Susceptibilities of Legionella spp. to Newer Antimicrobials In Vitro期刊界 All Journals 搜尽天下杂志传播学术成果专业期刊搜索期刊信息化学术搜索

Susceptibilities of Legionella spp. to Newer Antimicrobials In Vitro

Authors:

T. Schülin C. B. Wennersten M. J. Ferraro R. C. Moellering Jr. G. M. Eliopoulos

Affiliation:

Department of Medicine, West Campus, Beth Israel Deaconess Medical Center, Boston, Massachusetts 02215,1. Harvard Medical School, Boston, Massachusetts 02115,2. and Massachusetts General Hospital, Boston, Massachusetts 021143.

Abstract:

The in vitro activities of 13 antimicrobial agents against 30 strains of Legionella spp. were determined. Rifapentine, rifampin, and clarithromycin were the most potent agents (MICs at which 90% of isolates are inhibited [MIC₉₀s], ≤0.008 μg/ml). The ketolide HMR 3647 and the fluoroquinolones levofloxacin and BAY 12-8039 (MIC₉₀s, 0.03 to 0.06 μg/ml) were more active than erythromycin A or roxithromycin. The MIC₉₀s of dalfopristin-quinupristin and linezolid were 0.5 and 8 μg/ml, respectively. Based on class characteristics and in vitro activities, several of these agents may have potential roles in the treatment of Legionella infections.The array of antimicrobial agents useful for the treatment of serious infections caused by Legionella spp. is limited. This is in part due to the relative resistance of Legionella spp. to a variety of antimicrobial agents and to the fact that these organisms are obligate intracellular pathogens and, thus, to be effective, the drugs must be able to penetrate into phagocytic cells ().Erythromycin, rifampin, and fluoroquinolones have proven in vitro and in vivo efficacies and are used to treat clinical Legionella infections (, ). Mortality is still high in those with nosocomial pneumonia, especially immunocompromised and bacteremic patients (), so there is a need for a wider range of suitable antibiotics to treat severe Legionella infections.This study examined the in vitro activities of several newer antimicrobial agents, including a ketolide, two fluoroquinolones, two oxazolidinones, rifapentine, and dalfopristin-quinupristin, against Legionella spp., an initial step in assessing their potential usefulness as therapeutic agents.(This work was presented in part at the 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Ontario, Canada, 28 September to 1 October 1997 [33]).Thirty clinical isolates of Legionella spp. were tested, including 21 L. pneumophila, 3 L. longbeacheae, 2 L. bozemanii, and 2 L. dumoffii isolates and one strain each of L. micdadei and L. gormanii. Strains were referred to our collection from various sources over several years and were kept frozen at −80°C.Antimicrobial test substances and their sources are as follows. HMR 3647, erythromycin A, clarithromycin, roxithromycin, rifampin, rifapentine, and levofloxacin were gifts from Hoechst-Marion-Roussel, Romainville, France; eperezolid (U-100592) and linezolid (U-100766) were gifts from Pharmacia & Upjohn Company Laboratories, Kalamazoo, Mich.; dalfopristin-quinupristin was provided by Rhône-Poulenc Rorer Pharmaceuticals, Collegeville, Pa.; and BAY 12-8039 was a gift of Bayer Inc., West Haven, Conn. Clindamycin hydrochloride and doxycycline hydrochloride were purchased from Sigma Chemical Company, St. Louis, Mo.Agar dilution susceptibility testing was performed on the buffered starch-yeast extract (BSYE) agar medium described by Saito et al. (). Buffered charcoal-yeast extract medium (BCYE) has been shown to impair the activities of several antimicrobial substances (i.e., macrolides, rifampin, and fluoroquinolones) in earlier studies (, , ), so this medium was only used to subcultivate strains twice after thawing them from their −80°C storage temperature and as a growth control.To prepare inocula, several colonies were taken from BCYE plates (Remel, Lenexa, Kans.) after 48 h of incubation and were suspended in sterile water to a turbidity corresponding to a 0.5 McFarland standard, which yielded a cell density of approximately 10⁸ CFU/ml. Suspensions of bacteria were then further diluted 1:10 in sterile water for the smaller inoculum. Final inocula of 10⁵ and 10⁴ CFU/spot were applied to freshly made antibiotic-containing plates with a multiprong replicator device. Between each antibiotic, antibiotic-free plates were stamped to avoid carryover, and a blood agar plate was also inoculated at the end of each run to exclude contamination by other bacteria.Plates were incubated at 35°C in ambient air and were read after 48 and 96 h. Spots yielding the growth of single colonies and those with a faint haze were considered to be negative.Table Table11 shows the results for the 48-h incubation time for both inocula. For most agents, a twofold increase in the MIC at which 90% of the isolates were inhibited (MIC₉₀) was observed when the plates were examined after 96 h of incubation (Table (Table2).2). Such results may reflect either incomplete inhibition of growth at a particular antibiotic dilution or the loss of antimicrobial potency with prolonged incubation. Subsequent comments will be directed to results of the 48-h readings. With the larger inoculum, all strains grew on BSYE agar as well as on BCYE agar, whereas with the smaller inoculum, three to six strains yielded insufficient growth on control plates and therefore were excluded from the analysis. These findings are consistent with results from other studies, which showed that BSYE agar does not support the growth of some Legionella species as well as does BCYE agar (, ). Table Table33 compares the MICs of several antimicrobial agents tested against Legionella spp., obtained in different studies using different media and methods.

TABLE 1

Comparison of in vitro activities of 13 antimicrobial agents against Legionella spp., determined at 48 h of incubation

Antibiotic	No. of strains tested with inoculum of 10⁴^a	MIC (μg/ml) at inoculum size shown (CFU/spot)
		Range		MIC₅₀		MIC₉₀
		10⁴	10⁵	10⁴	10⁵	10⁴	10⁵
HMR 3647	27	≤0.004–0.06	≤0.004–0.12	0.008	0.015	0.03	0.03
Erythromycin A	27	0.008–0.25	0.06–0.5	0.03	0.12	0.12	0.12
Clarithromycin	27	≤0.004	≤0.004–0.03	≤0.004	≤0.004	≤0.004	≤0.004
Roxithromycin	27	≤0.004–0.06	0.03–0.25	0.015	0.06	0.03	0.12
Levofloxacin	27	≤0.004–0.03	0.015–0.06	0.008	0.015	0.015	0.03
BAY 12-8039	27	0.015–0.06	0.03–0.12	0.03	0.03	0.06	0.06
Rifampin	27	≤0.0005–0.015	≤0.0005–0.015	≤0.0005	≤0.0005	0.002	0.008
Rifapentine	27	≤0.001–0.002	≤0.001–0.002	≤0.001	≤0.001	≤0.001	0.002
Dalfopristin-quinupristin	24	0.015–0.05	0.12–1	0.12	0.25	0.5	0.5
Doxycycline	24	0.5–2	1.0–8.0	1	4	2	8
Eperezolid	24	1.0–8	2–≥16	1	4	4	≥16
Linezolid	24	1.0–4	4.0–8	2	4	4	8
Clindamycin	27	0.008–8	1.0–16	2	8	8	8

Open in a separate window^an = 30 strains for the larger inoculum in all cases.

TABLE 2

Comparison of MIC₉₀s at 48 and 96 h of incubation for inocula of 10⁴ and 10⁵ CFU/spot

Antibiotic	MIC 90 (μg/ml) at time point and inoculum indicated
	48 h		96 h
	10⁴	10⁵	10⁴	10⁵
HMR 3647	0.03	0.03	0.06	0.06
Erythromycin A	0.12	0.12	0.25	0.25
Clarithromycin	≤0.004	≤0.004	≤0.004	0.015
Roxithromycin	0.03	0.12	0.06	0.12
Dalfopristin-quinupristin	0.5	0.5	0.5	1
Levofloxacin	0.015	0.03	0.03	0.03
BAY 12-8039	0.06	0.06	0.12	0.12
Rifampin	0.002	0.008	0.004	0.008
Rifapentine	≤0.001	0.002	0.002	0.004
Doxycycline	2	8	8	16
Eperezolid	4	≥16	16	≥16
Linezolid	4	8	8	16
Clindamycin	8	8	16	16

Open in a separate window

TABLE 3

Comparison of methods used and MICs of antimicrobial agents tested against Legionella spp. in different studies

Medium^a	Inoculum (CFU/spot)	Incubation time (h)	MIC₉₀, MIC₉₀ range, or geometric mean MIC (μg/ml) for:					Reference
Medium^a	Inoculum (CFU/spot)	Incubation time (h)	Erythromycin	Rifampin	Ciprofloxacin	Levofloxacin	Dalfopristinquinupristin	Reference
BSYE	10⁴/10⁵	48	0.12	0.002/0.008	d.n.a.^d	0.015/0.03	0.5	This study
BCYE	10⁴	48	1	≤0.004	d.n.a.	0.125	d.n.a.	2^e
BCYE	10⁴	72	d.n.a.	0.03–0.125	d.n.a.	d.n.a.	d.n.a.	3^e
LHRC-BYE	10⁴	72	d.n.a.	0.00035–0.00075	d.n.a.	d.n.a.	d.n.a.	3^e
BAM₁	6 × 10⁵	72	0.25^b	d.n.a.	d.n.a.	d.n.a.	d.n.a.	4
BCYE	6 × 10⁵	72	0.12^b	d.n.a.	d.n.a.	d.n.a.	d.n.a.	4
BYE	6 × 10⁵	72	0.12^b	d.n.a.	d.n.a.	d.n.a.	d.n.a.	4
BCYE	10⁵	48	1	0.008	2	d.n.a.	d.n.a.	6
BSYE	10⁵	48	0.5	≤0.002	0.125	d.n.a.	d.n.a.	6
BYE	10⁴	48	0.25–1^c	≤0.004–0.008^c	0.01–0.12^c	d.n.a.	0.12–0.5^c	10^e
CYE	10⁴/10⁵	72	0.25/0.5	0.125/0.125	d.n.a.	d.n.a.	d.n.a.	15^e
BCYE	10⁴	96	≥2	0.5	2	d.n.a.	d.n.a.	17
BSYE	10⁴	96	0.5	≤0.015	≤0.06	d.n.a.	d.n.a.	17
BSYE	10⁶	48	0.5	0.008	0.06	d.n.a.	1	24^e
BCYE	10⁴	48	1	d.n.a.	1	0.5	d.n.a.	29^e
BSYE	10⁶	48	0.06	d.n.a.	0.015	0.015	d.n.a.	29^e

Open in a separate window^aAbbreviations: LHRC, lysed horse red cell; BYE, buffered yeast extract; BAM₁, buffered antibiotic medium no. 1; CYE, charcoal-yeast extract. ^bGeometric mean MIC. ^cMIC₉₀ range. ^dd.n.a., data not available. ^eData for L. pneumophila. Erythromycin, probably the most widely used drug for treatment of Legionella pneumonia (, ), inhibited all strains at ≤0.25 and ≤0.5 μg/ml with the small and the large inocula, respectively. Those data were comparable to erythromycin A MICs obtained previously in our laboratory ().A new ketolide designated HMR 3004 has been shown to reach high intracellular concentrations in phagocytes; therefore, agents of this class may be of potential therapeutic use against intracellular pathogens like Legionella spp. (1). The ketolide tested here, HMR 3647, inhibited 90% of all organisms at concentrations of 0.03 μg/ml and thus showed fourfold-higher activity than erythromycin A. These data complement a study by Bornstein et al. (5), who found HMR 3004 to be active against Legionella spp. with a range of MICs virtually identical to those obtained for HMR 3647 in our study (MIC, ≤0.03 to 0.12 μg/ml) when performed by the agar dilution technique on a different medium (buffered antibiotic medium no. 1). Clarithromycin was the most potent macrolide in our study, exhibiting an MIC₉₀ of ≤0.004 μg/ml with both inocula.Rifampin is used in combination with other drugs in severe or refractory cases of legionellosis (). In a number of comparative studies, it was the most active drug tested (, , ). In the present study, 90% of isolates were inhibited at concentrations of 0.008 μg/ml with the larger inoculum. Rifapentine is a newly developed agent related to rifampin. The MIC₉₀ of this drug was 0.002 μg/ml, fourfold lower than that of rifampin, with the large inoculum. All strains of L. pneumophila were inhibited at the lowest concentrations of rifampin and rifapentine tested, 0.0005 and 0.001 μg/ml, respectively. The MICs of rifampin and rifapentine for other species ranged from 0.0005 to 0.015 μg/ml and from 0.001 to 0.002 μg/ml, respectively, after 48 h of incubation.Fluoroquinolones have been shown to be highly effective in vitro (17), and they have also been shown to inhibit the growth of legionellae in alveolar macrophage systems and in experimental treatment models of L. pneumophila pneumonia in guinea pigs (, , ). Moreover, fluoroquinolones have been used clinically for treatment of Legionella pneumonia (). In the present study, the MIC₉₀s for BAY 12-8039 and levofloxacin were 0.06 and 0.03 μg/ml, respectively, with the larger inoculum. The MICs for levofloxacin were two to three times higher in a study by Baltch et al. (), but their study utilized BCYE agar, which is known to inhibit the activity of certain antimicrobial agents, especially fluoroquinolones (17). In experimental Legionnaires’ disease in guinea pigs, levofloxacin appeared to be as active as ofloxacin, which was superior to ciprofloxacin and erythromycin (, ). Our data for BAY 12-8039 were comparable to those reported by Ruckdeschel et al. (30); in the latter study, a larger inoculum was used (10⁸ CFU/spot).The streptogramin combination dalfopristin-quinupristin inhibited 90% of all isolates at a concentration of 0.5 μg/ml. The overall MICs were two- to fourfold higher than those of erythromycin A, which is consistent with a report by Johnson et al. (), in which they showed the same correlation between those two drugs with a larger inoculum (10⁶ CFU/spot). In contrast, in a study by Dubois and Joly (), dalfopristin-quinupristin demonstrated twofold-higher activity than erythromycin against some Legionella species. A possible role for this drug in the treatment of legionellosis is supported by reports which showed high intracellular accumulation and activity against intracellular staphylococci (); however, the activity of dalfopristin-quinupristin against intracellular enterococci was modest ().The oxazolidinones linezolid (U-100766) and eperezolid (U-100592) are recently developed antimicrobial agents which have shown therapeutic potential based on in vitro activity against various respiratory pathogens, including multidrug-resistant pneumococci, streptococci, staphylococci, Haemophilus spp., and Moraxella spp. (, , ). At the smaller inoculum, 90% of the legionellae tested were inhibited by a 4-μg/ml concentration of each drug, a concentration equivalent to MICs for other presumptively susceptible organisms. However, at the larger inoculum, the MIC₉₀s of the agents were 8 and ≥16 μg/ml, respectively. Unless there was evidence for intracellular accumulation in phagocytes, such in vitro data would not suggest that these specific oxazolidinones would be likely candidates for treatment of Legionella infections.At the smaller and larger inocula, doxycycline inhibited 90% of strains at 2 and 8 μg/ml, respectively, after 48 h of incubation. MICs of 8 μg/ml would indicate intermediate susceptibility (). Nevertheless, doxycycline showed activity against L. pneumophila in intracellular monocyte experiments when added at concentrations of 0.4 μg/ml (). It was also shown to be therapeutically effective in a guinea pig model of experimental legionellosis () and showed clinical efficacy in the treatment of human legionellosis (). It is known that tetracyclines accumulate in human neutrophils (), and such discrepancies between in vitro activity and therapeutic results exemplify the potential pitfalls in predicting clinical effectiveness from in vitro data alone.This study identified several new antimicrobial agents with in vitro activities against legionellae that were higher than that of the widely used agent erythromycin. To further explore the potential applicability of these in vitro findings to the clinical setting, intracellular susceptibility testing and animal model studies would be of interest.

Keywords:

设为首页 | 免责声明 | 关于勤云 | 加入收藏