Imipenem and meropenem: Comparison of in vitro activity, pharmacokinetics, clinical trials and adverse effects

Abstract

OBJECTIVE:

To compare and contrast imipenem and meropenem in terms of in vitro activity, pharmacokinetics, clinical efficacy and adverse effects.

DATA SELECTION:

MEDLINE search from 1975 to 1997 and follow-up of references.

DATA EXTRACTION:

Clinical trials comparing imipenem with meropenem, or either imipenem or meropenem with standard therapy in the treatment of serious infections were selected.

DATA SYNTHESIS:

Imipenem, the first carbapenem, was first marketed in 1987; meropenem was introduced to the market in 1996. In general, imipenem is more active against Gram-positive cocci while meropenem is more active against Gram-negative bacilli. The agents display similar pharmacokinetics. Clinical studies in patients with serious infections (intra-abdominal infection, respiratory infection, septicemia, febrile neutropenia) report similar bacteriological and clinical cure rates with imipenem and meropenem. Meropenem is approved for the treatment of bacterial meningitis, whereas imipenem is not. Adverse effects are similar.

CONCLUSIONS:

Current literature supports the use of imipenem at a dose of 500 mg every 6 h and meropenem at 1 g every 8 h for the treatment of severe infections. For the treatment of serious infections, imipenem (500 mg every 6 h or 2 g/day [$98/day]) is more economical than meropenem (1 g every 8 h or 3 g/day [$142/day]) based on acquisition cost.

Imipenem was the first of a new class of antibiotics, the carbapenems (1). The parent substance, thienamycin, was isolated from the soil organism Streptomyces cattleya (2). However, because thienamycin spontaneously breaks down at high concentrations, the more stable amidine derivative N-formimidoyl thienamycin, known as imipenem, was developed (Figure 1). Imipenem is administered in combination with an equal amount of cilastatin (Primaxin, Merck Sharp & Dohme), a compound that was specifically developed to inhibit metabolism of imipenem by the kidney. Imipenem, first marketed in Canada in 1987, immediately attracted attention because it possessed the most broad antibacterial spectrum of any antibiotic available at that time (3,4). Imipenem is indicated for the treatment of a variety of serious infections, but not for meningitis.

Figure 1.

Chemical structures of meropenem and imipenem

Meropenem (Merrem, Zeneca), first marketed in Canada in 1996, was the second semisynthetic parenteral carbapenem available in Canada. Meropenem differs structurally from imipenem, primarily by the addition of a methyl group in the 1-position of the carbapenem moiety (5) (Figure 1). This results in greater stability in vivo to inactivation by human renal dehydropeptidase-1 (DHP-1) than imipenem, with the result that meropenem need not be co-administered with cilastatin (6,7). Meropenem is indicated for the treatment of a variety of serious infections and meningitis.

The purpose of this paper is to compare and contrast the in vitro activity, pharmacokinetics, clinical uses and adverse effects of imipenem and meropenem. The focus of this paper is the comparison of imipenem and meropenem in the treatment of serious infections (intra-abdominal infections, respiratory tract infections, septicemia, bacterial meningitis and febrile neutropenia). For simplicity, imipenem and imipenem/cilastatin are used interchangeably.

MECHANISM OF ACTION

Meropenem and imipenem are bactericidal against susceptible organisms as demonstrated by time-kill curve studies with Enterobacteriaceae (8–10). Both agents cause bacterial lysis in susceptible organisms by binding with high affinity to high molecular weight penicillin-binding proteins (PBPs). It is generally agreed that PBP 2 is the primary target of both meropenem and imipenem in Escherichia coli (2,11,12). Additionally, with Gram-negative bacilli such as E coli and Pseudomonas aeruginosa, meropenem demonstrates a high affinity for PBP 3 not found with imipenem, which may account for its enhanced activity against Gram-negative bacilli compared with imipenem (11). The relative PBP affinities of meropenem and imipenem in strains of Staphylococcus aureus are quite similar, each demonstrating high affinity for PBP 2, PBP 1 and PBP 4 (12,13).

Both meropenem and imipenem are highly resistant to hydrolysis by most clinically important beta-lactamases, plasmid- or chromosomally mediated, of S aureus, E coli, Enterobacter species, Citrobacter freundii, Proteus species, Serratia marcescens, Klebsiella species, P aeruginosa and Bacteroides fragilis (2,3,8,14–17).

In addition, meropenem and imipenem are unaffected by strains of Enterobacteriaceae that produce plasmid-mediated beta-lactamases derived from TEM and SHV enzymes which are capable of hydrolyzing third-generation cephalosporins, ie, extended spectrum beta-lactamases (8,18,19). Both meropenem and imipenem are readily hydrolyzed by carbapenem-hydrolyzing beta-lactamases produced by Stenotrophomonas maltophilia and occasionally in strains of Bacteroides species, Bacillus cereus and Aeromonas hydrophila (1,7,20,21).

Unlike most beta-lactams (penicillins, cephalosporins, monobactams), carbapenems demonstrate a long, dose-dependent postantibiotic effect (PAE) against Gram-negative organisms (22,23). This effect makes carbapenems more similar to fluoroquinolones and aminoglycosides than beta-lactams. PAEs for meropenem and imipenem range from 2 to 9 h depending on the organism and concentration of antimicrobial studied (22–24). In addition, limited data suggest that carbapenems, unlike penicillin and cephalosporins, may demonstrate an element of dose-dependent bacterial killing (22–24). These data suggest that optimal dosing of these agents may be obtained if they are administered less frequently (longer dosing interval) but at higher doses. Higher dosages increase bacterial killing while longer dosing intervals are possible due to the long PAE.

MECHANISM OF RESISTANCE

Resistance to beta-lactams including carbapenems arises in one or a combination of three ways: beta-lactam hydrolysis by beta-lactamases, reduced permeability through the outer membrane (Gram-negatives only) and PBPs (6–8).

High level expression of carbapenem hydrolyzing beta-lactamases in S maltophilia and occasionally in strains of Bacteroides species, P aeruginosa and Serratia marcescens is associated with resistance to both meropenem and imipenem (14,20,21,25). These metallo-beta-lactamases require zinc for maximal activity and exhibit a broad hydrolytic profile including carbapenems, penicillins and cephalosporins (21). Carbapenem-hydrolyzing beta-lactamases that preferentially hydrolyze carbapenems (‘true carbapenemases’), are found in A hydrophila and, occasionally, in strains of Burkholderia cepacia (21).

The majority of published data describing carbapenem resistance concerns P aeruginosa (26–28). Penicillins and cephalosporins, due to their low molecular weight and zwitterionic nature, readily penetrate the outer membrane of Gram-negative bacilli through porin proteins (29). The major porins involved are outer membrane proteins (OMPs) F and C. Carbapenems, however, use an unconventional route of entry, OMP D₂ (27,30). Because carbapenems use a different entry mechanism and they are not very susceptible to hydrolysis by penicillinases and cephalosporinases, cross-resistance between carbapenems, and penicillins and cephalosporins does not occur (unless due to a carbapenem-hydrolyzing enzyme) (29). High level imipenem resistance in P aeruginosa appears to be due to a combination of decreased uptake due to reduction or lack of OMP D₂, along with concomitant slow hydrolysis by beta-lactamases (26). Although the majority of data suggest cross-resistance between meropenem and imipenem, Gram-negative bacillary resistance due to reduced permeability is less of a concern with meropenem because meropenem is transported more rapidly through OMP D₂ (7). Finally, in a few species such as Enterococcus faecium (which uses PBP 5 and PBP 6 to make peptidoglycan) and methicillin-resistant S aureus (MRSA) (which uses PBP 2a to make peptidoglycan), carbapenems do not readily bind to these PBPs, leading to resistance (29).

IN VITRO ACTIVITY

Susceptibility results (minimum inhibitory concentration of 90% of isolates [MIC₉₀]) of meropenem and imipenem in comparison with common alternative antimicrobials against Gram-positive and Gram-negative aerobes, and anaerobes are listed in Tables 1 to 3 (1,3–11,14,16–18,21,30–50). The values reflect the mean MIC₉₀s using standard susceptibility techniques for clinical isolates obtained worldwide.

TABLE 1.

Antibacterial activity (minimum inhibitory concentration of 90% of isolates μg/mL) of imipenem and meropenem in comparison with ceftazidime, piperacillin/tazobactam, gentamicin and ciprofloxacin against Gram-positive aerobes

Gram-positive aerobes	Imipenem	Meropenem	Ceftazidime	Piperacillin/tazobactam	Gentamicin	Ciprofloxacin
Staphylococcus aureus (MS)	0.06	0.25	>16	1	16	0.5
S aureus (MR)	8	16	64	128	16	2
Staphylococcus epidermidis (MS)	0.25	0.5	16	8	16	16
S epidermidis (MR)	16	16	64	128	16	16
Staphylococcus saprophyticus	0.03	0.25	32	0.25	0.06	8
Streptococcus pyogenes	<0.06	<0.06	0.12	0.25	16	1
Streptococcus agalactiae	0.03	0.1	0.12	0.25	16	1
Streptococcus pneumoniae (PS)	0.03	0.06	0.25	0.06	16	2
S pneumoniae (PR)	0.25	1	32	2	16	2
Enterococcus faecalis	2	8	>128	4	128	4
Enterococcus faecium	32	64	>128	128	128	16
Listeria monocytogenes	0.25	0.25	128	2	4	1

Adapted from references 1,3–8,10,11,16,17,30–40. In vitro susceptibility of imipenem and meropenem: susceptible 4 μg/mL or less, intermediate 8 μg/mL, resistant 16 μg/mL or more. MR Methicillin-resistant; MS Methicillin-susceptible; PR Penicillin-resistant; PS Penicillin-susceptible

TABLE 3.

Antibacterial activity (minimum inhibitory concentration of 90% of isolates μg/mL) of imipenem and meropenem in comparison with cefoxitin, clindamycin and metronidazole against anaerobes

Anaerobes	Imipenem	Meropenem	Cefoxitin	Clindamycin	Metronidazole
Bacteroides fragilis	0.5	0.25	16	8	2
B fragilis group	1	0.5	32	16	2
Clostridium difficile	8	2	128	32	0.5
Clostridium perfringens	0.5	0.12	0.5	2	1
Fusobacterium species	0.5	0.5	2	2	2
Peptostreptococcus species	0.25	0.5	2	2	32

Adapted from references 1,3,5–8,21,34,36,48–50. In vitro susceptibility of imipenem and meropenem: susceptible 4 μg/mL or less, intermediate 8 μmg/mL, resistant 16 μg/mL or more

Against Gram-positive aerobes, imipenem is in general two- to fourfold more active than meropenem (Table 1) (1,3–11,16,17,30–40). Neither agent is active against MRSA or methicillin-resistant Staphlyococcus epidermidis. Imipenem is more active than meropenem against streptococcal species including Streptococcus pneumoniae. Penicillin-resistant S pneumoniae are less sensitive to both imipenem and meropenem than penicillin-susceptible S pneumoniae. Nevertheless, both imipenem and meropenem are active against penicillin-resistant S pneumoniae, with imipenem being fourfold more active than meropenem. Imipenem is fourfold more active than meropenem versus Enterococcus faecalis, and neither agent is active against E faecium.

Generally, meropenem is two- to 16-fold more active than imipenem against Gram-negative aerobes (Table 2) (1,3–11, 14,16–18,31–47). Against Enterobacteriaceae, meropenem is four- to 16-fold more active than imipenem. Meropenem is approximately eightfold more active than imipenem against Haemophilus influenzae and Neisseria gonorrhoeae. B cepacia displays intermediate resistance, and S maltophilia is resistant to both agents.

TABLE 2.

Antibacterial activity (minimum inhibitory concentration of 90% of isolates μg/mL) of imipenem and meropenem in comparison with ceftazidime, piperacillin/tazobactam, gentamicin and ciprofloxacin against Gram-negative aerobes

Gram-negative aerobes	Imipenem	Meropenem	Ceftazidime	Piperacillin/tazobactam	Gentamicin	Ciprofloxacin
Acinetobacter anitratus	0.25	1.0	8	16	1	8
Citrobacter freundii	1	0.13	32	16	1	0.5
Enterobacter aerogenes	1	0.13	16	32	1	0.25
Enterobacter cloacae	1	0.25	16	32	8	0.25
Escherichia coli	0.13	0.03	1	1	8	0.13
Haemophilus influenzae (BLN)	0.5	0.06	0.06	0.13	8	0.016
H influenzae (BLP)	0.5	0.06	0.06	0.25	8	0.016
Klebsiella pneumoniae	0.25	0.03	0.25	4	4	0.25
Klebsiella species	0.5	0.06	0.25	2	4	0.25
Moraxella catarrhalis	0.06	0.008	0.5	2	2	0.06
Morganella morganii	4	0.25	16	4	4	0.13
Neisseria gonorrhoeae (PS, PR)	0.25	0.03	0.03	1	16	0.008
Neisseria meningitidis	0.03	0.016	0.25	0.25	8	0.008
Proteus mirabilis	2	0.13	0.13	0.5	4	0.13
Proteus vulgaris	4	0.25	0.25	2	4	0.06
Proteus rettgeri	1	0.12	4	4	32	8
Providencia stuartii	2	0.25	4	4	16	8
Pseudomonas aeruginosa	4	2	8	8	16	2
Burkholderia cepacia	8	8	16	128	128	8
Salmonella species	0.12	0.03	0.5	2	0.5	0.06
Serratia marcescens	2	0.25	4	2	16	2
Shigella species	0.25	0.06	0.5	4	1	0.06
Stenotrophomonas maltophilia	128	128	16	256	64	16
Yersinia enterocolitica	0.25	0.03	0.5	2	2	0.13

Adapted from references 1,3–11,14,16–18,31–47. In vitro susceptibility of imipenem and meropenem: susceptible 4 μg/mL or less, intermediate 8 μg/mL, resistant 16 μg/mL or more. BLN Beta-lactamase negative; BLP Beta-lactamase positive; PR Penicillin-resistant; PS Penicillin-susceptible

Meropenem and imipenem are active against most strains of clinically significant anaerobes (Table 3) (1,3,5–8,21, 34,36,48–50). Both agents demonstrate similar activity against B fragilis and Bacteroides group organisms. In addition, they demonstrate similar activity against Gram-positive anaerobes.

Both imipenem and meropenem are more active than ceftazidime, gentamicin or ciprofloxacin, and similarly active to piperacillin/tazobactam against Gram-positive aerobes. All of the antimicrobials listed in Table 2 have potent activity against Gram-negative aerobes. Imipenem and meropenem have very good anaerobic activity along with metronidazole, clindamycin, cefoxitin and piperacillin/tazobactam.

PHARMACOKINETICS

Pharmacokinetic profiles of imipenem and meropenem in healthy volunteers are shown in Table 4 (51–64). Imipenem and meropenem are not absorbed orally; however, after intravenous administration, both antimicrobials achieve peak serum concentrations far in excess of reported MICs for most Gram-positive, Gram-negative and anaerobic organisms (Tables 1–3) (51–54).

TABLE 4.

Pharmacokinetic comparison of imipenem and meropenem

Parameter	Imipenem	Meropenem
Orally absorbed	No	No
Vdss	0.25 L/kg	0.23 to 0.35 L/kg
Cmax (500 mg infusion, adult)	33 μg/mL	26 μg/mL
Cmax (1 g dose infusion, adult)	52 μg/mL	49 μg/mL
C (6 h after 1 g dose, adult)	1 μg/mL	–
C (8 h after 1 g dose, adult)	–	0.25 μg/mL
Half-life (normal renal function, adult)	1 h	1 h
Cmax (20 mg/kg, children younger age 12 years)	–	62μg/mL
Half-life (normal renal function, children younger than 12 years)	–	1 h
Plasma protein binding	20%	2%
Renal elimination as parent	60% to 70% (with cilastatin)	70%
Active metabolite	No	No

Adapted from references 51–64. Vd_ss Volume of distribution at steady state; C Serum concentration; C_max Maximum concentration in plasma

Like imipenem, meropenem distributes well into most bodily fluids (52,55–57). Hextall et al (58) reported that the intraperitoneal penetration of meropenem was 95% of the corresponding area under the curve (AUC) in plasma 2 h after a single intravenous infusion (58). In patients with meningitis, imipenem cerebrospinal fluid (CSF) concentrations of 0.5 to 11 μg/mL have been reported upon repeated dosing of 1 g every 6 h (59). Meropenem also appears to penetrate well into the CSF in the presence of inflamed meninges. Dagan et al (57) conducted a trial in 23 patients with bacterial or viral meningitis already receiving antimicrobial treatment. Patients receiving a single intravenous infusion of meropenem 20 mg/kg or 40 mg/kg achieved CSF concentrations of 0.1 to 2.8 μg/mL and 0.3 to 6.5 μg/mL, respectively. There was significant interpatient variability highlighted by CSF to plasma concentration ratios from 1% to 42% and from 2% to 52%, respectively (57). It should be mentioned that because CSF to plasma concentration ratios can show considerable variation, AUC comparisons between CSF and plasma are more likely to represent the true degree of penetration of an antibiotic into CSF.

Imipenem is hydrolyzed by renal dehydropeptidase-1 enzymes located on the brush border of the proximal renal tubules. Administration with cilastatin prevents imipenem destruction by dehydropeptidase and potential nephrotoxicity that occurs due to the metabolites. Sixty to seventy per cent of imipenem is excreted unchanged in the urine in the presence of cilastatin. Approximately 70% of meropenem is renally eliminated as the parent compound.

Because of extensive renal elimination, doses of both imipenem and meropenem must be adjusted in renal dysfunction. Imipenem’s half-life increases to 4 h and meropenem’s to 7 h in patients with creatinine clearances less than 10 mL/min (Table 4). Both imipenem and meropenem are removed by dialysis, therefore supplemental doses (regimen based on infection and severity) should be administered after dialysis (60,54). Tables 5 and 6 summarize the recommended dose changes for imipenem and meropenem, respectively, in patients with renal dysfunction. Neither antimicrobial undergoes appreciable hepatic metabolism, therefore no dose adjustment is necessary in patients with hepatic dysfunction.

TABLE 5.

Imipenem: Dose adjustment in renal dysfunction^*

Creatine clearance (mL/min/1.73 m2)	Dose	Dosing interval
31 to 70	500 mg	Every 6 to 8 h
21 to 30	500 mg	Every 8 to 12 h
0 to 20	250 to 500 mg	Every 12 h

^*Manufacturer’s recommendations: Imipenem/Cilastatin product monograph. Rahway: Merck Sharpe and Dohme, 1985

TABLE 6.

Meropenem: Dose adjustment in renal dysfunction^*

Creatine clearance (mL/min/1.73 m2)	Dose	Dosing interval
51 or greater	500 mg to 2 g (ie, recommended dose)	Every 8 h
26 to 50	Recommended dose	Every 12 h
10 to 25	1/2 recommended dose	Every 12 h
Less than 10	1/2 recommended dose	Every 24 h

^*Manufacturer’s recommendations: Meropenem product monograph. Mississauga: Zeneca Pharma Inc, 1996

Pharmacokinetic studies conducted in other disease states, including surgical patients with moderate or severe infections and patients with intra-abdominal infections, report no clinically significant changes in pharmacokinetic parameters and do not require specific dose adjustments (55,56).

Figure 2 demonstrates the average plasma concentration-time profiles for various steady-state dosing regimens of impenem and meropenem. The enhanced activity of meropenem against Enterobacteriaceae species allows for an extension in the dosing interval to every 8 h compared with every 6 h for imipenem. Whether the increased activity of imipenem over meropenem against Gram-positive cocci allows the imipenem dosing interval to be extended to every 8 h is unclear. For the average patient, the dosing regimen for meropenem of 1 g every 8 h provides for trough concentrations of approximately 0.25 μg/mL (Figure 2), while 500 mg every 8 h provides for trough concentrations around 0.13 μg/mL and a lower peak concentration. Whether the slightly lower peak and trough concentrations (with 500 mg every 8 h), and 1 h less time above the MIC than 1 g every 8 h result in clinically significant differences in bacteriological or clinical outcome is unknown. It would, however, appear that 500 mg every 6 h of meropenem would provide trough serum concentrations higher than when using 500 mg every 8 h of meropenem and similar to those using 500 mg every 6 h of imipenem.

Figure 2.

Plasma concentration-time profile comparisons for steadystate dosing regimens of imipenem and meropenem. IV Intravenous; q Every

CLINICAL TRIALS

Many comparative trials have been published using imipenem or meropenem for the treatment of serious infections in hospitalized patients. This review focuses on the comparison of imipenem with meropenem, and the comparison of either agent with standard therapy for potentially life-threatening infections including intra-abdominal infections, respiratory infections, septicemia, bacterial meningitis and febrile neutropenia. Imipenem and meropenem are most likely to be used in the treatment of serious infections. Only prospective, randomized trials published in peer reviewed journals were considered, and are summarized in Tables 7 to 9. Abstracts presented before 1994 without subsequent publication were not included.

TABLE 7.

Imipenem verus selected antibiotics for treatment of serious infections in hospitalized patients

Study (reference)	Design	Patients (clinically evaluable patients)	Indication	Regimen	Duration*	Results
Solomkin et al (75)	Prospective, randomized, open	290 (162)	Intra-abdominal infections	I: 500 mg every 6 h Clinda: 600 mg every 6 h and Tobra: 1.5 mg/kg (adjusted to peak ≥6 μg/mL, trough 1 to 2 μg/mL)	NR	I: success†: 67/81 (83%) Clinda/Tobra: success:57/81 (70%) P=0.043
Poenaru et al (73)	Prospective, randomized, open	104 (104)	Intra-abdominal infections (APACHE scores approximately 11 to 13)	I: 500 mg every 6 h Tobra: 1.5 mg/kg every 8 h (peak 6 to 10 μg/mL, trough not less than 1.5 μg/mL) and either Clinda: 600 mg every 6 h (n=32) or Metro: 500 mg every 6 h (n=20)	NR	I: success‡: 41/52 (79%) Tobra +/− Clinda or Metro: success‡: 35/52 (67%)
Niinikoski et al (74)	Prospective, randomized, open	86 (56)	Intra-abdominal infection (primarily peritonitis)	I: 1 g every 8 h Pip/Tazo: 4 g/0.5 g every 8 h	14 days 14 days	I: 20/26 (77%)§ Pip/Tazo: 26/29 (89%)§ P=0.37
Fink et al (84)	Prospective, randomized, double-blind	402 (205)	Severe pneumonia (primarily intensive care unit)	I: 1 g every 8 h (500 mg every 6 h for highly susceptible pathogens) Cipro: 400 mg IV every 8 h (400 mg every 12 h for highly susceptible pathogens)	10.1 days 10.5 days	Efficacy evaluable¶ I: 58/104 (56%) Cipro: 68/98 (69%) P=0.021 Intent to treat I: 90/162 (56%) Cipro: 92/144 (64%) P=0.123
Norrby et al (87)	Prospective, randomized, single-blind (septicemia subgroup)	91 (66)	Septicemia	I: 500 mg every 6 h Ceft: 2 g every 12 h	9.7 days 8.8 days	I: 26/32 (81%)** Ceft: 25/34 (74%)** (not statistically significant, no P reported)
Leyland et al (100)	Prospective, randomized, single-blind	234 (252 evaluable febrile episodes)	Febrile neutropenia	I: median dose 3.5 g/day in four divided doses Pip: median dose 16 g/day in four divided doses and Gent: 240 mg /day in three divided doses	Median 7 days Median 7 days	At 72 h†† I: 68/116 (59%) Pip/Gent: 65/117 (56%) At end of treatment‡‡ I: 58/103 (55%) Pip/Gent: 58/110 (53%) No significant difference in either group
Cornelissen et al (101)	Prospective, randomized, open	87 (94 evaluable febrile episodes)	Febrile neutropenia	I: 500 mg every 6 h Gent: 80 mg every 8 h and cefuroxime: 1.5 g every 8 h (n=35) or Cephalothin 1 g every 4 h (n=12) (Note: 65% of patients on cipro prophylaxis)	At least 7 days for initial responders	I: 43/49 (91%)§§ Gent + cefuroxime or cephalothin: 35/47 (74%)§§ P=0.05
Liang et al (102)	Prospective, randomized, open	89 (100 evaluable fever episodes)	Febrile neutropenia	I: 500 mg every 6 h Ceft: 2 g every 8 h	At least 7 days for initial responders, or 4 days postfebrile episode	I: 37/48 (77%)¶¶ Ceft: 29/52 (56%)¶¶ P=0.04

^*Mean number of days of treatment.

^†“Success” defined as initial intervention resolved the intra-abdominal infectious process with no infectious wound complications.

^‡“Resolution of the infection without additional antimicrobials”.

^§Cured or improved at four to 14 days post-therapy.

^¶Disappearance of signs and symptoms related to the infection three to seven days after completion of therapy.

^**Clinical signs and symptoms subside with complete resolution of active infection.

^††Response defined as temperature less than 37.5°C for 48 h.

^‡‡Response defined as temperature less than 37.5°C to 38°C based on initial response at 72 h.

^§§Response defined as improvement or resolution of signs and symptoms of infection with no need for other antimicrobials.

^¶¶Response to initial monotherapy defined as complete disappearance of all clinical and laboratory evidence of infection including fever. Ceft Ceftazidime; Cipro Ciprofloxacin; Clinda Clindamycin; Gent Gentamicin; I Imipenem; IV Intravenous; Metro Metronidazole; NR Not reported; Pip Piperacillin; Tazo Tazobactam; Tobra Tobramycin

TABLE 9.

Direct comparisons of imipenem and meropenem for treatment of serious infections in hospitalized patients

Study	Design	Number of patients enrolled in study	Indication	Regimen	Duration*	Results
Kanellakopoulou et al (65)	Prospective, randomized, open	62	Diffuse or local peritonitis	I: 1 g every 8 h versus M: 1 g every 8 h	8.6 days 7.7 days	30/31 (96.8%)†, 29/31 (93.5 %)‡ 28/28 (100%)†, 27/28 (96.4%)‡ P=“not significant”
Geroulanos et al (66)	Prospective, randomized, open	232	Moderate-severe intra-abdominal infection	I: 1 g every 8 h versus M: 1 g every 8 h	8.3 days 7.8 days	83/88 (94%)§, 58/66 (88%)¶ 79/82 (96%)§, 57/63 (90%)¶ P=0.534**
Brismar et al (67)	Prospective, randomized, open	249	Primarily “low risk” intra-abdominal infection (APACHE II= 0–10)	I: 500 mg every 8 h versus M: 500 mg every 8 h	5.1 days 5.4 days	86/90 (96%)†† 97/99 (98%)†† P=0.342**

^*Mean number of days of treatment.

^†Clinical “cure” at the end of therapy.

^‡Clinical “cure” at follow-up longer than 30 days.

^§Significant clinical response defined as “cure or improvement” at the end of therapy; “cure” defined as complete resolution of signs and symptoms of infection without addition of antibiotics or recurrence of symptoms; “improvement” defined as significant improvement in signs and symptoms without complete resolution of infection but allowing study treatment to be stopped.

^¶Significant clinical response defined as “cure or improvement” at two to four weeks’ follow-up.

^**Not statistically significant (for results at end of therapy).

^††“Cure” defined as complete remission of signs and symptoms of infection without further surgical intervention, addition of other antibiotics and without recurrence of symptoms. I Imipenem; M Meropenem

Although the trials were prospective and randomized, no direct comparisons of imipenem and meropenem were double-blinded (65–68). In addition, none of the trials comparing imipenem with meropenem demonstrated a statistically significant difference between treatment groups. Although two of three (67%) of the trials listed in Table 9 had a total sample size of more than 200 patients, approximately 140 patients per arm would be required to show a statistically significant difference if a 10% difference in cure rate existed (assuming the proportion of subjects expected to have clinical cure is 0.85, α=0.05, β=0.20) (69). None of the three studies presented had the required number of patients per arm.

INTRA-ABDOMINAL INFECTIONS

Imipenem:

Several clinical trials have compared imipenem with various antimicrobial combinations effective against a combination of anaerobic and aerobic bacteria for the treatment of intra-abdominal infections (70–75). Solomkin et al (75) reported a clinically and statistically significant improvement in favour of imipenem (83%) over a combination of clindamycin/tobramycin (70%) for the resolution of intra-abdominal infections. Differences were explained by a high failure rate in patients with Gram-negative organisms (primarily E coli and Enterobacter species) and an increased incidence of fasciitis requiring operative debridement for patients in the tobramycin/clindamycin group (75). Eklund et al (72) conducted a prospective, randomized, open trial comparing piperacillin/tazobactam (4 g/500 mg every 8 h) with imipenem (500 mg every 8 h) for treatment of severe intra-abdominal infections (72). Although piperacillin/tazobactam was statistically more effective than imipenem (91% cured versus 69%, respectively), the dose of imipenem was smaller than the currently recommended doses for serious infections (72). All trials investigating intra-abdominal infections presented in Table 7 used imipenem at doses of 500 mg every 6 h or 1 g every 8 h. Other trials investigating intra-abdominal infections demonstrated no statistically significant difference between imipenem and piperacillin/tazobactam or tobramycin plus clindamycin (or metronidazole) (73,74).

Meropenem:

In an open trial conducted by Huizinga et al (76) for the treatment of intra-abdominal infections, cefotaxime plus metronidazole achieved a significantly higher cure rate than meropenem (100% versus 91%, respectively; P=0.008) (Table 8). Clinically, both agents achieved a cure rate higher than 90%, and three of the failures in the meropenem group were attributed to ‘surgical misadventure’. Bacteriological cure rates were not significantly different (92% versus 90%, respectively) (76).

TABLE 8.

Meropenem versus selected antibiotics for treatment of serious infections in hospitalized patients

Study	Design	Patients (clinically evaluable patients)	Indication	Regimen	Duration*	Results
Huizinga et al (76)	Prospective, randomized, open	160 (148)	Intra-abdominal infection, diffuse or local peritonitis	C: 2 g every 8 h and Metro: 500 mg every 8 h M: 1 g every 8 h	C/Metro: 6.0 days M: 6.5 days	C/Metro: 78/78 (100%)†, 62/64 (97%)§ M: 64/70 (91%)†,‡, 53/54 (98%)§ P=0.008¶
Condon et al (77)	Prospective, randomized, double-blind	177 (127)	Majority: perforated appendicitis or intra-abdominal abscess	Clinda: 900 mg IV every 8 h and Tobra: 5 mg/kg/day divided every 8 h M: 1 g q8h	Clinda/Tobra 7 days M: 7.5 days	Clinda/Tobra 56/63 (89%)** M: 59/64 (92%)** χ2=0.41; not significant
Berne et al (78)	Prospective, randomized, double-blind	228 (129)	Advanced appendicitis (gangrenous or perforated)	Clinda: 900 mg IV every 8 h and Tobra: 5 mg/kg/day divided every 8 h M: 1 g every 8 h	Clinda/Tobra: 7.3 days M: 6.1 days	Clinda/Tobra: 60/66 (91%)†† M: 58/63 (92%)††
Wilson et al (79)	Prospective, randomized, double-blind	427 (191)	Intra-abdominal infection (majority complicated appendicitis); majority of APACHE II scores ≤10	Clinda: 900 mg IV every 8 h and Tobra: 5 mg/kg/day divided every 8 h M: 1 g every 8 h	Clinda/Tobra: 7.5 days M: 7.2 days	Clinda/Tobra: 81/94 (86%)*,‡‡ 115/134 (86%)§§ M: 89/97 92%)†,‡‡, 120/132 (91%)§§
Byrne et al (85)	Prospective, randomized 2:1 M: Ceft, open	40 (75 clinically evaluable episodes of Pseudomonas species infections)	Bacterial exacerbation of cystic fibrosis (Pseudomonas species infection)	Ceft 50 mg/kg every 8 h M: 25 mg/kg every 8 h	Ceft: 15 days M: 15 days	Ceft: 19/21 (90%)¶¶, 17/20 (85%)*** M: 53/54 (98%)¶¶, 43/50 (86%)***
Solberg, Sjursen (88)	Pooled subgroups from four prospective, randomized, open studies	153 (131)	Bacteremia; 108/153 defined as serious, originating from various sites	Ceft variable: 250 mg to 2 g every 8 h Ceft/Ami: 2 g every 8 h/15mg/kg/day M: variable 0.5 to 1 g every 8 h	Ceft +/– Ami: 9.6 days M: 9.8 days	Ceft+/– Ami: 66/70 (94%)† 45/45 (100%)§ M: 56/61 (92%)†, 32/33 (97%)§
Klugman et al (93)	Prospective, randomized, open	190 (139)††† median patient age 1 year	Bacterial meningitis	C: 75 to 100 mg/kg every 8 h M: 40 mg/kg every 8 h (dexamethasone 0.15 mg/kg every 6 h x four days in each group)	C: 9.7 days M: 9.9 days	With pre-existing neurological abnormalities before antibiotic:
							M (n=17)	C (n=5)
						Cure	8	3
						Cure + A	1	1
						Cure + N	6	0
						Cure + AN	2	1
						Without pre-existing neurological abnormalities:
							M (n=58)	C (n=59)
						Cure	46	49
						Cure + A	9	7
						Cure + N	2	1
						Cure + AN	1	0
								2 deaths
Schmutzard et al (94)	Pooled data from two prospective, randomized, open trials	56 (45) adult patients	Bacterial meningitis	Cef:100 mg/kg load then 80 mg/kg OD C: 75 to 100 mg/kg every 8 h M: 40 mg/kg every 8 h (dexamethasone 0.15 mg/kg every 6 h for 4 days)	Cef: 10.5 days C: 14.4 days M: 10.6 days		M (n=28)	C (n=17)	Cef (n=11)
						Cure	7	6	5
						Cure + A	12	1	3
						Cure +N	3	0	0
						Cure + AN	1	2	0
						Worse	0	3	2
						UE	5	5	1
Meropenem Study Group of Leuven London and Nijmegen (103)	Prospective, randomized, open	248 (304 evaluable episodes)	Febrile neutopenia	Ceft: 2 g tid M: 1 g tid	Ceft: 11.3 days M: 10.7 days	62/151 (41%)‡‡‡ 67/153 (44%)‡‡‡ Odds ratio 1.1 95% CI = 0.71 to 1.76

^*Mean number of treatment days.

^†Clinically cured or improved at end of therapy.

^‡Three failures attributed to “surgical misadventure”.

^§Clinically cured or improved at two to four weeks follow-up.

^¶Statistically significant difference at end of therapy.

^**“Success” defined as no clinical evidence of infection at any body site at the completion of antibiotic therapy, no relapse of abdominal infection during follow-up (28 to 42 days posthospitalization) and no adverse drug reaction requiring termination.

^††“Failure” defined as subsequent development of intra-abdominal abscesses, persistent or recurrent clinical signs of infection (without an identifiable septic source), or addition or change of antibiotic regimen.

^‡‡“Cured” defined as no clinical signs/symptoms of infection; “improved” defined as significant abatement of signs and symptoms of infection.

^§§Intent-to-treat analyses.

^¶¶“Satisfactory response” defined as improvement in lung function, ease of breathlessness, weight gain and general well-being at end of therapy.

^***“Satisfactory response” defined as improvement in lung function, ease of breathlessness, weight gain and general well-being at follow-up in four to six weeks.

^†††Cerebrospinal fluid culture confirmed bacterial meningitis included for efficacy analysis.

^‡‡‡All signs and symptoms of infection resolved without modification of empirical regimen (ie, addition of antifungal or modification of the antibiotic regimen). C Cefotaxime; c + A Cure with audiological sequelae; c + AN Cure with audiological and neurological sequelae; c + N Cure with neurological sequelae; Cef Ceftriaxone; Ceft Ceftazidime; Ceft/Ami Ceftazidime and amikacin; Clinda Clindamycin; IV Intravenous; M Meropenem; Metro Metronidazole; OD Once daily; Tobra Tobramycin; UE Unevaluable

Prospective, randomized, double-blind studies conducted by Condon et al (77) and Berne et al (78) reported no significant difference in clinical cure rates when clindamycin (900 intravenous every 8 h) and tobramycin (5 mg/kg/day divided every 8 h) were compared with meropenem (1 g every 8 h) for the treatment of serious intra-abdominal infections (77,78). Most recently, Wilson (79) confirmed the similar clinical cure rates of meropenem (1 g intravenous every 8 h) compared with clindamycin (900 mg intravenous every 8 h) plus tobramycin (5 mg/kg/day in three divided doses) for treatment of intra-abdominal infections (Table 8).

Imipenem versus meropenem:

All trials presented in Table 9 comparing imipenem with meropenem for treatment of low risk to severe intra-abdominal infections were prospective, randomized, open trials (65–67). Doses used were 1 g every 8 h in all cases except for the study conducted by Brismar et al (67) with low risk patients, which used 500 mg every 8 h. No clinically or statistically significant difference was reported between imipenem and meropenem in any of the trials, all of which achieved cure rates higher than 90% (65–67).

RESPIRATORY TRACT INFECTIONS

Imipenem:

Imipenem has been studied extensively for the treatment of severe lower respiratory tract infections (80–84). In the largest prospective, randomized, double-blind trial of clinically evaluable patients with severe pneumonia, Fink et al (84) reported a significantly higher clinical response rate with ciprofloxacin compared with imipenem (69% versus 56%, respectively; P=0.021) (Table 7). The subsequent intent-to-treat analysis showed a greater clinical response rate in favour of ciprofloxacin that was not statistically significant (64% versus 56%, P=0.123). Bacteriological eradication rates were higher in patients treated with ciprofloxacin (76%) than treated with imipenem (68%), a result that was primarily attributed to a superior eradication of Enterobacteriaceae. The isolation of P aeruginosa from initial respiratory tract cultures was associated with failure to achieve bacteriological eradication in 67% of patients receiving ciprofloxacin and 59% receiving imipenem. Development of resistance to P aeruginosa occurred during therapy in 33% and 53% of patients treated with ciprofloxacin and imipenem, respectively (84). Krilov et al (82) also reported that P aeruginosa rapidly developed resistance to imipenem in 11 of 19 patients treated for acute pulmonary exacerbations of cystic fibrosis. Monotherapy with imipenem should be discouraged in the treatment of severe lower respiratory tract infections if P aeruginosa is isolated.

Meropenem:

No prospective, randomized trials specifically regarding treatment for pneumonia are available. Byrne et al (85) conducted a prospective, randomized, open trial to determine the efficacy of meropenem compared with ceftazidime for the treatment of Pseudomonas species infections in cystic fibrosis patients (Table 8). Meropenem produced a “satisfactory response”, defined as improvement in lung function, ease of breathlessness, weight gain and general well-being in 98% of 54 evaluable episodes. Monotherapy with ceftazidime produced a 90% success rate in 21 evaluable episodes. Total bacterial counts were reduced by 73% in the meropenem group (n=59) and by 65% in the ceftazidime group (n=20). No statistical analysis was performed. Although this study had an open design with a small sample size, a 98% “satisfactory response” suggests that meropenem may show promise as an effective treatment option for Pseudomonas species infections in patients with cystic fibrosis (85).

Imipenem versus meropenem:

Comparisons of imipenem versus meropenem for respiratory tract infections are limited to acute bacterial exacerbations of chronic obstructive pulmonary disease. Although Hamacher et al (68) reported “cure or improvement” rates greater than 95% with either antibiotic, the authors absolutely do not recommend the use of either carbapenem for acute exacerbations of chronic bronchitis.

BACTEREMIA

Imipenem:

There are a limited number of clinical trials investigating the efficacy of imipenem in the treatment of bacteremia (86,87). Thirty-four patients requiring admission to the intensive care unit for treatment of bacteremia were studied by Linton et al (86) in a prospective, open, noncomparative trial. Imipenem doses ranging from 2 to 3 g/day in divided doses achieved a clinical cure in 28 of 34 patients (82%) and bacteriological eradication in 24 of 34 patients (71%) (86). In a more rigorous prospective, randomized, single-blind trial of serious hospital-acquired infections, Norrby et al (87) compared imipenem (500 mg every 6 h) and ceftazidime (2 g every 12 h) (Table 7). Ninety-one of 393 patients were diagnosed with septicemia (defined as clinical deterioration and rigors in association with unstable hemodynamic parameters and/or coagulopathy consistent with sepsis with or without bacteremia), of which 66 were clinically evaluable. There was no statistically significant difference between treatment groups, with 81% of imipenem patients and 74% of ceftazidime patients achieving a complete resolution of active infection (87). E coli, Klebsiella species and P aeruginosa were the most commonly isolated organisms. Overall, there was “no significant difference” in bacteriological eradication rates between the treatment groups in septic patients (P value was not provided).

Meropenem:

Ceftazidime with or without amikacin was compared with meropenem for the treatment of 153 patients with bacteremia (Table 8). Solberg and Sjursen (88) pooled the results of four prospective, randomized, open trials using a common protocol. There was no clinically significant difference between the treatment groups, with a clinical response rate of 92% in the meropenem group (n=61) and 94% in the ceftazidime with or without amikacin group (n=70) at the end of therapy (88). No statistical analysis was performed. Although a common protocol was used in all four studies, there is potential for variability in pooled results if the protocols were not followed in precisely the same fashion among studies (89).

Imipenem versus meropenem:

No direct comparative, prospective, randomized trials have been published comparing imipenem with meropenem.

BACTERIAL MENINGITIS

Imipenem:

Although imipenem has been investigated for the treatment of bacterial meningitis, development of seizures has been a concern in trials conducted thus far (90–92). In a small, prospective, noncomparative, open trial (n=21), Wong et al (91) reported a 33% incidence of seizures after administration of imipenem (25 mg/kg every 6 h) in children without seizures before therapy. This incidence is much higher than that reported in postmarketing surveillance of all indications (1.5% to 2%) (92).

Meropenem:

Meropenem has proven to be effective in the treatment of bacterial meningitis in clinical trials. Prospective, randomized, open studies conducted by Klugman and Dagnan (93) and Schmutzard et al (94) reported clinical cure with and without audiological and neurological sequelae in all clinically evaluable patients treated with meropenem (Table 8). In the largest of the two studies, 139 of 190 children had positive pretherapy CSF cultures. Of these, 121 (63 in the meropenem group and 58 in the cefotaxime group) underwent repeat lumbar puncture within 18 to 36 h, with a bacterial eradication rate of greater than 95% in both groups (93). Overall, in patients with no pre-existing seizures before therapy, seizures occurred in five of 82 patients (6%) receiving meropenem and one of 86 (1%) patients receiving cefotaxime (93). Although the difference was not statistically significant, continued monitoring for seizure potential in future trials is recommended.

Imipenem versus meropenem:

No direct comparative, prospective, randomized trials have been published comparing imipenem with meropenem.

FEBRILE NEUTROPENIA

Imipenem:

Imipenem has been extensively studied in the management of febrile neutropenia (94–99). In the largest prospective, randomized, single-blind study, Leyland et al (100) treated 252 febrile episodes with either imipenem or a combination of piperacillin and gentamicin (Table 7). Success, defined as a temperature less than 37.5°C by 72 h and maintained for 48 h, was achieved by 59% of patients in the imipenem group and by 56% of the piperacillin/gentamicin group (100). Patients requiring additional antimicrobials or antifungal agents to manage fever were considered treatment failures. In contrast, Cornelissen et al (101) reported an imipenem success rate of 91% in 94 febrile neutropenic episodes. Only 18% of episodes were considered fever of unknown origin, with 82% classified as acquired infections. Ciprofloxacin (500 mg bid) was used prophylactically in 65% of febrile episodes, in patients with hematological malignancy expected to be profoundly neutropenic for more than one week. Seventy-six per cent of causative microorganisms were Gram-positive. The most pronounced difference in efficacy was reported for microbiologically documented infections caused primarily by Gram-positive organisms (imipenem 89% versus gentamicin plus cefuroxime or cephalothin 53%, P=0.025) (101). Another trial reported success rates, defined as complete disappearance of all clinical and laboratory evidence of infection, in 77% of 100 evaluable febrile episodes treated with imipenem (500 mg every 6 h) (Table 7) (102).

Meropenem:

Few studies are available investigating meropenem for the management of febrile neutropenic patients. A recent study of 304 evaluable episodes of febrile neutropenia compared ceftazidime (2 g tid) with meropenem (1 g tid) (103). Response rates, defined as resolution of all signs and symptoms without modification of the empirical regimen, were 41% and 44%, respectively, and were not statistically or clinically different (95% CI 0.71 to 1.76) (103).

Imipenem versus meropenem:

No direct comparative, prospective, randomized trials have been published comparing imipenem with meropenem.

SUMMARY OF CLINICAL TRIALS

The majority of prospective, randomized trials published in peer reviewed journals had an open design and used meropenem in doses of 1 g every 8 h or imipenem at 500 mg every 6 h or 1 g every 8 h. In direct comparisons of imipenem and meropenem with identical dosing regimens, there does not appear to be a clinically or statistically significant difference for treatment of intra-abdominal infections or acute bacterial exacerbations of chronic obstructive pulmonary disease. Direct comparisons were not available for other severe life-threatening infections including septicemia, bacterial meningitis and febrile neutropenia.

There appears to be no clinically significant difference in the treatment of serious, life-threatening infections with imipenem or meropenem compared with standard antimicrobial regimens. Although only four trials provide a direct comparison of imipenem with meropenem, clinical cure rates from several trials against standard comparators suggest that imipenem and meropenem have similar efficacy for the treatment of serious life-threatening infections. Imipenem dose regimens of 500 mg every 6 h (73,75,84,87,101,102) and 1 g every 8 h (74,84,100) and meropenem 1 g every 8 h (65,66,76–79,88,103) have been reported to be effective in the treatment of serious infections. One exception involves using imipenem for the treatment of bacterial meningitis because imipenem can cause seizures. In this case, meropenem should be considered the preferred carbapenem.

ADVERSE EFFECTS

Mild, self-limiting adverse effects reported with meropenem are similar to those reported with imipenem (1,4,5,104, 105). The most common adverse effects reported with meropenem and imipenem include local irritation at the injection site, diarrhea, rash, nausea, vomiting and pruritus (105). All of these adverse effects reversed upon discontinuation of the antibiotic, and none were reported to exceed incidences associated with other beta-lactams. Adverse events requiring drug withdrawal occurred in 1.4% of patients treated with meropenem and 1.8% of patients treated with imipenem (105).

Both imipenem and meropenem can affect laboratory tests. Like other beta-lactams, imipenem and meropenem can cause mild, transient increases in hepatic enzymes such as alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase and lactate dehydrogenase (less than 5%) (104,105). In addition, meropenem and imipenem have been reported to cause increases in serum creatinine and serum urea (less than 1%). The most frequent drug-induced hematological changes are thrombocytosis and eosinophilia (less than 2%). No significant differences in the frequency of these altered laboratory tests have been reported between meropenem and imipenem (104,105). Both meropenem and imipenem display cross-allergenicity with other beta-lactam antibiotics and, thus, are contraindicated in patients with a history of anaphylactic reactions to penicillins and cephalosporins.

A notable adverse effect associated with imipenem is the development of seizures. Phase 3 trials and postmarketing surveillance have documented the incidence of imipenem induced seizures to be 1.5% to 2.0% (105,106). Risk factors for seizures include impaired renal function, pre-existing central nervous system disease or infection, stroke or past history of seizures (105,106). As mentioned previously, in one small clinical trial of meningitis, imipenem was associated with a 33% incidence of seizures (91). In contrast, both animal data and noncomparative clinical trials demonstrate that meropenem has a lower propensity than imipenem to cause seizures (105,107). As a result of these data, meropenem is indicated for the treatment of meningitis, while imipenem is not (108).

SUMMARY AND PHARMACOECONOMIC CONSIDERATIONS

Meropenem and imipenem are clearly, equally efficacious (bacteriologically and clinically) for the treatment of serious infections. This is not surprising because they display similar in vitro activity and pharmacokinetics. Meropenem, however, offers the advantages of use for the treatment of meningitis because it has a lower likelihood of causing seizures than imipenem. In addition, meropenem’s recommended dose regimen for the treatment of serious infections (1 g every 8 h) (108) requires one fewer dose per day than imipenem’s regimen (500 mg every 6 h) (109). This translates into cost savings (approximately $5.00/day) in nursing and pharmacy preparation time and in materials (110). We believe that there is no clinical relevance of the increased stability of meropenem to dehydropeptidase 1 and consequently no need to administer cilastatin concurrently, which is required with each dose of imipenem. Acquisition cost comparison between meropenem (1 g every 8 h or 3 g/day) and imipenem (500 mg every 6 h or 2 g/day) for the treatment of serious infections (in patients with normal renal function) suggests that meropenem is significantly more expensive (approximately $142/day) than imipenem (approximately $98/day) (108–110). However, the higher acquisition cost for meropenem is slightly offset by the approximate saving of $5.00/day and convenience with meropenem as a result of fewer administered doses per day (every 6 h for imipenem versus every 8 h for meropenem).

Present studies support the use of meropenem 500 mg every 8 h only for the treatment of mild to moderate infections (urinary tract, skin and soft tissue, low risk intra-abdominal infections and community acquired pneumonia requiring hospitalization) (67,111–113). Whether meropenem 500 mg every 8 h can be used for the treatment of serious infections is not known because no clinical data are available for evaluation.