Abstract
We investigated the in vitro postantibiotic effects (PAEs) of the ketolides telithromycin (HMR 3647) and HMR 3004 and analyzed the results using the sigmoid Emax model. Mean maximum telithromycin PAEs against erythromycin A-susceptible strains of Staphylococcus aureus, Streptococcus pyogenes, and Streptococcus pneumoniae were 3.7, 8.9, and 9.7 h, respectively, while maximum PAEs for erythromycin A-resistant strains were much shorter. Mean maximum HMR 3004 PAEs were 3.2 to 4.4 h for all species.
Ketolides are a novel class of oral antibiotics. They are semisynthetic macrolides that have good activity against erythromycin A-resistant bacteria, including multidrug-resistant pneumococci, Streptococcus pyogenes, and Staphylococcus aureus (1, 8, 14, 15). We investigated the in vitro postantibiotic effects (PAEs) of the ketolides telithromycin (HMR 3647) and HMR 3004 against Streptococcus pneumoniae, Streptococcus pyogenes, and Staphylococcus aureus and compared the duration of PAEs and the potency of the agents in producing PAEs against erythromycin A-susceptible and erythromycin A-resistant strains.
(Part of this work was presented at the 37th Interscience Conference on Antimicrobial Agents and Chemotherapy [abstr. F-252] and at the 21st International Congress of Chemotherapy [abstr. P92].)
The bacteria tested were four strains each of S. pneumoniae and S. pyogenes and five strains of S. aureus, including the reference strains S. aureus ATCC 25923, S. pyogenes ATCC 19615, and S. pneumoniae ATCC 49619. Apart from the ATCC strains, all were recent clinical isolates from the Monash Medical Centre. Two isolates of each species were erythromycin A susceptible, and two or three isolates were erythromycin A resistant. S. aureus strains 3 and 4 were resistant to erythromycin A by an inducible mechanism, detected by a standard disk approximation method (2); all other erythromycin A-resistant strains had constitutive erythromycin A resistance.
Telithromycin and HMR 3004 were obtained from Hoechst Marion Roussel, Romainville, France, and were stored and prepared according to the manufacturer's guidelines. Ketolide MICs were determined by the broth macrodilution technique, according to the National Committee for Clinical Laboratory Standards (NCCLS) guidelines (13). The broths used in experiments were Mueller-Hinton broth (BBL, Becton Dickinson Co., Cockeysville, Md.) for S. aureus, and brain heart infusion broth (Oxoid, Basingstoke, Hampshire, England) for the streptococci. The MICs were determined in ambient air not supplemented with CO2, because the presence of 5 to 7% CO2 has been shown to consistently increase MICs of ketolide for pneumococci by two- to fourfold (14). MICs of erythromycin A were determined by E-test (4), using a suspension of 0.5 McFarland turbidity plated onto Mueller-Hinton agar (S. aureus) or Mueller-Hinton agar supplemented with 5% blood (streptococci).
The in vitro PAEs were determined by the viable plate count method (7), using Mueller-Hinton broth (S. aureus) or brain heart infusion broth (streptococci). Logarithmic-phase organisms (106 CFU) were exposed for 1 h at 37°C to seven concentrations of the ketolide (0.5, 1, 4, 8, 16, 32, and 64 times the MIC). After 1 h, the drug was removed by centrifuging the solution for 10 min at 2,000 × g, decanting the supernatant, and resuspending the organisms in fresh broth prewarmed to 37°C. This washing procedure was performed twice. Washing was selected as the preferred method of drug removal to avoid carryover of the drug from the high concentrations of drug used in the experiments. After drug removal, viable counts were plated hourly until visible regrowth had occurred. The following controls were included for each experiment: (i) a growth control, prepared and treated identically to the test solution but without exposure to antibiotic, and (ii) a residual antibiotic control, to which 1/1,000 of the test antimicrobial concentration was added after centrifugation and washing. The latter tube was included to ensure that, after centrifugation and washing, residual drug in the tubes containing the treated organism did not affect the rate of growth. The PAE was calculated with the standard formula of Craig and Gudmundsson: PAE = T − C, where T is the time required for the count of CFU in the test culture to increase 1 log10 above the count observed immediately after drug removal and C is the time required for the count of CFU in an untreated control culture to increase 1 log10 above the count observed immediately after completion of the same procedure used on the test culture for drug removal (7).
As the methodology of determining PAE in the laboratory is time-consuming and labor-intensive, we used a mathematical model to predict the PAE for any particular antimicrobial concentration or duration of exposure, based on the data obtained experimentally for the same organism and antimicrobial at other concentrations and durations of exposure. The Hill (sigmoid Emax) equation (9) is a mathematical model that has been used to describe this relationship: PAE = (PAEmax × AUCn)/(E50n + AUCn), where PAEmax is the estimated maximum PAE, E50 is the estimated area under the concentration-time curve (AUC) of drug exposure at which 50% of the maximum PAE is reached, and n is a constant associated with the steepness of the exposure-response curve. When this equation has been applied to results obtained in some of the original publications on PAE (5, 11), the correlation between the PAEmax estimated using this equation and the experimentally obtained PAEmax was very good, with correlation coefficients in the range of 0.96 to 0.995 (16). This equation has therefore subsequently been used by us to describe the relationship between PAE and AUC of drug exposure (12). In this study, the sigmoid Emax equation was used to mathematically model the exposure-response curve for the ketolides. The parameters PAEmax, E50, and n for each bacterial strain were estimated with the nonlinear regression module of Systat (version 8.0; SPSS Inc., Chicago, Ill.). MICs were then compared to each of the three parameters by linear regression performed using Systat, with and without the inclusion of a constant.
The PAEs after 1 h of exposure to telithromycin for strains of S. aureus, S. pyogenes, and S. pneumoniae are shown in Table 1 and Fig. 1 and 2. Maximum PAEs for the streptococci ranged from 3.2 to 10.1 h, while maximum PAEs for S. aureus were shorter than the streptococcal PAEs, ranging from 3.5 to 4.3 h. Table 1 also shows the correlation between the PAEs determined experimentally and those estimated using the Hill (sigmoid Emax) equation, with corrected r2 ranging from 0.857 to 0.978.
TABLE 1.
Telithromycin (HMR 3647) PAE parameters
| Strain | Erythromycin A susceptibilitya | MIC (mg/liter) | Range of PAEsb (h) | PAEmax (h)c | E50 (mg · h/liter)c | Corrected r2c |
|---|---|---|---|---|---|---|
| S. aureus | ||||||
| ATCC 25923 | S | 0.125 | 0.7–3.4 | 3.7 | 1.03 | 0.923 |
| 3B | S | 0.125 | 1.1–3.5 | 3.5d | 0.23 | 0.857 |
| 4B | S | 0.063 | 0.5–3.8 | 4.4 | 0.53 | 0.978 |
| 2E | Ri | 0.125 | 0.8–3.5 | 3.8d | 2.65 | 0.879 |
| 2F | Ri | 0.063 | 0.1–3.2 | 4.3 | 1.46 | 0.969 |
| S. pyogenes | ||||||
| ATCC 19615 | S | 0.016 | 0.8–10.8 | 10.1 | 0.05 | 0.955 |
| 3B | S | 0.031 | 0.6–8.3 | 7.7 | 0.07 | 0.950 |
| 2C | Rc | 0.5 | 2.2–3.2 | 3.2 | 0.24 | 0.856 |
| 2B | Rc | 0.5 | 2.2–3.8 | 3.8 | 0.22 | 0.972 |
| S. pneumoniae | ||||||
| ATCC 49619 | S | 0.008 | 0–9.5 | 9.9 | 0.09 | 0.989 |
| 5306e | S | 0.016 | 0.9–9.3 | 9.5 | 0.11 | 0.982 |
| 5231 | Rc | 0.125 | 1.6–6.3 | 6.6 | 0.20 | 0.960 |
| 5050e | Rc | 0.0625 | 1.0–6.6 | 6.2 | 0.12 | 0.949 |
S, susceptible; Rc, constitutively resistant; Ri, inducibly resistant.
0.5 to 64 times the MIC.
PAEmax, n, and E50 were estimated by nonlinear least-squares regression, and corrected r2 shows the degree of fit with the experimental data, using Systat (version 8.0).
To achieve a satisfactory curve fit, PAEmax was fixed at the maximum value observed experimentally.
These strains also had reduced susceptibility to penicillin.
FIG. 1.
PAEs induced by 1 h of exposure to various concentrations of telithromycin (HMR 3647). (a) S. aureus ATCC 25923; (b) S. pyogenes ATCC 19615; (c) S. pneumoniae ATCC 49619.
FIG. 2.
Relationship between duration of PAE after 1 h of exposure to telithromycin (HMR 3647) and telithromycin concentration (expressed as multiples of the MIC). (a) Strains of S. aureus; (b) Strains of S. pyogenes; (c) Strains of S. pneumoniae.
PAEs after 1 h of exposure to HMR 3004 are shown in Table 2 and Fig. 3 and 4. Maximum PAEs for S. aureus, S. pyogenes, and S. pneumoniae ranged from 2.7 to 4.9 h. PAEs against streptococci were much shorter than those of telithromycin, but there was little difference between the drugs in duration of PAEs against staphylococci. Once again, there was good correlation between actual PAEs and PAEs estimated using the mathematical model, with corrected r2 ranging from 0.931 to 0.995.
TABLE 2.
HMR 3004 PAE parameters
| Strain | Erythromycin A susceptibilitya | MIC (mg/liter) | Range of PAEsb (h) | PAEmax (h)c | E50 (mg · h/liter)c | Corrected r2c |
|---|---|---|---|---|---|---|
| S. aureus | ||||||
| ATCC 25923 | S | 0.063 | 0–3.6 | 3.3 | 0.16 | 0.990 |
| 3B | S | 0.125 | 0.3–3.0 | 3.1 | 0.54 | 0.995 |
| 2A | Ri | 0.063 | 0.3–3.0 | 2.7 | 0.15 | 0.951 |
| 2E | Ri | 0.063 | 0.2–3.6 | 4.5 | 1.03 | 0.928 |
| S. pyogenes | ||||||
| ATCC 19615 | S | 0.016 | 0.3–3.7 | 3.8 | 0.06 | 0.981 |
| 3A | S | 0.031 | 0.6–3.7 | 4.0 | 0.16 | 0.976 |
| 2A | Rc | 0.125 | 0.5–3.8 | 3.9 | 0.25 | 0.931 |
| 2B | Rc | 0.125 | 1.0–4.2 | 4.9 | 0.47 | 0.960 |
| S. pneumoniae | ||||||
| ATCC 49619 | S | 0.008 | 0.3–4.0 | 4.3 | 0.04 | 0.988 |
| 3Ad | S | 0.016 | 0.2–3.5 | 3.7 | 0.05 | 0.993 |
| 4A | Rc | 0.016 | 0.3–3.5 | 3.9 | 0.09 | 0.987 |
| 4Bd | Rc | 0.016 | 0.4–4.0 | 4.4 | 0.09 | 0.992 |
S, susceptible; Rc, constitutively resistant; Ri, inducibly resistant.
0.5 to 64 times the MIC.
PAEmax, n, and E50 were estimated by nonlinear least-squares regression, and corrected r2 shows the degree of fit with the experimental data, using Systat (version 8.0).
These strains also had reduced susceptibility or resistance to penicillin.
FIG. 3.
PAEs induced by 1 h of exposure to various concentrations of HMR 3004. (a) S. aureus ATCC 25923; (b) S. pyogenes ATCC 19615; (c) S. pneumoniae ATCC 49619.
FIG. 4.
Relationship between duration of PAE after 1 h of exposure to HMR 3004 and HMR 3004 concentration (expressed as multiples of the MIC). (a) Strains of S. aureus ATCC 25923; (b) Strains of S. pyogenes ATCC 19615; (c) Strains of S. pneumoniae ATCC 49619.
A comparison of MIC, PAEmax, and E50 between erythromycin A-susceptible and -resistant strains is shown in Table 3. For S. aureus, the effects of resistance were similar for both drugs, with a negligible effect on MIC or PAEmax but a two- to fourfold increase in E50. For S. pyogenes and telithromycin, although the MICs for resistant strains were increased more than 20-fold, PAEmax was reduced only about twofold, and E50 was increased only about fourfold. In contrast, with HMR 3004, MICs for resistant strains increased about sixfold but there was no effect on either PAEmax or E50. For S. pneumoniae, the only observable effect of erythromycin A resistance was an eightfold increase in the MIC of telithromycin; the PAE parameters remained unaffected.
TABLE 3.
Comparison of MICs and PAE parameters between erythromycin A-resistant and susceptible strains
| Agent and species | Erythromycin A susceptibility | Geometric mean MIC (mg/liter) | R/S ratio for MICa | Arithmetic mean PAEmax (h) | S/R ratio for PAEmaxb | Geometric mean E50 (mg · h/liter) | R/S ratio for E50c | E50/MIC ratiod |
|---|---|---|---|---|---|---|---|---|
| Telithromycin | ||||||||
| S. aureus | S | 0.099 | 0.89 | 3.9 | 0.96 | 0.50 | 3.91 | 5.1 |
| R | 0.088 | 4.0 | 1.97 | 22.3 | ||||
| S. pyogenes | S | 0.022 | 22.63 | 8.9 | 2.52 | 0.06 | 3.88 | 2.7 |
| R | 0.5 | 3.5 | 0.23 | 0.5 | ||||
| S. pneumoniae | S | 0.011 | 8.00 | 9.7 | 1.51 | 0.10 | 1.58 | 9.0 |
| R | 0.088 | 6.4 | 0.16 | 1.8 | ||||
| HMR 3004 | ||||||||
| S. aureus | S | 0.088 | 0.71 | 3.2 | 0.89 | 0.38 | 2.35 | 4.3 |
| R | 0.063 | 3.6 | 0.89 | 14.2 | ||||
| S. pyogenes | S | 0.022 | 5.66 | 3.9 | 0.89 | 0.78 | 0.80 | 35.2 |
| R | 0.125 | 4.4 | 0.62 | 5.0 | ||||
| S. pneumoniae | S | 0.011 | 1.41 | 4.0 | 0.98 | 0.90 | 0.92 | 81.2 |
| R | 0.016 | 4.1 | 0.82 | 53.0 |
Ratio of geometric mean MICs of erythromycin A-resistant to -susceptible strains.
Ratio of geometric mean PAEmaxs of erythromycin A-susceptible to resistant strains.
Ratio of geometric mean E50s of erythromycin A-resistant to -susceptible strains.
Ratio of geometric mean E50s to geometric mean MICs.
In a previous study, we demonstrated a significant relationship between E50 and the MIC of imipenem (12). However, resistant strains were not included. For the ketolides, when examined by linear regression, correlation between MIC and PAEmax, n, and E50 in particular was poor except for HMR 3004 when regression was performed through the origin (without a constant), yielding a moderate correlation with all three parameters. In many instances regression detected outliers. Both of these findings, and an examination of E50/MIC ratios for erythromycin A-susceptible and -resistant strains, as shown in Table 3, suggest that the MIC changes observed with erythromycin-resistant strains do not necessarily predict changes in postantibiotic growth suppression. This may relate to differences in resistance mechanisms within and between species.
Others have previously examined postantibiotic growth suppression of telithromycin, but in less detail. One study examining PAEs of HMR 3647 (now named telithromycin) found in vitro PAEs of 3.7 to 8.2 h after exposure of S. aureus, S. pyogenes, and S. pneumoniae to 10 times the MICs of the drug, with S. pneumoniae having longer PAEs than the other bacteria (3). However, this study did not analyze strains by erythromycin A resistance phenotype and did not test concentrations of drug greater than 10 times the MIC, which we predict would not have detected the maximum effect. Jacobs et al. have also demonstrated PAEs after 1 h of exposure to telithromycin at 10 times the MIC of between 0.4 and 2.7 h for S. pyogenes, 1.5 and 3.8 h for S. pneumoniae, and 0.3 and 2.4 h for S. aureus, and also showed a sub-MIC PAE for all species (M. R. Jacobs, S. Bajaksouzian, J. Chuang, M. P. Ronchetti, and P. C. Appelbaum, Abstr. 37th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-253, 1997). This study included a variety of resistant strains.
There are limited pharmacokinetic and pharmacodynamic data in humans for the ketolides telithromycin and HMR 3004. Typical steady-state plasma AUCs in humans after 800-mg once-a-day oral dosing of telithromycin are 4 to 11 mg · h/liter (B. Lenfant, E. Sultan, C. Wable, M. H. Pascual, and B. H. Meyer, Abstr. 38th Intersci. Conf. Antimicrob. Agents Chemother., abstr. A-49, 1998; E. Sultan, B. Lenfant, C. Wable, M. H. Pascual, and B. H. Meyer, Abstr. 21st Int. Congr. Chemother., abstr. P66, 1999). In our study, AUCs of telithromycin achieved with 64 times the MIC and an exposure time of 1 h ranged from 3.84 to 8 mg · h/liter (S. aureus), 1.024 to 32 mg · h/liter (S. pyogenes), and 0.5 to 8 mg · h/liter (S. pneumoniae). The duration of PAEs in our study generally peaked with concentrations of 32 to 64 times the MIC, concentrations which, according to the pharmacokinetic data, can readily be achieved in humans. Other human pharmacokinetic studies show that levels of telithromycin are 10- to 100-fold higher in lung than in plasma (C. Serleys, C. Cantalloube, P. Soler, F. Lemaitre, H. P. Gia, F. Brunner, and A. Andremont, Abstr. 21st Int. Congr. Chemother., abstr. P78, 1999) and that intracellular levels of the drug are 350 times higher in polymorphonuclear leukocytes than in the extracellular space (17). Thus, the concentrations of telithromycin we used in vitro are readily achievable both in human plasma and in tissues, and results obtained are likely to be clinically relevant.
Studies performed in the murine neutropenic thigh-infection model have shown that the 24 h AUC/MIC ratio is the pharmacokinetic/pharmacodynamic parameter that best correlates with the activity of telithromycin against S. pneumoniae and that this parameter is a much better predictor of efficacy than time above MIC (O. Vesga, W. A. Craig, and C. Bonnat, Abstr. 37th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-255, 1997). This study did not measure the in vivo PAE, but if the prolonged PAE of telithromycin against streptococci is present in vivo, it may help explain this finding. Similar studies do not appear to have been performed for HMR 3004.
Although the presence or absence of in vivo PAE is usually predicted by in vitro studies (6), in vitro studies such as ours have several limitations, including the exposure of bacteria to fixed concentrations of drug, and the immediate rather than gradual removal of drug, neither of which occur in a living system. Our results should therefore be confirmed in animal models, as host defense mechanisms, concentrations of drug within the cell, tissue binding of drug, and sub-MIC effects may be important. Also, we have not tested durations of exposure to drug other than 1 h. In vivo, the duration of exposure is likely to be significantly longer than this. However, previous studies have shown that the duration of the PAE is strongly correlated with the AUC of the drug, the product of concentration of drug versus duration of exposure to drug (5, 10, 12). Therefore, for example, doubling the concentration of the drug is likely to have the same effect as doubling the duration of the exposure to the drug.
In conclusion, the ketolides telithromycin and HMR 3004 exhibit long concentration-dependent in vitro PAEs against S. aureus, S. pyogenes, and S. pneumoniae. When results for the same strain were directly compared, telithromycin had substantially longer PAEs than HMR 3004, particularly against streptococci. Also, telithromycin had much longer PAEs against erythromycin A-susceptible strains of streptococci than against erythromycin A-resistant strains, although these differences were not seen with staphylococci.
Acknowledgments
This work was supported by a grant from Hoechst Marion Roussel, Division of Clinical Anti-Infectives, Paris, France.
We are grateful for the assistance of André Bryskier in the design of this study and the supply of some resistant strains.
REFERENCES
- 1.Agouridas C, Bonnefoy A, Chantot J F. Antibacterial activity of RU64004 (HMR 3004), a novel ketolide derivative active against respiratory pathogens. Antimicrob Agents Chemother. 1997;41:2149–2158. doi: 10.1128/aac.41.10.2149. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bonnefoy A, Girard A M, Agouridas C, Chantot J F. Ketolides lack the inducibility properties of the MLSB resistance phenotype. J Antimicrob Chemother. 1997;40:85–90. doi: 10.1093/jac/40.1.85. [DOI] [PubMed] [Google Scholar]
- 3.Boswell F J, Andrews J M, Wise R. Pharmacodynamic properties of HMR 3647, a novel ketolide, on respiratory pathogens, enterococci and Bacteroides fragilis demonstrated by studies of time-kill kinetics and postantibiotic effect. J Antimicrob Chemother. 1998;41:149–153. doi: 10.1093/jac/41.2.149. [DOI] [PubMed] [Google Scholar]
- 4.Brown D F J, Brown L. Evaluation of the Etest, a novel method for quantifying antimicrobial activity. J Antimicrob Chemother. 1991;27:185–190. doi: 10.1093/jac/27.2.185. [DOI] [PubMed] [Google Scholar]
- 5.Bundtzen R W, Gerber A U, Cohn D L, Craig W A. Postantibiotic suppression of bacterial growth. Rev Infect Dis. 1981;3:28–37. doi: 10.1093/clinids/3.1.28. [DOI] [PubMed] [Google Scholar]
- 6.Craig W A. Postantibiotic effects in experimental infection models: relationship to in vitro phenomena and to treatment of infections in man. J Antimicrob Chemother. 1993;31(Suppl. D):138–147. doi: 10.1093/jac/31.suppl_d.149. [DOI] [PubMed] [Google Scholar]
- 7.Craig W A, Gudmundsson S. The postantibiotic effect. In: Lorian V, editor. Antibiotics in laboratory medicine. Baltimore, Md: Williams and Wilkins; 1996. pp. 296–329. [Google Scholar]
- 8.Hamilton-Miller J M T, Shah S. Comparative in vitro activity of ketolide HMR 3647 and four macrolides against Gram-positive cocci of known erythromycin susceptibility status. J Antimicrob Chemother. 1998;41:649–653. doi: 10.1093/jac/41.6.649. [DOI] [PubMed] [Google Scholar]
- 9.Holford N H G, Sheiner L B. Understanding the dose-effect relationship: clinical application of pharmacokinetic-pharmacodynamic models. Clin Pharmacokinetics. 1981;6:429–453. doi: 10.2165/00003088-198106060-00002. [DOI] [PubMed] [Google Scholar]
- 10.Kuenzi B, Segessemann C, Gerber A U. Postantibiotic effect of roxithromycin, erythromycin and clindamycin against selected Gram-positive bacteria and Haemophilus influenzae. J Antimicrob Chemother. 1987;20(Suppl. B):39–46. doi: 10.1093/jac/20.suppl_b.39. [DOI] [PubMed] [Google Scholar]
- 11.McDonald P J, Craig W A, Kunin C M. Persistent effect of antibiotics on Staphylococcus aureus after exposure for limited periods of time. J Infect Dis. 1977;135:217–223. doi: 10.1093/infdis/135.2.217. [DOI] [PubMed] [Google Scholar]
- 12.Munckhof W J, Olden D, Turnidge J D. The postantibiotic effect of imipenem: relationship with drug concentration, time of exposure, and MIC. Antimicrob Agents Chemother. 1997;41:1735–1737. doi: 10.1128/aac.41.8.1735. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.National Committee for Clinical Laboratory Standards. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. 3rd ed. Vol. 10. 1997. , no. 8. Approved standard M7-A3. National Committee for Clinical Laboratory Standards, Villanova, Pa. [Google Scholar]
- 14.Reinert R R, Bryskier A, Lutticken R. In vitro activities of the new ketolide antibiotics HMR 3004 and HMR 3647 against Streptococcus pneumoniae in Germany. Antimicrob Agents Chemother. 1998;42:1509–1511. doi: 10.1128/aac.42.6.1509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Schülin T, Wennerstein C B, Moellering R C, Jr, Eliopoulos G M. In vitro activity of RU 64004, a new ketolode antibiotic, against gram-positive bacteria. Antimicrob Agents Chemother. 1997;41:1196–1202. doi: 10.1128/aac.41.5.1196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Turnidge J D. Prediction of antibiotic dosing intervals from in vitro susceptibility, pharmacokinetics and postantibiotic effect: theoretical considerations. Scand J Infect Dis. 1991;74(Suppl.):137–141. [PubMed] [Google Scholar]
- 17.Vazifeh D, Preira A, Bryskier A, Labro M T. Interactions between HMR 3647, a new ketolide, and human polymorphonuclear neutrophils. Antimicrob Agents Chemother. 1998;42:1944–1951. doi: 10.1128/aac.42.8.1944. [DOI] [PMC free article] [PubMed] [Google Scholar]