The Critical Role of Pathology in the Investigation of Bioterrorism-Related Cutaneous Anthrax

Abstract

Cutaneous anthrax is a rare zoonotic disease in the United States. The clinical diagnosis traditionally has been established by conventional microbiological methods, such as culture and gram staining. However, these methods often yield negative results when patients have received antibiotics. During the bioterrorism event of 2001, we applied two novel immunohistochemical assays that can detect Bacillus anthracis antigens in skin biopsy samples even after prolonged antibiotic treatment. These assays provided a highly sensitive and specific method for the diagnosis of cutaneous anthrax, and were critical in the early and rapid diagnosis of 8 of 11 cases of cutaneous anthrax during the outbreak investigation. Skin biopsies were obtained from 10 of these 11 cases, and histopathological findings included various degrees of ulceration, hemorrhage, edema, coagulative necrosis, perivascular inflammation, and vasculitis. Serology was also an important investigation tool, but the results required several weeks because of the need to test paired serum specimens. Other tests, including culture, special stains, and polymerase chain reaction assay, were less valuable in the diagnosis and epidemiological investigation of these cutaneous anthrax cases. This report underscores the critical role of pathology in investigating potential bioterrorism events and in guiding epidemiological studies, a role that was clearly demonstrated in 2001 when B. anthracis spores were intentionally released through the United States postal system.

Anthrax captured worldwide attention and aroused serious public health concerns following the intentional release of the etiological agent Bacillus anthracis in the United States postal system during Fall 2001, 1,2 resulting in 11 cases of inhalational anthrax and 11 cases of cutaneous anthrax. 3-7 In its conventional form, cutaneous anthrax accounts for 95% of all naturally occurring B. anthracis infections in the United States. 8-11 Patients often have a history of occupational contact with animals or animal products contaminated with B. anthracis spores. 12-14 These pathogenic spores are introduced through a cutaneous cut or abrasion, with the most common areas of exposure are the head, neck, and extremities, although any area can be involved. Bacteremia and toxemia following cutaneous infection can occur with a fatality rate of 20% to 25% among untreated cases. 15-17

Cutaneous anthrax is characterized by the formation of a black eschar surrounded by prominent edema and vesicles, which may resemble many other skin lesions, such as the brown recluse spider bite, 18-21 ulceroglandular tularemia, 22,23 plague, 24,25 ecthyma gangrenosum, 26,27 various spotted fever group rickettsial infections, 28-31 and scrub typhus. 32,33 The clinical diagnosis of cutaneous anthrax is traditionally established by microbiological methods (eg, demonstrating gram-positive, capsulated bacilli on the smear of the lesion or isolating B. anthracis in culture). 34,35 However, gram stain and culture for B. anthracis can be unrevealing for patients who receive antibiotic therapy before specimens are obtained. 36,37 Historically, skin biopsies of cutaneous anthrax lesions have seldom been performed, and the histopathological features have been described in only a few untreated human cases and in experimental animals. 38-41

In this report, we describe the histopathological features and the immunohistochemical (IHC) findings of bioterrorism-related cutaneous anthrax cases of 2001, and compare the results of IHC assays with other laboratory diagnostic methods, including culture, special stains, polymerase chain reaction (PCR), and serology. The critical role of pathology in directing the investigation of the outbreak is also discussed.

Materials and Methods

From October 1 through December 31, 2001, 117 formalin-fixed skin biopsies from patients with clinically suspected cutaneous anthrax were submitted for testing to the Infectious Disease Pathology Activity (IDPA), Centers for Disease Control and Prevention (CDC). Some clinical and demographic data accompanied each specimen, although they varied in completeness. Serum, blood, wound swab, and other clinical samples were available in some cases. Serology and reverse transcription (RT)-PCR were performed by other laboratories at CDC. 42-44 All skin biopsies were examined by using hematoxylin and eosin (H&E)-stained sections for histopathological evaluation. Special stains, including gram stain and Steiner’s silver stain, were used to detect bacterial agent in tissue sections. A colorimetric immunoalkaline phosphatase IHC method was developed by using two monoclonal antibodies as described below. In brief, 3-μm sections from formalin-fixed, paraffin-embedded tissues were deparaffinized, rehydrated, and placed in a DAKO autostainer (DAKO Corporation, Carpinteria, CA). Three sections were incubated for 1 hour with a mouse monoclonal IgM antibody reactive with B. anthracis cell wall antigen at 1:200 dilution (USA Military Research Institute of Infectious Disease, Frederick, MD). 45 Three other sections were incubated for 1 hour with a mouse monoclonal IgM antibody reactive with B. anthracis capsule antigen at 1:1000 dilution. Slides to be incubated with anti-cell wall antibody were first digested in 0.1 mg/ml proteinase K (Boehringer-Mannheim Corporation, Indianapolis, IN), but slides incubated with anti-capsule antibody were not pre-digested. Optimal dilutions of the antibodies and the requirement for pre-digestion were determined by a series of pilot studies performed on positive control samples (described later). After incubation, slides were washed and incubated with a biotinylated anti-mouse IgM antibody. Antigens were visualized by using a strepavidin-alkaline phosphatase complex followed by naphthol/fast red substrate for colorimetric detection (DAKO Corporation). Sections were counterstained with Mayer’s hematoxylin (Fisher Scientific, Pittsburgh, PA).

Positive controls included sections prepared from formalin-fixed, paraffin-embedded B. anthracis grown in culture and tissues obtained from confirmed animal and fatal human anthrax cases. 46 Negative controls consisted of the following: 1) each patient’s tissue sections incubated with an IgM antibody reactive with an irrelevant infectious agent, 2) cutaneous lesions caused by non-anthrax etiologies, including human herpesvirus 1 (Herpes simplex 1), human herpesvirus 3 (Varicella-zoster virus), human herpesvirus 5 (Cytomegalovirus), Fransicella tularensis, spotted fever group rickettsiae, and spider bite, and 3) formalin-fixed, paraffin-embedded Bacillus subtilis, Bacillus cereus, Clostridium novyi, Clostridium sordelli, Streptococcus pneumoniae, Group A Streptococcus, Staphylococcus aureus, Staphylococcus epidermidis, Pseudomonas, Klebsiella pneumoniae, Coccidiodes, Blastomyces dermatitidis, Cryptococcus neoformans, and Histoplasma capsulatum.

A confirmed case of cutaneous anthrax was defined as a clinically compatible case that was either laboratory confirmed by isolation of B. anthracis from an affected tissue or site, or accompanied with other laboratory evidence of B. anthracis infection based on at least two supportive laboratory tests, including 1) evidence of B. anthracis DNA by PCR from specimens collected from an affected tissue or site, 2) demonstration of B. anthracis in a clinical specimen by IHC, or 3) four-fold rise in anti-protective antigen (PA) IgG. A suspect case of cutaneous anthrax was defined as a clinically compatible illness with no alternative diagnosis and no isolation of B. anthracis, but with either 1) laboratory evidence of B. anthracis by one supportive laboratory test or 2) an epidemiological link to an environmental B. anthracis exposure. 6,7,47

Results

During the bioterrorism event of 2001, skin biopsy samples of 117 cases were tested at IDPA and 8 were positive for B. anthracis by IHC. There were two cases with detectable levels of serum antibody to anthrax toxin PA, but IHC and other laboratory tests were negative. Selected demographic information, histopathological features, laboratory test results, and case status for the 8 IHC-positive cases and the 2 IHC-negative, serology-reactive cases are summarized in Table 1 ▶ and Table 2 ▶ . Based on the case definition for cutaneous anthrax, there were 7 confirmed cases: 2 based on culture and other diagnostic tests, including IHC (numbers 5 and 8); 2 based on IHC and serology (numbers 1 and 3); 2 based on all three supportive tests: IHC, serology and PCR (numbers 4 and 7); and 1 based on IHC and PCR (number 2). There were 3 suspect cases: 1 based only on IHC (number 6); 2 based only on serology (numbers 9 and 10). One suspect case, described in another report, 7 was excluded because no skin biopsy sample was available for histopathological evaluation and IHC testing.

Table 1.

Demographic Information, Histopathologic Features, Special Stains and Bacillus anthracis Immunostaining Results

Case no./case definition and status*	Age/sex	Duration of illness/ treatment† (days)	Routine histopathologic examination		Bacillus anthracis immunostaining
			H&E stain	Special stains
1/Confirmed	38/F	14/9	Epidermis: mild acantholysis, mild PMN infiltrate	Negative	Cell wall: rare bacilliform, focal granular
			Dermis: edema, focal necrosis, focal hemorrhage, vasculitis, intense mononuclear perivascular infiltrate
						Capsule: focal rare granular
2/Confirmed	0.6/M	14/13	Epidermis: necrosis, hemorrhage, acantholysis, focal mild PMN infiltrate	Negative	Cell wall: focal rare bacilliform and granular
			Dermis: edema, necrosis, focal mild mononuclear perivascular infiltrate		Capsule: widespread bacilliform and granular
3/Confirmed	27/F	15/12	Epidermis: necrosis, hemorrhage, ulceration	Negative	Cell wall: focal bacilliform and granular
			Dermis: edema, necrosis, hemorrhage, vasculitis, intense mononuclear perivascular infiltrate
						Capsule: focal bacilliform and granular
4/Confirmed	46/F	19/16	Epidermis: necrosis, hemorrhage, ulceration, PMN infiltrate	Gram stain: mixed gram-positive and gram-negative cocci	Cell wall: focal abundant bacilliform and granular
			Dermis: edema, necrosis, hemorrhage, vasculitis, intense focal mononuclear perivascular infiltrate	Steiner’s stain: mixed cocci and bacilli
							Capsule: focal abundant bacilliform and granular
5/Confirmed	35/M	3/1	Epidermis: necrosis, hemorrhage, acantholysis, suprabasilar cleft, ulceration, mild PMN infiltrate	Gram stain: Negative	Cell wall: widespread abundant bacilliform and granular
						Steiner’s stain: abundant bacilli in dermis
			Dermis: edema, necrosis, hemorrhage, vasculitis, intense mixed perivascular infiltrate		Capsule: widespread abundant bacilliform and granular
6/Suspect	34/M	5/3	Epidermis: necrosis, hemorrhage, acantholysis, bullous, intense PMN infiltrate	Gram stain: Negative	Cell wall: widespread abundant bacilliform and granular
						Steiner’s stain: focal bacilli in superficial dermis
			Dermis: edema, necrosis, hemorrhage, vasculitis, intense diffuse PMN infiltrate		Capsule: widespread abundant bacilliform and granular
7/Confirmed	51/F	7/6	Epidermis: necrosis, hemorrhage, ulceration, mild focal PMN infiltrate	Gram stain: a few gram-positive bacilli in epidermis and superficial dermis	Cell wall: focal abundant bacilliform and granular
			Dermis: edema, necrosis, hemorrhage, vasculitis, intense mixed perivascular infiltrate
					Steiner’s stain: abundant bacilli in epidermis and superficial dermis	Capsule: focal abundant bacilliform and granular
8/Confirmed	38/M	5/0	Epidermis: necrosis, suprabasillary cleft, mild PMN infiltrate, bacteria	Gram stain: abundant gram-positive bacilli in focal epidermis and superficial dermis	Cell wall: focal abundant bacilliform, less granular
			Dermis: edema, necrosis, hemorrhage, vasculitis, intense focal mononuclear perivascular infiltrate
							Capsule: focal abundant bacilliform and granular
					Steiner’s stain: abundant bacilli in focal epidermis and superficial dermis
9/Suspect	23/F	10/8	Epidermis: necrosis, acantholysis, superficial crust, mild PMN infiltrate	Negative	Negative‡
			Dermis: edema, necrosis, hemorrhage, vasculitis, moderate mixed perivascular infiltrate
10/Suspect	31/F	20/13	Epidermis: small biopsy, no significant change	Negative	Negative§
			Dermis: mild edema, mild mixed perivascular infiltrate

*Case definition and status are described in the Methods section. One suspect case was excluded because skin biopsy was unavailable for pathologic evaluation.

^†Duration of illness/antibiotic treatment prior to biopsy.

^‡Patients had multiple lesions (cheek, buttock, thigh, and lower leg); the biopsy was taken from left thigh unrelated to the primary lesion in the cheek.

^§Contained a small, superficial, shave biopsy that was inadequate for pathologic evaluation.

Table 2.

Test Results of IHC, Culture, PCR, and Serology

Case number/status*	IHC	Culture	PCR	Serology
1/Confirmed	Positive	Negative (nasal and wound swab)	Negative (serum, swab, formalin-fixed skin biopsy)	Reactive
2/Confirmed	Positive	Negative (blood and wound swab)	Positive (one of eight blood samples) Negative (formalin-fixed skin biopsy)	NA
3/Confirmed	Positive	Negative (serum)	Negative (serum and frozen skin biopsy)	Reactive
4/Confirmed	Positive	Negative (blood)	Positive (formalin-fixed skin biopsy) Negative (blood and serum)	Reactive
5/Confirmed	Positive	Positive (blood) Negative (nasal swab and fresh skin); Staphylococcus species (wound)	Negative (frozen skin biopsy)	Reactive
6/Suspect	Positive	Negative (blood and wound swab)	Negative (blood and frozen skin biopsy)	Negative
7/Confirmed	Positive	Negative (blood, wound swab, and fresh skin)	Positive (frozen skin biopsy) Negative (serum)	Reactive
8/Confirmed	Positive	Positive (wound swab) Negative (blood and fresh skin)	Negative (serum and frozen skin biopsy)	Reactive
9/Suspect	Negative†	Negative (blood, serum, and nasal swab)	Negative (blood, serum, nasal swab, and formalin-fixed skin biopsy)	Reactive
10/Suspect	Negative‡	NA	Negative (serum and frozen skin biopsy)	Reactive

NA, not available.

*Case definition and status are described in the Methods section. One suspect case was excluded because skin biopsy was unavailable for pathologic evaluation.

^†Patients had multiple lesions (cheek, buttock, thigh, and lower leg); the biopsy was taken from left thigh unrelated to the primary lesion in the cheek.

^‡Contained a small, superficial, shave biopsy that was inadequate for pathologic evaluation.

Microscopic examination of the epidermis of these 10 biopsy specimens showed 8 with necrosis, 6 with hemorrhage, 6 with acantholysis or bullous formation, and 4 with various degrees of epidermal ulceration (Figure 1 ▶ , Figure 2 ▶ , and Figure 3 ▶ ). Examination of the dermis showed various degrees of edema in all 10 specimens, coagulation necrosis in 9, and hemorrhage and vasculitis in 8 each (Figures 1 to 3) ▶ . Various degrees of inflammation were present in the lower epidermis and dermis of all case specimens (Figure 1A ▶ , Figure 2B ▶ , and Figure 3, A and C ▶ ). Polymorphonuclear infiltrates were often present with prominent spongiosis in the lower epidermis (Figure 3A) ▶ . The dermis was usually markedly edematous, with separation of collagen bundles, and often contained an intense perivascular inflammatory infiltrate (Figure 1A ▶ and Figure 3C ▶ ). Inflammatory changes were not consistent among cases, and varied from mixed to predominantly mononuclear (Table 1) ▶ .

Figure 1.

Histopathology and immunohistochemical staining on cutaneous anthrax index case (case 1). This patient received antibiotics treatment for 9 days before the biopsy was taken. A: Skin biopsy from the edge of eschar showed intact epidermis; the dermis showed edema, focal hemorrhage, and perivascular mononuclear cell infiltrates. Superficial ulceration and fibrinopurulent membrane typically associated with acute lesions were not seen. B: Focal, rare immunohistochemical staining of B. anthracis cell wall antigen was present in the dermis. H&E stain (A); immunoalkaline phosphatase with napthol fast red substrate and hematoxylin counterstain (B). Original magnifications: ×25 (A); ×250 (B).

Figure 2.

Histopathology, special stains, and immunohistochemical staining on case 7. This patient received antibiotics treatment for 6 days before the biopsy was taken. A: Skin biopsy showed superficial ulceration, dermal edema, necrosis, and hemorrhage. B: Higher-power magnification showed a mixed inflammatory infiltrate in the dermis. C: Steiner’s silver stain showed many bacilli in the dermis. D: Gram stain in the same area showed only a few gram-positive bacilli (arrowheads). E: Abundant immunohistochemical staining of granular and bacilliform antigens were demonstrated in the dermis by using anti-B. anthracis cell wall antibody. F: Immunohistochemical staining of B. anthracis capsule antigens was present in fibrohistiocytic cells. H&E stain, A and B; Steiner’s silver stain, C; gram stain, D; immunoalkaline phosphatase with napthol fast red substrate and hematoxylin counterstain, E and F. Original magnifications: ×12.5 (A); ×100 (B and F); ×250 (C and D); ×158 (E).

Figure 3.

Histopathology and immunohistochemical staining on case 6. This patient received antibiotics treatment for 3 days before the biopsy was taken. A: Skin biopsy showed acantholysis, diffuse polymorphonuclear cell infiltrate, dermal edema, necrosis, and hemorrhage. B: Immunohistochemical staining showed focal bacilliform antigen staining by anti-B. anthracis cell wall antibody in the same area as A. C: Vasculitis and intense polymorphonuclear cell infiltrate were present in the dermis. D: Immunohistochemical staining of granular and bacilliform antigens were demonstrated by using anti-B. anthracis capsule antibody in the same area as C. H&E stain, A and C; immunoalkaline phosphatase with napthol fast red substrate and hematoxylin counterstain, B and D. Original magnifications: ×100 (A and C); ×250 (B); ×158 (D).

Bacilli morphologically compatible with B. anthracis were detected by routine H&E stain in only 1 case (number 8) (Figure 4C) ▶ . Gram staining of tissue demonstrated unequivocal gram-positive bacilli in 2 cases (numbers 7 and 8) (Figure 2D ▶ and Figure 4D ▶ ), and Steiner’s silver stain showed bacilli compatible with B. anthracis in 4 cases (numbers 5, 6, 7, and 8) (Figure 2C) ▶ . Case number 4 showed mixed cocci and bacilli in the biopsy samples and was difficult to evaluate. IHC staining revealed B. anthracis antigens in 8 cases, and all of them were positive by using both anti-cell wall and anti-capsule antibody (Table 1) ▶ . Extracellular bacilliform and granular antigens were distributed in both epidermis and dermis (Figures 1 to 4) ▶ . Intracellular immunostaining was also observed in fibrohistiocytic cells, and usually more prominent by anti-capsule antibody (Figure 3D) ▶ . The amount and distribution of bacterial antigens among cases varied from rare (Figure 1B ▶ and Figure 3B ▶ ) to abundant (Figure 2E ▶ and Figure 4A ▶ ) and from focal to widespread (Table 1) ▶ . Of the two skin biopsies with negative IHC results (numbers 9 and 10), one was obtained from a skin lesion unrelated to the primary lesion, and the other was an extremely small, superficial, shave biopsy that was inadequate for pathological evaluation.

Figure 4.

Histopathology, special stain, and immunohistochemical staining on case 8. This patient did not receive antibiotics treatment before the biopsy was taken. A: Abundant immunohistochemical staining of B. anthracis capsule antigens was present extracellularly and intracellularly. Inset: Higher-power magnification showed bacilliform and granular antigens. B: Immunohistochemical staining of B. anthracis cell wall antigens was present focally. The amount of immunostaining is less than capsule antigens in A. Inset: Higher-power magnification showed bacilliform and granular antigens. C: H&E stain showed encapsulated bacilli in the dermis. This is the only case in which morphologically compatible bacilli were discernable by routine H&E stain. D: Gram stain demonstrated abundant gram-positive bacilli in the same area as C. Immunoalkaline phosphatase with napthol fast red substrate and hematoxylin counterstain, A and B. H&E stain, C; gram stain, D. Original magnifications: ×50 (A and B); ×250 (A, inset; B, inset; C and D).

Discussion

Since 1950, human anthrax in the United States was confined to those occupationally at risk and its incidence has been very low owing to improvements in animal husbandry and the handling of animal products. 48,49 There were 235 confirmed cases reported from 1955 to 2002, and most of them were cutaneous anthrax. 8,10 Before autumn 2001, all cases of cutaneous anthrax in the United States were related to agricultural or industrial exposures. However, the epidemiology and laboratory diagnosis of cutaneous anthrax changed after October 2001 because of the intentional release of B. anthracis spores through the U.S. postal system and the subsequent identification of anthrax in 22 patients, including 11 with cutaneous disease. 6,7

The index cutaneous anthrax case-patient from the 2001 outbreak, a 38-year-old woman who worked at a major news network, initially developed an ulcerative lesion on the chest and was seen by an infectious disease clinician. She reported handling “threat” letters at work. Bacterial cultures of specimens both from the lesion and from a suspicious letter (dated September 25) tested negative for B. anthracis, and the patient was administered ciprofloxacin.

After learning of a patient with inhalational anthrax in Florida, 4,6 the patient contacted the New York City Department of Health, and a skin biopsy was obtained 9 days after antibiotic treatment began. The biopsy was negative for B. anthracis by PCR, and gram stain did not reveal any bacteria. However, a rapid IHC test for B. anthracis showed focal granular and rare bacilliform antigen staining in the dermis (Figure 1B) ▶ establishing the diagnosis of cutaneous anthrax.

Identification of this index cutaneous anthrax case prompted a soaring public health concern with ensuing actions, such as mobilization of multidisciplinary teams for onsite investigations and initiation of antibiotic distribution at clinics. The diagnosis also led to the identification of another threat letter (dated September 18) containing B. anthracis spores. In the following days, several other possible cases of anthrax were reported at other major media outlets, and in most of them the diagnosis was established by IHC. 4,5 The epidemiology of anthrax in this outbreak was shifted from the traditional occupational hazard of zoonotic contact to a deliberate bioterroristic action involving the U.S. postal system. 50

Cutaneous anthrax lesions can be difficult to diagnose clinically and histopathologically. During investigation of the cutaneous anthrax cases, skin lesions were frequently mistaken for other diseases, including those caused by human herpes simplex virus, Varicella zoster virus, rickettsial pox, and arthropod bite. These diseases were diagnosed by IHC for specific agents and by histopathological evaluation as well. In this report, we found that histopathological features of cutaneous anthrax lesions can vary with the duration of illness, modality of treatment, and site of biopsy in relation to the lesion. In general, perivascular infiltrates are predominantly mononuclear in cases with longer duration of illness and treatment course, while polymorphonuclear infiltrates are more prominent in cases with shorter duration (Table 1) ▶ . Biopsies obtained from the edge of an eschar did not show superficial ulceration with fibrinopurulent membrane typically associated with biopsies taken from the center of eschar of cutaneous anthrax. 38,41

Laboratory evidence was critical in the diagnosis of suspect cutaneous anthrax cases during investigation of this outbreak. Traditionally, cutaneous anthrax was diagnosed by microbiological methods and skin biopsy was seldom done for histopathological examination. Although culture and gram stain of specimens from suspicious skin lesions remain essential methods for diagnosing cutaneous anthrax, their sensitivity decreases significantly when the patient has received antibiotic treatment before specimen collection. In general, the longer the duration of illness and treatment course, the less chance of detecting B. anthracis organism by culture and special stains. The organism is best demonstrated when the lesion is in the vesicular stage, so swab exudates for gram stain and culture should be performed before initiation of antibiotic therapy. Similar observation in histopathological evaluation and tissue special stains in this study is shown in Table 1 ▶ . The only case showing encapsulated bacilli in the skin by H&E (number 8, Figure 4C ▶ ) was also the only case that skin biopsy was taken before initiation of antibiotic treatment. The 4 cases (numbers 5, 6, 7, and 8) showing bacilli morphologically compatible with B. anthracis by special stains were those with skin biopsies taken within 1 week of antibiotic treatment. In these cases, Steiner’s silver stain was more useful than gram stain for demonstrating these bacilli in the cutaneous lesion. Because IHC testing of skin biopsies can detect bacterial antigens in tissues regardless of the treatment, it provides a more sensitive and specific way to establish the diagnosis of cutaneous anthrax. Using two monoclonal antibodies against different epitopes of B. anthracis further enhances the sensitivity and specificity of IHC assays.

Other laboratory methods, such as serology and PCR, are also important investigation tools. Serology is very useful for surveillance and diagnosis of cutaneous anthrax, but the results require several weeks because of the need to test paired serum specimens. PCR is a sensitive and specific method to detect bacterial DNA in clinical samples, 43,51-53 but it does not provide a morphological correlation with the result. Moreover, even real-time PCR is not as sensitive as IHC in this study for B. anthracis detection, probably because of the treatment issues.

The following recommendations for obtaining skin biopsies were proposed based on a combination of histopathological findings and localization of bacterial antigens: 1) a full-thickness punch biopsy fixed in 10% buffered formalin from a papule or vesicle lesion and including adjacent skin, and 2) biopsies should be taken from both vesicle and eschar, if present. All biopsy specimens should be accompanied by pertinent clinical information, including a brief history, description, and chronology of the lesion(s); specific treatment; and date of biopsy in relation to antibiotic treatment. A photograph, digital image, or diagram indicating the site of each biopsy in relation to the lesion would be particularly helpful. The above recommendations were adapted by the American Academy of Dermatology and published as part of the management algorithm. 54 Based on the study results described in this report, a skin biopsy should be obtained whenever cutaneous anthrax is suspected, preferably before initiation of antibiotic treatment, and evaluated by IHC testing. Although PCR is usually performed on fresh tissue, it can be done on formalin-fixed tissue as well. Dividing a single biopsy specimen for multiple tests is not recommended because of the focal nature of immunostaining.

Pathology has been a critical component in a multidisciplinary team for detection, surveillance, and research of emerging and reemerging pathogens. 55-59 With the increasing concern of bioterrorism, pathologists are now playing an even more prominent role in both responding to a known terrorist event and in conducting surveillance for unusual deaths or clusters of critical illness that may represent unannounced terrorist activity. The investigation of both inhalational and cutaneous anthrax cases clearly exemplifies these roles. 6,7,60,61 In addition to B. anthracis, several rapid and sensitive IHC assays have been developed for detecting other bioterrorism-related agents, such as Ebola virus, 62,63 Fransicella tularensis, 64 Yersinia pestis, 65 and smallpox virus. These assays provide a method for rapid diagnosis and have fortified the preparedness of responses to possible bioterrorism events.