Article Images
Antimicrob Agents Chemother. 2003 Nov; 47(11): 3442–3447.
G. Zanetti,1,* F. Bally,1 G. Greub,1 J. Garbino,2 T. Kinge,2 D. Lew,2 J.-A. Romand,2 J. Bille,1 D. Aymon,1 L. Stratchounski,3 L. Krawczyk,4 E. Rubinstein,5 M.-D. Schaller,1 R. Chiolero,1 M.-P. Glauser,1 A. Cometta,1 and the Cefepime Study Group†
Abstract
In a randomized, evaluator-blind, multicenter trial, we compared cefepime (2 g three times a day) with imipenem-cilastatin (500 mg four times a day) for the treatment of nosocomial pneumonia in 281 intensive care unit patients from 13 centers in six European countries. Of 209 patients eligible for per-protocol analysis of efficacy, favorable clinical responses were achieved in 76 of 108 (70%) patients treated with cefepime and 75 of 101 (74%) patients treated with imipenem-cilastatin. The 95% confidence interval (CI) for the difference between these response rates (−16 to 8%) failed to exclude the predefined lower limit for noninferiority of −15%. In addition, therapy of pneumonia caused by an organism producing an extended-spectrum β-lactamase (ESBL) failed in 4 of 13 patients in the cefepime group but in none of 10 patients in the imipenem group. However, the clinical efficacies of both treatments appeared to be similar in a secondary intent-to-treat analysis (95% CI for difference, −9 to 14%) and a multivariate analysis (95% CI for odds ratio, 0.47 to 1.75). Furthermore, the all-cause 30-day mortality rates were 28 of 108 (26%) patients in the cefepime group and 19 of 101 (19%) patients in the imipenem group (P = 0.25). Rates of documented or presumed microbiological eradication of the causative organism were similar with cefepime (61%) and imipenem-cilastatin (54%) (95% CI, −23 to 8%). Primary or secondary resistance of Pseudomonas aeruginosa was detected in 19% of the patients treated with cefepime and 44% of the patients treated with imipenem-cilastatin (P = 0.05). Adverse events were reported in 71 of 138 (51%) and 62 of 141 (44%) patients eligible for safety analysis in the cefepime and imipenem groups, respectively (P = 0.23). Although the primary end point for this study does not exclude the possibility that cefepime was inferior to imipenem, some secondary analyses showed that the two regimens had comparable clinical and microbiological efficacies. Cefepime appeared to be less active against organisms producing an ESBL, but primary and secondary resistance to imipenem was more common for P. aeruginosa. Selection of a single agent for therapy of nosocomial pneumonia should be guided by local resistance patterns.
Nosocomial pneumonia is the most frequent nosocomial infection among patients in intensive care units (ICUs) (30). It occurs in 10 to 25% of ICU patients (5, 8, 11, 12, 20, 28, 30) and is associated with high overall rates of mortality, which range from 22 to 71% (11, 27, 28).
The adequacy of the initial antibiotic therapy is an independent predictor of both overall mortality and mortality attributable to pneumonia (1, 18, 22, 25). Since the causative pathogens are not usually known at the time of diagnosis, it is critical for an early empirical therapy to be active against the most common agents of nosocomial pneumonia in ICUs, such as Pseudomonas aeruginosa, members of the family Enterobacteriaceae, or Staphylococcus aureus (17). Several broad-spectrum antibiotics, including imipenem-cilastatin (7, 13, 15), piperacillin-tazobactam (15), ceftazidime (9, 14), and fluoroquinolones (13), were recently successfully used as monotherapy in this setting.
Cefepime is a “fourth-generation” cephalosporin (26) with good activity against most nosocomial gram-negative pathogens, as well as gram-positive cocci such as Streptococcus pneumoniae and methicillin-susceptible S. aureus (21). In contrast to other cephalosporins, cefepime is a weaker inductor of chromosomal β-lactamases, and it shows good stability against both chromosomal and most plasmidic β-lactamases (16). Cefepime is therefore an appealing option for monotherapy of nosocomial pneumonia. Several trials with relatively small sample sizes comparing cefepime with ceftazidime or cefotaxime suggested that cefepime is as effective as broad-spectrum cephalosporins for the treatment of community-acquired lower respiratory tract infections (2, 10, 19).
The present study was performed to compare cefepime with imipenem-cilastatin for monotherapy of nosocomial pneumonia in ICU patients with or without mechanical ventilation.
MATERIALS AND METHODS
Patients.
We evaluated for enrollment all patients hospitalized in the ICUs of 13 hospitals in six European countries between April 1997 and May 1999. To be eligible for the study, patients with or without mechanical ventilation had to be age 16 years or older and had to have nosocomial pneumonia. Nosocomial pneumonia was defined as a new or progressive infiltrate on the chest X-ray film that occurred at least 48 h after hospital admission in the presence of two or more of the following findings: temperature, ≥38°C; cough, sputum production, or tracheal secretions; leukocyte count, ≥10 × 109/liter or >15% band forms; hypoxemia (defined as a partial O2 pressure <70 mm Hg while the patient was breathing normal air or a decrease in the partial O2 pressure of ≥25% from an initial value); and rales and/or evidence of consolidation on pulmonary auscultation.
Patients were excluded if they had hypersensitivity to carbapenems or β-lactams; a neutrophil count <109/liter; were infected with human immunodeficiency virus; were infected with a microorganism known to be resistant to one of the study antibiotics; had a primary diagnosis of viral, fungal, or mycobacterial infection; had been treated with one of the study antibiotics within 4 weeks before randomization or with any investigational drug within 30 days before randomization; had hepatic failure; had a high likelihood of death within 48 h; or were pregnant or lactating.
All patients or their legally authorized representative gave written informed consent. The protocol was approved by the institutional review boards of all participating centers.
Study design and treatment.
This was a randomized, controlled, open-label, evaluator-blind trial. Patients were assigned to receive one of the following regimens by using a centralized telephone randomization system: cefepime at 2 g three times a day (which is the recommended regimen for life-threatening infections) or imipenem-cilastatin at 500 mg four times a day. The dosages of both antibiotics were adjusted according to the patient's renal function when appropriate. The physician in charge of the patient determined the duration of therapy.
Assessment and monitoring.
At the baseline, a complete history was obtained and a physical examination was performed by a local investigator in each participating center. The APACHE II score and the clinical pulmonary infection score (CPIS) (24) were determined. CPIS is based on body temperature, leukocyte count, the volume and character of tracheal secretions, arterial oxygenation, findings on the chest X-ray film, and Gram stain and culture of a tracheal aspirate. A good correlation between CPIS and quantitative bacteriology with samples obtained by invasive diagnostic methods has been described (24). The APACHE II score at the time of admission to the ICU was also calculated. Clinical data were then collected twice weekly and within 72 h after completion of study drug therapy. Blood chemistry and hematology evaluations were performed at the baseline and then twice weekly and within 48 h after the completion of study drug therapy. The baseline microbiological evaluation included two sets of blood cultures for all patients and a Gram stain and culture of tracheal secretions for mechanically ventilated patients. In patients without mechanical ventilation, a Gram stain and culture of sputum were performed when sputum was available. Collection of specimens by the use of invasive techniques was performed as deemed necessary by the physicians in charge of the patients. The microbiological evaluation was repeated between 48 and 96 h after the initiation of study drug therapy and within 72 h after the completion of study drug therapy. The evaluation included a new sampling of lower respiratory tract secretions, if available, and one set of blood cultures if the culture result for a previous blood sample was positive. Culture and antibiotic susceptibility testing were done at each participating center according to the recommendations of the American Society for Microbiology and the Kirby-Bauer method, respectively. The National Committee for Clinical Laboratory Standards recommendations for breakpoints (23) were used to define the antibiotic susceptibility levels of the pathogens isolated. Bacteria assessed as causative were sent to a central laboratory (the laboratory of J.B.) for confirmation of the identification. Their susceptibility profiles were confirmed by use of E-tests (AB Biodisk, Solna, Sweden). When confirmation by the central laboratory was not available, the results from the participating centers were reported.
Microbiological documentation of nosocomial pneumonia was considered definite if a pathogen was isolated from lung tissue or simultaneously from blood or pleural fluid and a respiratory tract specimen. Documentation was considered probable when >104 CFU of a possible pathogen per ml was isolated from bronchoalveolar lavage fluid or if Haemophilus influenzae, S. pneumoniae, or Legionella pneumophila was the predominant organism in respiratory tract secretions. Pneumonia was considered a possible microbiologically documented infection when a possible pathogen was semiquantitatively predominant in any lower respiratory tract specimen in the presence of more than 25 polymorphonuclear leukocytes and less than 10 squamous epithelial cells per magnification field (10 × 10).
End points.
The primary end point of the study was comparison of the clinical response to cefepime with that to imipenem-cilastatin, as assessed blindly by the scientific coordinating committee (G.Z., G.G., and A.C.) according to guidelines on the evaluation of anti-infective drugs published by experts (3, 4, 6). This comparison was made for patients who have been treated according to the protocol (per-protocol analysis) after exclusion of those for whom circumstances precluded classification of treatment as a success or a failure. We also conducted a modified intent-to-treat analysis that included all patients who fulfilled the eligibility criteria, even if they had not been treated according to the protocol. Different from classical intent-to-treat analysis, randomized patients who were later found to violate the inclusion criteria were excluded. A therapeutic success was defined as either a cure (a complete resolution of symptoms and signs and an improvement or the lack of progression of abnormalities on the chest X-ray film) or an improvement (the same as for a cure, but with partial resolution of symptoms and signs). A minimal therapy duration of 5 days was required for assessment of the treatment as a success. The clinical response rate was defined as the number of patients whose clinical response was a success divided by the total number of patients included in the analysis. A treatment failure was defined as the absence of a response to treatment (a persistence or progression of signs and symptoms after 72 h of therapy, the development of new symptoms consistent with active infection, or the progression of radiographic abnormalities after 5 days of therapy), the documentation of a bacterium resistant to the allocated regimen resulting in the introduction of another antibiotic, a death due to pneumonia, or the inability to complete the study due to severe adverse effects. Patients who were not treated as required for efficacy assessment were also considered treatment failures in the modified intent-to-treat analysis, which avoided a biased comparison of the two groups in case they would differ regarding nonadherence to the protocol.
One secondary end point was comparison of the bacteriologic response obtained with each of the two drugs. Eradication was defined as the elimination of the causative organism(s) during or at the completion of therapy. When samples were not available for evaluation of the bacteriologic response because of clinical improvement (e.g., resolution of sputum production), the case was assessed as a presumed eradication. The bacteriologic response could also be classified as persistence (failure to eradicate the original causative organism from the sites previously listed), a relapse (recurrence of pulmonary infection with the same organism, i.e., an organism of the same genus and species and with the same antibiotic susceptibility profile, within 5 days after the discontinuation of treatment or during treatment after two consecutive cultures had been negative), a superinfection (a new lower respiratory tract infection during treatment or within 3 days after the completion of treatment due to an organism not previously recognized as causative), colonization (at least two cultures yielded a bacterium other than the primary causative isolate, in the absence of evidence of a new or progressive infection), and indeterminate (circumstances prevented classification of the response).
Finally, safety was assessed twice weekly until the end of treatment and within 72 h after the completion of treatment for all patients who received at least one dose of study medication.
Statistical analyses.
We assumed that the proportion of patients with a favorable clinical response in the imipenem-cilastatin group would be 80%. The null hypothesis was that the absolute difference in the proportion of patients with a favorable clinical response between the imipenem-cilastatin and cefepime groups would exceed 15%. Noninferiority would be established in case of rejection of the null hypothesis, which required that the upper limit of the two-sided 95% confidence interval (CI) for the difference in clinical response rates was less than or equal to 15%. The choice of the 15% margin was recommended in 1992 by the U.S. Food and Drug Administration for trials of anti-infective therapy when the expected response rate was 80%. By using an alpha level of 5% and a power of 80% and assuming that 10% of the randomized patients would not be treated according to the protocol, a total of 250 patients were required. The sample size was increased to 280 patients during the study because a higher than expected proportion of patients were not evaluable for the primary efficacy analysis. The decision to increase the sample size was made before any comparison of the clinical responses of the two groups.
Clinical response rates were further compared after adjustment for predictors of clinical response. For that purpose, significant predictors at a 0.1 level of significance were candidates for inclusion in a logistic regression model that was built through a forward selection process.
Noninferiority of the bacteriological response to cefepime compared with that to imipenem-cilastatin was tested by using the method mentioned above for clinical efficacy.
Baseline characteristics in the two groups were compared by use of the two-sided Fisher exact test for proportions and the Wilcoxon rank-sum test for continuous and ordinal data. All tests were performed at a significance level of 5%.
Statistical analyses were performed with STATA statistical software (version 6.0; Stata Corp., College Station, Tex.).
RESULTS
Patient characteristics at randomization.
Two hundred eighty-one patients from 13 centers were randomly assigned. Of these, 209 (108 in the cefepime group and 101 in the imipenem-cilastatin group) were evaluable for clinical response by the per-protocol analysis. Eleven patients were excluded because of violation of the eligibility criteria (9 patients) or withdrawal of consent before the start of therapy (2 patients). These 11 patients were also excluded from the modified intent-to-treat analysis. The reasons for exclusion of 61 other patients were (i) a length of therapy that was too short in 29 patients (cefepime group, 12 patients; imipenem group, 17 patients) due to early deaths unrelated to pneumonia (15 patients), protocol violation (12 patients), or the patient's decision (2 patients); (ii) concomitant administration of another antibiotic in 18 patients (cefepime group, 8 patients; imipenem group, 10 patients); (iii) use of the wrong dosage of the study drug or the dose was omitted in 6 patients (cefepime group, 2 patients; imipenem group, 4 patients); and (iv) a clinical course that precluded assessment in 8 patients (cefepime group, 2 patients; imipenem group, 6 patients).
The two groups were balanced with respect to demographic and prestudy clinical characteristics (Table 1). The mean APACHE II score was nearly 15, and two-thirds of the patients in both groups were mechanically ventilated at the time of inclusion in the study.
TABLE 1.
Patients' characteristics at baseline
| Characteristic | Cefepime group (n = 108) | Imipenem-cilastatin group (n = 101) |
|---|---|---|
| Mean ± SDa age (yr) | 55 ± 18 | 53 ± 18 |
| Mean ± SD wt (kg) | 73.0 ± 13.9 | 76.0 ± 17.6 |
| No. (%) male | 72 (67) | 67 (66) |
| Mean ± SD APACHE II score at ICU admissionb | 15.6 ± 6.6 | 14.8 ± 6.3 |
| Mean ± SD APACHE II score at randomizationb | 15.4 ± 6.6 | 14.8 ± 6.7 |
| Mean ± SD CPIS scoreb | 7.1 ± 2.8 | 6.7 ± 2.0 |
| Mean ± SD previous hospital stay (days) | 10.2 ± 8.2 | 10.8 ± 8.2 |
| Mean ± SD previous ICU stay (days) | 6.3 ± 5.1 | 6.5 ± 5.8 |
| Mean ± SD serum creatinine (μmol/liter) | 116 ± 125 | 110 ± 118 |
| No. (%) of patients with previous surgery or trauma | 63 (58) | 65 (64) |
| No. (%) of patients with mechanical ventilation (%) | 71 (66) | 67 (66) |
| Mean ± SD PaO2/FiO2c | 189 ± 85 | 181 ± 90 |
| No. (%) of patients with concomitant bacteremia | 9 (8) | 11 (11) |
| No. (%) of patients who received prestudy antibiotics | 86 (80) | 85 (84) |
Nosocomial pneumonia was documented microbiologically in 148 (71%) of 209 patients, 77 of 108 (71%) in the cefepime group and 71 of 101 (70%) in the imipenem-cilastatin group (Table 2). The most frequent pathogen was P. aeruginosa, which was present in 59 (40%) patients with microbiological documentation of pneumonia. An extended-spectrum β-lactamase (ESBL)-producing causative organism was documented in 23 (16%) patients (cefepime group, 13 patients; imipenem group, 10 patients): Klebsiella pneumoniae was found in 22 patients (19 of whom were hospitalized in the same participating center), Enterobacter aerogenes was found in 1 patient, and Acinetobacter baumannii was found in 1 patient (together with K. pneumoniae).
TABLE 2.
Documentation and microbiology of nosocomial pneumonia
| Characteristic or causative agent | No. (%) of patients
| |
|---|---|---|
| Cefepime (n = 108) | Imipenem-cilastatin (n = 101) | |
| Clinical documentation | 31 (29) | 30 (30) |
| Microbiological documentation | 77 (71) | 71 (70) |
| Definite | 10 (9) | 12 (12) |
| Probable | 24 (22) | 19 (19) |
| Possible | 43 (40) | 40 (40) |
| Gram-negative bacteria | 59 (55) | 55 (55) |
| P. aeruginosa | 23 | 29 |
| K. pneumoniae | 18 | 16 |
| A. baumannii | 12 | 8 |
| H. influenzae | 5 | 3 |
| E. coli | 7 | 2 |
| Others | 25a | 17b |
| Gram-positive bacteria | 9 (8) | 8 (8) |
| S. aureus | 6 | 6 |
| S. pneumoniae | 2 | 3 |
| Others | 1 | |
| Mixed gram-positive and gram-negative bacteria | 9 (8) | 8 (8) |
| P. aeruginosa | 4 | 3 |
| S. aureus | 8 | 6 |
| Others | 10 | 12 |
Clinical efficacy.
The mean ± standard deviation lengths of therapy were 9.1 ± 4.2 days for the cefepime group and 9.4 ± 4.3 days for the imipenem-cilastatin group. The clinical response rate in the per-protocol analysis was 76 of 108 (70%) patients in the cefepime group and 75 of 101 (74%) patients in the imipenem-cilastatin group (95% CI for the difference in clinical response, −16 to 8%). Similarly, the clinical response rates in both groups were close in the modified intent-to-treat analysis (78 of 132 [59%] patients in the cefepime group, compared with 78 of 138 [57%] patients in the imipenem-cilastatin group; 95% CI for the difference, −9 to 14%); these rates were lower than those observed in the per-protocol analysis because the outcomes for patients for whom the protocol could not be completed were considered failures in the modified intent-to-treat analysis. Finally, the clinical response rates were not associated with the treatment group by multivariate analysis after adjustment for the following significant predictors of response: study center and APACHE II score at study entry (adjusted odds ratio for success with cefepime, 0.91; 95% CI, 0.47 to 1.75). The causes of treatment failure were similar in the two groups (Table 3).
TABLE 3.
Clinical outcomes
| Outcome | No. (%) of patients
| |
|---|---|---|
| Cefepime (n = 108) | Imipenem-cilastatin (n = 101) | |
| Success | 76 (70) | 75 (74) |
| Cured | 42 (39) | 43 (43) |
| Improved | 34 (31) | 32 (32) |
| Failure | 32 (30) | 26 (26) |
| Causes of failure: | ||
| Resistant pathogen | 7a | 8b |
| New infection | 7c | 5d |
| Lack of improvement of signs or symptoms | 6 | 7 |
| Worsening of signs or symptoms | 6 | 4 |
| Progression of X-ray infiltrates | 1 | |
| Death due to pneumonia | 3 | 1 |
| New signs or symptoms | 2 | |
| Adverse event | 1 | |
The clinical response rates of patients with P. aeruginosa infection were similar in the two groups: 23 of 27 (85%) patients in the cefepime group and 23 of 32 (72%) patients in the imipenem-cilastatin group.
Among patients infected with an ESBL-producing causative pathogen, the clinical response rates were 9 of 13 (69%) patients in the cefepime group and 10 of 10 (100%) patients in the imipenem-cilastatin group. The reasons for failure in four patients treated with cefepime were an absence of improvement in two patients, death attributed to pneumonia in one patient, and early withdrawal because of the intermediate susceptibility of the organism to the antibiotic in one patient (although the reference laboratory later measured an MIC of 4 μg/ml for the organism from this patient; i.e., the MIC was below the breakpoint for in vitro susceptibility). The MICs of cefepime for the organisms from these four patients ranged from 2 to 4 μg/ml.
The all-cause 30-day mortality rates were 28 of 108 (26%) patients in the cefepime group and 19 of 101 (19%) patients in the imipenem-cilastatin group (P = 0.25). Pneumonia contributed to the deaths of 10 (9%) patients treated with cefepime and 4 (4%) patients treated with imipenem-cilastatin, including 3 and 1 patients, respectively, for whom death was considered the reason of treatment failure according to the protocol.
Microbiological efficacy.
Antibiotic susceptibility to both study drugs was determined in 219 of the 226 causative pathogens (97%), which represented 143 of 148 pathogens from patients with microbiologically documented pneumonia. One hundred ninety-two of 219 (88%) pathogens were susceptible to cefepime at the baseline, and 204 of 219 (93%) pathogens were susceptible to imipenem-cilastatin at the baseline (P = 0.07). The causative pathogens from 121 of 143 (85%) patients were susceptible to cefepime, and the causative pathogens from 131 of 143 (92%) patients were susceptible to imipenem-cilastatin (P = 0.10). Eradication of the causative organism was proved or presumed in 47 (61%) of the 77 patients with microbiological documentation in the cefepime group and 38 of 71 (54%) patients in the imipenem-cilastatin group (95% CI for the difference in bacteriological response, −8 to 23%) (Table 4). The development of secondary resistance during cefepime therapy was observed in 6 of the 63 (10%) patients whose pathogens were initially susceptible to this drug, whereas the development of secondary resistance during imipenem-cilastatin therapy was observed in 10 of 61 (16%) patients (P = 0.29). Among patients with P. aeruginosa infections, resistance to the allocated regimen was initially present among the organisms from 2 of 27 patients in the cefepime group and 5 of 32 patients in the imipenem-cilastatin group. Secondary resistance was detected among the organisms from 3 of 25 patients during therapy with cefepime and 9 of 27 patients during therapy with imipenem-cilastatin. In total, primary resistance and secondary resistance were detected in the organisms from 5 of 27 patients (19%) treated with cefepime and 14 of 32 patients (44%) treated with imipenem (P = 0.05).
TABLE 4.
Microbiological outcomes among patients with microbiological documentation
| Outcome | No. (%) of patients
| |
|---|---|---|
| Cefepime (n = 77) | Imipenem-cilastatin (n = 71) | |
| Eradication | 13 (17) | 10 (14) |
| Presumed eradication | 34 (44) | 28 (39) |
| Persistence | 19 (25) | 22 (31) |
| Superinfection | 3 (4) | 3 (4) |
| Colonization | 3 (4) | 4 (6) |
| Not assessable | 5 (6) | 4 (6) |
Safety analysis.
Among 138 patients in the cefepime group eligible for safety analysis, 71 (51%) experienced 148 adverse events, whereas 62 of 141 (44%) patients in the imipenem-cilastatin group had 120 events. Adverse events possibly or probably related to the study drug were more frequent in the cefepime group than in the imipenem-cilastatin group (41 events in 33 patients and 22 events in 20 patients, respectively). One such event led to withdrawal of the drug (acute renal failure possibly related to cefepime in one patient). Specific therapy was necessary in 22 patients (14 patients in the cefepime group and 8 patients in the imipenem-cilastatin group). In both groups, the most frequent adverse event possibly related to a study drug was diarrhea (seven patients treated with cefepime and nine patients treated with imipenem-cilastatin). Diarrhea was due to Clostridium difficile in six patients treated with imipenem-cilastatin but in none of the patients treated with cefepime (P = 0.03). Renal problems developed in five patients in the cefepime group (four had a deterioration of renal function, defined as a serum creatinine level of more than 200 μmol/liter, and one had a possible case of interstitial nephritis) but in none of the patients in the imipenem-cilastatin group. The peak serum creatinine concentration in patients with deterioration of renal function ranged from 210 to 317 mmol/liter. In one of these patients the deterioration was associated with multiple organ failure; in another patient, gentamicin had been administered for 4 days before study entry. The serum creatinine levels were still elevated at the end of the 72-h follow-up in three of the four patients.
DISCUSSION
Several clinical studies (7, 13, 15) have previously shown the feasibility of monotherapy with imipenem-cilastatin for nosocomial pneumonia in critically ill patients. The clinical efficacy of cefepime was close to that of imipenem-cilastatin in the present study with patients from a panel of European ICUs, 66% of whom had ventilator-associated pneumonia. Indeed, the proportions of favorable clinical responses in the per-protocol analysis were 70% among patients treated with cefepime and 74% among patients treated with imipenem-cilastatin. These results are within the range of those obtained with other therapeutic regimens in comparable settings published previously (13, 15). These results do not formally exclude the inferiority of cefepime. Indeed, the lower limit of the CI for the difference in clinical efficacy (−16%) is slightly lower than −15%, which was the difference considered clinically irrelevant and was required in the protocol to claim noninferiority. However, the confidence interval should be interpreted with caution, because the number of evaluable patients in this study was smaller than anticipated (209 instead of the 225 required in the protocol), which resulted in a wider CI. In addition, the observed response rate was lower than the expected 80% assumed in the protocol for the sample size calculation. Therefore, the power of the study is not as high as the 80% that was planned. Moreover, both the result of the modified intend-to-treat analysis (59% success in the cefepime group compared with 57% success in the imipenem-cilastatin group; 95% CI for the difference, −9% to 14%) and the result of the multivariate analysis (adjusted odds ratio for success with cefepime, 0.91; 95% CI, 0.47 to 1.75) support the hypothesis that cefepime is not inferior to imipenem-cilastatin in terms of the clinical response rate.
The efficacy of cefepime for the therapy of nosocomial pneumonia is of interest in an era of increasing bacterial resistance that requires novel therapeutic options. Cefepime can also be administered less frequently on a daily basis than imipenem-cilastatin and showed interesting microbiological features. The causative organism was documented for 71% of the patients in the study. Of these organisms, 88 and 93% were susceptible to cefepime and imipenem-cilastatin, respectively. The bacteriological response to cefepime (61% of microbiologically documented infections) was not inferior to that of imipenem (54%). As is usually the case, P. aeruginosa was the most frequent pathogen and was found in 40% of patients with microbiological documentation of pneumonia. The propensity of P. aeruginosa to develop secondary resistance during therapy with imipenem-cilastatin has been reported previously (7, 15, 29), and this propensity was not reduced by the simultaneous use of aminoglycosides in one study (7). We confirmed that the resistance of P. aeruginosa prior to or during therapy was more frequent with imipenem-cilastatin than with cefepime. However, no difference in clinical outcome was observed among patients infected with this bacterium.
Another interesting characteristic of cefepime is its resistance to β-lactamase hydrolysis in vitro, which is different from the case for broad-spectrum cephalosporins. Our study provided an opportunity to compare cefepime with imipenem-cilastatin for the treatment of pneumonia attributed to ESBL-producing bacteria. All ESBL-positive strains were susceptible to cefepime in vitro when a threshold of 8 μg/ml was used. Among the 13 patients in the cefepime group infected with an ESBL-positive pathogen, however, 1 died because of the pneumonia and 2 did not experience improvements with therapy. There were no such unfavorable outcomes among the 10 patients treated with imipenem-cilastatin. These data may be of poor generalizability, because 19 of the 23 ESBL-positive strains were K. pneumoniae isolates from a single center, which likely represented an epidemic cluster. While more data are needed before advice on the use of cefepime for the treatment of infections caused by ESBL-positive organisms can be provided, caution seems appropriate on the basis of our findings.
Adverse reactions possibly or probably attributable to a study drug were more frequently reported in patients treated with cefepime. We observed a slight excess of renal adverse events in the cefepime group, which was unexpected on the basis of existing data on the safety profile of the drug. The adverse events section of the product insert states that transient elevations in blood urea nitrogen and/or serum creatinine levels are observed in 0.5 to 1% of patients. This adverse event did not result either from an excessive dosage of cefepime or from a baseline imbalance in serum creatinine levels. Two of the five cases of renal adverse events reported in our study could also have been related to other predisposing factors. However, a study with an open-label design primarily aimed at a comparison of efficacy is not appropriate for investigation of this problem. We therefore recommend that this issue be addressed in future investigations of the drug. Treatment with imipenem-cilastatin, but not treatment with cefepime, was associated with symptomatic C. difficile infections in 4% of the patients. This difference may result from a more important alteration of the anaerobic flora with imipenem-cilastatin than with cefepime.
The absence of a prospectively required microbiological investigation at randomization and during follow-up was a limitation of the present study, resulting in a lack of a precise denominator for the results of follow-up cultures. However, the use of a clinical, poorly specific definition of nosocomial pneumonia mimics the real clinical situation, in which decisions must be made on an empirical basis (27), and several recent studies have stressed the importance of appropriate initial antibiotic coverage (1, 18, 22, 25). The two study drugs have shown comparable efficacies in this setting.
In conclusion, our study suggested but could not prove the similar clinical efficacies of cefepime and imipenem-cilastatin for monotherapy of nosocomial pneumonia in ICU patients. Therapy with imipenem-cilastatin, but not with cefepime, was complicated by C. difficile infection. The moderate renal failure associated with cefepime treatment in some patients deserves further consideration. Other observations from this study that require further evaluation include the higher rate of clinical failures in patients infected with an ESBL-producing organism in the cefepime group and the higher rate of resistant P. aeruginosa strains in the imipenem-cilastatin group.
Acknowledgments
This work was supported by Bristol-Myers Squibb, Waterloo, Belgium.
Members of the Cefepime Study Group are as follows: A Mertens, J. Nagler, and P. Rogiers, Antwerp, Belgium; J. C. Preiser, J. P. Thys, and J. L. Vincent, Brussels, Belgium; E. Rubinstein and S. Segev, Tel Hashomer, Israel; W. Hustinx, Utrecht, The Netherlands; L. Krawczyk, Sosnowiec, Poland; J. Jastrzebski and J. Kacki, Warsaw, Poland; A. Grigoriev, Y. Grigoriev, and A. Sinopalnikov, Moscow, Russia; R. Koslov, V. Osipova, L. Strachunsky, and A. Zuzova, Smolensk, Russia; L. Matter, B. Regli, H. U. Rothen, and M. Taüber, Bern, Switzerland; J. Garbino, T. Kinge, D. Lew, and J. A. Romand, Geneva, Switzerland; D. Aymon, F. Bally, J. Bille, R. Chiolero, A. Cometta, M. P. Glauser, G. Greub, M. D. Shaller, and G. Zanetti, Lausanne, Switzerland; and M. Betschart, D. von Ow, and P. Vernazza, St. Gallen, Switzerland.
REFERENCES
1. Alvarez-Lerma, F., et al. 1996. Modification of empiric antibiotic treatment in patients with pneumonia acquired in the intensive care unit. Intensive Care Med. 22:387-394. [PubMed] [Google Scholar]
2. Barckow, D., and C. D. Schwigon. 1993. Cefepime versus cefotaxime in the treatment of lower respiratory tract infections. J. Antimicrob. Chemother. 32(Suppl. B):187-193. [PubMed] [Google Scholar]
3. Beam, T. R., D. N. Gilbert, C. M. Kunin, et al. 1992. General guidelines for the clinical evaluation of anti-infective drug products. Clin. Infect. Dis. 15(Suppl. 1):S5-S32. [PubMed] [Google Scholar]
4. Beam, T. R., D. N. Gilbert, and C. M. Kunin. 1993. European guidelines for anti-infective drug products. Introduction. Clin. Infect. Dis. 17:787-788. [PubMed] [Google Scholar]
5. Chastre, J., J. Y. Fagon, and J. Y. Fagon. 2002. Ventilator-associated pneumonia. Am. J. Respir. Crit. Care Med. 165:867-903. [PubMed] [Google Scholar]
6. Chow, A. W., C. B. Hall, J. O. Klein, R. B. Kammer, R. D. Meyer, J. S. Remington, et al. 1992. Evaluation of new anti-infective drugs for the treatment of respiratory tract infections. Clin. Infect. Dis. 15(Suppl. 1):S62-S88. [PMC free article] [PubMed] [Google Scholar]
7. Cometta, A., J. D. Baumgartner, D. Lew, W. Zimmerli, D. Pittet, P. Chopart, U. Schaad, C. Herter, P. Eggimann, and O. Huber. 1994. Prospective randomized comparison of imipenem monotherapy with imipenem plus netilmicin for treatment of severe infections in nonneutropenic patients. Antimicrob. Agents Chemother. 38:1309-1313. [PMC free article] [PubMed] [Google Scholar]
8. Cook, D. J., S. D. Walter, R. J. Cook, L. E. Griffith, G. H. Guyatt, D. Leasa, R. Z. Jaeschke, and C. Brun-Buisson. 1998. Incidence of and risk factors for ventilator-associated pneumonia in critically ill patients. Ann. Intern. Med. 129:433-440. [PubMed] [Google Scholar]
9. Croce, M. A., T. C. Fabian, R. M. Stewart, F. E. Pritchard, G. Minard, L. Trenthem, and K. A. Kudsk. 1993. Empiric monotherapy versus combination therapy of nosocomial pneumonia in trauma patients. J. Trauma-Injury Infect. Crit. Care 35:303-309. [PubMed] [Google Scholar]
10. Edelstein, H., V. Chirurgi, S. Oster, R. Karp, K. Cassano, S. Aiken, and R. McCabe. 1991. A randomized trial of cefepime (BMY-28142) and ceftazidime for the treatment of pneumonia. J. Antimicrob. Chemother. 28:569-575. [PubMed] [Google Scholar]
11. Fagon, J. Y., J. Chastre, A. J. Hance, P. Montravers, A. Novara, and C. Gibert. 1993. Nosocomial pneumonia in ventilated patients: a cohort study evaluating attributable mortality and hospital stay. Am. J. Med. 94:281-288. [PubMed] [Google Scholar]
12. Fagon, J. Y., J. Chastre, A. Vuagnat, J. L. Trouillet, A. Novara, and C. Gibert. 1996. Nosocomial pneumonia and mortality among patients in intensive care units. JAMA 275:866-869. [PubMed] [Google Scholar]
13. Fink, M. P., D. R. Snydman, M. S. Niederman, K. V. Leeper, Jr., Johnson, R. H., S. O. Heard, R. G. Wunderink, J. W. Caldwell, J. J. Schentag, G. A. Siami, and The Severe Pneumonia Study Group. 1994. Treatment of severe pneumonia in hospitalized patients: results of a multicenter, randomized, double-blind trial comparing intravenous ciprofloxacin with imipenem-cilastatin. Antimicrob. Agents Chemother. 38:547-557. [PMC free article] [PubMed] [Google Scholar]
14. Hartenauer, U., L. S. Weilemann, K. F. Bodmann, W. W. Ritzerfeld, S. Asmus, and E. M. Koch. 1990. Comparative clinical trial of ceftazidime and imipenem/cilastatin in patients with severe nosocomial pneumonias and septicaemias. J. Hosp. Infect. 15(Suppl. A):61-64. [PubMed] [Google Scholar]
15. Jaccard, C., N. Troillet, S. Harbarth, G. Zanetti, D. Aymon, R. Schneider, R. Chiolero, B. Ricou, J. Romand, O. Huber, P. Ambrosetti, G. Praz, D. Lew, J. Bille, M. P. Glauser, and A. Cometta. 1998. Prospective randomized comparison of imipenem-cilastatin and piperacillin-tazobactam in nosocomial pneumonia or peritonitis. Antimicrob. Agents Chemother. 42:2966-2972. [PMC free article] [PubMed] [Google Scholar]
16. Jones, R. N., and S. A. Marshall. 1994. Antimicrobial activity of cefepime tested against Bush group I beta-lactamase-producing strains resistant to ceftazidime. A multilaboratory national and international clinical isolate study. Diagn. Microbiol. Infect. Dis. 19:33-38. [PubMed] [Google Scholar]
17. Kollef, M. H. 1999. Antimicrobial therapy of ventilator-associated pneumonia: how to select an appropriate drug regimen. Chest 115:8-11. [PubMed] [Google Scholar]
18. Kollef, M. H., and S. Ward. 1998. The influence of mini-BAL cultures on patient outcomes: implications for the antibiotic management of ventilator-associated pneumonia. Chest 113:412-420. [PubMed] [Google Scholar]
19. Leophonte, P., A. Bertrand, G. Nouvet, J. F. Muir, F. Lucht, P. Delaval, A. Depierre, F. Hughes, M. Ulmer, and J. J. Gres. 1993. A comparative study of cefepime and ceftazidime in the treatment of community-acquired lower respiratory tract infections. J. Antimicrob. Chemother. 32(Suppl. B):165-173. [PubMed] [Google Scholar]
20. Leu, H. S., D. L. Kaiser, M. Mori, R. F. Woolson, and R. P. Wenzel. 1989. Hospital-acquired pneumonia. Attributable mortality and morbidity. Am. J. Epidemiol. 129:1258-1267. [PubMed] [Google Scholar]
21. Liu, Y. C., W. K. Huang, and D. L. Cheng. 1994. Antibacterial activity of cefepime in vitro. Chemotherapy 40:384-390. [PubMed] [Google Scholar]
22. Luna, C. M., P. Vujacich, M. S. Niederman, C. Vay, C. Gherardi, J. Matera, and E. C. Jolly. 1997. Impact of BAL data on the therapy and outcome of ventilator-associated pneumonia. Chest 111:676-685. [PubMed] [Google Scholar]
23. National Committee for Clinical Laboratory Standards. 1987. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. National Committee for Clinical Laboratory Standards Villanova, Pa.
24. Pugin, J., R. Auckenthaler, N. Mili, J. P. Janssens, P. D. Lew, and P. M. Suter. 1991. Diagnosis of ventilator-associated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic “blind” bronchoalveolar lavage fluid. Am. Rev. Respir. Dis. 143:1121-1129. [PubMed] [Google Scholar]
25. Rello, J., M. Gallego, D. Mariscal, R. Sonora, and J. Valles. 1997. The value of routine microbial investigation in ventilator-associated pneumonia. Am. J. Respir. Crit. Care Med. 156:196-200. [PubMed] [Google Scholar]
26. Sanders, C. C. 1993. Cefepime: the next generation? Clin. Infect. Dis. 17:369-379. [PubMed] [Google Scholar]
27. Timsit, J. F., S. Chevret, J. Valcke, B. Misset, B. Renaud, F. W. Goldstein, P. Vaury, and J. Carlet. 1996. Mortality of nosocomial pneumonia in ventilated patients: influence of diagnostic tools. Am. J. Respir. Crit. Care Med. 154:116-123. [PubMed] [Google Scholar]
28. Torres, A., R. Aznar, J. M. Gatell, P. Jimenez, J. Gonzalez, A. Ferrer, R. Celis, and R. Rodriguez-Roisin. 1990. Incidence, risk, and prognosis factors of nosocomial pneumonia in mechanically ventilated patients. Am. Rev. Respir. Dis. 142:523-528. [PubMed] [Google Scholar]
29. Troillet, N., M. H. Samore, and Y. Carmeli. 1997. Imipenem-resistant Pseudomonas aeruginosa: risk factors and antibiotic susceptibility patterns. Clin. Infect. Dis. 25:1094-1098. [PubMed] [Google Scholar]
30. Vincent, J. L., D. J. Bihari, P. M. Suter, H. A. Bruining, J. White, M. H. Nicolas-Chanoin, M. Wolff, R. C. Spencer, M. Hemmer, et al. 1995. The prevalence of nosocomial infection in intensive care units in Europe. Results of the European Prevalence of Infection in Intensive Care (EPIC) Study. JAMA 274:639-644. [PubMed] [Google Scholar]
Articles from Antimicrobial Agents and Chemotherapy are provided here courtesy of American Society for Microbiology (ASM)