Abstract
In vitro studies were designed to investigate the influence of peak drug concentration (Cmax), the area under the concentration-time curve (AUC), and, consequently, the trough concentration on the bactericidal effects of gentamicin against Enterobacter cloacae (MIC, 0.5 mg/liter) by simulating bolus versus infusion administration and bolus dosing with altered drug clearance. Bacteria in the lag phase were exposed to gentamicin concentration-time profiles modelling either bolus or infusion dosing (AUC constant, Cmax changing) with 30-min postdose peak concentrations (Cpeak30) of 4, 6, 8, and 10 mg/liter or bolus dosing with normal and double drug clearance (Cmax constant, AUC changing) corresponding to normal clearance profiles with Cpeak30 of 6 and 8 mg/liter. Exposure to gentamicin caused early bactericidal effects apparent by 2 h, followed by variable bacteriostatic and recovery phases. Exposure to bolus profiles resulted in greater bactericidal activity than the corresponding infusion profile up to a Cpeak30 of 8 mg/liter. At a Cpeak30 of 10 mg/liter, there were no differences in bactericidal effect. Double clearance profiles had a reduced bactericidal effect at 6 mg/liter compared to the corresponding normal clearance profile, but no differences in bactericidal effect were observed for 8-mg/liter double and normal clearance profiles. These results suggest that the initial exposure (i.e., 0 to 30 min) is a more important determinant for bacterial killing than the AUC or trough concentration for this bacterium. Subject to confirmation of these findings with other gram-negative bacteria, to optimize aminoglycoside efficacy the initial exposure (Cmax) should be maximized by giving higher doses or bolus administration at intervals which may not produce detectable trough concentrations. Clinical trials with a broad range of patients, especially those with higher clearance, would confirm these in vitro observations and define optimal dosing recommendations.
Recent interest in alternative dosing practices indicates that the past use of aminoglycosides has been suboptimal. Dosing nomograms and therapeutic drug monitoring guidelines have not been well based on a comprehensive understanding of the relevant pharmacodynamics. By identifying those pharmacodynamic parameters that determine efficacy and toxicity and relating them to pharmacokinetic principles, it should be possible to optimize aminoglycoside use and improve patient outcomes.
Aminoglycosides demonstrate concentration-dependent killing (5, 6, 8, 12) and a prolonged concentration-dependent postantibiotic effect which varies according to bacterium (3, 15). For aminoglycosides, antibacterial efficacy has been shown to be dependent on the maximum concentration attained in serum (Cmax) (1, 9, 14), the area under the concentration-time curve (AUC) (4, 13), and some trough concentration contribution (7, 10). Optimal dosage design relies on the determination of the relative importance of these pharmacokinetic parameters.
Dependence on Cmax suggests that the method of administration (rapid bolus rather than slow infusion) as well as the size of the dose is important for bacterial killing, whereas dependence on the AUC suggests that the size of the dose alone is important. AUC is also a function of drug clearance, and dependence on AUC demands higher doses for patients with increased rates of clearance, i.e., patients with cystic fibrosis, burns, and severe sepsis. Conversely, dependence on Cmax suggests that, provided the dose achieves an adequate Cmax, the dose would not need to be altered in response to increased clearance.
We report studies investigating independently the contributions of Cmax and AUC to the antibacterial effect of gentamicin against Enterobacter cloacae. The methods involved in vitro concentration-time modelling of profiles with constant AUC and variable Cmax (i.e., the same dose given by bolus or infusion administration) and constant Cmax and variable AUC (i.e., the same dose given by bolus administration with altered rates of clearance).
MATERIALS AND METHODS
E. cloacae ATCC 13047 with a MIC and an MBC of 0.5 mg of gentamicin per liter was cultured overnight in brain heart infusion broth (Oxoid, Basingstoke, England). The culture was diluted to 107 CFU/ml in 0.1% peptone water (Difco Laboratories, Detroit, Mich.), and a 1-ml sample was added to the experimental culture broth, resulting in an initial density of 106 CFU/ml.
Firstly, in vitro concentration-time modelling of clinical concentration-time bolus and infusion profiles was performed as previously described by Bastone and coworkers (1). The 8-mg/liter bolus profile had a target 30-min-postdose peak concentration (Cpeak30) of 8.1 mg/liter, a Cmax of 28 mg/liter, and an AUC of 36.6 mg · h/liter. The corresponding infusion profile had a target Cpeak30 of 7.7 mg/liter, a Cmax of 12.6 mg/liter, and an AUC of 34.9 mg · h/liter. Linear extrapolations of the above profiles were performed to generate profiles targeting Cpeak30s of 4, 6, and 10 mg/liter (Table 1).
TABLE 1.
Summary of results for bolus versus infusion profilesa
| Dose exposure levelb | Cmax (mg/liter) | AUC0–48.5 (mg · h/liter) | Cpeak30 (mg/liter) | Bacterial CFU/ml (log10) at h:
|
||||
|---|---|---|---|---|---|---|---|---|
| 0 | 2.5 | 10.5 | 24.5 | 48.5 | ||||
| Control (no drug exposure) | ||||||||
| Bolus | 0 | 0 | 0 | 5.97 ± 0.13 | 6.60 ± 0.16 | 8.86 ± 0.21 | 9.11 ± 0.12 | 9.25 ± 0.20 |
| Infusion | 0 | 0 | 0 | 5.86 ± 0.11 (NSc) | 6.86 ± 0.25 (NSc) | 8.86 ± 0.09 (NSc) | 9.16 ± 0.12 (NSc) | 9.37 ± 0.12 (NSc) |
| 4-mg/liter profile | ||||||||
| Bolus | 14.28 ± 0.91 | 19.68 ± 1.7 | 3.88 ± 0.45 | 6.01 ± 0.15 | 1.11 ± 0.86 | 0.43 ± 1.04 | 6.28 ± 2.63 | 8.44 ± 1.70 |
| Infusion | 6.98 ± 0.67 (P < 0.001c) | 17.88 ± 1.06 (NSc) | 3.37 ± 0.72 (NSc) | 5.89 ± 0.10 (NSc) | 2.12 ± 0.29 (P = 0.012c) | 3.05 ± 2.49 (P = 0.04d) | 7.05–3.65 (NSc) | 8.76 ± 1.66 (NSc) |
| 6-mg/liter profile | ||||||||
| Bolus | 21.22 ± 3.60 | 28.21 ± 1.49 | 5.78 ± 0.35 | 5.86 ± 0.07 | NGD | 0.29 ± 0.70 | 1.65 ± 1.81 | 3.61 ± 3.06 |
| Infusion | 8.20 ± 0.20 (P < 0.001c) | 27.20 ± 1.69 (NSc) | 5.22 ± 0.82 (NSc) | 5.81 ± 0.10 (NSc) | 0.61 ± 1.00 (NSd) | 1.76 ± 1.55 (P = 0.047d) | 8.33 ± 0.85 (P < 0.001c) | 9.27 ± 0.09 (P = 0.001d) |
| 8-mg/liter profile | ||||||||
| Bolus | 30.55 ± 3.35 | 34.46 ± 2.19 | 7.67 ± 0.37 | 5.99 ± 0.17 | NGD | NGD | NGD | NGD |
| Infusion | 12.52 ± 0.57 (P < 0.001c) | 36.20 ± 1.23 (NSc) | 7.15–0.78 (NSc) | 5.87 ± 0.11 (NSc) | 0.29 ± 0.70 (NSd) | 0.75 ± 1.17 (NSd) | 4.03 ± 2.68 (P = 0.002c) | 8.44 ± 1.39 (P < 0.001c) |
| 10-mg/liter profile | ||||||||
| Bolus | 37.52 ± 1.50 | 41.58 ± 1.60 | 10.55 ± 0.62 | 5.90 ± 0.07 | NGD | NGD | NGD | NGD |
| Infusion | 14.92 ± 0.54 (P < 0.001c) | 39.56 ± 2.07 (NSc) | 9.85 ± 0.38 (P = 0.04c) | 5.92 ± 0.05 (NSc) | NGD (NSc) | NGD (NSc) | NGD (NSc) | NGD (NSc) |
n = 6. Values are shown as means ± standard deviations. NS, not significant (>0.05); NGD, no growth detected.
Based on target Cpeak30.
Independent t test.
Mann-Whitney test.
The second series of studies involved characterizing the bolus profile with a two-compartment pharmacokinetic model, doubling the elimination rate constant, and simulating a bolus profile with a doubled rate of drug clearance by nonlinear regression analysis (Sigmaplot; SPSS Scientific Software, Chicago, Ill.). In vitro concentration-time modelling crudely emulated the pharmacokinetic simulation (1). Double clearance profiles targeted the same Cmax as the corresponding bolus profile with half the AUC. As a result of higher clearance, the targeted Cpeak30s were lower than the corresponding normal clearance bolus profile Cpeak30s (Table 2).
TABLE 2.
Summary of results for normal versus double clearance bolus profilesa
| Dose exposure levelb | Cmax (mg/liter) | AUC0–48.5 (mg · h/liter) | Cpeak30 (mg/liter) | Bacterial CFU/ml (log10) at h:
|
||||
|---|---|---|---|---|---|---|---|---|
| 0 | 2.5 | 10.5 | 24.5 | 48.5 | ||||
| Control (no drug exposure) | ||||||||
| Normal clearance | 0 | 0 | 0 | 5.97 ± 0.13 | 6.60 ± 0.16 | 8.86 ± 0.21 | 9.11 ± 0.12 | 9.25 ± 0.20 |
| Double clearance | 0 | 0 | 0 | 5.96 ± 0.07 (NSc) | 6.38 ± 0.31 (NSc) | 8.71 ± 0.35 (NSc) | 9.11 ± 0.13 (NSc) | 9.24 ± 0.17 (NSc) |
| 6-mg/liter profile | ||||||||
| Normal clearance | 21.22 ± 3.60 | 28.21 ± 1.49 | 5.78 ± 0.35 | 5.86 ± 0.07 | NGD | 0.29 ± 0.70 | 1.65 ± 1.81 | 3.61 ± 3.06 |
| Double clearance | 20.73 ± 0.61 (NSc) | 14.79 ± 0.63 (P < 0.001c) | 3.93 ± 0.35 (P < 0.001c) | 5.86 ± 0.07 (NSc) | 0.95 ± 1.05 (NSd) | 2.80 ± 0.41 (P = 0.001d) | 8.71 ± 0.06 (P < 0.001c) | 9.27 ± 0.21 (P = 0.001d) |
| 8-mg/liter profile | ||||||||
| Normal clearance | 30.55 ± 3.35 | 34.46 ± 2.19 | 7.67 ± 0.37 | 5.99 ± 0.17 | NGD | NGD | NGD | NGD |
| Double clearance | 28.4 ± 1.36 (NSc) | 18.66 ± 1.43 (P < 0.001c) | 5.65 ± 0.46 (P < 0.001c) | 5.92 ± 0.05 (NSc) | 0.30 ± 0.74 (NSd) | NGD (NSc) | NGD (NSc) | NGD (NSc) |
n = 6. Values shown are means ± standard deviations. NS, not significant (>0.05); NGD, no growth detected.
Based on normal clearance target Cpeak30.
Independent t test.
Mann-Whitney test.
Viable cell counts were performed as previously described by Bastone and coworkers (1). Gentamicin concentrations were measured by an EMIT immunoassay (Solaris; Syva Laboratories, San Jose, Calif.) at 0, 0.16, 0.33, 0.5, 2.5, 4.5, 6.5, 8.5, 10.5, 24.5, and 48.5 h for the bolus and double clearance profiles and at −0.5, −0.25, 0, 0.17, 0.33, 0.5, 2.5, 4.5, 6.5, 8.5, 10.5, 24.5, and 48.5 h for the infusion profiles. All results were compared to those for control cultures not exposed to gentamicin. All in vitro experiments involved six replicates.
The AUC from 0 to 48.5 h was calculated with the trapezoidal rule. The Student t test and the Mann-Whitney test were used for statistical analyses depending on the distribution of data.
RESULTS
For all profiles, bacterial growth in the absence of gentamicin showed a plateau at or near 109 CFU/ml. Within each exposure level, a graded response was observed in the presence of aminoglycoside (Fig. 1).
FIG. 1.
Target gentamicin concentration-time profiles (i) and time course of bacterial counts (ii) in studies with E. cloacae ATCC 13047 modelling bolus and infusion dosing associated with 30-min-postdose concentrations of 6 (a) and 8 (b) mg/liter (n = 6).
Bolus and infusion profiles with target Cpeak30s of 4 and 6 mg/liter and infusion profiles with target Cpeak30 of 8 mg/liter caused an early bactericidal effect, followed by a bacteriostatic phase and regrowth. Exposure to bolus profiles with target Cpeak30s of 8 and 10 mg/liter and to an infusion profile with a target Cpeak30 of 10 mg/liter caused complete bacterial killing with no regrowth. There were significant differences in mean bacterial counts among the target 4-mg/liter Cpeak30 profiles at 2.5 and 10.5 h; the 6-mg/liter profiles at 10.5, 24.5, and 48.5 h; and the 8-mg/liter profiles at 24.5 and 48.5 h. The differences in Cmax values between bolus and infusion profiles were significant for all profiles investigated (P < 0.05). Cpeak30s and AUC0–48.5 were not significantly different over the range of exposures investigated, except for the Cpeak30s at the 10-mg/liter level (Fig. 1 and Table 1).
The 6-mg/liter double clearance profile caused an early bactericidal effect followed by a bacteriostatic phase and regrowth, whereas the 8-mg/liter double clearance profile caused complete killing with no regrowth to 48.5 h. There were significant differences in the mean bacterial counts between the 6-mg/liter double clearance and 6-mg/liter bolus profiles at 10.5, 24.5, and 48.5 h. There were no significant differences in bacterial counts observed between the 8-mg/liter profiles. The differences in Cmax values between the double clearance and bolus profiles were not significant for all profiles investigated. Cpeak30s and AUC0–48.5 were significantly different for all profiles investigated (P < 0.001) (Fig. 2 and Table 2).
FIG. 2.
Target gentamicin concentration-time profiles (i) and time course of bacterial counts (ii) in studies with E. cloacae ATCC 13047 modelling bolus dosing with single clearance and doubled gentamicin clearance corresponding to 30-min-postdose concentrations of 6 (a) and 8 (b) mg/liter from single clearance bolus dosing (n = 6).
DISCUSSION
In vitro concentration-time modelling of bolus (high Cmax) versus infusion (low Cmax) dosing allowed the AUC to be held constant as the Cmax varied. Following exposure to 8-mg/liter profiles, there was evidence of complete bacterial killing with the bolus profile but regrowth with the infusion profile. This suggests that administration of gentamicin as a bolus (high Cmax) dose should be more effective than administration as a 30-min infusion (low Cmax). At 10 mg/liter, the difference between exposure profiles disappeared, suggesting that the method of administration may become irrelevant in higher-dose regimens. These results are consistent with previous observations involving gentamicin and Escherichia coli and tobramycin and Pseudomonas aeruginosa (1, 14). In this study, as with E. coli, a Cpeak30/MIC ratio of ∼16:1 was necessary for total killing (1), whereas other workers have reported Cpeak30/MIC ratios of 10:1 to be necessary for killing of other bacteria including P. aeruginosa (2, 11, 14).
In vitro concentration-time modelling of bolus administration (i.e., 1× AUC) with normal clearance versus bolus administration with doubled drug clearance (i.e., 0.5× AUC) allowed the Cmax to be held constant but the AUC to vary. At 6 mg/liter, the bolus profile resulted in significantly reduced bacterial growth from 10.5 to 48.5 h, compared to that for the corresponding double clearance profile. However, there were no significant differences observed in bacterial counts for the 8-mg/liter profiles. These results suggest that the antibacterial effect of gentamicin against E. cloacae may be unaffected for doubled drug clearance with doses of ≥2.0 mg/kg of body weight given as an intravenous bolus injection. However, at lower doses or in tissue compartments where drug concentrations are more attenuated than those in blood, a doubled rate of clearance may result in less-effective suppression of bacterial regrowth.
The independent effect of Cmax has been reported elsewhere (1, 9, 14), but this paper distinguishes it as a significant determinant of efficacy, separate from the AUC. As manipulation of Cmax, with constant AUC, resulted in marked changes in antibacterial response and halving AUC, with constant Cmax, had minimal effect on antibacterial activity, it appears that Cmax exerts an independent and more important effect than does AUC in determining the antibacterial activity of gentamicin versus E. cloacae.
The role of a trough concentration as a determinant of efficacy is minimized where an optimal Cmax is achieved. However, in tissue infections a lower Cmax would be achieved, and therefore trough concentrations may be an important determinant of bacterial killing (7). This demonstration of significant dependence of bacterial killing on Cmax and lesser importance of the subsequent exposure including trough concentration requires confirmation with other gram-negative bacteria both in vitro and in vivo. If confirmed, however, this would endorse administration regimens in which higher peak concentrations are attained at extended dose intervals despite the lower-to-negligible trough concentrations.
In summary, these in vitro studies suggest that the initial exposure (i.e., 0 to 30 min) is a more important determinant for bacterial killing than the AUC or trough concentration. The general applicability of these results to the clinical setting awaits confirmation in studies with other gram-negative bacteria. Clinical interpretation of these findings would suggest that attempts to optimize aminoglycoside efficacy would require maximizing the initial exposure (Cmax and Cpeak30) by bolus administration or giving higher doses at intervals which may not produce detectable trough concentrations. Clinical trials with a broad range of patients, including those with higher clearance, would confirm these in vitro observations and help define precise dosing recommendations.
REFERENCES
- 1.Bastone E B, Li S C, Ioannides-Demos L L, Spicer W J, McLean A J. Kill kinetics and regrowth patterns of Escherichia coli exposed to gentamicin concentration-time profiles simulating in vivo bolus and infusion dosing. Antimicrob Agents Chemother. 1993;37:914–917. doi: 10.1128/aac.37.4.914. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Begg E J, Peddie B A, Chambers S T, et al. Comparison of gentamicin dosing regimens using an in vitro model. J Antimicrob Chemother. 1992;29:427–433. doi: 10.1093/jac/29.4.427. [DOI] [PubMed] [Google Scholar]
- 3.Craig, W. A. 1993. Post antibiotic effects in experimental infection models: relationship to in vitro phenomena and treatment of infections in man. J. Antimicrob. Chemother. 31(Suppl. D):149–158. [DOI] [PubMed]
- 4.Craig, W. A., J. Redington, and S. C. Ebert. 1991. Pharmacodynamics of amikacin in vitro and in mouse thigh and lung infections. J. Antimicrob. Chemother. 27(Suppl. C):29–40. [DOI] [PubMed]
- 5.Hancock R E W. Aminoglycoside uptake and mode of action with special reference to streptomycin and gentamicin. I. Antagonists and mutants. J Antimicrob Chemother. 1981;8:249–276. doi: 10.1093/jac/8.4.249. [DOI] [PubMed] [Google Scholar]
- 6.Hancock R E W. Aminoglycoside uptake and mode of action with special reference to streptomycin and gentamicin. II. Effects of aminoglycosides on cells. J Antimicrob Chemother. 1981;8:429–445. doi: 10.1093/jac/8.6.429. [DOI] [PubMed] [Google Scholar]
- 7.Ioannides-Demos L L, Liolios L, Wood P, Spicer W J, McLean A J. Changes in MIC alter responses of Pseudomonas aeruginosa to tobramycin exposure. Antimicrob Agents Chemother. 1998;42:1365–1369. doi: 10.1128/aac.42.6.1365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kadurugamuwa J L, Clarke A J, Beveridge T J. Surface action of gentamicin on Pseudomonas aeruginosa. J Bacteriol. 1993;175:5798–5805. doi: 10.1128/jb.175.18.5798-5805.1993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.McLean A J, Ioannides-Demos L L, Li S C, Bastone E B, Spicer W J. Bactericidal effect of gentamicin peak concentration provides a rationale for administration of bolus doses. J Antimicrob Chemother. 1993;32:301–305. doi: 10.1093/jac/32.2.301. [DOI] [PubMed] [Google Scholar]
- 10.McLean A J, Bastone E B, Ioannides-Demos L L, Spicer W J. Bactericidal effect of gentamicin trough concentration provides a rationale for administration of bolus doses and maintenance of trough levels. J Antimicrob Chemother. 1994;33:999–1004. doi: 10.1093/jac/33.5.999. [DOI] [PubMed] [Google Scholar]
- 11.Moore R D, Lietman P S, Smith C R. Clinical response to aminoglycoside therapy: importance of the ratio of peak concentration to minimal inhibitory concentration. J Infect Dis. 1987;155:93–99. doi: 10.1093/infdis/155.1.93. [DOI] [PubMed] [Google Scholar]
- 12.Taber H W, Mueller J P, Miller P F, Arrow A S. Bacterial uptake of aminoglycoside antibiotics. Microbiol Rev. 1987;51:439–457. doi: 10.1128/mr.51.4.439-457.1987. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Vogelman B S, Gudmundsson S, Turnidge J, Leggett J, Ebert S, Craig W A. Correlation of antimicrobial pharmacokinetic parameters with therapeutic efficacy in an animal model. J Infect Dis. 1988;158:831–847. doi: 10.1093/infdis/158.4.831. [DOI] [PubMed] [Google Scholar]
- 14.Wood P J, Ioannides-Demos L L, Bastone E B, Spicer W J, McLean A J. Kill kinetics and regrowth patterns of Pseudomonas aeruginosa exposed to concentration-time profiles of tobramycin simulating in vivo infusion and bolus dosing. Antimicrob Agents Chemother. 1996;40:1321–1324. doi: 10.1128/aac.40.5.1321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zhanel G G, Hoban D J, Harding G K M. The postantibiotic effect: a review of in vitro and in vivo data. DICP-Ann Pharmacother. 1991;25:153–163. doi: 10.1177/106002809102500210. [DOI] [PubMed] [Google Scholar]