The rapid emergence of high level gentamicin resistance in enterococci

The proportion of enterococci isolated from blood and urine culture that were highly resistant to gentamicin and streptomycin were determined. No blood or urine isolates highly resistant to gentamicin were seen in 1983. whereas by 1986-87 25% of blood and 17% of urine isolates were highly resistant. The ENTEROCOCCI REPRESENT AN INCREASINGLY IMPORtant cause of infection, especially in hospitalized patients. In recent years, there has been an increase in both the absolute number of infections and the proportion of nosocomial infections due to Enterococcus species (1-3). En terococcal bacteremia is the cau se of significant morbidity and mortality, especially when associated with endoDepartments of Laboratory Medicine and Medicine. St Boniface General Hospital: and Departments of Medica l Microbiology and Medicine. University of Manitoba. Winnipeg.

carditis or meningitis (2,(4)(5)(6)(7).Life threatening enterococcal infections are usually treated with a combination of an aminoglycoside and either a penicillin or vancomycin (2) .These combinations have a lso been used to prevent enterococcal endocarditis in predisposed patients following genitourinary manipulation.
In recent years.strains highly resistant to aminoglycosides h ave been reported with increasing frequ ency (8)(9)(10)(11)(12)(13) .These strains are not susceptible to the synergistic action of an aminoglycosid e and either a penicillin or vancomycin (13).In order to determine whether such strains were prevalent in Manitoba, a study of blood and urine isolates recovered at two large Winnipeg teaching hospitals over a five year period was conducted.

MATERIALS AND METHODS
Enterococcus species isolated from patients cared for at St Boniface General Hospital and the Health S ciences Centre over a six year period ( 1982-87) were studied.Blood a nd urinary isolates had been collected and stocked in skim milk at -70°C.Urinary isolates were predominantly from patients seen in outpatient infectious diseases clinics.Duplicate isolates from the same patient were excluded from analysis.Isolates were speciated using the identification scheme of FalkJam (14).Susceptibilities to gentamicin and streptomycin were determined using th e agar dilution method on Mueller-Hinton agar (Scott Laboratories Inc .Carson.California).Strain were considered to possess high level resistance to gentamicin if th e minimal inhibitory concentration (MIC) was in excess of 500 mg/L and to streptomycin if the MIC was in excess of 2000 mg /L.In order to validate results of susceptibility testing.time-kill synergy studies were performed on strains isolated from blood with MICs within two dilutions of the breakpoint.Time-kill synergy studies were performed in cation-supplemented 40 ------

YEAR
Mueller-Hinton broth using gentamicin 5 mg/L or streptomycin 15 mg/L, each with penicillin 10 mg/L.Synergy was said to be present when there was a 100-fold reduction in colony count in th e presence of both agents, compared with each individually.

RESULTS
Susceptibilities of 149 blood cu lture isolates and 153 urinary isolates of Enterococcus species were determined .Ninety-three per cent of blood and 98% of urine isolates were revived from stock cultures.Results are shown in Table 1 and Figure 1.One hundred and thirty blood culture isolates were Ejaecalis, 15 were Ejaecium and four were other species.Of the 153 urinary isolates.14 7 were E jaecalis, three were E Jaecium and three were other species.In the first two years of the study, there were no blood culture isolates highly resistant to gentamicin.In 1986 and 1987.17 of 69 isolates (24.6%) recovered from blood were resistant to gentamicin.During the same period of time, nine of 54 strains (16.7%) from urine were highly resistant to gentamicin.Three of 16 blood culture isolates ( 18. 7%) were highly resistant to streptomycin in 1983, whereas 10 of 24 (41.6%) were highly resistant by 1987.The proportion of isolates from urine highly resistant to streptomycin did not change significantly over U1e period studied.and ranged from 16.7% in 1983 to 40.7% in 1987 2 (X 3.55; P=0.62).
Of the 27 gentamicin-resistant E jaecalis strains recovered from blood , 24 were highly resistant to streptomycin; three were not.There were 13 urinary isolates highly resistant to gentamicin, eight of which were a lso highly streptomycinresistant.In each case, when penicillin -gentamicin synergy testing was performed.r esults confirmed that the MICs used correctly identified highly resistant strains.

DISCUSSION
The use of both an a minoglycoside and a penicillin together h as been considered standard therapy for the treatment of enterococcal infections since synergy between both classes of antibiotic was first d escribed ( 15) .Strains highly resistant to streptomycin h ave been recognized for a number of years.However, strains highly resistant to gentamicin were not r ecognized prior to 1979 (8).Since that time, enterococci highly resistant to gentamicin have been recognized with increasing frequency.These strains produce aminoglycoside-inactivating enzymes capable of adenylation, phosphorylation or acetylation of aminoglycosides and render th em resistant to the synergistic interaction of aminoglycoside and cell wall-active agents such as th e penicillins and vancomycin ( 16).
Patients with serious infections due to highly resistant Enterococcus species may not respond to combined therapy and may be unnecessarily exposed to potentially toxic aminoglycosides.Therefore.such isolates should be identified by the clinical laboratory as highly resistant.
The usual breakpoints for determining susceptibility of other Gram-positive and Gram-negative bacteria are inapproptiate for determining the presence of aminoglycoside high level resistance (17).Other methods have been developed to detect this form of resistance.Th e firs t is the use of a disk diffusion test with high potency antibiotic-impregnated disks (1 8 , 19).These disks are not yet commercially available and must be produced inhouse.A second method involves determining the ability of enterococci to grow in a broth containing either 500 mg/L of gentamicin or 2000 mg/L of streptomycin ( 1 7) .This m ethod has been incorporated into at least one commercially available system (American Microscan Inc, Sacramento.California).
The prevalence of highly resistant enterococci in Canadian hospitals has not been well established.Mederski-Samoraj et a l (20) found that 15.5% of blood culture isolates a nd 12.5% of isolates from other sterile body sites were highly resistant.to genta micin.They noted that prevalence varied considerably between hospitals.Clinicians shou ld b e made aware of the presen ce of strains highly resistant to a minoglycosides, especia lly those isolated from blood and cerebrospina l fluid.It is important to note that three blood culture isolates highly resistant to gentamicin were susceptible to streptomycin a nd that streptomycin can probably be u sed in combination with a penicillin when severely ill patients a re encountered.Although altern ative treatment regimens have not yet been established, other therapies may be considered in such cases.No two agents have shown synergy against strains of E jaecaLis highly resistant to a minoglycosides .Studies in the authors' laboratory have shown that th e new lipopeptide a ntibiotic da ptomycin tested at 5 mg/L is more rapidly bactericidal than penicillin, vancomycin or teicoplanin in vitro (unpublished observations).Wh eth er this observation will equate to in vivo effi cacy in huma ns remains to be determined.Clinicia ns caring for serious ly ill patients with infection du e to highly resistant enterococci may choose to administer longer courses of a ntibiotics, may be more likely to intervene surgically and may provide more careful follow-up.
In summary.highly resistant E jaecaLis were recovered with increasing frequency from both th e urinary tract of a predominantly outpatient population and from th e blood of hospita lized patients.Clinicians should be aware of this emerging problem and laboratories should provide appropriate susceptibility data on enterococci isolated from blood or cerebrospinal fluid , or repeatedly from other extra-u rinary sites in order to optimize the care of infected patients.Treatment of urinary tract infections du e to enterococci usually does not require an a minoglycoside.and screening for high level resistance is probably unnecessary in this setting.

TABLE 1
Summary of enterococcal strains tested