotypic expression of methicillin (A), vancomycin (B), and teicoplanin (C) resistance of consecutive isolates of S. epidermidis strains recovered from the PD patient.

Selection and stability of the highly vancomycin-resistant subpopulations. It was relatively easy to increase the proportion of these subpopulations under laboratory conditions. Colonies of isolate RR4 that were capable of growing on 6 µg of vancomycin per ml were dispersed in drug-free growth medium and used at very low cell concentrations (approximately 100 cells per ml) as inocula. Upon replating for population analysis, the majority of the culture showed an increase in the vancomycin MIC from 6 µg/ml for the original isolate, RR4, to 12 µg/ml. A separate colony of RR4 picked from the 25-µg/ml vancomycin plate yielded a culture in which the MIC of vancomycin shifted up to 25 µg/ml. Both of these cultures also became more homogeneous with respect to vancomycin susceptibility (Fig. 3A).

View larger version (14K): [in this window] [in a new window]   FIG. 3.   Enrichment of S. epidermidis cultures for subpopulations of bacteria for which the vancomycin MICs were elevated. Isolate RR4 was plated for population analysis on agar containing increasing concentrations of vancomycin (). Bacterial cells capable of forming colonies on the agar containing 6 and 25 µg of vancomycin/ml (A, 6* and 25*, respectively) were picked, diluted in drug-free TSB, grown overnight, and subsequently plated for population analysis. Symbols indicate cultures grown from a colony picked from the 6-µg/ml plate () or from the 25-µg/ml plate (). A culture of RR4 was diluted into TSB con- taining increasing concentrations of vancomycin (eventually 25 µg/ml), grown overnight, and then plated for population analysis on agar plates containing vancomycin (B) or teicoplanin (C). Open circles () represent population profiles of the original culture.

Enrichment of S. epidermidis cultures for bacteria for which the vancomycin MIC was increased during exposure to a bactericidal concentration of the antibiotic. S. epidermidis isolates RR4 and RR8 were grown in TSB. At a cell concentration of about 5 × 10⁶ CFU/ml, both cultures received 30 µg of vancomycin per ml, and viable titers were determined at hourly intervals (Fig. 4 [top]). After 7 h of incubation with the antibiotic, the viable titers of the cultures were reduced to about 5 × 10⁴ and 5 × 10³ CFU/ml in strains RR4 and RR8, respectively. The drop in viability was not accompanied by any change in optical density of the cultures. Upon further incubation (6 to 7 days), both cultures increased in viable titer (and optical density), to reach a stationary-phase concentration of about 5 × 10⁹ CFU/ml. Population analysis done at the beginning and at the end of this experiment showed increases in the vancomycin MICs from 6 µg/ml to 50 µg/ml in RR4 and from 3 µg/ml to 25 µg/ml in RR8 (Fig. 4).

View larger version (24K): [in this window] [in a new window]   FIG. 4.   Selection for bacteria with increased vancomycin resistance during the bactericidal action of vancomycin. (Top panel) Effect of vancomycin (30 µg/ml added at the time indicated by arrow) on the viable titers of strains RR4 () and RR8 (). (Bottom panels) Population analysis of the two strains before () and after () the massive population shift caused by vancomycin treatment.