Spread of Extended-Spectrum b-lactamase Producing Klebsiella pneumoniae: Are b-lactamase Inhibitors of Therapeutic Value?

76 Spread of Extended-Spectrum b-lactamase Producing Klebsiella pneumoniae: Are b-lactamase Inhibitors of Therapeutic Value? Lionel Piroth, Hervé Aubé, Jean-Marc Doise, and Michel Vincent-Martin From Service de Réanimation Médicale, Centre Hospitalo-Universitaire de Dijon, Dijon, France Because of recurrent colonization by Klebsiella pneumoniae strains producing type SHV-4 extended-spectrum b-lactamases (ESBLs), a case-control study was conducted in an intensive care unit to investigate the risk of acquisition, with special reference to antibiotic therapy and resuscitation procedures. Fifty-one patients colonized or infected by ESBL-producing K. pneumoniae (cases) were matched with 51 noncolonized patients (controls). Duration of intubation was significantly longer for cases than for controls, while duration of b-lactamase inhibitor therapy was significantly shorter. By means of multivariate analysis, intubation was the only risk factor identified (odds ratio [OR] Å 1.19), while b-lactamase inhibitor therapy was shown to be a protective factor (OR Å 0.849). During outbreaks of SHV-4 type ESBL producing K. pneumoniae in intensive care units, preferential use of b-lactamase inhibitors may help control the emergence and spread of these pathogens even if essential hand washing and isolation procedures are adhered to. variate analysis only [2]. Most of these studies [17 19] were not designed to standardize the main risk factors for nosocomial infections as duration of stay or severity of disease. Nonetheless, it is of particular interest to know which drugs are poten- tially associated with acquisition of ESBL-producing K. pneumoniae to determine preferential empirical treatment. A case-control study was conducted in our ICU to investigate the risk factors for transmission of ESBL-producing K. pneu- moniae with special reference to antibiotic therapy and technical procedures; the study was designed to standardize known risk factors such as duration of exposure and severity of disease was also investigated. Nosocomial infection due to enterobacteria producing extended-spectrum b-lactamases (ESBLs), particularly Klebsiella species, has become more prevalent. These infections are of particular concern in intensive care units (ICUs) but also in other wards such as geriatric wards and nursing homes where stays are medium and long term [1]. Nosocomial infections with ESBL-producing Klebsiella pneumoniae are difficult to eradicate because these organisms develop resistance to multiple antibiotics. Effective control can be achieved only through more thorough understanding of the transmission process. See editorial response by Jacoby on pages 81 3. The study was carried out at the medical ICU of the Dijon University Hospital (Dijon, France), an 11-bed unit with an average annual admittance of Ç550 adult patients. An active control program was introduced in December 1993, including systematic identification (by culturing of stool specimens ob- tained at admission and at least once a week) of colonization with ESBL-producing K. pneumoniae and isolation of all pa- tients with infection or colonization until their discharge from the ICU. In the study by Schiappa et al. [2], one risk factor for acquiring ESBL-producing K. pneumoniae was duration of stay in the hospital. Equipment such as artificial ventilation apparatuses, catheters, and cannulas [2 6] and nursing care work load [2] were also suggested as likely risk factors. In many studies [5, 7 14], the impact of antibiotics on acquiring ESBL-producing K. pneumoniae was questioned on the basis of historical or indirect arguments: etiologic investigations that included antibiotic administration as a risk factor gave conflicting results. In some studies, antibiotic therapy and colonization by ESBL-producing K. pneumoniae were not significantly associated [15]. In other studies, colonization was demonstrated to be associated with antibiotic therapy with no precision of antibiotic class [3, 6, 16], with sulfonamides and fluoroquinolones [4], or with cephalosporins by means of uni- Materials and Methods Setting Bacteriologic Data In every suspected case of infection or colonization, multiple routine bacteriologic specimens were tested for Klebsiella by Received 14 October 1997; revised 16 March 1998. means of the API System (biomérieux, Marcy l Etoile, France). Reprints or correspondence: Dr. Hervé Aubé, Service de Réanimation Médi- Susceptibility of each K. pneumoniae strain to antibiotics was cale, Hôpital Général, B.P. 1519, 21033 Dijon Cedex, France. tested by using the agar disk diffusion method, the results Clinical Infectious Diseases 1998;27:76 80 of which were interpreted according to French recommended 1998 by the Infectious Diseases Society of America. All rights reserved. 1058 4838/98/2701 0016$03.00 standards [20]. The double-disk synergy test described by Le-

CID 1998;27 (July) b-lactamase Inhibitors for ESBL-Producing K. pneumoniae 77 grand et al. [21] was used to detect ESBL-producing strains. Strains of ESBL-producing K. pneumoniae were also identified by pulsed field gel electrophoresis. ESBLs were typed by comparing their isoelectric points with those of reference ESBLs. Clinical Data All medical records for every patient admitted to the ICU from 1 January 1994 to 1 September 1996 were examined. Gender, age, admission route, simplified acute physiological score (SAPS) at the time of admission, presentation disorder(s), duration of stay, microorganisms identified in bacteriologic specimens and site and date of isolation, technical procedures, number of days of treatment, and course of the disease were all recorded. Results were recorded in terms of percentage for qualitative variables and mean { SD for quantitative variables. With use of the standardized mortality ratio (SMR), the mortality rate among patients infected with ESBL-producing K. pneumoniae was compared with the expected rate among the ICU population as a whole (reference population) following indirect standardization. Univariate analysis was performed by using Wilcoxon s testing of variables for matched pairs to compare and identify factors significantly associated with development of infection or colonization with ESBL-producing K. pneumoniae. Multivariate analysis was carried out by means of logistic regression of variables for matched pairs with backward stepping. The final model included all matching variables. Results Cases and Controls A patient from whom ESBL-producing K. pneumoniae was Descriptive Epidemiology isolated at least once during the ICU stay, whatever the isola- From 1 January 1994 to 1 September 1996, 1,533 patients tion site and the inflammatory response, was defined as a case. were admitted to the medical ICU of Dijon University Hospital. Colonized and infected patients were included in the study. Of these patients, 54 (3.5%) were ESBL-producing K. pneumoniae Cases were regarded as community acquired when evidence carriers. The mean age { SD of these 54 patients was of ESBL-producing K. pneumoniae was demonstrated within 66 { 14 years; 63% were men; and the average SAPS { SD the first 48 hours of admission to the ICU. After this time, was 15.1 { 4. The mean duration of stay in the ward { SD the case was considered as hospital acquired. Only hospital- was 23 { 27 days. One of the 54 carriers had a communityacquired acquired cases were included in the study. case; the remaining 53 had hospital-acquired cases. Every case was matched with a control in terms of age ({4 The average time for acquisition was 15.8 days (range, 3 87 years), gender, SAPS ({3), presentation disorder(s) (including days). presence or absence of infection), and length of exposure. K. pneumoniae represented 8.7% of the total number of Length of exposure for controls was defined as a period that bacterial strains isolated, irrespective of site. Of these strains, was at least equal to the period before contamination for cases. 55% produced ESBLs. ESBL-producing K. pneumoniae was For controls, absence of ESBL-producing K. pneumoniae was isolated from the following specimens from the 54 carriers: demonstrated by bacteriologic specimens throughout their stay feces, 42 patients (78%); urine, 12 (22%); bronchi, 11 (20%); in the ward. and blood or other sites, 4 (7%). Three different antibiotypes Controls were selected by an independent observer on the were demonstrated. The most frequent antibiotype (60% of sole basis of matching criteria, other variables being unknown. strains) showed resistance to b-lactam agents (with the excep- Should several potential controls meet the matching criteria, tion of cephamycin and imipenem), resistance to aminoglycosides that control whose admission date was closest to that of the (except for gentamicin), and resistance to quinolones and case was selected. When no suitable control met the matching fluoroquinolones, sulfonamides, and chloramphenicol. Only criteria, the case was excluded from the study. one ESBL-producing strain was identified by pulsed field gel electropheresis throughout the ward. The ESBL was SHV-4. Risk Factors Of the 1,533 inpatients, 49% received antibiotic therapy during their stay in the ICU (64% of 872 patients who stayed for Risk factors such as number of days of use of third-generation ú48 hours). The most frequently used antibiotic classes in the cephalosporins, penicillins, b-lactamase inhibitors, amino- ICU during the study period were penicillins (alone in 15% of glycosides, fluoroquinolones, deep-venous catheters, indwelling cases or in combination with b-lactamase inhibitors in 19%), urinary catheters, and endotracheal tubes were investigated. third-generation cephalosporins (22%), aminoglycosides (19%), and systemic fluoroquinolones (16%). Of the 53 patients Data Analysis who had hospital-acquired cases, 84% had received antibiotic treatment prior to colonization. Statistical analysis was performed by using Epi-Info version The crude mortality rate was 32% among patients colonized 5.1 (Centers for Disease Control and Prevention, Atlanta) and or infected with ESBL-producing K. pneumoniae and 26% BMDP (BMDP, West Los Angeles). among noninfected patients. Following indirect standardization

78 Piroth et al. CID 199827 (July) Table 1. General characteristics of cases and controls in a study Table 3. Results of multivariate analysis with logistic regression in of b-lactamase inhibitor therapy and infection or colonization with a case-control study of b-lactamase inhibitor therapy and infection extended-spectrum b-lactamase producing Klebsiella pneumoniae. or colonization with extended-spectrum b-lactamase producing Klebsiella pneumoniae. Characteristic Cases Controls P value Variable Coefficient { SE OR P value No. of patients 51 51 Mean age (y) { SD 65.8 { 13.4 65.9 { 13.8.86 b-lactamase inhibitor (/1 d) 00.1632 { 0.0807 0.849.04 No. of males/no. of females 32/19 32/19 1.00 Intubation (/1 d) 0.1768 { 0.0775 1.19.03 Mean SAPS { SD 15.5 { 4.8 15.0 { 4.5.16 No. (%) with infectious NOTE. /1 då augmentation of the coefficient for 1 day of treatment. disease at admission 12 (24) 12 (24) 1.00 Mean length of exposure (d) { SD 13.1 { 9.5 12.6 { 8.3.09 No. (%) treated with By means of multivariate analysis with logistic regression, antibiotics 43 (84) 45 (88).57 the significant factors were intubation (risk factor) and previous use of b-lactamase inhibitors (protective factor) (table 3). No NOTE. SAPS Å simplified acute physiological score. interaction was found among use of b-lactamase inhibitors, use of third-generation cephalosporins, and intubation. of SAPS and duration of stay, the SMR was not statistically significant. Discussion Epidemiological data (especially bacteriologic data) re- Case-Control Study corded in our ward during the period 1 January 1994 to 1 There were 51 cases. Apart from the patient with a commusame September 1996 revealed major characteristics that were the nity-acquired case, two other patients colonized by ESBL-probreaks as those previously reported following recurrent out- ducing K. pneumoniae were excluded because there was no of ESBL-producing K. pneumoniae since 1983: namely, suitable control who met the matching criteria. Fifty-one of resistance to third-generation cephalosporins, aminoglycosides, 186 potential controls were selected for the study. The general sulfonamides, and chloramphenicol [10, 12, 22]; type of encharacteristics of these groups are shown in table 1. zyme (SHV-4, which is commonly isolated in France) [5, 10, By means of univariate analysis (table 2), only the intubation 23 25]; main isolation sites [5, 9 12, 15, 22, 26, 27]; and period was significantly longer for cases than for controls, average contamination period [4, 5, 9, 11, 26]. An increase whereas the duration of b-lactamase inhibitor therapy was sigproducing in the mortality rate associated with infection due to ESBL- nificantly shorter. The b-lactamase inhibitors used by either K. pneumoniae was reported previously [2], but this group were clavulanic acid (80% of patients) and tazobactam increase was not documented in this study following standard- (20%). Agents used in association with b-lactamase inhibitors ization because it was attributable to the severity of underlying were amoxicillin (76% of patients) and piperacillin or ticarcillin diseases. (24%). Third-generation cephalosporins were never used in Intubation was the major risk factor in the spread of ESBLcombination with b-lactamase inhibitors. producing K. pneumoniae, implying that colonized patients be- Table 2. Results of univariate analysis in a case-control study of long to the most critically ill ICU population. This factor is independent from the type of antibiotic therapy given. Mouth care and enteral feeding of intubated patients are likely to b-lactamase inhibitor therapy and infection or colonization with ex- increase the risk of cross-infection by contact transfer. This tended-spectrum b-lactamase producing Klebsiella pneumoniae. observation points out the importance of hand transmission of ESBL-producing K. pneumoniae and therefore hand washing Mean no. of days in the critical care setting. of use { SD Whether patients were intubated, previous use of b-lacta- Variable Cases Controls P value mase inhibitors was a protective factor against acquisition of SHV-4 type ESBL producing K. pneumoniae. Penicillins were Penicillins 4.6 { 4.7 5.9 { 5.1.15 always given in combination with b-lactamase inhibitors in Third-generation cephalosporins 1.4 { 3.0 1.4 { 2.8 1.00 our study, but these agents had no protective effect as shown Aminoglycosides 1.4 { 2.6 2.0 { 2.9.20 Fluoroquinolones 1.7 { 3.3 1.5 { 3.1.80 in the univariate and multivariate analyses. To our knowledge, b-lactamase inhibitors 3.0 { 3.7 4.5 { 4.1.03 this specific role of b-lactamase inhibitors has not been re- Intubation 11.4 { 9.0 9.1 { 8.8.006 ported in any etiologic study, since these agents are seldom Catheter 9.9 { 8.5 9.9 { 7.5.95 analyzed on their own. It is not however surprising that patients Urinary catheter 8.1 { 7.4 6.3 { 6.5.08 receiving an active drug would not be infected with the suscep-

CID 1998;27 (July) b-lactamase Inhibitors for ESBL-Producing K. pneumoniae 79 tible pathogen. By inhibiting ESBLs, b-lactamase inhibitors association with penicillins may help control outbreaks of appear to impair the emergence and spread of Klebsiella carrying ESBL-producing K. pneumoniae, in particular SHV-4 type resistance plasmids [28, 29]. Furthermore, administration ESBL producing K. pneumoniae. In the absence of an out- of b-lactamase inhibitors and penicillin may exert in vitro pressure break, the risk for the emergence of resistant pathogens seems on ESBLs, thereby facilitating their reverse mutation into to be far higher than the expected benefit to recommend the less harmful enzymes [30]. On the other hand, some clinical systematic use of b-lactamase inhibitors in empirical treat- isolates of ESBL-producing K. pneumoniae were found to be ments. The results of this study call for further investigations less susceptible to b-lactam b-lactamase inhibitor combinations such as multicenter and prospective trials. because of hyperproduction of ESBLs [31, 32], but the role of previous b-lactamase inhibitor treatment in this emerging resistance has not yet been established. There is also risk for References a selection effect on gram-negative rods producing plasmidic or 1. Jacoby GA, Medeiros AA. More extended-spectrum b-lactamases. Antimichromosomic AmpC type b-lactamases [28]. crob Agents Chemother 1991;35:1697 704. Findings of this study must be put into perspective because 2. Schiappa DA, Hayden MK, Matushek MG, et al. Ceftazidime-resistant of several limitations. First, only one clonal strain of ESBLcase-control Klebsiella pneumoniae and Escherichia coli bloodstream infection: a and molecular epidemiologic investigation. J Infect Dis producing K. pneumoniae that produced a single type of ESBL 1996;174:529 36. (SHV-4) was observed and investigated. It is not possible to 3. Lucet JC, Chevret S, Vanjak D, et al. Risk factors for acquiring extendedstate if the conclusion is valid for other types of ESBL-produc- spectrum beta-lactamases Enterobacteriaceae (ESBLE) in ICU [abstract ing K. pneumoniae or ESBLs. Second, the small population no 637]. In: Program and abstracts of the 32nd Interscience Conference size generated a lack of power, especially regarding the use of on Antimicrobial Agents and Chemotherapy (Anaheim, CA). Washing- ton, DC: American Society for Microbiology, 1992:217. third-generation cephalosporins, aminoglycosides, and fluoro- 4. De Champs C, Rouby D, Guelon D, et al. A case-control study of an quinolones. Last, cases included patients who acquired SHVoutbreak of infections caused by Klebsiella pneumoniae strains produc- 4 type ESBL producing K. pneumoniae following either alter- ing CTX-1 (TEM-3) beta-lactamase. J Hosp Infect 1991;18:5 13. ation of the endogenous flora or patients colonized via exogeinfections 5. Arlet G, Sanson Le Pors MJ, Rouveau M, et al. Outbreak of nosocomial nous routes. Since mechanisms of acquisition are different, risk due to Klebsiella pneumoniae producing SHV-4 beta-lacta- mase. Eur J Clin Microbiol Infect Dis 1990;9:797 803. factors are likely to be different too. However, one would be 6. Pena C, Pujol M, Ricart A, et al. Risk factors for faecal carriage of inclined to think that antibiotic therapy prior to acquisition Klebsiella pneumoniae producing extended-spectrum beta-lactamase alters the modes of both endogenous development and exoge- (ESBL-KP) in the intensive care unit. J Hosp Infect 1997;35:9 16. nous spread of infection due to ESBL-producing K. pneumon- 7. De Champs C, Sauvant MP, Chanal C, Sirot D, Gazuy N, Malhuret R. iae. Furthermore, since presentation disorders, severity of disspectrum-b-lactamase-producing Prospective survey of colonisation and infection caused by expandedease, members of the family Enterobacteri- and length of exposure were fairly similar in both groups, aceae in an intensive care unit. J Clin Microbiol 1989;27:2887 90. it could be supposed that nursing care work load was similar 8. Meyer KS, Urban C, Eagan JA, Berger BJ, Rahal JJ. Nosocomial outbreak too and that exposure to cross-infection through contact transfer of Klebsiella infection resistant to late-generation cephalosporins. Ann was equal in each matched case-control pair. Intern Med 1993;119:353 8. 9. Philippon A, Ben Redjeb S, Fournier G, Ben Hassen A. Epidemiology of extended-spectrum b-lactamases. Infection 1989;17:347 54. Conclusion 10. Brun-Buisson C, Legrand P, Philippon A, Legrand P, Montravers F, Duval This study suggests that b-lactamase inhibitors may have a J. Transferable enzymatic resistance to third-generation cephalosporins during nosocomial outbreak of multiresistant Klebsiella pneumoniae. protective effect against acquiring SHV-4 type ESBL produc- Lancet 1987;2:302 6. ing K. pneumoniae, especially in intubated patients in ICUs. 11. Sirot J, Chanal C, Petit A, Sirot D, Labia R, Gerbaud G. Klebsiella pneu- moniae and other Enterobacteriaceae producing novel plasmid-mediated The therapeutic value of b-lactamase inhibitors must be pointed b-lactamases markedly active against third-generation cephalosporins: out not only for the management of infection or colonization epidemiologic studies. Rev Infect Dis 1988;10:850 9. with ESBL-producing K. pneumoniae [28, 33 36] but also for 12. Naumovski L, Quinn JP, Miyashiro D, et al. Outbreak of ceftazidime conventional antibiotic therapy. It is of particular interest that resistance due to a novel extended-spectrum b-lactamase in isolates some investigators [14] recently used this approach (with piper- from cancer patients. Antimicrob Agents Chemother 1992;36:1991 6. 13. Quinn JP. Clinical significance of extended-spectrum beta-lactamases. Eur acillin and tazobactam rather than ceftazidime) in the empirical J Clin Microbiol Infect Dis 1994;13(suppl 1):S39 42. treatment of nosocomial infections. Such an approach resulted 14. Rice LB, Eckstein EC, De Vente J, Shlaes DM. Ceftazidime-resistant in a significant decrease in colonizations with K. pneumoniae Klebsiella pneumoniae isolates recovered at the Cleveland Department (producing SHV-4 and TEM-derived ESBLs) without in- of Veterans Affairs Medical Center. Clin Infect Dis 1996;23:118 24. creased resistance to b-lactamase inhibitors in the clinical iso- 15. Venezia RA, Scarano FJ, Preston KE, et al. Molecular epidemiology of an SHV-5 extended-spectrum b-lactamase in Enterobacteriaceae isolated lates. from infants in an neonatal intensive care unit. Clin Infect Dis 1995; Even if essential hand washing and isolation procedures are 21:915 23. adhered to for preventing and limiting nosocomial infections, 16. Hibberd PL, Jacoby GA. Multiple drug resistant Klebsiella pneumoniae we conclude that preferential use of b-lactamase inhibitors in (MDRKP) strains: predictors of acquisition and mortality [abstract no

80 Piroth et al. CID 199827 (July) C46]. In: Program and abstracts of the 34th Interscience Conference on 3-year survey in France. Antimicrob Agents Chemother 1992;36:1677 Antimicrobial Agents and Chemotherapy (Orlando, FL). Washington, 81. DC: American Society for Microbiology, 1994:87. 28. Livermore DM. b-lactamases in laboratory and clinical resistance. Clin 17. Freeman J, Mc Gowan JE. Risk factors for nosocomial infection. J Infect Microbiol Rev 1995;8:557 84. Dis 1978;138:811 9. 29. Goldstein FW, Pean Y, Rosato A, Gertner J, Gutmann L, the Vigil Roc 18. Maki DG. Risk factors for nosocomial infection in intensive care. Arch Study Group. Characterization of ceftriaxone-resistant Enterobacteria- Intern Med 1989;149:30 3. ceae: a multicentre study in 26 French hospitals. J Antimicrob Chemoter 19. French GL, Cheng AFB, Wong SL, Donnan S. Repeated prevalence sur- 1993;32:595 603. veys for monitoring effectiveness of hospital infection control. Lancet 30. Du Bois SK, Mariott MS, Amyes SGB. TEM- and SHV-derived extended- 1989;2:1021 3. spectrum b-lactamases: relationship between selection, structure and 20. Acar J, Bergogne-Berezin E, Chabbert Y, et al. Communiqué 1992 du function. J Antimicrob Chemother 1995;35:7 22. Comité de l Antibiogramme de la société Française de Microbiologie. 31. Rice LB, Carias LL, Bonomo RA, Shlaes DM. Molecular genetics of Pathol Biol (Paris) 1992;40:741 8. resistance to both ceftazidime and b-lactam b-lactamase inhibitor com- 21. Legrand P, Fournie G, Bure A, et al. Detection of extended broad-spectrum binations in Klebsiella pneumoniae and in vivo response to b-lactam beta-lactamases in Enterobacteriaceae in four French hospitals. Eur J therapy. J Infect Dis 1996;173:151 8. Clin Microbiol Infect Dis 1989;8:527 9. 32. French GL, Shannon KP, Simmons N. Hospital outbreak of Klebsiella pneumoniae resistant to broad-spectrum cephalosporins and b-lactam- 22. Decre D, Arlet G, Pernet M, Calvez H, Bergogne-Berezin E, Regnier B. b-lactamase inhibitor combinations by hyperproduction of SHV-5 b- Epidémiologie des souches de Klebsiella pneumoniae productrices de lactamase. J Clin Microbiol 1996;34:358 63. beta-lactamase à spectre élargi en réanimation. Rean Urg 1993;2:719. 33. Rice LB, Carias LL, Shlaes DM. In vivo efficacies of b-lactam-b-lacta- 23. Garau J. Beta-lactamases: current situation and clinical importance. Intenmase inhibitor combinations against a TEM-26-producing strain of sive Care Med 1994;20(suppl 3):S5 9. Klebsiella pneumoniae. Antimicrob Agents Chemother 1994;38: 24. Arlet G, Rouveau M, Casin I, et al. Molecular epidemiology of Klebsiella 2663 4. pneumoniae strains that produce SHV-4 beta-lactamase and which were 34. Jett BJ, Ritchie DJ, Reichley R, Bailey TC, Sahm SF. In vitro activities isolated in 14 French hospitals. J Clin Microbiol 1994;32:2553 8. of various b-lactam antimicrobial agents against clinical isolates of 25. Thabaut A, Acar J, Allouch P, et al. Fréquence et distribution des betalacta- Escherichia coli and Klebsiella spp. resistant to oxyiminocephalospormases chez 1792 souches de Klebsiella pneumoniae isolées en France ins. Antimicrob Agents Chemother 1995;39:1187 90. entre 1985 et 1988. Pathol Biol (Paris) 1990;38:459 63. 35. Bauernfeind A. Perspectives of beta-lactamase inhibitors in therapy of 26. Jacoby GA. Characteristics of an outbreak of multidrug resistant Klebsiella infections caused by Escherichia coli or Klebsiella with plasmidic resispneumoniae (MDRKP) infections [abstract no C48]. In: Program and tance to third-generation cephalosporin. Infection 1990;18:48 52. abstracts of the 34th Interscience Conference on Antimicrobial Agents 36. Leleu G, Kitzis MD, Vallois JM, Gutmann L, Decazes JM. Different ratios and Chemotherapy (Orlando, FL). Washington, DC: American Society of the piperacillin-tazobactam combination for treatment of experimenfor Microbiology, 1994:88. tal meningitis due to Klebsiella pneumoniae producing the TEM-3 ex- 27. Sirot DL, Goldstein FW, Soussy CJ, et al. Resistance to cefotaxime and tended-spectrum b-lactamase. Antimicrob Agents Chemother 1994;38: seven other b-lactams in members of the family Enterobacteriaceae: a 195 9.