In Vivo Pharmacodynamic Profiling of Doripenem against Pseudomonas aeruginosa by Simulating Human Exposures

Abstract

Doripenem is a new broad-spectrum carbapenem with activity against a range of gram-negative pathogens, including nonfermenting bacteria such as Pseudomonas aeruginosa. The objective of this study was to evaluate simulated human exposures to doripenem using a neutropenic murine thigh infection model against 24 clinical P. aeruginosa isolates with a wide range of MICs. Dosing regimens in mice were designed to approximate the free time above MIC (fT>MIC) observed with 500 mg doripenem every 8 h given as either a 1-h or 4-h intravenous infusion in humans. Maximal antibacterial killing was associated with doripenem exposures of ≥40% fT>MIC; bacteriostatic effects were noted at ≈20% fT>MIC. The simulated 1-h infusion provided bactericidal effects for isolates with MICs of ≤2 μg/ml, while variable killing was noted for isolates with MICs of 4 to 8 μg/ml and regrowth for isolates with an MIC of 16 μg/ml. The 4-h infusion regimen displayed similar killing for isolates with MICs of ≤2 μg/ml and enhanced activity for two of the four isolates with an MIC of 4 μg/ml. Given that the 4-h regimen yields negligible fT>MIC for MICs of ≥8 μg/ml, regrowth was generally observed. Simulated doses of 500 mg doripenem every 8 h infused over 1 h demonstrated antibacterial killing for P. aeruginosa isolates with MICs of 0.125 to 8 μg/ml. Exposures of ≥40% fT>MIC resulted in the most pronounced bactericidal effects, while killing was variable for 20 to 30% fT>MIC. Infusing doses over 4 h enhanced efficacy against selected pseudomonal isolates with an MIC of 4 μg/ml.

The resistance crisis confronting today's practice for treating infectious diseases is well exemplified by the pathogen Pseudomonas aeruginosa. Afflicting immunocompromised hosts as the leading gram-negative pathogen in nosocomial infections, P. aeruginosa also possesses traits that favor antimicrobial resistance (23). An inherently resistant organism already, it is capable of adapting to antibiotics and consequently can develop resistance during therapy (4, 17). The ample stock of defensive mechanisms ranges from impermeability to numerous β-lactamases and, most notably, efflux pumps, all of which can occur concurrently and further enhance resistance (4, 17). Resistance to broad-spectrum agents like imipenem and ceftazidime, desirable for their pseudomonal activity, increased over the period from 1998 to 2003 to rates of 21.1% and 31.9%, respectively, in the United States (21). Moreover, the proportion of multidrug-resistant P. aeruginosa isolates, defined as isolates resistant to at least three different drug classes, has risen by 10% over a period of 10 years nationwide (22).

Despite the desperate need for novel antibacterial agents, attention has been diverted away from bacterial infectious diseases, evidenced by the 57% drop in new antimicrobials to arrive on the market over the past 25 years (5). Currently, there are only 12 antimicrobial compounds in development and fewer still offered against nonfermenting gram-negative organisms (25). Doripenem (Johnson & Johnson Pharmaceutical Research & Development, LLC, Raritan, NJ) is a recently Food and Drug Administration-approved carbapenem antibiotic with activity against gram-positive and gram-negative organisms reminiscent of that of imipenem and meropenem, respectively (13, 16). It displays in vitro potency against P. aeruginosa that is two- to fourfold greater than that of available carbapenems (13, 16). In a previously published time-kill study, doripenem was shown to exhibit time-dependent antibacterial activity, consistent with what has been discovered and widely accepted for the carbapenem class (9, 16, 28). With the advent of dose optimization, innovative dosing strategies like prolonged and continuous infusion are being considered for time-dependent β-lactams, in efforts to maximize pharmacodynamic exposure and enhance the utility of antibiotics (19). Various prolonged infusion dosing schemes have been evaluated for doripenem in phase 1 studies and have been carried over into phase 3 clinical trials for the treatment of nosocomial infections (6, 12).

The primary objective of this study was to investigate the efficacy of doripenem against P. aeruginosa by simulating human exposures in a neutropenic murine thigh infection model over a wide range of MICs. As dosing strategies such as prolonged infusion of β-lactams are being increasingly used, we also aimed to compare efficacies of standard doses of doripenem administered as a 1-h infusion against that of a prolonged 4-h infusion.

MATERIALS AND METHODS

Antimicrobial test agents.

Standard analytical-grade doripenem (Johnson & Johnson Pharmaceutical Research & Development, LLC, Raritan, NJ) was used for all in vitro and in vivo analyses. Doripenem powder was weighed out and reconstituted with normal saline to achieve desired concentrations immediately prior to each in vivo experiment. Final solutions were kept at room temperature, protected from light, and discarded 10 h after reconstitution.

Bacterial isolates and in vitro susceptibility.

Twenty-four clinical P. aeruginosa isolates collected from Hartford Hospital (Hartford, CT) in 2006 were used throughout the study. Doripenem MICs were determined using the microdilution method according to CLSI guidelines with cation-adjusted Mueller-Hinton broth (CAMHB [20 to 25 mg/liter calcium, 10 to 12.5 mg/liter magnesium]) at a standard inoculum (10⁵ CFU/ml) in ambient air (8). P. aeruginosa ATCC 27853 was used for quality control purposes. A minimum of three independent tests were performed for each isolate, from which the modal MIC was obtained and utilized.

Thigh infection model.

Specific-pathogen-free, female ICR mice weighing approximately 25 g were acquired from Harlan Sprague Dawley, Inc. (Indianapolis, IN), and utilized throughout these experiments. The animals were maintained and used in accordance with National Research Council recommendations, and they were provided with food and water ad libitum. Mice were rendered transiently neutropenic by intraperitoneal injections of cyclophosphamide at 150 and 100 mg/kg of body weight at 4 days and 1 day, respectively, prior to inoculation (1). Three days before inoculation, the mice also received a single intraperitoneal injection of uranyl nitrate at 5 mg/kg to induce a predictable degree of renal impairment (1). A suspension of each test isolate was prepared from a fresh subculture that had been incubated overnight and diluted to achieve a final inoculum of 10⁶ CFU/ml. Thigh infection was produced by a single 0.1-ml intramuscular injection of inoculum into each mouse thigh 2 h prior to the initiation of antimicrobial therapy.

Pharmacokinetic studies and dosing regimen determination.

Mice were prepared as described for the thigh infection model. P. aeruginosa ATCC 27853 was used to produce thigh infection for the pharmacokinetic studies. Single doses of doripenem at 10, 50, and 150 mg/kg in 0.2-ml volumes were administered subcutaneously 2 h after inoculation. Blood samples were collected by intracardiac puncture from groups of 6 to 12 mice per time point for a total of eight time points per dose over 4 h. Serum samples were separated after centrifugation and stored at −80°C until analysis.

Concentrations of doripenem in murine serum were determined using a previously validated high-performance liquid chromatography assay (24). The assay was linear over a range of 0.5 to 40 μg/ml (R² = 0.995). Intraday coefficients of variation for the low (1-μg/ml) and high (30-μg/ml) quality control samples were 4.02% and 4.44%, respectively. Interday coefficients of variation for the quality control samples were 5.44% and 5.93%, respectively.

Pharmacokinetic parameters for single doses of doripenem were calculated using first-order elimination, by a nonlinear least-squares techniques (WinNonlin version 5.0.1; Pharsight, Mountain View, CA). Compartment model selection was based on visual inspection of the fit and correlation between the observed and calculated concentrations by Akaike's information criterion. Mean pharmacokinetic parameters were calculated from the individual parameter estimates and applied to WinNonlin in order to simulate various doripenem exposures expressed as free time above the MIC (fT>MIC) over a wide range of MICs. Dosing regimens in mice were designed to approximate fT>MIC observed in humans following 500 mg doripenem every 8 h given as either a 1-h or a prolonged 4-h intravenous infusion. Human concentration-time profiles were derived from data for 24 healthy human volunteers (data on file, Johnson & Johnson Pharmaceutical Research & Development, LLC), and protein binding was previously estimated to be 8.5% in humans and 25.2% in mice (3, 14).

The resulting regimen for 1-h infusions of doripenem involved six doses at concentrations of 10, 15, 2.5, 1.25, 0.5, and 0.25 mg/kg of body weight administered at 0, 0.5, 2.5, 4, 5.5, and 7 h every 8 h over 24 h. The 4-h infusion regimen comprised three 8-h intervals of eight doses at concentrations of 5.5, 2.25, 4.5, 4.5, 4.5, 4.5, 0.75, and 0.375 mg/kg given at 0, 0.5, 1.5, 2.5, 3.5, 4.5, 6, and 7.5 h over 24 h. Prior to the efficacy studies, confirmation of the exposures attained by these simulated dosing regimens was performed in a separate pharmacokinetic experiment with infected mice.

Efficacy as assessed by bacterial density.

Two hours after infection, 1-h and 4-h doripenem infusion regimens were administered as subcutaneous 0.2-ml injections and studied for each isolate in groups of three mice over a 24-h period. Twenty-four isolates were studied with the 1-h infusion, while 11 isolates were selected for the 4-h infusion based on doripenem MICs of ≥2 μg/ml, in order to accurately characterize the differences between the two dosing regimens. Control animals received sterile normal saline with the same volume, route, and schedule as the active-drug regimens. Untreated control mice (three per group) were sacrificed just prior to antibiotic initiation (0 h) and after 24 h. After the 24-h treatment period, all animals were euthanized by CO₂ exposure, followed by cervical dislocation. After sacrifice, thighs were removed and individually homogenized in normal saline. Serial dilutions of the thigh homogenate were plated on Trypticase soy agar with 5% sheep blood for CFU determination. Efficacy, designated as the change in bacterial density, was calculated as the log₁₀ change in bacterial CFU/ml obtained for doripenem-treated mice after 24 h from the preantibiotic CFU/ml measured for 0-h control animals. Differences in efficacy between the 1-h and 4-h infusions were assessed with the Student t test or Mann-Whitney rank sum test for rank-based data using SigmaStat version 2.0 (SPSS Inc., Chicago, IL). A P value less than 0.05 was considered significant for the statistical analysis.

Supplemental in vitro studies: MIC fractionation and time-kill.

Additional in vitro experiments were performed with the P. aeruginosa isolates with an MIC of 4 μg/ml, where the distinction between the 1-h and prolonged 4-h infusion was most significant. MIC fractionation tests were executed by the microdilution method as mentioned previously, but drug concentrations were varied purposely to produce MICs with segmenting intervals of 0.25 and 0.5 μg/ml (i.e., 2, 2.25, 2.5, 2.75, 3…4, 4.5, 5, 5.5, 6 μg/ml). Time-kill analyses were conducted by using different doripenem concentrations with a fixed inoculum in test tubes that were incubated at 37°C in ambient air for 24 h. Concentrations of 1×, 2×, and 7× MIC were examined, along with positive growth controls devoid of antibiotic. Doripenem concentrations were chosen to evaluate the effects of the peak concentration (C_max) for the different infusion regimens. The inoculum was prepared from an overnight growth of bacteria in CAMHB that had been adjusted to a 0.5 McFarland standard (1 × 10⁸ CFU/ml) and was then added to drug- and broth-containing tubes to achieve a final inoculum of 10⁶ CFU/ml. Throughout the 24-h incubation period, 0.1-ml samples were extracted at 0, 2, 4, 6, 12, and 24 h. Samples were then serially diluted, plated onto Trypticase soy agar with 5% sheep blood, and incubated for 24 h in order to obtain viable-colony counts. The lower limit of detection was set at 10² CFU/ml.

RESULTS

In vitro susceptibility.

Doripenem MICs for all P. aeruginosa study isolates are shown in Table 1. The depicted MICs of the 24 organisms provided for a wide range (0.125 to 16 μg/ml) of susceptibilities, against which an array of pharmacodynamic exposures could be evaluated.

TABLE 1.

MICs of P. aeruginosa test isolates with corresponding fT>MICs for doripenem (500 mg) as 1- and 4-h infusions in humans and mice

Isolate	MICa (μg/ml)	fT>MIC (%) of doripenem
		1-h infusion		4-h infusion
		Humansb	Micec	Humansb	Micec
993	0.125	100	100	100	100
1029	0.125	100	100	100	100
869	0.25	90	90	100	100
886	0.25	90	90	100	100
1055	0.5	72.5	75	95	95
1062	0.5	72.5	75	95	95
935	1	55	55	80	82.5
1005	1	55	55	80	82.5
1052	1	55	55	80	82.5
1082	1	55	55	80	82.5
906	2	40	42.5	65	70
979	2	40	42.5	65	70
1036d	2	40	42.5	65	70
1058d	2	40	42.5	65	70
931d	4	27.5	30	55	52.5
944d	4	27.5	30	55	52.5
1050d	4	27.5	30	55	52.5
1060d	4	27.5	30	55	52.5
896	8	17.5	20	0	0
936d	8	17.5	20	0	0
988d	8	17.5	20	0	0
1006d	8	17.5	20	0	0
927d	16	7.5	10	0	0
968d	16	7.5	10	0	0

^aMode of three or more independent MIC experiments.

^bDerived from data on file by Johnson & Johnson Pharmaceutical Research & Development, LLC, Raritan, NJ.

^cDerived from pharmacokinetic modeling with WinNonlin.

^dAlso utilized in the 4-h-infusion experiments.

Pharmacokinetic determination.

A one-compartment model was used to characterize the free concentration-time profile of doripenem in murine serum. Linear, dose-proportional pharmacokinetics were observed over the range of single doses tested (Table 2), and mean parameters were determined from these doses and used to design dosing regimens in mice. The free concentration-time profiles of the resulting murine regimens (equivalent to 1-h and 4-h infusions of 500 mg doripenem) and the corresponding exposures in humans are displayed in Fig. 1. The C_max for the 1-h infusion regimen in mice (28.4 μg/ml) was approximately equal to that observed in humans (21.8 μg/ml). The prolonged 4-h infusion regimen had a less pronounced C_max at 7.99 and 7.79 μg/ml for mice and humans, respectively. The free areas under the curve for 1-h and 4-h infusions of doripenem were relatively similar in mice (37.6 and 34.4 μg·h/ml, respectively) and humans (32.6 and 32.4 μg·h/ml, respectively). However, these dosing regimens were customized to match specifically the percentage of time that free drug concentrations are maintained above the MIC (fT>MIC), within 2.5% of that in humans at each MIC, as presented in Table 1.

TABLE 2.

Pharmacokinetic parameters of single-dose doripenem in mice^a

Dose (mg/kg)	Cmax (μg/ml)	Tmax (h)	AUC0-24 (μg·h/ml)	V (ml/kg)	kel (h−1)	t1/2 (h)	CL (ml/h/kg)
10	16.3	0.25	13.5	400.7	1.68	0.41	673.9
50	59.7	0.38	62.4	384.7	2.04	0.34	783.3
150	194.4	0.22	168.0	545.8	1.56	0.44	850.5

^aAUC_0-24, area under the concentration-time curve for 0 to 24 h; CL, clearance; C_max, peak concentration; k_el, elimination rate constant; t_1/2, half-life; T_max, time of peak concentration; V, volume of distribution.

FIG. 1.

Free concentration-time profiles of 500 mg doripenem in healthy human volunteers versus mice. (A) One-hour infusion; (B) 4-h infusion. Each value is the mean ± standard deviation for six to eight infected mice; dotted lines represent the 95% confidence intervals for human data.

Efficacy as assessed by bacterial density.

The mean starting bacterial density in the P. aeruginosa-infected thighs at 0 h was 5.14 ± 0.24 log₁₀ CFU/ml, with a range from 4.69 to 5.50 log₁₀ CFU/ml for all of the study isolates. The degree of bacterial growth was generally 3 logs, which progressed to a mean of 8.60 ± 0.45 log₁₀ CFU/ml (7.40 to 9.46 log₁₀ CFU/ml) at 24 h. Due to the aggressively virulent features of the pseudomonal infection, a percentage of untreated control animals failed to survive through the 24-h period and were subsequently harvested at an earlier time. However, bacterial counts obtained from these animals were still viable, as there were no differences relative to those that survived to the 24-h time point. All animals in the treatment groups were able to receive the entire 24-h course of doripenem for both regimens, as there was no premature mortality.

The dose-response relationship between fT>MIC and the corresponding change in bacterial density for the simulated 1-h infusion is pictured in Fig. 2. The efficacy of doripenem was correlated with various fT>MIC exposures (R² = 0.788), as bactericidal effects, a 3-log killing by definition, was associated with ≥40% fT>MIC. Static exposures were established at ≈20% fT>MIC, where there is no net change in bacterial density at the end of treatment from starting bacterial counts, though actual results fluctuated between growth and killing for 20 to 30% fT>MIC.

FIG. 2.

Relationship between change in log₁₀ CFU/ml and fT>MIC for a 1-h doripenem infusion regimen. Each value represents 6 to 12 infected thighs.

The efficacy of the 1-h infusion regimen is shown in Fig. 3, defined as the change in log₁₀ number of CFU/ml from 0-h control data. Overall, a 1- to 3-log reduction in bacterial load was observed at 24 h for MICs of 0.125 to 8 μg/ml. Antibacterial activity was fairly consistent with approximately a 3-log decrease, specifically at MICs of 0.125 to 2 μg/ml, while results were decidedly variable at 4 to 8 μg/ml, for which fT>MIC was moderately less (20 to 30%). At the highest MIC (16 μg/ml), bacteria steadily grew by 2 to 3 logs.

FIG. 3.

Change in log₁₀ CFU/ml for the 1-h doripenem infusion regimen at 24 h compared with 0-h controls for P. aeruginosa isolates (n = 24) with an MIC range of 0.125 to 16 μg/ml. Each value is the mean ± standard deviation for 6 to 12 infected thighs for an individual isolate.

Additionally, the comparative efficacies of simulated 1-h and 4-h infusions of doripenem are depicted in Fig. 4. At an MIC of 2 μg/ml, the 4-h infusion produced bactericidal results nearly identical to those of the 1-h infusion, despite statistically significant differences (P < 0.05). This, however, is reflective only of differences in the 0-h control data, as both dosing regimens decreased the bacterial load to the limit of detection. Moving up a dilution to 4 μg/ml, the prolonged infusion had significantly greater bacterial killing (−1.5 to −2.5-log decrease) for two of the four isolates (1050 and 1060), attributable to the higher fT>MIC (P < 0.05). The remaining two organisms (931 and 944) demonstrated ≈1-log regrowth rather than reduction, contrary to the predicted enhanced efficacy with the 4-h infusion. Lastly, there was an overall rise in bacterial density at MICs of 8 to 16 μg/ml, given the absence of any fT>MIC with the 4-h infusion. Bacterial effects of the 4-h infusion varied significantly from those of the 1-h infusion for one of the three organisms with an MIC of 8 μg/ml, while differences between dosing regimens at an MIC of 16 μg/ml were attributable only to the degree of regrowth.

FIG. 4.

Comparison of change in log₁₀ CFU/ml from 0-h controls for 1-h versus 4-h doripenem infusion regimens at 24 h for P. aeruginosa isolates (n = 11) with an MIC range of 2 to 16 μg/ml. Each value is the mean ± standard deviation for 6 to 18 infected thighs; a pair of values was obtained for each individual isolate. §, 931; §§, 944; †, 1050; ††, 1060. Asterisks indicate statistically significant differences (P < 0.05).

Supplemental in vitro studies.

MIC fractionation experiments revealed similar results among the four pseudomonal isolates that were previously classified by standard doubling dilutions as having an MIC of 4 μg/ml. The fractionated MICs for isolates 931, 944, 1050, and 1060 were 3, 3, 4, and 2.75 μg/ml, respectively. Moreover, time-kill curves for these isolates revealed similar killing profiles over the 24 h of observation.

DISCUSSION

As doripenem is one of the latest antimicrobials to combat P. aeruginosa, we characterized the pharmacodynamic profile of doripenem against this pathogen in simulated human exposures, with a secondary goal of assessing the efficacy of a prolonged infusion dosing scheme. Traditional time-dependent antibiotics like β-lactams are driven by the pharmacodynamic parameter fT>MIC in order to generate desirable antibacterial action (9, 11, 19, 28). Though no β-lactam necessitates a full 100% fT>MIC, the extent of fT>MIC does vary for each β-lactam class; the carbapenems appear to require the least, with 20% and 40% for static and bactericidal effects, respectively (9, 11, 19, 28). Like that of other carbapenems, doripenem efficacy was related to fT>MIC (9, 28). Bactericidal activity was noted at 40% fT>MIC, whereas static results occurred at approximately 20 to 30% fT>MIC. This efficacy relationship to fT>MIC has been reported by other investigators using a similar animal model (2).

Dose optimization of β-lactams by extending the duration of infusion has been shown to enhance pharmacodynamic exposures against organisms residing at higher MICs, which is highly probable against a pathogen such as P. aeruginosa (18, 19, 20). Based on the comparative fT>MICs between the simulated 1-h and 4-h infusions, the greatest divergence occurred at MICs of 4 and 8 μg/ml, where the fT>MIC was considerably higher with the prolonged infusion for 4 μg/ml (55%), exceeding the pharmacodynamic target of 40%, and effectively nonexistent for 8 μg/ml (0%), as the peak concentration becomes less prominent with a more slowly infused dose. The predicted enhanced efficacy at 4 μg/ml was observed in two of the four isolates, and regrowth was produced, as expected, by the majority of the isolates at 8 μg/ml. Of note was a single exception at an MIC of 8 μg/ml, where the 4-h infusion resulted in bacterial reduction in spite of the deficient fT>MIC. This discrepancy was likely due to the nature of the model, since concentrations of the simulated 4-h infusion so closely border 8 μg/ml for a substantial period that given even the small pharmacokinetic variations in the animals, it would be possible to have not only bacterial killing but also improved killing activity. Additionally, while the MIC of this particular organism was determined to be 8 μg/ml by doubling dilutions, the actual MIC may have been closer to 4 μg/ml, a concentration amenable to the pharmacokinetic profile of the prolonged infusion.

With respect to the isolates with an MIC of 4 μg/ml, supplementary MIC fractionation studies offered no explanation for the discordant results observed with the prolonged infusion. All four isolates with regrowth (931 and 944) and with killing (1050 and 1060) revealed highly similar MICs that ranged from 2.75 to 4 μg/ml. If the specific MICs for those opposing organisms had ventured nearer to 8 μg/ml, one could logically contribute the failure of the 4-h infusion to its phenotypic profile. Another consideration that was investigated was the impact of concentration, as the C_max for the 1-h infusion was ≈28 μg/ml (seven times the MIC of 4 μg/ml), while the C_max was far less for the 4-h infusion, only ≈8 μg/ml (twice the MIC of 4 μg/ml). However, time-kill curves for these isolates revealed a profile expected of a time-dependent agent, with minimal increases in killing activity alongside rising concentrations (10). Moreover, kill curves for the isolates with regrowth (931 and 944) were nearly indistinguishable from those with killing (1050 and 1060) with the 4-h infusion. Interestingly, there was minor regrowth of bacteria seen at the 24-h time point for organism 944. However, this occurred with the 7× MIC concentration, which is indicative of the C_max of the 1-h and not the 4-h infusion.

Other time-kill investigations with β-lactams and P. aeruginosa have agreed that the killing activity of these time-dependent drugs is not extensively enhanced with increasing concentrations (10, 26, 29). It is important to note, however, that this holds true in the presence of concentrations that are ≥4× MIC, at which point maximum antibacterial effect is generally achieved (10, 26, 29). In one study, a 3-log killing was consistently maintained for 24 h with imipenem and meropenem concentrations that were 4× MIC or greater, while regrowth for any concentration above the MIC occurred only at 2× MIC with meropenem (29). This consideration of partial concentration-dependent activity may help explain why regrowth was exhibited by some study isolates with the prolonged doripenem infusion, despite the greater fT>MIC associated with the MIC of 4 μg/ml, since concentrations were only 2× MIC or less for the entire 24-h interval.

While proactive efforts are being considered for doripenem, with dose-optimizing strategies being incorporated up front into clinical trials, a prolonged infusion of a higher dose (greater than the 500 mg we studied) warrants exploration and may prove advantageous in nosocomial infections, where higher MICs will assuredly be encountered. Our data for the 4-h infusion of 500 mg doripenem every 8 h indicates a lack of antibiotic exposures for MICs of 8 and 16 μg/ml, which have been demonstrated for some, albeit infrequent, pseudomonal isolates (7, 15, 16, 27). Prolonged infusions of doripenem at 1,000 mg every 12 h over 6 h and 1,000 mg every 8 h over 4 h have been examined in phase 1 studies, and Monte Carlo simulations have analyzed target attainments for 4- to 6-h infusions of 1,000 mg every 12 h and 1- to 5-h infusions of doripenem 1,000 mg every 8 h (3, 12). Extended infusions (500 mg doripenem every 8 h over 4 h) have also been investigated in a phase 3 study for the treatment of ventilator-associated pneumonia, with similar efficacy as conventional high-dose imipenem (500 mg every 6 h or 1,000 mg every 8 h for 0.5 or 1 h) (6).

It should be noted that our simulation of human exposures was based on pharmacokinetic data obtained from healthy volunteers and not from ill patients. As such, these exposures would not apply to a patient population known to have pharmacokinetics significantly differing from those utilized in this study. Additionally, genotypic profiling to determine underlying resistance mechanisms was not performed for our study isolates. P. aeruginosa is apt at acquiring resistance, and therefore, sequential emergence of mutations can occur in the presence of an existing resistance mechanism (17). There may have been microbiological differences among the strains, which could have contributed to variability in the data, as well as the discordant results for the 4-h infusion at an MIC of 4 μg/ml.

In conclusion, doripenem is a much-needed addition to the antibacterial armamentarium, with a broad-spectrum activity that encompasses nonfermenting gram-negative bacteria, including P. aeruginosa. Doripenem displayed in vivo efficacy predictable for a time-dependent agent, and incorporating dose optimization schemes like prolonged infusion will serve to enhance its efficacy.