Anthrax: Current, comprehensive information on pathogenesis, microbiology, epidemiology, diagnosis, treatment, and prophylaxis

Last updated May 30, 2008

Agent
Pathogenesis
Epidemiology
—Naturally Occurring Anthrax
—Weaponized Anthrax
Clinical Features
—Cutaneous Anthrax
—Inhalational Anthrax
—Gastrointestinal and Oropharyngeal Anthrax
—Anthrax Meningitis
Differential Diagnosis
—Cutaneous Anthrax
—Inhalational Anthrax
—Gastrointestinal and Oropharyngeal Anthrax
—Anthrax Meningitis
Staging of Inhalational Anthrax
Distinguishing Inhalational Anthrax from Influenza-Like Illness and Community-Acquired Pneumonia (CAP)
Pediatric Considerations
Anthrax During Pregnancy
Laboratory Diagnosis
—Specimen Collection and Transport
—Laboratory Biosafety and Biosecurity Information
—Laboratory Response Network
—Standard Tests for Detection of B anthracis
—Stepwise Identification and Confirmation
—Other Tests for B anthracis Detection
—Antimicrobial Susceptibility Studies
—Tests for Exposure
Anthrax Countermeasures
— Treatment of Inhalational, Gastrointestinal, and Oropharyngeal Anthrax
—Treatment of Cutaneous Anthrax
—Treatment of Anthrax Meningitis
—PEP
—Project BioShield
—New Therapeutic Approaches
—Anthrax Vaccine
Infection Control and Worker Protection
Issues Related to Autopsies and Burial
Management of Exposure Events
—Potential Types of Exposures
—Handling of Suspicious Packages
—Mass Exposure Events
—Surveillance During Exposure Events or for Early Detection of Outbreaks
Environmental Testing
Use of Autonomous Detection Systems in the Workplace
Decontamination
Public Health Reporting and Case Definition
References

Agent

Key microbiologic characteristics of Bacillus anthracis follow (see References: CDC: Basic laboratory protocol for identification of Bacillus anthracis):

Vegetative cell: large, gram-positive bacillus (1.0 to 1.5 mcm by 3.0 to 5.0 mcm), "jointed bamboo-rod" appearance
Endospore: oval, central-to-subterminal, does not usually swell (1.0 x 1.5 mcm); CO₂ levels within the body inhibit sporulation
Forms long chains of vegetative cells in vitro; single cells or short chains of 2 to 4 cells in direct clinical samples
Aerobic or facultatively anaerobic
Nonmotile
Catalase-positive
Nonhemolytic on sheep blood agar
Susceptible to lysis by gamma phage
Rapid growth on sheep blood agar (SBA); comma-shaped projections may give "Medusa head" appearance on SBA
Colonies have been described as having a "ground glass" or curled hair" appearance and have the consistency of beaten egg whites
Colonies are 2 to 5 mm in diameter after 16 to 18 hours of incubation
Forms mucoid capsule when grown on agar with sodium bicarbonate and incubated in CO₂-enriched atmosphere; capsule can be visualized with India ink preparation

Spores germinate and form vegetative cells in environments rich in nutrients (eg, glucose, amino acids, nucleosides). Vegetative bacteria generally survive poorly outside of mammalian hosts. Conversely, vegetative cells form spores when nutrients in the environment are exhausted. Spores are protected by a morphologically complex protein coat (see References: Giorno 2007). Spores have been shown to have heat-resistance characteristics similar to other Bacillus species, and survive in the environment for more than 40 years (see References: Manchee 1990, Montville 2005). Microarray analyses of transposon-mediated mutations in B anthracis may allow identification of additional genes that play a role in spore germination and organism growth. The method identified two conserved hypothetical genes in the genomes analyzed (see References: Day 2007).

Pathogenesis

Virulence Factors

The primary virulence factors produced by B anthracis are plasmid-coded and include:

A poly-D-glutamic acid capsule (coded for on plasmid pX02) that inhibits phagocytosis of vegetative bacilli
Three exotoxins that combine to produce two binary toxins:

Protective antigen (PA) is a binding protein, which permits entry of toxin into host cells via endocyte formation. Once in the endocyte, PA toxin forms a pore, which creates a small passageway in the endosomal membrane that allows the enzymatic components of the toxin to enter the cytoplasm.

Analysis of host anthrax toxin receptor-2 determinants demonstrated that selected receptor domains influence binding of PA and subsequently, protein rearrangements that accompany anthrax toxin pore formation (see References: Scobie 2007).
The three-dimensional structure of a PA–host receptor complex has recently been determined. The receptor-pathogen interaction surface closely resembles that of the normal receptor–extracellular matrix interaction (see References: Santelli 2004).
Anthrax toxin is capable of causing hemolysis in the presence of polymorphonuclear cells (PMNs); this effect appears primarily to be mediated by PA (see References: Wu 2003).

Edema factor (EF) is a calmodulin-dependent adenylate cyclase. EF combines with PA to form edema toxin (coded by the cya gene of plasmid pX01).

Edema toxin converts adenosine triphosphate to cyclic adenosine monophosphate (cAMP); high intracellular levels of cAMP lead to impaired maintenance of water homeostasis and characteristic edema. The cAMP response generated by edema toxin requires an influx of calcium into the affected cells (see References: Kumar 2002).
Edema toxin also inhibits neutrophil function. In a rabbit model, edema toxin acting indirectly or directly on an uncharacterized target stimulates production or release of multiple inflammatory mediators, including neurokinins, prostanoids, and histamine (see References: Tessier 2007).

Lethal factor (LF) is a zinc metalloprotease. LF combines with PA to form lethal toxin (coded by the lef gene on pX01). Lethal toxin is thought to stimulate overproduction of cytokines (eg, tumor necrosis factor [TNF] alpha and interleukin [IL] 1-beta), which leads to lysis of macrophages. Rapid release of inflammatory mediators also may contribute to the sudden death that can occur with anthrax.

Experiments in mice show that lethal toxin suppresses cytokine secretion during infection (see References: Drysdale 2007).
Lethal toxin also apparently impairs the function of dendritic cells by disrupting the mitogen-activated protein (MAP) kinase intracellular signaling network, which may suppress host immunity (see References: Agrawal 2003).
Lethal toxin has been shown to cause endothelial cell apoptosis and endothelial barrier dysfunction, which may contribute to vascular destruction (see References: Kirby 2004, Warfel 2005).

A recent study using a mouse model demonstrated that both EF and LF can inhibit activation of T lymphocytes by T-cell receptor-mediated stimulation (see References: Comer 2005).
A gene encoding the RNA polymerase sigma factor, sigma H, is required for toxin gene expression. B anthracis mutants lacking the gene were asporogenous and toxin deficient (see References: Hadjifrangiskou 2007).
Recombinant preparations of edema toxins and lethal toxins, either singly or in combination, have been used to determine the relative contributions of these proteins to lethality. In a rat model, edema toxins were shown to be 10 times less toxic than lethal toxins but produced greater hypotension and contributed to lethal toxin harmful effects (see References: Cui 2007).

Virulence of B anthracis appears to be related to clonality and to the numbers of copies of the pX01 and PX02 plasmids within each bacterial cell (see References: Coker 2003). The complete genomic sequence of B anthracis (Ames strain) has been analyzed. Several chromosomally encoded potential virulence factors were identified, including hemolysin, phospholipases, and iron acquisition proteins (see References: Read 2003).

A newly described toxin, anthrolysin O (ALO), a cholesterol-dependent cytolysin, may represent a previously unidentified virulence factor of B anthracis. When human neutrophils, monocytes, and macrophages are treated with native or recombinant ALO, the protein provoked dose- and time-dependent cytotoxicity (see References: Mosser 2006). Experiments that paired ALO with three phospholipase C proteins (PLCs) in a murine mouse model revealed that ALO and PLCs may have overlapping roles in pathogenesis (see References: Heffernan 2007).

Anthrax scavanges iron via two siderophores, bacillibactin (BB) and petrobactin (PB). These proteins help the pathogen evade the mammalian immune system and contribute to virulence. PB may be the only siderphore required to ensure full virulence (see References: Abergel 2006). Organisms with mutated genes had limited growth in cultures, and addition of exogenous PB restored growth (see References: Lee 2007).

Anthrax toxin genes were identified in a naturally occurring Bacillus cereus isolate obtained from a patient with an illness similar to inhalational anthrax (see References: Hoffmaster 2004). More recently, anthrax-like illnesses were identified in two metalworkers who had inhalation exposure to B cereus that contained B anthracis toxin genes (see References: Avashia 2007).

Inhalational Anthrax

Pathogenesis of inhalational anthrax involves the following steps (see References: Hanna 1998, Abramova 1993):

Endospores are introduced into the body via inhalation. Endospores are 1 mcm by 1.5 mcm in size (see References: CDC: Basic laboratory protocol for identification of Bacillus anthracis) and are, therefore, able to reach the alveoli (ie, <5 mcm).
Endospores are phagocytosed by macrophages and carried to regional lymph nodes.
Alveolar epithelial cells and macrophages actively participate in the innate immune response to infection via cell signal-mediated production of cytokines and chemokines (see References: Chakrabarty 2007).
The endospores then germinate inside macrophages and become vegetative cells, which leave the macrophages and multiply in the lymphatic system.
Bacteria enter the bloodstream and lead to septic shock and toxemia; hematogenous spread can lead to hemorrhagic meningitis.
Regional hemorrhagic lymphadenitis of mediastinal and peribronchial lymph nodes causes the occurrence of hemorrhagic mediastinitis (see References: Abramova 1993). A widened mediastinum may be noted on chest radiograph or enlarged lymph nodes may be directly visualized on chest computed tomography.
Pulmonary lymphatic drainage can be blocked, leading to pulmonary edema.
Pleural effusions are common and may be massive. B anthracis bacilli, bacillary fragments, and antigens can be noted with immunohistochemistry (IHC) testing of pleural effusions (see References: Guarner 2003).
True pneumonia rarely occurs, although a focal, hemorrhagic, necrotizing pneumonic lesion (similar to the Gohn complex of tuberculosis) may be seen (see References: Abramova 1993). Intraalveolar edema, focal areas of hyaline membrane formation, and interstitial mononuclear inflammation may be noted (see References: Guarner 2003).
Compression of the lungs and septic shock are the major causes of death.
The ID₅₀ for inhalational anthrax is estimated at 8,000 to 50,000 spores (see References: Franz 1997), although the minimum infective dose may be considerably less. On the basis of experimental studies involving primates, the US Department of Defense has estimated that the LD₅₀ for inhalational anthrax in humans from weapons-grade anthrax is 2,500 to 55,000 spores (see References: Defense Intelligence Agency).
Extrapolation of dose-response curves involving cynomolgous monkeys suggest that the LD₁₀ in humans following exposure to airborne anthrax spores may be as low as 50 to 98 spores, the LD₅ may be only 14 to 28 spores, and the LD₁ may be only one to three spores (see References: Peters 2002).
Mathematical modeling of airborne anthrax infection (based on observations from the US Postal Service experience during the 2001 anthrax outbreak and an assumption that 10,000 people may have been exposed) suggests that exposures ranged from 18 to 863 spores and may have been as low as 2 to 9 spores (see References: Fennelly 2004).

Cutaneous anthrax

The pathogenesis of cutaneous anthrax involves the following process:

Endospores are introduced through the skin (usually via preexisting skin lesions or abrasions).
Low-level germination at the site of introduction produces localized necrosis with eschar formation and soft-tissue or mucosal edema (which can be massive in some cases). Epithelial damage appears to be required for germination of spores. Germination begins 1 to 3 hours after inoculation, but spore germination by itself is not sufficient to produce infection in undamaged skin (see References: Bischof 2007).
Endospores often are phagocytosed by macrophages and carried to regional lymph nodes, causing painful lymphadenopathy and lymphangitis.
Hematogenous spread with resultant toxemia can occur, although such spread is not common with appropriate antibiotic therapy.
The infective dose for cutaneous anthrax is not known.

Gastrointestinal anthrax

The pathogenesis of gastrointestinal anthrax is not clear, since this condition is relatively rare. Historically, illness has been thought to result from ingestion of B anthracis spores; however, recently experts have postulated that illness predominantly results from ingestion of large numbers of vegetative cells (such as may be found in poorly cooked meat from infected animals) (see References: Inglesby 2002).

Two forms of gastrointestinal anthrax have been recognized: oropharyngeal and gastrointestinal.
In oropharyngeal anthrax, the portal of entry is the oral or pharyngeal mucosa. A mucosal ulcer occurs initially, followed by regional lymphadenopathy and localized edema.
In gastrointestinal anthrax, the portal of entry often is the terminal ileum or cecum. Intestinal lesions occur and are followed by regional lymphadenopathy. Edema of the bowel wall and ascites (sometimes massive) may be present.
Hematogenous spread with resultant toxemia can occur.
The infective dose for gastrointestinal anthrax is not known. In animal models (guinea pigs, rabbits, and rhesus monkeys), investigators failed to induce infection following oral challenge with 10⁸ spores; these animals are all thought to be more susceptible to anthrax than humans (see References: Beatty 2003).

Epidemiology: Naturally Occurring Anthrax

Reservoir

B anthracis is found in soil in many areas of the world. Ecologic factors (such as abundant rainfall following a period of drought) may enhance spore density in soil, although the exact influence of such factors remains poorly understood.

The organism generally exists in the endospore form in nature; germination of spores outside an animal host may occur when the following conditions are met (see References: Turnbull 1998):

Temperature between 8°C and 45°C
pH between 5 and 9
Relative humidity >95%
Presence of adequate nutrients

Endospores are resistant to drying, heat, ultraviolet light, gamma radiation, and some disinfectants. Spores can persist in soil for decades, as illustrated by biological warfare experiments during World War II on the Scottish island of Gruinard (see References: Manchee 1990). During 1943 and 1944, an estimated 4 x 10¹⁴anthrax spores were dispersed on the island through explosive means. Spores were still detectable more than 40 years later. Disinfection of the island was finally completed in 1987, using a combination of seawater and formaldehyde.

Anthrax in Animals

Anthrax is predominantly a disease of animals. Livestock or other herbivores (eg, cattle, sheep, goats, pigs, bison, water buffalo) acquire infection from consuming contaminated soil or feed. Anthrax in animals is hyperendemic or endemic in the following areas of the world (see References: WHOCC):

Most areas of the Middle East
Most areas of equatorial Africa
Mexico and Central America
Chile, Argentina, Peru, and Bolivia
Certain Southeast Asian countries (eg, Myanmar, Vietnam, Cambodia, Thailand)
Papua New Guinea
China
Some Mediterranean countries

In most of the rest of the world, anthrax occurs only sporadically. In the United States, outbreaks in animals have occurred since 1990 in the Midwest (Kansas, Minnesota, Nebraska, North Dakota, South Dakota, Missouri); in the West (California, Nevada); and in Texas and Oklahoma (see References: WHOCC; MBAH 2006). Outbreaks have also recently occurred in Saskatchewan and Manitoba, Canada, affecting more than 930 animals (see References: CFIA 2006).

Recently, anthrax has been reported as the cause of death among chimpanzees in Ivory Coast (see References: Leendertz 2004) and chimpanzees and a gorilla in Cameroon (see References: Leendertz 2006). Investigators postulated that the chimps became ill either through consumption of an infected animal or through drinking contaminated water. Isolates from the wild apes in both outbreaks showed that the strains were clearly different from those of any previously described. Molecular analyses of eight variable number of tandem repeat (VNTR) segments in DNA revealed that they were distinct from those in group A and group B strains. The isolates established a new "forest anthrax cluster," termed F, suggesting that B anthracis is a far less homogeneous species than currently believed (see References: Leendertz 2006).

Anthrax spores were detected in 2 of 6 species of raptors (road-side hawks and chimango caracaras [a bird of prey that is in the falcon family]) in central Argentina, suggesting that scavenger and nonscavenger bird species may influence anthrax epidemiology in some countries (see References: Saggese 2007).

Modes of Transmission

Illness in humans most commonly occurs following exposure to infected animals or contaminated animal products; such exposures include:

Contact with infected tissues of dead animals (eg, butchering, preparing contaminated meat), which generally leads to cutaneous anthrax
Consumption of contaminated undercooked meat, which can lead to gastrointestinal or oropharyngeal anthrax
Contact with contaminated hair, wool, or hides (particularly during processing) or contact with products made from them, which can lead to either inhalational or cutaneous anthrax

Cases following laboratory exposure have been recognized (see References: Brachman 1980; CDC 2002: Suspected cutaneous anthrax in a laboratory worker). Person-to-person transmission of B anthracis has been reported rarely with cutaneous anthrax, but has not been recognized with gastrointestinal or inhalational disease (see References: Weber 2001, Weber 2002).

Anthrax in Humans—United States

Approximately 130 cases occurred annually in the United States in the early 1900s. The incidence has gradually declined over time, with less than 10 cases reported each year since the early 1960s (see References: CDC: Summary of notifiable diseases).
About 95% of naturally occurring cases in the United States are cutaneous and 5% are inhalational. Gastrointestinal infection has not been recognized in this country (see References: Brachman 1980).
Only 18 cases of naturally occurring inhalational anthrax were reported in the US during the 20th century (see References: Brachman 1980). All but three were associated with industrial exposures; two of the remaining cases were laboratory-acquired and the source of exposure for the third case remains unknown.
Between 1990 and 2000, only two cases of naturally occurring anthrax were reported in the United States (one in 1992 and one in 2000); both patients had cutaneous disease. The latter case occurred in North Dakota and resulted from agricultural exposure (see References: CDC: Human anthrax associated with an epizootic among livestock—North Dakota; CDC: Summary of notifiable diseases).
In February 2006, a case of inhalational anthrax was recognized in a New York City resident (see References: CDC: Inhalational anthrax associated with dried animal hides—Pennsylvania and New York City, 2006; Walsh 2007). This was the first new case of naturally occurring inhalational anthrax in the United States since 1976. The patient made traditional African drums using hard-dried animal hides. In December 2005, he traveled to Africa and brought back four goat hides. He worked with the hides between December 2005 and February 2006. Environmental sampling of his workspace, van, and home yielded B anthracis spores. Four potential exposed persons received prophylaxis; no additional cases were identified through surveillance efforts. In response to this case, the Centers for Disease Control and Prevention (CDC) posted a document in February 2006 on safety issues related to anthrax and animal hides (see References: CDC: Anthrax Q & A: Anthrax and Animal Hides).

Anthrax in Humans—Global Perspective

An estimated 2,000 to 20,000 human cases of anthrax occur globally each year (see References: Brachman 1984).
Most cases are cutaneous, with inhalational and gastrointestinal cases occurring less frequently.
Human cases generally follow disease occurrence in ruminants and are most prevalent in Africa, the Middle East, and parts of Southeast Asia.
A recent systematic review of MEDLINE (1996-2005) and selected journal indexes (1900-1966) for inhalational anthrax cases worldwide between 1900 and 2005 identified 82 cases of inhalational anthrax (see References: Holty 2006: Systematic review: a century of inhalational anthrax cases from 1900 to 2005).

Seventy-one of the cases were naturally occurring (11 were part of the 2001 bioterrorism outbreak in the United States, and cases from the 1979 Sverdlovsk outbreak in Russia [see section below on the epidemiology of weaponized anthrax] were excluded from analysis because symptoms, treatment, and disease progression were not described).
Among naturally occurring cases, most involved exposure to contaminated wool, goat hair, or animal hides.

A similar literature review of 42 patients who had atypical anthrax presentations (ie, patients with confirmed anthrax infection who did not have known cutaneous, gastrointestinal, or inhalational ports of entry) revealed that these patients were significantly less likely to have cough, chest pain, or abnormal lung findings, even though they most likely had inhalational anthrax (see References: Holty 2006: Anthrax: a systematic review of atypical presentations).

Outbreaks of Naturally Occurring Disease

Outbreaks have been reported in industrial settings where animal products are processed and in agricultural settings following consumption of or exposure to ill animals. Notable examples of outbreaks include the following:

A major outbreak involving nearly 10,000 cases (most of them cutaneous infection) occurred in Zimbabwe during the late 1970s and early 1980s (see References: Davies 1982). An epizootic in cattle occurred at that time in the same area.
An outbreak involving nine cases (five inhalational and four cutaneous) occurred in 1957 in the United States in a New Hampshire goat-hair processing plant (see References: Brachman 1960, Plotkin 1960). This was the last recognized outbreak of naturally occurring infection in this country.
An outbreak of oropharyngeal anthrax involving 24 cases occurred in Thailand in 1982 following consumption of contaminated meat (see References: Sirisanthana 1984). Oropharyngeal disease is an unusual manifestation of infection, which makes this outbreak of particular interest.

Epidemiology: Weaponized Anthrax

Anthrax as a Bioterrorist Weapon

Aerosol release of weaponized spores is the most likely mechanism for use of anthrax as a biological weapon (see References: Inglesby 2002). However, deliberate contamination of food also potentially could occur. During World War II, the Japanese reportedly impregnated chocolate with anthrax to kill Chinese children. The apartheid government of South Africa also experimented with anthrax in chocolate (see References: Sirisanthana 2002).

Although there is no formal definition of weaponized anthrax, weaponization for aerosol release generally involves:

Use of small particle size
A high concentration of spores
Treatment to reduce clumping
Neutralization of the electrical charge
Use of antimicrobial-resistant strains or genetic modification of the organism to increase virulence or escape vaccine protection

In 1972, more than 140 countries signed the Biological and Toxin Weapons Convention, which called for termination of all offensive biological weapons research and development and destruction of existing biological weapons stocks. However, the former Soviet Union continued to expand its biological weapons program (which included weaponization of anthrax) throughout the 1980s and early 1990s.

After the demise of the Soviet Union, many of the scientists who worked in the biological weapons program left the country. The status of those scientists remains unknown; however, Iraq, Iran, Syria, Libya, and North Korea actively have recruited such experts (see References: Henderson 1999). These countries and others have been suspected of ongoing development of offensive bioweapons programs. Recent reports suggest that at least five countries have offensive biological weapons programs and at least an additional five have research programs with possible production of offensive weapons (see References: Monterey Institute for International Studies).

The impact of a large aerosol release of weaponized anthrax remains unknown; however, scenarios have been hypothesized, including:

A 1970 WHO report estimated that an aerosol release of 50 kg of dried powder containing 6 x 10¹⁵ anthrax spores over a city of 5 million people in an economically developed country (such as the United States) would produce 250,000 incapacitating illnesses and up to 100,000 deaths (see References: WHO 1970).
A 1993 Office of Technology Assessment (OTA) study estimated that up to 3 million deaths could occur following the release of 100 kg of B anthracis (see References: OTA 1993).

Recent experience with aerosol spraying of Bacillus thuringiensis in Canada to control the European gypsy moth demonstrated the following pertinent findings (see References: Levin 2003):

Approximately 5 to 6 hours after the spray application began, indoor concentrations of airborne B thuringiensis actually exceeded outdoor concentrations, suggesting that the organisms entered homes and buildings after the aerosol release.
Although most of the particles were relatively large, particles with a medium diameter of 2 to 7 microns were detected both inside and outside of homes. The authors estimated that, at 5 to 6 hours after spraying, adults in the spray zone inhaled approximately 203 spores per hour.
Nasal swabs from asthmatic children in the spray zone were collected after spraying; Bacillus species with genetic patterns consistent with B thuringiensis were identified in 76.6% of isolates obtained from nasal passages.

The findings from this study suggest that it is technologically feasible to disseminate biological agents from aircraft; however, the applicability of this information to an intentional aerosol release of B anthracis is unknown. These data also raise concerns about indoor air safety should an intentional release of a biological agent occur. Modeling studies of viability of bioweapons agents in the environment indicate that anthrax is among the most resilient agents, since spores are capable of surviving longer in the environment than are vegetative cells (see References: Stuart 2005).

Deliberate contamination of food or water with anthrax spores also is a possibility. Even though contamination of a water supply is unlikely owing to the large volume of water involved and the chlorination process, contamination of smaller water sources is theoretically feasible since spore counts remain stable in water for at least several days following inoculation (see References: Beatty 2003). Since B anthracis spores are not destroyed by pasteurization, contamination of milk is another theoretical possibility (see References: Perdue 2003).

Weaponized anthrax has caused two outbreaks of disease (key points from each are outlined in the sections below). In addition, in July 1993, the religious group Aum Shinrikyo aerosolized a liquid suspension of B anthracis from the roof of an eight-story building in the Kameido district of Tokyo, but the release did not cause any human cases. Factors contributing to failure of this bioterrorist act included use of an attenuated B anthracis strain, low spore concentration, ineffective dispersal, a clogged spray device, and inactivation of spores by sunlight (see References: Takahashi 2004).

The Sverdlovsk Outbreak—1979

This outbreak in the Union of the Soviet Socialist Republics resulted from accidental release of anthrax spores from a military microbiologic facility where weaponized anthrax was being produced (see References: Meselson 1994).
Seventy-seven human cases were reported and 66 of the patients died, for a case-fatality rate of 86%; 75 cases were inhalational and two were cutaneous (one on the back of the neck and one on the shoulder). A recent statistical analysis of available data suggests that 250 cases with 100 fatalities may actually have occurred (see References: Brookmeyer 2001).
The mean incubation period was 9 to 10 days (range, 2 to 43 days), and the mean time between illness onset and death was 3 days.
Mean patient age for male cases was 42 years and for female cases was 55 years, and no cases occurred in children.
Approximately 2 weeks after the presumed date of exposure, a vaccination campaign of 59,000 eligible residents was begun; an estimated 80% of the target population received at least one dose of vaccine.
Investigators postulated that the weight of spores released as aerosol "could have been as little as a few milligrams or as much as nearly a gram."
Recent modeling studies suggest that the incubation period for anthrax is dose-dependent. The authors postulate that the relatively long incubation period for cases associated with Sverdlovsk was related to the level of exposure (see References: Wilkening 2006).

United States—2001

An outbreak of cutaneous/inhalational anthrax occurred in the United States in 2001.
The outbreak predominantly involved direct exposure to mail that was deliberately contaminated with anthrax spores. Several contaminated letters were sent and one was reported to contain 2 g of powder, with 100 billion to 1 trillion anthrax spores per gram (see References: Inglesby 2002).
The following features were noted in an epidemiologic report that summarized the outbreak findings (see References: Jernigan 2002):

Twenty-two cases (11 inhalational and 11 cutaneous) were identified. Five of the patients with inhalational anthrax died, for a case-fatality rate of 45% among that group.
Cases occurred in residents of seven states along the East Coast of the United States (Connecticut, Florida, Maryland, New Jersey, New York, Pennsylvania, and Virginia), with illness onsets between September 22 and November 16, 2001.
Four contaminated letters were recovered; all four were mailed in or around Trenton, New Jersey. Two were postmarked September 18, 2002, and two were postmarked October 9, 2002.
Twenty of the patients were either mail handlers or were exposed to work sites where mail was handled or received; one of these cases was a 7-month-old infant. The remaining two cases had no apparent association with contaminated mail. It is likely that these persons became exposed through cross-contamination of bulk mail that passed through contaminated mail facilities (see References: Griffith 2003, Holtz 2003).
B anthracis isolates were obtained from the four contaminated letters, 17 clinical specimens from cases, and 106 environmental samples collected along the mail path of the contaminated letters; all were identical by molecular subtyping.

The case of a 7-month-old infant with cutaneous anthrax was complicated by quick progression to severe microangiopathic hemolytic anemia despite early antibiotic treatment. The source is thought to be the workplace of the infant's mother, which the infant visited the day before the onset of symptoms. One possible exposure scenario, according to the authors, is that spores present on the hands of someone in the workplace who lifted or held the child may have contacted an exposed or possibly abraded area of the child's skin (see References: Freedman 2002).
Following recognition of anthrax cases in postal workers, air sampling was conducted before and during activation of a contaminated mail-sorting machine at a Washington, DC, postal facility. This testing (which was conducted several weeks after the contaminated mail passed through the machine) demonstrated that a mail-sorting machine can remain contaminated for many days after processing mail contaminated with B anthracis (see References: Dull 2002).
The outbreak demonstrated several important points about weaponized anthrax:

Mail can be an effective vehicle for disseminating anthrax spores.
Cross-contamination of mail likely can occur within postal facilities.
Persons who handle or process unopened contaminated mail are at risk of acquiring anthrax.
Substantial environmental contamination can occur in facilities handling contaminated mail or in offices where contaminated mail is opened.

Following the outbreak, a case of cutaneous anthrax occurred in a laboratory worker in Texas who was working at a private laboratory that was processing environmental samples from the CDC investigations (see References: CDC: Suspected cutaneous anthrax in a laboratory worker).
A 1-year health assessment of adult anthrax survivors demonstrated that survivors continued to report moderate to severe health complaints affecting multiple organ systems and had significantly greater overall psychological distress compared with US referent populations (see References: Reissman 2004). Fifty-three percent had not returned to work since their infection.
The source of the anthrax in this outbreak remains unknown, although Federal Bureau of Investigation (FBI) investigators believe that the perpetrator was domestic and had laboratory access to the Ames strain of B anthracis. The original Ames strain came from a laboratory in College Station, Texas (rather than Ames, Iowa). Several distinct Ames strains have been identified.
As a result of this outbreak, the US Postal Service developed and is deploying an autonomous detection system for anthrax in all mail-processing systems across the United States (see References: CDC: Responding to detection of aerosolized Bacillus anthracis by autonomous detection systems in the workplace). Completion of the project is expected by the end of 2005. See Use of Autonomous Detection Systems in the Workplace below for more information.

Clinical Features

Clinical Features of Cutaneous Anthrax
Feature	Characteristics
Incubation period	1-7 days (may be as long as 12 days)
Signs and symptoms*	—Initial lesion is small papule or vesicle, which may be pruritic. —By second day, papule ulcerates with central necrosis and drying. —Painless, localized, nonpitting edema surrounds ulcerated area. —Fine vesicles may encircle ulcer; these enlarge over next 1- 2 days and may discharge serosanguinous fluid. —After 1 to 2 days, painless black eschar forms over ulcerated area. —Eschar sloughs off after 12-14 days. —Lesions resolve without complications or scarring in 80%-90% of patients. —Extensive nonpitting edema, lymphangitis, and painful lymphadenopathy may occur. —Malignant edema is rare complication and is characterized by severe edema, multiple bullae, and shock.† —Fever and malaise are common.* —Lesions tend to occur on exposed areas of body (eg, face, hands, arms, neck). —One outbreak in Thailand demonstrated the following cutaneous findings for 13 patients with cutaneous anthrax‡: ~Eschar (85%) ~Blister (92%) ~Ulcer (23%) ~Edema around lesion (77%) ~Lymphadenopathy (100%)
Case-fatality rate	—Currently <1%* (most patients recover with appropriate antimicrobial therapy) —In preantibiotic era, case-fatality rates of about 20% were reported.§ —A literature review of pediatric anthrax cases identified between 1900 and 2005 demonstrated an overall mortality rate for cutaneous disease of 14% (5 of 37 cases).** Not all cases in this report received antimicrobial therapy.
Laboratory findings	—Gram stain of lesion may reveal gram-positive rods; neutrophils are uncommon. —WBC count often is normal or may be slightly elevated.* —Histologic examination shows necrosis, edema, and lymphocytic infiltrate.††
Abbreviation: WBC, white blood cell. See References: Gold 1955. †See References: Amadi 1974. ‡See References: Kunanusont 1990. §See References: Smyth 1941. *See References: Bravata 2007. ††See References: Dixon 1999.

Clinical Features of Inhalational Anthrax
Feature	Characteristics
Incubation period	2-43 days (may be as short as 1 day or may be longer)*
Signs and symptoms	—Illness may be biphasic, with an initial prodrome (that includes symptoms such as fever, malaise, fatigue, anorexia) followed by sudden increase in fever, severe respiratory distress, diaphoresis and shock, if left untreated. —Symptoms for 10 patients with inhalational anthrax identified during the 2001 US outbreak†: ~Fever, chills (100%) (7 were febrile on initial presentation) ~Sweats, often drenching (70%) ~Fatigue, malaise, lethargy (100%) ~Cough (minimally or nonproductive) (90%) ~Nausea or vomiting (90%) ~Dyspnea (80%) ~Chest discomfort or pleuritic pain (70%) ~Myalgias (50%) ~Headache (50%) ~Confusion (40%) ~Abdominal pain (30%) ~Sore throat (20%) ~Rhinorrhea (10%) —In the 2001 US outbreak, no evidence of a mild form of inhalational anthrax was detected through follow up serologic testing of exposed persons.† —A systematic review of 82 inhalational anthrax cases reported between 1900 and 2005 found that the most common symptoms or findings on admission included the following‡: ~Abnormal lung findings (80%) ~Fever or chills (67%) ~Tachycardia (66%) ~Fatigue or malaise (64%) ~Cough (62%) ~Dyspnea (52%) ~All 26 patients who had chest radiography had abnormal findings, including pleural effusion (69%) or widened mediastinum (54%)
CFR	—Sverdlovsk outbreak: 86%* —US outbreak: 45%**(lower observed CFR in the US outbreak likely was due to early diagnosis and aggressive therapy) —In a systematic review of 82 cases of inhalational anthrax, the overall CFR was 85%; however, patients in the US 2001 outbreak who received early antibiotic therapy (ie, during the prodromal phase [<4.7 days after illness onset]) had a CFR of 40% and those who received antibiotics >4.7 days after illness onset had a CFR of 75%‡ —A literature review of pediatric anthrax cases identified between 1900 and 2005 demonstrated an overall mortality rate for inhalational disease of 60% (3 of 5 cases).§ Not all cases in this report received antimicrobial therapy.
Laboratory findings	Findings for 10 patients with inhalational anthrax identified during 2001 US outbreak: —Median WBC count at presentation was 9,800/mm³(range, 7,500/mm³ to 13,300/mm³) —Differential WBC count >70% neutrophils (70%) —Band forms present (4 of 5; 40%) —Peak WBC during illness was 26,400/mm³(range, 11,900/mm³ to 49,600/mm³) —Elevated transaminases (SGOT or SGPT) >40 (90%) —Hypoxemia†† (60%) —Metabolic acidosis (20%) —Abnormal chest radiograph (100%): ~Mediastinal widening (70%) ~Infiltrates, consolidation (70%) ~Pleural effusion (80%) —Abnormal CT scan (8 of 8; 100%): ~Mediastinal lymphadenopathy, widening (7 of 8; 88%) ~Pleural effusion (8 of 8; 100%) ~Infiltrates, consolidation (6 of 8; 75%)
Abbreviations: CFR, case-fatality rate; CT, chest-computed tomography; SGOT, serum glutamic oxalacetic transaminase; SGPT, serum glutamic pyruvic transaminase; WBC, white blood cell. See References: Meselson 1994. †See References: Baggett 2005. ‡See References: Holty 2006. §See References: Bravata 2007. *See References: Jernigan 2001. ††Alveolar-arterial oxygen gradient >30 Hg on room air; O₂ saturation <94%.

Clinical Features of Gastrointestinal and Oropharyngeal Anthrax
Feature	Characteristics
Incubation period	1-7 days (usually 2-5 days)
Signs and symptoms	—One outbreak in Uganda demonstrated the following findings in 143 patients*: ~Fever (may be low-grade) (70%) ~Abdominal tenderness (85%) ~Diarrhea (80%; bloody in only 5%) ~Vomiting (may be coffee-ground or blood-tinged) (90%) ~Headache (100%) ~Pharyngeal edema (10%) —Ascites may develop 2-4 days after onset (fluid may be clear or purulent)† —Ulcerations can occur anywhere along the GI tract and may cause hemorrhage, obstruction, or perforation‡ —If the patient survives, symptoms last about 2 wk —One outbreak of oropharyngeal anthrax in Thailand demonstrated the following findings for 24 patients§: ~Neck swelling (100%) ~Fever (96%) ~Sore throat, dysphagia (63%) ~Mouth or pharyngeal ulcerative or necrotic lesions (100%) (pseudomembranes also were noted in some patients) ~Respiratory distress (25%) ~Hoarseness (8%) ~Sensation of a "lump in throat" (8%) ~Diarrhea (4%) ~Bleeding from the mouth (4%)
Case-fatality rate	—Rate for GI anthrax is between 25% and 60%.** In outbreaks where patients received antibiotic therapy, rates have ranged from 0% to 29%.†† —A literature review of pediatric anthrax cases identified between 1900 and 2005 demonstrated an overall mortality rate for gastrointestinal disease of 65% (13 of 20 cases).‡‡ Not all cases in this report received antimicrobial therapy. —In Thailand outbreak of oropharyngeal disease, rate was 13%.§ In another report of 6 cases of pharyngeal anthrax, rate was 50%.§§
Laboratory findings	—Gram stain of peritoneal fluid or oropharyngeal ulcers may demonstrate gram-positive rods. —Median WBC count for 13 patients with oropharyngeal anthrax in Thailand outbreak was 15,635/mm³ (range, 5,100/mm³ to 30,570/mm³). Mean percentage of neutrophils was 79.6% (range, 73% to 91%). —B anthracis has been cultured from oropharyngeal swaps and stool specimens in patients with GI anthrax.††
Abbreviations: GI, gastrointestinal; WBC, white blood cell. See References: Ndyabahinduka 1984. †See References: Dixon 1999. ‡See References: Sirisanthana 2002. §See References: Sirisanthana 1984. *See References: CDC: Investigation of anthrax associated with intentional exposure and interim public health guidelines. ††See References: Beatty 2003. ‡‡See References: Bravata 2007. §§See References: Doganay 1986.

Clinical Features of Anthrax Meningitis
Feature	Characteristics
Incubation period	Varies according to primary source of infection.
Signs and symptoms*†‡§	—May occur as complication of cutaneous, inhalational, or gastrointestinal anthrax and symptoms of primary site of infection usually will be present; however, meningitis may be the presenting illness. —Characteristic features of bacterial meningitis usually present (eg, fever, nuchal rigidity, headache, change in mental status, seizures). —Nausea and/or vomiting are common. —Hemorrhagic meningoencephalitis is a characteristic presentation.
Case-fatality rate**	—Illness fatal in >90% of cases. —A literature review of pediatric anthrax cases identified between 1900 and 2005 demonstrated an overall mortality rate for meningoencephalitis of 100% (6 of 6 cases).‡‡ Not all cases in this report received antimicrobial therapy. —Death usually occurs 1 to 6 days after illness onset, and 75% of patients die within 24 hr of presentation.
Laboratory findings	—Gram stain of CSF reveals many gram-positive rods. —CSF is usually bloody. —Report of two cases demonstrated elevated WBC counts in both patients at presentation (24,000 with 84% neutrophils and 18,000 with 90% neutrophils).† —CT or MRI of head may show focal intracerebral hemorrhage, subarachnoid hemorrhage, diffuse cerebral edema, intraventricular hemorrhage, and/or leptomeningeal enhancement.§
Abbreviations: CSF, cerebrospinal fluid; CT, computed tomography; MRI, magnetic resonance imaging; WBC, white blood cell. See References: Dixon 1999. †See References: Rangel 1975. ‡See References: Meyer 2003. §See References: Sejvar 2005 *See References: Lanska 2002 ‡‡See References: Bravata 2007.

Differential Diagnosis

The differential diagnosis for anthrax depends upon the clinical syndrome (inhalational, cutaneous, gastrointestinal, or meningeal). Other diagnoses to consider are outlined in the tables below.

Differential Diagnosis for Cutaneous Anthrax
(Note:* Two key features that distinguish cutaneous anthrax from other conditions in differential diagnosis are painlessness of the lesion and the relatively large extent of associated edema.)*
Diagnosis*†‡	Distinguishing Features
Ecthyma gangrenosum	—Usually in neutropenic patients with Pseudomonas aeruginosa bacteremia —Edema usually not present
Ulceroglandular tularemia (Francisella tularensis)	—Clinical course usually indolent; disease often self-limited —Systemic toxicity uncommon
Bubonic plague (Yersinia pestis)	—Systemic toxicity common —Extremely tender regional lymphadenopathy present —Ulceration and eschar formation usually absent
Staphylococcal or streptococcal cellulitis	—May be history of trauma or preexisting lesion at site of infection —Eschar formation does not occur —Usually painful
Necrotizing soft tissue infections (particularly Group A streptococcus and Clostridium species)	—Severe systemic toxicity often present —Early in course, pain usually more severe than clinical findings would indicate
Bite of brown recluse spider (Loxosceles reclusa)§	—Brown recluse spiders prefer warm temperatures and are not native to Northern half of United States —Spiders tend to hide in barns, woodpiles, and similar places —Bite usually causes painful blister which progresses to necrosis (unlike anthrax, which is painless) —Edema generally absent
Rickettsialpox (Rickettsia akari)	—Initial presentation involves painless papule which forms black eschar —Generalized maculopapular rash appears 2-3 days later
Scrub typhus (Orientia tsutsugamushi; formerly Rickettsia tsutsugamushi)	—Zoonotic infection from chigger bites; occurs in endemic areas (Asia and Western Pacific) —Often associated with generalized maculopapular rash
Orf (orf virus, a parapox virus)	—Occurs in farm workers —Characterized by pustule that progresses to weeping nodule —Eschar formation does not occur —Edema usually absent
Necrotic herpes simplex infection	—More likely to occur in immunocompromised host
*See References: Dixon 1999. †See References: Swartz 2001. ‡See References: Bell: Clinical issues in the prophylaxis, diagnosis, and treatment of anthrax. §See References: Nelson 2002.

Differential Diagnosis for Inhalational Anthrax
(Note: Features that distinguish inhalational anthrax from other conditions in differential diagnosis include presence of widened mediastinum and pleural effusions on chest radiograph or CT scan with minimal evidence of pneumonia)
Diagnosis*†	Distinguishing Features
Pneumonic plague (Yersinia pestis)	—Hemoptysis relatively common with pneumonic plague, but rare with inhalational anthrax
Tularemia (Francisella tularensis)	—Clinical course usually indolent, lasting weeks —Less likely to be fulminant
Community-acquired bacterial pneumonia —Mycoplasmal pneumonia (Mycoplasma pneumoniae) —Pneumonia caused by Chlamydia pneumoniae —Legionnaires' disease (Legionella pneumophila or other Legionella species) —Psittacosis (Chlamydia psittaci) —Other bacterial agents (eg, Staphylococcus aureus, Streptococcus pneumoniae, Haemophilus influenzae, Klebsiella pneumoniae, Moraxella catarrhalis)	—Rarely as fulminant as inhalational anthrax —Legionellosis and many other bacterial agents (S aureus, S pneumoniae, H influenzae, K pneumoniae, M catarrhalis) usually occur in persons with underlying pulmonary or other disease or in elderly —Bird exposure with psittacosis —Gram stain of sputum may be useful —Community outbreaks caused by other etiologic agents not likely to be as explosive as pneumonic plague outbreak —Outbreaks of S pneumoniae usually institutional —Community outbreaks of Legionnaires' disease often involve exposure to cooling towers
Viral pneumonia —Influenza —Hantavirus —RSV —CMV	—Influenza generally seasonal (October-March in United States) or involves history of recent cruise ship travel or travel to tropics —Exposure to mice droppings, feces with Hantavirus —RSV usually occurs in children (although may be cause of pneumonia in elderly); tends to be seasonal (winter/spring) —CMV usually occurs in immunocompromised patients
—Q fever (Coxiella burnetii)	—Exposure to infected parturient cats, cattle, sheep, goats —Severe pneumonia not prominent feature
Abbreviations: CMV, cytomegalovirus; CT, computed tomography; RSV, respiratory syncytial virus. *See References: Dixon 1999. †See References: Bell: Clinical issues in the prophylaxis, diagnosis, and treatment of anthrax.

Differential Diagnosis for Gastrointestinal and Oropharyngeal Anthrax
Diagnosis	Distinguishing Features
Gastrointestinal Anthrax*
Typhoid fever (Salmonella typhi)	—Ascites usually not present —Other clinical features may be similar
Intestinal tularemia (Francisella tularensis)	—Illness often less severe than that seen with gastrointestinal anthrax —Ascites not present —Less likely to resemble acute abdomen —Fever may be less prominent
Bacillary dysentery (Shigella dysenteriae)	—Ascites usually not present —Other clinical features may be similar
Acute bacterial gastroenteritis caused by other agents (eg, Campylobacter jejuni, Shiga toxin–producing Escherichia coli, Yersinia enterocolitica)	—Illness often less severe than that seen with gastrointestinal anthrax —Ascites not present —Less likely to resemble acute abdomen —Fever may be less prominent —Hemolytic uremic syndrome may occur with infection caused by Shiga toxin–producing E coli
Bacterial peritonitis	—Gastrointestinal symptoms (nausea, vomiting, gastrointestinal bleeding, diarrhea) not prominent features —Tends to occur in persons with underlying medical conditions (eg, alcoholism, other liver disease)
Acute abdomen (eg, appendicitis)†	—Anthrax generally begins with vague systemic symptoms rather than abdominal pain —Ascites is relatively common with gastrointestinal anthrax and less common with appendicitis and similar conditions
Oropharyngeal Anthrax
Diphtheria (Corynebacterium diphtheriae)	—Primarily occurs in nonimmune children under 15 yr of age —Pharyngeal membrane is prominent feature; ulcerative or necrotic lesions generally not present —Removal of pharyngeal membrane often causes bleeding of submucosa
Pharyngeal tularemia (Francisella tularensis)	—Neck swelling usually absent —Exudative pharyngitis common; ulcerative lesions may occur
Streptococcal pharyngitis (Streptococcus pyogenes)	—Exudative pharyngitis most prominent feature; necrotic ulcers generally absent —Neck edema usually absent, although cervical lymphadenopathy may be prominent
Infectious mononucleosis	—Most common in young adults —Splenomegaly commonly occurs —Neck edema usually absent, although cervical lymphadenopathy may be prominent
Enteroviral vesicular pharyngitis (coxsackievirus)	—Small vesicles noted on soft palate, uvula, or anterior tonsillar pillars —Generally occurs in children —Neck edema usually absent
Acute herpetic pharyngitis (herpes simplex virus)	—Vesicles, shallow ulcers may be noted, but lesions usually not necrotic —Neck edema usually absent, although cervical lymphadenopathy may be prominent
Anaerobic pharyngitis (Vincent's angina)	—Purulent exudate covers posterior pharynx —Tonsillar abscesses may occur —Neck edema usually absent
Yersinia enterocolitica pharyngitis	—Exudative pharyngitis most prominent feature —Neck edema usually absent —Cervical adenopathy, abdominal pain may occur
*See References: Dixon 1999. †See References: Kanafani 2003.

Differential Diagnosis for Anthrax Meningitis
Diagnosis	Distinguishing Features
Subarachnoid hemorrhage	—Fever not usually prominent feature —Can be distinguished by computed tomography without contrast*


Home	Mission & Activities	About Us	Center Support	Contact Us