Clinical manifestations

Presentation and course

There are three principal clinical forms of anthrax, which are categorized by the route of exposure: (1) cutaneous, (2) gastrointestinal, and (3) inhalational. A new form of anthrax, injectional anthrax, has emerged among heroin users in Europe (16; 65; 113). Regardless of the route of exposure, B anthracis can spread through the bloodstream, causing systemic disease, septic shock, and death. Systemic anthrax is typically fatal.

Cutaneous anthrax. Cutaneous anthrax is the predominant clinical form in naturally acquired anthrax, accounting for more than 95% of naturally occurring B anthracis infections (14). Cutaneous anthrax is associated with handling infected animals or animal products (eg, meat, wool, hides, leather, or hair products) (146; 101; 151; 159; 118; 133; 03; 78; 135; 153; 08; 91) and can sometimes be acquired by contamination of preexisting skin lesions with B anthracis spores from the soil (70). Cutaneous anthrax typically develops in exposed skin areas (eg, face, neck, arms, hands) from 2 to 12 days after endospores are introduced into minor skin abrasions, but the incubation period can range from one to over 19 days (141; 67; 216; 146; 159). Pruritic, erythematous papules from 1 to 2 cm in size progress to a vesicle or vesicles containing serosanguineous fluid and then evolve to a painless necrotic ulcer with a black eschar, associated with progressive regional edema and lymphadenitis (71; 76; 79; 175; 105; 52; 146).

Clinical features in severe cases typically include fever, headache, malaise, hemorrhagic bullous lesions surrounded by an extensive erythema and edema, and leukocytosis; in occasional cases, patients develop septic shock (67).

Predictors of developing meningitis or dying include the following: fever or chills; diastolic hypotension; headache; anxiety; nausea or vomiting; abdominal pain; malignant pustule edema; thoracic edema; lymphadenopathy; leukocytosis; bacteremia; and coagulopathy (200).

Gastrointestinal anthrax. Gastrointestinal anthrax occurs from 1 to 7 days following ingestion of large numbers of anthrax endospores, usually associated with eating undercooked contaminated meat (33; 11; 216; 118; 148; 150; 19; 08; 88). Infection may be associated with preexisting lesions in the alimentary tract, especially in the oropharynx (216). Initial symptoms may include fever, anorexia, nausea, vomiting, hematemesis, abdominal pain, bloody diarrhea, and hemorrhagic ascites (57; 33; 105; 216). In a review of the clinical features of hospitalized patients with anthrax, more than half with ingestion of anthrax developed ascites (91). Toxins destroy the mesenteric lymph nodes and the blood supply to the small bowel, producing bowel ischemia, peritonitis, and septicemia.

Inhalational anthrax. Naturally acquired inhalational anthrax is rare in the United States; only 19 cases were identified over the last century, with most occurring in special occupational risk groups (26; 51). Inhalational anthrax typically develops from 2 to 10 days of respiratory exposure to aerosolized anthrax endospores, but cases may present up to 6 or 7 weeks after exposure (141; 216). The mediastinal lymph notes are the nidus of bacterial proliferation. Initial symptoms usually resemble a viral respiratory illness and may include fever, chills, diaphoresis, headache, myalgias, dyspnea, sore throat, cough (usually nonproductive), a feeling of heaviness in the chest, pleuritic pain, malaise, and weakness (141; 57; 43; 40; 35; 51; 105; 10). Some may present with symptoms suggestive of a gastroenteritis or cholecystitis, with nausea, vomiting, diarrhea, and abdominal pain before developing more serious respiratory symptoms (165; 40; 35; 105). Some may present with chest pain of sufficient acuteness and severity to suggest myocardial infarction (02). Chest radiography often reveals a widened mediastinum but does not show pneumonia (10). In a review of the clinical features of hospitalized patients with anthrax, most patients with inhalation of anthrax developed pleural effusions (91). Apparent clinical improvement over several days may be followed by abrupt worsening, with severe respiratory distress, septic shock, and death within 36 hours (57). About half of cases ultimately develop hemorrhagic meningitis (02). Physical findings are usually nonspecific. Based on animal studies using nonhuman primates, the lethal dose is estimated to range from 2500 to 760,000 spores (216).

Injectional anthrax. Since 2009, injectional anthrax has emerged among heroin users in Europe (16; 65; 113). More than 70 cases have been reported from Denmark, France, Germany, and the United Kingdom. Many of these cases presented as a severe soft tissue infection, without a history of animal contact. The unfamiliar characteristics of injectional anthrax have led to diagnostic delays, inadequate treatment, and high case fatality rates (16). Cases of injectional anthrax commonly require surgical debridement (16).

Genome-based characterization of injectional anthrax isolates has identified two tight clusters that imply at least two separate disease events spanning more than 12 years, with one group exclusively associated with the 2009 to 2010 outbreak and located primarily in Scotland, and the second comprised of later (2012 to 2013) cases but also including a single Norwegian case from 2000 (113). Because of the genomic similarity of the two clusters, they were both probably caused by separate contamination events originating from the same geographic region (113).

Anthrax meningoencephalitis. Meningoencephalitis may develop with any clinical type of anthrax, including cutaneous (197; 169; 166; 68; 80; 117; 130; 134; 219; 159; 167), gastrointestinal (196; 130), mixed gastrointestinal and cutaneous (117; 130), injectional (95), and inhalational (05; 165; 02; 29; 130). Meningitis is most common among cases of inhalational anthrax, where it may occur in up to half of cases (05; 165; 25; 26; 02; 57; 130); however, among the 10 cases of confirmed or suspected inhalational anthrax related to the bioterrorist outbreak of anthrax in the United States in 2001, only the index case had documented meningitis (29; 38; 43; 40; 108; 130). Meningitis develops in 5% or fewer of cases of cutaneous anthrax (182; 134); however, most cases of anthrax meningitis develop from cutaneous anthrax, as the vast majority of naturally acquired cases of anthrax are cutaneous (130). Meningitis may occur rarely without a clinically apparent primary focus (69; 169; 166; 76; 130; 11; 81).

Anthrax meningoencephalitis generally presents with fever, headache, vomiting, and confusion or agitation, but there may also be symptoms related to the source of infection (eg, cutaneous, gastrointestinal, or inhalational) (05; 165; 69; 197; 169; 166; 196; 68; 80; 117; 220; 04; 110; 130; 198; 17; 31). Many patients present in extremis following a prodromal period of 1 to 6 days, and two thirds die within 24 hours of admission. Clinical signs on presentation frequently include fever (39.5°C to 41°C), malaise, meningeal signs, hyperreflexia often with bilateral Babinski signs, and delirium or stupor. Other clinical findings may include focal or generalized seizures, myoclonus, fasciculations, generalized rigidity possibly progressing to trismus and opisthotonos, lateralizing signs including unilateral fixed and dilated pupil with contralateral hemiplegia, and decerebrate posturing. Pathologic findings include hemorrhagic meningitis, with extensive edema, inflammatory infiltrates, and numerous gram-positive rods (66; 130; 142). The extensive hemorrhage of the leptomeninges gives them a dark red appearance on gross examination at autopsy, a finding referred to as a "cardinal's cap" (02). Uncommonly, there may be multiple intraparenchymal heterogeneous lesions or areas of ring contrast enhancement and a central diffusion restriction, compatible with abscesses (167).

Atypical anthrax presentations. Atypical presentations of anthrax can occur in patients without known cutaneous, gastrointestinal, or inhalational ports of entry (99). Patients with atypical presentations are less likely to have cough, chest pain, or abnormal lung examination findings (99). Rarely, atypical presentations of anthrax can manifest as meningoencephalitis, so-called “primary anthrax meningitis” (17; 81).

Prognosis and complications

Untreated, the case fatality from cutaneous anthrax is about 20% to 30%, compared to 25% to 60% with gastrointestinal anthrax, and nearly 100% with inhalational anthrax. With treatment, mortality is less than 1% with cutaneous anthrax (57; 216; 92), whereas mortality remains high, even with aggressive treatment, with gastrointestinal or inhalational anthrax (98; 92). Inhalational anthrax was generally thought to be fatal in 80% to over 90% of cases (165; 26; 02; 216), but the mortality in the 2001 U.S. outbreak has been much better than anticipated (5 of 11 cases or 45%) with the use of aggressive treatment and intensive care unit support, including mechanical ventilation and dialysis as necessary (40; 216). Since 2001, 8 of 15 (53%) known patients with inhalation anthrax have survived with early diagnosis, combination antimicrobial drug treatment to eradicate the bacteria and inhibit toxin production, and aggressive pleural effusion management (92). Initiation of antibiotic or anthrax antiserum therapy during the prodromal phase of inhalational anthrax is associated with markedly improved survival compared with initiation of treatment of fulminant cases (98). Nevertheless, anthrax survivors report significant health problems and poor life adjustment one year after the onset of bioterrorism-related anthrax (172).

Anthrax meningoencephalitis is usually rapidly fatal with roughly two-thirds of affected patients dying within 24 hours of presentation (130; 160), although there a small number of reported survivals following anthrax meningoencephalitis (61; 182; 209; 197; 199; 116; 196; 130; 17). In a systematic review, survival of anthrax meningitis was predicted by treatment with a bactericidal agent and the use of multiple antimicrobials (111). There is limited information on long-term outcomes among the few survivors, but several cases were reported to have fully recovered (197; 196; 130; 17).

Clinical vignette

A 63-year-old man presented with a 4-day history of fever, myalgias, malaise, nausea, vomiting, and confusion (29; 108). He was treated with intravenous cefotaxime and vancomycin for presumed bacterial meningitis. Examination revealed a temperature of 39°C, no nuchal rigidity, absent Kernig and Brudzinski signs, and no cranial neuropathies or focal neurologic findings. The white blood cell count was 9400 with 77% neutrophils. Chest x-ray showed a widened mediastinum and basilar infiltrates. A computed tomography scan of the head without intravenous contrast was normal. A lumbar puncture showed cloudy CSF. Gram stain of CSF showed many polymorphonuclear leukocytes and numerous large gram-positive rods, singly and in chains. CSF laboratory studies disclosed relative hypoglycorrhachia (with CSF glucose 57 mg/dl and blood glucose 174 mg/dl), increased protein (666 mg/dl), red cells (1375 per cubic ml), and a polymorphonuclear pleocytosis (4750 white cells with 81% polymorphonuclear). A diagnosis of anthrax meningoencephalitis was considered based on the CSF results, and high-dose intravenous penicillin therapy was added. Within 6 hours, CSF cultures grew colonies of gram-positive rods, which were presumptively identified as Bacillus anthracis within 18 hours and confirmed as such within 36 hours. His hospital course was complicated by coma, generalized convulsive seizures, hypotension, azotemia, hyperkalemia, and acidosis. Cultures of blood and CSF grew Bacillus anthracis. Autopsy was performed and was consistent with inhalational anthrax; the central nervous system was not included in the autopsy. Subsequent investigation identified the probable source of exposure as an anthrax-contaminated letter, sent as part of an outbreak of anthrax bioterrorism in the United States. This was the index case of that outbreak.

Biological basis

Etiology and pathogenesis

The Bacillus anthracis bacterium. Anthrax is caused by a large, encapsulated, gram positive, aerobic, nonmotile, spore-forming bacillus, Bacillus anthracis. The spore is the infectious form commonly found in the environment (216).

Cultured colonies of Bacillus anthracis typically exhibit a roughly textured surface and irregular edges and may also exhibit a classic "medusa-head" or "judge's wig" morphology, a term given to colonies whose edges resemble a tangled mass of curly hair.

Colonies also exhibit a characteristic "tenacity," such that a small spire of colonial matter remained standing on its own support without toppling.

Bacillus anthracis may have different morphologic characteristics exhibited by the same microorganism, when grown on different growth media, a phenomenon that is applied in the "encapsulation test.”

Bacillus anthracis tends to grow in long filamentous strands, which have a characteristic appearance on Gram-stained specimens.

Other stains may bring out additional features. For example, using the malachite green spore staining technique can highlight green-stained endospores amongst red-colored, Bacillus anthracis bacteria.

Some additional ultrastructural morphology of Bacillus anthracis bacteria can be visualized with light microscopy using special stains and growth media.

Greater ultrastructural detail is obtained using transmission electron microscopy.

Anthrax meningitis. Anthrax meningitis is histologically characterized by an influx of neutrophils and monocytic cells, hemorrhage, edema, vascular congestion, and the presence of B anthracis bacteria in the cerebrospinal fluid. There may be associated focal intracerebral hemorrhage, subarachnoid hemorrhage, intraventricular hemorrhage, and diffuse cerebral edema.

Inhalation anthrax. Following inhalation, anthrax spores are phagocytized by alveolar macrophages and transported to tracheobronchial lymph nodes, causing some to label the alveolar macrophages as “Trojan horse cells” (177; 57; 87; 14). The spores germinate beyond the lungs in the lymphatics. At the tracheobronchial lymph nodes, local toxin production causes a necrotizing edematous lymphadenitis, progressing to mediastinitis and pulmonary edema, possibly with a bloody pleural effusion (177; 83; 77; 57).

Phagocytic dendritic cells (CD11c+ cells) probably serve as conduits of B anthracis spores from the lung to the draining lymph nodes (183). The bacteria escape the lymphatic system and enter the blood stream (bacteremia). About half of the cases of inhalational anthrax develop a concurrent hemorrhagic meningoencephalitis with bloody, purulent spinal fluid (83; 02; 77; 57). Septicemia, toxic shock, and death rapidly follow in less than 48 hours (177; 57). Bacillus anthracis efficiently invades human brain microvascular endothelial cells, comprising the blood-brain barrier (204).

Clinically similar presentations, particularly fatal pneumonia, can occur with inhalation exposure to Bacillus cereus containing Bacillus anthracis toxin genes, particularly in immunocompromised hosts (97; 96; 155; 09). Bacillus cereus (a soil organism that is also an opportunistic pathogen) is closely related to Bacillus anthracis.

Gastrointestinal anthrax. Pathological examination of infected tissue reveals edema, necrosis, and bacilli in the lymphatics with lymphadenitis.

“The jail-break model” in cutaneous and gastrointestinal anthrax. An alternate model, the so-called “jail-break model,” has been proposed for initial growth and dissemination of anthrax in cutaneous and gastrointestinal anthrax, and possibly for inhalational anthrax. In this model, (1) spores germinate at the site of initial entry into capsulated bacteria and start producing toxins; (2) bacterial growth and toxins cause epithelial/endothelial lining disruption; (3) bacteria pass through the damaged barrier, enter the lymphatics, and then reach the blood stream without requiring phagocytic transport (84; 208; 14).

Bacterial toxins and virulence factors. Bacterial toxins have long been felt to play a major role in clinical anthrax, because many effects of Bacillus anthracis cannot be attributed to microscopically evident tissue changes (21; 71). By the time that clinical symptoms appear, deadly toxins have already been produced, and it may be too late for antibiotics to alter the outcome.

Full virulence of Bacillus anthracis requires an antiphagocytic (poly-γ-D-glutamic acid) capsule and the tripartite anthrax toxin, which is comprised of three proteins: (1) protective antigen, (2) edema factor, and (3) lethal factor (107; 23). These toxins are thought to be responsible for the primary clinical manifestations of hemorrhage, edema, necrosis, and death. Protective antigen serves as an essential carrier molecule for edema factor and lethal factor (in binary combinations) and enables penetration of the toxins into cells by a complex process involving actin-dependent endocytosis (01; 176; 218).

Protective antigen is activated proteolytically in the blood and tissues, and once activated is able to bind to anthrax toxin receptors on cell surfaces (23). Specifically, to reach the endosomes, the anthrax toxin hijacks the endocytic pathway of CMG2 and TEM8, the two anthrax toxin receptors (75; 192). The pore-forming subunit of the toxin then inserts itself into endosomal membranes, preferentially into intraluminal vesicles (rather than the limiting membrane of the endosome), leading to the translocation of the enzymatic subunits in the lumen of these vesicles (75; 192). Then the enzymatic subunits reach the cytoplasm either by direct translocation or by back-fusion of intraluminal vesicles of late endosomes. In the cytoplasm, lethal factor cleaves members of the mitogen-activated protein (MAP)/extracellular signal-regulated kinase (ERK) family (or MEK) to inhibit nuclear protein synthesis (213; 75). The enzymatic subunits can also be released from the cell in exosomes.

Edema factor and lethal factor actively suppress chemokine production and neutrophil recruitment during infection, which allows the anthrax bacillus to reproduce and spread unimpeded (204). Edema factor is a calmodulin-dependent adenylate cyclase that combines with protective antigen to form edema toxin; edema toxin acts to increase cyclic AMP and produce local edema and impair neutrophil function (76; 66; 193; 23). Lethal factor is a zinc metalloprotein that combines with protective antigen to form lethal toxin, a protein kinase; lethal toxin stimulates macrophages to release large quantities of pro-inflammatory cytokine molecules (ie, tumor necrosis factor alpha and interleukin-1 beta), resulting in shock (76; 66; 193; 217). Lethal toxin is critical to bacterial penetration of the blood brain barrier and development of meningoencephalitis because it disrupts brain microvascular endothelial monolayer integrity and endothelial tight junction protein zone occludens-1 (ZO-1) (73). Lethal toxin also attacks host immunity through multiple mechanisms, including attenuation of macrophage pro-inflammatory responses, suppression of dendritic cell function, and blocking actions of T- and B-lymphocytes, such as antigen-receptor dependent lymphocyte proliferation, cytokine production, and immunoglobulin production (13; 217). Among other actions, these toxins collectively depress cerebral cortical electrical activity (119; 205), depress central respiratory center activity (119; 173; 205), cause bleeding and destruction of the brain and vital organs in the chest (02), and induce cardiovascular collapse (119). Toxin-induced vascular dysfunction is mediated in part by disruption of the endothelial cell barrier (86).

Progression of anthrax after infection is largely mediated by two native virulence plasmids, pXO1 and pXO2 (72). The genes for lethal toxin and edema toxin are located on pXO1, and genes for the anti-phagocytic poly-Υ-D-glutamic acid capsule are located on pXO2. Attachment to (adhesion) and invasion of brain microvascular endothelial cells by blood-borne B anthracis is also mediated by the pXO1 plasmid-encoded envelope protein, BslA. BslA is necessary and sufficient to promote adherence to brain endothelium and also contributes to blood-brain barrier breakdown by disrupting the tight junction protein ZO-1.

Additional virulence factors are secreted as well, including two zinc metalloproteases: neutral protease 599 and immune inhibitor A. Immune inhibitor A contributes to blood-brain barrier disruption associated with anthrax meningitis, by additional proteolytic attack on ZO-1 (149). The increase in blood-brain barrier permeability begins with internalization of immune inhibitor A into human brain microvascular endothelial cells, followed by degradation of ZO-1. The expression of Immune inhibitor A contributes to hemorrhagic brain damage associated with fatal meningoencephalitis.

Epidemiology

Anthrax in animals. Anthrax is endemic in sheep, cattle, goats, and horses in multiple countries, particularly in Africa, Asia, and the Middle East. Spores can remain dormant in some types of soil for many decades, serving as a potential source of infection for grazing livestock (66; 144; 216; 221; 152; 194; 156; 214). Animals acquire spores through direct contact with contaminated soil (144).

Human anthrax. Humans may become infected by contact with, ingestion, or inhalation of spores from infected animals or their products (eg, wool or hides) (51; 221; 152; 194; 156; 214).

Cutaneous anthrax most commonly results when spores from contaminated animal products enter breaks in the skin, whereas "woolsorter disease" is an occupational inhalational anthrax infection acquired by wool mill workers (05; 27; 165; 15).

Naturally occurring cases outside of agricultural or industrial settings have resulted from exposure to products from anthrax-infected animals, eg, anthrax-contaminated bristle shaving brushes, animal hair or yarn, bone meal, and drums made from contaminated hides (53; 216). Two Turkish boys developed cutaneous anthrax after contaminated cow's blood was smeared on their foreheads as part of a traditional ritual (78). Laboratory-acquired cases have been reported and may be fatal (49; 50; 185). Human-to-human transmission of clinical anthrax has not been clearly documented (32), although historically in Gambia, antibodies to B anthracis were found in people living in close contact with patients, and B anthracis was isolated from communal loofahs (94).

Since 2000, cases of cutaneous and systemic anthrax have occurred in Scotland due to injection of heroin contaminated with anthrax (174; 22; 56; 168).

Anthrax is rarely found in most developed countries because of vaccines for animals and at-risk people. There has been a dramatic reduction in livestock and human cases of anthrax in the United States since 1900, in part due to livestock and at-risk human immunization, proper disposal of infected animals, and restriction of imported wool, hides, and other products (25; 26).

Since 1960, adult anthrax mortality has ranged from 31% for cutaneous to 90% for primary meningitis (91). Median incubation periods ranged from 1 day (interquartile range [IQR]: 0–4) for injection anthrax to 7 days (IQR: 4–9) for inhalation anthrax (91).

Anthrax meningoencephalitis. Among patients with anthrax meningoencephalitis, males outnumber females by 3 to 1 (130). This represents, in part, differences in occupational exposure because many affected individuals have been wool mill workers or farmers (05; 165; 166; 80; 117).

Anthrax bioweapon or bioterrorism. Because of the complexities of development and dispersal of an anthrax bioweapon, it had been anticipated that small-scale sabotage or attempts at personal murder were more likely than large-scale attempts at inflicting mass casualties with anthrax, and that crude dispersal in a close area was the most likely mode of attack. In the 2001 U.S. outbreak of anthrax acquired through intentional contamination of mail, most cases involved media, government, and postal service employees; several civilian cases remain unexplained (43).

Epidemiologic clues to bioweapon use include (1) an outbreak or multiple, simultaneous outbreaks with large numbers of patients complaining of respiratory symptoms; (2) high case fatality; (3) sick or dead animals of different species; (4) multidrug-resistant pathogens (57). A World Health Organization report estimated that release of 50 kg of anthrax spores upwind of a city of 500,000 population would result in 125,000 infections and nearly 95,000 deaths (215). A report by the United States Congressional Office of Technology Assessment estimated between 130,000 and 3 million deaths following aerosolized release of 100 kg of anthrax spores upwind of Washington, D.C. (154).

Prevention

Vaccination of livestock in endemic areas is the most effective method of preventing naturally acquired anthrax in animals and humans (33; 194; 214). Management of anthrax in livestock includes quarantine of the herd, removal of the herd from contaminated pastures, vaccination of livestock, treatment of affected livestock, and incineration and, later, burial of infected carcasses, bedding, and other material found around the carcasses (33). Other factors that contributed to the control of naturally occurring human anthrax in the United States include improved industrial hygiene, strict importation guidelines for animals, and regulations that minimized the importation and use of contaminated animal products.

In anthrax endemic areas, livestock are seldom presented for vaccination, and unsafe handling of diseased animals and carcasses can greatly augment the frequency and lethality of human cases. Knowledge remains inadequate to prevent unnecessary livestock and human cases in many endemic areas, including African countries and India (147; 161). In a cross-section study of community knowledge, attitudes, and practices toward anthrax in Kenya, 23% reported that they would slaughter and sell potentially contaminated beef to neighbors, whereas 63% would bury or burn the carcass (147); in addition, although 94% believed vaccination prevents anthrax, only 55% would present livestock for vaccination.

Prevention of anthrax in civilians as a result of bioterrorist activity is limited to prevention of exposure and postexposure prophylaxis (postexposure prophylaxis is considered under management). In military personnel and selected at-risk populations, prevention has also included vaccination (48); however, legal challenges resulted in a temporary halt of military vaccinations as of December 2003. Although previous Centers for Disease Control guidance (32) indicated that anthrax vaccine could be requested through the Centers for Disease Control, such a vaccine was not provided to civilians during the 2001 outbreak in the United States (43; 130). Subsequently, civilians deemed at high risk of exposure who had completed a 60-day course of antibiotics were offered postexposure anthrax vaccination (34; 105).

Anthrax Vaccine Adsorbed (AVA), a cell-free, noninfectious anthrax vaccine prepared from a noncapsulating, nonproteolytic anthrax strain, was licensed by the FDA in 1970 for soldiers, at-risk scientists and laboratory workers (ie, those working with anthrax), at-risk veterinarians and livestock handlers, and at-risk wool mill workers (203). AVA was reapproved for licensure by the Food and Drug Administration in 1985 and remains the only licensed human anthrax vaccine in the United States (216). AVA, now marketed as BioThrax (Emergent BioSolutions, Lansing, Michigan), is licensed for use in non-pregnant adults aged 18 to 65 years who are at high risk of exposure.

AVA is a considerable improvement over previously available vaccines, including the uncharacterized whole-cell vaccines developed initially in 1881 (180). Originally it was administered as a series of six subcutaneous exposures over 18 months, followed by annual boosters for persons with continuing potential exposure risk (57; 203). In 2008, the Centers for Disease Control and Prevention published the results of a 4-arm randomized controlled trial comparing the original regimen administered by subcutaneous administration with the original regimen administered by intramuscular injection, a reduced dose schedule administered by intramuscular injection, and a placebo that received saline injections at the same intervals; intramuscular injection significantly reduced the occurrence of injection-side reactions, and immunological priming of reduced-dose regimens proved noninferior to the original licensed regimen (137). For pre-event or preexposure prophylaxis, the United States Advisory Committee on Immunization Practices now recommends five 0.5-ml doses of AVA administered intramuscularly at 0 and 4 weeks and 6, 12, and 18 months, followed by annual 0.5-ml booster injections after the 18-month dose (216). A booster dose of AVA for preexposure prophylaxis can be given every 3 years instead of annually to persons not at high risk for exposure to Bacillus anthracis who have previously received the initial AVA 3-dose priming and 2-dose booster series and want to maintain protection (24). People with contraindications to intramuscular injections (eg, coagulopathies) may continue to receive AVA by subcutaneous administration (216). In contrast to the revised vaccine administration schedule and route for pre-event or preexposure prophylaxis, the FDA decided in June 2009 to continue the preexisting postexposure prophylaxis protocol until new data become available; therefore, for postexposure prophylaxis, AVA should only be administered under an Investigative New Drug protocol or Emergency Use Authorization in a 3-dose series (at 0, 2, and 4 weeks) by the subcutaneous route in conjunction with antimicrobial therapy for a minimum of 60 days (216).

The Department of Defense Anthrax Vaccine Immunization Program, announced in 1997 and instituted in 1998, administered 8.4 million doses to approximately 2.1 million military personnel from March 1998 through December 2008 (216). Although previously mandatory for military servicemen and women, a United States District Court ruled on October 27, 2004, that the Department of Defense can no longer inoculate troops without their consent. The vaccine is associated with frequent local reactions but only rare serious side effects sufficient to require hospitalization (approximately once per 200,000 doses) (203). In fact, in a study employing a historical cohort before the inception of the Department of Defense Anthrax Vaccine Immunization Program in 1998, hospitalization following anthrax vaccination was significantly less likely from any cause, diseases of the blood and blood-forming organs, and diseases of the respiratory system among 170,723 active duty U.S. service members who were vaccinated with the anthrax vaccine (210).

The Vaccine Adverse Event Reporting System has monitored adverse events reported following vaccination with AVA. The most common adverse events that occurred after AVA administration (either alone or concurrently with other vaccines) were arthralgia (17%), headache (16%), pruritus (15%), pain (14%), injection-site erythema (13%), fever (11%), erythema (10%), pain at the injection site (10%), rash (10%), and myalgia (10%) (216). There is no convincing evidence of long-term adverse health effects associated with AVA administration (216). Vaccine postmarketing surveillance has identified no association between optic neuritis and receipt of the anthrax vaccine by members of the U.S. military (162).

AVA is neither recommended nor licensed for use during pregnancy, and the Department of Defense continues a policy of avoiding anthrax vaccination in pregnant women (48). AVA is now available under an investigational new drug application in combination with extended antibiotic treatment for civilians exposed to inhalational anthrax who have completed a 60-day course of antibiotics (34; 105).

Other vaccines are used in people and animals in other countries with diverse adaptive immune responses to different anthrax antigens (131). Britain and China have vaccines derived from toxins filtered from a bacterial culture (184; 104). The United Kingdom uses the Anthrax Vaccine Precipitated (AVP) vaccine, an alum-precipitated cell-free filtrate of culture supernatant from a nonencapsulated strain of B anthracis (from which bacterial cells and some LF and EF are removed by downstream filtering) (131; 145). Russia's vaccine involves inoculation of live spores of a weakened strain of anthrax originally developed for livestock vaccines in the 1930s (184; 104); this vaccine, referred to as a live anthrax vaccine (LAV) or live attenuated anthrax vaccine (LAAV), uses live attenuated STI strain B anthracis spores that lack an essential virulence factor (ie, the plasmid that encodes the capsule), and consequently the vaccine is “attenuated” (131). At least a dozen countries manufacture anthrax vaccines for livestock, which are similar to the Russian vaccine for people.

New vaccines are in development, including at least one based on highly purified, genetically engineered protective factor administered with a modern adjuvant (179; 139). The AV7909 anthrax vaccine being developed for postexposure prophylaxis may require fewer vaccinations and a reduced amount of antigen to achieve an accelerated immune response compared with the AVA (BioThrax®) vaccine (100).

Much work needs to be done internationally to develop capabilities to manage victims of biological or chemical terrorism (114). At present, most physicians and hospital emergency departments in the United States are ill prepared to treat victims of such terrorism (112; 211; 190).

Differential diagnosis

Confusing conditions

A high index of suspicion is essential for prompt diagnosis of inhalational anthrax, especially in identified or likely at-risk groups (ie, employees of public figures, media employees, mail handlers, and those with known or likely exposure). Inhalational anthrax does not typically cause nasal congestion, rhinorrhea, or sore throat like the common cold or most other influenza-like respiratory viruses (35; 103), although it can (165; 35). Clinical features suggestive of inhalational anthrax, as opposed to viral respiratory tract infection, include the presence of nonheadache neurologic symptoms, dyspnea, nausea or vomiting, and any abnormality on lung auscultation (103). Early cases of inhalational anthrax may resemble viral respiratory illness or, in some cases, gastroenteritis. In the only 20th-century outbreak of inhalational anthrax in the United States, which occurred in a textile mill in 1957, clinical diagnoses among the four (of five) listed cases included (1) influenza, (2) a severe cold or (3) possible bronchitis, and (4) possible cholecystitis. In the Sverdlovsk outbreak in 1979, cases of inhalational anthrax were most commonly felt to be cases of pneumonia; in some cases, though, severe chest pain resulting from hemorrhagic thoracic lymphadenitis and mediastinitis resulted in a clinical diagnosis of myocardial infarction. Other possible considerations in the differential diagnosis of inhalational anthrax include viral pneumonia, mycoplasma pneumonia, Legionnaire disease, histoplasmosis, coccidioidomycosis, psittacosis, tularemia, and Q fever. Both tularemia and anthrax may produce similar mediastinal lymphadenopathy (193).

Cutaneous anthrax has been confused with a brown recluse spider bite; it may also be confused with staphylococcal or streptococcal furuncles, ecthyma gangrenosum, cutaneous tuberculosis, ulceroglandular tularemia, plague, and other skin infections (66; 175; 193).

Gastrointestinal anthrax must be differentiated from other causes of acute abdomen, including peritonitis, acute gastroenteritis, mechanical obstruction, peptic or duodenal ulcer, typhoid fever, and tularemia (66).

The differential diagnosis of anthrax meningoencephalitis includes subarachnoid and intracerebral hemorrhage, nonanthrax bacterial or aseptic meningitis, herpes simplex virus encephalitis, other forms of infectious meningoencephalitis in immunocompromised individuals, and cerebral malaria (05; 165; 69; 71; 197; 26; 66; 29; 117; 130). Listeria monocytogenes is the only common gram-positive rod causing meningitis; it does not produce hemorrhagic meningitis and is rarely seen on gram stain.

Diagnostic workup

Treatment should not be delayed pending a diagnostic workup for possible anthrax. Detailed interim guidelines for investigation of, and response to, Bacillus anthracis exposures have been published (35; 37).

No rapid screening test is available to diagnose inhalational anthrax in the early stages (35). Rapid tests to detect conditions with similar clinical presentations (eg, group A beta-hemolytic streptococcal infections, influenza infections, and respiratory syncytial virus infections in infants) have been encouraged (40; 35). White blood cell counts may be elevated with anthrax (40; 105) but are generally normal (or possibly low) with influenza and other respiratory viral infections. Gram stain and culture of blood, CSF, or pleural fluid can establish the diagnosis of anthrax, but these are late clinical findings. Sputum gram stain is not often positive. Enzyme-linked immunosorbent assay, immunohistology, serum serology (with sequential blood samples at least a week apart), and polymerase chain reaction demonstration of Bacillus anthracis DNA can be used to confirm the diagnosis but are of little help clinically, as these typically must be performed at a reference laboratory, the results are generally not available in time to affect the clinical outcome (126). Nasal swabs can help establish exposure, but anthrax spores are rapidly cleared, which means that an exposed person may have a negative swab; nasal swabs are not indicated to diagnose anthrax, to determine risk of exposure and need for antimicrobial prophylaxis, or to determine when antimicrobial prophylaxis may be stopped (35; 37). With inhalational anthrax, a chest x-ray may demonstrate hilar or mediastinal adenopathy, a widened mediastinum, pleural effusions, or infiltrates (57; 40; 35).

Anthrax should be diagnosed if a chest x-ray demonstrates a widened mediastinum in the setting of fever and other constitutional symptoms, in the absence of another obvious cause (eg, severe blunt chest trauma or postsurgical infection).

Diagnostic recommendations for anthrax meningitis vary by setting (23). In a conventional (noncrisis) setting, patients with systemic anthrax should undergo a lumbar puncture to determine whether they have anthrax meningitis, provided that no contraindications exist. In a contingency setting (ie, when resource availability or other circumstances shift evaluation and management to nonpreferred methodologies, medications, and locations), patients with systemic anthrax who have two or more clinical features of meningitis (ie, severe headaches, altered mental status including confusion, meningeal signs, or other neurologic symptoms/signs) or a contraindication to a lumbar puncture should be presumed to have anthrax meningitis, whereas patients with fewer of the screening symptoms and signs should have lumbar punctures. Identification of Gram-positive rods, pleocytosis, visible turbidity, or visible hemorrhage in cerebrospinal fluid is sufficient for a diagnosis of probable anthrax meningitis. In a crisis setting (ie, when resource limitations require medical care prioritization, allocating care to those with the highest likelihood of survival or benefit) a clinical case definition (ie, without lumbar puncture) may be used to identify patients with probable anthrax meningitis: under these circumstances, patients with systemic anthrax and severe headache, altered mental status, meningeal signs, or other neurologic signs should be considered to have anthrax meningitis.

An evidence-based assessment tool has been developed for screening patients for meningitis during an anthrax mass casualty incident (111). Severe headache, altered mental status, meningeal signs, and other neurologic signs at presentation independently predicted meningitis in a derivation cohort. The presence of one of these signs and symptoms on admission had a sensitivity of 89% for finding anthrax meningitis in the adult validation cohort and 83% in the pediatric validation cohort. Anthrax meningitis was unlikely in the absence of any of these signs or symptoms, whereas anthrax meningitis was very likely in the presence of two or more of these signs or symptoms.

Quick and reliable assessment of patients for anthrax meningitis using clinical symptoms and signs is possible (18). If low-resource diagnostic testing is feasible for some patients, pairing algorithms improves specificity. Pairing algorithms with differing symptoms and signs minimizes the proportion of patients requiring additional diagnostics.

Hemorrhagic meningitis should raise strong suspicion of anthrax infection (104; 110; 130; 198; 142; 81). In the index case in the 2001 bioterrorist outbreak in the United States, the initial diagnosis was made from CSF (29). Findings are fairly consistent across reported cases (69; 197; 169; 166; 196; 68; 80; 29; 130). CSF in anthrax may appear cloudy, with yellowish or pinkish coloration, or may be grossly bloody; evolution across this spectrum has been reported in serial lumbar punctures in an individual patient. The supernatant is frequently xanthochromic. Opening pressure may be normal or increased (300 to 500 mm of CSF). Hypoglycorrhachia is common with CSF glucose frequently from 20 to 40 mg/dl. CSF protein is generally elevated, with initial CSF protein values from 120 to 1150 mg/dl (median 400 mg/dl) and with maximum reported values of 1333 mg/dl. CSF is frequently hemorrhagic, but in the early stages there may be no red cells; the range of red cell concentration on initial lumbar puncture is from 0 cells per cubic mm to 1375 cells per cubic mm (median 400), with maximum reported values as high as 1,800,000 cells per cubic mm. There is a polymorphonuclear CSF pleocytosis with a range of 560 to 4750 white cells per cubic mm on initial lumbar puncture (median 2500), with maximum reported values as high as 58,000 white cells per cubic mm; the percentage of polymorphonuclear cells ranges from 67% to 95% (median 81%). CSF cultures are usually positive, but CSF may be sterile after several days of antibiotics. Gram stain generally shows numerous polymorphonuclear leukocytes as well as many large gram-positive rods, either singly or in short or long chains; indeed, virtually all case reports of anthrax meningitis that give gram stain results report these findings. Serial lumbar punctures may document an evolution from cloudy to grossly bloody CSF, with increasing protein concentration, numbers of red cells and polymorphonuclear leukocytes, and proportion of polymorphonuclear leukocytes among white cells.