INTRODUCTION — Anaerobic bacteria are the major constituents of normal human flora at virtually all anatomic sites and have been recovered from a wide array of different infections. Isolating the organisms from specimens requires the use of specialized methods in the microbiology laboratory, but it has often been problematic to determine when these bacteria represent true pathogens or merely commensals. This decision is facilitated by the anaerobe recovered and its established pathogenic potential and concentration as well as the appropriateness of the specimen source.
The pathophysiology of anaerobic infections, clinical clues to suspect anaerobes, and methods for recovering the organisms will be reviewed here. Clinical syndromes involving anaerobes (including treatment), the history of anaerobes, and their role in the normal flora are discussed separately. (See "Anaerobic bacterial infections" and "Anaerobic bacteria: History and role in normal human flora".)
PATHOPHYSIOLOGY OF ANAEROBIC INFECTIONS — Anaerobic infections nearly always arise from leakage of endogenous bacteria into contiguous or other sites. Important exceptions are some of the clostridial syndromes, including botulism, Clostridium perfringens food poisoning, enteritis necroticans, tetanus, some cases of gas gangrene, and Clostridioides difficile-associated diarrhea. (See related topics.)
The usual pathophysiologic mechanism for anaerobic infection is a breach in the mucocutaneous barrier resulting in displacement of the normal flora. Host defense mechanisms are presumably important, but the compromised host is not unusually susceptible. Exceptions are infections associated with defects of mucocutaneous barriers, such as carcinoma with obstruction, mucositis, perirectal lesions, or compromised consciousness with aspiration.
Three major issues arise in the pathogenesis of anaerobic infections:
●Virulence factors of the organisms
●Mechanisms of abscess formation
●Bacterial synergy
Some of the pathogenic mechanisms of anaerobes also apply to aerobic infections, especially virulence factors. However, others figure far more prominently in infections involving anaerobes. As an example, abscesses represent an intriguing combination of pathology and host defense. Animal models of intra-abdominal sepsis have contributed substantial insights into the pathogenesis of abscesses. Synergy among bacterial species is also a common feature of anaerobic infections. These three important components of the pathogenesis of anaerobic infections are discussed below.
Virulence factors — Anaerobic bacteria contain a number of components that have been defined as virulence factors in facultative bacteria, including toxins, polysaccharide capsules, and lipopolysaccharides. Most of these also have been shown to be virulence factors in anaerobic species. The capacity to survive and grow under anaerobic conditions has also led to the identification of other properties of anaerobes that may play a role in virulence.
Toxins — The most clearly identified virulence factors for anaerobic bacteria are the exotoxins produced by clostridial species, including botulinum toxins, tetanus toxin, C. difficile toxin B [1,2], and five toxins produced by C. perfringens (as well as many other clostridial species); these are among the most virulent bacterial toxins in mouse lethality assays. Alpha toxin is the major toxin produced by C. perfringens, but studies have shown a complex interaction between alpha and theta toxin in the production of experimental gas gangrene [3,4]. Both toxins appear to be involved in upregulation of intercellular adhesion molecule (ICAM)-1 and platelet aggregating factor (PAF), which contributes to vascular leukostasis and the absence of a polymorphonuclear leukocyte (PMN) response to the infection [5]. Alternatively, clostridial species have been shown to produce leukocytosis-inducing factors that result in marked leukocytosis in some infected patients ("leukemoid reaction") [6,7].
Clinical expression of histotoxic clostridial syndromes depends upon the site of toxin production and the physiologic effects of the toxin. (See "Clostridioides difficile infection in adults: Epidemiology, microbiology, and pathophysiology" and "Clostridial myonecrosis" and "Botulism".)
A molecular subgroup of Bacteroides fragilis secrete a metalloprotease enterotoxin [8]. Strains producing the toxin are associated with diarrheal illnesses, particularly in children <5 years old, but not uncommonly also asymptomatically colonize healthy individuals [9-11]. There is evidence from mouse models that enterotoxin-producing strains of B. fragilis may play a role in colon carcinoma [12,13].
Capsular polysaccharides — In the same way that Clostridium spp serve as the prototypes for toxins produced by anaerobes, B. fragilis has been studied as a uniquely virulent anaerobe. The organism constitutes only 0.5 percent of cultivable normal colonic flora, yet it is the most common anaerobe isolated from intra-abdominal infections; animal models show it to be important in intra-abdominal abscess formation, and it is the most frequent anaerobic isolate in cases of bacteremia [14,15].
Based upon a rat model of intra-abdominal sepsis, it became apparent that B. fragilis alone among the anaerobes studied was capable of provoking abscesses without a synergistic facultative organism and that heat-killed B. fragilis retained this capacity [16]. Following extraction and purification of the polysaccharide capsule of a prototype strain, this sugar was also able to provoke abscesses in the model system. Further study of the capsular polysaccharide has led to an appreciation that it is actually a complex of polysaccharides with zwitterionic properties (containing alternating oppositely charged sugars) [17,18]. Adherence of the capsular polysaccharide complex to mesothelial cells in vitro stimulates ICAM-1 and tumor necrosis factor (TNF)-alpha; pretreatment of mice with antibodies to ICAM-1 or TNF-alpha resulted in failure of the animals to develop intraperitoneal abscesses in a mouse model of intra-abdominal sepsis [19]. Polysaccharide A of Bacteroides spp has also been shown to be important for abscess formation via activation of toll-like receptor (TLR)2, since abscess formation is impaired in TLR2-deficient mice [20].
Lipopolysaccharides — Anaerobic gram-negative bacteria, like all gram-negative bacteria, contain lipopolysaccharide (LPS) that can be extracted from the envelope, but the biologic activity of this endotoxin (mouse lethality assays, the chick embryo death test, and the Shwartzman reaction) is 100 to 1000 times less than that of LPS from Enterobacteriaceae [21]. The LPS of B. fragilis contains a lipid A moiety (the endotoxin portion of LPS), but there are structural and chemical composition differences that render this LPS less potent than the LPS of Escherichia coli [21]. The inability of B. fragilis LPS to activate TLR2 may be responsible for this difference [22].
Other anaerobic gram-negative bacteria, such as Fusobacteria, are thought to contain endotoxin with substantial biologic activity, including activation of TLR4 [23]. One study of extracted LPS from a strain of Fusobacterium nucleatum found biologic activity equal to the LPS of Enterobacteriaceae and correlated this with inhibition of gingival fibroblasts [24]. Thus, LPS of Fusobacteria may play a pathogenic role in periodontal disease and presumably accounts for the severity of illness associated with Lemierre disease [25,26]. (See "Anaerobic bacterial infections".)
Volatile fatty acids — Another virulence factor of anaerobes is the production of short-chain volatile fatty acids. Production of these fatty acids is a characteristic feature of the metabolic system of anaerobes that is used to identify these microbes in the microbiology laboratory and may be responsible for the characteristic putrid drainage. These volatile fatty acids have been shown to inhibit phagocytic killing of bacteria [27,28].
Ability to tolerate oxygen — A number of anaerobic bacteria, including B. fragilis and some Clostridia, can tolerate exposure to oxygen but do not replicate in atmospheric oxygen. These organisms contain varying concentrations of superoxide dismutase, an enzyme present in aerobic bacteria, which protects against the toxic effects of oxygen [29,30]. The ability to survive exposure to oxygen is a different type of factor that facilitates the survival and thus pathogenicity of the organism.
Formation of and host defense to abscesses — As noted above, an abscess is a unique biologic phenomenon, representing the combination of a pathologic process and a host defense to contain infection. Within the fibrous capsule of a well-formed abscess, bacteria and neutrophils coexist; these same neutrophils in vitro will kill the bacteria.
Intra-abdominal sepsis model — Rat and mice models of intra-abdominal sepsis have been used extensively to study the role of anaerobes and mechanisms of formation and host defense to abscesses. These models closely simulate human infection. As an example, when rats are challenged with an intraperitoneal implant of fecal contents with an adjuvant, a two-stage infection results [31,32]. The initial phase of the infection is characterized by generalized peritonitis, followed by a second phase with abscess formation. Initial work showed a 43 percent mortality rate; all deaths occurred during the first four days after challenge and were associated with E. coli bacteremia. Abscesses were initially detected five days after challenge, and all animals sacrificed seven days or more after challenge had abscesses [32]. This model represents a classic example of microbial synergy.
Role of anaerobes — Studies of the rat model were initially designed to distinguish the role of various bacteria in abscess formation through quantitative cultures, antimicrobial probes (such as gentamicin for its selective activity against coliforms and clindamycin for its selective activity against anaerobes), and monomicrobial challenge with the organisms recovered from infected sites. This work supported the role of E. coli as the major pathogen in the initial phase of infection characterized by generalized peritonitis, bacteremia, and death. B. fragilis appeared critical for the second phase of the infection characterized by abscess formation. Further testimony to the "abscessogenic potential" of B. fragilis was the demonstration of typical abscesses after challenge with the capsular polysaccharide of B. fragilis alone without facultative bacteria [31,32].
The conclusion is that both coliforms and anaerobic bacteria represent pathogens in this model of intra-abdominal sepsis, although they appear to be responsible for different biologic events as the infection evolves through its two stages. The practical application of these data is that antimicrobial therapy should be directed against both coliforms and anaerobes, a thesis that is well supported by the two published clinical trials that test outcome when anaerobes are treated or not treated [33,34].
Role of T cells in abscesses — The animal models of intra-abdominal sepsis were also used to identify cells responsible for abscess formation. Early work with nude or cyclophosphamide-depleted mice suggested that T cells played a critical role in abscess formation due to B. fragilis [35]. Further studies demonstrated that the zwitterionic capsular polysaccharide of the organism described above activated CD4 T cells in vitro, and abscess formation in vivo was inhibited by CTLA4Ig, which blocks the CD28-B7 pathway activation of T cells [36-38]. Another study has shown that interleukin-17, a proinflammatory cytokine produced by CD4 T cells, is necessary for abscess formation following challenge with B. fragilis or abscess-inducing zwitterionic polysaccharides [39].
Activation of T cells, in turn, appears to play a role in the prevention of abscess formation. Adoptive transfer of splenic T cells from mice immunized with B. fragilis capsular polysaccharide protects the recipient mice from abscesses caused by B. fragilis [40]. The unusual charge of the B. fragilis capsular polysaccharide directly affects T cell activation, which also needs major histocompatibility complex II on antigen-presenting cells [41]. Thus, B. fragilis (specifically, the capsular polysaccharide of the organism) is involved in both the induction of abscesses and protection against abscess formation, with T cell activation being an important component of both arms of the pathway. (See 'Capsular polysaccharides' above.)
Bacterial synergy — The polymicrobial nature of the majority of anaerobic infections was appreciated by French investigators in the early 1900s. One of the most common observations is that they are polymicrobial, often with aerobic bacteria. In these settings, oxygen utilization by aerobes is thought to facilitate the growth of anaerobes, representing a form of synergy. However the molecular mechanism(s) of synergy is usually unclear [31]. (See "Anaerobic bacteria: History and role in normal human flora".)
CLUES TO ANAEROBIC INFECTIONS — Clinical clues to the probable presence of anaerobic bacteria at infected sites are summarized in the table (table 1). Infections that occur in continuity with mucosal surfaces where anaerobic bacteria compose the normal flora often involve these microbes. Thus, the assumption is that most cases of orodental infection, aspiration pneumonia, and intra-abdominal infection involve anaerobes. These assumptions are usually correct.
Anaerobes are often associated with tissue necrosis and abscess formation, as supported by their frequency in abscesses at virtually all anatomic sites, including cerebral, dental, peritonsillar, lung, intra-abdominal, tubo-ovarian, prostatic, and cutaneous abscesses. (See "Anaerobic bacterial infections".)
The putrid odor of infections or discharges, although diagnostic of infection containing anaerobes, may be a relatively late feature and is seen in only approximately one-third to one half of patients [42,43]. The chemical basis for the odor is not well established, but it presumably reflects the metabolic products of anaerobic bacteria, including volatile fatty acids (eg, succinic and butyric acid and methylmercaptan). As noted, these short-chain fatty acids are used to identify anaerobic bacteria; direct detection of these acids in exudate by gas-liquid chromatography may be used as an early clue to the presence of anaerobes [44]. For clinical application purposes, the putrid odor is considered diagnostic of anaerobic infection.
Gas in the tissue is another clue to the presence of anaerobic bacteria, but it is not considered diagnostic because occasional aerobic bacteria produce gas [44]. Gas may be detected by palpation, radiography, or scanning techniques. This may reflect not only gas production by microbes but also air introduced during irrigations or other manipulations, such as the release of carbon dioxide with hydrogen peroxide [45].
Anaerobic infections often involve a polymicrobial flora, so a Gram stain of exudate usually shows multiple different morphotypes at infected sites. The Gram stain may also show the unique morphologic features of many anaerobes, especially Bacteroides spp (small, delicate gram-negative bacilli), Fusobacterium spp (F. nucleatum: fusiform bacteria with pointed ends; F. necrophorum: long, "ropy" gram-negative bacilli), and Clostridia (large "box car"-like gram-positive bacilli). By contrast, Peptostreptococcus cannot be distinguished from aerobic gram-positive cocci on the basis of Gram stain appearance.
METHODS FOR RECOVERING ANAEROBES — The paucity of anaerobes in clinical laboratories reflect several inter-related factors:
●Specimens for clinical culture are most successful when devoid of contamination by the normal flora.
●The organisms require anaerobic conditions for growth.
●Most anaerobic bacterial infections involve multiple different bacteria that grow slowly in complex communities and require extensive testing for identification.
●Prior antibiotic therapy substantially reduces cultivability for these bacteria.
Culturing specimens for anaerobes is not always performed because anaerobic cultures are tedious and identification is expensive and slow. Many clinicians also feel that it is fairly easy to predict the presence of anaerobes and appropriate treatment is usually empiric based upon the site of the infection.
Specimen selection — Optimal clinical specimens for anaerobic culture are normally sterile body fluids and aspirates or biopsy material from normally sterile sites (table 2) [46]. On occasion, the problem of contamination may be obviated by quantitative cultures, although most laboratories do not provide this service. Some laboratories will report growth as being heavy or light, which can be clinically useful information. As a general rule, liquid or tissue specimens are preferred; swab specimens should be avoided.
Specimen transport — The optimal way to transport specimens is with anaerobic transport tubes or by immediate delivery to permit prompt microbiologic processing. When specimens need to be held for 24 hours, they should be stored at 4ºC [47,48]. A variety of specialized transport devices are available that provide an atmosphere of oxygen-free gas, such as mixtures of carbon dioxide, hydrogen, and nitrogen [49-51]. Most of these devices also include an indicator, such as resazurin, to document anaerobic conditions and a reducing agent, such as cysteine, to eliminate small amounts of oxygen that are inadvertently introduced.
Tissue specimens may be placed in a sterile tube flushed with carbon dioxide; if the tube is held upright with the stopper removed, the heavier carbon dioxide will exclude oxygen until the stopper is replaced [46]. Although they are theoretically attractive, there is little evidence that such techniques are actually necessary for fluid specimens [52]. Swabs are different due to drying rather than oxygen sensitivity. When swab specimens are necessary, the best transport medium is a specially prepared, pre-reduced, and anaerobically sterilized semisolid medium such as Cary-Blair medium.
Laboratory processing — Gram stain of exudate is an important clue to the presence of anaerobic bacteria due to their often unique morphology. The Gram stain is also an important method of quality assurance of microbiology culture technique. Two systems are generally advocated for recovering anaerobes in culture:
●The anaerobic jar (GasPak, evacuation-replacement, and Bio-Bags)
●The anaerobic glovebox
Several studies indicate comparable results in the yield of anaerobes from clinical specimens with these various systems [46]. Consequently, the decision of which system to use depends upon previous training of personnel, the volume of specimens, and the resources of the laboratory. Many clinical laboratories have found the GasPak jar method particularly convenient, although the jar should remain inviolate for at least 48 hours after the GasPak has been generated. An alternative, more convenient method for one or two plates is Bio-Bags. Anaerobic chambers may be preferred when a large number of specimens are processed.
Quality assurance with anaerobic technology indicates that the major discrepancies in most clinical laboratories are:
●Failure to use the initial Gram-stained specimen to ensure that culture results account for all recognized morphotypes
●Improper use of anaerobic jars, especially by opening before 48 hours
●Failure to use pre-reduced plate media
●Premature discarding of plates
●Inadequate picking of colonies from primary isolation plates
●Inadequate selective and nonselective plate media or excessive dependence on broth cultures
Identification — The major clinically significant anaerobes are summarized in the table (table 3). Most clinical laboratories should be able to identify these organisms. Peptostreptococci, which account for approximately 25 percent of all anaerobic isolates in most microbiology laboratories, rarely merit speciation because these organisms lack distinctive virulence properties and appear to be predictably susceptible to the same antimicrobial agents. The non-spore-forming gram-positive bacilli require chromatographic analysis for genus designation and extensive biochemical testing for speciation. With the exception of Actinomyces, these organisms have minimal documented pathogenic potential, and cursory identification is generally adequate.
A rational approach is to separate Cutibacterium (formerly Propionibacterium), a common contaminant, from the others simply by a catalase test and indole reaction. Clostridia are identified by spores seen on Gram stain examination, by survival with exposure to ethanol for 30 minutes, or by survival following heating to 80ºC for 10 minutes. Extensive biochemical testing is required for speciation, which generally is unnecessary except in selected cases, such as gas gangrene. C. difficile may be cultured from stool, but antigen assays, toxin assays, and polymerase chain reaction are more practical. (See "Clostridioides difficile infection in adults: Clinical manifestations and diagnosis".)
The important gram-negative anaerobes for identification include the B. fragilis group, Prevotella spp, and Fusobacterium spp. The B. fragilis group (B. fragilis, B. thetaiotaomicron, B. distasonis, B. ovatus, and B. vulgatus) is distinguished by the ability to grow in the presence of 20 percent bile, and most are catalase positive. Formerly classified as B. melaninogenicus, pigmented Prevotella (P. melaninogenica, P. corporis, P. denticola, P. intermedia, P. loescheii, and P. nigrescens) are a clinically important group of anaerobes that are relatively fastidious. The frequency of recovery of these species depends upon the expertise of the laboratory. Fusobacteria are also relatively fastidious. These organisms can be distinguished from Bacteroides spp because they are susceptible to the 1000 mcg kanamycin disk, are indole positive and nonmotile, produce butyric acid, and show distinctive morphologic features on Gram stain (slender, elongated gram-negative rods).
Cutibacterium is usually a contaminant in clinical samples but may play an important role in septic arthritis complicating shoulder prostheses or infections following neurosurgery [53].
Susceptibility testing — The methods to test the antibiotic susceptibility of anaerobes are well standardized but not advocated for clinical laboratories. The reasons are that the standard method is difficult, it often takes many days to separate and test the components of these mixed infections, the testing is expensive, and most anaerobes have predictable susceptibility profiles.
SUMMARY
●Anaerobic bacteria are the major constituents of normal human flora and have been recovered from a wide array of different infections. Isolating the organisms from specimens required discovery of specialized methods in the microbiology laboratory, but it has been more problematic to determine when these bacteria represent true pathogens or merely commensals. (See 'Introduction' above.)
●Anaerobic infections nearly always arise from contamination by endogenous bacteria into contiguous or other sites. Important exceptions are some of the clostridial syndromes, including botulism, Clostridium perfringens food poisoning, enteritis necroticans, tetanus, some cases of gas gangrene, and Clostridioides difficile-associated diarrhea. (See 'Pathophysiology of anaerobic infections' above.)
●Anaerobic bacteria contain a number of virulence factors, including toxins, polysaccharide capsules, and lipopolysaccharides. (See 'Virulence factors' above.)
●An abscess is a unique biologic phenomenon, representing the combination of a pathologic process and a host defense to contain infection. Within the fibrous capsule of a well-formed abscess, bacteria and neutrophils coexist; these same neutrophils in vitro will kill the bacteria. (See 'Formation of and host defense to abscesses' above.)
●The polymicrobial nature of the majority of anaerobic infections was appreciated by investigators in the early 1900s. One of the most common observations is that they are polymicrobial, often with aerobic bacteria. This may represent synergy, but the mechanism of synergy is usually unclear. (See 'Bacterial synergy' above.)
●Clinical clues to the probable presence of anaerobic bacteria at infected sites are summarized in the table (table 1). Infections that occur in continuity with mucosal surfaces where anaerobic bacteria compose the normal flora often involve these microbes. Thus, the assumption is that most cases of orodental infection, aspiration pneumonia, and intra-abdominal infection involve anaerobes. Anaerobes are often associated with tissue necrosis and abscess formation as supported by their frequency in abscesses at virtually all anatomic sites, including cerebral, dental, peritonsillar, lung, intra-abdominal, tubo-ovarian, prostatic, and cutaneous abscesses. (See 'Clues to anaerobic infections' above.)
●The putrid odor of infections or discharges, while diagnostic of infection containing anaerobes, may be a relatively late feature and is seen in only approximately one-third to one half of patients. Gas in the tissue is another clue to the presence of anaerobic bacteria, but it is not considered diagnostic because occasional aerobic bacteria produce gas. Anaerobic infections usually involve a polymicrobial flora, so a Gram stain of exudate usually shows multiple different morphotypes at infected sites. (See 'Clues to anaerobic infections' above.)
●The paucity of anaerobes in clinical laboratories reflects several interrelated factors:
•Specimens for meaningful culture must be devoid of contamination by the normal flora.
•The organisms require anaerobic conditions for growth.
•Most anaerobic bacterial infections involve multiple different bacteria that grow slowly and require extensive testing for identification.
•Prior antibiotic therapy substantially reduces cultivability for these bacteria. (See 'Methods for recovering anaerobes' above.)
●Culturing specimens for anaerobes is not always performed because anaerobic cultures are tedious and identification is expensive and slow. Many clinicians also feel that it is fairly easy to predict the presence of anaerobes, and appropriate treatment is usually empiric based upon the site of the infection. (See 'Methods for recovering anaerobes' above.)
●Optimal specimens are normally sterile body fluids and aspirates or biopsy material from normally sterile sites (table 2). (See 'Specimen selection' above.)
●The optimal way to transport specimens is with anaerobic transport tubes or by immediate delivery to permit prompt microbiologic processing. (See 'Specimen transport' above.)
●Gram stain of exudate is an important clue to the presence of anaerobic bacteria due to their often unique morphology. The Gram stain is also an important method of quality assurance of microbiology culture technique. (See 'Laboratory processing' above.)
●The major clinically significant anaerobes are summarized in the table (table 3). Most clinical laboratories should be able to identify these organisms. (See 'Identification' above.)
●In vitro susceptibility tests are usually not done by clinical laboratories for anaerobes because it is technically difficult, results are usually not available in sufficient time to impact antibiotic decisions, and treatment based on empiric selection is usually effective. (See 'Susceptibility testing' above.)
ACKNOWLEDGMENT — The editorial staff at UpToDate would like to acknowledge John G Bartlett, MD, who contributed to an earlier version of this topic review.
