lig/ml. The spectrum of cefmenoxime was similar to that of other, newer cephalosporins,

ANTIMICROBLAL AGENTS AND CHEMOTHERAPY, Dec. 1981, p. 747-759 Vol. 2, No. 6 66-484/81/12747-13$2./ Cefinenoxime (SCE-1365), a new Cephalosporin: In Vitro Activity, Comparison with Other Antimicrobial Agents, Beta- Lactamase Stability, and Disk Diffusion Testing with Tentative Interpretive Criteria PETER C. FUCHS,`* RONALD N. JONES,2 CLYDE THORNSBERRY,3 ARTHUR L. BARRY,4 E. HUGH GERLACH,5 AND HERBERT M. SOMMERS6 Department of Pathology, St. Vincent Hospital and Medical Center, Portland, Oregon 97225'; Department of Pathology, Kaiser Foundation Laboratory, Clackamas, Oregon 9715F; Antimicrobics Investigations Section, Centers for Disease Control, Atlanta, Georgia 3333; Clinical Microbiology Laboratories, University of California, Davis, Medical Center, Sacramento, California 978174; Department of Laboratories, St. Francis Hospital, Wichita, Kansas 672145; and Department of Pathology, Northwestern Memorial Hospital, Chicago, Illinois 661 Received 23 April 1981/Accepted 18 August 1981 The in vitro activity of cefmenoxime (SCE-1365) was evaluated in a multiphased collaborative investigation. Over 7,5 consecutive clinical isolates were tested in five laboratories, and greater than 9%o of the following organisms were inhibited by cefmenoxime at the following concentrations: Enterobacteriaceae and non-enterococcal streptococci,.125,ug/ml; Staphylococcus aureus, <2.,ug/ml; and nonfermenting gram-negative bacilli and Bacteroides fragilis group, s ug/ml. Both beta-lactamase-producing and -nonproducing Haemophilus influenzae and Neisseria gonorrhoeae were inhibited by cefmenoxime at s.3 lig/ml. The spectrum of cefmenoxime was similar to that of other, newer cephalosporins, particularly cefotaxime. A pronounced inoculum effect was found with some species upon increasing inocula from 15 to 17 colony-forming units per ml, resulting in an approximate eightfold increase in minximum inhibitory concentrations. Cefmenoxime was bactericidal when tested with inocula of 15 colonyforming units per ml, and mean differences between the minimum inhibitory concentration and the miniimum lethal concentration were less than one log2 dilution. No significant hydrolysis of cefmenoxime by five different beta-lactamases was detectable, and cefmenoxime exhibited marked inhibition of type I beta-lactamases. Regression and error rate-bounded analyses of results of disk diffusion and reference broth microdilution susceptibility tests were performed on 421 bacterial isolates, and the following tentative zone size breakpoints are proposed: 222 mm, susceptible; c14 mm, resistant; and 15 to 21 mm, moderately susceptible (indeterminate). These data and cross-resistance studies with other newer cephalosporins indicate marked similarity of in vitro activity within this group of drugs, particularly between cefmenoxime, moxalactam, and cefotaxime. Any one of these could serve as the representative for the disk diffusion testing of this group of drugs if comparable minimum inhibitory concentration breakpoints were used for each drug. Cefmenoxime, formerly known as SCE-1365, over 7,5 recent clinical bacterial isolates, bacis a new cephalosporin structurally similar to tericidal activity against a select population of cefotaxime, differing from the latter by having a clinical isolates, effect of inoculum size on minimethyltetrazolthio group in the 3 position in- mum inhibitory concentrations (MICs), comparstead of an acetoxymethyl group. Cefmenoxime ison of in vitro activity of cefmenoxime with that is reported to have broad-spectrum activity of related cephalosporins and gentamicin against against both gram-positive and gram-negative selected groups of susceptible and resistant bacbacteria (2). In this report, we present the teria, evaluation of the effects of various betaresults of a collaborative in vitro evaluation of lactamases on cefmenoxime by hydrolysis and cefmenoxime which include: activity against susceptibility studies, study of beta-lactamase- 747

41 --Aw FUCHS ET AL. inhibitory activity of cefmenoxime, and comparison of cefinenoxime MIC with disk diffusion susceptibility results and regression analysis. MATERIALS AND METHODS Antibiotics. Cefmenoxime was provided by Abbott Laboratories, North Chicago, Ill., as standard powder., cefamandole, moxalactam, and gentamicin standard powders were obtained from Eli Lilly Research Laboratories, Indianapolis, Ind. Other drug standard powders used were cefotaxime (Hoechst- Roussel Pharmaceuticals, Somerville, N.J.), cefoperazone (Pfizer Pharmaceuticals, New York, N.Y.), cefoxitin (Merck, Sharp & Dohme, West Point, Pa.) and nitrocefin (Glaxo, Inc., Fort Lauderdale, Fla.). Cefmenoxime 3-pg disks were provided by Difco Laboratories, Detroit, Mich. Organisms. The bacteria tested in phase 1 were consecutive clinical isolates encountered between 1 June and 31 July 198 in the clinical microbiology laboratories of Kaiser Foundation Hospital, Portland, Oreg.; Northwestern Memorial Hospital, Chicago, Ill.; St. Francis Hospital, Wichita, Kans.; St. Vincent Hospital and Medical Center, Portland, Oreg.; and University of California, Davis, Medical Center, Sacramento, Calif. In this phase there were 7,675 isolates, which are listed in Tables 1 through 3. The 11 selected clinical isolates in the bactericidal studies were contributed by the collaborating laboratories and selected to represent common pathogens. They included 1 srains each of: Escherichia coli, Klebsiella pneumoniae, Enterobacter spp., Serratia marcescens, Citrobacter spp., Proteus mirabilis, indole-positive Proteus spp., Providencia spp., Pseudomonas aeruginosa, Staphylococcus aureus, and Streptococcus faecalis. Beta-lactamase-positive and - negative Haemophilus influenzae and Neisseria gonorrhoeae from the stock collection at the Centers For Disease Control were similarly tested (see Table 8). Seven bacterial strains with well-documented specific beta-lactamases were utilized to evaluate cefmenoxime susceptibility to beta-lactamases (see Table 9) and beta-lactamase-inhibitory activity. Selected clinical isolates (436) obtained from the five laboratories named above and the Cleveland Clinic Foundation, Cleveland, Ohio, were utilized to establish disk criteria and to compare cefmenoxime MICs with those of six other beta-lactams and gentamicin. These include: 15 Acinetobacter spp., 21 Citrobacter app., 48 Enterobacter spp., 25 E. coli, 25 enterococci, 25 K. pneumoniae, 1 Morganella morganii, 25 P. mirabilis, 1 Proteus vulgaris, 29 Providencia app., 5 P. aeruginosa, 3 Pseudomonas spp., 25 S. marcescens, 48 S. aureus, 1 methicillin-resistant S. aureus, 2 Streptococcuspneumoniae, and 2 Strepcococcus pyogenes. Twenty-five N. meningitidis strains were included in the betalactam comparison, but not in the disk study. Aintimicrobial agent susceptibility testing. In the clinical isolate study, MICs were determined by the microdilution broth method at the Kaiser Foundation Hospital and at the University of California, Davis, Medical Center as previously described (8) and by agar dilution at Northwestern Memorial Hospital, St. Francis Hospital, and St. Vincent Hospital utilizing ANTIMICROB. AGENTS CHEMOTHER. the method recommended by the National Committee for Clinical Laboratory Standards (13). These were conducted in Mueller-Hinton broth and agar (Difco), respectively. The concentrations tested by both methods were:.125,.5,, 2., 4.,, and. jg/ml. The inoculum size for the microdilution broth method was approximately 5 x 15 colony-forming units per ml, and that for agar dilution was approximately 14 colony-forming units per spot. In other phases of the study, the MICs and minimum lethal concentrations were determined by microdilution broth methods as previously described (2). Concentrations tested were serial twofold dilutions ranging from 256 to.8 jig/ ml. H. influenzae, Neisseria meningitidis, and N. gonorrhoeae strains were tested at the Centers for Disase Control by broth microdilution with Mueller- Hinton broth supplemented with 5% peptic digest of blood (Fildes reagent) (H. influenzae and N. meningitidis) and by agar dilution using proteose peptone agar no. 3 supplemented with 1% hemoglobin (BBL Microbiology Systems, Cockeysville, Md.) and 1% Kellogg supplement (23) (N. gonorrhoeae). These procedures have been previously described in detail (1, 18). The susceptibility of anaerobes was tested in Wilkins- Chalgren broth in microdilution trays incubated for 48 h anaerobically at 35 C in GasPak jars (BBL) as previously described (1). The disk diffusion susceptibility testing was performed in accordance with standardized procedures (12). Disk zone size interpretive criteria were determined by both regression analyses according to the formula of least squares as adopted for computer computation and the error ratebounded method of Metzler and DeHaan (11). The beta-lactamase hydrolysis rates of cefmenoxime and three other cephalosporins were determined by the spectrophotometric method described by Neu (14) at 255- to 273-nm wavelengths; The reaction mixtures contained ml of.4 mm beta-lactam substrate in.5 M phosphate buffer (ph 7.) at 37 C and 1 to 2 id of crude enzyme preparation. Decrease in optical density was recorded for 3 min, and the hydrolysis rates in micromoles per minute were converted to percentage rates relative to that of nitrocefin. Inhibition of nitrocefin (87/312) hydrolysis was measured by spectrophotometer at 37 C at a wavelength of 482 nm (15). Reaction mixtures of nitrocefin and cefoxitin, and nitrocefin and cefotaxime were compared with nitrocefin and cefienoxime using crude enzyme preparation (inhibition studies). RESULTS Among the Enterobacteriaceae the susceptibility to cefmenoxime was much greater than susceptibility to cefamandole or cephalothin (Table 1). The MIC9o (concentration at which 9% of isolates are inhibited) of cefmnenoxime for Enterobacteriaceae was.125,ag/ml, whereas that of cefamandole was Ag/ml-a -fold difference. Compared with cephalothin, with an MICoo of >,ug/ml, the superiority was even greater. Over 99% of 4,768 Enterobacteriaceae isolates were inhibited by 5 ug of cefmenoxime per ml. Serratia, with an MICgo of 4,ug/ml,

VOL. 2, 1981 TABLE 1. CEFMENOXIME EVALUATION AND DISK STUDY 749 Comparative susceptibility of 4,768 consecutive Enterobacteriaceae isolated at five institutions to cefmenoxime, cefamandole, and cephalothin % In- Organism n Drug MIC5a MIC9 at hii nml E. coli 2,73 Cefmenoxime.12 c.12 >99.5 2. 98 4. 86 Shigella spp. Klebsiella spp. Enterobacter spp. Serratia spp. Citrobacter spp. Salmonella spp. P. mirabilis P. vulgaris, Morganella spp., and Providencia spp. Miscellaneousb Total 76 Cefmenoxime 696 Cefmenoxime 468 Cefmenoxime 119 Cefmenoxime 13 Cefmenoxime 13 Cefmenoxime 379 Cefmenoxime 144 Cefmenoxime 15 Cefmenoxime 4,768 Cefmenoxime.12 :.12.5 4. c.12 2. > 5.12 > > c.12 -.12 <.12 4.!.12 2. > c.12 c.12.5 4. <.12 2. <.12 2. > > 4. > > > > 2 -c.12 2..5 > >.5 > c.12 > a MIC5o and MIC9, MICs at which 5 and 9%, respectively, of isolates are inhibited; expressed in micrograms per milliliter. b Includes: 2 Arizona hinshawii, 11 Hafnia alvei, and 2 Alkalescens-dispar group. 87 >99 96 9 97 829 97 122 98 79 37 98 79 37 >99 >99 88 >99 78 1 87 4 >99 93 73 was the genus least susceptible to cefmenoxime, but was still far more susceptible than to cefamandole (MIC9o, >,ug/ml). Other genera showing relative resistance to second-generation cephalosporins (cefamandole MIC9o, >,ig/ml) were Enterobacter, Citrobacter, Morganella, and Providencia, for which the cefmenoxime MIC9 values were,,.5, and.5 ug/ml respectively. The activity of cefmenoxime against nonfermenting gram-negative bacilli, although somewhat greater than that of cefamandole and cephalothin (Table 2), was still poor. The mean MIC of cefmenoxime for P. aeruginosa and nonfermenters in general was > ug/ml. However, nearly one-fourth of P. aeruginosa and other Pseudomonas spp. were susceptible to cefme- and an additional 62 and noxime at 8,Ag/ml,

75 FUCHS ET AL. 42%, respectively, of those isolates fell in the moderately susceptible category of to,ug/ ml. Similarly, 67% of Acinetobacter anitratus TABLE 2. Comparative susceptibility of 772 consecutive nonenteric gram-negative bacilli isolated at five institutions to cefinenoxime % In- Organism Organmn IC5aMIC9 MICs n MCso hibited at c8 jig/nil Aeromonas hydrophila 34.12.5 P. aeruginosa 5 > 23 Pseudomonas spp. 41 > 24 Acinetobacter cakcoace- 1 > 15 ticus subsp. anitratus Miscellaneousb 49 4. 63 Total 772 > a See footnote a of Table 1. bincludes: 6 Achromobacter xylosoxidans, 17 Acinetobacter cakoaceticus var. lwoffii, 1 Agrobacterium radiobacter, 5 Alcaligenes app., 11 Flavobacterium spp., 6 Moraxella spp., and 2 Pasteurella multocida. TABLE 3. ANTIMICROB. AGENTS CHEMOTHER. were susceptible to cefmenoxime at c,ug/ml. Against gram-positive cocci the mean MIC of cefmenoxime was slightly greater than that of cefamandole and cephalothin. This was more apparent with the Micrococcaceae, among which a fourfold-greater MICOo was observed with cefmenoxime than with the streptococci; the differences were less apparent at the relatively high concentrations tested, e.g., lowest concentration =.125,ug/ml (compare Tables 3 and 6). The in vitro activity of cefmenoxime was compared with that of six other beta-lactams and gentamicin against 461 clinical strains contributed by the collaborating laboratories. This study showed marked similarity in the activity of cefmenoxime and the other third-generation cephalosporins, especially cefotaxime and moxalactam (Tables 4 through 6). The MICOgo values of these three drugs were within one dilution of each other against all genera of Enterobacteriaceae, except the Proteus-Morganella-Provi- Comparative susceptibility of2,135 consecutive gram-positive clinical isolates at five institutions to cefinenoxime, cefamandole, and cephalothin % In- Organism n Drug MIC5 MICg hibited at s8 JAg/ml S. aureus 99 Cefmenoxime 2..5 >99.5.5 >99 Staphylococcus epidermidis 4 Cefmenoxime 4. 96.5 97.5 97 Staphylococcus saprophyticus 2 Cefmenoxime 2. 9.5 2. 95.5 2. 95 Enterococci 539 Cefmenoxime > > 18 > 4 > 3 S. agalactiae 71 Cefmenoxime so.12 so.12 s.12 s.12 s.12 SO.12 Other Streptococcus spp.b 89 Cefinenoxime <.12 5.12 99.12.5 99 so.12.5 98 Miscellaneous, gram-positive bacteriac 26 Cefmenoxime.5 > 77.5 73.5 > 81 Total 2,135 Cefmenoxime > 78.5 75.5 74 asee footnote a of Table 1. b Includes: 7 S. pyogenes, 7 S. pneumoniae, 4 S. bovis, 28 S. viridans group, 14 beta-hemolytic streptococci not in groups A, B, or D, and 29 nonhemolytic streptococci. eincludes: 1 Micrococcus spp., 1 Listeria monocytogenes, 7 Corynebacterium spp., and 8 Bacillus spp.

VOL. 2, 1981 dencia group (Table 4). was 8 to times less active than cefmenoxime and cefotaxine against P. mirabilis, but was more active than the other drugs against the indole-positive species. Cefmenoxime had lowest MIC9o values against S. marcescens and Citrobacter spp. It also had greater activity than gentamicin against all Enterobacteriaceae, except that gentamicin was slightly superior against Enterobacter spp. had the best activity against P. aeruginosa (MIC9o, 8,ug/ml) and was slightly more active than gentamicin (Table 5). The majority of P. aeruginosa was moderately susceptible (MIC, to,g/ml) to cefmenoxime (78%), cefotaxime (76%), and moxalactam (66%). Other Pseudomonas spp. could be divided into two groups on the basis of their susceptibility to third-generation cephalosporins: P. acidovorans and P. stutzeri were extremely susceptible, and the remaining species were resistant (Table 5). was the most active cephalosporin against A. anitratus (MlCo,,ug/ml), but was less active than the aminoglycoside, gentamicin (MIC9o, 4,ug/ml). and cefamandole were fourfold more active against S. aureus than were thirdgeneration drugs, but a marked increase in MICs of all drugs tested was seen with methicillinresistant strains (Table 6). Enterococci were not susceptible to any of these drugs, but cefoperazone and cefamandole had the best activity (MIC9,,ug/ml). Cefmenoxime and cefotaxime exhibited the best activity against S. pneumoniae (MIC9,.6 and.12,ug/ml) and S. pyogenes (MlC9o,.3 Ag/ml). and cefoxitin were the least active of all cephalosporins tested against these two species (MICso, and 2.,ug/ml). N. meningitidis was extremely susceptible to all cephalosporins, with the third-generation drugs being somewhat more active (MIC9o, '.8 yig/ml) than the earlier compounds (MICgo,.125,ug/ml). When tested against 48 anaerobic isolates, cefmenoxime was intermediate among the other cephalosporins compared (Table 7). The activity against the Bacteroides fragilis group (44% with MIC of c8,ug/ml) was significantly greater than that of cefamandole, but less than that of cefotaxime, moxalactam, and cefoxitin. No one drug was consistently superior against all groups of anaerobes, e.g., moxalactam was most active against B. fragilis and gram-positive cocci; cefoxitin was most active against other Bacteroides spp.; and cefamandole was most active against Clostridium spp. The effect of inoculum concentration on the MICs of cefmenoxime was tested against 11 selected isolates from 11 species. There was a relatively small increase in MICs between the CEFMENOXIME EVALUATION AND DISK STUDY 751 use of inocula of 15 versus 13 colony-forming units per ml (less than 1 log2 difference in mean MICs). By contrast, increasing the inoculum from 15 to 17 colony-forming units per ml produced a substantial increase in MICs (2 to 5 log2 dilution steps). The greatest increases occurred with Citrobacter freundii and Enterobacter spp. Cefmenoxime demonstrated excellent bactericidal activity against these same isolates with an inoculum of 15 colony-forming units per ml. The minimum lethal concentration exceeded the MIC by less than.5 log2, and above 4 utg/ml the cefmenoxime miniimum lethal concentration equalled the MIC. Cefmenoxime and other third-generation cephalosporins were most active against H. in-. fluenzae and N. gonorrhoeae; this potency was not appreciably affected by beta-lactamase production (Table 8). The resistance of cefmenoxiine to beta-lactamases was further tested by determining the hydrolysis rates of cefmenoxime and other cephalosporins by a series of wellcharacterized beta-lactamases by spectrophotometric methods (14). Cefmenoxime was generally refractory to hydrolysis by all tested enzymes (Table 9). The least resistance to hydrolysis occurred with type IV beta-lactamase produced by a Klebsiella spp., with a relative hydrolysis rate (compared with nitrocefin) of only 2%. Furthermore, cefmenoximne markedly inhibited the hydrolysis of nitrocefin by type I beta-lactamase, but had little or no inhibitory effect on hydrolysis by other enzyme types (Fig. 1). This inhibitory spectrum was comparable to that of cefoxitin and cefotaximne. The comparison of cefmenoxime MICs and zone sizes by disk diffusion testing is summarized in the scattergram in Fig. 2. The regression line illustrated is based on values in the MIC range of 2 to 256,ug/ml. Based on the limited cefmenoxime pharnacological data currently available (19), MIC breakpoints for susceptible, resistant, and moderately susceptible (indeterminate) were selected as c8, >, and to,ig/ml, respectively. With these MIC breakpoints, regression analysis yielded corresponding zone size breakpoints of -22 mm for susceptible, c14 mm for resistant, and 15 to 21 mm for the moderately susceptible range. Error ratebounded analysis likewise yielded a susceptible breakpoint for -22 mm, at which point there is a.5% very major error rate (false-susceptible) and a 1.2% major error rate (false-resistant). However, with the introduction of the indeterminate category, these major and very major errors are reduced to minor errors. These categories effectively segregated the major bacterial populations such that most enterococci are resistant, nonfermenters are moderately suscepti-

TABLE 4. E. coli K. pneumoniae Comparative susceptibility of 218 Enterobacteriaceae to cefmenoxime, six other cephalosporins, and gentamicin Enterobacter Spp.b S. marcescens Organism % Inn Drug MICr5a MIC9a hibited ml 25 Cefmenoxime.6.25.6.12.6.25.12 88 2. 96.5 88 68 2. 25 Cefmenoxime.12.25.6.12.12.25.25 92 2. 96 92 4. 84.5 48 Cefmenoxime.12 4. 92.12 9.25 4. 96.25 83 17 4. 58 1.5 25 Cefmenoxime.25 2..5 4..5 4. 2. 84 8 2. Citrobacter spp.c 21 Cefmenoxime.12.25.12.5.25.5.25 95 43 9 43 P. mirabilis 25 Cefmenoxime.3.3.15.15.25.25.5 2. 96.5 4. 4. Morganella, Providencia, and Proteus spp.d 49 Cefmenoxime.25 4. 96.12 2. 94.25.5 98 4. 256 69 4. 8 29 4. 67 a See footnote a of Table 1. bincludes: 21 E. aerogenes, E. cloacae, 9 E. agglomerans, and 2 E. gergoviae. 'Includes: 1 C. diversus and 11 C. freundii. d Includes: 1 P. vulgaris, 1 Providencia rettgeri, 19 P. stuartii, and 1 M. morganii. 752

VOL. 2, 1981 TABLE 5. CEFMENOXIME EVALUATION AND DISK STUDY 753 Comparative susceptibility ofgram-negative nonfermenters to cefmnenoxime, six other cephalosporis, and gentamicin % In- Organism n Drug MIC5a MIC"a hibited -_s jagl Organism MICgo ~~~~~~~~~~~~~~~~~~ ml A. calcoaceticus subsp. anitratus 15 Cefmenoxime 13 47 1377 128 4. 93 P. aeruginosa Pseudomonas spp. (group 1)b Pseudomonas spp. (group 2)C 5 Cefmenoxime 12 Cefmenoxime 18 Cefmenoxime See footnote a of Table 1. b Includes: 3 P. acidovorans and 9 P. stutzeri. 'Includes: 4 P. cepacia, 3 P. maltophilia, 6 P. fluorescens, and 5 P. putida. ble, and Enterobacteriaceae and S. aureus are susceptible (Fig. 3). Furthermore, these categories segregated the organisms in the phase 1 clinical study (Fig. 3C) in a fashion similar to that of the organisms in the regression analysis study (Fig. 3B)-thus demonstrating that the latter population was representative of current clinical isolates. MICs of cefmenoxime, cefotaxime, and moxalactam against the same organisms used for the disk study showed no cross-resistance between cefmenoxime and cefotaxime and less than 1% cross-resistance with moxalactam by using the MIC breakpoints described above (Table 1). Of 421 isolates tested, 5 were susceptible to cefotaxine (MIC, c8,ug/ml) of which 318 were susceptible, 7 were moderately susceptible 1. 4. 2..5 4. 256 >256 256 128 12 12 28 9 82 92 92 25 75 11 17 44 (MIC, to Ag/ml), and none were resistant (MIC, > ug/ml) to cefmenoxime. Only one of these was resistant to moxalactam. There were 38 strains resistant to cefotaxine, of which 33 were also resistant to cefinenoxime and 5 were moderately susceptible to this drug. Three of these isolates were susceptible to moxalactam; all three were strains of Pseudomonas maltophilia. DISCUSSION The in vitro activity of cefmenoxime in this report (Tables 1 through 6), as well as that of a previous report (2), was considerably greater than that of cephalothin and cefamandole, and was remarkably similar to the activity reported for cefotaxime (1, 6, 7) and moxalactam (1, 3, 9,

TABLE 6. _ Comparative susceptibility ofgram-positive cocci and N. meningitidis to cefmenoxime, six other cephalosporins, and gentamicin % In- Organism n Drug MICwa MIC,b athibited lg/ ml 754 FUCHS ET AL. ANTimICROB. AGENTS CHEMOTHER. S. aureus (methicillin susceptible) S. aureus (methicillin resistant) S. faecalis S. pneumoniae S. pyogenes N. meningitidis a See footnote a of Table 1. 48 Cefmenoxime 1 Cefmenoxime 25 Cefmenoxime 2 Cefmenoxime 2 Cefmenoxime 25 Cefmenoxime 4. 2. 2..25.25.25 128 128 256 >256.15.15.12.25 O.12.15 s.8.12.5.25 SO.12 5.8 5.8 5.8 5.8 so.12 so.12 5.12 2. 2. 4. 4..5.5 128 256 256 256 >256 >256.6.12.25 2..25 SO.12.3.3 2..12.25 so.12 s.8 s.8 5.8.15 s.12 s.12 so.12 5 2 1 1 6 6 4 4 4 8 11 21, 22). Whereas the Enterobacteriaceae, nonenterococcal streptococci, and S. aureus were highly susceptible to cefmenoxime, the majority of P. aeruginosa and Acinetobacter sp. required to jig/ml for inhibition. Conversely, these two species constituted the bulk of all organisms in this moderately susceptible range (Fig. 3). Because these levels of cefmenoxime in serum

VOL. 2, 1981 are easily exceeded (19), and because this MIC range for cefotaxime (6) and moxalactam (3) is considered indeterminate or moderately susceptible, the same is tentatively applied to cefmenoxime. That this category includes primarily P. aeruginosa and Acinetobacter sp. may be significant in clinical evaluations when the efficacy of the moderately s(iceptible category concept is tested. Despite the considerable similarity of in vitro TABLE 7. Comparative susceptibility of 48 anaerobic isolates to cefnenoxime, cefamandole, cefotaxime, moxalactam, and cefoxitin % In- Organism n Drug MICsoa MICra MIC9 at d8 CEFMENOXIME EVALUATION AND DISK STUDY 755 MI9ahibited jg/ml B. fragilis Cefmenoxime 44 group b 69 81 69 Bacteroides 18 Cefmenoxime 2. 83 spp. 56 83 4. 83 2. Clostridium 6 Cefmenoxime 4. 67 spp.d.5 67 67 2. 5 67 Gram-positive 8 Cefmenoxime.5 4. cocci.5 88.5 4..5 2. 88 a See footnote a of Table 1. 'Includes: 11 B. fragilis, 2 B. distasonis, 2 B. ovatus, and 1 B. thetaiotaomicron. c Includes: 3 B. melaninogenicus and 15 Bacteroides spp. NOS. d Includes: 2 C. difficile, 1 C. perfringens, and 3 Clostridium spp. NOS. activity of these third-generation cephalosporins, preliminary pharmacological data in experimental animals suggest some potential advantages of cefmenoxime (19). Although plasma levels of cefmenoxime and cefotaxime in animals were comparable, cefmenoxime tissue levels were significantly higher than those of cefotaxime, and the half-life in both plasma and tissue was greater for cefmenoxime (19). Furthermore, cefmenoxime showed greater activity than ce- TABLE 8. Comparative MICs of cefmenoxime and other cephalosporins against beta-lactamasenegative and -positive H. influenzae and N. gonorrhoeae Organism n Drug MIC5a MIC9a H. influenzae (beta- 2 Cefmenoxime c.8.3 lactamase negative) <.8.3.3.12 co.8.3 2. 4..25.5 2. H. influenzae (beta- 2 Cefmenoxime c.8.3 lactamase positive) 5.8.3.3.6.12.25 4. > > N. gonorrhoeae 22 Cefmenoxime.8.3 (beta-lactamase S.8.3 negative) 5.8.3 5.8.12.12 4..5 4. N. gonorrhoeae 26 Cefmenoxime 5.8.15 (Beta-lactamase 5.8 s.8 positive).15.6.15.6 2. 4. 2. a See footnote a of Table 1. TABLE 9. Comparison of beta-lactamase hydrolysis of cefmenoxime with that of other cephalosporins Relative rate of hydrolysis in % (MIC)b Organism Oae Type of enzyme' Cepha epha- Cefmenoxime loridine Enterobacter sp. I, inducible 15.3 () <.1 () <.1 () E. coli III, TEM 1 57 14 () < (.6) < (.6) E. coli III, TEM 2 2 () < (.6) < (.12) Klebsiella ap. IV, inducible 69 75 () 17 (2.) 2 (4.) Bacillus cereus 87 1 (ND)C <.1 (ND) <.1 (ND) a Enzyme classifications from references and 17. b The hydrolysis rates in micromoles per minute were converted to percentage relative rates compared with that of labile substrate nitrocefin (87/312). The MICs by the reference broth microdilution procedure are in parentheses. c ND, Not determined.

756 FUCHS ET AL. LU I- -,.75 > w -5 z -25 LA. 4.4.4 A CEFMENOXIME (PM) FIG. 1. Inhibition of hydrolysis of nitrocefin by cefmnenoximne. Symbols:, E. cloacae type I;- A, E. coli type IIa;, E. coli type 111b; and A, Kiebsiella sp. type IV. 512 "Iin.4 mj x,l 4 2.5 25 12 U 3 15 N. N~~~~~~~~~~~~~~ \"I" N~~~~ N _ I -2 4412 ANTIMICROB. AGENTS CHEMOTHER. fotaxime in mice infected with several Enterobacteriaceae spp. (2). In view of the cefmenoxime structural similarity to cefotaxime at the 7 position, it is not surprising to find cefmenoxime to be comparably resistant to inactivation by bacterial beta-lactamases. When hydrolysis rates of cefmenoxime and other cephalosporins by several specific types of beta-lactamases were determined (Table 9), cefotaxime, as previously reported (22), and cefmenoximne were virtually unaffected by the enzymes tested; the MICs for the corresponding enzyme-producing bacteria were significantly lower than those of cefamandole, which was much more susceptible to these enzymes. Because of its high resistance to beta-lactamases, cefmenoxime was tested for its ability to inhibit beta-lactamase hydrolysis of a susceptible beta-lactam substrate, in this case, nitrocefin. The previously reported inhibitory spectrum of cefoxitin and cefotaxime (4, 5) was confirmed, and the spectrum of cefmenoxime was found to be essentially the same, i.e., all effectively inhibit Richmond and Sykes type I beta-lactamases, but not types II, III, IV or V. and 1*3 2 3 2 5 233 2 I 215 2 2 3 2 54 4 663 2 2 1 2 2 7 6 7 * 4' sx t X 1 2 3 I 1 4 1 4 N- 2 4 2 5 22 44 2 2 4 A N* 3 2 4 4 5 3"5,13 ZONE SIZE IN MM FIG. 2. Correlation between cefmenoxime MICs and zone diameters, using 3-ptg cefmenoxime disks. Horizontal lines represent proposed MIC resistant (upper) and susceptible (lowei) breakpoints. Vertical lines represent the proposed tentative zone size susceptible (right) and resistant (left) breakpoints. The long diagonal line represents standard regression line using all on-scale data: slope = -.458, y intercept - 11.6 (expressed as log2), and correlation coefficient = -.87. The short diagonal line represents regression line using on-scale values with MIC 2 pg/ml: slope = -.2687, y intercept = 8.8 (expressed as log2), and correlation coefficient = -.91.

VOL. 2, 1981 I- 5 25 I 1, I "1 ii /l / 1/ II 68 9n 12-14 15-17 3-2 21-22 2526WU29,31-34 F35 ZONE SIZES IN l %12 1A LUC 6 5 4 2 1 L5 54.sr'Gim. %I, 'I I I \ I I i %. ' I Id "- I:- -. 22 8 4 2 1.5.5 WC N FIG. 3. Cefmenoxime zone size (A) and MlrC (B) distribution of isolates of major organism groups used in regression analysis, and clinical isolate study (C). Symbols: (---) S. aureus, (...) S. faecalis, (---a-) nonfermenters, ( ) Enterobacteriaceae. cefmenoxime showed slight inhibition of type III-b beta-lactamase compared with cefotaxime, which exhibited none. Since moxalactam has the same inhibitory spectrum as cefotaxime (5), this is an additional property shared by the three third-generation drugs. Resistance of cefmenoxime, cefotaxime, and moxalactam to the beta-lactamases of H. influ- CEFMENOXIME EVALUATION AND DISK STUDY 757 enzae and N. gonorrhoeae is documented in Table 8, which shows that the MICs for betalactamase-positive and -negative strains are essentially the same. By contrast, cefamandole A showed considerable susceptibility to the betalactamase of H. influenzae, but not to that of N. gonorrhoeae. The slightly greater resistance of beta-lactamase-negative N. gonorrhoeae to the cephalosporins shown in Table 8 is an interesting phenomenon partially explained by the inclusion in this group of seven strains exhibiting increased resistance to several other antibiotics (possible cell membrane permeability mutants). The activity of cefmenoxime was most comparable to that of cefotaxime or moxalactam, which were shown to be the most active cephalosporins among 13 beta-lactam antibiotics previously tested against both beta-lactamase-positive and -negative H. influenzae and N. gonorrhoeae (1). The extremely low MICs of cefinenoxime and the third-generation cephalosporins against H. influenzae, N. meningitidis, and N. gonorrhoeae plus their resistance to the beta-lactamases produced by two of these species should qualify these drugs for clinical trials against infection due to these microbes. The disk diffusion zone sizes versus MIC values of cefmenoxime (Fig. 2) demonstrate a marked accentuation of the parabolic distribution of endpoints below an MIC of 1,ug/ml. Standard regression analysis utilizing all onc scale data yields a line with a rather steep slope of -.458 which crosses the 8-,ug/ml line at the 19-mm zone size. A 219-mm susceptible breakpoint would yield a 7.1% false-susceptible (very major error) rate, which is unacceptable. On the other hand, when regression analysis is performed on those values along the straighter segment of the regression curve which includes three MIC values above and below the MiC TABLE 1. Cross-resistance analysis of cefotaxime versus cefmenoxime and moxalactam with organisms used to establish disk criteria MICs (yg/ Drug MIC (tg/ ml) s8 to > Cefmenoxime c8 318c 3-7 54 5 > 1 33 c8 317 11 3-7 34 5 > 1 13 3 a Numbers in table represent numbers of isolates. A total of 421 isolates were tested.

758 FUCHS ET AL. breakpoints (i.e., 2 to 256,ug/ml), the regression line then has a slope of -.2677 and crosses the 8-,ug/ml level at 21.5 mm. This susceptible breakpoint of a22 mm is also supported by error rate-bounded analysis, which yields an acceptable.5% very major error rate. By both analytical methods, the resistant zone size breakpoint was determined to be c14 mm. The major error (false-resistant) rate was 2.1% which is less than the 5% acceptable limit. All very major and major errors became minor errors when the moderately susceptible (indeterminate) ranges (MICs of to,ug/ml and zone sizes of 15 to 21 mm) were applied, since all values 222 mm with MICs >8,ug/ml fell in the indeterminate MIC range, as did all values s14 mm with MICs s,ug/ml. The regression data and analysis are very similar to those reported for cefotaxime (6) and moxalactam (3). In fact, the suggested zone size breakpoints for cefotaxime (c14 mm, resistant; 223 mm, susceptible) differ by only 1 mm at the susceptible breakpoint from that suggested herein for cefmenoxime. With all three drugs the proportions of different bacterial species in the three susceptibility categories were similar. These features, together with the data from the cross-resistance studies, indicate that susceptibility and resistance to cefmenoxime are predictable from the results of disk diffusion susceptibility testing of cefotsxime or moxalactam. It would appear that any one of these drugs could serve successfully as the class representative for disk susceptibility testing to predict susceptibility to this grouping of third-generation cephalosporins. This assumes, of course, that the human pharmacokinetics, usual dosage regimens, and clinical efficacy against organisms in the categories of susceptible and moderately susceptible will prove to be comparable; these studies are yet to be completed for cefmenoxime. LITERATURE CIE 1. Baker, C. N., C. Thornsberry, and R. N. Jones. 198. In vitro antimicrobial activity of cefoperazone, cefotaxime, moxalactam (LY127935), azlocillin, mezlocillin, and other,b-lactam antibiotics against Neisseria gonorrhoeae and Haemophilus influenzae, including 6-lactamase producing strains. Antimicrob. Agents Chemother. 17:757-761. 2. Barry, A. L, C. Thornsberry, R. N. Jones, P. C. Fuchs, T. L Gavan, and E. H. Gerlach. 1977. Cefuroxime, an in vitro comparison with six other cephalosporins. Proc. R. Soc. Med. 7(Suppl. 9):63-7. 3. Barry, A. L, C. Thornsberry, R. N. Jones, and E. H. Gerlach. 198. Tentative interpretive standards for disk susceptibility tests with moxalactam (LY127935). Antimicrob. Agents Chemother. 18:7-721. 4. Fu, K. P., and H. C. Neu. 1978. Beta-lactamase stability of HR 756, a novel cephalosporin, compared to that of cefuroxime and cefoxitin. Antimicrob. Agents Chemother. 14:2-6. ANTIMICROB. AGENTS CHEMOTHER. 5. Fu, K. P., and H. C. Neu. 1979. The comparative betalactamase resistance and inhibitory activity of 1-oxacephalosporin, cefoxitin and cefotaxime. J. Antibiot. : 99-914. 6. Fuchs, P. C., A. L Barry, C. Thornsberry, R. N. Jones, T. L. Gavan, E. H. Gerlach, and H. M. Sommers. 198. : in vitro activity and tentative interpretive standards for disk susceptibility testing. Antimicrob. Agents Chemother. 18:88-93. 7. Hamilton-Miller, J. M. T., W. Brumfitt, and A. V. Reynolds. 1978. (HR 756): a new cephalosporin with exceptional broad spectrum activity in vitro. J. Antimicrob. Chemother. 4:437-444. 8. Jones, R. N., P. C. Fuchs, T. L Gavan, E. H. Gerlach, A. L Barry, and C. Thornsberry. 1977. Cefuroxime, a new parenteral cephalosporin: collaborative in vitro susceptibility comparison with cephalothin against 5,887 clinical bacterial isolates. Antimicrob. Agents Chemother. 12:47-5. 9. Jones, R. N., P. C. Fuchs, H. M. Sommers, T. L. Gavan, A. L Barry, and E. H. Gerlach. 198. (LY127935), a new semisynthetic 1-oxa-betalactam antibiotic with remarkable antimicrobial activity: in vitro comparison with cefamandole and tobramycin. Antimicrob. Agents Chemother. 17:75-756. 1. Jones, R. N., P. C. Fuchs, C. Thornsberry, and N. Rhodes. 1978. Antimicrobial susceptibility tests for anaerobic bacteria. Comparison of Wilkin-Chalgren agar reference method and a microdilution method, and determination of stability of antimicrobics frozen in broth. Curr. Microbiol. 1:81-87. 11. Metzler, C. M., and R. M. DeHaan 1974. Susceptibility tests of anaerobic bacteria: statistical and clinical considerations. J. Infect. Dis. 13:588-594. 12. National Committee for Clinical Laboratory Standards Subcommittee for Antimicrobial Susceptibility Tests. 1975, 1979. Performance standards for antimicrobial disc susceptibility tests, approved standard, M2-A2. National Committee for Clinical Laboratory Standards, Villanova, Pa. 13. National Committee for Clinical Laboratory Standards Subcommittee for Anthimcrobial Susceptibility Tests. 198. Standard methods for dilution antimicrobial susceptibility tests for bacteria which grow aerobically, proposed standards, M7-P. National Committee for Clinical Laboratory Standards, Villanova, Pa. 14. Neu, H. C. 198. Antibiotic inactivating enzymes and bacterial resistance, p. 454-473. In V. Lorian (ed.), Antibiotics in laboratory medicine. The Williams & Wilkins Co., Baltimore. 15. O'Callaghan, C. H., A. Morris, S. ML Kirby, and A. H Shingler. 1972. Novel method for detection of betalactmases by using a chromogenic cephalosporin substrate. Antimicrob. Agents Chemother. 1:283-288.. Richmond, M. HL, and R. B. Sykes. 1973. The betalactamases of gram-negative bacteria and their possible physiological role. Adv. Microb. Physiol. 9:3148. 17. Syke8, R. B., and M. Matthew. 1976. The beta-lactamases of gram-negative bacteria and their role in resistance to beta-lactam antibiotics. J. Antimicrob. Chemother. 2:115-157. 18. Thornsberry, C., C. N. Baker, and R. N. Jones. 1979. In vitro antimicrobiol activity of piperacillin and seven other beta-lactam antibiotics against Neisseria gonorrhoeae and Haemophilus influenzae including betalactamase producing strains. J. Antimicrob. Chemother. 5:137-142. 19. Tsuchiya, K., Y. Kita, I. Yamaaki, K Kondo, Y. Noji, and T. Fugono. 198. Absorption, distribution and excretion of cefmenoxime (SCE-1365), a novel broad-spectrum cephalosporin, in mice, rats, rabbits and dogs. J. Antibiot. 33:15-1544. 2. Tsuchiya, K., M. Kondo, M. Kida, M. Nakao, T. Iwahi,

VOL. 2, 1981 T. Nishi, Y. Noji, M. Takeuchi, and Y. Nozaki. 1981. Cefmenoxime (SCE-1365), a novel broad spectrum cephalosporin: in vitro and in vivo antibacterial activities. Antimicrob. Agents Chemother. 19:56-65. 21. VanLanduyt, H. W., and M. Pyckavet. 1979. In vitro activity of cefotaxime against cephalothin-resistant clinical isolates. Antimicrob. Agents Chemother. : 19-111. CEFMENOXIME EVALUATION AND DISK STUDY 759 22. Watanakunakorn, C., and C. Glotzbecker. 1979. Comparative in vitro activity of LY127935 (659-S), seven cephalosporins, cefoxitin and cefotaxime. J. Antibiot. :119-124. 23. White, L A., and D. S. Kellogg, Jr. 1965. Neisseria gonorrhoeae identification in direct smears by a fluorescent antibody-counterstain method. Appl. Microbiol. 13:171-174.