206
In Vitro Bactericidal Activity of Amoxicillin, Gentamicin, Rifampicin, Ciprofloxacin and Trimethoprim-Sulfamethoxazole Alone or in Combination against Listeria monocytogenes A. B o i s i v o n 1., C. G u i o m a r 1, C. C a r b o n 2 The in vitro bactericidal activity of amoxiciilin, gentamicin, rifampicin, ciprofloxacin and trimethoprim-sulfamethoxazole alone or in combination was determined against seven strains of Listeria monocytogenes by the killing curve method. Amoxicillin plus gentamicin was the most rapidly bactericidal combination, whereas trimethoprim-sulfamethoxazole was less bactericidal at 6 h but as bactericidal at 24 h. The combination of trimethoprim-sulfamethoxazole with either amoxicillin, ciprofloxacin or rifampicin did not result in antagonism, but the combinations were no more active than trimethoprim-suifamethoxazole alone. The interaction of amoxicillin with rifampin was fairly antagonistic (1 log10 difference). The combination of amoxicillin and ciprofloxacin, although producing antagonism during the first 6 h, was more active at 24 h than amoxiciilin alone and prevented the regrowth observed with ciprofloxacin alone. Ciprofloxacin and rifampicin interacted antagonistically during the first 6 h, and the combination was not very bactericidal (3 log10) but prevented the emergence of mutants, as observed with each drug alone, when used at concentrations greater than the MICs for the strain tested. These regimens merit evaluation in in vivo models of Listeria monocytogenes meningitis. Listeria monocytogenes infects both the immunocompromised and non-immunocomproraised host. Clinical infections include bacteremia and meningoencephalitis. The therapy of choice for Listeria monocytogenes infections is ampicillin or a combination of ampicillin with an aminoglycoside, as has been demonstrated in vitro and clinically (1). The prognosis of the most severe forms of meningoencephalitis re1Laboratoire de Microbiologie, Centre Hospitalier de Saint Germain-en-Laye, 78104 Saint Germain-en-Laye, France. 2Service de M6decine Interne, Hbpital Bichat. Universit6 Paris VII, 75018Paris, France.
Eur. J. Clin. Microbiol. Infect Dis.
mains poor. Thus alternative regimens have to be considered. Since Listeria monocytogenes is able to multiply in cells, antibiotics must penetrate easily into macrophages. Combining antibiotics possessing this property could improve the cure rate for these infections. Furthermore, in cases of allergy to beta-lactams, contra-indications to use of aminoglycosides or clinical failure with the combination used, other drugs have to be administered. In allergic patients trimethoprim-sulfamethoxazole, which is bactericidal within 24 hours and diffuses well into tissues, can be considered as an alternative (27). Chloramphenicol on the other hand is only bacteriostatic (8). Rifampicin and ciprofloxacin, which concentrate in macrophages and can kill intracellular bacteria, need to be tested alone and in combination with other drugs to assess the bactericidal activity and the effect on emergence of resistant mutants. The main purpose of the present study was to evaluate the bactericidal activity of different combinations of six drugs against Listeria rnonocytogenes in vitro. Materials and Methods. The seven Listeria monocytogenes strains used were isolated between 1986 and 1988 from adult patients with meningitis (n = 2) and septicaemia (n = 2), and infants with infection (n = 3). Five strains were serotype 4 and two serotype 1. The antibiotics tested were amoxicillin (Bristol, France), trimethoprim-sulfamethoxazole (Hoffmann-La Roche, France), gentamicin (Unicet, France), ciprofloxacin (Bayer, France), and rifampicin (Le Petit, France). Susceptibility tests were performed in MuellerHinton broth with inocula of 106-107 CFU/ml; the medium was supplemented with 5 % lysed horse blood for tests with the combinations with trimethoprim-sulfamethoxazole. The MIC was defined as the lowest concentration of antibiotic which inhibited visible growth after overnight incubation at 37 °C. MBCs were determined by plating 0.01 ml of the bacterial suspension onto antibiotic free agar after incubation for 24 h. The MBC was defined as the concentration giving 99.9 % killing, except in the case of amoxieillin where the MBC was defined as the concentration giving the best bactericidal effect. The killing rate was expressed as the log10 reduction in the inoculum. The time-kill curves were determined using the same broth and inoculum conditions; at 2, 4, 6 and 24 h an 0.1 mt aliquot was removed and plated onto antibiotic free agar after dilution. A carry-over effect was prevented by dilution of the specimens before plating and in the case of trimethoprim-sulfamethoxazole by incorporating 1 mg/1 of thymidine in the agar. The antibiotic concentrations
Vol. 9, 1990
207
used to determine the killing curves were as follows: amoxicillin 10 mg/1, gentamicin 0.5 mg/1, trimethoprim 1 mg/1, sulfamethoxazole 20 mg/1, ciprofloxacin 1 mg/l, and rifampicin 2 mg/l. These concentrations were chosen because they represent the mean concentrations which can be obtained in cerebrospinal fluid or brain tissue, as reported in the literature (4, 9, 10, 11). The frequency of emergence of resistant mutants was determined for one strain by plating 0.1 ml of an inoculum of 109 CFU/ml on MuellerHinton agar containing increasing concentrations of rifampicin (0.03-32 rag/l) or ciprofloxacin (0.12-32 rag/l) alone, or rifampicin (0.0332 mg/1) in combination with a fixed concentration of ciprofloxacin (0.12, 0.25, 0.5, 1 or 2 rag/l). Resistant mutants growing at the different concentrations were counted after 48 h. Ten mutants per plate were subcultured twice on antibiotic free agar and the MICs determined by the agar dilution method using a Steers replicator giving 104 CFU/spot. For statistical analysis of data Student's t test was used, p < 0.05 being considered to indicate a significant difference. Results and Discussion. The MICs and MBCs of the six antibiotics tested at 24h and their bactericidal activity are shown in Table 1. The bactericidal activity of the antibiotics listed in decreasing order was as follows: gentamicin, trimethoprim-sulfamethoxazole, ciprofioxacin, trimethoprim, sulfamethoxazole, rifampicin and amoxicillin. In the killing curve experiments (Figure 1), gentamicin and amoxicillin plus gentamicin showed a killing rate of 3.86 log10 CFU/ml at 2 h. Trimethoprim-sulfamethoxazole was less bactericidal during the first 6 h, with a killing rate of 1.9 log10 at 2 h, 2.85 at 4 h and 3.45 at 6 h. At 24 h the two combinations had similar activity, with
killing rates of 6 and 5.6 log10 respectively. The results of the combination of trimethoprim-sulfamethoxazole with either amoxicillin, rifampicin or ciprofloxacin are shown in Figure 1A and D. The combinations did not appear to be antagonistic but were no more bactericidal than trimethoprim-sulfamethoxazole alone. The combination of amoxicillin with trimethoprim produced the same result (data not shown). The combination of amoxicillin and ciprofloxacin (Figure 1B) exhibited antagonism during the first 6 h (1.5 log10 difference at 6 h between the combination and ciprofloxacin alone, p < 0.01). However, at 24 h the combination prevented the regrowth observed with ciprofloxacin alone and was more active than either drug tested alone (1.05 log10 difference). The bactericidal activity of the combination of amoxicillin plus rifampicin (Figur e 1C) paralleled the activity of amoxicillin alone during the first 6 h; the combination appeared somewhat less active than rifampicin alone (0.5 log10 difference at 6 h). At 24 h the combination was significantly less active than both drugs alone (1 log10 difference, p < 0.05). Ciprofloxacin plus rifampicin (Figure IE) was less rapidly bactericidal than ciprofloxacin alone during the first 6 h, whereas at 24 h the combination prevented the regrowth observed with ciprofloxacin alone. However, the killing rate of this combination (3.3 log10) was less pronounced than the killing rate obtained with amoxicillin plus gentamicin (6 log10)or trimethoprim-sulfamethoxazole (5.6 loglo). The emergence of resistant mutants was evaluated for rifampicin and ciprofloxacin alone and in combination. The rate was 10-7 for rifampicin and ciprofloxacin alone. The emergence of rifampicin-resistant mutants was not prevented by increasing the concentration of rifampicin up to 64 mg/1. MICs of rifampicin for the mutants
T a b l e 1: MIC and M B C range, and bactericidal activity of six antimicrobial agents against seven strains of Listeria monocytogenes. MIC range (mg/1) Amoxicillin Gentamicin Rifampicin Ciprofloxacin Trimethoprim Sulfamethoxazole Trimethoprim/sulfamethoxazole
0.06- 0.25 0.12-0.5 0.06- 0.12 0.25 - 0.5 0.12- 0.25 3 2 - 64 0.03/0.5 - 0.06/1
aExpressed as the mean 1%10 reduction in the inoculum.
M B C range (rag/l) 0.12 - 0.5 1-2 0.06 - 0.25 1- 2 0.12- 0_5 3 2 - 64 0.03/0.5 - 0.12/2
Killing rate a 2 4 2.8 3.5 3 3 4
208
Eur. J. Clin. Microbiol. Infect Dis__=.
10
I0
9
9
R
8
7
{+ ?
~
i,
5
3
3
2
2 ! 0
5
"• 0
n 1 ]
I 4
I
I
Io
6
II
24 h
0
10
I0
9
9
B
8
7
+7
I 2
I 4
t 6
I 24 h
,,_--~
I' 4
4
3 2
C .............I....
I
J
I
2
4
6
8
.........
I
24 h
0
2
4
6
8
1 24 h
Figure 1: Bactericidal activity of six antimicrobiaI agents
alone and in combination against seven strains of Listeria monocytogenes. Each point represents the mean CFU/ml
E 0
.......... I
I
I
2
4
6
| 24 h
were 4 and 256 mg/l and remained unchanged for ciprofloxacin. The ciprofloxacin-resistant mutants disappeared when the antibiotic concentration was equal to or greater than 16 times the MIC (8 mg/1). MICs of ciprofloxacin for these mutants were 4 and 8 mg/l and remained unchanged for rifampicin. The concentration of both drugs had to be greater than the MICs to prevent the emergence of mutants. We did not observe mutants resistant to both drugs. The optimal regimen for treatment of Listeria monocytogenes meningoencephalitis is still a matter of debate since this organism is able to multiply in macrophages. Amoxicillin is not
for seven strains. A- amoxicillin 10 mg/1 (rn), gentamicin 0,5 mg/l (*), amoxicillin 10 mg/l + gentamicin 0.5 mg/l (I), trimethoprim 1 mg/1 + sulfamethoxazole 20 mgll (0), amoxiciilin 10 mg/l + trimethoprim I mg/! + sulfamethoxazole 20 mg/l (D). B: amoxicillin 10 mg/l (12), ciprofloxacin 1 mg/l (0), amoxicillin 10mg/l + ciprofloxacin 1 mg/l (I). C: amoxicillin 10 mg/I (F1),rifampicin2 mg/l (,), amoxicillin 10 mg/l + rifampiein 2 mg/l (151). D: trimethoprim 1 mg/l + sulfamethoxazole 20 mg/1 (0), rifampicin 2 mg/l + trimethoprim 1 mg/l + sulfamethoxazole 20 mgll (1:3), ciprofloxacin i mg/I + trimethoprim 1 rag/1 + sulfamethoxazole 20 mg/i (I). E: ciprofloxacin 1 mg/I (0), rifampicin 2 mg/l (*), ciprofloxacin 1 mg/l + rifampicin 2 mg/l ([]). Control: ,',.
highly bactericidal at 24 h but this antibiotic, like ampicillin, appears to be more effective at 48 h (1, 2). Our results with amoxicillin and gentamicin confirm previous findings (1, 2, 3). We found the combination of amoxicillin and rifampicin to be antagonistic at 6 h and not rapidly bactericidal at 24 h. These results do not differ from those obtained previously (6, 8, 9). However, in experimental models rifampicin has given contradictory results. Thus it was 200 times more active than ampicillin in a murine septicemia model, but seemed to be less active than ampicillin in a rabbit meningitis model (9, 11, 12, 13). The emergence of mutants resistant
Vol. 9, 1990
to rifampicin using a high inoculum definitely limits the possibility of single drug therapy with this compound. Ciprofloxacin appeared to be bactericidal. Its combination with amoxicillin was antagonistic during the first 6 h, as reported by L. Berkowitz et al. (27th Interscience Conference on Antimicrobial Agents and Chemotherapy, New York, 1987, Abstract no. 626), but at 24 h it was more active than amoxicillin alone and prevented the regrowth observed with ciprofioxacin at concentrations close to the MIC. Thus ciprofloxacin might enhance the action of amoxicillin and kill the bacteria enclosed in the cells. Trimethoprim-sulfamethoxazole has good bactericidal activity at 24 h, equivalent to that of ampicillin-gentamicin and quicker than that of ampicillin alone (2). Trimethoprim-sulfamethoxazole has been used successfully in vivo alone or combined with amoxicillin or ampicillin (5). Trimethoprim or trimethoprim-sulfamethoxazole could replace an aminoglycoside after a few days of therapy, since their combination with amoxicillin was not antagonistic. A possible advantage of this combination is its ability to diffuse into tissues and macrophages (3). For the same reason the combination of trimethoprim-sulfamethoxazole with ciprofloxacin or rifampicin could be of value. In vitro, these combinations gave no better results than trimethoprim-sulfamethoxazole alone. However, their efficacy might be enhanced in vivo due to their good penetration into the site of infection. The combination of rifampicin with ciprofioxacin was antagonistic compared to ciprofloxacin alone during the first 6 h, but reduced the inoculum by 3.3 loglo by 24 h. Thus this combination seems less bactericidal than the abovementioned combinations. However, it was as bactericidal as amoxicillin at 24 h. Another problem considered in the present study was the emergence of resistant mutants in the presence of rifampicin or ciprofloxacin alone. Ciprofloxacin and rifampicin concentrations greater than the MICs for the strain were necessary to prevent the emergence of resistant mutants. Such concentrations of ciprofloxacin cannot be obtained accurately in cerebrospinal fluid for treatment of meningitis, but can be obtained in tissues (10, 14). Both regimens have to be evaluated in in vivo models of macrophage infection and experimental disease before clinical investigations can be undertaken. In conclusion, trimethoprim-sulfamethoxazole administered in conjunction with rifampicin is a potential alternative to the classical regimens for the treatment of Listeria monocytogenes meningoencephalitis. The combination of tri-
209
methoprim-sulfamethoxazole with ciprofloxacin might be useful in treatment of other types of Listeria monocytogenes infections.
References 1. Moellering R, Medoff G, Leech I, Wennersten C, Kunz L: Antibiotic synergism against Listeria monocytogenes. Antimicrobial Agents and Chemotherapy 1972, I: 30-34. 2. Boisivon A: Action bact~riostatique et baet6ricide compar6e des p6nicillines, c6phalosporines, aminosides et des associations ampicillin-gentamicine et trim6thoprime-sulfam6thoxazole sur Listeria m o n o cytogenes. Annales de Mierobiologie 1980, 131 B: 267276. 3. P r i c h a r d M G , Miles HM, Pavillard ER: L i s t e r i a meningitis in vitro sensitivities to cotrimoxazole, penicillins and gentamicin. Australian and New Zealand Journal of Medicine 1983, 13: 76-77. 4. Spitzer PG, Hammer SM, Karchmer AW: Treatment of Listeria monocytogenes infection with trimethoprim-sulfamethoxazole: case report and review of the literature. Reviews of Infectious Diseases 1986, 8, 3: 427--430. 5. Slam JP, Gaillal J, Croize J, Bru JP, Carpentier F, Micoud M: Traitement des m6ningites ~t L i s t e r i a r n o n o c y t o g e n e s par le cotrimoxazole seul. In: Trimethoprime et sulfamides. Edition Arnette, 1984, Paris, p. 177-179. 6. Tuazon CU, Shamsuddin D, Miller H" Antibiotic susceptibility and synergy of clinical isolates of Listeria monocytogenes. Antimicrobial Agents and Chemotherapy 1982, 21: 525-527. 7. WinslowDL, Steele ML: Listeria baeteremia and peritonitis associated with a peritoneo-venous shunt: successful treatment with sulfamethoxazole and trimethoprim. Journal of Infectious Diseases 1984, 149: 820. 8. Winslow DL, Damme J, Diechman E: Delayed bactericidal activity of beta-lactam antibiotics against Listeria monocytogenes: antagonism of ehloramphenicol and rifampicin. Antimicrobial Agents and Chemotherapy 1983, 23: 555--558. 9. Seheld M: Evaluation of rifampicin and other antibiotics against Listeria monocytogenes in vitro and in vivo. Reviews of Infectious Diseases 1983, 5, Supplement 3: 593-597. 10, Wolff M, Boutron L, Singlas E, Clair B, Decazes JM, R6gnier B: Penetration of ciprofloxacin into cerebrospinal fluid of patients with bacterial meningitis. Antimicrobial Agents and Chemotherapy 1987, 31: 899-902. 11. Scheld MW, Fletcher DD, Fink FN, Sande M: Response to therapy in an experimental rabbit model of meningitis due to Listeria monocytogenes. Journal of Infectious Diseases 1979, 140: 287-294. 12. Hof H, Emmerling P: Murine model for therapy of listeriosis in the compromised host. Chemotherapy 1984, 30: 125-130. 13. VisherWA, Rominger C: Rifampicin against experimental listeriosis in the mouse. Chemotherapy 1978, 24: 104-111. 14. Desplaces N, Gutmann L, Carlet J, Guibert J, Acar JF: The new quinolones and their combinations with other agents for therapy of severe infections. Journal of Antimierobiat Chemotherapy 1986, 17, Supplement A: 25-29.
In Vitro Bactericidal Activity of Amoxicillin, Gentamicin, Rifampicin, Ciprofloxacin and Trimethoprim-Sulfamethoxazole Alone or in Combination against Listeria monocytogenes A. B o i s i v o n 1., C. G u i o m a r 1, C. C a r b o n 2 The in vitro bactericidal activity of amoxiciilin, gentamicin, rifampicin, ciprofloxacin and trimethoprim-sulfamethoxazole alone or in combination was determined against seven strains of Listeria monocytogenes by the killing curve method. Amoxicillin plus gentamicin was the most rapidly bactericidal combination, whereas trimethoprim-sulfamethoxazole was less bactericidal at 6 h but as bactericidal at 24 h. The combination of trimethoprim-sulfamethoxazole with either amoxicillin, ciprofloxacin or rifampicin did not result in antagonism, but the combinations were no more active than trimethoprim-suifamethoxazole alone. The interaction of amoxicillin with rifampin was fairly antagonistic (1 log10 difference). The combination of amoxicillin and ciprofloxacin, although producing antagonism during the first 6 h, was more active at 24 h than amoxiciilin alone and prevented the regrowth observed with ciprofloxacin alone. Ciprofloxacin and rifampicin interacted antagonistically during the first 6 h, and the combination was not very bactericidal (3 log10) but prevented the emergence of mutants, as observed with each drug alone, when used at concentrations greater than the MICs for the strain tested. These regimens merit evaluation in in vivo models of Listeria monocytogenes meningitis. Listeria monocytogenes infects both the immunocompromised and non-immunocomproraised host. Clinical infections include bacteremia and meningoencephalitis. The therapy of choice for Listeria monocytogenes infections is ampicillin or a combination of ampicillin with an aminoglycoside, as has been demonstrated in vitro and clinically (1). The prognosis of the most severe forms of meningoencephalitis re1Laboratoire de Microbiologie, Centre Hospitalier de Saint Germain-en-Laye, 78104 Saint Germain-en-Laye, France. 2Service de M6decine Interne, Hbpital Bichat. Universit6 Paris VII, 75018Paris, France.
Eur. J. Clin. Microbiol. Infect Dis.
mains poor. Thus alternative regimens have to be considered. Since Listeria monocytogenes is able to multiply in cells, antibiotics must penetrate easily into macrophages. Combining antibiotics possessing this property could improve the cure rate for these infections. Furthermore, in cases of allergy to beta-lactams, contra-indications to use of aminoglycosides or clinical failure with the combination used, other drugs have to be administered. In allergic patients trimethoprim-sulfamethoxazole, which is bactericidal within 24 hours and diffuses well into tissues, can be considered as an alternative (27). Chloramphenicol on the other hand is only bacteriostatic (8). Rifampicin and ciprofloxacin, which concentrate in macrophages and can kill intracellular bacteria, need to be tested alone and in combination with other drugs to assess the bactericidal activity and the effect on emergence of resistant mutants. The main purpose of the present study was to evaluate the bactericidal activity of different combinations of six drugs against Listeria rnonocytogenes in vitro. Materials and Methods. The seven Listeria monocytogenes strains used were isolated between 1986 and 1988 from adult patients with meningitis (n = 2) and septicaemia (n = 2), and infants with infection (n = 3). Five strains were serotype 4 and two serotype 1. The antibiotics tested were amoxicillin (Bristol, France), trimethoprim-sulfamethoxazole (Hoffmann-La Roche, France), gentamicin (Unicet, France), ciprofloxacin (Bayer, France), and rifampicin (Le Petit, France). Susceptibility tests were performed in MuellerHinton broth with inocula of 106-107 CFU/ml; the medium was supplemented with 5 % lysed horse blood for tests with the combinations with trimethoprim-sulfamethoxazole. The MIC was defined as the lowest concentration of antibiotic which inhibited visible growth after overnight incubation at 37 °C. MBCs were determined by plating 0.01 ml of the bacterial suspension onto antibiotic free agar after incubation for 24 h. The MBC was defined as the concentration giving 99.9 % killing, except in the case of amoxieillin where the MBC was defined as the concentration giving the best bactericidal effect. The killing rate was expressed as the log10 reduction in the inoculum. The time-kill curves were determined using the same broth and inoculum conditions; at 2, 4, 6 and 24 h an 0.1 mt aliquot was removed and plated onto antibiotic free agar after dilution. A carry-over effect was prevented by dilution of the specimens before plating and in the case of trimethoprim-sulfamethoxazole by incorporating 1 mg/1 of thymidine in the agar. The antibiotic concentrations
Vol. 9, 1990
207
used to determine the killing curves were as follows: amoxicillin 10 mg/1, gentamicin 0.5 mg/1, trimethoprim 1 mg/1, sulfamethoxazole 20 mg/1, ciprofloxacin 1 mg/l, and rifampicin 2 mg/l. These concentrations were chosen because they represent the mean concentrations which can be obtained in cerebrospinal fluid or brain tissue, as reported in the literature (4, 9, 10, 11). The frequency of emergence of resistant mutants was determined for one strain by plating 0.1 ml of an inoculum of 109 CFU/ml on MuellerHinton agar containing increasing concentrations of rifampicin (0.03-32 rag/l) or ciprofloxacin (0.12-32 rag/l) alone, or rifampicin (0.0332 mg/1) in combination with a fixed concentration of ciprofloxacin (0.12, 0.25, 0.5, 1 or 2 rag/l). Resistant mutants growing at the different concentrations were counted after 48 h. Ten mutants per plate were subcultured twice on antibiotic free agar and the MICs determined by the agar dilution method using a Steers replicator giving 104 CFU/spot. For statistical analysis of data Student's t test was used, p < 0.05 being considered to indicate a significant difference. Results and Discussion. The MICs and MBCs of the six antibiotics tested at 24h and their bactericidal activity are shown in Table 1. The bactericidal activity of the antibiotics listed in decreasing order was as follows: gentamicin, trimethoprim-sulfamethoxazole, ciprofioxacin, trimethoprim, sulfamethoxazole, rifampicin and amoxicillin. In the killing curve experiments (Figure 1), gentamicin and amoxicillin plus gentamicin showed a killing rate of 3.86 log10 CFU/ml at 2 h. Trimethoprim-sulfamethoxazole was less bactericidal during the first 6 h, with a killing rate of 1.9 log10 at 2 h, 2.85 at 4 h and 3.45 at 6 h. At 24 h the two combinations had similar activity, with
killing rates of 6 and 5.6 log10 respectively. The results of the combination of trimethoprim-sulfamethoxazole with either amoxicillin, rifampicin or ciprofloxacin are shown in Figure 1A and D. The combinations did not appear to be antagonistic but were no more bactericidal than trimethoprim-sulfamethoxazole alone. The combination of amoxicillin with trimethoprim produced the same result (data not shown). The combination of amoxicillin and ciprofloxacin (Figure 1B) exhibited antagonism during the first 6 h (1.5 log10 difference at 6 h between the combination and ciprofloxacin alone, p < 0.01). However, at 24 h the combination prevented the regrowth observed with ciprofloxacin alone and was more active than either drug tested alone (1.05 log10 difference). The bactericidal activity of the combination of amoxicillin plus rifampicin (Figur e 1C) paralleled the activity of amoxicillin alone during the first 6 h; the combination appeared somewhat less active than rifampicin alone (0.5 log10 difference at 6 h). At 24 h the combination was significantly less active than both drugs alone (1 log10 difference, p < 0.05). Ciprofloxacin plus rifampicin (Figure IE) was less rapidly bactericidal than ciprofloxacin alone during the first 6 h, whereas at 24 h the combination prevented the regrowth observed with ciprofloxacin alone. However, the killing rate of this combination (3.3 log10) was less pronounced than the killing rate obtained with amoxicillin plus gentamicin (6 log10)or trimethoprim-sulfamethoxazole (5.6 loglo). The emergence of resistant mutants was evaluated for rifampicin and ciprofloxacin alone and in combination. The rate was 10-7 for rifampicin and ciprofloxacin alone. The emergence of rifampicin-resistant mutants was not prevented by increasing the concentration of rifampicin up to 64 mg/1. MICs of rifampicin for the mutants
T a b l e 1: MIC and M B C range, and bactericidal activity of six antimicrobial agents against seven strains of Listeria monocytogenes. MIC range (mg/1) Amoxicillin Gentamicin Rifampicin Ciprofloxacin Trimethoprim Sulfamethoxazole Trimethoprim/sulfamethoxazole
0.06- 0.25 0.12-0.5 0.06- 0.12 0.25 - 0.5 0.12- 0.25 3 2 - 64 0.03/0.5 - 0.06/1
aExpressed as the mean 1%10 reduction in the inoculum.
M B C range (rag/l) 0.12 - 0.5 1-2 0.06 - 0.25 1- 2 0.12- 0_5 3 2 - 64 0.03/0.5 - 0.12/2
Killing rate a 2 4 2.8 3.5 3 3 4
208
Eur. J. Clin. Microbiol. Infect Dis__=.
10
I0
9
9
R
8
7
{+ ?
~
i,
5
3
3
2
2 ! 0
5
"• 0
n 1 ]
I 4
I
I
Io
6
II
24 h
0
10
I0
9
9
B
8
7
+7
I 2
I 4
t 6
I 24 h
,,_--~
I' 4
4
3 2
C .............I....
I
J
I
2
4
6
8
.........
I
24 h
0
2
4
6
8
1 24 h
Figure 1: Bactericidal activity of six antimicrobiaI agents
alone and in combination against seven strains of Listeria monocytogenes. Each point represents the mean CFU/ml
E 0
.......... I
I
I
2
4
6
| 24 h
were 4 and 256 mg/l and remained unchanged for ciprofloxacin. The ciprofloxacin-resistant mutants disappeared when the antibiotic concentration was equal to or greater than 16 times the MIC (8 mg/1). MICs of ciprofloxacin for these mutants were 4 and 8 mg/l and remained unchanged for rifampicin. The concentration of both drugs had to be greater than the MICs to prevent the emergence of mutants. We did not observe mutants resistant to both drugs. The optimal regimen for treatment of Listeria monocytogenes meningoencephalitis is still a matter of debate since this organism is able to multiply in macrophages. Amoxicillin is not
for seven strains. A- amoxicillin 10 mg/1 (rn), gentamicin 0,5 mg/l (*), amoxicillin 10 mg/l + gentamicin 0.5 mg/l (I), trimethoprim 1 mg/1 + sulfamethoxazole 20 mgll (0), amoxiciilin 10 mg/l + trimethoprim I mg/! + sulfamethoxazole 20 mg/l (D). B: amoxicillin 10 mg/l (12), ciprofloxacin 1 mg/l (0), amoxicillin 10mg/l + ciprofloxacin 1 mg/l (I). C: amoxicillin 10 mg/I (F1),rifampicin2 mg/l (,), amoxicillin 10 mg/l + rifampiein 2 mg/l (151). D: trimethoprim 1 mg/l + sulfamethoxazole 20 mg/1 (0), rifampicin 2 mg/l + trimethoprim 1 mg/l + sulfamethoxazole 20 mgll (1:3), ciprofloxacin i mg/I + trimethoprim 1 rag/1 + sulfamethoxazole 20 mg/i (I). E: ciprofloxacin 1 mg/I (0), rifampicin 2 mg/l (*), ciprofloxacin 1 mg/l + rifampicin 2 mg/l ([]). Control: ,',.
highly bactericidal at 24 h but this antibiotic, like ampicillin, appears to be more effective at 48 h (1, 2). Our results with amoxicillin and gentamicin confirm previous findings (1, 2, 3). We found the combination of amoxicillin and rifampicin to be antagonistic at 6 h and not rapidly bactericidal at 24 h. These results do not differ from those obtained previously (6, 8, 9). However, in experimental models rifampicin has given contradictory results. Thus it was 200 times more active than ampicillin in a murine septicemia model, but seemed to be less active than ampicillin in a rabbit meningitis model (9, 11, 12, 13). The emergence of mutants resistant
Vol. 9, 1990
to rifampicin using a high inoculum definitely limits the possibility of single drug therapy with this compound. Ciprofloxacin appeared to be bactericidal. Its combination with amoxicillin was antagonistic during the first 6 h, as reported by L. Berkowitz et al. (27th Interscience Conference on Antimicrobial Agents and Chemotherapy, New York, 1987, Abstract no. 626), but at 24 h it was more active than amoxicillin alone and prevented the regrowth observed with ciprofioxacin at concentrations close to the MIC. Thus ciprofloxacin might enhance the action of amoxicillin and kill the bacteria enclosed in the cells. Trimethoprim-sulfamethoxazole has good bactericidal activity at 24 h, equivalent to that of ampicillin-gentamicin and quicker than that of ampicillin alone (2). Trimethoprim-sulfamethoxazole has been used successfully in vivo alone or combined with amoxicillin or ampicillin (5). Trimethoprim or trimethoprim-sulfamethoxazole could replace an aminoglycoside after a few days of therapy, since their combination with amoxicillin was not antagonistic. A possible advantage of this combination is its ability to diffuse into tissues and macrophages (3). For the same reason the combination of trimethoprim-sulfamethoxazole with ciprofloxacin or rifampicin could be of value. In vitro, these combinations gave no better results than trimethoprim-sulfamethoxazole alone. However, their efficacy might be enhanced in vivo due to their good penetration into the site of infection. The combination of rifampicin with ciprofioxacin was antagonistic compared to ciprofloxacin alone during the first 6 h, but reduced the inoculum by 3.3 loglo by 24 h. Thus this combination seems less bactericidal than the abovementioned combinations. However, it was as bactericidal as amoxicillin at 24 h. Another problem considered in the present study was the emergence of resistant mutants in the presence of rifampicin or ciprofloxacin alone. Ciprofloxacin and rifampicin concentrations greater than the MICs for the strain were necessary to prevent the emergence of resistant mutants. Such concentrations of ciprofloxacin cannot be obtained accurately in cerebrospinal fluid for treatment of meningitis, but can be obtained in tissues (10, 14). Both regimens have to be evaluated in in vivo models of macrophage infection and experimental disease before clinical investigations can be undertaken. In conclusion, trimethoprim-sulfamethoxazole administered in conjunction with rifampicin is a potential alternative to the classical regimens for the treatment of Listeria monocytogenes meningoencephalitis. The combination of tri-
209
methoprim-sulfamethoxazole with ciprofloxacin might be useful in treatment of other types of Listeria monocytogenes infections.
References 1. Moellering R, Medoff G, Leech I, Wennersten C, Kunz L: Antibiotic synergism against Listeria monocytogenes. Antimicrobial Agents and Chemotherapy 1972, I: 30-34. 2. Boisivon A: Action bact~riostatique et baet6ricide compar6e des p6nicillines, c6phalosporines, aminosides et des associations ampicillin-gentamicine et trim6thoprime-sulfam6thoxazole sur Listeria m o n o cytogenes. Annales de Mierobiologie 1980, 131 B: 267276. 3. P r i c h a r d M G , Miles HM, Pavillard ER: L i s t e r i a meningitis in vitro sensitivities to cotrimoxazole, penicillins and gentamicin. Australian and New Zealand Journal of Medicine 1983, 13: 76-77. 4. Spitzer PG, Hammer SM, Karchmer AW: Treatment of Listeria monocytogenes infection with trimethoprim-sulfamethoxazole: case report and review of the literature. Reviews of Infectious Diseases 1986, 8, 3: 427--430. 5. Slam JP, Gaillal J, Croize J, Bru JP, Carpentier F, Micoud M: Traitement des m6ningites ~t L i s t e r i a r n o n o c y t o g e n e s par le cotrimoxazole seul. In: Trimethoprime et sulfamides. Edition Arnette, 1984, Paris, p. 177-179. 6. Tuazon CU, Shamsuddin D, Miller H" Antibiotic susceptibility and synergy of clinical isolates of Listeria monocytogenes. Antimicrobial Agents and Chemotherapy 1982, 21: 525-527. 7. WinslowDL, Steele ML: Listeria baeteremia and peritonitis associated with a peritoneo-venous shunt: successful treatment with sulfamethoxazole and trimethoprim. Journal of Infectious Diseases 1984, 149: 820. 8. Winslow DL, Damme J, Diechman E: Delayed bactericidal activity of beta-lactam antibiotics against Listeria monocytogenes: antagonism of ehloramphenicol and rifampicin. Antimicrobial Agents and Chemotherapy 1983, 23: 555--558. 9. Seheld M: Evaluation of rifampicin and other antibiotics against Listeria monocytogenes in vitro and in vivo. Reviews of Infectious Diseases 1983, 5, Supplement 3: 593-597. 10, Wolff M, Boutron L, Singlas E, Clair B, Decazes JM, R6gnier B: Penetration of ciprofloxacin into cerebrospinal fluid of patients with bacterial meningitis. Antimicrobial Agents and Chemotherapy 1987, 31: 899-902. 11. Scheld MW, Fletcher DD, Fink FN, Sande M: Response to therapy in an experimental rabbit model of meningitis due to Listeria monocytogenes. Journal of Infectious Diseases 1979, 140: 287-294. 12. Hof H, Emmerling P: Murine model for therapy of listeriosis in the compromised host. Chemotherapy 1984, 30: 125-130. 13. VisherWA, Rominger C: Rifampicin against experimental listeriosis in the mouse. Chemotherapy 1978, 24: 104-111. 14. Desplaces N, Gutmann L, Carlet J, Guibert J, Acar JF: The new quinolones and their combinations with other agents for therapy of severe infections. Journal of Antimierobiat Chemotherapy 1986, 17, Supplement A: 25-29.
