Cs-99/9_umbr 21.09.2000 8:51 Uhr Seite 318
(Schwarz/Process Black Bogen)
Brief Reports
Cefminox versus Metronidazole plus Gentamicin in Intra-abdominal Infections: A Prospective Randomized Controlled Clinical Trial A.J. Torres, L. Díez Valladares, J.M. Jover, A. Sánchez -Pernaute, J. Frías, A.J. Carcas, P. Coronel, E. Ródenas, I. Pérez-Balcabao, R. Fernández-Roblas, M. Moreno, J.L. Balibrea
Summary Background: The aim of this prospective study was to compare the safety and efficacy of a new cephamycin, cefminox 2 g/12 h, to those of the usual regimen combining metronidazole 500 mg/8 h and gentamicin 80 mg/8 h (M+G). Patients and Methods: 160 patients with clinically proven intra-abdominal infection were prospectively included in an open parallel randomized comparative multicenter trial. Antibiotics were started preoperatively and discontinued after clinical and laboratory evidence of resolution of the infection. Serum and peritoneal fluid levels and serum bactericidal activities were also studied. Results: 150 patients were clinically evaluable. There was one failure in the cefminox group and three in the M+G group (not significant, RR: 1.07, 95% CI: 1–1.15). No differences were found in the number of wound infections, length of stay or duration of antibiotic therapy. Adverse effects were reported in 11 cases, all of them mild to moderate. Escherichia coli and Bacteroides fragilis were the most frequently found microorganisms. Conclusion: Cefminox is as effective and as safe as M+G in the treatment of intra-abdominal infections.
Key Words Cefminox · Intra-abdominal infections · Clinical trial
having a similar spectrum of activity to the previously mentioned combination, but with fewer side effects. Cefminox, a new cephamycin antibiotic with potent bactericidal activity [8–11], is characterized by a second lithic target of binding which differs from that of the classic penicillin binding proteins (PBPs). Its antimicrobial spectrum covers both important aerobic and anaerobic organisms [12]. Cefminox has a half-life of 2.35 h, and achieves peak serum levels of 117 mg/l following administration of a 2 g dose [13]. It penetrates extensively into retroperitoneal pelvic exudate and large areas under the curve of bactericidal powers are obtained against reference strains [13, 14].Therefore, cefminox is an antibiotic with antibacterial activity, pharmacokinetic profile and in vivo efficacy, which makes it a suitable candidate for the treatment of intra-abdominal infections [8, 15]. The purpose of this study was to compare the clinical and microbiological efficacy and the safety of cefminox with those of metronidazole and gentamicin (M+G) in the empirical treatment of intra-abdominal infections.
Patients and Methods Study Design Two surgical departments participated in the study which was designed as an open parallel randomized prospective trial. 160 patients over 18 years of age with signs and symptoms of intra-abdominal infection were consecutively included in the trial. Both surgery and percutaneous drainage under radiological control were classified as
Infection 2000; 28: 318–322
Introduction The efficacy of several antibiotic regimens in the treatment of intra-abdominal infections is well established [1–4]. The choice of one treatment or another is based principally on patient factors. The combined regimens of an anaerobicide with an aminoglycoside have been used and are effective for empirical treatment of these patients. The aminoglycosides show a series of toxic effects especially in elderly, renal failure and shock patients [5–7]. This is the current reason for the tendency to use treatment regimens based on an combinations with fewer toxic effects or a single antibiotic
318
A.J. Torres (corresponding author), L. Díez Valladares, A. SánchezPernaute, J.L. Balibrea Dept. of Surgery II, Clínico San Carlos Hospital, Complutense University, C/ Martín Lagos, s/n, E-28040 Madrid, Spain; Phone: (+34/91) 3303-444, Fax: -3179, e-mail:
[email protected] J.M. Jover, M. Moreno Dept. of Surgery, Getafe University Hospital, Madrid, Spain J. Frías, A.J. Carcas Clinical Pharmacology Dept., La Paz University Hospital, Madrid, Spain P. Coronel, E. Ródenas R & D Dept., Tedec-Meiji Farma, S.A., Madrid, Spain I. Pérez-Balcabao Microbiology Dept., Gómez Ulla Hospital, Madrid, Spain R. Fernández-Roblas Microbiology Dept., Fundación Jiménez Díaz Hospital, Madrid, Spain. Received: March 29, 1999 • Revision accepted: July 1, 2000
Infection 28 · 2000 · No. 5 © URBAN & VOGEL
Cs-99/9_umbr_kopie 21.09.2000 8:53 Uhr Seite 319
(Schwarz/Process Black Bogen)
A.J. Torres et al.: Cefminox in Intra-abdominal Infections
surgical procedures. The study was previously reviewed and approved by the Clinical Trials Committee at each participating center, and authorized by the Spanish Department of Health. All patients gave their informed consent before being enrolled in the study. Patients were excluded from the study according to the following criteria: known or suspected allergy to beta-lactam antibiotics, gentamicin or metronidazole; prior antibiotic therapy within 72 h of the start of the study, or 1 month in the case of slow-release drugs; platelet count < 100,000/mm3; pregnancy or lactation; treatment with an investigational drug within 30 days prior to the start of the study; hemodialysis or immunosuppresive therapy; clinical symptoms suggestive of acute mild abdominal infection without perforation; probability of death greater than 90% in the hours following or Acute Physiology and Chronic Health Evaluation (APACHE) II score ≥ 35 [16]; serum creatinine level > 2.5 mg/dl; cirrhosis or ascites; presence of an extra-abdominal septic focus. All patients were allocated blindly using sealed envelopes to one of the treatment groups: cefminox (Tedec-Meiji Farma, S.A.) 2 g/12 h or metronidazole (Rhône-Poulenc Rorer, S.A.) 500 mg/8 h plus gentamicin (Shering-Plough, S.A.) 80 mg/8 h. Prior to administration of the first dose of antibiotic and during the followup period blood samples were taken for culture and for laboratory tests. In addition, intraoperative samples of peritoneal exudate and blood were taken for microbiological analysis. Culture of wound exudate was performed if wound infection was suspected. Blood samples from 79 patients were also taken for determining the peak and trough levels of antibiotics. In addition, a sample of peritoneal fluid was taken on the day of surgery. The treatment response was assigned as cure, failure or undetermined according to the Solomkin criteria [17].
Sample Analysis
Results and Discussion 160 patients were enrolled in the study: 69 women and 91 men with a mean age of 46.6 years. The distribution of patients by diagnoses and baseline characteristics is shown in table 1. There were no differences between treatment groups in this respect (p > 0.05). The most common diagnoses were acute perforated appendicitis, followed by gastroduodenal perforation and acute perforated cholecystitis. In 68.1% of patients treatment was for localized peritonitis and in the remaining 31.9% for diffuse peritonitis. Every patient was submitted to a complete surgical treatment, including drainage of the peritoneal effusion and specific treatment of the primary process, being resective or not. No significant statistical differences in the pathology causing the infection and in the patient conditions defined by the APACHE II system between the treatment groups were observed. Of the 160 randomized patients, 152 were clinically evaluable for the per protocol analysis. There were eight non-evaluable patients, four in the cefminox group and four in the M+G group. Among the 76 clinically evaluable patients in the cefminox group, 75 were cured (98.6%) and only one case was considered as a failure. This was a patient with diffuse peritonitis caused by colon perforation. The antibiotic treatment had to be changed due to a delayed response on the 5th day; Pseudomonas aeruginosa and Enterococcus spp. were isolated from peritoneal fluid. Among 76 patients in the M+G group, 70 patients were cured (92.11%): in three patients the response was considered as undetermined and three cases were failures.The first failure was a patient with diffuse peritonitis by peptic perforated duodenal ulcer. It was necessary to change the antibiotic treatment due to a delayed response to therapy on the 5th day of treatment. The peritoneal fluid culture was negative. The second patient showed diffuse peritonitis by colon perforation, and Clostridium spp. and E. coli were isolated from peritoneal fluid; there was no therapeutical response on the 5th day of treatment. The third
Cefminox concentrations in sera were determined by reverse phase high-pressure liquid chromatography [18]. To determine cefminox concentrations in peritoneal fluid, the samples were first centrifuged and the supernatant was then analzed in the same way as serum. Levels of gentamicin in serum and peritoneal fluid were determined using a fluorescent polarization immunoassay with an automatic analyzer TDX|FLX (Abbott Diagnostic Division, North Chicago, IL). Metronidazole concentrations were measured using a microbiological method according to Levinson [19]. In a subset of these patients (14 patients treated with cefminox and 11 patients treated with M+G) serum samples (intraoperative and peak and trough at steady state) were tested for bactericidal activity. Serum bactericidal titers were determined Table 1 against Escherichia coli ATCC 25922 and Bacteroides fragDiagnosis and baseline characteristics in patients participating in the trial. ilis ATCC 25285 using previously described methods [20]. No. of patients treated with Diagnosis
Cefminox (%)
Appendicitis
42 (52.5)
Cholecystitis
12 (15)
Gastric perforation
15 (18.8)
Statistical Analysis Student’s t test was used for intergroup comparisons in the case of quantitative variables with a normal curve fit, and the Mann-Whitney test for non-gaussian variables. The Chi-square test or Fisher’s exact test were used for qualitative variables.The main variable for analysis was the clinical cure rate. All computations and analyses were carried out using the SAS statistics program (version 6.10). Statistical analysis of the clinical cure rate was performed by the Chi-square test with the continuity correction of Yates using the program Epi Info 6 (version 6.04 a). The relationship between serum bactericidal titers and drug concentrations was analyzed by Spearman’s rank correlation.
Infection 28 · 2000 · No. 5 © URBAN & VOGEL
Metronidazole+ gentamicin (%)
Total (%)
p
53 (66.3)
95 (59.4)
3 (3.8)
15 ( 9.4)
0.027
32 (20)
0.844
17 (21.3)
0.107
Intestinal perforation
5 (6.3)
2 (2.5)
7 (4.4)
0.443
Colon perforation
5 (6.3)
4 (5)
9 (5.6)
1.000
Other
1 (1.3)
1 (1.3)
2 (1.3)
1.000
APACHE II score Mean (SD)
3.68 (3.45)
3.18 (2.65)
3.43 (3.08)
319
Cs-99/9_umbr_kopie 21.09.2000 8:53 Uhr Seite 320
(Schwarz/Process Black Bogen)
A.J. Torres et al.: Cefminox in Intra-abdominal Infections
Table 2 Pathogens isolated from microbiologically evaluable patients.
Cefminox
Metronidazole + gentamicin
Gram-negative aerobes E. coli Klebsiella spp. Enterobacter spp. Proteus spp. Citrobacter spp. Serratia spp. Aeromonas spp. P. aeruginosa
16 3 1 1 0 0 1 1
19 6 1 3 1 1 1 2
4 8 20
0 4 9
2
2
Peptostreptococcus Clostridium spp. Bacteroides fragilis Other Bacteroides
2 3 10 4
0 3 4 1
Total
76
57
Gram-positive aerobes Staphylococcus coagulase negative Enterococcus spp. Streptococcus spp. Yeasts Candida albicans Anaerobes
case required further surgery due to focal recurrence of infection. Initially this patient had diffuse peritonitis due to appendicitis; Streptococcus viridans was isolated from peritoneal fluid.
Differences in the rate of clinical cure between the two treatment regimens following a per protocol analysis did not reach statistical significance; thus, the relative risk of cure rate between the cefminox group and the M+G group was 1.07 with 95% CI of 1.00–1.15 favoring cefminox. In the intention to treat analysis, all randomized patients were included. In this analysis a favorable response was considered for cured patients only; unfavorable response grouped patients were classified as failure, non-evaluable and undetermined. The statistical evaluation revealed no significant differences in the outcome between the two treatment groups (93.8% in the cefminox group versus 87.5% in the M+G group: Chi-square test; p = 0.28). Relative risk was also 1.07 with a 95% confidence interval of 0.97–1.18. Three patients in the M+G group required further surgery for the following reasons: evisceration, adenocarcinoma of appendix and intra-abdominal abscess. The latter was evaluated as a clinical failure. There were eight (10.53%) postoperative wound infections in the cefminox group and 14 (18.42%) in the M+G group. No significant differences were found between the two treatment groups (Fisher’s test p = 0.248). The mean length of hospital stay and the mean duration of antibiotic therapy were 7.9 ± 5.2 days and 4.9 ± 1.9 days for localized peritonitis and 9.5 ± 4.1 days and 6.3 ± 2.8 days for diffuse peritonitis, respectively. The clinical results obtained in this study show that there are no statistically significant differences in treatment efficacy when either cefminox or M+G are used. Clinical success rates were high in both groups; this may be explained not only by the high efficacy of the antibiotics and the characteristics of the patient population, but also by the large number of surgical operations in which localized peritonitis was treated (two-thirds).
Table 3 Relationship between serum concentration of antibiotics and serum bactericidal titer.
Concentration (mg/l)
Serum bactericidal titer E. coli
Antibiotic (no. of samples)
Time
Range
Median
Range
Cefminox (n=14)
Peak Trough Intraoperative
74.4 –302.8 1.75– 52.4 16.9 –160.4
139.9 10.9 68.6
128–2048 2–256 32– > 4096
Metronidazole (n=11)
Peak Trough Intraoperative
7.4 – 28.1 0.8 – 19.3 6.9 – 21
18.6 8.7 12.8
Gentamicin (n=11)
Peak Trough Intraoperative
1.7 – 6.6 0.2 – 3.8 0.4 – 3.3
3.6 0.5 2.3
B. fragilis Median 256 16 128
ND ND ND 2–16 < 2–8 < 2–8
Range
Median
32–1024 < 2–64 32–512
128 8 128
8–32 8–32 16–64 8
