Journal of Antimicrobial Chemotherapy (1992) 30, 489-496
Frequencies of subpopulations of amiooglycoside- and vancomycinresistant variants in Staphylococcus awreus and Staphylococcus epidermidis G. R&dberg, L. E. Nilsson, E. Klhlstrtm, R. MaDer and L. Soren
Selection and regrowth of resistant variants, which are present in low frequencies in the initial inoculum, were seen when large inocula of five strains of Staphylococcus aureus and four strains of Staphylococcus epidermidis were incubated in broth with amilcacin, gentamicin, netilmicin and tobramycin. Statistical analysis showed no significant difference between the aminoglycosides in the selective growth of resistant variants (P > 05). Vancomycin differed significantly from the aminoglycosides in both the frequency of, and selection of resistant variants (P < 0-001). No bacteria resistant to > 1 x MIC was seen in the vancomycin-exposed cultures of S. aureus and S. epidermidis, while in most aminoglycoside-exposed cultures, bacteria resistant to 4-16 x MIC were seen. Introduction Cultures of Gram-positive and Gram-negative bacteria contain subpopulations which are less susceptible to aminoglycosides than the rest of the culture (Wise & Spink, 1954; Quie, 1969; Musher et al., 1977; Pelletier, Richardson & Feist, 1979; Soren & Nilsson, 1984; Nilsson & Soren, 1986; Nilsson, Soren & Ridberg, 1987). Incubation of large inocula of Enterobacteriaceae and Pseudomonas aeruginosa with amilcacin, gentamicin, netilmicin or tobramycin results in a rapid and considerable initial killing which is followed by regrowth during prolonged incubation. This regrowth is due to the selection and multiplication of resistant variants, which are present in low frequencies in the initial inoculum (Soren & Nilsson, 1984; Nilsson & Soren, 1986; Nilsson et al., 1987). Attempts to select and characterize subpopulations of vancomycin-resistant variants in staphylococci have been relatively unsuccessful (Archer, 1978; Watanakunakorn, 1988, 1990). To our knowledge no study, that compares the occurrence and frequencies of subpopulations of resistant variants in Staphylococcus aureus or Staphylococcus epidermidis for different aminoglycosides, has been performed. The aim of this study was to compare different aminoglycosides, and to test and compare vancomycin with the aminoglycosides for the selection and regrowth of resistant variants. Materials and methods Bacterial strains Five strains of S. aureus and four strains of S. epidermidis recovered from randomly selected blood cultures were used. 489 03O5-7453/92/10O489+08 $08.00/0
© 1992 The British Society for Antimicrobial Chemotherapy
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
Department of Clinical Microbiology, University Hospital, S-S81 85 Lmkoping, Sweden
490
G. RMberg et aL
Antibiotics Stock aqueous solutions of 1000 mg/L of active drug were prepared from amikacin base (Bristol Laboratories, Syracuse, NY, USA), gentamicin sulphate (Schering Corp., Bloomfield, NJ, USA), netilmicin sulphate (Schering Corp.). tobramycin sulphate (Eli Lilly and Co., Indianapolis, IN, USA) and vancomycin (Eli Lilly and Co.). These stock solutions were kept in a frozen state at — 20°C. MIC determination
Bacterial population analysis Large inocula (107 cfu/mL) were incubated in supplemented Mueller-Hinton broth without antibiotic or with serial dilutions of amikacin, gentamicin, netilmicin, tobramycin and vancomycin respectively for 24 h at 37°C. After thorough mixing 100/iL was taken from cultures without antibiotic and 100 /iL from the tube with the highest antibiotic concentration which showed visible growth (1/2 MIC). These samples were diluted serially in 0-9 mL of physiological saline and 100 /iL from appropriate dilutions were placed on Mueller-Hinton agar containing different concentrations of aminoglycoside or vancomycin. The drops were allowed to diffuse and dry at room temperature, then the plates were incubated overnight at 37°C. The colonies were counted and the number of viable bacteria calculated. The frequencies of variants resistant to different concentrations of amikacin, gentamicin, netilmicin, tobramycin and vancoTable. The minimum inhibitory concentrations (MIC) of amikacin, gentamicin, netilmicin, tobramycin and vancomycin in agar Tor five strains of S. aureus and four strains of S. epidermidis with inocula of 104cfu/spot Strains
gentamicin
MIC (mg/L) tobramycin
2
05 025 025 006 05
1 05 1 05
025 006 006 006
amikacin
netilmicin
vancomycin
025 025 025 025 025
05 05
05
1 05 1 2 1
4 006 O06 006
4 025 012 012
2 2 1 2
S. aureus LU5
LU 31 LU 32 LU39 LU47 S. epidermidis LU 15 LU 16 LU44 LU 48
2 1 4 2
05 012
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
Mueller-Hinton agar plates (Difco Laboratories, Detroit, MI, USA), containing serial two-fold dilutions of the different aminoglycosides or vancomycin were prepared. Overnight cultures of the bacteria in Mueller-Hinton broth (viable count ~ 10s cfu/mL) were diluted in physiological saline to approximately 10s cfu/mL and samples (drops of 100/iL) were manually placed on the antibiotic-containing agar plates, allowed to dry at room temperature, and incubated 24 h at 37°C. The MIC of each agent was interpreted as the lowest concentration to inhibit growth.
Aminoglycoside- and rancomydn-resistaiit variants in staphyiococci
491
mycin were calculated by dividing the number of colonies on plates with antibiotic by the number of colonies on plates without antibiotic. All population analysis were performed twice. Statistical methods
Results MIC values The MIC values of the aminoglycosides and vancomycin for 5. aureus and 5. epidermidis are shown in the Table. Population analysis For 5. aureus LU 47 (Figure l(a)) and S. epidermidis LU 44 (Figure l(b)) similar frequencies of resistant variants were obtained for gentamicin and tobramycin at the same concentrations. The frequency curve of variants resistant to netilmicin was shifted towards higher concentrations of the drug, and the frequency curve of amikacin was shifted towards even higher concentrations (Figure 1). The frequency of bacteria resistant to 1 x MIC for 5. aureus LU 47 and S. epidermidis LU 44 were 10~4 to 10~J for all aminoglycosides, but no bacteria resistant to 1 x MIC of vancomycin were detected (Figure 1). The frequencies in S. aureus (Figures 2(a) to 6(a)) and S. epidermidis (Figures 2(b) to 6(b)), of variants resistant to different multiples of MICs are shown for amikacin (Figure 2), gentamicin (Figure 3), netilmicin (Figure 4), tobramycin (Figure 5) and vancomycin (Figure 6). The values are based on the pooled data for five strains of S. aureus and four strains of 5. epidermidis. For S. aureus and S. epidermidis not preexposed to antibiotic, low frequencies ( < lO^-lO" 6 ) of resistant variants were seen at concentrations 2-8 x MIC of all aminoglycosides (Figures 2 to 5). For vancomycin, no variants resistant to concentrations above the MIC were detected for either species (Figure 6). After incubation with 1/2 MIC of each aminoglycoside or vancomycin the selection and multiplication of aminoglycoside-resistant variants was pronounced for all aminoglycosides and both species (Figures 2 to S), but was not seen for vancomycin (Figure 6). Statistical analysis showed no significant (P > OS) difference between the aminoglycosides when comparing differences in frequencies of resistant variants obtained from unexposed cultures and cultures pre-exposed to each aminoglycoside. In contrast, vancomycin differed significantly from the aminoglycosides (P < 0-001), since pre-exposure to vancomycin did not select for resistant variants (Figure 6). No
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
A comparison of the frequencies of resistant variants of S. aureus and S. epidermidis was made for concentrations of 1/4-2 x MIC (the part of the frequency-curves depicting a declining slope). The frequencies of resistant variants in cultures incubated without antibiotic and those exposed to subinhibitory concentrations (1/2 MIC) of each tested antibiotic, were logarithmically transformed. The difference in frequencies between exposed and unexposed cultures was calculated and a comparison of these differences between the antibiotics was made, using analysis of variance. When differences between means were studied, a double-sided Student /-test was used.
492
G. RAdberg et aL
(b)
(o)
0
0
-1
s
-2
In -In
-3
-1
-2
-3
v\V
-5
-4
-5
-6
-6 O-OI5 0-06
O23
I
4
i- \\ - IV\
-, Av
0-015 O-O6
0-25
I
4
Concentration of ontibtotlc in aqar ptotes (mg/L)
Figure 1. Frequencies of variants resistant to different concentrations of amikacin ( • ) , gentamicin ( • ) , netilmicin (A), tobramycin (afr), and vancomycin (T), in cultures of S. aureus LU 47 (a) and S. epidermidis LU 44 (b) incubated without drug for 24 h. A log, 0 value of 0 is where equal numbers of colonies grew an agar with and without antibiotic.
bacteria resistant to concentrations higher than 1 x MIC were seen in the vancomycinexposed cultures of either species while in the aminoglycoside-exposed cultures bacteria resistant to 4-16 x MIC were seen for all the aminoglycosides (Figures 2 to 5). Discussion
Aminoglycoside-resistant subpopulations in Gram-negative (Gerber & Craig, 1982; Soren & Nilsson, 1984; Nilsson & Soren, 1986; Nilsson et al., 1987) and Gram-positive (Wise & Spink, 1954; Quie, 1969; Musher et al., 1977; Pelletier et al., 1979) bacteria have been previously reported. Quie (1969) showed that small colonies of resistant variants can be selected from human sources and discussed the possibility that these variants could contribute to therapeutic failures in patients treated with apparently appropriate antimicrobial therapy. Animal model studies have demonstrated that gentamicin induced small colony variants of S. aureus produced subcutaneous abscesses in rats and haematogeneous pyelonephritis in mice (Musher et al., 1977), as well as endocardial infection in rabbits, at about the same rate as the parent strain (Pelletier et al., 1979). Earlier studies have also shown that populations of P. aeruginosa contain greater numbers of variants resistant to high concentrations of amikacin, gentamicin, netilmicin and tobramycin than populations of Escherichia coli (Nilsson & Soren, 1986; Nilsson et al., 1987). In this study it was found that the frequencies of variants resistant to aminoglycosides in 5. aureus and S. epidermidis are similar to those of E. coli. Variants resistant to up to 8 mg/L of the aminoglycosides occurred when strains of 5. aureus and S. epidermidis were pre-exposed to the aminoglycosides and later tested. These data
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
l\\
-4
•u
Amlnoglycofide- and vaocomydn-redstant variants in stapfaytococd
493
I
1/64 1/16
1/4
I
4
16
1/64
1/16
1/4
Concent rot ion of amjkocm in multiples of MIC
Figure 2. Frequencies of variants resistant to different multiples of MIC of amikacin. The means of five strains of S. aureus (a) and four strains of 5. cpidermidis (b), are shown. Unex posed cultures ( # ) , and cultures exposed to subinhibitory concentrations (1/2 MIQ of anulriotic (A)- The vertical bars denote standard deviation. Note: For figures 2-6, when frequencies of resistant variants of a certain strain is below the level of detection (10"7). these strains have been excluded from calculation of the mean. (b) 0
r
%
IT I
1
-1
\
-2
-3
—4
-5
-a
1/64 1/16
s
' 1
[ \
: I i
1/64
.
i
1/16
•
i
1/4
•
j
I
i
16
Concentration of gentomicin in muHiptesof MIC
Figure 3. Frequencies of variants resistant to different multiples of MIC of gentamicin. The means of five strains of S. aureus (a) and four strains of 5. epkkrmidis (b), are shown. Unexposed cultures ( • ) , and cultures exposed to subinhibitory concentrations (1/2 MIQ of antibiotic (A). The vertical bars denote standard deviation.
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
-6
G. RAdbergrf aL
494
l""
k
0
-1
'— \,-
\1
-2
1
-3
-5
•
H 1
-6
-6
1/64
1/64
1/16
1/16
1/4
I
16
Concentration of netilmicln in multiples of MIC
Figure 4. Frequencies of variants resistant to different multiples of MIC of netilmidn. The means of five strains of S. mtreus (a) and four strains of S. epidermidis (b), are shown. Unexposed cultures ( # ) , and cultures exposed to subinhibitory concentrations (1/2 MIC) of antibiotic (A). The vertical bars denote standard deviation.
1/64
1/16
1/4
I
4
16
1/64
1/16
1/4
Concentration of tobromycin in multiple* of MIC Figure 5. Frequencies of variants resistant to different multiples of MIC of tobramycin. The means of five strains of 5. mtreus (a) and four strains of S. epidermidis (b), are shown. Unexposed cultures ( • ) and cultures exposed to subinhibitory concentrations (1/2 MIC) of antibiotic (A)- The vertical bars denote standard deviation.
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
-4
Amioogtycoside- and vaocomycfo-nsistaiit variants in staphytocood
495
(a)
o-I
-
\
f
1-3 I
-5
•
1/64
1/16
1/4
I
4
16
1/64
1/16
1/4
I
Concentration of vancomycin tn multiple! of MIC
Flgm 6. Frequencies of variants resistant to different multiples of MIC of vancomycin. The means of five •trains of S. aweus (g) and four strains of 5. epidermidis (b), are shown. Unexposed cultures ( • ) and cultures exposed to subinhibitory concentrations (1/2 MIQ of antibiotic ( • ) . The vertical bare denote standard deviation.
and earlier in-vivo data (Quie, 1969; Musher et al., 1977; Pelletier et al., 1979), indicate that single therapy with aminoglycosides in the treatment of serious infections caused by staphylococci probably should be avoided. When the frequencies of resistant variants in unexposed cultures and those exposed to subinhibitory concentrations of each aminoglycosides were compared, there was no significant difference between the aminoglycosides in selective regrowth of resistant variants in the tested species of staphylococci. These data suggest that there is no preference for any of the tested aminoglycosides in the treatment of patients with staphylococcal infections. Low level aminoglycoside resistance has been shown not to be due to aminoglycosidc inactivating enzymes (Miller, Wexler & Steigbigel, 1978). The general development of resistance for all aminoglycosides in this study indicates an impairment of aminoglycoside uptake in the selected aminoglycoside-resistant variants. It has been suggested that this resistance is due to defects in the oxidative metabolism which is required for aminoglycoside uptake in bacteria (Kaplan & Dye, 1976; Miller et al., 1978). Vancomycin has been used for many years to treat severe infections caused by staphylococci and both coagulase-positive and -negative staphylococci are still usually susceptible to this agent. In this study, vancomycin differed significantly from the aminoglycosides in frequencies and selection of resistant variants (P < 0-001) in strains of S. aureus and S. epidermidis. In fact no bacteria resistant to higher concentrations than 1 x MIC was seen in the vancomycin exposed cultures. However, there have been a few reports describing the emergence of vancomycinresistant staphylococci in vivo (Schwalbe, Stapleton & Gilligan, 1987) or in vitro (Watanakunakorn, 1988, 1990). Watanakunakorn (1988, 1990) performed several
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
-4 -
496
G. RAdberg et aL
References Archer, G. L. (1978). Antimicrobial susceptibility and selection of resistance among Staphylococcus epidermidis isolates recovered from patients with infections of indwelling foreign devices. Antimicrobial Agents and Chemotherapy 14, 353-9. Gerber, A. U. & Craig, W. A. (1982). Aminoglycoside-selected subpopulations of Pseudomonas aeruginosa: characterization and virulence in normal and leukopenic mice. Journal of Laboratory and Clinical Medicine 100, 671-81. Kaplan, M. L. & Dye, W. E. (1976). Growth requirements of some small-colony-forming variants of Staphylococcus aureus. Journal of Clinical Microbiology 4, 343-8. Miller, M. H., Wexler, M. A. & Steigbigel, N. H. (1978). Single and combination antibiotic therapy of Staphylococcus aureus experimental endocarditis: emergence of gentamicinresistant mutants. Antimicrobial Agents and Chemotherapy 14, 336-43. Musher, D. M., Baughn, R. E., Templeton, G. B. & Minuth, J. N. (1977). Emergence of variant forms of Staphylococcus aureus after exposure to gentamicin and infectivity of the variants in experimental animals. Journal of Infectious Diseases 136, 360-9. Nilsson, L. & Soren, L. (1986). Selective growth of resistant variants during incubation of Enterobacteriaceae with four aminoglycosides. Journal of Antimicrobial Chemotherapy 18, 317-24. Nilsson, L., S6ren, L. & Radberg, G. (1987). Frequencies of variants resistant to different aminoglycosides in Pseudomonas aeruginosa. Journal of Antimicrobial Chemotherapy 20, 255-9. Pelletier, L. L., Richardson, M. & Feist, M. (1979). Virulent gentamicin-induced small colony variants of Staphylococcus aureus. Journal of Laboratory and Clinical Medicine 94, 324-34. Quie, P. G. (1969). Microcolonies (G-variants) of Staphylococcus aureus. Yale Journal of Biology and Medicine 41, 394-403. Schwalbe, R. S., Stapleton, J. T. & Gilligan, P. H. (1987). Emergence of vancomycin resistance in coagulase-negative staphylococci. New England Journal of Medicine 316, 927-31. Sdren, L. & Nilsson, L. (1984). Regrowth of aminoglycoside-resistant variants and its possible implication for determinations of MICs. Antimicrobial Agents and Chemotherapy 26, 501-6. Watanakunakorn, C. (1988). In-vitro induction of resistance in coagulase-negative staphylococci to vancomycin and teicoplanin. Journal of Antimicrobial Chemotherapy 22, 321-4. Watanakunakorn, C. (1990). In-vitro selection of resistance of Staphylococcus aureus to teicoplanin and vancomycin. Journal of Antimicrobial Chemotherapy 25, 69-72. Wise, R. I. & Spink, W. W. (1954). The influence of antibiotics on the origin of small colonies (G variants) of Micrococcus pyogenes var. aureus. Journal of Clinical Investigation 33, 1611-22. (Received 20 January 1992; revised version accepted 29 April 1992)
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
passages of coagulasc-positive and -negative staphylococci in subinhibitory concentrations of vancomycin. Only a few strains were found to be four-fold less susceptible, but only for coagulase-negative staphylococci was this increase stable. The mechanism by which vancomycin-resistance develops remains unclear. The in-vitro studies of Watanakunakorn (1988, 1990) stimulates the conditions that may occur in vivo, when only subinhibitory concentrations reach the focus of infection during long-term treatment. This may occur under certain circumstances such as the presence of a foreign device in a patient, when bacteria produce extracellular glycocalyx acting as a permeability barrier to antibiotics and antibiotic regimens with long intervals between the doses. These factors may result in prolonged exposure of the bacteria to antibiotic concentrations below the MIC. In spite of the hitherto few reports of development of resistance to vancomycin in staphylococci this possibility should be considered in clinical conditions. In conclusion, there was no difference between the aminoglycosides in the occurrence and selective growth of resistant variants of 5. aureus and S. epidermidis, while vancomycin differed significantly from the aminoglycosides in that respect.
Frequencies of subpopulations of amiooglycoside- and vancomycinresistant variants in Staphylococcus awreus and Staphylococcus epidermidis G. R&dberg, L. E. Nilsson, E. Klhlstrtm, R. MaDer and L. Soren
Selection and regrowth of resistant variants, which are present in low frequencies in the initial inoculum, were seen when large inocula of five strains of Staphylococcus aureus and four strains of Staphylococcus epidermidis were incubated in broth with amilcacin, gentamicin, netilmicin and tobramycin. Statistical analysis showed no significant difference between the aminoglycosides in the selective growth of resistant variants (P > 05). Vancomycin differed significantly from the aminoglycosides in both the frequency of, and selection of resistant variants (P < 0-001). No bacteria resistant to > 1 x MIC was seen in the vancomycin-exposed cultures of S. aureus and S. epidermidis, while in most aminoglycoside-exposed cultures, bacteria resistant to 4-16 x MIC were seen. Introduction Cultures of Gram-positive and Gram-negative bacteria contain subpopulations which are less susceptible to aminoglycosides than the rest of the culture (Wise & Spink, 1954; Quie, 1969; Musher et al., 1977; Pelletier, Richardson & Feist, 1979; Soren & Nilsson, 1984; Nilsson & Soren, 1986; Nilsson, Soren & Ridberg, 1987). Incubation of large inocula of Enterobacteriaceae and Pseudomonas aeruginosa with amilcacin, gentamicin, netilmicin or tobramycin results in a rapid and considerable initial killing which is followed by regrowth during prolonged incubation. This regrowth is due to the selection and multiplication of resistant variants, which are present in low frequencies in the initial inoculum (Soren & Nilsson, 1984; Nilsson & Soren, 1986; Nilsson et al., 1987). Attempts to select and characterize subpopulations of vancomycin-resistant variants in staphylococci have been relatively unsuccessful (Archer, 1978; Watanakunakorn, 1988, 1990). To our knowledge no study, that compares the occurrence and frequencies of subpopulations of resistant variants in Staphylococcus aureus or Staphylococcus epidermidis for different aminoglycosides, has been performed. The aim of this study was to compare different aminoglycosides, and to test and compare vancomycin with the aminoglycosides for the selection and regrowth of resistant variants. Materials and methods Bacterial strains Five strains of S. aureus and four strains of S. epidermidis recovered from randomly selected blood cultures were used. 489 03O5-7453/92/10O489+08 $08.00/0
© 1992 The British Society for Antimicrobial Chemotherapy
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
Department of Clinical Microbiology, University Hospital, S-S81 85 Lmkoping, Sweden
490
G. RMberg et aL
Antibiotics Stock aqueous solutions of 1000 mg/L of active drug were prepared from amikacin base (Bristol Laboratories, Syracuse, NY, USA), gentamicin sulphate (Schering Corp., Bloomfield, NJ, USA), netilmicin sulphate (Schering Corp.). tobramycin sulphate (Eli Lilly and Co., Indianapolis, IN, USA) and vancomycin (Eli Lilly and Co.). These stock solutions were kept in a frozen state at — 20°C. MIC determination
Bacterial population analysis Large inocula (107 cfu/mL) were incubated in supplemented Mueller-Hinton broth without antibiotic or with serial dilutions of amikacin, gentamicin, netilmicin, tobramycin and vancomycin respectively for 24 h at 37°C. After thorough mixing 100/iL was taken from cultures without antibiotic and 100 /iL from the tube with the highest antibiotic concentration which showed visible growth (1/2 MIC). These samples were diluted serially in 0-9 mL of physiological saline and 100 /iL from appropriate dilutions were placed on Mueller-Hinton agar containing different concentrations of aminoglycoside or vancomycin. The drops were allowed to diffuse and dry at room temperature, then the plates were incubated overnight at 37°C. The colonies were counted and the number of viable bacteria calculated. The frequencies of variants resistant to different concentrations of amikacin, gentamicin, netilmicin, tobramycin and vancoTable. The minimum inhibitory concentrations (MIC) of amikacin, gentamicin, netilmicin, tobramycin and vancomycin in agar Tor five strains of S. aureus and four strains of S. epidermidis with inocula of 104cfu/spot Strains
gentamicin
MIC (mg/L) tobramycin
2
05 025 025 006 05
1 05 1 05
025 006 006 006
amikacin
netilmicin
vancomycin
025 025 025 025 025
05 05
05
1 05 1 2 1
4 006 O06 006
4 025 012 012
2 2 1 2
S. aureus LU5
LU 31 LU 32 LU39 LU47 S. epidermidis LU 15 LU 16 LU44 LU 48
2 1 4 2
05 012
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
Mueller-Hinton agar plates (Difco Laboratories, Detroit, MI, USA), containing serial two-fold dilutions of the different aminoglycosides or vancomycin were prepared. Overnight cultures of the bacteria in Mueller-Hinton broth (viable count ~ 10s cfu/mL) were diluted in physiological saline to approximately 10s cfu/mL and samples (drops of 100/iL) were manually placed on the antibiotic-containing agar plates, allowed to dry at room temperature, and incubated 24 h at 37°C. The MIC of each agent was interpreted as the lowest concentration to inhibit growth.
Aminoglycoside- and rancomydn-resistaiit variants in staphyiococci
491
mycin were calculated by dividing the number of colonies on plates with antibiotic by the number of colonies on plates without antibiotic. All population analysis were performed twice. Statistical methods
Results MIC values The MIC values of the aminoglycosides and vancomycin for 5. aureus and 5. epidermidis are shown in the Table. Population analysis For 5. aureus LU 47 (Figure l(a)) and S. epidermidis LU 44 (Figure l(b)) similar frequencies of resistant variants were obtained for gentamicin and tobramycin at the same concentrations. The frequency curve of variants resistant to netilmicin was shifted towards higher concentrations of the drug, and the frequency curve of amikacin was shifted towards even higher concentrations (Figure 1). The frequency of bacteria resistant to 1 x MIC for 5. aureus LU 47 and S. epidermidis LU 44 were 10~4 to 10~J for all aminoglycosides, but no bacteria resistant to 1 x MIC of vancomycin were detected (Figure 1). The frequencies in S. aureus (Figures 2(a) to 6(a)) and S. epidermidis (Figures 2(b) to 6(b)), of variants resistant to different multiples of MICs are shown for amikacin (Figure 2), gentamicin (Figure 3), netilmicin (Figure 4), tobramycin (Figure 5) and vancomycin (Figure 6). The values are based on the pooled data for five strains of S. aureus and four strains of 5. epidermidis. For S. aureus and S. epidermidis not preexposed to antibiotic, low frequencies ( < lO^-lO" 6 ) of resistant variants were seen at concentrations 2-8 x MIC of all aminoglycosides (Figures 2 to 5). For vancomycin, no variants resistant to concentrations above the MIC were detected for either species (Figure 6). After incubation with 1/2 MIC of each aminoglycoside or vancomycin the selection and multiplication of aminoglycoside-resistant variants was pronounced for all aminoglycosides and both species (Figures 2 to S), but was not seen for vancomycin (Figure 6). Statistical analysis showed no significant (P > OS) difference between the aminoglycosides when comparing differences in frequencies of resistant variants obtained from unexposed cultures and cultures pre-exposed to each aminoglycoside. In contrast, vancomycin differed significantly from the aminoglycosides (P < 0-001), since pre-exposure to vancomycin did not select for resistant variants (Figure 6). No
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
A comparison of the frequencies of resistant variants of S. aureus and S. epidermidis was made for concentrations of 1/4-2 x MIC (the part of the frequency-curves depicting a declining slope). The frequencies of resistant variants in cultures incubated without antibiotic and those exposed to subinhibitory concentrations (1/2 MIC) of each tested antibiotic, were logarithmically transformed. The difference in frequencies between exposed and unexposed cultures was calculated and a comparison of these differences between the antibiotics was made, using analysis of variance. When differences between means were studied, a double-sided Student /-test was used.
492
G. RAdberg et aL
(b)
(o)
0
0
-1
s
-2
In -In
-3
-1
-2
-3
v\V
-5
-4
-5
-6
-6 O-OI5 0-06
O23
I
4
i- \\ - IV\
-, Av
0-015 O-O6
0-25
I
4
Concentration of ontibtotlc in aqar ptotes (mg/L)
Figure 1. Frequencies of variants resistant to different concentrations of amikacin ( • ) , gentamicin ( • ) , netilmicin (A), tobramycin (afr), and vancomycin (T), in cultures of S. aureus LU 47 (a) and S. epidermidis LU 44 (b) incubated without drug for 24 h. A log, 0 value of 0 is where equal numbers of colonies grew an agar with and without antibiotic.
bacteria resistant to concentrations higher than 1 x MIC were seen in the vancomycinexposed cultures of either species while in the aminoglycoside-exposed cultures bacteria resistant to 4-16 x MIC were seen for all the aminoglycosides (Figures 2 to 5). Discussion
Aminoglycoside-resistant subpopulations in Gram-negative (Gerber & Craig, 1982; Soren & Nilsson, 1984; Nilsson & Soren, 1986; Nilsson et al., 1987) and Gram-positive (Wise & Spink, 1954; Quie, 1969; Musher et al., 1977; Pelletier et al., 1979) bacteria have been previously reported. Quie (1969) showed that small colonies of resistant variants can be selected from human sources and discussed the possibility that these variants could contribute to therapeutic failures in patients treated with apparently appropriate antimicrobial therapy. Animal model studies have demonstrated that gentamicin induced small colony variants of S. aureus produced subcutaneous abscesses in rats and haematogeneous pyelonephritis in mice (Musher et al., 1977), as well as endocardial infection in rabbits, at about the same rate as the parent strain (Pelletier et al., 1979). Earlier studies have also shown that populations of P. aeruginosa contain greater numbers of variants resistant to high concentrations of amikacin, gentamicin, netilmicin and tobramycin than populations of Escherichia coli (Nilsson & Soren, 1986; Nilsson et al., 1987). In this study it was found that the frequencies of variants resistant to aminoglycosides in 5. aureus and S. epidermidis are similar to those of E. coli. Variants resistant to up to 8 mg/L of the aminoglycosides occurred when strains of 5. aureus and S. epidermidis were pre-exposed to the aminoglycosides and later tested. These data
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
l\\
-4
•u
Amlnoglycofide- and vaocomydn-redstant variants in stapfaytococd
493
I
1/64 1/16
1/4
I
4
16
1/64
1/16
1/4
Concent rot ion of amjkocm in multiples of MIC
Figure 2. Frequencies of variants resistant to different multiples of MIC of amikacin. The means of five strains of S. aureus (a) and four strains of 5. cpidermidis (b), are shown. Unex posed cultures ( # ) , and cultures exposed to subinhibitory concentrations (1/2 MIQ of anulriotic (A)- The vertical bars denote standard deviation. Note: For figures 2-6, when frequencies of resistant variants of a certain strain is below the level of detection (10"7). these strains have been excluded from calculation of the mean. (b) 0
r
%
IT I
1
-1
\
-2
-3
—4
-5
-a
1/64 1/16
s
' 1
[ \
: I i
1/64
.
i
1/16
•
i
1/4
•
j
I
i
16
Concentration of gentomicin in muHiptesof MIC
Figure 3. Frequencies of variants resistant to different multiples of MIC of gentamicin. The means of five strains of S. aureus (a) and four strains of 5. epkkrmidis (b), are shown. Unexposed cultures ( • ) , and cultures exposed to subinhibitory concentrations (1/2 MIQ of antibiotic (A). The vertical bars denote standard deviation.
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
-6
G. RAdbergrf aL
494
l""
k
0
-1
'— \,-
\1
-2
1
-3
-5
•
H 1
-6
-6
1/64
1/64
1/16
1/16
1/4
I
16
Concentration of netilmicln in multiples of MIC
Figure 4. Frequencies of variants resistant to different multiples of MIC of netilmidn. The means of five strains of S. mtreus (a) and four strains of S. epidermidis (b), are shown. Unexposed cultures ( # ) , and cultures exposed to subinhibitory concentrations (1/2 MIC) of antibiotic (A). The vertical bars denote standard deviation.
1/64
1/16
1/4
I
4
16
1/64
1/16
1/4
Concentration of tobromycin in multiple* of MIC Figure 5. Frequencies of variants resistant to different multiples of MIC of tobramycin. The means of five strains of 5. mtreus (a) and four strains of S. epidermidis (b), are shown. Unexposed cultures ( • ) and cultures exposed to subinhibitory concentrations (1/2 MIC) of antibiotic (A)- The vertical bars denote standard deviation.
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
-4
Amioogtycoside- and vaocomycfo-nsistaiit variants in staphytocood
495
(a)
o-I
-
\
f
1-3 I
-5
•
1/64
1/16
1/4
I
4
16
1/64
1/16
1/4
I
Concentration of vancomycin tn multiple! of MIC
Flgm 6. Frequencies of variants resistant to different multiples of MIC of vancomycin. The means of five •trains of S. aweus (g) and four strains of 5. epidermidis (b), are shown. Unexposed cultures ( • ) and cultures exposed to subinhibitory concentrations (1/2 MIQ of antibiotic ( • ) . The vertical bare denote standard deviation.
and earlier in-vivo data (Quie, 1969; Musher et al., 1977; Pelletier et al., 1979), indicate that single therapy with aminoglycosides in the treatment of serious infections caused by staphylococci probably should be avoided. When the frequencies of resistant variants in unexposed cultures and those exposed to subinhibitory concentrations of each aminoglycosides were compared, there was no significant difference between the aminoglycosides in selective regrowth of resistant variants in the tested species of staphylococci. These data suggest that there is no preference for any of the tested aminoglycosides in the treatment of patients with staphylococcal infections. Low level aminoglycoside resistance has been shown not to be due to aminoglycosidc inactivating enzymes (Miller, Wexler & Steigbigel, 1978). The general development of resistance for all aminoglycosides in this study indicates an impairment of aminoglycoside uptake in the selected aminoglycoside-resistant variants. It has been suggested that this resistance is due to defects in the oxidative metabolism which is required for aminoglycoside uptake in bacteria (Kaplan & Dye, 1976; Miller et al., 1978). Vancomycin has been used for many years to treat severe infections caused by staphylococci and both coagulase-positive and -negative staphylococci are still usually susceptible to this agent. In this study, vancomycin differed significantly from the aminoglycosides in frequencies and selection of resistant variants (P < 0-001) in strains of S. aureus and S. epidermidis. In fact no bacteria resistant to higher concentrations than 1 x MIC was seen in the vancomycin exposed cultures. However, there have been a few reports describing the emergence of vancomycinresistant staphylococci in vivo (Schwalbe, Stapleton & Gilligan, 1987) or in vitro (Watanakunakorn, 1988, 1990). Watanakunakorn (1988, 1990) performed several
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
-4 -
496
G. RAdberg et aL
References Archer, G. L. (1978). Antimicrobial susceptibility and selection of resistance among Staphylococcus epidermidis isolates recovered from patients with infections of indwelling foreign devices. Antimicrobial Agents and Chemotherapy 14, 353-9. Gerber, A. U. & Craig, W. A. (1982). Aminoglycoside-selected subpopulations of Pseudomonas aeruginosa: characterization and virulence in normal and leukopenic mice. Journal of Laboratory and Clinical Medicine 100, 671-81. Kaplan, M. L. & Dye, W. E. (1976). Growth requirements of some small-colony-forming variants of Staphylococcus aureus. Journal of Clinical Microbiology 4, 343-8. Miller, M. H., Wexler, M. A. & Steigbigel, N. H. (1978). Single and combination antibiotic therapy of Staphylococcus aureus experimental endocarditis: emergence of gentamicinresistant mutants. Antimicrobial Agents and Chemotherapy 14, 336-43. Musher, D. M., Baughn, R. E., Templeton, G. B. & Minuth, J. N. (1977). Emergence of variant forms of Staphylococcus aureus after exposure to gentamicin and infectivity of the variants in experimental animals. Journal of Infectious Diseases 136, 360-9. Nilsson, L. & Soren, L. (1986). Selective growth of resistant variants during incubation of Enterobacteriaceae with four aminoglycosides. Journal of Antimicrobial Chemotherapy 18, 317-24. Nilsson, L., S6ren, L. & Radberg, G. (1987). Frequencies of variants resistant to different aminoglycosides in Pseudomonas aeruginosa. Journal of Antimicrobial Chemotherapy 20, 255-9. Pelletier, L. L., Richardson, M. & Feist, M. (1979). Virulent gentamicin-induced small colony variants of Staphylococcus aureus. Journal of Laboratory and Clinical Medicine 94, 324-34. Quie, P. G. (1969). Microcolonies (G-variants) of Staphylococcus aureus. Yale Journal of Biology and Medicine 41, 394-403. Schwalbe, R. S., Stapleton, J. T. & Gilligan, P. H. (1987). Emergence of vancomycin resistance in coagulase-negative staphylococci. New England Journal of Medicine 316, 927-31. Sdren, L. & Nilsson, L. (1984). Regrowth of aminoglycoside-resistant variants and its possible implication for determinations of MICs. Antimicrobial Agents and Chemotherapy 26, 501-6. Watanakunakorn, C. (1988). In-vitro induction of resistance in coagulase-negative staphylococci to vancomycin and teicoplanin. Journal of Antimicrobial Chemotherapy 22, 321-4. Watanakunakorn, C. (1990). In-vitro selection of resistance of Staphylococcus aureus to teicoplanin and vancomycin. Journal of Antimicrobial Chemotherapy 25, 69-72. Wise, R. I. & Spink, W. W. (1954). The influence of antibiotics on the origin of small colonies (G variants) of Micrococcus pyogenes var. aureus. Journal of Clinical Investigation 33, 1611-22. (Received 20 January 1992; revised version accepted 29 April 1992)
Downloaded from http://jac.oxfordjournals.org/ at RMIT University Library on July 2, 2015
passages of coagulasc-positive and -negative staphylococci in subinhibitory concentrations of vancomycin. Only a few strains were found to be four-fold less susceptible, but only for coagulase-negative staphylococci was this increase stable. The mechanism by which vancomycin-resistance develops remains unclear. The in-vitro studies of Watanakunakorn (1988, 1990) stimulates the conditions that may occur in vivo, when only subinhibitory concentrations reach the focus of infection during long-term treatment. This may occur under certain circumstances such as the presence of a foreign device in a patient, when bacteria produce extracellular glycocalyx acting as a permeability barrier to antibiotics and antibiotic regimens with long intervals between the doses. These factors may result in prolonged exposure of the bacteria to antibiotic concentrations below the MIC. In spite of the hitherto few reports of development of resistance to vancomycin in staphylococci this possibility should be considered in clinical conditions. In conclusion, there was no difference between the aminoglycosides in the occurrence and selective growth of resistant variants of 5. aureus and S. epidermidis, while vancomycin differed significantly from the aminoglycosides in that respect.
