Epidemiology, antimicrobial susceptibility, pathogenicity, and significance of Bacteroides fragilis group organisms isolated at Los Angeles County-University of Southern California Medical Center

The epidemiology of species of the Bacteroides fragilis groups isolated at Los Angeles County-University of Southern California Medical Center was examined. In addition, frequency of resistance to six beta-lactam antibiotics (cefmetazole, cefotetan, ceftizoxime, imipenem, penicillin, and cefoxitin) and to clindamycin, chloramphenicol, and metronidazole was determined for each species. While B. fragilis was most commonly isolated, the other species of the B. fragilis group accounted for half of the isolates. Seven percent of 1,128 patients with infections due to species of the B. fragilis group were bacteremic. A review of bacteremic cases indicated that non-fragilis species were highly pathogenic. Resistance to clindamycin ranged from 8% to 22% among species and was most common among isolates of Bacteroides distasonis and Bacteroides thetaiotaomicron. Significant differences in antimicrobial activity were noted among the agents tested. Only imipenem, chloramphenicol, and metronidazole were predictably effective against non-fragilis species of the B. fragilis group. Prompt identification of species and susceptibility testing of clinical isolates of this group are needed if a newer beta-lactam agent or clindamycin is to be used for initial therapy.     

Activity of combinations of antimicrobial agents against Bacteroides fragilis.

Thirty-two clinical isolates of Bacteroides fragilis were tested against nine pairs of antimicrobial agents by means of an agar dilution technique. A synergistic effect was observed with the combination of clindamycin and metronidazole. End points were achieved with 17 strains. Of these strains, 13 (76%) were inhibited by the combination of clindamycin and metronidazole; each drug was present at a concentration of less than or equal to 25% of its minimal inhibitory concentration when tested alone. This combination also showed synergistic bactericidal activity against three of six strains examined by a tube dilution technique. No antagonism was noted with any strain. The other eight combinations tested failed to show a consistent synergistic effect, although no antagonism was observed. These in vitro data indicate that antagonism is not likely to be encountered when combination therapy is used for B. fragilis infections. For selected B. fragilis infections, the combination of clindamycin and metronidazole may be useful.     

Carbenicillin for treatment of Bacteroides fragilis infections: why not penicillin G?

Carbenicillin has been advocated for treatment of infections caused by Bacteroides fragilis and other anaerobic bacteria. Wide-scale use of the drug in this setting could result in a substantial increase in carbenicillin-resistant Pseudomonas aeruginosa, an effect that would have serious implications. Thirty-four strains of B. fragilis, one-half from bacteremic infections, were tested in vitro, and penicillin G was found to be twice as active as carbenicillin on an equal weight basis; 94% of the strains were inhibited by 32 microgram of penicillin/ml, a level easily achieved therapeutically. Penicillin killed B. fragilis organisms as rapidly as carbenicillin. In two subjects given equivalent doses (100 mg/kg intravenously) of carbenicillin and aqueous penicillin G, the bactericidal activity of serum against B. fragilis after administration of each drug was the same. Controlled clinical trials of treatment of anaerobic bacterial infections with penicillin G in high dosage, carbenicillin (or closely related ticarcillin), clindamycin, and chloramphenicol should be undertaken. Carbenicillin (and ticarcillin) for the present would seem better reserved for P. aeruginosa infections.     

Plasmid transfer from Escherichia coli to Bacteroides fragilis: differential expression of antibiotic resistance phenotypes.

A unique shuttle plasmid, pDP1, has been constructed to mediate gene transfer between Escherichia coli and the Gram-negative anaerobe Bacteroides fragilis. pDP1 contains the pBR322 replicon and the Bacteroides clindamycin resistance plasmid pCP1 linked to the transfer origin of the broad host range plasmid RK2. pDP1 can be transferred from E. coli to B. fragilis by the RK2 conjugation system even though RK2 itself is not maintained in the Bacteroides recipients. The antibiotic resistance and replication functions of pDP1 have been mapped by deletion analysis, and a 5-kilobase portion of the plasmid has been identified as the essential region for maintenance in Bacteroides. Comparison of the resistance conferred by pDP1 on E. coli and B. fragilis shows that antibiotic resistance genes are expressed differently in aerobic and anaerobic bacteria. These results document the feasibility of gene transfer from E. coli to B. fragilis and demonstrate the usefulness of this conjugation system to study genetic structure and expression in Bacteroides.     

Detection of specific IgG antibody in sera from patients infected with Bacteroides fragilis by enzyme-linked immunosorbent assay.

During a period of 13 months, 28 serious infections caused by Bacteroides were seen in 27 patients. Sixteen patients yielded Bacteroides fragilis; sera from 13 (81%) of these 16 had increased levels of IgG specific for B. fragilis lipopolysaccharide (LPS) antigens by enzyme-linked immunosorbent assay (ELISA). Sera from 20 normal controls did not have increased specific IgG. Sera from 22 of 23 patients with bacteremia caused by other gram-negative rods also failed to yield increased levels of specific antibody (P less than 0.0012). Analysis of sera from patients with B. fragilis infections disclosed a significant correlation between the levels of specific IgG to B. fragilis LPS measured by ELISA and the IgG antibody to the infecting B. fragilis by indirect immunofluorescence (r = 0.84, P less than 0.012). Two of the remaining 12 infections caused by Bacteroides not apparently due to B. fragilis organisms were also associated with increased levels of specific IgG to B. fragilis LPS antigens. Specific IgG antibody response may be an important adjunct in diagnosis of common B. fragilis infections and may allow better management of antimicrobial agents.     

Aerobic and anaerobic bacterial interactions in the development of anaerobic bacterial pneumonia.

The interactions between aerobic and anaerobic bacteria in the development of anaerobic bacterial pneumonia were studied by introducing Bacteroides fragilis and Escherichia coli alone or in combination into guinea pigs by tracheal infusion. The lung lesions induced by B. fragilis were mainly located near the pleura, unlike those induced by E. coli, and were accompanied by pneumonia, lung abscess, and pleuritis. The lung lesions produced by mixed infection with B. fragilis (10(9) cfu) and E. coli (10(7) cfu) were significantly more severe than those induced by either microbe alone, and the redox potentials at the foci of inflammation were markedly reduced (max: -330 mV). Analysis of lung lesions after treatment with aztreonam and clindamycin and neutrophil phagocytic activity suggested that E. coli was primarily responsible for the lung lesions and that B. fragilis promoted the accompanying inflammation, resulting in increased pathogenicity of the mixed infections.     

Treatment of anaerobic infections with metronidazole.

Metronidazole (Flagl), a compound widely used in man with minimal side-effects, has been used in the treatment of anaerobic infections caused by Bacteroides fragilis and other Bacteroides species. Seven patients were treated and all were restored to full health. Four of them did not respond to lincomycin or clindamycin which so far have been the drugs of choice against anaerobic infections. The reasons for the good results during metronidazole therapy such as the good penetration through the blood/brain barrier and into abscess cavities are discussed.     

beta-lactamases in bacteroides.

Bacteroides fragilis is responsible for most anaerobic infections in man. Most isolates of B. fragilis show resistance to beta-lactam antibiotics. This resistance might be due to beta-lactamase production or permeability barrier in the cell wall. B. fragilis produce beta-lactamase with mainly cephalosporinase activity. Other Bacteroides species such as B. clostridiformis, B. melaninogenicus and B. oralis also produce beta-lactamase but with different biochemical characteristics.     

In vitro susceptibility vs. in vivo efficacy of various antimicrobial agents against the Bacteroides fragilis group.

In vitro susceptibility testing is only one step in the evaluation of the potential efficacy of antimicrobial agents against the Bacteroides fragilis group. An assessment of in vivo efficacy, with a consideration of the factors that can best be studied in an infected host, is also an integral part of this process. Abscess models in rodents have been used to correlate in vitro activity with in vivo efficacy against this group of microorganisms. For metronidazole, clindamycin, moxalactam, and cefoxitin, the correlation was strong; for chloramphenicol and carbenicillin, it was not. In vivo studies of mixed infection with the B. fragilis group and Escherichia coli showed that cefoxitin and imipenem were effective; in contrast, cefotetan was not effective against resistant strains. Only strains susceptible to ceftizoxime in the agar dilution test were also affected by this drug in vivo. The so-called inoculum effect noted with ceftizoxime may explain this finding. In vivo elimination of encapsulated organisms of the B. fragilis group was found to be more difficult than elimination of unencapsulated isolates. The beta-lactamase produced by Bacteroides species can protect the enzyme-producing organism as well as its partners in mixed infections from the effects of beta-lactam antibiotics. These data illustrate the complexity and difficulties encountered when in vitro activity is correlated with in vivo efficacy.     

Surface antigens as virulence factors in infection with Bacteroides fragilis

Organisms of the genus Bacteroides represent the major group of obligate anaerobes involved in human infections. Bacteroides usually cause either bacteremia or localized abscesses. Of the numerous species of Bacteroides, Bacteroides fragilis is the single most frequent clinical isolate. B. fragilis and Bacteroides melaninogenicus have chemically incomplete lipopolysaccharides as compared with the lipopolysaccharides (endotoxins) of aerobic bacteria, and the lipopolysaccharides of Bacteroides lack the biologic potency characteristic of endotoxin. This inactivity may account for the very infrequent occurrence of disseminated intravascular coagulation or purpura that can accompany sepsis due to these organisms. Furthermore, strains of B. fragilis have an immunologically common capsular polysaccharide. In an animal model of intraabdominal sepsis, the encapsulated strains caused abscesses when given without other organisms, but abscess formation from unencapsulated strains of Bacteroides generally required the administration of a synergistic aerobe. The abscesses caused by encapsulated strains were shown to be directly attributable to the capsular polysaccharide, which is an important virulence factor of this organism. Patients or experimental animals infected with B. fragilis develop antibodies to the capsular polysaccharide, and these antibodies can be detected in a radioactive antigen-binding assay.     

Mezlocillin in the therapy of serious infections

Mezlocillin, a new semisynthetic penicillin chemically related to ampicillin which is more active than carbenicillin against Ps. aeruginosa, B. fragilis and Strep. faecalis and which inhibits many Klebsiella, was evaluated in the therapy of 34 episodes of infection in 26 patients. Infection sites included pulmonary, urinary tract and tissue infections, including peritonitis. Seven patients had bacteremia. Clinical cures were achieved in 83 per cent and bacteria cures in 76 per cent of infections. Cure was achieved with mezlocillin in patients with infections caused by carbenicillin-resistant species. Adverse effects of therapy were minimal, one rash and one episode of reversible neutropenia. Serum and body fluid levels were easily maintained above the inhibitory levels of susceptible organisms. Mezlocillin was a safe, well tolerated and effective antibiotic in the treatment of infections due to susceptible organisms.     

Clindamycin and gentamicin for aerobic and anaerobic sepsis.

Thirty-eight adult patients with serious pleuropulmonary, soft-tissue, bone, and intra-abdominal infections caused by combinations of aerobic, facultative, and anaerobic bacteria were treated with parenterally given clindamycin phosphate and gentamicin sulfate and surgery when appropriate. Nine had associated bacteremia. In 29, infections failed to respond to other therapeutic regimens, which included penicillins, cephalosporins, aminoglycosides, and chloramphenicol. Results with clindamycin and gentamicin were excellent and were attributed primarily to the activity of clindamycin against anaerobes, particularly Bacteroides fragilis. Serum concentrations of clindamycin surpassed by manyfold the minimal inhibitory concentrations (MICs) for anaerobes. Serum concentrations of gentamicin did not consistently surpass the MICs for Enterobacteriaceae and Pseudomonas aeruginosa, although those organisms were consistently gentamicinsusceptible by disk diffusion susceptibility tests. Persistent colonization with Enterobacteriaceae, P aeruginosa, enterococci, or Candida were common, and occasionally they were significant in prolonging the clinical courses of patients with extensive infections.     

The polysaccharide capsule of Bacteroides fragilis subspecies fragilis: immunochemical and morphologic definition.

A large-molecular-weight capsular polysaccharide was isolated from strains of Bacteroides fragilis subspecies fragilis. By means of electron microscopy and staining with ruthenium red, the thick polysaccharide capsule was also visualized. With use of a radioactive antigen-binding assay, antibody to this capsular polysaccharide was demonstrated in antisera prepared in rabbits to each of eight strains of B. fragilis fragilis. Antibody of similar specificity was not found in antisera prepared to Bacteroides melaninogenicus or to strains of Bacteroides fragilis subspecies vulgatus and Bacteroides fragilis subspecies distasonis; such antibody was found in antisera to only one of two strains of Bacteroides fragilis subspecies thetaiotaomicron. The radioactive antigen-binding assay is a sensitive test for the detection of antibody to capsular polysaccharide. This polysaccharide antigen may form the basis of a serogrouping system for B. fragilis.     

Effect of subinhibitory doses of clindamycin on the virulence of Bacteroides fragilis: role of lipopolysaccharide

The capsular polysaccharide (CP) of Bacteroides fragilis is an important virulence factor in the formation of experimental intraabdominal abscesses. Incubation of this organism with subinhibitory doses of clindamycin induced morphological changes in the bacteria, including elongation and loss of CP, detected by ferritin-labeled antibody to capsule. Pretreatment of bacteria with subinhibitory doses of clindamycin, however, did not affect the ability of live or heat-killed organisms to produce intraabdominal abscesses in a mouse model of intraabdominal sepsis. Dose-response experiments with purified CP as well as lipopolysaccharide (LPS) from B. fragilis ATCC strain 23745 mixed with sterile cecal contents as adjuvant revealed that both surface components of the organism were capable of causing abscesses in the mouse model. The dose of LPS required to induce abscesses was five times higher than the required dose of CP. Nevertheless, these studies suggested that B. fragilis LPS is another virulence factor in the formation of intraabdominal abscesses.     

Perforated and gangrenous appendicitis: an analysis of antibiotic failures

The relationships between resistant pathogens, serum levels of gentamicin, and the outcomes of gangrenous or perforated appendicitis were analyzed in 147 patients. Failure to cure the infection occurred significantly more frequently among patients treated with cefoperazone or cefamandole than among those treated with clindamycin and gentamicin in combination. The failures were associated with recovery of resistant Bacteroides fragilis from intraoperative cultures. Pseudomonas species were also associated with failures, their in vitro susceptibility not correlating with clinical cure. Patients with gentamicin peak serum levels of less than 6 micrograms/ml in the first three days were not more likely to be associated with failure than were patients with higher levels. These clinical observations indicate that antibiotic therapy of intra-abdominal sepsis should include antibiotics with in vitro activity against B fragilis and that precise adjustments of gentamicin levels may not improve outcome. In addition, Pseudomonas species may play a significant role in some of these infections.     

Susceptibility of isolates of Bacteroides to the bactericidal activity of normal human serum.

Seventy-one strains of species from the Bacteroides fragilis group, including 46 isolates of B. fragilis, were tested for susceptibility to the bactericidal effect of serum from healthy subjects. Twenty-seven (38%) of the isolates were killed by serum. Isolates from feces were significantly more sensitive to serum than were isolates from patients with clinical infections. Killing of bacteria required heat-labile serum components and was an exponential function of serum concentration. Among the various species tested, B. fragilis was clearly the most resistant to bactericidal activity of serum. These observations may be important to the understanding of infections caused by the B. fragilis group, which contains the anaerobes of greatest clinical importance.     

Isolation and identification of encapsulated strains of Bacteroides fragilis.

One hundred three clinical isolates of Bacteroides fragilis were identified during a two-year period. Most of these isolates were strains of B. fragilis subspecies fragilis, which constitutes a minor component of the fecal flora in comparison with the other subspecies of B. fragilis. By use of several techniques for demonstration of capsules, it was found that only B. fragilis strains classified as subspecies fragilis were encapsulated. An indirect immunofluorescence assay was developed for identification of clinical isolates possessing capsular material that was immunologically similar to that found in the reference strain of B. fragilis subspecies fragilis. All strains examined that were classified as subspecies fragilis were positive in this assay for the capsular material, whereas strains of the other subspecies were negative. This tests represents a rapid and sensitive means of identifying the most prevalent anaerobic gram-negative bacillus involved in human infections. The capsular polysaccharide of B. fragilis subspecies fragilis is a unique factor associated with the predominant subspecies of B. fragilis isolated from clinical material.     

Susceptibility of anaerobic bacteria to metronidazole: relative resistance of non-spore-forming gram-positive baccilli.

Susceptibility of 358 clinical isolates of obligate anaerobes to metronidazole was determined by an agar-dilution technique. Only 66% of all isolates were inhibited by 6.25 mug/ml, whereas 30% required larger than or equal to 50 mug/ml. Considerable variation in susceptibility was observed among different genera and species of bacteria. Fusobacterium was most senstitive, followed by Clostridium, Bacteroides and Peptococcus, Peptostreptococcus, Veillonella and Acidaminococcus, and non-spore-forming gram-positive bacilli. Bacteroides fragilis was more sensitive than other species of Bacteroides. Similarly, Clostridium perfringens was more susceptible than other species of Clostriduim. While metronidazole appears to be a promising antimicrobial agent for infections caused by Fusobacterium, Clostrididium, and B.fragilis, therapy for infections with other anaerobic bacteria should be guided by in vitro tests of sensitiivity.     

Bacteroides fragilis septicemia originating in the middle ear and the lungs

Summary Two unusual sources of Bacteroides fragilis septicemia are described. In the first case, a young man, the septicemia orginated in an infection of the middle ear. In the second case, an older man with chronic pulmonary disease, lung abcsesses were probably the cause of bacteremia. The older man died before appropriate antibiotics could be administrated. The younger patient recovered remarkably well under therapy with doxycyclin and clindamycin. In both cases presented Bacteroides fragilis, which is frequently resistant to tetracycline, was susceptible to doxycyclin. Chloramphenicol, lincomycin and clindamycin exhibited the greatest in vitro antibacterial activity.

---

Das Mittelohr und die Lunge als Herd einer Bacteroides fragilis-Sepsis

Zusammenfassung Es werden zwei ungewöhnliche Ausgangsherde einer Bacteroides fragilis-Sepsis beschrieben. Im ersten Fall ging bei einem jungen Mann eine Sepsis vom Mittelohr aus. Beim zweiten handelt es sich um einen älteren männlichen Patienten mit einer chronischen Lungenerkrankung. Wahrscheinlich war hierbei ein Lungenabszeß die Ursache für eine Bakteriämie. Der letztere Patient starb, bevor wirksame Antibiotika gegeben werden konnten, der jüngere Patient erholte sich bemerkenswert gut unter der Therapie mit Doxycyclin und Clindamycin. In den vorliegenden Fällen erwies sich Bacteroides fragilis, das sonst häufig tetracyclin-resistent ist, beidemal doxycyclin-empfindlich. Die Stämme wiesen gegenüber Chloramphenicol, Lincomycin und Clindamycin die höchste in-vitro-Sensibilität auf.

     

Cefoxitin: its role in treatment and prophylaxis of obstetric and gynecologic infections

Cefoxitin has become one of the most used parenteral antibiotics in the United States, perhaps because of a broad spectrum of activity, including activity against Bacteroides fragilis, which makes the drug suitable for prevention and treatment of intraabdominal and pelvic infections. This review focuses on the use of cefoxitin in obstetric and gynecologic infections, with comparisons to older and newer antibiotics. Numerous studies have shown that cefoxitin is clearly effective; in most of these studies, however, either the initial infection rates were low or the sample sizes were small--circumstances making it difficult to establish the superiority of any one agent. Thus, the necessity of using a drug with activity against B. fragilis for prevention and treatment of pelvic infections has not been proven. Several antibiotics without such activity have been equally effective. Cefoxitin may be of particular value when combined with surgical drainage of pelvic abscesses, infections in which control of B. fragilis may be especially important to outcome.