Bacteria requiring an environment without oxygen for growth and replication have been known since the 17th century observations of Leeuwenhoek, but their significance in clinical infection has become widely appreciated only since relatively simple anaerobic culture techniques have become available. ' Most investigators agree that anaerobic infections are less frequent in children than in adults, but several studies have documented that anaerobic organisms, when sought, are responsible for 5% to 10% of all clinically significant bacteremias and for varying proportions of other deep organ and soft tissue infections. 2 Thus, these infections should be considered in high-risk situations or in cases of unexplained clinical sepsis.
My view of anaerobic infections is that they are largely opportunistic, with the special caveat that in this context the susceptible patient is one whose normal barriers to local tissue invasion by bacteria has been breached by trauma, surgery, vascular embarrassment or primary inflammatory process. This situation may and often does occur in an otherwise healthy child, although children immunocompromised by drugs or malignancy, etc., may also succumb to anaerobic infections. The observation that most anaerobic infections are caused by endogenous flora from skin or mucus membranes which have invaded adjacent tissues and possibly produced dissemination, is consistent with the thesis that, by and large, protective "immunity" to these infections is offered by intact mucocutaneous barriers.3
The primary characteristic which differentiates anaerobic organisms from aerobic bacteria is that oxygen is actively toxic to them. This appears to be due to the organism's need for a low environmental oxidationreduction (redox) potential, a requirement which does not hamper growth in living tissues where reduced substances are the natural products of metabolism.4 In vitro cultivation however, has required the developmentof systems which can create a sufficiently reduced and/or O2-free atmosphere, a variety of which are currently available to clinical laboratories.5 Whatever system is used for cultivation, it is critical that culture material obtained from patients be handled with care and dispatch so that organisms are not exposed to air prior to inoculation in the laboratory.
The clinically important anaerobic bacteria that are likely to be isolated in culture are listed in Table 1. It is important to note, also, that anaerobic infections are commonly polymicrobial, involving both grampositive and gram-negative strains.6
The host factors which predispose to anaerobic infection have been outlined previously. The particular infectious diseases commonly caused by anaerobes, shown in Table 2, are largely predictable due to their occurrence adjacent to skin or mucus membranes, or because they result from contamination from those surfaces. Some of these infections are common in adults and have received a great deal of "press" as well as attention from pharmaceutical manufacturers. By contrast, some such as cervical adenitis, peritonsillar abscess, and human and animal bite infections, are much more characteristically pediatrie diseases which may involve anaerobes.7'10
ANAEROBIC BACTERIA RESPONSIBLE FOR CLINICAL INFECTION
The particular characteristics which should alert the clinician to the possible presence of anaerobes in an infection include a foul odor (from the pus, sputum, wound drainage, etc. ), a particularly necrotizing infectious process with extensive tissue destruction, gas production and bacteria which are visible on gramstained smears but not isolated from routine cultures. While none of these characteristics are pathognomontc or unique to anaerobic infections, in the aggregate they strongly suggest disease due to this class of organism.
Making the diagnosis of an anaerobic infection requires a high degree of suspicion together with the acquisition of appropriate laboratory tests to support and confirm the diagnosis. The type of patient likely to develop such an infection and the type of infection likely to involve anaerobes have been described. Laboratory tests which may support the diagnosis of anaerobic infections include gram stains of purulent material revealing mixed bacterial flora and radiographic studies showing air in the tissues.
Confirmation of the diagnosis as in most infectious diseases, is achieved by culture and identification of anaerobic bacteria from infected material. Reliable cultures are usually obtained from aspirated material rather than from secretions or spontaneous wound drainage since the latter may be contaminated by endogenous flora or air. In addition to culture, several new methods of "immunodetection" have been described recently. While not yet generally available or uniformly sensitive and specific, these procedures for the detection of bacterial antigens in body fluids using fluorescent antibodies hold promise for rapid diagnosis in the near future.11*12
CLINICAL INFECTIONS CAUSED BYANAEROBES IN CHILDREN
The management of anaerobic infections usually involves appropriate surgical debridement and antibiotic therapy. Experience with other modalities, such as hyperbaric oxygen, is limited in children and recommendations cannot be made at present.
Incision and drainage is required of most anaerobic abscesses. Irrigation with physiologic saline may be coupled with drainage when feasible, particularly in the abdominal cavity. Small, multiple or inaccessible abscesses, occasionally seen in the brain or liver, may respond to medical therapy alone, although firm data confirming the superiority of this approach are not available. Even in the absence of abscess formation, necrotizing anaerobic cellulitis may require extensive surgical debridement in order to control the infection.13
The choice of antibiotics for the treatment of anaerobic infections usually must be based on the recognition of which organisms are likely to be involved in various anatomic sites and on published reports of susceptibilities. While susceptibility testing should be performed on any pathogens isolated, these tests generally require several days and immediate choices of therapeutic agents must be made empirically. Most gram-positive and non-Bacteroides gram-negative anaerobes are, to date, sensitive to clinically achievable levels of penicillin. Sem i -synthetic derivatives, such as ampicillin, methicillin, oxacillin and nafcillin are less active than the parent compound, but the very high levels of these drugs achieved with the parenteral administration of large doses, appear to be adequate in some cases. Penicillin G, the preferred choice, should be administered in doses of 100,000 to 200,000 units/ kg/24 H.
The organism which most commonly presents the problem of penicillin resistance is Bacteroides /ragilis, although a growing proportion of B. melarunogenicus, other Bacteroides species and an occasional Clostridia species now demonstrate resistance.14 The mechanism of penicillin resistance is usually beta- lac tamase production, which may render the beta-lactamase inhibitors useful for treatment.
In general, clindamycin has become the drug of choice for suspected penicillin-resistant, non-CNS anaerobic infections in children because of its overall efficacy and low potential for toxicity. Pseudomembranous enterocolitis occurs rarely in children, being more frequently associated with other antimicrobial agents. Cefoxitin is at present probably the second choice agent for non-CNS infections with a level of efficacy approaching that of clindamycin and a low potential for toxicity.15
Chloramphenicol remains the drug of choice for penicillin-resistant CNS infections because of its excellent penetration, whereas, neither clindamycin nor cefoxitin cross the blood-brain barrier. Recently, metronidazole has been used increasingly for anaerobic infections, including brain abscess, with good results. Comparative clinical trials in children are not available, but its bactericidal activity against the penicillin-resistant anaerobes makes metronidazole an attractive drug for these serious infections.16
The third-generation cephalosporins on the market and "in the wings" exhibit variable efficacy against all groups of anaerobes and should not be relied upon without culture results demonstrating susceptibility to these agents. The newer ureidopenicillins demonstrate broader activity than the cephalosporins but their inhibitory concentrations are significantly higher than those of any of the "first-line" drugs. Older, first-generation cephalosporins and aminoglycosides, frequently used for aerobic gram-negative infections, have no activity against B. fraglis.
The dose of clindamycin is 20 to 40 mg/kg/24 H in three or four divided intravenous doses, with the higher dose being administered in severe illnesses. Similar doses of clindamycin can be administered orally when the clinical response and condition of the patient warrant. Cefoxitin is administered in doses of 100 to 150 mg/kg/24 H in four or six divided intravenous doses. Doses should be modified slightly if significant renal impairment is present. Metronidazole has been used in the same doses administered to adults although safety and efficacy have not been determined. The dose is 15 mg/kg as an intravenous loading dose and a maintenance dose of 30 mg/kg/24 H in four divided doses. Accumulation of drug has been demonstrated in patients with significant hepatic dysfunction.
Chloramphenicol succirvate is administered to normal children and infants over 3 months of age in adose of 100 mg/kg/24 H in four divided intravenous doses. Younger infants, and children with hepatic dysfunction, should receive 25 to 50 mg/kg/24 H in two or four divided intravenous infusions, the lower dose being administered to premature infants and those less than 2 weeks of age. It is imperative that serum levels of chloramphenicol be monitored in young infants receiving the drug, in order to maintain therapeutic and nontoxic levels. Serum levels between L5 and 30 µg/ml meet this requirement. Concurrent concentrations of chloramphenicol in the CSF are 40% to 60% of blood levels.17 Assessment of bone marrow function by complete blood count, reticulocyte count, and platelet count should be performed frequently during chloramphenicol therapy. Neutropenia generally is the first sign of bone marrow suppression; we consider an absolute neutrophil count of less than 1000/mm3 an indication to discontinue the drug.
The duration of antimicrobial therapy for anaerobic infections does not differ from that recommended for the same infections of aerobic etiology. Minor soft tissue infections generally are treated for 10 to 14 days, septicemia for 10 days to 3 weeks, bone and joint infections for 6 or more weeks and brain abscesses for 6 to 8 weeks.
The outcome of patients with severe anaerobic infections generally is related to the rapidity with which the infection is treated effectively with surgery and/or appropriate antibiotics and to the severity of underlying disease. When endotoxic shock, disseminated intravascular coagulopathy, or metastatic foci of infection supervene, the prognosis for favorable outcome diminishes and reflects the effectiveness of treatment for these complications. Thus, as is so commonly the case, recognition of the high-risk clinical situation and institution of early and appropriate therapy are likely to result in a good outcome.
1. Tally TB, Stewatt PR. Suttei VL. et al. Oxygen tolerance of fresh clinical anaerobic bacteria. J Clin Microbiol [975; 1:161.
2. Brook I. The role of anaerobic bacteria in pediatrie infections. Adv Pediaterr 1980: 27:163-197.
3. Dunkle LM, Brotherton TJ. Feicin RD Anaerobic infections in children; A prospective survey. Jfefaana 1976: 57i31l-J20.
4. Moms JG: The biochemical basis of oxygen sensitivity. J Gm MicRihtif 1970; 60;3.
5. Hansen SL. Steward BJ: Comparison of API and Minitek to CDC methods tui the biochemical chracterization of anaerobes. j Clin Microbiol 1976; 4:227-211.
6. Dunkle LM: Anaerobic infections in infants anj children, in Feigin RD. Cherry JD (ed): Textbook of Pediatr Infections Disease. Philadelphia. WB Saunders, 1971. pp 808-819.
7. Brook 1: Aerobic and anaerobic bacteriology of ci'rvical .idenitis in children. Cim ftdidir I960; 19:69)-696.
8. Brook I: AeniNc and anaerobic bacteriology of pentnnsillar arracess in children. ACIB Radiati Scand 1981; 70;831-835.
9. Brook I. Chronic ixiris media in children: Microbiological siudies. Am J Dis Child 1980; 134:564-566.
10. Golilstein E)C, Citron DM. Finegold SM; Rule of anaerobic hacreri.i in hite-WHind infections. Rev Inject Da 1984; 6:SI77-S813.
11. Citron DM: Specimen collectiun and transport, anaerobic culture techniques and idemilication of anaerobes. Rev Infect Da 1984; 6:S51-S58.
12. Wessfelf AS. Sonnenwirtth AC: Rapid derection and identification of B. fraglis and S. melanininogenicus by immunoflouresnence. J Clin Microbiol 1981; 1;798-800.
13. Stamenkovk I, Lew PD: Early recognition of potentially ratal necnrtizing f.istiitis. N Engl Mid 1984: 310:1689-1693.
14. Edson RS. Rosenblatt JE, Lee DT. et al· Recent experience with untinticrobial susceptibility of anaerobic bacteria. Mir«! CJm Pwe 1982; 57:7734-741.
15. Onderdonlt AB, flaniert JG. Louie T, et al: Mictobial synergy m experimental inrraabdominal abscess. Infect Immun 1976; ?:22-26.
16. Ingharn HR, Se Ikon JB, Roxby CM: Bacteriological study of otogenic cerebral abscesses-. Cnerftulherapeutic role of metronidazole. BT Med J 1977; 2:991-991.
17. Dunkle LM: Central nerwus system chknamphenicol concentration in premature infants. Antimicrob Agerus Chemother 1978;427-429.
ANAEROBIC BACTERIA RESPONSIBLE FOR CLINICAL INFECTION
CLINICAL INFECTIONS CAUSED BYANAEROBES IN CHILDREN