Heteroresistance to Vancomycin in Enterococcus faecium

Abstract

This study presents the first report of vancomycin heteroresistance in an Enterococcus faecium isolate from a patient. The original isolate was susceptible in vitro to vancomycin. E-tests showed growth of subcolonies in a zone of inhibition with a vancomycin MIC of >256 μg/ml. Both the susceptible and resistant colonies were from the same strain as determined by PFGE, and both contained the vanA gene as determined by PCR.

Enterococci are recognized to be important human pathogens that are responsible for serious nosocomial infections, including bacteremia, endocarditis, and intra-abdominal and urinary tract infections (5, 9). Recent data suggest that 50 to 90% of Enterococcus faecium isolates are resistant to vancomycin (1, 6). Treatment of vancomycin-resistant enterococci (VRE) has become a clinical challenge, since E. faecium is resistant to multiple antimicrobial agents.

Resistance to vancomycin among enterococci is known to be homogenous within a culture. However, heteroresistance to vancomycin has been previously observed in staphylococci (10). This is the first report of a hetero-vancomycin-resistant E. faecium (hetero-VREF) isolate; importantly, it was isolated from a patient with endocarditis. This isolate was reported to be susceptible to vancomycin, in vitro; however, E-test (AB Biodisk, Solna, Sweden) results showed a subpopulation of isolates resistant to vancomycin. Standardized automated quantitative testing methods may not detect the presence of resistant subpopulations. Therefore, a subsequent adverse outcome is possible when an inappropriate use of vancomycin is combined with quantitative methods based on broth microdilution, especially if a rapid reading of results is performed.

The patient in the present study was a 31-year-old female with a history of intravenous drug abuse who was transferred to William Beaumont Hospital in March 2000; for tricuspid valve replacement surgery after recurrent VRE tricuspid valve endocarditis. In October of 1999, the patient was hospitalized and received 6 weeks of intravenous vancomycin treatment for tricuspid valve endocarditis caused by methicillin-resistant Staphylococcus aureus. In December, she was readmitted and treated for 4 weeks with quinupristin-dalfopristin plus rifampin for both VRE bacteremia and possible endocarditis. She was hospitalized again in February 2000, for recurrent fever, night sweats, and buttock abscess. The blood culture collected on February 20 2000 grew vancomycin-susceptible E. faecium (VSEF; vancomycin MIC, <2 μg/ml by microtiter broth dilution), and quinupristin-dalfopristin therapy was restarted subsequently with no improvement. Follow-up blood cultures 2 days later grew vancomycin-resistant E. faecium (VREF; vancomycin MIC, >16 μg/ml), and treatment with quinupristin-dalfopristin plus rifampin was continued. Subsequently, the patient was treated with tricuspid valve replacement and 6 weeks of quinupristin-dalfopristin plus rifampin, which cured the endocarditis.

Although the original isolate was susceptible in vitro when tested by microtiter broth dilution, follow-up E-tests showed subcolonies present in the clear zone of inhibition with a vancomycin MIC of >256 μg/ml (Fig. 1).

FIG. 1.

E-test result for VSEF isolate (WBH22608) with resistant subcolonies present in the clear zone of inhibition.

Isolates.

Two clinical isolates of E. faecium collected from the blood of the patient on 20 February 2000 (VSEF isolate WBH22608) and 22 February 2000 (VREF isolate WBH22609) were evaluated.

Susceptibility testing.

The MIC of vancomycin (Eli Lilly & Co.) was determined by broth microdilution according to National Committee for Clinical Laboratory Standards guidelines (8). Vancomycin MIC determinations were also repeated by E-test according to the manufacturer's specification. E. faecalis strain ATCC 29212 was used as the control for the in-vitro susceptibility studies. The MICs for both the original isolates and the subcolonies from within the E-test zone of inhibition were evaluated. The MIC of the VSEF isolate (WBH22608) was reassessed after serial passages in broth containing a subinhibitory concentration (0.125 μg/ml) of vancomycin.

Strain typing by PFGE.

Genomic DNA was prepared in agarose plugs, and digested with the enzyme SmaI (New England BioLabs, Beverly, Mass.), and pulsed-field gel electrophoresis (PFGE) was performed using a CHEF-DRIII apparatus (Bio-Rad Laboratories, Richmond, Calif.) as previously described (3). Interpretation of gels was performed by visual inspection using the criteria of Tenover et al. (11).

Detection of vancomycin resistance genes by (PCR).

The vancomycin resistance gene content of each strain type was determined by PCR, using methods previously described (4, 12). The oligonucleotide primers used for amplification of vanA and vanB genes were described by Clark et al. (2). The oligonucleotide primers used to amplify the vanRS, vanSH, vanHAX, vanXY, and vanYZ regions of Tn1546 were described previously (7).

By broth microdilution, the MICs of vancomycin for isolates WBH22609 and WBH22608 were 256 and 0.25 μg/ml, respectively. By E-test methods, the vancomycin MICs were >256 μg/ml for WBH22609 and 1.0 μg/ml for WBH22608. Although the VSEF isolate was susceptible by both methods, a subpopulation of resistant isolates (WBH22610) was observed in the E-test zone of inhibition (Fig. 1). The vancomycin MIC for this isolate, as determined by broth microdilution and E-test, was >256 μg/ml. The vancomycin MIC for WBH22608 increased to 256 μg/ml after eight serial passages in vancomycin-containing broth.

The results of PFGE are shown in Fig. 2a. All isolates belonged to the same strain type, with the PFGE pattern for WBH22607 differing from those of the other two isolates in only two bands. Both the VSEF and the VREF isolates were positive for the vanA gene, and all three isolates were negative for the vanB gene (Fig. 2b). The vanA transposon in the vancomycin-sensitive isolate had no detectable changes compared to either the vancomycin-resistant isolate or the control isolate EF228.

FIG. 2.

(a) PFGE of genomic DNA. Lanes: 1, lambda ladder; 2, VSEF isolate WBH22608; 3, resistant WBH22610 subcolonies. (b) Agarose gel electrophoresis of PCR-amplified vanA genes of strain EF228. Lanes: 1, 1-Kb ladder; 2, vanA strain EF228; 3, vanB strain EF6133; 4, VSEF isolate WBH22608; 5, VREF isolate WBH22610.

Over the past few years there have been reports of vancomycin heteroresistance among staphylococcus; (10). This is the first report of heteroresistance to vancomycin in enterococci. The initial isolates were originally susceptible in vitro to low concentrations of vancomycin. Subsequent isolates yielded VREF after the hetero-VREF isolate was cultured. Initial treatment with vancomycin may have resulted in the selection of a resistant subpopulation, leading to the identification of VREF in this patient. The heterogeneity of the VSEF isolate was observed only by E-test, and standardized automated microscan was not adequate for detecting the heteroresistance. The vanA gene was present in both the susceptible and the resistant subcolonies. PFGE analysis also showed that all colonies evaluated were in the same PFGE group, eliminating the possibility that mixed cultures were responsible for the findings observed. The mechanism for resistance was undetermined.

Since heteroresistance to vancomycin in enterococci has not been reported previously, it is not known what percentage of VSEF isolates are heteroresistant and what clinical impact this heteroresistance may have. Further studies are needed to determine not only the frequency of infections due to hetero-VREF but also the clinical impact of infections caused by these isolates. If the use of vancomycin may lead either to the selection of resistant subpopulations that subsequently cause VREF infections or to treatment failure, performance of E-tests on clinically significant isolates may be appropriate. Since VRE infections have been associated with high mortality rates, early therapeutic intervention may decrease the mortality, the morbidity, and the cost associated with treatment of VRE. The results of the present study have important implications for patients with serious enterococcal infection where treatment with vancomycin is being considered.