Washington, John A.

Data derived from testing the bactericidal activity of combinations of penicillin with gentamicin or streptomycin and of clindamycin with gentamicin on nine isolates of Streptococcus mutans were analyzed by preparing isobolograms to determine the presence of additive, synergistic, or antagonistic effects. Synergy with penicillin-aminoglycoside combinations was found in two strains; additive effects occurred in seven instances with penicillin-gentamicin combinations; and antagonism occurred in eight instances with clindamycin-gentamicin combinations.

The in vitro activities of verdamicin and gentamicin were studied in parallel against 1,049 bacterial isolates. Verdamicin demonstrated activity similar to that of gentamicin against members of the family Enterobacteriaceae and against Pseudomonas aeruginosa at 2 and 8 μg/ml, respectively. Proteus rettgeri and Providencia stuartii were notably more susceptible to verdamicin. The new aminoglycoside was also highly active against staphylococci but was not effective against group D streptococci.

Minimal inhibitory concentrations of clindamycin, minocycline, metronidazole, penicillin, and carbenicillin were determined by agar dilution against 150 recent clinical isolates of anaerobic bacteria. Ninety-nine percent of Bacteroides fragilis and all B. melaninogenicus, Clostridium perfringens , and Fusobacterium were inhibited by clindamycin at 3.1 μg/ml. Only 58% of other clostridial species were inhibited by this concentration of clindamycin. Minocycline at 3.1 μg/ml inhibited 72% of C. perfringens , 81% of other Clostridium species, and 66, 75, and 100% of B. fragilis, B. melaninogenicus , and Fusobacterium , respectively. Metronidazole at 12.5 μg/ml inhibited all bacteria tested. B. fragilis was resistant to both penicillin and carbenicillin at 6.2 μg/ml. Concentrations of 25 μg/ml for penicillin and 100 μg/ml for carbenicillin were needed to inhibit more than 90% of B. fragilis . Organisms other than B. fragilis were moderately or extremely susceptible to the penicillins.

Kanamycin, BB-K8, and gentamicin were tested in parallel against 1,037 bacterial strains isolated from clinical material. The activity of BB-K8 at 20 μg/ml was comparable to that of gentamicin at 8 μg/ml against Pseudomonas aeruginosa and against Enterobacteriaceae , with the exception of gentamicin-resistant strains of Proteus rettgeri and Providencia stuartii , in which case the activities of BB-K8 and kanamycin were the same. BB-K8 exhibited little or no activity against streptococci. The activity of BB-K8 was affected by pH and inoculum size. A regression graph for inhibition data with a 10-μg disk of BB-K8 was developed, and synergistic activity of penicillin and BB-K8 against Streptococcus faecalis was tested.

The in vitro inhibitory and bactericidal activities of spectinomycin hydrochloride were tested against a variety of bacteria. The antibiotic was inhibitory at 31.2 μg/ml to most strains of Escherichia coli, Klebsiella, Enterobacter , and Staphylococcus epidermidis . Concentrations of antibiotic exhibiting bactericidal activity exceeded the inhibitory concentration by at least fourfold. Regression graphs were plotted for results obtained with 30-, 100-, 200-, and 300-μg spectinomycin discs; tentative interpretative standards are proposed.

Because excessively high rates of false resistance have been encountered with the 10-μg amikacin disk in diffusion susceptibility tests, a study was performed to examine existing zone diameter interpretative criteria and to compare the accuracy of 10- and 30-μg amikacin disks by the error rate-bounded classification scheme. Although current zone diameter interpretative criteria eliminate false susceptibles, there is an unacceptably high rate of false resistants. This problem can be resolved in most instances by revising the zone diameter interpretative criteria for the 10-μg disk (resistant, ≤9 mm; indeterminate, 10 to 11 mm; susceptible, ≥12 mm) or, preferably, by replacing the 10-μg disk with a 30-μg disk and adopting new interpretative criteria (resistant, ≤14 mm; indeterminate, 15 to 16 mm; susceptible, ≥17 mm). Because of significant differences in performance among media, it is necessary to include Pseudomonas aeruginosa ATCC 27853 among controls routinely tested and to exclude from use lots of Mueller-Hinton agar yielding results outside the 75% tolerance (90% confidence) limits for amikacin.

A multi-center study of 1,838 clinical isolates established the accuracy of diffusion susceptibility tests with 30-μg cephalothin disks and 30-μg cefamandole disks. The same interpretive zone standards can be applied to tests with either disk but the two drugs cannot be tested interchangeably.

The penetrations of methicillin, oxacillin, and cephalothin into cortical bone and synovial tissues were studied 1 h after their intravenous administration in 105 patients having arthroplasty of the hip. Although the lowest serum levels were noted with cephalothin ( P < 0.01), more patients receiving cephalothin achieved osseous drug levels inhibitory to staphylococci ( P < 0.01). Differences in the penetration of the three agents into synovial tissues were not statistically significant.

The effect of combinations of gentamicin with trimethoprim or sulfamethoxazole-trimethoprim against clinical isolates of Escherichia coli (11 strains) and Klebsiella pneumoniae (12 strains) was examined by using a microdilution checkerboard technique. All isolates were susceptible to each antimicrobial agent. Synergism, defined as at least a 2-log 2 -dilution lowering of the minimal inhibitory concentration of either antibiotic in the combination compared with the minimal inhibitory concentration of the antibiotic alone, was observed with 15 of 23 (65%) isolates tested against trimethoprim and gentamicin and 14 of 23 (61%) isolates tested against sulfamethoxazole-trimethoprim and gentamicin. A 3-log 2 -dilution lowering of the minimal inhibitory concentration of either antibiotic was observed in 7 of 23 (30%) trimethoprim and gentamicin trials and 3 of 23 (13%) sulfamethoxazole-trimethoprim and gentamicin trials. Antagonism was observed in 3 of 46 combination trials and only with strains of K. pneumoniae .

Fifty-one clinical isolates of ticarcillin-resistant, gram-negative bacilli were tested for susceptibility to combinations of ticarcillin and clavulanic acid (BRL 14151), a potent beta-lactamase inhibitor. Minimal inhibitory concentrations (MICs) were measured by a microdilution method, and minimal bactericidal concentrations for selected strains were measured by the broth dilution method. Ticarcillin MICs were ≥128 μg/ml for all and ≥512 μg/ml for 38 (75%) of the strains. Thirty-nine strains of Enterobacteriaceae tested included Escherichia coli (14), Klebsiella pneumoniae (16), Citrobacter sp. (3), Enterobacter sp. (3), Salmonella enteritidis (1), Serratia marcescens (1), and Proteus mirabilis (1). Ticarcillin MICs for 34 strains (88%) were lowered at least threefold by the addition of 1.0 μg of clavulanic acid per ml. Against 33 strains (85%), the MICs were 64 μg or less per ml in the presence of 5 μg of clavulanic acid per ml. In contrast, the MICs for seven of eight strains of Pseudomonas aeruginosa were unaffected by the addition of up to 10 μg of clavulanic acid per ml. Ticarcillin with 5 μg of clavulanic acid per ml was bactericidal against ticarcillin-resistant (MIC ≥ 2,048 μg/ml) E. coli, K. pneumoniae, Enterobacter , and P. mirabilis .