Penicillin failure in the treatment of Bacteroides fragilis lung abscess. Experimental study in rabbits.

Carbenicillin, chloramphenicol, doxycycline, and clindamycin were compared with penicillin for the treatment of lung abscess in an animal model produced by transtracheal inoculation of a mixture of anaerobes: Bacteroides fragilis, Peptococcus morbillorum, Eubacterium lentum and Fusobacterium nucleatum. Both chloramphenicol and doxycycline eliminated the bacteria but failed to close the abscess cavity. Carbenicillin, although it eradicated B. fragilis, failed to close the abscess cavity in 3 of 6 animals. In all animals tested, clindamycin sterilized the abscess cavities and healed the lung abscesses, while penicillin failed to eradicate the infection. Clindamycin was significantly more effective than penicillin in the elimination of anaerobic bacteria from the lung (p less than 0.05). Clindamycin also closed the abscess cavity faster than penicillin (p less than or equal to 0.02). The superior efficacy of clindamycin may have been the result of accumulation in the lung tissue in concentrations four- to eightfold higher than in the serum.

Animal model for anaerobic lung abscess.

There are no satisfactory animal models for the study of anaerobic lung abscess. Aspiration of food, gastric mucin, or hydrochloric acid, or any combination of these, along with oropharyngeal bacteria, is commonly believed to cause aspiration pneumonia and lung abscess. In the animal model described, none of the adjuvants was effective in producing anaerobic lung abscesses. Anaerobic bacteria derived from dental scrapings of a healthy adult (Peptococcus morbillorum, Fusobacterium nucleatum, Eubacterium lentum, and Bacteroides fragilis), when inoculated transtracheally without any adjuvants into New Zealand male white rabbits, consistently produced lung abscesses. Neither B fragilis by itself nor a mixture of P. morbillorum, F. nucleatum, and E. lentum without the addition of B. fragilis produced lung abscesses. The bacterial isolates used in this study were stored in prereduced chopped-meat-glucose medium and subcultured several times and were found effective in reproducing anaerobic lung abscesses repeatedly. This animal model is suitable for the study of pathogenesis, diagnosis, and treatment of B. fragilis-associated anaerobic lung abscess.

Is penicillin alone effective in enterococcal endocarditis? An experimental study in rabbits.

Enterococcal endocarditis in man is traditionally treated with penicillin and an aminoglycoside. Whether penicillin alone is adequate has not been fully evaluated. Experimental enterococcal endocarditis (EEE) in male New Zealand rabbits, when untreated, was fatal in all animals within 2 weeks. When crystalline penicillin G was given at 30,000 U/day in three equally divided doses, the fatality rate was 63.3%; when the dosage was increased to 600,000 U/day, the mortality ws 37.5%, demonstrating the ineffectiveness of crystalline penicillin alone in EEE. Procaine penicillin at 300,000 U/day given as a single dose provided increased protection, but was still associated with unacceptably high mortality (17.5%). A higher dose level, 600,000 U/day, of procaine penicillin protected all animals with EEE. High and sustained levels of penicillin in the serum and the myocardium cured enterococcal endocarditis in rabbits.

Antimicrobial activity and synergism of cefuroxime on anaerobic bacteria.

In vitro efficacy of cefuroxime was tested against 333 clinical isolates of anaerobic bacteria, including 80 strains of Bacteroides fragilis. Its activity was compared with that of cefamandole and cefoxitin. In addition, the antimicrobial synergism of cefuroxime was tested in combination with either penicillin or carbenicillin against B. fragilis. At 16 microgram/ml, cefuroxime inhibited 70% of all anaerobic bacteria, but it was ineffective against 64% of B. fragilis strains tested. Cefuroxime in combination with penicillin or carbenicillin exerted excellent synergistic activity against two of the three isolates of B. fragilis tested and had an additive effect on the other. Although cefuroxime by itself is ineffective against B. fragilis, it acts synergistically with penicillin or carbenicillin.

Tissue levels of doxycycline in the human lung and pleura.

Antibiotic concentrations in the human lung and pleural tissue were measured following intravenous administration of doxycycline. The average serum levels were 9.3, 7.0 and 3.0 micrograms/ml at 1, 2, and 3 hours; correspondingly, the lung tissue levels were 6.8, 3.5 and 2.3 micrograms/g. The pleural tissue antibiotic concentrations at the same time intervals were 2.5, 1.5, and 1.5 micrograms/g. This suggests that doxycycline is retained well by the lung and the pleura, well above the minimal inhibitory concentration of most of the common bacterial pathogens of the lung, and may be used in the treatment of lung infections caused by susceptible organisms.

In vitro susceptibility of anaerobic flora of the female genital tract to ampicillin compared to spectinomycin.

Pelvic inflammatory disease (PID) may be caused by anaerobic bacteria and or gonococcus and therefore the efficacy of two antibiotics, ampicillin (AMP) and spectinomycin (SPM) commonly used in this setting was examined against 370 isolates of anaerobic bacteria. At the highest therapeutically achievable serum levels, AMP (16 micrograms/ml) inhibited 83% of all anaerobes and 70% of Bacterioides fragilis, and SPM (128 micrograms/ml) inhibited 98% of all anaerobes and 95% of B. fragilis strains tested. It therefore appears that both AMP and SPM may have a place in the treatment of PID. In our study, greater percentage of anaerobes were susceptible to SPM than previously reported. We used the 'glove box' technique used for testing the susceptibility of anaerobes.

Susceptibility of anaerobic bacteria in vitro to 23 antimicrobial agents.

589 isolates of anaerobic bacteria, including 127 isolates of Bacterioides fracilis, were tested against 23 antimicrobial agents. At their usually achievable serum levels, carbenicillin, ticarcillin, mezlocillin and piperacillin; cefoxitin, minocycline, doxycycline, clindamycin, chloramphenicol and rifampin were effective against more than 80% of B. fragilis isolates. At usual therapeutic doses, penicillin (PEN), ampicillin (AMP), cyclacillin (CYC), cephalothin (CT), cefazolin (CZ), cephradine (CPD), cefamandole (CMD), cefaclor (CCL), cefuroxime (CRX), spectinomycin (SPT), tetracycline (TET), ethambutol (EMB) and isoniazid (INH) were ineffective against B. Fragilis. However, high and clinically tolerated doses of PEN, CMD and SPT inhibited more than 80% of B. fragilis isolates, but AMP, CYC, CT, CZ, CPD, CCL and CRX, TET, EMB and INH remained totally ineffective against B. fragilis. The therapeutic merits of these antibiotics are discussed.

Clinical evaluation of piperacillin.

28 patients with aerobic (7), anaerobic (7), and mixed (14) infections were treated with intravenous piperacillin for an average duration of 14 days. All bacterial isolates tested in this study were susceptible to piperacillin less than or equal to 128 micrograms/ml. 27 patients were treated with 20 g of piperacillin/day in four divided doses. 1 patient with renal failure received only 8 g/day. 26 patients were treated with piperacillin and two with Staphylococcus aureus infection received gentamicin in addition. The peak serum levels were 266 micrograms/ml at 1 h and the trough 16 micrograms/ml at 5 h. In all, 26 patients (93%) were 'cured', one had 'recurrence', and one 'failed'. Adverse effects were pruitus in two patients, transient elevation of LDH in one, and transient eosinophilia in another. In our preliminary study, piperacillin was found to be safe and effective in the treatment of clinical infections.

Antimicrobial synergism of piperacillin and gentamicin against pseudomonas aeruginosa, Staphylococcus aureus and Streptococcus faecalis.

The potential synergistic effect of piperacillin with gentamicin was examined against five isolates each of Pseudomonas aeruginosa, Staphylococcus aureus and Streptococcus faecalis isolated from patients with endocarditis. Piperacillin and gentamicin synergistically inhibited all strains of P. aeruginosa and S. aureus tested. They exerted no synergistic effect against S. faecalis. Piperacillin and gentamicin synergism may be clinically tested in the treatment of serious pseudomonas and staphylococcal infections but not against infections caused by S. faecalis.

Microflora of the human small intestine.

The human small intestine is normally sterile in nearly one half of North American subjects. In this study the duodenum, jejunum, and ileum were sterile in 82, 69, and 55 per cent of the cases, respectively. Gram-positive cocci were the most frequent finding. E. coli, Enterobacter, and Klebsiella were present in the small bowel in nearly 7, 15, and 35 per cent of duodenal, jejunal, and ileal samples, respecatively. They were present in significant numbers (greater than 1 X 10(5)/ml) in the mid-jejunum in two patients and in the mid-ileum in seven patients (23 per cent). Even with modern anaerobic techniques, anaerobes are scarce in the small bowel; 4 to 6 per cent of persons may have aerotolerant anaerobes like clostridia, but strict anaerobes like bacteroides are rare. Our study provides baseline data for use in interpreting the intestinal bacterial overgrowth associated with certain postoperative disorders.

Antimicrobial activity of antituberculosis agents against anaerobic bacteria.

Anaerobic infections may coexist with tuberculosis, and can be mistaken for one another. The effect of therapy with antituberculosis chemotherapeutic agents against anaerobic bacteria (with the exception of rifampin) is unknown. We therefore examined the in vitro efficacy of certain commonly used antituberculosis agents (rifampin, isoniazid, and ethambutol) against 370 strains of anaerobic bacteria, including 86 isolates of Bacteroides fragilis. Rifampin at a concentration of 2 microgram/ml inhibited 91 percent of all anaerobic isolates. Both ethambutol and isoniazid were totally ineffective against any of the anaerobes tested, even at 64 microgram/ml. Therapy with rifampin in an unsuspected anaerobic infection can be misdiagnosed for tuberculosis. Therefore, when tuberculosis is suspected, isoniazid and ethambutol can be used and rifampin withheld until the acid-fast bacilli are demonstrated by additional diagnostic procedures, such as transtracheal aspiration.

In vitro activity of mezlocillin and its related compounds against aerobic and anaerobic bacteria.

A total of 900 clinical isolates of aerobic (462 isolates) and anaerobic (438 isolates) bacteria were tested against mezlocillin in comparison with other penicillins, and the aerobes were also tested against cephalothin, cefoxitin, and cefamandole. Among penicillins, mezlocillin was the most effective against Escherichia coli, Klebsiella pneumoniae, Salmonella species, Pseudomonas aeruginosa, and Bacteroides fragilis. Mezlocillin was more effective than cephalothin against Klebsiella pneumoniae.

Ticarcillin disodium in anaerobic infections.

Twenty-five patients were treated with ticarcillin disodium, 18 of whom had anaerobic infections that included pleuropulmonary infections (seven), mandibular osteomyelitis (four), perirectal abscess (two), sepsis, primary site unknown (one), liver abscess (one), pelvic abscess (one), decubitus ulcer (one), and synergistic gangrene (one). Seven had no anaerobic infections. Three had anaerobic septicemia. Culture results included anaerobes: peptococci (ten), peptostreptococci (ten), Bacteroides fragilis (six), Bacteroides not fragilis (ten), eubacteria (three), fusobacteria (two), Clostridium (one), Veillonella (one), and acidaminococcus (one); aerobes: Proteus (three), Klebsiella (two), Escherichia coli (two), and streptococci (two). Six patients with mixed aerobic infections initially received gentamicin sulfate in addition. The serum levels were 110 +/- 20 microgram/ml one hour after intravenous infusion of 5 g of ticarcillin disodium. All anaerobic isolates were susceptible at less than or equal to 100 microgram/ml and 85% by less than or equal to 25 microgram/ml of ticarcillin. Sixteen patients responded well to ticarcillin and two failed to respond. Our study suggests that ticarcillin is useful in the treatment of anaerobic infections.

Antimicrobial effect of amniotic fluid.

The antimicrobial effect of amniotic fluid (AF) obtained during the first (AF1) and second (AF2) trimesters was compared with the third (AF3) against anaerobic bacteria, such as Bacteroides fragilis ss. fragilis (6 strains), Eubacterium lentum (3 strains), and Peptostreptococcus anaerobius (4 strains). Escherichia coli (5 strains) served as a positive control. AF1 supported the growth of all 4 anaerobes (except B fragilis for 4 hours) for the entire 24-hour period tested. AF2 supported the growth of E coli and B fragilis for 24 hours but temporarily inhibited P anaerobius and E lentum. In contrast, AF3 inhibited all bacteria tested for 8 hours or more. It is concluded that AF1 is the least inhibitory, AF3 the most, and AF2 intermediate for the organisms tested. Lack of antimicrobial effect of AF on anaerobic bacteria may be one explanation for the higher incidence of anaerobic infections during absortion than during the prenatal period.

In vitro activity of cefaclor against aerobic and anaerobic bacteria.

Cefaclor (CCL), a new cephalosporin, was tested in vitro against 602 (271 anaerobic and 331 aerobic) clinical isolates in comparison with cephalothin, cefazolin, cephradine, and cefamandole. Sixteen micrograms of CCL per ml inhibited 68% of all aerobes tested and 80% of the 211 enteropathogenic organisms (Escherichia coli, Salmonella, and Shigella) isolated from cases of infantile diarrhoea. CCL inhibited 88% of gram-positive anaerobic cocci and 72% of Bacteroides other than B. fragilis at a concentration of 16 mug/ml. B. fragilis and Clostridia were resistant to CCL. Increased inoculum of E. coli from 10(5) to 10(9) increased the minimal inhibitory concentration of CCL and cefamandole by fourfold against 7 of the 64 strains tested. All seven were beta-lactamase negative. No antimicrobial synergism was noted between CCL and penicillin. The in vitro efficacy of CCL, an oral cephalosporin, against enteropathogenic E. coli, if proven safe, may be tested in vivo against such infections.

Susceptibility of anaerobic bacteria to cofoxitin and related compounds.

Cephalothin, cefazolin, cephradine, and cefoxitin inhibited more than 75% of 326 anaerobic bacteria in vitro, but cefoxitin was the most effective against anaerobes in general and Bacteroides fragilis in particular.