Definitive diagnosis of most bacterial diseases depends on specific identification of the pathogenic organisms. This must often be accomplished by appropriate cultures or molecular techniques because special bacterial stains frequently fail to detect sparse numbers of bacteria. Studies of infectious cellulitis show that even cultures yield organisms in less than 10% of cases. Cultures from actual biopsies have a higher yield than cultures originating from injected saline aspirates. Swab cultures from pustules, ulcers, or crusted lesions are suitable for only the most superficial infections, such as impetigo. An ideal method for deeper infections is to send half of the biopsy for culture and the other half for hematoxylin and eosin (H&E) and special stains. Before biopsy, excess antiseptic, such as chlorhexidine or povidone-iodine, which may inhibit the culture, should be wiped off the skin. Cultures are best performed on tissue that is rushed to the laboratory before it desiccates. It should be minced and added immediately to the culture medium. If there is a chance that a biopsy might dry during transport, the biopsy should be placed on sterile gauze soaked in normal saline (without preservatives such as benzyl alcohol).
Even if the organisms are stained, it may be impossible to distinguish similar pathogens from normal flora. The location of the bacteria is important to note. Gram-positive cocci and coccobacilli, such as staphylococci and corynebacteria, are normal flora within the follicles or secondarily on the surface of crusted, eroded, or ulcerated conditions (Fig. 19-1A).1,2 In contrast, gram-positive cocci in the dermis or deeper tissue indicate a significant infectious process such as infectious cellulitis (dermis), panniculitis (fat), or fasciitis (fascia). If the follicles are disrupted and infiltrated by neutrophils, then bacteria in the follicles may be pathogenic. They can also be secondary invaders in cases of fungal folliculitis and noninfectious folliculitis; therefore, in such cases, it is important for pathologists finding bacteria to consider looking for other etiologic agents and to remember that the primary condition may not be infectious. In immunosuppressed patients, particularly those with acquired AIDS, it is well known that a single biopsy specimen may contain more than 1 pathogen; consequently, the pathologist should not stop after identifying 1 organism.3
Bacterial colonization. (A) Numerous bacteria on the surface of a crust is colonization, not true infection. For this reason, swab cultures of ulcers have limited value in assessing deep infection. (B) Impetigo. Subcorneal pustule containing neutrophils and occasional acantholytic keratinocytes.
Smears or touch preparations from tissue may improve the chances of detecting organisms. This is particularly true for acid-fast bacillus (AFB) infections, chancroid, and granuloma inguinale, in which stains of tissue sections have a lower yield. The Fite modification of the AFB stain may be better than the ...