ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Nov. 1978, p. 643-649 0066-4804/78/0014-0643$02.00/0 Copyright 1978 American Society for Microbiology Vol. 14, No. 5 Printed in U.S.A. Susceptibility of the Anaerobic Bacteria, Group D Streptococci, Enterobacteriaceae, and Pseudomonas to Semisynthetic Penicillins: Carbenicillin, Piperacillin, and Ticarcillin GILLES R. G. MONIF,1 PENELOPE R. CLARK,' JOHN J. SHUSTER,2 AND HERMAN BAER3 Laboratory of Infectious Diseases, Departments of Obstetrics and Gynecology,' Pathology,3 and Biostatistics,2 University of Florida College of Medicine, Gainesville, Florida 32610 Received for publication 11 August 1978 Sodium piperacillin T-1220, a new semisynthetic penicillin, was tested in vitro against 297 clinical isolates of anaerobic bacteria and 669 aerobic bacteria by the conventional agar dilution method and compared with carbenicillin and ticarcillin. At a 100-jig/ml concentration the three drugs showed comparable effectiveness against the anaerobes tested. However, at 20 jig/ml, piperacillin was the most effective drug against Bacteroides fragilis, peptostreptococci, and group D streptococci. At this drug concentration only 48% of the B. fragilis strains exhibited susceptibility to carbenicillin only, 64% exhibited susceptibility to ticarcillin but 90% exhibited susceptibility to piperacillin. Similar findings were observed with peptostreptococci and group D streptococci. On a weight basis piperacillin was statistically shown to be the most effective antibiotic of the three tested against these anaerobes. At 20 jig/ml, piperacillin exhibited a statistically significant difference (P < 0.01) over carbenicillin and ticarcillin for Serratia marcescens, Escherichia coli, Klebsiella species, Klebsiella pneumoniae, Pseudomonas isolates, and Citrobacter diversus. At both 20- and 100-jig/ml concentrations, piperacillin appeared to be the most effective (calculated P < 0.01) upon Klebsiella species, K. pneumoniae, S. marcescens, and C. freundii in activity over ticarcillin and carbenicillin. The predominance of polymicrobial infection in obstetrics and gynecology has prompted the development of three-drug therapy for serious life-threatening infection. This regimen developed out of the necessity to provide antibiotic coverage for four major categories of bacteria: (i) the penicillin-susceptible anaerobic bacteria; (ii) the non-penicillin-susceptible Bacteriodiaceae; (iii) the enterococci; and (iv) the Enterobacteriaceae (3). The broad-spectrum semisynthetic penicillins exemplified by carbenicillin, by virtue of the penicillin coverage against anaerobic bacteria and most of the enterococci and partial efficacy at 100 jig against many of the non-penicillin-susceptible Bacteroidiaceae and Enterobacteriaceae, come closest to the required spectrum of coverage of any single antibiotic (1, 6-8). The principal problem associated with single-drug therapy with the carbenicillinlike broad-spectrum semisynthetic penicillins is that a significant number of isolates in three of the four categories exhibited relative or absolute resistance, making it difficult in case of antibiotic failure to identify the "hole" in the antibiotic 643 coverage. The purpose of this study was to compare the susceptibility of 297 strains of anaerobic bacteria, including strains of Bacteroides fragilis and 50 strains of group D streptococci (enterococci), and 669 strains of Enterobacteriaceae and Pseudomonas to carbenicillin, ticarcillin, and a new broad-spectrum semisynthetic penicillin, piperacillin (T-1220). MATERIALS AND METHODS Anaerobic bacteria. A total of 297 anaerobic isolates from clinical specimens obtained at Shands Teaching Hospital were tested. These included 50 Bacteroides species; 50 B. fragilis, including the five subspecies B. fragilis subsp. ovalis, B. fragilis subsp. thetaiotaomicron, B. fragilis subsp. vulgatus, B. fragilis subsp. fragilis, and B. fragilis subsp. diatasonis; 50 peptostreptococci; 25 peptococci; 50 Clostridium perfringens; 10 Veillonella; 11 Propionibacterium; and 51 group D streptococci. These were grown on anaerobic blood agar plates and thioglycollate broth, tested for purity, and stored at -70 C until the time of testing. Aerobic bacteria. A total of 669 isolates from clinical specimens obtained at Shands Teaching Hos-
644 MONIF ET AL. pital were tested. The gram-negative isolates selected were: 25 Escherichia coli; 52 E. coli resistant to ampicillin and/or cephalothin and/or aminoglycosides and/or tetracycline; 31 Klebsiella species resistant to ampicillin and carbenicilin; 24 Klebsiellapneumoniae resistant to cephalothin and/or aminoglycoside in addition to carbenicillin and ampicillin; 44 Serratia marcescens; 31 Pseudomonas species; 105 Pseudomonas aeruginosa susceptible to gentamicin and tobramycin; 102 P. aeruginosa resistant to gentamicin and/or tobramycin; 22 Proteus vulgaris; 3 Proteus mirabilis with multiple-antibiotic resistance; 10 Providencia stuartii; 60 Enterobacter species resistant to ampicillin, cephalothin, tetracycline, and furadantin; 18 Enterobacter species resistant to aminoglycosides and/or tetracycline, chloramphenicol, and sulfamethoxaoltrimethoprim; 48 Citrobacter diversus; and 6 Citrobacter freundii. Resistance to a given antibiotic was determined by the Kirby-Bauer method. These organisms were grown on blood agar plates and Trypticase soy broth, and then isolated and identified by conventional bacteriological tests. Control bacteria included Staphylococcus aureus ATCC 25923, E. coli ATCC 25922, and P. aeroginosa ATCC 27853. Antibiotics. Piperacillin was obtained from Lederle Laboratories; carbenicillin was from the Roerig Division of Pfier Laboratories; and ticarcillin was from Beecham Laboratories. Stock solutions of 25,000 pug of piperacillin, ticarcillin and carbenicillin per ml were prepared and divided into 2.5-ml portions and froen at -20 C. Just before use, a portion was thawed and discarded after use. Dilutions used for each antibiotic were as follows: 0.1, 0.5, 1.0, 5, 10, 20, 50, 100, 200 pg/ml per plate. Test procedure. (i) Anaerobic bacteria. Susceptibility testing was performed by the conventional agar plate dilution method for determining minimum inhibitory concentrations. The plating medium consisted of Mueller-Hinton agar, ph 7.4, enriched with 5% yeast extract and 5% lysed sheep erythrocytes. Dilutions of antibiotics were incorporated into the molten agar medium before pouring to yield plates with the appropriate antibiotic concentration. Cultures to be tested were grown for 48 h in thioglycolate broth supplemented with hemin (5 pg/ml) and menadione (0.5 pug/ml). These were diluted in thioglycolate broth to a turbidity equivalent to a McFarland no. 1 barium sulfate standard and used to fill the wells of a Steers replicator. The plates were inoculated and incubated in anaerobic jars at 370C for 48 h, and then examined for the presence or absence of growth. The lowest concentrations of antibiotic producing complete inhibition of growth were taken as the minimum inhibitory concentrations. All anaerobic procedures were performed in an anaerobic glove box flled with a gas mixture of 85% nitrogen, 5% carbon dioxide, and 10% hydrogen. (ii) Aerobic bacteria. Serial dilutions of stock antibiotics were prepared in Trypticase soy broth. Suffi'cient amounts of each antibiotic dilution were incorporated in Trypticase soy agar cooled to 50 C and poured into plates. The inoculum sie was adjusted to contain 108 organisms per ml and applied onto the surface of the antibiotic-containing plate by the Steers replication method. Plates were incubated at 370C for 16 to 18 h and examined for presence or ANTIMICROB. AGENTS CHEMOTHER. absence of growth. The lowest concentration of antibiotic producing complete inhibition of growth was accepted as the minimum inhibitory concentration. Statistical analysis. The overall significance was determined by the Shuster-Downing chi-square method and then repeated for each statistical hypothesis at two levels provided the overall test was significant (5). This method enabled the simultaneous comparison of the three drugs at 20 and 100 pg/ml and then adjusted for cluster effects of the isolate, whereas the Pearson chi-square did not. Given that significance was established at a given dosage, a multiple comparison test of Shuster-Boyett (4) was run for each organism to test for the superiority of piperacillin. Under the null hypothesis that the three drugs are equivalent, the P value was the probability of observing a margin of superiority of piperacillin (over the other drugs) as large or larger than the actual margin of superiority. RESULTS In a 100-pg/ml concentration the three drugs piperacillin, ticarcillin and carbenicillin showed comparable effectiveness in vitro against the anaerobes tested (Table 1). The majority of the anaerobic isolates required minimum inhibitory concentrations of 5 ug of each drug per ml. At levels as low as 20 jig/ml, close to 100% of strains were inhibited, except with B. fragilis, the peptostreptococci, and the D streptococci. At this drug concentration 48% of the strains of B. fragilis were susceptible to carbenicillin, and 64% were susceptible to ticarcillin, but 90% were susceptible to piperacillin (Fig. 1). At 20,ug/ml, 72% of strains of peptostreptococci were susceptible to carbenicillin, 86% were susceptible to ticarcillin, and 100% were susceptible to piperacillin (Fig. 2). Similar findings were obtained with group D streptococci (Fig. 3). Piperacillin was the most active drug; 98% of the isolates were susceptible at concentrations of 20,ug/ml, whereas only 80% were susceptible to carbenicillin and ticarcillin at this concentration. The differences were even more pronounced at concentrations lower than 20 pg/ml. On a weight basis piperacillin was statistically (calculated P < 0.01) shown to be the most effective of these antibiotics against B. fragilis, peptostreptococci, and the enterococci. Figure 4 is a bar graph representing the cumulative percentage of aerobic bacteria that were susceptible to 20 and 100 pag of ticarcillin, carbenicillin, and piperacillin per mil. At concentrations of 100 pg/ml these three antibiotics were equally effective in inhibiting P. aeruginosa (resistant to gentamicin and/or tobramycin), P. aeruginosa (susceptible to tobramycin and gentamicin), P. vulgaris, Enterobacter species, P. stuartii, C. freundii, and E. coli. (For P. stuartii and C. freundii, the number of isolates tested was less than 18 and not included in the statis-
VOL. 14, 1978 TABLE 1. SUSCEPTIBILITY OF ANAEROBES TO PENICILLINS 645 Susceptibilities of anaerobic strains to piperacillin, carbenicillin, and ticarcillin in concentrations of 20 and 100 pg/ml Cumulative % of strains susceptible at: Organism No. of strains 20 Ag/ml 100,ug/rnl Piperacillin Carbenicillin Ticarcillin Piperacillin Carbenicillin TicarcilWin Bacteroides sp. 50 98 96 92 100 100 100 B. fragilis 25 88 40 56 88 88 88 Peptostreptococci 50 100 72 70 100 100 100 Peptococci 25 100 100 100 100 100 100 C. perfringens 50 100 92 92 100 100 100 Enterococci (group D 51 98 80 80 100 98 98 streptococci) Veillonella 10 100 100 100 100 100 100 Propionibacterium 11 100 100 100 100 100 100 o 90-90 90 T 80. 70 60-~ ~ ~ 0~ ~ ~ ~ ~ ~ us w~ 50 ~ 0 48 4 90* 30 40 ~~~~~40 a. o 20- w~~~~~ 10 12 ~BCEO DES FRGLI8 (0 0.1 0.5 0 500 20 5'0 l'00 > 00 20'0 > 200 MINIMUM INHIBITORY CONCENTRATION, MICROGRAM PER ML. BACTEROIDES FRAGILIS - (50) FIG. 1. Susceptibility of 50 strains of B. fragilis to piperacillin T-1220, ticarcillin, and carbenicillin. tical analysis.) Close to 100% of these strains exhibited susceptibility to the 100-,ug/ml level, with the exception of one set of Enterobacter (18 isolates tested). Differences in susceptibility to the three semisynthetic penicillins were distinctly obvious at 20,tg/ml. At the 20-,ug/ml level, piperacillin exhibited a statistically significant difference (P < 0.01) over the two other drugs against S. marcescens, E. coli, Klebsiella species, K. pneumoniae, all Pseudomonas isolates, and C. diversus. At 20 and ltg/ml, 100 piperacillin appeared to be the most effective (calculated P < 0.01) on Klebsiella species, K. pneumoniae, and S. marcescens and C. freundii in activity over both ticarcillin and carbenicillin. The differences in activity of the three antibiotics against P. aeruginosa and K. pneumoniae are demonstrated in Fig. 5 and 6. Even in a concentration as low as 5,ug, piperacillin showed distinctly greater inhibition than did ticarcillin and carbenicillin. Piperacillin was twice as active against Pseudomonas species at this level and at least ten times more potent against Klebsiella. This inhibitory effect continued to increase until a maximum difference was attained at 20,tg/ml. Activity at this low level is 100
646 MONIF ET AL. ANTIMICROB. AGENTS CHEMOTHER. 0 a: IL 0 -J C) GI 0.5 1.0 5.0 1b 20 50 100 100 260 MINIMUM INHIOTING CONCENTRATION, MICROGRAM PER ML. PEPTOSTREPTOCOCCI - (50) FIG. 2. Susceptibility of50 strains ofpeptostreptococci topiperacillin T-1220, ticarcillin, and carbenicillin. 100 90. 80. 60-1-- CL 0. 30 200 Susceptibility of 51 strains ofgroup D streptococci to piperacillin T-1220, ticarcillin, and carbenii- FIG. 3. cit/in. 0.1 0.5 1.0 5.0o 10 20 100i >10 200io >200o MINIMUAM INHIBITORY CONCENTRATION, MICROGRAM PER ML. ENTEROCOCCUS (GROIN' D STREPTXOcOcCI- (5I)
VOL. 14, 1978 SUSCEPTIBILITY OF ANAEROBES TO PENICILLINS 647 P.perac.li.n CorbenicIllin Ticorcelin El El Z;60 20,wg- 40-20 T cccfliin [ Carbencill, n Piperocillin En 100 p P I Citrobocter Proteu Enterobocter speciesm Enterobcttr sp(l8) Providencio E coli (25) freundi. (6) vulqaris(22) resestont to oanpcilin, resistn to amino- stuortii (10) cepholothin,tetrocyclote glycosides ond/or fturadontin tetrocyclne. chlor- Omptenlcol.sulfomethosool ond trimethoprim TC~~~~~~~~~~~~~~~~~~~~~~~ p P Sersotoo E Cali (2) Kebsiella species)3)) Kiebsiell pneumon ae Psuomonas Pseudomonas Pseudomonos C trobactr morcescens resistant to to (24) resistnt to aeruginosa(102) aeruginosa (105) species (31) diversus ampl- CiliPn and/or cepitalrffunion? omnpciilln and carbenicilin ceplalotiss and/or r@5.stant togeitt- susceptible to (48) ottin an/or amino- aminoglycosides in micen and to tobra- tobramycin and glycoside and/or addition to carbeni- mycin gertamicir tetrocycln cillin and ampicillin FIG. 4. Distribution of cumulative percentage of inhibition of indiv,idual strains of Enterobacteriaceae at 20- and 100-p.g/ml concentrations of ticarcillin, carbenicillin, and piperacillin. especially advantageous because these isolates are characteried by multiple-drug resistance including resistance to some aminoglycosides. DISCUSSION On a weight basis piperacillin was statistically (calculated P < 0.01) shown to be the most effective of these antibiotics against B. fragilis, peptostreptococci, and the enterococci. Piperacillin was the most active drug; 98% of the isolates were susceptible at concentrations of 20,ug/ml, whereas only 80% were susceptible to carbenicillin and ticarcillin at this concentration. The differences were even more pronounced at concentrations lower than 20,ig/ml. At a concentration of 100 ytg/ml the three drugs were equally effective. At the 100-,ug/ml concentration, piperacillin was statistically more effective in vitro against K. pneumoniae, Klebsiella species, S. marcescens, and C. diversus. Piperacillin on a weight basis appears to have the greatest spectrum of activity among the susceptible
648 MONIF ET AL. ANTIMICROB. AGENTS CHEMOTHER. boc IV D 70..s Sc 60 0 50 a 40. 30. E 20 10 0 C1 0 5 0 5.0 10 20 50 100 200 Minimum inhibiting concentration, microgrom per ml. Pseudomonos oeruginoso(102) resistant to gentomicin and/or tobromycin FIG. 5. Susceptibility of 102 gentamicin- and/or tobramycin-resistant strains of P. aeruginosa to piperacillin T-1220, ticarcillin, and carbenicillin. 100. 90. 80 70- C c 60. -0 50 CE 451 84 iv~~~~~~~~~c, B 30 ' 29-3 E "20 7 100% 10.~~~~~~~~~~~~~~~ IC~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I 01 0 5 0 550 10 20 50 100 200 Minimim inhibiting concentration, microgrom per ml Klebsiella pneumoniae resistont to ompicillin ard corbenicillir FIG. 6. Susceptibility of ampicillin- and carbenicillin-resistant K. pneumoniae to piperacillin T-1220, ticarcillin, and carbenicillin.
VOL. 14, 1978 SUSCEPTIBILITY OF ANAEROBES TO PENICILLINS 649 strains of Enterobacteriaceae and has an augmented spectrum of coverage as it pertains to Klebsiella, where 90% of isolates are sensitive to 20,g/ml. The data reported corroborate the recently published observations of Fu and Neu and Verbist (2, 9). Piperacillin offers excellent coverage for P. aeruginosa, K. pneumoniae, and Klebsiella species without sacrificing efficacy against other members of Enterobacteriaceae. If confinned in vivo the augxnented coverage obtained with piperacillin would support its use in cases of polymicrobial aerobic-anaerobic infection. LTERATURE CITED 1. Bodey, G. P., V. Rodrigue, and J. K. Luce. 1969. Carbenicillin therapy of gram-negative bacilli infections. Am. J. Med. Sci. 257:408-414. 2. Fu, K. P., and H. C. Neu. 1978. Piperacillin, a new penicillin active against many bacteria resistant to other penicilhins. Antimicrob. Agents Chemother. 13:358-367. 3. Monif, G. R. G., S. L. Welkos, and H. Baer. 1977. The bacteriological spectrum of isolates obtained from the cul-de-sac of patients with endometritis/salpingitis/peritonitis. Excerpta Med. 3:26-32. 4. Shuster, J. J., and J. M. Boyett. 1978. Non-parametric multiple comparison procedures. J. Am. Stat. Assoc. 73:665-668. 5. Shuster, J. J., and D. J. Downing. 1976. Two-way contingency tables for complex sampling schemes. Biometrika 63:271-276. 6. Solberg, C.O., K. M. Kjellstraud, and J. M. Matsen. 1971. Carbenicillin therapy of severe Pseudomonas aeruginosa infections. J. Chronic Dis. 24:19-28. 7. Swenson, R. M., and B. Lorher. 1977. Clindamycin and carbenicillin in treatment of patients with intraabdominal and female genital tract infections. J. Infect. Dis. (Suppl.) 135:S40-S45. 8. Thadepalli, H., and T. H. John. 1977. Treatment of anaerobic infections: carbenicillin alone compared with clindamycin and gentamicin. Curr. Ther. Res. 22(4):549-555. 9. Verbist, L. 1978. In vitro activity of piperacillin, a new semi-synthetic penicillin with an unusually broad spectrum of activity. Antimicrob. Agents Chemother. 13:349-357. Downloaded from http://aac.asm.org/ on August 22, 2018 by guest