In Vivo Efficacy of the New Ketolide Telithromycin (HMR 3647) in Murine Infection Models

Abstract

We compared the oral antibacterial activities of telithromycin (HMR 3647), a new ketolide drug, in different infections induced in mice by Staphylococcus aureus, Streptococcus pneumoniae, streptococci, enterococci, and Haemophilus influenzae with those of various macrolides and pristinamycin. Unlike all other comparators, telithromycin displayed a high therapeutic activity, particularly in septicemia induced by erythromycin A-resistant pathogens, where the ketolide was the only active compound, displaying effective doses between 3 and 26 mg/kg of body weight. Against H. influenzae, telithromycin was the most effective compound. Telithromycin displayed bacteriostatic behavior against S. pneumoniae and H. influenzae. The ketolide was also active against thigh muscle infection induced by S. aureus. The pharmacokinetic properties of telithromycin accounted for its outstanding well-balanced oral in vivo efficacy against both gram-positive cocci, whatever their phenotype of resistance, and H. influenzae.

Telithromycin (HMR 3647) is an antibiotic belonging to a new class of 14-membered-ring macrolides called ketolides which has shown marked in vitro activity against a large bacterial spectrum extending to multidrug-resistant pneumococci, staphylococci, streptococci, Haemophilus influenzae, Moraxella catarrhalis, Branhamella pertussis, and intracellular respiratory pathogens, such as Chlamydia pneumoniae and Legionella spp. (4; C. Agouridas, A. Bonnefoy, and J. F. Chantot., Abstr. 37th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-112, p. 165, 1997).

We describe here the in vivo antibacterial efficacy of telithromycin in comparison to those of erythromycin A, clarithromycin, azithromycin, josamycin, pristinamycin, and vancomycin against enterococci in a murine model of septicemia induced by intraperitoneal injection of virulent bacteria, as well as in a localized Staphylococcus aureus thigh infection. In vivo bactericidal activity and pharmacokinetic parameters are also included.

The present studies were approved by the Internal Animal Ethics Committee.

(Part of this work was previously presented [C. Agouridas, A. Bonnefoy, and J. F. Chantot, Abstr. 37th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-257, p. 189, 1997]).

MATERIALS AND METHODS

Antibiotics.

Telithromycin, clarithromycin, and azithromycin were prepared by Hoechst Marion Roussel (Romainville, France). Free-base telithromycin was used as an amorphous powder. Erythromycin A was obtained from Hoechst Marion Roussel. Pristinamycin and josamycin were from Rhône-Poulenc-Rorer (Vitry, France) and ICN Biomedicals (Costa Mesa, Calif.), respectively. Rokitamycin was from Pierre Fabre Medicament (Boulogne, France), and vancomycin was from Sigma (Saint Louis, Mo.).

Bacterial strains.

Most of the strains tested were clinical isolates from various European and U.S. hospitals. Inducible and constitutive macrolides-lincosamides-streptogramin B (MLS) resistance among staphylococci was defined by using josamycin, a 16-membered-ring macrolide active against inducibly resistant staphylococci. With pneumococci, enterococci, and streptococci, rokitamycin, another 16-membered-ring macrolide, was tentatively used to distinguish between inducibly and constitutively resistant strains (7). The MICs of rokitamycin were ≤1 and >4 mg/liter for inducibly and constitutively MLS-resistant strains, respectively. All isolates were stored frozen at −80°C except staphylococci, which were kept at −20°C.

Susceptibility testing.

MICs were measured by a twofold agar dilution method. Mueller-Hinton (MH) agar medium (pH 7.4; Diagnostic Pasteur, France) was used throughout the study (1). The test medium was supplemented to support the growth of some fastidious microorganisms (4% globular extract [Diagnostic Pasteur] for H. influenzae; 7% horse blood for streptococci and pneumococci). A standard inoculum of 10⁴ CFU/spot was used throughout. All plates were incubated at 37°C for 24 h. The MIC was defined as the lowest concentration at which no visible growth could be detected on agar plates.

Infection models. (i) Systemic infection.

Male Charles River (St. Aubin, France) mice were used to study the antibacterial activities of compounds in the septicemia model. BALB/c and C₃H/HeOuJ strains were used in infections induced by pneumococci and by H. influenzae or Enterococcus faecalis, respectively. Infections were induced by Enterococcus faecium in C57BI/6 strains. Otherwise, CD1 mice were used. Each dosing group was composed of 7 to 10 animals weighing 20 to 22 g. The mice were infected intraperitoneally with 0.5 ml of an overnight culture appropriately diluted in physiological buffer or 5% hog mucin (Sigma) to a final cell density corresponding to 10 to 100 times the minimal lethal dose (6). Except for vancomycin, suspensions of the compounds (0.5 ml) in carboxymethyl cellulose were administered by the oral route (p.o.) immediately and 4 h postinfection. Vancomycin was subcutaneously dispensed in saline buffer. The mice were observed for 8 to 10 days following the start of the infections, and the 50% protective doses (PD₅₀s), expressed as the unit dose that protected 50% of the animals from death, were calculated by the probit method of Litchfield and Wilcoxon (9).

(ii) Thigh muscle infection.

A 1/10-diluted exponential-phase growing culture density of erythromycin A-susceptible S. aureus 011UC4 in MH agar corresponding to 10⁷ CFU/ml was injected (0.1 ml) into the right thighs of slightly ether-anesthetized immunocompetent male CD1 mice. Then, the five animals of each group received p.o. 10 mg of telithromycin or clarithromycin/kg of body weight immediately after infection. One control group received carboxymethyl cellulose instead of the tested antibiotics. Twenty-four hours later, the mice were euthanized by asphyxiation with CO₂. The thigh muscles were removed and homogenized in 0.9% NaCl. Viable-cell counts were determined on MH agar by plating correctly diluted samples. The limit of quantification was 400 CFU per ml. The results were expressed as the geometric mean log₁₀ ± standard error. Statistical comparisons were performed by Student's t test.

Time-kill curve studies in vivo.

In vivo time-kill studies with constitutively erythromycin A-resistant Streptococcus pneumoniae MV2 and H. influenzae RD7 utilized intraperitoneal challenge of CD1 mice as described above. The animals were treated p.o. 2 h after infection. Blood samples were obtained from five infected animals by retroorbital bleeding after ether anesthesia at regular intervals within 5 h. Samples were plated on adequate agar after appropriate dilution for the determination of viable bacterial-cell counts following overnight incubation in 5% CO₂. The limit of quantification was 400 CFU per ml.

Pharmacokinetic study.

Two groups of 70 approximately 6-week-old male Swiss mice (Iffa Credo, St. Germain, France) were used after being maintained on a water-only diet for 21 h before administration and 6 h afterwards. The animals were given intravenously (i.v.) in a vein of the tail or p.o. by gastric intubation a single dose of 10 mg of telithromycin/kg. The drug was administered in an aqueous solution of 2.5 mM hydrochloric acid and in a volume of 2 ml/kg. Subgroups of five mice each were anesthetized with ether and euthanized immediately at different times after i.v. and p.o. treatments. Blood was collected by carotid exsanguination and centrifuged, and the plasma was kept frozen pending analysis. Plasma concentrations were determined by a validated reverse-phase high-performance liquid chromatography method, following precipitation of plasma proteins by acetonitrile, and by fluorescence detection (excitation at 263 nm and emission at 460 nm). The method has a standard curve range of 0.05 to 10 mg/liter using 100 μl of plasma and a limit of quantification of 0.05 mg/liter. Calibration and quality controls were included over the standard range. The interbatch percent coefficient of variation for the quality control samples was between 0.0 and 2.3%. The pharmacokinetic parameters were determined from the mean plasma concentration of telithromycin using a noncompartmental model with WinNonLin software version 1.5.

C_max was defined as the maximum mean plasma drug concentration observed after p.o. administration. T_max was the time corresponding to C_max. AUC_0–z was the area under the concentration-time curve from 0 to z h. Calculated from the mean plasma concentrations by the trapezoidal rule, the AUCs were determined from time zero until the last measurable concentration-time point (C_z). T_z was the time corresponding to C_z. The terminal elimination half-life (t_1/2) the total clearance (CL), and the volume of distribution (V_z) based on the terminal phase were calculated after i.v. administration. The absolute bioavailability was calculated from the ratio of the AUC_0–z of telithromycin after p.o. and i.v. administrations, assuming that CL was the same for the two routes of administration.

RESULTS

Protection against systemic infection.

Table 1 gives the PD₅₀s of telithromycin and several reference macrolides obtained in 19 different lethal infections in mice. In infections caused by gram-positive cocci susceptible to erythromycin A, telithromycin exhibited an efficacy similar to that of clarithromycin but considerably superior to that of erythromycin A (PD₅₀s ranged from 1 to 16 mg/kg). In infections caused by gram-positive cocci resistant to erythromycin A, 14- or 16-membered-ring reference macrolides showed complete cross inactivity, except for clarithromycin against one erythromycin A-inducibly resistant S. aureus isolate.

TABLE 1.

Comparative in vivo oral antibacterial activities of telithromycin and several compounds in a murine septicemia model^a

Infective strain (CFU/mouse) and antibiotic	MIC (mg/liter)	PD50 (mg/kg) (95% CI)
S. aureus 011UC4 Erys (1.6 × 108)
Telithromycin	0.04	8 (6.5–10)
Erythromycin A	0.3	>50
Clarithromycin	0.15	6 (4.5–8)
Azithromycin	1.2	>30
Josamycin	ND	ND
Pristinamycin	ND	ND
S. aureus GO3 Eryri (2 × 108)
Telithromycin	0.08	4.5 (2–10)
Erythromycin A	>40	>50
Clarithromycin	>40	55 (25–110)
Azithromycin	>40	>100
Josamycin	2.5	>50
Pristinamycin	0.15	50 (20–165)
S. aureus 011UC4 Erys (1.6 × 108)
Telithromycin	0.04	8 (6.5–10)
Erythromycin A	0.3	>50
Clarithromycin	0.15	6 (4.5–8)
Azithromycin	1.2	>30
Josamycin	ND	ND
Pristinamycin	ND	ND
S. aureus GO3 Eryri (2 × 108)
Telithromycin	0.08	4.5 (2–10)
Erythromycin A	>40	>50
Clarithromycin	>40	55 (25–110)
Azithromycin	>40	>100
Josamycin	2.5	>50
Pristinamycin	0.15	50 (20–165)
S. aureus GR56 Eryri Oxar (1.6 × 108)
Telithromycin	0.04	6.5 (4.5–10)
Erythromycin A	>40	>50
Clarithromycin	>40	10 (0.5–15)
Azithromycin	>40	>100
Josamycin	0.6	>50
Pristinamycin	0.15	>50
S. pneumoniae UC1 Erys (5.1 × 104)
Telithromycin	≤0.01	1 (0.5–4)
Erythromycin A	0.02	>50
Clarithromycin	0.01	7.5 (0.5–16)
Azithromycin	0.04	6 (4.5–8)
Josamycin	ND	ND
Pristinamycin	ND	ND
S. pneumoniae RO1 Eryri (5 × 103)
Telithromycin	≤0.002	4 (2–6)
Erythromycin A	>40	>50
Clarithromycin	>40	>50
Azithromycin	>40	>50
Josamycin	0.04	>50
Pristinamycin	0.04	>50
S. pneumoniae SJ6 Eryrc (5 × 104)
Telithromycin	0.01	6.5 (5.5–8)
Erythromycin A	>40	>40
Clarithromycin	>40	>50
Azithromycin	>40	>50
Josamycin	>40	>50
Pristinamycin	0.15	>50
S. pneumoniae MV2 Eryrc Oxar (1.1 × 106)
Telithromycin	0.08	4 (2–9)
Erythromycin A	>40	>50
Clarithromycin	>40	>50
Azithromycin	>40	>50
Josamycin	>40	>50
Pristinamycin	0.15	>50
S. pneumoniae SJ1 Eryrc (1 × 107)
Telithromycin	0.01	24 (16.5–37)
Erythromycin A	>40	>50
Clarithromycin	>40	>50
Azithromycin	>40	>50
Josamycin	>40	>50
Pristinamycin	0.08	>50
S. pneumoniae CR29 Eryrc (1 × 105)
Telithromycin	0.3	3 (1–6)
Erythromycin A	>40	>50
Clarithromycin	>40	>50
Azithromycin	>40	>50
Josamycin	>40	>50
Pristinamycin	0.3	>50
Streptococcus pyogenes UC1 Erys (2.5 × 105)
Telithromycin	0.01	16 (9.5–33)
Erythromycin A	0.02	>50
Clarithromycin	0.08	8 (4–16.5)
Azithromycin	0.08	12.5 (9–18.5)
Josamycin	ND	ND
Streptococcus agalactiae HT3 Erys (5 × 104)
Telithromycin	0.01	9 (6.5–13)
Erythromycin	0.04	>30
Clarithromycin	ND	ND
Azithromycin	0.08	10.5 (7.5–14)
Josamycin	ND	ND
H. influenzae R1 Amps (2.5 × 107)
Telithromycin	0.6	68 (49–95)
Erythromycin A	1.2	225
Clarithromycin	5	175 (115–255)
Azithromycin	0.6	56 (69–81)
H. influenzae 14 Amps (1 × 107)
Telithromycin	1.2	25 (10–35)
Erythromycin A	2.5	>300
Clarithromycin	5	>300
Azithromycin	1.2	100 (90–110)
H. influenzae TO19 Ampr (Blact−) (1 × 107)
Telithromycin	0.6	40 (30–50)
Erythromycin A	1.2	230
Clarithromycin	5	>300
Azithromycin	1.2	145 (120–180)
H. influenzae RD7 Ampr (Blact+) (1 × 107)
Telithromycin	0.3	52 (38–70)
Erythromycin A	2.5	>200
Clarithromycin	1.2	120 (70–230)
Azithromycin	0.6	75 (45–145)
E. faecalis HT6 Erys Vans (1.6 × 108)
Telithromycin	0.08	12 (7.5–19.6)
Vancomycin	ND	ND
Pristinamycin	0.6	>25
E. faecium HT12 Eryr Vanr (1 × 109)
Telithromycin	0.04	26 (13–56)
Vancomycin	>40	>50
Pristinamycin	ND	ND
E. faecium AP9 Erys Vanr (9 × 108)
Telithromycin	≤0.04	20.5 (9–32)
Vancomycin	>40	>150
Pristinamycin	ND	ND
E. faecium IP2 Eryr Vanr (1.5 × 107)
Telithromycin	0.6	<5
Vancomycin	>40	>50
Pristinamycin	0.3	40.5 (30.5–53.5)

^aMice were infected intraperitoneally, and the drugs were administered p.o. (except vancomycin, which was administered by the subcutaneous route) immediately and 4 h postinfection. PD₅₀s were calculated by the probit method. CI, confidence interval; ND, not determined; Ery^s, erythromycin A susceptible; Ery^rc, constitutively erythromycin A resistant; Ery^ri, inducibly erythromycin A resistant; Oxa^r, oxacillin resistant; Amp^s, ampicillin susceptible; Amp^r (Blact⁻), ampicillin resistant and β-lactamase negative; Amp^r (Blact⁺), ampicillin resistant and β-lactamase positive; Van^s, vancomycin susceptible; Van^r, vancomycin resistant. 

For septicemia caused by erythromycin A-resistant pneumococci, macrolides displayed PD₅₀s which most of the time were well above 50 mg/kg. The corresponding effective doses of telithromycin ranged between 3 and 24 mg/kg, which was similar to the range of values found with erythromycin A-susceptible pathogens. Pristinamycin was unable to show measurable activity at the range of doses used.

In H. influenzae-induced infections, the ketolide was mostly 2 to more than 10 times more potent than erythromycin A or clarithromycin. The PD₅₀s ranged between 25 and 68 mg/kg, while the lowest PD₅₀ of azithromycin, the reference macrolide against this pathogen, was only 56 mg/kg.

Against enterococci, the activity of telithromycin did not depend on the phenotypes of resistance of the four strains tested and was higher than that of pristinamycin, with PD₅₀s ranging from lower than 5 to 26 mg/kg.

In vivo bactericidal activity.

After challenge with around 10⁴ CFU of erythromycin A-resistant S. pneumoniae cells/ml by the intraperitoneal route (Fig. 1), telithromycin exhibited bacteriostatic behavior for 5 h at 100 mg/kg while pristinamycin, a streptogramin B, remained inactive at the same dose. Against Haemophilus strains (Fig. 2), azithromycin, the “gold standard” macrolide against H. influenzae, and telithromycin were bacteriostatic; neither of the doses tested (100 and 200 mg/kg) led to a dramatic decrease in the initial 10⁷-CFU/ml inoculum.

FIG. 1.

In vivo time-kill curves with S. pneumoniae MV2 from the onset of treatment. Male Charles River mice were infected intraperitoneally. Compounds were administered p.o. 2 h after infection. ▴ and ▵, controls without antibiotic; ■, telithromycin (50 mg/kg); □, telithromycin (100 mg/kg); ⧫, pristinamycin (50 mg/kg); ◊, pristinamycin (100 mg/kg). The error bars indicate standard errors of the mean.

FIG. 2.

In vivo time-kill curves with H. influenzae RD7 from the onset of treatment. Male Charles River mice were infected intraperitoneally. Compounds were administered p.o. 2 h after infection. ▴ and ▵, controls without antibiotic; ■, telithromycin (100 mg/kg); □, telithromycin (200 mg/kg); ●, azithromycin (100 mg/kg); ○, azithromycin (200 mg/kg). The error bars indicate standard errors of the mean.

Protection against thigh infection.

The efficacy of telithromycin versus clarithromycin was tested in mice intramuscularly challenged with 10⁶ CFU of erythromycin A-susceptible S. aureus/thigh. Antibiotics were administered p.o. to five mice per group immediately after bacterial challenge, and viable bacteria in the thigh muscles were measured versus untreated controls 24 h after the treatment. Both telithromycin (MIC, ≤0.04 mg/liter) and clarithromycin (MIC, 0.3 mg/liter) at 10 mg/kg produced a significant decrease (P < 0.01) in bacterial counts after 24 h (2.75 ± 0.48 and 2.05 ± 0.65 log₁₀ CFU/thigh, respectively, versus 6.92 ± 0.54 CFU/thigh for the untreated controls).

Pharmacokinetic parameters.

The pharmacokinetic parameters of telithromycin following a single i.v. or p.o. administration of 10 mg/kg in mice are reported in Table 2. The mean precision of the plasma telithromycin assay was 6.5 to 11.6% (percent coefficient of variation). After i.v. administration, the mean concentration of telithromycin was 7.30 ± 0.44 mg/liter at 5 min. A short distribution phase and a slow decrease of concentrations were then observed. Two hours later, the mean concentration of telithromycin was still around 1/3 of the initial value. The last measurable concentration was observed 8 h after administration and represented around 1% of the initial value (Fig. 3). The AUC_0–z was 14.82 mg · h/liter. CL was 0.024 liters/h. t_1/2 was 1.2 h, and the volume of distribution associated with the elimination phase was 0.043 liters, corresponding to 1.6 times the animals' body size. After p.o. administration, the observed C_max was 2.91 mg/liter, with a corresponding T_max of 1.5 h. From that time on, the concentrations decreased slowly and regularly. The last measurable concentration (0.094 ± 0.028 mg/liter) was observed 8 h after administration and still represented 3% of C_max (Fig. 3). The AUC_0–z was 7.15 mg · h/liter, and the bioavailability of telithromycin was 53%.

TABLE 2.

Pharmacokinetic parameters of telithromycin following a single i.v. or p.o. administration of 10 mg/kg in the male Swiss mouse

Parameter	Units	Value
		i.v.	p.o.
Cmax	mg/liter	7.30 ± 0.44a	2.91 ± 0.22
Tmax	h		1.5
Cz	mg/liter	0.086 ± 0.027	0.094 ± 0.028
Tz	h	8	8
AUC0–z	mg · h/liter	14.82	7.15
CL	liter/h	0.024
	liter · h−1 · kg−1	0.80
t1/2	h	1.2
Vz	liters	0.043
	liters/kg	1.41
Bioavailability	%		53

^aConcentration at first sampling time (5 min). 

FIG. 3.

Mean concentration-time curves for telithromycin following a single i.v. (▴) or p.o. (□) administration of 10 mg/kg in the male Swiss mouse. Error bars, ± standard error of the mean.

DISCUSSION

Ketolides are new antibacterial agents specifically designed for treating respiratory tract infections induced by macrolide-resistant pathogens. Focusing only on S. pneumoniae, and depending on the country, more than 30% of isolates are now resistant to macrolides, including recent molecules such as clarithromycin and azithromycin (10, 12). Because of the emergence of penicillin-resistant strains, β-lactams can no longer be used as the first-choice therapy. Even though new quinolones are more active against pneumococci, they are still contraindicated for pregnant women or young children. Thus, a true medical need exists for new drugs active against S. pneumoniae.

Telithromycin has been extensively shown to be active in vitro against all of the respiratory pathogens (2–4, 14). We report here the in vivo activities of telithromycin in systemic and localized animal models, together with its pharmacokinetic profile.

In murine septicemia caused by erythromycin A-resistant isolates, all macrolides were completely inactive, as expected. Even pristinamycin, despite its good MICs, did not show any therapeutic efficacy at the doses tested, probably due to disadvantageous kinetics. For staphylococci, streptococci, and pneumococci, telithromycin was the most active compound against all the strains tested, whatever their phenotype of macrolide resistance. The range of PD₅₀s was from 1 to 24 mg/kg, while the MICs rank from ≤0.002 to 0.3 mg/liter. Consequently, the amount of the drug in the blood, shown as the major determinant of ketolide efficacy (O. Vesga, W. A. Craig, and C. Bonnat, Abstr. 37th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-255, p.189, 1997), can easily explain this pharmacodynamic efficacy by leading to large AUCs with peak concentrations that remain much higher than the MICs. Moreover, in four systemic H. influenzae infections, the increasing MICs (0.3 to 1.2 mg/liter) gave similarly increased PD₅₀s, ranging from 25 to 68 mg/kg. In these infections, telithromycin was at least as active as azithromycin, as reported by others in murine H. Influenzae pneumonia, demonstrating the predictive value of the septicemia model (13; S. Miyazaki, H. Okamoto, and K. Yamaguchi, Abstr. 38th Intersci. Conf. Antimicrob. Agents Chemother., abstr. E-139, p. 209, 1998).

The emergence of vancomycin resistance in Enterococcus spp. has become a major problem in antibiotherapy. The septicemia model allowed us to highlight the in vivo activity of telithromycin against enterococci, including vancomycin-resistant E. faecium. The highest PD₅₀ calculated in four different infections was only 26 mg/kg. Our results are in good agreement with data obtained in a similar model in mice (B. E. Murray, K. V. Singh, and K. K. Zscheck, Abstr. 38th Intersci. Conf. Antimicrob. Agents Chemother., abstr. B-13, p. 48, 1998).

In the murine septicemia model induced by intraperitoneal injection of bacteria, telithromycin was bacteriostatic against erythromycin A-resistant S. pneumoniae or H. influenzae only within the first 5 h of treatment. Such results were predicted by in vitro studies (5, 11) which reported only slow time-dependent bactericidal activity for the ketolide (99.9% kill at 24 h) at 10 times the MIC, but not with all strains studied.

In soft tissue localized infections, such as thigh infection, telithromycin was still efficient, as previously presented in this model induced by pneumococci (O. Vesga, D. Andes, and W. A. Craig, Abstr. 37th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-258, p. 189, 1997). Our experiment reports an efficacy for telithromycin similar to that of clarithromycin against S. aureus susceptible to macrolides, demonstrating the capability of the ketolide to diffuse in and cure infected tissues.

In conclusion, we confirmed in experimental infections the in vitro activity of the ketolide antimicrobial telithromycin. Its efficacy favorably combined high activity against important respiratory pathogens, in particular multidrug-resistant pneumococci and H. influenzae, and good pharmacokinetic properties.