INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

An important mechanism of bacterial resistance to -lactam antibiotics is inactivation by existing and evolving -lactam-hydrolyzing enzymes (-lactamases) (2, 16). The evolution of extended-spectrum -lactamases (ESBLs) is associated with extensive use of -lactam antibiotics, particularly broad-spectrum cephalosporins, and represents a serious threat to reliable therapy with these antibiotics (6, 8, 15). ESBLs are plasmid-encoded enzymes frequently found in gram-negative organisms such as Escherichia coli and Klebsiella pneumoniae. ESBLs are found throughout the world in hospital, long-term care, and community settings.

MK-826 (formerly L-749,345) is a new broad-spectrum 1--methyl carbapenem antibiotic with an extended in vivo half-life (t_1/2) (Fig. 1). Its antimicrobial activity demonstrates that this antibiotic is resistant to hydrolysis by bacterial, plasmid, and chromosomally mediated non-metallo--lactamases and is able to penetrate the cell wall of clinically significant human pathogens (11). MK-826 exhibits extensive reversible protein binding, which is a basis for its extended duration of in vivo activity. Due to its improved stability to renal dehydropeptidase-I (DHP-I), MK-826 can be administered as a single agent, achieving high enough levels in urine to provide more than adequate coverage of bacteria in the urinary tract (21).

View larger version (8K): [in this window] [in a new window]   FIG. 1.   Chemical structure of MK-826.

This report compares the in vivo efficacy of this novel compound with those of other -lactam antibiotics against a number of gram-positive and gram-negative organisms in a localized (thigh) methicillin-susceptible Staphylococcus aureus (MSSA) infection as well as in systemic infection models. Also the distribution in tissue and pharmacokinetic properties of MK-826 were studied in both mouse and rat models.

(Part of this work was presented in abstract form at the 36th Interscience Conference on Antimicrobial Agents and Chemotherapy, 15 to 18 September 1996, New Orleans, La. [3, 4].)

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains and preparation. A 10-h broth culture of S. aureus MB2985 (Smith) was washed once by centrifugation (5,000 rpm for 20 min at 4°C, Sorvall RC5 rotor). The cell pellet was reconstituted in half the original volume (~10⁹ CFU/ml) and further diluted (1:5) with Trypticase soy broth for use as the challenge inoculum in the thigh infection model. Except for Streptococcus pneumoniae (MB212, CL4997, CL4983, and CL5631) and Streptococcus pyogenes (MB2874), all bacterial cultures used in the systemic infection models were grown in brain heart broth (BHB) for 10 h on a shaker (250 rpm) at 35°C. The streptococci were grown stationary, overnight (16 to 18 h), under 5% CO₂ in BHB plus 10% horse serum. Further dilutions of the cultures to prepare the infectious inocula are listed in Table 1.

Antibiotic preparation. Imipenem (IPM), meropenem (MER), cefepime (FEP), ceftriaxone (CTRX), ceftazidime (CAZ), cefazolin (CEF), cefonicid (CID), cefotaxime (CTX), and penicillin G (PEN) were either commercially available or were synthesized at Merck Research Laboratories. MK-826 (ZD4433) was made by Merck or Zeneca Pharmaceutical Co. Ltd. Antibiotic stock solutions were prepared in 10 mM morpholinopropanesulfonic acid (MOPS) buffer, pH 7.1, and stored at 70°C. For all carbapenems, unless otherwise stated, the MOPS buffer described above also contained cilastatin at a final concentration of 2.0 mg/ml. Injection of 0.5 ml of this solution yields a final concentration of cilastatin of 40 mg/kg of body weight/mouse for each antibiotic treatment regardless of antibiotic concentration. In one experiment, the efficacy of MER was evaluated in the presence and absence of cilastatin. In the thigh model, antibiotic test concentrations generally ranged from 0.5 to 10 mg/kg/dose. In the systemic models, frozen stock solutions were diluted to yield a series of fourfold antibiotic concentrations. In the pharmacokinetic studies, 10-mg/kg solutions of MK-826-cilastatin and CTRX were prepared in 10 mM MOPS buffer.

Animals. DBA/2 female mice (Taconic Laboratories, Germantown, N.Y.) weighing 25 ± 1 g were used in the thigh infection model. DBA/2 female mice (19 to 21 g; Taconic) and viral antibody-free CD-1 female mice (Charles River Laboratories, Wilmington, Mass.) weighing 19 to 21 g were used in the systemic infection models. CD-1 mice (19 to 21 g) and female Sprague-Dawley rats (Sasco, St. Louis, Mo.) weighing 235 to 305 g were used in the pharmacokinetic assays.

Localized soft tissue infection model. On day 0, 0.2 ml of the challenge inoculum was injected intramuscularly into the right thigh of each mouse. Antibiotic therapy was initiated 2 h after challenge (0 h), i.e., after the first of eight antibiotic treatments, was administered subcutaneously (s.c.) (0.5 ml) into the upper back. Subsequent antibiotic dosing was performed at 6, 10, 24, 48, 72, 96, and 120 h postinfection for a total of eight doses over a 5-day period. Two days after the final antibiotic dose (day 7) mice were euthanatized, the lower abdominal areas and thighs were flushed with 70% ethanol, and the outer skin was removed. Denuded thighs were then surgically removed aseptically and placed in sterile tubes containing 4 ml of sterile phosphate-buffered saline-10% glycerol for temporary storage at 70°C or for same-day processing. Usually, plating for bacterial counts was done at a later date. On the day of plating, the tubes were thawed in cold water and thighs were ground with a Polytron homogenizer (Brinkmann Instruments) for ~5 to 10 s each and placed on ice to await further dilution. Tenfold serial dilutions of each sample were made in tubes containing 0.9 ml of cold TSB prior to plating 0.1 ml from selected dilutions onto staphylococcus-selective mannitol salt agar plates. Plates were incubated for 2 days at 35°C to determine the number of CFU remaining per thigh. The geometric mean of the CFU remaining and the log₁₀ change in CFU were also determined for each group.

Systemic infection models. Mice were challenged intraperitoneally (i.p.) with the inocula described above in 0.5 ml. All antibiotics were prepared from frozen stock cultures to yield a series of fourfold antibiotic concentrations, which were maintained at 4°C for dosing (0.5 ml; s.c.) of five mice per group. Mice were monitored daily for morbidity and mortality over the 7-day test period.

Mouse tissue drug distribution and pharmacokinetic studies. At time zero, groups of 50 CD-1 female mice were injected (0.5 ml; i.p.) with a 10-mg single dose of either MK-826 or CTRX. Cilastatin was coadministered with MK-826 to improve its -terminal elimination plasma t_1/2 (t_1/2) and urinary recovery due to extra renal DHP-I metabolism of carbapenems encountered in rodent species (12). It may be important that in mice, the half-life of IPM in plasma was increased approximately twofold when administered with cilastatin and that of MK-826 was increased nearly threefold (16 to 44 min). Mice were given a 0.5-ml oral dose of water to stimulate urine flow just prior to placement in metabolism cages (five per cage) designed to collect urine free from fecal contamination. At 1, 5, 15, 30, 60, 120, 180, 240, 360, and 1,440 min post-antibiotic treatment, pooled blood samples from each group were allowed to clot, and serum was collected at 4°C, following centrifugation at 10,000 rpm for 10 min (Beckman Microfuge 12 rotor). The volume and pH of urine samples collected during the same time intervals also were recorded. Following blood collection, the left kidney, liver, lung, spleen, brain, heart, and right thigh muscle were removed from individual mice, weighed, and pooled by test groups and tissue types prior to homogenization in four volumes (wt/vol) of buffer to a concentration of ~200 mg of tissue/ml. CTRX samples were homogenized in Sorensen's buffer, pH 7.1, while MK-826 samples were prepared in 25 mM 2-(N-morpholino)ethanesulfonic acid (MES) buffer (pH 5.5)-3% MeOH containing cilastatin (50 µg/ml). Urine samples were stored at 70°C until bioassayed, while serum and other tissue samples were placed on ice for immediate assay.

Determination of antibiotic content in fluids and tissues. A direct-injection, column-switching, reversed-phase high-performance liquid chromatography technique was used to assay serum and other tissue samples. Stabilized samples were injected onto a short pellicular C₁₈ column and washed with 0.15% trifluoroacetic acid (clean-up mobile phase). This clean-up column was then switched in line with an analytical C₁₈ column (4.6 by 250 mm; 40°C; Zorbax SB-C18) and the chromatogram was developed at 1 ml/min with an acetonitrile-methanol gradient in 0.1% trifluoroacetic acid. MK-826 and CTRX were detected at 307 and 254 nm, respectively, with a 165 Variable Wavelength Detector (Beckman Instruments, Palo Alto, Calif.). The limit of detection was ~0.39 µg/ml of serum for both MK-826 and CTRX and 0.25 to 0.63 and 0.49 to 0.98 µg/g of tissue for MK-826 and CTRX, respectively.

Rat plasma, bile, and urine antibiotic concentration studies. Female Sprague-Dawley rats were anesthetized i.p. with ~50 mg of pentobarbitol sodium (Nembutol; Abbott, Chicago, Ill.) per kg. The surgical areas of each rat were shaved and swabbed with providone-iodine-alcohol. A small horizontal incision (3 to 4 cm) was made below the xiphoid process. The bile duct was exposed and a cannula of Silastic tubing was threaded into the duct and anchored with ligatures. The abdominal incision was closed with stainless steel wound clips. To facilitate blood collection, an incision was made in the right hind leg of the rat, exposing the femoral artery. A sterile PE 10 catheter was fed into the abdominal aorta and secured in place with silk ligatures. Once normal bile flow was established, each rat was given either an s.c. dose of cilastatin (40 mg/kg) followed by an intravenous (i.v.) injection of MK-826 or an i.v. injection of CTRX in the femoral artery at a dose equivalent to 10 mg/kg. Heparinized blood samples were taken at 5, 30, 60, 120, 240, 360, and 480 min postdose. The plasma was separated by centrifugation and stabilized with an equal volume of ethylene glycol-MES buffer mixture and frozen at 70°C until bioassayed. Bile samples were collected continuously at 0 to 30, 30 to 120, 120 to 360, and 360 to 480 min post-antibiotic treatment in tubes suspended in an ice-water bath. Each sample was stabilized with an equal volume of ethylene glycol-MOPS buffer and frozen at 70°C. At the termination of the study, residual urine was removed from the bladder.

0-

1/2

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Localized infection model. In the S. aureus thigh tissue infection model MK-826, IPM, FEP, and CTRX were all efficacious at 10 mg/kg, with >3 log₁₀ CFU reduction of organism compared to non-antibiotic-treated controls. However, at a fivefold-lower concentration (2 mg/kg), only MK-826 and IPM, with 3.3 and 4.4 log₁₀ CFU eliminated, respectively, maintained this activity (Fig. 2). MER, while not as active as MK-826 or IPM, did exhibit some activity at all levels tested. Interestingly, >4 logs of bacterial clearance was observed with 5 mg of MER per kg in the presence of cilastatin while a <1 log₁₀ CFU decrease was observed with the same dose of MER in the absence of cilastatin (Fig. 3). CAZ and CTX were not efficacious in this MSSA localized infection model.

View larger version (60K): [in this window] [in a new window]   FIG. 2.   Efficacy of MK-826 and other agents in the mouse thigh model: experiment 1. DBA/2 mice were challenged (0.2 ml; i.p.) with S. aureus MB2985 in the right thigh at 0 h. Antibiotic treatment (0.5 ml; s.c.) was given at 2, 6, 10, 24, 48, 72, 96, and 120 h after challenge. Circles show individual values for mouse 1 to 5; bars show geometric means.

View larger version (56K): [in this window] [in a new window]   FIG. 3.   Efficacy of MK-826 and other agents in the mouse thigh model: experiment 2. DBA/2 mice were challenged (0.2 ml; i.p.) with S. aureus MB2985 in the right thigh at 0 h. Antibiotic treatment (0.5 ml; s.c.) was given at 2, 6, 10, 24, 48, 72, 96, and 120 h after challenge. Cilastatin was used except where noted. Circles show individual values for mouse 1 to 5; bars show geometric means.

Systemic model. MK-826 exhibited good activity against all gram-positive organisms tested in the systemic infection model, including highly PEN-resistant pneumococci (Table 2). While not as potent as IPM or CEF against S. aureus (MSSA), MK-826 was two- to fourfold more active than FEP, CID, and CTRX. All four compounds tested against a PEN-susceptible strain of S. pneumoniae were equally efficacious, while MK-826 and CTRX were slightly more active than IPM and PEN against the moderately susceptible strain. IPM was the most active agent against the two PEN-resistant S. pneumoniae strains tested, while MK-826 and CTRX were equipotent against these strains. MK-826 was approximately twofold more active than FEP and CTRX against S. pyogenes. MK-826 was also very active against the gram-negative organisms tested with ED₅₀s of <0.25 mg/kg/dose for Enterobacter cloacae, E. coli, K. pneumoniae, M. morganii, Proteus mirabilis, and Serratia marcescens (Table 3). MK-826 exhibited good activity against two Pseudomonas aeruginosa isolates tested, with ED₅₀s of 1.2 and 3.0 mg/kg/dose, respectively.

Mouse pharmacokinetic studies. The serum and urine parameters of MK-826 (with cilastatin) and CTRX following i.p. administration of a 10-mg/kg dose in CD-1 mice were similar, with peak concentrations in serum of 62.8 µg/ml at 30 min for MK-826 and 42.6 µg/ml at 15 min for CTRX. The AUC_0- was 150.8 µg · hr/ml for MK-826, compared to 90.0 µg · hr/ml for CTRX. The t_1/2s in serum were also similar for MK-826 (1.27 h) and CTRX (1.22 h). The Clp for MK-826 was 1.11 ml/min/kg, while a slightly higher rate was seen for CTRX (1.85 ml/min/kg). The volume of distribution for CTRX, 170 ml/kg, was slightly higher than that of MK-826 (107 ml/kg). Urinary recoveries over a 24-h period were similar, at 46% for MK-826 and 49% for CTRX.

Rat pharmacokinetic studies. Results of the pharmacokinetic distribution of MK-826 (with cilastatin) and CTRX in rats are shown in Table 5. MK-826 was cleared very slowly from the plasma of all three rats, with an average concentration of 15.10 µg/ml remaining 8 h posttreatment. The t_1/2 averaged 3.2 h. The Clp averaged 0.47 ml/min/kg. The biliary recoveries in two of three rats were the same, at 12%, compared with only 6% in the third rat. This may have been due to cannulation of the bile duct above the bifurcation, resulting in collection of only half the available bile. Difficulties in establishing normal bile flow in one rat resulted in CTRX analysis in only two rats. Five minutes following i.v. administration of CTRX, an average concentration of 175 µg/ml was observed. At 8 h postdose measurable levels of CTRX in plasma were considerably lower than those seen with MK-826. In addition, the average plasma t_1/2 was 2.3 h, compared to 3.2 h for MK-826. As expected, >55% of CTRX was recovered in the bile over the 8-h test period. Urinary recoveries of MK-826 and CTRX were similar, with 18 and 16%, respectively, recovered in 8 h.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Multidrug-resistant bacteria, especially those harboring ESBLs, are of increasing clinical concern (1). These emerging nosocomial pathogens leave clinicians with few treatment options. A once-a-day antibiotic with broad-spectrum gram-positive and gram-negative activity, including activity against organisms with ESBLs, could offer a solution to this problem. MK-826 is a long-acting carbapenem antibiotic which possesses such a broad antibacterial spectrum (5, 7, 10).

In a thigh tissue infection model, MK-826 was highly efficacious, at a level as low as 2 mg/kg, in organism clearance. Neither the cephalosporins nor MER was as active at this level. CTRX and FEP were equally efficacious as MK-826 at 10 mg/kg, but their activities were significantly reduced at lower concentrations. MK-826 also exhibited potent activity in the systemic models of infection against a variety of both gram-positive and gram-negative organisms. Except for one PEN-resistant S. pneumoniae isolate (MCL4997), IPM was the most active agent against all the gram-positive isolates tested. MK-826 in most cases was either equipotent or more efficacious than either CTRX or FEP against these organisms. MK-826 also was highly efficacious against the gram-negative organisms in this infection model. With the exception of P. aeruginosa the ED₅₀s of MK-826 obtained for these organisms were all <0.25 mg/kg/dose. MK-826, while not as efficacious as IPM, did exhibit some activity against two P. aeruginosa isolates tested. These in vivo outcomes accurately confirm the in vitro susceptibility endpoints. In vitro studies have found MK-826 to be very potent against other bacterial species (3, 18, 19). The extensive protein binding of MK-826, while a major factor in its long t_1/2, is not deleterious to its in vivo efficacy against organisms for which the MICs are low. As Craig et al. (1) have shown, the most important factor influencing the efficacy of -lactams, including carbapenems, is the amount of time when drug concentrations are above the MIC. Carbapenems were found to exert maximal cell killing when levels in serum were above the MIC for at least 40% of the dosing interval. Thus, the lower the MIC is, the more negligible the effect of protein binding is, since less free drug would be required to eradicate the organism and more would be available, extending the time when the concentration is above the MIC.

Pharmacokinetically the serum and urine parameters of MK-826 (with cilastatin) and CTRX in mice were similar (4). A slightly higher peak concentration in serum and a larger AUC was observed for MK-826 compared to that of CTRX. Urinary recovery was the same for both compounds, and distribution in tissue was effective and sustained. In rats, the average AUC for MK-826 (with cilastatin) was greater than that of CTRX and the calculated t_1/2 of MK-826 was approximately twice as long as that for CTRX. Urinary recovery again was similar for both antibiotics. Like CTRX, MK-826 was eliminated via the hepato-biliary route; however, the total amount of drug excreted over 8 h (55 to 64% versus 6 to 12%) was greater for CTRX than for MK-826. These levels for CTRX were slightly higher than the 40 to 50% obtained in humans by other investigators (17, 20). Similar comparative results of MK-826 and CTRX were obtained in nonhuman primates (22). Presently, results from phase-I clinical trials indicate a mean t_1/2 of 4.5 h and an average 48-h urinary excretion of intact MK-826 of 30 to 45% (13).

In summary, MK-826 is a new, long-acting, 1--methyl carbapenem presently in phase-II clinical trials. It exhibits potent activity which is comparable or superior to established agents against both gram-positive and gram-negative organisms in systemic and tissue infection models of disease. Importantly, MK-826 exhibits a broad antibacterial spectrum that includes activities against organisms which harbor ESBLs. Advantageous pharmacokinetics, including an extended t_1/2 and improved stability to renal DHP-I, support the argument for further development of MK-826, perhaps as a single once-a-day dosing agent.