Antimicrobial activity and spectrum of LY146032, a lipopeptide antibiotic, including susceptibility testing recommendations

LY146032, a new lipopeptide, was found to have a spectrum of gram-positive antimicrobial activity that includes activity against staphylococci (methicillin susceptible and resistant), beta-hemolytic Streptococcus spp., pneumococci, viridans group Streptococcus spp., anaerobic gram-positive cocci, Clostridium spp., and enterococci. The new lipopeptide was generally bactericidal and showed more rapid killing of Listeria spp. (MIC, 1 to 2 micrograms/ml) and staphylococci than either vancomycin or teicoplanin. The 30-micrograms disk was preferred to the 15-micrograms disk on the basis of the preliminary interpretive criteria for susceptibility which indicated zone diameters of greater than or equal to 16 mm for susceptible strains (MIC, less than or equal to 2.0 micrograms/ml) and greater than or equal to 12 mm for resistant strains (MIC, greater than or equal to 8.0 micrograms/ml). These criteria are valid pending the testing of additional gram-positive strains which have LY146032 MICs of greater than or equal to 8 micrograms/ml.

Broth microdilution test panels (Prepared Media Laboratory, Tualatin, Oreg.) were prepared, and tests were performed by the method of the National Committee for Clinical Laboratory Standards (NCCLS) (9). Anaerobic organisms were tested in Wilkins-Chalgren broth (WCB) by the microdilution method described by Jones et al. (5). Agar diffusion disk tests were also performed by NCCLS procedures (8). Interpretive zone standards were selected by correlating zone diameters with MICs by using both regression analysis (method of least-squares) and the error-rate-bounding method of Metzler and DeHaan (7). LY146032 MICs were determined in Mueller-Hinton broth (MHB) supplemented with calcium and magnesium (9). A lot of WCB without agar was also used to determine LY146032 MICs for both the anaerobic bacteria and the facultative strains (5).
The MBC was defined as the lowest concentration producing a .99.9% reduction in the CFU from the initial 5 x 105 CFU/ml inoculum concentration at 24 h. The CFU per milliliter determinations were done by the method of Pearson et al. (11), by replicate subculturing of l0-pul inoculum volumes to drug-free agar plates. MBCs were determined for 25 strains at 24 h only, and time-kill curves were done for 15 isolates at 0, 4, and 24 h of incubation. The effects on LY146032, vancomycin, and teicoplanin MICs of altering the inoculum concentrations were determined by using inoculum sizes of approximately 104 5 x 105, and 107 CFU/ml (see Table 3).
shown in Table 1. An additional 20 facultative gram-negative bacilli were tested, and the LY146032, vancomycin, and teicoplanin MICs were all >32 ,ug/ml, i.e., the bacilli were categorically resistant (data not shown). With the 75 methicillin-resistant strains, LY146032 had only a slightly higher MIC for 90% of the strains tested (MIC90), MIC range, and geometric mean MIC. All staphylococcal isolates were susceptible to LY146032, vancomycin, and coumermycin. Imipenem MICs for the methicillin-resistant strains were 190-fold higher (mean MIC, 3.8 ,ug/ml) than those for the methicillin-susceptible isolates. Teicoplanin inhibited 93.1% of methicillin-resistant Staphylococcus strains at s4.0 ,ug/ml; the MICs for all of the remaining isolates were 8.0 or 16 ,ug/ml. The highest LY146032 MICs were for the E. faecium (MIC range, 0.5 to 2.0 ,ug/ml) and E. faecalis (MIC range, 0.5 to 1.0 ,ug/ml) isolates. L. monocytogenes was inhibited by very low concentrations (MIC9o, 0.03 ,ug/ml) of coumermycin. LY146032 was also tested against 20 strict anaerobes (data not shown), with the following results for MIC for 50o of strains tested (MIC50): B. fragilis, >32 ,ug/ml; C. perfringens, 4.0 ,ug/ml; C. sporogenes, 8.0 p.g/ml; and anaerobic gram-positive cocci, 0.12 jxg/ml. Effect of medium on activity. LY146032 MICs were deter-mined in three broth media ( Table 2). The LY146032 MICs with cation-supplemented Mueller-Hinton broth (CSMHB) (9) were 16-to 64-fold lower than those produced in unsupplemented MHB or in the cation-deficient anaerobe testing broth, WCB. Comparisons of LY146032 MICs in different media by regression statistics produced results as follows: CSMHB versus MHB, r = 0.81 and y = -0.06 + 0.60x; CSMHB versus WCB, r = 0.77 and y = 0.46 + 0.57x; and MHB versus WCB, r = 0.92 and y = 1.44 + 0.89x. These  data imply an LY146032 MIC which is an average of 64-fold lower with CSMHB. Inoculum studies and MBCs. By geometric mean comparisons, LY146032 was bactericidal for all genera except the enterococci (Table 3). The LY146032 MBCs were at least 10-fold greater than the mean MICs. Increasing inoculum concentrations generally had little effect on the MICs of the drug when inocula of 104 and 5 x i05 CFU/ml were compared. However, the LY146032 MICs increased greater than fivefold with the 107 CFU/ml inoculum concentration for the enterococci and Staphylococcus strains.
Time-kill studies. LY146032 was rapidly bactericidal (inoculum reduction of .2 log10 CFU at 4 h) for 13 of 15 strains and bactericidal (reduction of -3 logl0 at 24 h) against all tested isolates (Table 4). This killing activity was generally enhanced by the addition of gentamicin at therapeutic levels (4 ,ig/ml). In contrast, vancomycin alone at therapeutic concentrations (8.0 p.g/ml) demonstrated slower bactericidal action.
Disk diffusion tests. Scattergrams for the 15-and 30-,ugdisk zone diameters and LY146032 MICs determined in CSMHB are shown in Fig. 1. All LY146032 MICs for gram-positive organisms were s2.0 jig/ml, and the zones were .13 or .16 mm for the 15-and 30-pug disks, respectively. For all gram-negative strains tested with LY146032, MICs were >32 ,glml, and no zones were produced around the disks. Regression statistics were poor because of the widely separated populations of bacteria which are considered to be either within (susceptible) or outside (resistant) the spectrum of LY146032. Similarly, error-rate bounding could assure only low, minor, and false-resistant errors without absolutely minimizing the false-susceptible result, e.g., rare resistant gram-positive isolates. Organisms for which MICs are in the indeterminate or resistant ranges have not been widely identified by the CSMHB microdilution method. Absolute spectrum agreement (all 313 strains) was observed between LY146032 and vancomycin for the 293 gram-positive organisms and for the 20 selected gramnegative strains. A 98.4% interpretive agreement was seen for LY146032 and teicoplanin with interpretive variations from coagulase-negative, methicillin-resistant staphylococci categorized as susceptible to LY146032 and resistant or indeterminate to teicoplanin.
The NCCLS susceptibility tests confirmed that the LY146032 spectrum includes nearly all clinically important gram-positive organisms (2-4, 6). The organisms for which the LY146032 MICs were closest to the proposed susceptibility breakpoint of s2.0 ,ug/ml were E. faecium (MIC90, 2.0 ,ug/ml), E. faecalis (MIC90, 1.0 jig/ml), and L. monocytogenes (MIC90, 2.0 ,ug/ml). The results for antimicrobial activity and spectrum were most similar to results previously found for glycopeptides, vancomycin, and teicoplanin (1-4, 6, 10). In this study, the clinically popular antistaphylococcal agent vancomycin was observed to have slower bactericidal activity against the staphylococci, regardless of methicillin susceptibility. LY146032 activity was greatly reduced (16to 64-fold) by the absence of physiologic concentrations of divalent cations in the test media (2). Since the NCCLS procedures recommend the presence of these concentrations, we believe that LY146032 activity will be accurately assessed in most clinical laboratories (9). When LY146032 is tested against anaerobes in a WCB medium, supplemental calcium should be added to produce accurate MICs for the gram-positive anaerobic species. Selection of in vitro susceptibility test breakpoints for the glycopeptide drugs has'been difficult because of the rarity of truly resistant gram-positive strains (1). Therefore, we suggest LY146Q32 MIC and zone diameter breakpoints for all gram-positive species that should be considered susceptible and which also represent a cons'ervative application of early human pharmacokinetic results' (H. R. Black, G. L. Brier, J. D. Wolny, and E. H. Nyhart, Jr., ICAAC, abstr. no. 894, 1986). The preliminary criteria for LY146032'tested by NCCLS standardized methods favor the use of a 30-,ug disk (the same as for vancomycin and teicoplanin). The criteria indicate zone diameters of -16 mm (MIC correlate, <2.0 ,ug/ml) for susceptible isolates and <12 mm (MIC correlate, .8.0 ,ug/ml) for resistant isolates (8,9). As gram-positive strains with resistant MICs emerge, these tentative LY146032 interpretive criteria may require revision.
We thank R. R. Packer, D. Preston, and C. Thornsberry for suggestions regarding the manuscript and for contributions of data. The word processing was provided by Jacqueline McClung.