to estimate the in vivo inoculum effect. Investigations were performed by using a TEM-3-producing strain of K

Transcription

1 ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Sept. 1992, p Vol. 36, No /92/ $02.00/0 Copyright 1992, American Society for Microbiology Piperacillin, Tazobactam, and Gentamicin Alone or Combined in an Endocarditis Model of Infection by a TEM-3-Producing Strain of Klebsiella pneumoniae or Its Susceptible Variant HERVE MENTEC,1 JEAN-MARIE VALLOIS,l ANNE BURE,1 AZZAM SALEH-MGHIR,l FRAN(;OIS JEHL,2 AND CLAUDE CARBON"* Institut National de la Sant,e et de la Recherche Medicale U13, H6pital Claude Bernard, Paris Cede-x 19,1 and Institut de Bacteriologie, Faculte de Medecine, 3 rue Koeberle, Strasbourg, France Received 30 September 1991/Accepted 3 July 1992 The efficacy of tazobactam, a 1-lactamase inhibitor, in combination with piperacillin, was studied in vitro and in rabbit experimental endocarditis due to a Klebsiella pneumoniae strain (KpR) producing an extendedspectrum 13-lactamase, TEM-3, or its nonproducing variant (KpS). In vitro, piperacillin was active against KpS (MIC = 4,g/ml, MBC = 8,Lg/ml with 107-CFU/ml inoculum) but not against KpR (MIC = MBC = 256 p.g/ml). Tazobactam (1 jg/ml) restored the activity of piperacillin against KpR (MIC = 2,ug/ml, MBC = 4 jig/ml). Gentamicin was active against both strains (MIC = 0.25 and 0.5,ug/ml for KpS and KpR, respectively). The piperacillin-tazobactam-gentamicin combination was synergistic in vitro. The piperacillin/ tazobactam ratio in plasma and in vegetations was always lower than the 4/1 injected dose ratio. In vivo, piperacillin (300 mg/kg of body weight four times a day [QID]) was active against KpS but not against KpR. Tazobactam (75 mg/kg QID) was able to restore the in vivo effect of piperacillin (300 mg/kg QID) against KpR (-3.0 log10 CFU/g of vegetation versus that of controls). Gentamicin (4 mg/kg twice a day [BID]) was active against both strains. Compared with controls, the combination of gentamicin plus piperacillin against KpS (-5.6 log1o CFU/g of vegetation), and the gentamicin-piperacillin-tazobactam combination against KpR (-4.4 loglo CFU/g of vegetation) achieved the greatest decrease in bacterial counts in vegetations and were the only regimens that significantly increased the proportion of sterile vegetations. It is concluded that (i) tazobactam was able to restore the effect of piperacillin against a TEM-3 extended-spectrum I-lactamase-producing strain of K. pneumoniae, both in vitro and in a severe experimental infection with high inoculum, when used in a 4/1 piperacillin/tazobactam dose ratio; (ii) gentamicin alone was effective because of the high peak/mbc ratio in plasma; (iii) piperacillin-tazobactam-gentamicin, probably because of the effect of gentamicin in reducing bacterial inoculum in vivo, as stressed by the results obtained by piperacillin-gentamicin against KpS, may be the most effective regimen against KpR. 1-Lactamase production is the main mechanism of resistance of gram-negative rods to beta-lactam antibiotics (31). Since 1983, extended-spectrum,b-lactamase-producing strains of the family Enterobacteriaceae, resistant to penicillins and expanded-spectrum cephalosporins, have been largely responsible for nosocomial infections (7, 17, 21, 25, 26, 33, 34). Such resistance often includes cross-resistance to antibiotics of other antimicrobial classes (7, 33). Therefore, the therapeutic regimens which are active against such multiresistant members of the family Enterobacteriaceae remain limited. One possible treatment for such strains is the use of,3-lactamase inhibitors. One potential combination includes piperacillin, a broad-spectrum penicillin, and tazobactam, a penicillanic acid-sulfone derivative, which is a potent irreversible P-lactamase inhibitor according to in vitro data (3-5, 18-20). The aim of this study was to evaluate the efficacy of piperacillin-tazobactam combination in the treatment of severe infections due to extended-spectrum,b-lactamase (type TEM-3)-producing strains of the multiresistant Enterobacteriaceae. For that purpose, we used the rabbit aortic endocarditis model, which provides a rigorous test for studying the efficacy of antimicrobial agents in severe infections with high local concentrations of bacteria. We chose a Klebsiella pneumoniae strain which produced a * Corresponding author TEM-3 3-lactamase for inducing endocarditis. We therefore made an attempt to correlate in vivo results to in vitro and kinetics data, to determine the conditions of in vivo efficacy of the piperacillin-tazobactam combination, and to test the benefit of the addition of gentamicin to piperacillin-tazobactam, to estimate the in vivo inoculum effect. Investigations were performed by using a TEM-3-producing strain of K pneumoniae and the susceptible clinical isolate. (This work was presented in part at the 30th Interscience Conference on Antimicrobial Agents and Chemotherapy, Atlanta, Ga., 21 to 24 October 1990 [28a].) MATERIALS AND METHODS Both strains produced, although at a very low level, the common chromosomal SHV1 P-lactamase found in K pneumoniae. Both strains had identical growth rates. In this work, KpS designates the susceptible strain and KpR designates its resistant variant. Both KpS and KpR Bacterial strains. Two strains were used in this study. The reference strain K pneumoniae 2222, a clinical isolate, was susceptible to piperacillin. Its resistant variant was obtained after transfer of a plasmid encoding one,b-lactamase only (an extended-spectrum P-lactamase-type TEM-3), a 6'-aminoglycoside acetyltransferase type IV, and resistance to sulfonamides and tetracyclines (25).

2 1884 MENTEC ET AL. were resistant to rabbit serum and able to reliably induce experimental endocarditis in the animal. Antibiotics. Piperacillin and tazobactam were supplied as separate compounds by Lederle Laboratories (Rungis, France), and gentamicin was supplied by the Pharmacie Centrale des H6pitaux de Paris (Paris, France). In vitro studies. (i) Susceptibility testing. MICs and MBCs were determined by the microdilution or the tube macrodilution method in Mueller-Hinton broth supplemented with Ca2" and Mg2" (Biomierieux, Craponne, France) (13, 32). Each value was determined with inocula in the logarithmic phase of growth. The MIC was defined as the lowest concentration of a drug that inhibited growth, i.e., visible turbidity after 24 h of incubation at 37 C. The MBC was determined by subculturing 0.01 ml from each clear cupule or tube onto agar plates and was defined as the lowest concentration of a drug that reduced the number of viable organisms by 99.9%. MICs and MBCs of piperacillin either alone or combined with a fixed concentration of tazobactam (1, 5, or 10 p,g/ml) and of tazobactam either alone or combined with a fixed concentration of piperacillin (1 or 4 p,g/ml) were determined by the microdilution method with an inoculum of 107 CFU/ ml İnoculum effect was investigated by the macrodilution method with inocula of 105 and 107 CFU/ml for piperacillin combined with a fixed concentration of tazobactam (1,g/ml) and for gentamicin, to test the susceptibility under standard conditions and also to reproduce in vitro the high bacterial concentrations observed in cardiac vegetations. (ii) Combination studies. The combination of gentamicin and piperacillin-tazobactam was studied through the checkerboard method, by the microdilution method, with two inocula of 105 and 107 CFU/ml. Piperacillin was combined with a fixed concentration of tazobactam (1 p,g/ml). Fractional inhibitory concentration (FIC) and fractional bactericidal concentration (FBC) indices were determined as previously described (13). (iii) Killing curves. The combination of gentamicin and piperacillin-tazobactam was also studied through the killing curve method. Bactericidal rates were determined in Mueller-Hinton broth, by using an inoculum of 2 x 105 or 2 x 107 CFU of log-phase KpR per ml in a final volume of 2 ml. Antibiotic concentrations were chosen to approximate trough levels achieved in rabbit plasma and cardiac vegetations. Tazobactam was used at a concentration of 2,ug/ml for both inocula. With the small inoculum, piperacillin was used at concentrations of 1 and 2 pg/ml and gentamicin was used at concentrations of 0.125, 0.25, and 0.5,ug/ml. With the large inoculum, piperacillin was used at concentrations of 2 and 4,ug/ml and gentamicin was used at concentrations of 0.25, 0.5, and 1 p,g/ml. Glass tubes were aerated and allowed to shake in a water bath for 18 h. At 0, 3, 6, and 18 h after inoculation of bacteria in the antibiotic-containing broth, serial dilutions of 0.1-ml samples (to 10-6) were subcultured in duplicate on agar plates containing 1-lactamase and incubated at 37 C for 24 h before CFU determination, as previously described (16, 24, 29). Serial dilutions were sufficient to reduce the concentration of gentamicin in samples far below the MIC. The lowest accurate count of CFU was 300 CFU/ml, i.e., 30 colonies on the agar plate on which a 0.1-ml sample of undiluted broth had been subcultured. Synergy was defined as a 22 log1o decrease in CFU per milliliter between the combination and its most active constituent after 18 h (27). In vivo studies. (i) Pharmacokinetic studies. Pharmacoki- ANTIMICROB. AGENTS CHEMOTHER. netics were studied with female New Zealand White rabbits. Plasma samples were collected before and at 15, 30, 60, 120, 240, and 360 min after a single intramuscular (i.m.) injection of piperacillin (300 mg/kg of body weight), either alone or combined with tazobactam (75 mg/kg), in five noninfected animals for each regimen. For animals with experimental endocarditis infected with either KpS (n = 21) or KpR (n = 14) and treated i.m. for 4 days, plasma samples were collected after the last antibiotic injection at peak (45 min) and at trough (6 h for piperacillin and tazobactam, 12 h for gentamicin). The antibiotic regimens were piperacillin, 300 mg/kg four times a day (QID) for 10 animals; piperacillin, 300 mg/kg QID combined with tazobactam 75 mg/kg QID for 15 animals; and gentamicin, 4 mg/kg twice a day (BID) for 10 animals. For animals with experimental endocarditis infected either with KpS (n = 16) or KpR (n = 4) and treated i.m. for 4 days, cardiac vegetations were collected 6 h after the last antibiotic injection, for determination of the trough concentration in vegetations. The antibiotic regimens were piperacillin, 300 mg/kg QID, for 7 animals and piperacillin, 300 mg/kg QID, combined with tazobactam, 75 mg/kg QID, for 13 animals. Plasma and vegetations were stored at -80 C until assayed. Antibiotic levels in vegetations were determined by assay of supernatant of vegetation samples homogenized in phosphate buffer, as previously described (10, 24, 29). Assay standards were prepared in normal rabbit plasma for plasma samples and in phosphate buffer for vegetation samples. The concentrations of piperacillin and tazobactam in plasma and in vegetations were determined by high-pressure liquid chromatography (HPLC) (22). The limit of sensitivity was 0.1,ug/ml. The day-to-day variability of the HPLC assay was characterized by coefficients of variation of 4.5% (2.5,ug/ ml), 2.1% (50,ug/ml), and 1.8% (150,ug/ml) for piperacillin (n = 10 each time) and 5.01% (0.5,ug/ml), 2.1% (7 p,g/ml), and 2% (30,ug/ml) for tazobactam (n = 10 each time). The concentrations of gentamicin in plasma were determined by fluorescence polarization immunoassay (Abbott Laboratories, Rungis, France) (23). The limit of sensitivity was 0.16,ug/ml. (ii) Experimental endocarditis. A modification of the method of Perlman and Freedman (30) was used to produce bacterial aortic endocarditis in female New Zealand White rabbits (weight range, 2.0 to 3.9 kg), as previously described (15, 16, 24, 29). Twenty-four hours after catheter insertion, an inoculum of 109 bacteria, either KpS or KpR, was injected in the marginal ear vein. Only animals with both weight loss of at least 10% of body weight and at least one positive result in blood cultures on day 4 were included in the study. Animals were treated i.m. for 4 days, from days 5 to 9. Seventy-six rabbits were randomly assigned one of the following regimens: untreated controls (KpS and KpR); piperacillin, 300 mg/kg QID (KpS and KpR); piperacillin, 300 mg/kg QID combined with tazobactam, 75 mg/kg QID (KpS and KpR); gentamicin, 4 mg/kg BID (KpS and KpR); piperacillin, 300 mg/kg QID, combined with gentamicin, 4 mg/kg BID (KpS); and piperacillin, 300 mg/kg QID, combined with tazobactam, 75 mg/kg QID, and gentamicin, 4 mg/kg BID (KpR). Animals were killed by chloroform inhalation 6 h after the last antibiotic injection. The heart was removed and the left-side chambers were inspected to confirm vegetative endocarditis. All vegetations from each animal were excised, pooled, rinsed in sterile saline, dried with a sterile compress, and weighed. Vegetations were homogenized in 0.5 ml of sterile saline, and 0.1-ml aliquots were quantitatively subcultured onto agar plates at 37 C for 24 h. An antibiotic

3 VOL. 36, 1992 TEM-3 KLEBSIELLA EXPERIMENTAL ENDOCARDITIS 1885 TABLE 1. MICs and MBCs (micrograms per milliliter) (microdilution method or tube macrodilution method) of piperacillin and tazobactam, either alone or combined, and gentamicin against a TEM-3 3-lactamase-producing strain of K pneumoniae (KpR) or its non-1-lactamase producing variant (KpS) KpS KpR Antibiotic (concn) 1Os io7 MIC MBC MIC MBC MIC MBC MIC MBC Piperacillin NDb ND 4 8 ND ND Tazobactam ND ND ND ND Piperacillin + Tazobactam (1pg/ml) Tazobactam (5 pg/ml) ND ND 1 2 ND ND 2 2 Tazobactam (10 pg/ml) ND ND 1 1 ND ND 2 2 Tazobactam + Piperacillin (1 pg/ml) ND ND 1 2 ND ND 1 1 Piperacillin (4 pg/ml) ND ND uc U ND ND Gentamicin a Inoculum (CFU/ml). b ND, not done. c U, unascertainable; 4 1.g of piperacillin per ml was bacteriostatic and bactericidal in these holes. carryover within vegetations able to modify the counts was eliminated since a 10-2 to 10-6 dilution of homogenized vegetation was used to determine the bacterial titers. Colony count results were expressed as log1o CFU per gram of vegetation. The lower limit of detection by this method was 2 log1o CFU/g. Vegetations were considered sterile when no growth occurred from a 0.1-ml subculture of the undiluted tissue homogenate. In calculating the means, such vegetations were considered to be 2 log1o CFU/g. The susceptibility pattern of surviving bacteria to the, various drugs was assessed by the disk method. Statistics. The Mann and Whitney test was used to compare pharmacokinetic parameters. For comparison of observed strengths, the chi-square test, possibly corrected by the Yates method, or the Fisher exact test was used, in accordance with group sizes. Bacterial concentrations in vegetations from various groups of animals were compared by nonparametric tests, by using the Kruskall and Wallis test followed by the Mann and Whitney test. A two-sided P value of <0.05 was considered significant. For multiple comparisons of bacterial concentrations in vegetations, a one-sided approach was used, and significance level was corrected according to the Bonferroni method. Results were expressed as means ± standard deviations. RESULTS Susceptibility testing. MICs and MBCs for the two study strains are shown in Table 1. Tazobactam was inactive against both strains. For piperacillin, MICs and MBCs against KpS were those commonly reported for susceptible strains, without inoculum effect, and KpR was resistant. Tazobactam restored the activity of piperacillin against KpR from the lowest concentration (1,ug/ml), with no further effect of highetr concentrations. Results of the combination were the same as those with a fixed concentration of either tazobactam or piperacillin. No inoculum effect was noted with tazobactam. The activity of piperacillin against KpS was also enhanced by tazobactam. Gentamicin was active against both strains, without inoculum effect. Combination studies. By the checkerboard method, the combination of gentamicin with piperacillin-tazobactam showed no synergistic effect. For KpS, FIC and FBC indices were and 0.75 at small and large inocula, respectively, and for KpR both indices were and 0.75 at small and large inocula, respectively. No inoculum effect was observed. Killing curves. Results of killing curve studies are shown in Fig. 1. With the standard inoculum (Fig. 1A), the addition of gentamicin (0.25,ug/ml) to the piperacillin (1,ug/ml)-tazobactam (2 p,g/ml) combination was synergistic (-5.8 log1o CFU/ml versus the most effective constituent after 18 h), but the addition of gentamicin (0.125,ug/ml) to the piperacillin (2,ug/ml)-tazobactam (2 pg/ml) combination was not synergistic. With the large inoculum (Fig. 1B), the addition of gentamicin (0.5 p,g/ml) to the piperacillin (2,ug/ml)-tazobactam (2 p1g/ml) combination was synergistic (-3.5 log1o CFU/ml versus the most effective constituent after 18 h, with a marked effect from early counts after 3 and 6 h), but the addition of gentamicin (0.5,g/ml) to the piperacillin (4 ig/ml)-tazobactam (2 jig/ml) combination was not synergistic Ṗharmacokinetic studies. Parameters derived from concentrations of piperacillin and tazobactam in plasma following a single i.m. injection of each drug separately or in combination in noninfected animals are summarized in Table 2. Simultaneous injection of tazobactam significantly reduced the maximum concentration and lengthened the mean retention time of piperacillin. During the therapeutic interval, the piperacillin/tazobactam ratio remained close to the 4/1 injected ratio, ranging from 4.2 after 15 min to 2.4 after 360 min; i.e., the tazobactam concentration in plasma was always at least a quarter of the piperacillin concentration. In infected animals, the injected piperacillin/tazobactam ratio was also 4/1. The mean peak ratio in plasma was 2.4, and the mean trough ratio in plasma was 3.0. The mean ratio in vegetations was 1.0. Table 3 shows antibiotic concentrations after 4 days of treatment, after the last antibiotic injection, at peak and trough for plasma levels and at trough for vegetation levels, in infected animals.

4 1886 MENTEC ET AL. ANTIMICROB. AGENTS CHEMOTHER. Aio B lo U. -ț U5 6 ~6 -I %O. XI I I I I~I~ I I "" s * 1 TME TIME lm FIG. 1. In vitro killing rates of K pneumoniae producing an extended-spectrum,-lactamase, TEM-3, incubated without antibiotic (control) and with various concentrations of gentamicin (G), piperacillin (P), and tazobactam (T). The inoculum was either standard (2 x 105 CFU/ml) (A) or large (2 x 107 CFU/ml) (B). Symbols and concentrations of each drug (in micrograms per milliliter) are indicated on the right of each part of the figure (e.g., Gl = 1 ILg of gentamicin per ml). Experimental endocarditis. In vivo results are shown in Table 4. Bacterial counts in vegetations did not differ between the two control groups (KpS and KpR). Piperacillin alone reduced bacterial counts in vegetations in animals infected with KpS but was inactive against KpR. Tazobactam was able to restore the in vivo effect of piperacillin against KpR. Gentamicin alone was active against both strains. The combinations of gentamicin with piperacillin against KpS and with piperacillin-tazobactam against KpR were the only regimens that significantly increased the proportion of sterile vegetations compared with that of the controls. The susceptibility pattern of bacteria isolated from blood cultures and vegetations never differed from that of the parental strain. No loss of resistance was observed for KpR. DISCUSSION Previous studies have investigated the efficacy of the combination of a beta-lactam and a P-lactamase inhibitor in the experimental endocarditis model: ceftazidime-dicloxacillin (6), ampicillin-sulbactam (14, 35), and ceftriaxone-sulbactam (8, 15). These studies yielded variable results that could be explained by differences between 3-lactamase types (1, 8). Our results can be compared with those obtained by Caron and colleagues (8) with the same model, for the same bacteria. Despite the less favorable kinetics of piperacillin compared with ceftriaxone in rabbits and despite the MIC and MBC of piperacillin-tazobactam being higher than those of ceftriaxone-sulbactam, tazobactam, in our work, was able to restore the activity of piperacillin and to allow an in vivo TABLE 2. Plasma kinetic parameters (means + standard deviations) of piperacillin (300 mg/kg) alone or combined with tazobactam (75 mg/kg) and of tazobactam combined with piperacillin following a single i.m. injection in noninfected rabbits (n = 5 for each regimen) Drug treatment Cm.a Tma AUCa MRTb (pg/mi) (min) (mg liter-' min-') (min) Piperacillin alone ± ± ,859 ± 6, ± 25.6 Piperacillin with tazobactam 68.0 ± 13.3c 33.8 ± ,861 ± ± 22.7c Tazobactam with piperacillin 21.9 ± ± , , ± 55.0 a Cmax, maximum concentration of drug in serum; Tmx, time to maximum concentration of drug in serum; AUC, area under the concentration-time curve. b MRT, mean retention time. c P < 0.05 versus piperacillin alone.

5 VOL. 36, 1992 TEM-3 KLEBSIELLA EXPERIMENTAL ENDOCARDITIS 1887 TABLE 3. Antibiotic concentrations (means + standard deviations) after the last antibiotic injection of a 4-day treatment, in plasma and in vegetations, of rabbits with K pneumoniae experimental endocarditis Antibiotic (dose) Peak Plasma Concn (p,g/ml) ina: Trough Vegetations Piperacillin alone (300 mg/kg QID) (0/10) 5.6 ± 4.3 (0/9) 3.6 ± 6.3 (5/7) Piperacillin with tazobactam (300 mg/kg QID) ± (0/15) 7.6 ± 6.7 (0/12) 9.8 ± 10.8b (5/13) Tazobactam with piperacillin (75 mg/kg QID) ± (0/15) 5.8 ± 7.0 (0/12) 5.4 ± 9.3 (9/13) Gentamicin (4 mg/kg BID) (0/7) 0.3 ± 0.6 (5/10) NIc P/T ratio' ± ± 0.6e a The number of assays with results below the limit of detectability/total number of assays done is indicated in parentheses. b p < 0.01 versus piperacillin alone. c ND, not done. d P/T ratio, piperacillin-to-tazobactam ratio. e Mean for four animals. efficacy, while sulbactam was not able to restore the efficacy of ceftriaxone. As previously reported (3-5, 18-20), tazobactam was able, in vitro, to restore completely the activity of piperacillin otherwise hydrolyzed by TEM-3 P-lactamase. This effect was observed from the lowest concentration tested (1,ug/ml), with no further effect of higher concentrations. The optimal piperacillin/tazobactam ratio seemed to be within the range of 2/1 to 4/1. The 4/1 piperacillin/ tazobactam dose ratio has already been shown to be more active, both in vitro and in vivo, than the 8/1 dose ratio recommended by the manufacturer (2, 28). In vivo, tazobactam was used in combination with piperacillin in a 4/1 dose ratio, in agreement with in vitro experiments. Tazobactam also entailed a fourfold reduction of the MIC and MBC of piperacillin on KpS. This effect was probably due to the inhibition of the common chromosomal SHV1,B-lactamase present in K pneumoniae (18-20, 28, 33, 36). Tazobactam has been shown to be an irreversible inhibitor of both plasmidic and chromosomal P-lactamases (18, 19). There was a discrepancy between the potent TEM-3 inhibitory effect exerted in vitro by tazobactam, which sharply lowered the piperacillin MIC and MBC on KpR, and the imperfect in vivo effect of the piperacillin-tazobactam combination, which left nonsterilized 3 of 7 vegetations infected with KpR. This could be partly explained by piperacillin and tazobactam pharmacokinetics in rabbits, taking into account the usually poor susceptibility of K pneumoniae strains to piperacillin. These considerations might promote the study of tazobactam in combination with a P-lactamase susceptible beta-lactam of longer half-life and of better intrinsic activity against K pneumoniae. Nevertheless, other factors that have an effect upon the activity of antimicrobial agents in experimental endocarditis should be considered. (i) A lack of penetration of tazobactam and/or piperacillin in vegetations was unlikely, as measured concentrations were sufficient to exert activity against KpR. A good diffusion in vegetation has been reported for piperacillin (12) and for another penicillanic acid-sulfone-derived P-lactamase inhibitor, sulbactam (8, 15). However, a heterogenous distribution of tazobactam in vegetations, as described for other drugs by using an autoradiography method, cannot be excluded (11). (ii) The high density of bacteria TABLE 4. Results of different 4-day therapeutic regimens for rabbits with endocarditis due to a TEM-3 P-lactamase-producing strain of K pneumoniae (KpR) or its nonproducing variant (KpS) KpS KpR Antibiotic regimen Log1l CFU/g No. sterile/ Log10 CFU/g No. sterile/ (unitary dose, mg/kg) of vegetation total no. of of vegetation total no. of (mean + SD) vegetations (mean ± SD) vegetations Control 7.9 ± 0.9 0/6 7.6 ± 1.3 0/9 Piperacillin (300 QID) 5.3 ± 2.5a 2/9 6.1 ± 2.7 0/8 Piperacillin (300 QID) + tazobactam (75 QID) 5.7 ± 3.0 2/8 4.6 ± 2.4a 3/9 Gentamicin (4 BID) 2.9 ± 1.Oa 2/5 3.7 ± 2.1a 4/8 Piperacillin (300 QID) + gentamicin (4 BID) 2.3 ± 0.7a 5/7a NDb ND Piperacillin (300 QID) + tazobactam (75 QID) ND ND 3.2 ± 1.9a 4/7a + gentamicin (4 BID) a Results statistically different from those of control. b ND, not done.

6 1888 MENTEC ET AL. inside vegetations can entail an in vivo inoculum effect. (iii) The high density of bacteria also produces a high concentration of P-lactamase in vegetations. Such an effect towards P-lactamase inhibitors has been documented previously (8, 15), even for piperacillin-tazobactam (9). In our study, however, the in vitro inhibitory effect of tazobactam was not concentration dependent and persisted at a large inoculum. (iv) The poor in vivo activity of piperacillin-tazobactam on KpS could be explained by the loss of activity of beta-lactam antibiotics on slowly growing microorganisms. In our study, these limits were strengthened by the short treatment duration, which was intentionally chosen to allow differences between groups to arise. This short treatment duration did not allow sterilization of all vegetations. In conclusion, the enhanced activity of piperacillin observed in the presence of an inhibitor of P-lactamases and tazobactam, in vitro and in an endocarditis model, seems promising for the treatment of serious infections due to members of the family Enterobacteriaceae producing an extended-spectrum 3-lactamase such as TEM-3. However, it must be emphasized that our study was carried out with a 4/1 piperacillin/tazobactam dose ratio to inhibit the great amount of,-lactamase present in vegetations. The use of an aminoglycoside is advocated to improve in vivo efficacy, through reduction of the inoculum. Further studies with other beta-lactam antibiotics combined with tazobactam would be of interest. ACKNOWLEDGMENTS This work was supported in part by a grant from La Fondation pour la Recherche Medicale (H.M.). We are indebted to Laurent Gutmann, who provided the two study strains. REFERENCES 1. Acar, J. F., L. Gutmann, and M. D. Kitzis Beta-lactamases in clinical isolates. Spectrum implications of sulbactam/ ampicillin. Drugs 35: American Cyanamid Company Piperacillin/tazobactam, brochure for clinical investigators. American Cyanamid Company, Pearl River, N.Y. 3. Appelbaum, P. C., M. R. Jacobs, S. K. Spangler, and S. Yamabe Comparative activity of P-lactamase inhibitors YTR 830, clavulanate, and sulbactam combined with,-lactams against P-lactamase-producing anaerobes. Antimicrob. Agents Chemother. 30: Aronoff, S. C., M. R. Jacobs, S. Johenning, and S. Yamabe Comparative activities of the 3-lactamase inhibitors YTR 830, sodium clavulanate, and sulbactam combined with amoxicillin or ampicillin. Antimicrob. Agents Chemother. 26: Aronoff, S. C., M. R. Jacobs, P. H. Labrozzi, and S. Yamabe Synergy of amoxicillin combined with clavulanate and YTR 830 in experimental infections in mice. J. Antimicrob. Chemother. 18: Bayer, A. S., M. Selecky, K. Babel, L. Hirano, J. Yih, and T. R. Parr, Jr Bactericidal interactions of a f3-lactam and P-lactamase inhibitors in experimental Pseudomonas aeruginosa endocarditis caused by a constitutive overproducer of type Id 3-lactamase. Antimicrob. Agents Chemother. 31: Brun-Buisson, C., P. Legrand, A. Philippon, F. Montravers, M. Ansquer, and J. Duval Transferable enzymatic resistance to third-generation cephalosporins during nosocomial outbreak of multiresistant Klebsiella pneumoniae. Lancet ii: Caron, F., L. Gutmann, A. Bure, B. Pangon, J.-M. Vallois, A. Pechinot, and C. Carbon Ceftriaxone-sulbactam combination in rabbit endocarditis caused by a strain of Klebsiella pneumoniae producing extended-broad-spectrum TEM-3 1-lactamase. Antimicrob. Agents Chemother. 34: Chiesa, C., P. H. Labrozzi, and S. C. Aronoff Decreased ANTIMICROB. AGENTS CHEMOTHER. baseline beta-lactamase production and inducibility associated with increased piperacillin susceptibility of Pseudomonas cepacia isolated from children with cystic fibrosis. Pediatr. Res. 20: Contrepois, A., J.-M. Vallois, J.-J. Garaud, B. Pangon, J. Mohler, A. Meulemans, and C. Carbon Kinetics and bactericidal effect of gentamicin and latamoxef (moxalactam) in experimental Escherichia coli endocarditis. J. Antimicrob. Chemother. 17: Cremieux, A. C., B. Maziere, J.-M. Vallois, M. Ottaviani, A. Azancot, H. Raffoul, A. Bouvet, J.-J. Pocidalo, and C. Carbon Evaluation of antibiotic diffusion into cardiac vegetations by quantitative autoradiography. J. Infect. Dis. 159: Daschner, F. D., M. Just, G. Spillner, and V. Schlosser Penetration of piperacillin into cardiac valves, subcutaneous and muscle tissue of patients undergoing open-heart surgery. J. Antimicrob. Chemother. 9: Dougherty, P. F., D. W. Yotter, and T. R. Matthews Microdilution transfer plate technique for determining in vitro synergy of antimicrobial agents. Antimicrob. Agents Chemother. 11: English, A. R., D. Girard, C. Cimochowski, J. Faiella, J. A. Retsema, and J. E. Lynch Activity of sulbactam/ampicillin in screening and discriminative animal models of infection. Rev. Infect. Dis. 8:S535-S Fantin, B., B. Pangon, G. Potel, F. Caron, E. Vallee, J.-M. Vallois, J. Mohler, A. Burk, A. Philippon, and C. Carbon Activity of sulbactam in combination with ceftriaxone in vitro and in experimental endocarditis caused by Escherichia coli producing SHV-2-like 3-lactamase. Antimicrob. Agents Chemother. 34: Fantin, B., B. Pangon, G. Potel, J.-M. Vallois, F. Caron, A. BurE, and C. Carbon Ceftriaxone-netilmicin combination in single-daily-dose treatment of experimental Escherichia coli endocarditis. Antimicrob. Agents Chemother. 33: Fournier, G., A. BurE, V. Jarlier, P. Legrand, M. H. Nicolas, P. Decrez, J. Duval, and A. Philippon P-lactamases mediated multiple P-lactam resistance in Enterobacteriaceae including Klebsiella, Escherichia and Salmonella, abstr. 521, p Program Abstr. 27th Intersci. Conf. Antimicrob. Agents Chemother. 18. Gutmann, L., M. D. Kitzis, S. Yamabe, and J. F. Acar Comparative evaluation of a new P-lactamase inhibitor, YTR 830, combined with different 1-lactam antibiotics against bacteria harboring known,-lactamases. Antimicrob. Agents Chemother. 29: Jacobs, M. R., S. C. Aronoff, S. Johenning, D. M. Shlaes, and S. Yamabe Comparative activities of the,b-lactamase inhibitors YTR 830, clavulanate, and sulbactam combined with ampicillin and broad-spectrum penicillins against defined 3-lactamase-producing aerobic gram-negative bacilli. Antimicrob. Agents Chemother. 29: Jacobs, M. R., S. C. Aronoff, S. Johenning, and S. Yamabe Comparative activities of the beta-lactamase inhibitors YTR 830, clavulanate, and sulbactam combined with extendedspectrum penicillins against ticarcillin-resistant Enterobacteriaceae and pseudomonads. J. Antimicrob. Chemother. 18: Jarlier, V., M. H. Nicolas, G. Fournier, and A. Philippon Extended broad-spectrum beta-lactamases conferring transferable resistance to newer beta-lactam agents in Enterobacteriaceae: hospital prevalence and susceptibility patterns. Rev. Infect. Dis. 10: Jehl, F., P. Birckel, and H. Monteil Hospital routine analysis of penicillins, third-generation cephalosporins and aztreonam by conventional and high-speed high-performance liquid chromatography. J. Chromatogr. 413: Jolley, M. E Fluorescence polarization immunoassay for the determination of therapeutic drug levels in human plasma. J. Anal. Toxicol. 5: Joly, V., B. Pangon, J.-M. Vallois, L. Abel, N. Brion, A. BurE, N. Phong Chau, A. Contrepois, and C. Carbon Value of antibiotic levels in serum and cardiac vegetations for predicting

7 VOL. 36, 1992 TEM-3 KLEBSIELLA EXPERIMENTAL ENDOCARDITIS 1889 antibacterial effect of ceftriaxone in experimental Eschenichia coli endocarditis. Antimicrob. Agents Chemother. 31: Kitzis, M. D., D. Billot-Klein, F. W. Goldstein, R. Williamson, G. Tran-Van Nhieu, J. Carlet, J. F. Acar, and L. Gutmann Dissemination of the novel plasmid-mediated 3-lactamase CTX-1, which confers resistance to broad-spectrum cephalosporins, and its inhibition by 13-lactamase inhibitors. Antimicrob. Agents Chemother. 32: Knothe, H., P. Shah, V. Krcmery, M. Antal, and S. Mitsuhashi Transferable resistance to cefotaxime, cefoxitin, cefamandole and cefuroxime in clinical isolates of Klebsiella pneumoniae and Serratia marcescens. Infection 11: Krogstad, D. J., and R. C. Moellering Antimicrobial combinations, p In V. Lorian (ed.), Antibiotics in laboratory medicine. The Williams & Wilkins Co., Baltimore. 28. Kuck, N. A., N. V. Jacobus, P. J. Petersen, W. J. Weiss, and R. T. Testa Comparative in vitro and in vivo activities of piperacillin combined with the 13-lactamase inhibitors tazobactam, clavulanic acid, and sulbactam. Antimicrob. Agents Chemother. 33: a.Mentec, H., J. M. Vailois, A. Bure, A. Saleh-Mghir, F. Jehl, and C. Carbon Program Abstr. 30th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 705, p Pangon, B., V. Joly, J.-M. Vallois, L. Abel, A. Bure, N. Brion, A. Contrepois, and C. Carbon Comparative efficacy of cefotiam, cefmenoxime, and ceftriaxone in experimental endocarditis and correlation with pharmacokinetics and in vitro efficacy. Antimicrob. Agents Chemother. 31: Perlman, B., and L. Freedman Experimental endocarditis. II. Staphylococcal infection of the aortic valve following placement of a polyethylene catheter in the left side of the heart. Yale J. Biol. Med. 44: Philippon, A., R. Labia, and G. Jacoby Extended-spectrum P-lactamases. Antimicrob. Agents Chemother. 33: Schoenknecht, F. D., L. D. Sabath, and C. Thornsberry Susceptibility tests: special tests, p In E. H. Lennette, A. Balows, W. J. Hausler, Jr., and H. J. Shadomy (ed.), Manual of clinical microbiology, 4th ed. American Society for Microbiology, Washington, D.C. 33. Sirot, D., J. Sirot, R. Labia, A. Morand, P. Courvalin, A. Darfeuille-Michaud, R. Perroux, and R. Cluzel Transferable resistance to third-generation cephalosporins in clinical isolates of Kiebsiella pneumoniae: identification of CTX-1, a novel beta-lactamase. J. Antimicrob. Chemother. 20: Sirot, J., C. Chanal, A. Petit, D. Sirot, R. Labia, and G. Gerbaud Kiebsiella pneumoniae and other Enterobacteriaceae producing novel plasmid-mediated beta-lactamases markedly active against third-generation cephalosporins: epidemiologic studies. Rev. Infect. Dis. 10: Washburn, R. G., and D. T. Durack Efficacy of ampicillin plus a beta-lactamase inhibitor (CP-45,899) in experimental endocarditis due to Staphylococcus aureus. J. Infect. Dis. 144: Williams, J. D Importance of beta-lactamases and clinical implications of their inhibitors. Drugs 35(Suppl. 7):3-11.