IDSA | Botulism: Current, comprehensive information on pathogenesis, microbiology, epidemiology, diagnosis, and treatment

Last updated May 23, 2008

Agent
Pathogenesis
Epidemiology
—Foodborne Botulism
—Wound Botulism
—Infant Botulism
—Adult Intestinal Toxemia Botulism
—Inhalational Botulism
—Iatrogenic Botulism
—Botulinum Toxin as a Biological Weapon
—Use of Therapeutic Botulinum Toxin
Clinical Features
Pediatric Considerations
Differential Diagnosis
Laboratory Diagnosis
—Specimen Collection and Transport
— Laboratory Biosafety
—Laboratory Response Network
—Diagnostic Tests for Detection of Botulinum Toxin and C botulinum
Prevention and Treatment Issues
—Therapy for Botulism
—Botulinum Toxoid
—Research on New Therapies and Vaccines
Emergency Response
Infection Control
Issues Related to Autopsies and Burial
Case Definitions and Public Health Reporting
References

Agent

Botulinum Toxin

Botulism is an intoxication caused by botulinum toxin, which is produced by Clostridium botulinum and, rarely, by other Clostridium species. Seven antigenically distinct toxin types (A, B, C, D, E, F, G) have been identified.

The following are key characteristics of botulinum toxins (see References: CDC 1998: Botulism in the United States, 1899-1996; Hatheway 1998; Lacy 1998; Schiavo 1994; Sneath 1986).

Botulinum toxins are the most lethal toxins known. For type A toxin, the toxic dose is estimated at 0.001 mcg/kg (see References: Franz 1997); the lethal dose for a 70-kg person by the oral route is estimated at 70 mcg, by the inhalational route 0.80 to 0.90 mcg, and by the intravenous route 0.09 to 0.15 mcg (see References: Sobel 2005). The toxins are identified by neutralization with type-specific antitoxin; minor cross-neutralization between types C and D and between types E and F has been observed (see References: Smith 1988).
The toxins are produced by vegetative cells (ie, germination of spores) and released by cell lysis.
Some toxins are fully activated by the bacteria that produce them (proteolytic strains of type A, B, and F), and some require exogenous proteolytic activation (types E and non-proteolytic types B and F). Research with two strains revealed variation in toxin production. Further research with these strains may provide insight into bacterial growth, toxin production, and toxin processing and release (see References: Rao 2007).
Types A, B, E, and F cause natural disease in humans. The vast majority of disease is caused by types A, B, and E; type F rarely occurs (ie, about 1% of US cases; [see References: Gupta 2005]).
In a recent study, a novel in vivo mouse assay was used to correlate toxin type and dosage with the duration of muscle paralysis for types A, B, and E (see References: Keller 2006).

Botulinum toxin A produced longer paralysis than botulinum toxin B, consistent with human observations.
For type A, duration of paralysis was exponentially related to toxin dose; the paralysis time doubled with every 25% increase of the toxin concentration.
For type B, the duration of paralysis was linear relative to the toxin dose.
Type E toxin had the shortest duration of action, but unlike the other two toxins, the dose of toxin did not influence recovery time.

Types C and D cause natural disease in birds, horses, and cattle; strains that produce these types reside in the intestinal tract of certain animals.
Toxin type G has never clearly been shown to cause human disease.
Toxin types C, D, and G cause botulism in primates when administered through aerosol challenge (see References: Middlebrook 1997). As a result of these experiments, experts generally believe that humans also are susceptible to these types.
Botulinum toxins are colorless, odorless, and presumably tasteless.
Aerosolized particles of toxin are approximately 0.1 to 0.3 mcm in size (see References: Shapiro 1997).
The toxins are inactivated by heating (>85°C for 5 minutes) (see References: Siegel 1993).
In the event of an intentional release of botulinum toxin, the causative organisms may or may not be present.

Clostridium botulinum

The following are key microbiologic characteristics of C botulinum (see References: CDC 1998: Botulism in the United States, 1899-1996; Hatheway 1998; Smith 1988; Sneath 1986).

Gram-positive spore-forming bacillus (may stain poorly)
Somewhat varying strain sizes but generally in the range of 0.5 to 2.0 mcm in width and 1.6 to 22.0 mcm in length (see References: CDC 1998: Botulism in the United States, 1899-1996)
Straight to slightly curved, with a peritrichous flagellum
Spores are oval, eccentric to subterminal, and usually swell the bacterial cell
Strict anaerobe
"Sluggishly" motile
Produce lipase on egg-yolk agar
Ferment glucose and liquefy gelatin (all strains)
Commonly isolated from soil and marine and lake sediments

The classification of C botulinum strains is based on metabolic activity (groups I to IV) and on toxin types (types A to G) (see References: Hatheway 1998, Sneath 1986, Smith 1988):

Group I includes type A strains and proteolytic strains of types B and F.
Group II includes type E strains and nonproteolytic strains of types B and F
Group III includes nonproteolytic strains of types C and D.
Group IV includes only strains that produce type G.
Strains that produce more than one toxin type or have genetic sequences encoding more than 1 toxin have been identified (see References: Barash 2004, Kirma 2004)
Each group has a different optimal growth temperature, but there are no colonial morphology features that allow distinction between groups or antigenic types.
Genetic homology has been demonstrated within antigenic groups of C botulinum, and there is minimal antigenic cross-reactivity between groups.
Antimicrobial susceptibilities of C botulinum strains vary somewhat by group, but most strains are susceptible to penicillin, metronidazole, rifampin, and erythromycin (see References: Smith 1988).
Recent publication of the complete genome of a C botulinum strain (Hall A, American Type Culture Collection [ATCC] 3502) may fuel additional investigations of bacterial metabolic processes and gene regulation (see References: Sebaihia 2007). The genome indicates that C botulinum is adapted to a saprophytic lifestyle in both soil and aquatic environments.

C botulinum spores have the following features (see References: Smith 1988):

Spores may survive boiling for up to 3 to 4 hours or temperatures of 105^oC for 100 minutes.
Spores are readily killed by chlorine (either as chlorinated water or as diluted solutions of hypochlorite).
Spores undergo maximum germination when activated by heat. For example, type A strains undergo maximum germination by heat treatment (or "heat shocking") at 80°C for 10 to 20 minutes.
Spores are resistant to desiccation and can survive in the dry state for 30 years or more.
Spores are resistant to ultraviolet light, alcohols, and phenolic compounds. They are relatively resistant to irradiation.

Other Neurotoxin-Producing Clostridium Species

Clostridium butyricum–producing type E toxin has been reported to cause intestinal botulism in infants and young adults in Italy and foodborne botulism in Asia (see References: Aureli 1986, Fenicia 1999, Schecter 1999).
Clostridium baratii producing type F toxin has caused intestinal botulism in infants and adults; in the latter it is usually associated with gastrointestinal pathology, recent gastrointestinal surgery, or recent use of antimicrobial agents (see References: Barash 2005, Gupta 2005, McCroskey 1991, Schecter 1999).

Pathogenesis

Exposure to botulinum toxin occurs through the following mechanisms (toxin is not absorbed through intact skin):

Ingestion of preformed toxin
Inhalation of preformed toxin
Local production of toxin by C botulinum organisms in the gastrointestinal tract
Local production of toxin by C botulinum organisms in devitalized tissue at the site of a wound

Following exposure, pathogenesis includes the following steps (see References: Arnon 2001; CDC 1998: Botulism in the United States, 1899-1996; Halpern 1995; Schiavo 1995; Simpson 2004):

Botulinum toxin is activated by proteolytic cleavage; the activated structure is a 150-kd polypeptide comprising two chains (a heavy chain [100 kd] and a light chain [50 kd]) that are connected by a single disulfide bond.
Botulinum toxin enters the circulation and is transported to the neuromuscular junction.
At the neuromuscular junction, the heavy chain of the toxin binds to the neuronal membrane on the presynaptic side of the peripheral synapse.
The toxin then enters the neuronal cell via receptor-mediated endocytosis. A recent study demonstrated that botulinum toxin type A binds to the synaptic vesicle protein SV2 (isoforms A, B, and C) on the neuronal surface (see References: Dong 2006). Recent analyses have defined the cell surface receptor and neurotoxin interactions for neurotoxin B and have clarified the three-dimensional relationships between toxin and receptor (see References: Chai 2006, Jin 2006). Mutational studies of neurotoxin B have identified amino acids critical for receptor binding (see References: Kohda 2007, Rummel 2007).
The light chain of the toxin crosses the membrane of the endocytic vesicle and enters the cytoplasm.
Once inside the cytoplasm, the light chain of the toxin (which is a zinc-containing endopeptidase) cleaves some of the proteins that form the synaptic fusion complex. Mutational experiments indicate that two glutamic acid residues form a critical part of the toxin's active site (see References: Kukreja 2007).The synaptic proteins, referred to as SNARE proteins, include synaptobrevin (cleaved by toxin types B, D, F, and G), syntaxin (cleaved by toxin type C), and synaptosomal-associated protein (SNAP-25; cleaved by toxin types A, C, E) (see References: Arnon 2001). The clostridial neurotoxin apparently first binds to the SNARE complex through substrate-recognition exosites before cleavage occurs (see References: Breidenbach 2004). Researchers have recently defined a multistep mechanism for recognition and cleavage of SNAP-25 by type A toxin that may offer insight into design of inhibitors (see References: Chen 2007); two candidates have been identified with high throughput methods (see References: Eubanks 2007).
The synaptic fusion complex allows the synaptic vesicles (which contain acetylcholine) to fuse with the terminal membrane of the neuron. Disruption of the synaptic fusion complex prevents the vesicles from fusing with the membrane, which in turn prevents release of acetylcholine into the synaptic cleft.
Without neuronal acetylcholine release, the affiliated muscle is unable to contract and becomes paralyzed.
The blockade of acetylcholine release lasts up to several months; normal functioning slowly resumes either through turnover of SNARE proteins within the cytoplasm or through production of new synapses.
Death from botulism results acutely from airway obstruction or paralysis of respiratory muscles. Death also can result from complications related to prolonged ventilatory support and intensive care.
Botulinum toxin apparently does not cross the blood-brain barrier; therefore, central nervous system functions remain intact.
A new yeast assay that employs chimeric soluble N-ethylmaleimide–sensitive factor attachment protein receptor (SNARE) proteins may help elucidate mechanisms of botulinum toxicity. The process involves analysis of intracellular botulinum neurotoxin and light-chain endoprotease interactions and proteolysis of SNARE proteins (see References: Fang 2006).

Epidemiology

Foodborne Botulism

Foodborne botulism is caused by ingestion of food contaminated with preformed botulinum toxin and subsequent absorption of toxin through the gastrointestinal tract. The following steps are necessary for a food item to cause botulism (see References: CDC 1998: Botulism in the United States, 1899-1996):

The food item must be contaminated with C botulinum spores, which are normally found in soil.
The spores must survive food preservation methods.
Adequate conditions for spore germination and neurotoxin production must be present (see next bullet).
The food must not be reheated adequately (>85°C for 5 minutes) to inactivate the heat-labile toxin before the food is consumed (see References: Siegel 1993).

Generally, adequate conditions for germination and neurotoxin production include the following, although various caveats exist (see References: CDC 1998: Botulism in the United States, 1899-1996; Solomon 2001, Smith 1988):

An anaerobic environment
Nonacidic pH (generally 4.6 to 4.8; pockets of different pH may be present within a single food source and allow toxin to be produced in a food that overall has an acidic pH)
Minimum temperature of 10°C (the optimum temperature for growth of proteolytic strains is close to 35°C; some nonproteolytic strains of types B, E, and F can produce toxin at refrigeration temperatures [3°C to 4°C])
Availability of water with limited solute concentration

Toxin types A, B, and E account for most cases of foodborne botulism, and toxin types tend to be geographically distributed within the United States. The outbreaks reported to the CDC between 1950 and 1996 (see References: CDC 1998: Botulism in the United States, 1899-1996) were distributed as follows:

144 (86%) of 167 type A outbreaks occurred west of the Mississippi River
37 (61%) of 61 type B outbreaks occurred east of the Mississippi River
56 (84%) of 67 type E outbreaks occurred in Alaska

An unusual outbreak of type F foodborne botulism was caused by home-prepared venison jerky (see References: CDC 1998: Botulism in the United States, 1899-1996; Midura 1972).
The median number of cases of foodborne botulism reported to the CDC annually between 1973 and 1996 was 24 (range, 8 to 86 cases) (see References: Shapiro 1998).
The mean number of foodborne botulism outbreaks per year between 1950 and 1996 was 9.4, with a mean number of 2.5 cases per outbreak (see References: CDC 1998: Botulism in the United States, 1899-1996).
Between 1990 and 2000, the median number of botulism events per year was 14 (range, 9 to 24) and the median number of cases per event was 1 (range, 1-17) (see References: Sobel 2004). During this time period, the highest incidence rates were in Alaska (19 per million population), Idaho (0.6 per million population), and Washington (0.3 per million population).
Improperly home-canned or home-prepared foods (particularly vegetables) continue to account for most of the food vehicles associated with foodborne botulism in the United States (see References: Sobel 2004). Two foodborne botulism cases in California in 2006 arose from home-fermented tofu (see References: CDC 2007: Brief report: foodborne botulism).
Over the past 20 years, a wide variety of commercially produced (preserved and nonpreserved) foods have caused botulism outbreaks.

Examples include foil-wrapped baked potatoes, sauteed onions held under a layer of butter, garlic in oil, commercially produced cheese sauce, commercially prepared chili, hazelnut yogurt, jarred peanuts, and matambre (Argentine meat roll) sealed in heat-shrinked plastic wrap (see References: Angulo 1998; Chou 1988; Kalluri 2003; MacDonald 1985: Type A botulism from sauteed onions; O'Mahony 1990; St Louis 1988; Townes 1996; Villar 1999).
In September 2006, six cases of botulism (four in the United States and two in Canada) were associated with consumption of commercially produced carrot juice that was believed to have been improperly stored; however, mishandling by patients could not be confirmed (see References: CDC 2006: Botulism associated with commercial carrot juice; WHO 2006). The FDA issued an advisory note about proper refrigeration of carrot juice after the incident occurred (see References: FDA 2006).
In mid-July 2007, four suspect cases of botulism (two in Texas and two in Indiana) were apparently associated with consumption of commercially produced hot dog chili sauce. Botulism type A was confirmed in two patients (see References: CDC 2007: Botulism associated with commercially canned chili sauce—Texas and Indiana July 2007 [MMWR]). The product was produced by Castleberry's, a company located in Augusta, Ga. In response to the cases, the company recalled the implicated product and then on Jul 21, 2007, expanded the recall to include more than 80 products (see References: Castleberry News Staff 2007; FDA 2007: FDA issues nationwide warning; FDA 2007: FDA expands its nationwide warning). As of late August 2007, eight cases of botulism (in Indiana, Texas, and Ohio) had been associated with this outbreak (see References: CDC 1998: Botulism associated with canned chili sauce, July-August 2007). The investigation traced contamination to commercial-scale pressure cookers for processing canned foods at a Georgia facility (see References: CDC 2007: Botulism associated with commercially canned chili sauce—Texas and Indiana, July 2007 [MMWR]).

A variety of salted, fermented, smoked, and canned fish sources have been implicated in type E botulism outbreaks in the United States and worldwide (see References: Lindstrom 2006, Telzak 1990). In the United States, five individuals contracted mild type E botulism after consuming home-salted fish. Their symptoms were mostly gastrointestinal (see References: Sobel 2007).
Foodborne botulism is a significant public health problem among Alaskan natives and is usually associated with consumption of fermented meat from aquatic mammals and fish (see References: McLaughlin 2004; Shaffer 1990, Wainwright 1988).
A recent large foodborne outbreak in Thailand involving 209 cases was associated with consumption of home-canned bamboo shoots (see References: CDC 2006). At least 42 people experienced respiratory failure and were treated in hospitals (see References: Kongsaengdao 2006).
Occasionally, unusual food preparation methods (particularly for home-prepared products) can lead to botulism. For example, an outbreak in Turkey (eastern Anatolia) in 2005 was associated with eating "suzme" (yoghurt buried under soil) (see References: Akdeniz 2007).
Sales of minimally heated, chilled foods have grown recently in Western countries, such as the United States and the United Kingdom, and have raised concerns about the potential for foodborne botulism (see References: Peck 2006).
Waterborne botulism has not been reported , most likely for the following reasons (see References: Arnon 2001, Siegel 1993):

Botulinum toxin is rapidly inactivated by standard treatment of potable water.
A very large amount of toxin would be needed to contaminate a water supply on account of the dilution factor.

Wound Botulism

Wound botulism is caused by infection of a contaminated wound with C botulinum and subsequent absorption into the circulation of locally produced toxin.
C botulinum is a natural contaminant of soil throughout the United States (see References: Smith 1978).
Wound botulism has been recognized with increasing frequency among injecting drug users, particularly in California, where the disease has been associated with use of black tar heroin (see References: MacDonald 1985: Botulism and botulism-like illness in chronic drug users; Passaro 1998; Werner 2000). Similarly, in the United Kingdom, bacterial infections (particularly wound botulism) have increased markedly since 2000 among injecting heroin users (see References: Brett 2005: Soft tissue infections caused by spore-forming bacteria in injecting drug users in the United Kingdom). The authors of this study observed that the major risk factor was skin- or muscle-popping. Cases also have been reported in Germany (see References: Preuss 2006) and in Sweden, where real-time PCR was used to diagnose a case of type wound E botulism (see References: Artin 2007).
Wound botulism in drug abusers can be misdiagnosed as drug intoxication (see References: Royl 2007); presenting features can alert physicians to the correct diagnosis (see References: Wenham 2008). It should be considered in injecting drug users who present with dysarthria and dysphagia (see References: Preuss 2006).
Cases may occur following traumatic injury to an extremity, such as a compound fracture, laceration, puncture wound, gunshot wound, severe abrasion ("road rash"), or crush injury (see References: Merson 1973, Werner 2000).
Sinusitis associated with intranasal cocaine use has been the source of wound botulism in a few cases (see References: Kudrow 1988; MacDonald 1985: Botulism and botulism-like illness in chronic drug users; Werner 2000). Although no evidence of sinusitis was found for two recent cases of botulism associated with intranasal cocaine use in France, the authors postulate that occult sinusitis was the most likely source of infection (see References: Roblot 2006). However, they also suggested that the two patients may have inhaled minute amounts of preformed toxin along with the cocaine, resulting in direct absorption of toxin through the nasal mucosa.
A few cases have occurred postoperatively (usually following intra-abdominal procedures) and an abscessed tooth was the source of C botulinum infection in one case (see References: Weber 1993).
Between 1943 (when the condition was first recognized) and 1985, 33 cases of wound botulism were reported to the CDC. Between 1986 and 1996, 78 cases were reported and most were associated with injecting drug use (see References: CDC 1998: Botulism in the United States, 1899-1996).

Infant Botulism

Infant botulism is caused by ingestion of C botulinum spores. The spores subsequently colonize the gastrointestinal tract, germinate, and produce toxin, which is absorbed into the circulation.
Most infants are well before illness onset (see References: Wigginton 1993). The disease characteristically begins with lethargy and poor feeding (with or without constipation), followed by neuromuscular paralysis, hypotonia, or weakness (see References: Clemmens 2007). Constipation may be subtle or overt.
The source of spores for most cases remains unknown, although the most common sources of infection for infants appear to be honey and environmental exposure (see References: Arnon 1979, Brook 2007).

One recent case report of infant botulism described a 5-month-old infant who was found to be infected with two different strains of C botulinum type B (as identified by DNA fingerprinting using amplified fragment length polymorphism [AFLP]). Culture of 14 food and drink items from the home yielded the same two strains of C botulinum type B from opened infant formula. Two other strains also were identified in the formula, and C botulinum type A was isolated from an opened container of dried rice pudding (see References: Brett 2005: A case of infant botulism with a possible link to infant formula milk powder).
In another situation, the same strain of C botulinum type B was identified in the intestinal contents of an 11-week-old infant who died suddenly and in vacuum cleaner dust from the infant's home, suggesting environmental exposure to dust as the source of infection (see References: Nevas 2005).
Spores have been identified in chamomile tea, which in some countries is used as an herbal remedy for intestinal colic in infants (see References: Bianco 2008).

Other risk factors identified in one study for infants 2 months of age and older included breast-feeding, less than one bowel movement per day in the 2 months before illness onset, and ingestion of corn syrup (see References: Spika 1989). In that study, the only identified risk factor among infants less than 2 months old was living in a rural area or on a farm.
Between 1976 (when infant botulism was first recognized) and 1996, 1,442 cases were reported to CDC (see References: CDC 1998: Botulism in the United States, 1899-1996).

Cases were reported from 46 states, with Delaware, Hawaii, Utah, and California having the highest incidence rates (9.0, 8.8, 6.3, and 5.7 per 100,000 live births, respectively).
Almost half of all cases were reported from California (680 cases; 47.2%).
The mean age at onset was 13 weeks (range, 1 to 63 weeks).

Five cases of infant botulism caused by C baratii type F have been identified; the youngest patient was just 38 hours old at presentation (see References: Barash 2005).
A review of charts of infant patients in California who were treated with Human Botulism Immune Globulin on the basis of clinical presentation but did not ultimately have laboratory-confirmed botulism (32 of the 681 who were treated) demonstrated that these patients fell into five categories: spinal muscular atrophy type I (5 patients), metabolic disorders (8 patients), infectious diseases (3 patients), miscellaneous (7 patients; includes Miller Fisher variant of Guillain-Barre syndrome, neuroblastoma stage III, and cerebral infarctions, among others), and probable infant botulism lacking laboratory confirmation (9 patients) (see References: Francisco 2007).

Adult Intestinal Toxemia Botulism

The pathogenesis of intestinal botulism in adults is similar to that of infant botulism. Disease is caused by ingestion of C botulinum spores, with subsequent colonization of the gastrointestinal tract. Spores germinate and produce toxin, which is then absorbed into the circulation.
Only a few cases have been recognized, and most have occurred postoperatively or in adults with underlying pathology of the gastrointestinal tract such as Crohn's disease (see References: Bartlett 1986, Chia 1986, Griffin 1997, Shapiro 1998).
Several cases caused by type F toxin produced by C baratii have been reported to the CDC (see References: McCroskey 1991) and cases caused by C butyricum producing type E toxin also have been recognized (see References: Fenicia 1999).

A recently published review of type F adult botulism in the United States between 1981 and 2002 demonstrated the following findings (see References: Gupta 2005):

Thirteen cases of adult type F botulism were reported to CDC during the study period, representing 1% of US cases.
A toxigenic C baratii organism producing type F toxin was isolated in 8 (80%) of 10 positive stool cultures. Type F toxin was identified in serum for nine of the cases.
In 5 (42%) of 12 cases, a history of gastrointestinal disease or an invasive gastrointestinal procedure was present before illness onset. Also in 5 (42%) of 12 cases, antimicrobials were reportedly taken before illness onset. A possible food source was only identified in one instance.

Inhalational Botulism

Disease is caused by inhalation of aerosolized preformed botulinum toxin with subsequent absorption through the lungs into the circulation.
Three cases of inhalational botulism were reported in 1962 in veterinary technicians in Germany who were working with aerosolized botulinum toxin in animals (see References: Arnon 2001). Symptoms occurred about 72 hours after exposure.
Inhalational disease also has been produced experimentally in animals. One study, involving primates, demonstrated that illness occurred 12 to 80 hours after exposure (see References: Franz 1993). Another study, involving mice, demonstrated that following inhalational challenge, the maximum concentration of botulinum toxin in blood occurred at 2 hours postexposure (see References: Park 2003).
A more recent mouse study characterized the pathological consequences of inhalational botulinum toxin exposure in mice given prophylactic pentavalent (ABCDE) toxoid. The authors found that the mice sustained severe histopathological lung damage despite protection from the lethal neurotoxic effects. Signs included "thickening of the alveolar septa and perivascular areas with a generalized spreading interstitial edema and a moderate intra-alveola/intrabronchiola hemorrhage" (see References: Taysse 2005). These findings suggest a direct toxic affect of botulinum toxin on lung tissues; however, more research is needed to better define this potential effect.

Iatrogenic Botulism

Iatrogenic botulism is caused inadvertently by injection of botulinum toxin for therapeutic or cosmetic reasons (see References: Sobel 2005).
Four cases of iatrogenic botulism occurred in December 2004 in Florida following cosmetic injection with a botulinum toxin that was not approved for use in humans (see Dec 15, 2004, CIDRAP News story). The injections contained much higher concentrations of botulinum toxin than the FDA-approved product Botox. A research firm in Arizona sold the raw botulinum toxin to healthcare practitioners as a Botox substitute.

Botulinum Toxin as a Biological Weapon

Past efforts to weaponize botulinum toxin include the following:

The United States produced botulinum toxin as a potential biological weapon beginning in World War II; however, the US offensive biological weapons program ended after the 1972 Biological and Toxin Weapons Convention (BTWC).
The former Soviet Union conducted research on use of botulinum toxin as a biological weapon as late as the early 1990s, despite having signed the BTWC.
At the time of the Gulf War, Iraq had produced 19,000 L of concentrated botulinum toxin, some of which was loaded into military weapons (see References: Zilinskas 1997).
The Japanese cult Aum Shinrikyo attempted to use aerosolized botulinum toxin in Japanese cities on at least three occasions between 1990 and 1995. Fortunately, these efforts were not successful.

Modes of dissemination for toxin used as a biological weapon:

Deliberate contamination of food or beverages with botulinum toxin is the most likely route of dissemination. Contamination of a commercially produced and widely distributed food or beverage product could result in a high number of casualties and fatalities across the country. In addition, such a bioterrorist act would produce severe civic disruption, economic loss, and social anxiety. Any food or beverage item that is not heat-processed at 85°C (185°F) for 5 minutes prior to consumption or is potentially contaminated following sufficient temperature processing must be considered a possible vehicle for botulinum toxin. For example, typical temperatures employed for pasteurization of commercially available beverage products will not sufficiently denature all botulinum toxin in the product. Mathematical modeling suggests that 1 g of botulinum toxin added to commercially distributed milk consumed by 568,000 people could result in 100,000 cases of botulism (see References: Wein 2005). Ten grams of toxin added to the same quantity of milk could result in over 500,000 cases in the exposed population.
Dispersion of aerosolized toxin also is possible. Aerosolized particles of botulinum toxin are approximately 0.1 to 0.3 mcm in size (see References: Shapiro 1997). Experts have estimated that 1 g of aerosolized botulinum toxin could kill up to 1.5 million people (see References: Shapiro 1997).
Although contamination of a water supply is feasible, this approach is unlikely since a large amount of toxin would be needed to initially contaminate water. Also, toxin is naturally inactivated in fresh water within 3 to 6 days, and toxin is rapidly (within 20 minutes) inactivated by standard potable water treatment (see References: Siegel 1993). A 2005 study found that two of seven small-scale water purification devices tested were able to effectively eliminate botulinum toxin from water. Those based on filtration (pore size 0.2 to 0.4 mcm) or irradiation from a UV-lamp (254 nm) failed toremove the toxin from inoculated water. Reverse osmosis and experimental sand filtration effectively eliminated the toxin (see References: Horman 2005).

The following features of a botulism outbreak would suggest deliberate toxin release (see References: Arnon 2001).

An outbreak involving a larger number of cases than previous outbreaks
An outbreak caused by an unusual toxin type (ie, C, D, F, or G) or an outbreak involving type E toxin without an apparent aquatic source
Multiple simultaneous outbreaks with or without an apparent source
For aerosol release, cases would not have a common food exposure but would have been in a common geographic location during the week before symptom onset

Use of Therapeutic Botulinum Toxin

Patients with a range of spastic or autonomic neuromuscular disorders may benefit from small amounts of purified botulinum toxin injected into affected muscles (see References: Schantz 1992): Examples include:

Spasmodic torticollis
Strabismus
Blepharospasm
Laryngeal dystonia
Focal dystonias of the hand
Limb spasticity
Hemifacial spasm
Cerebral palsy
Migraine headache
Hyperhydrosis
Post-stroke spasticity

Purified botulinum toxin type A (Botox, produced by Allergan, Inc) was originally approved by the Food and Drug Administration (FDA) in 1989 to treat blepharospasm and strabismus and was approved in December 2000 to treat cervical dystonia (see References: Allergan, Inc). In a recent report, botulinum type A toxin injections produced functional improvement in spastic diplegia among children with cerebral palsy (see References: Bjornson 2007). A retrospective study of 261 children with cerebral palsy (average age, 8 years 4 months) treated with botulinum toxin showed that adverse-event rates for higher-dose regimens (15 to 20 U/kg. 3.5%; 20 to 25 U/kg, 8.6%) were not significantly different than rates for standard regimens (5 to 10 U/kg, 3.9%; 10 to 15 U/kg, 7.6%). No patient developed botulism after treatment, and all doses were associated with a significant increase in passive range of motion using the Tardieu scale (see References: Willis 2007).
Purified botulinum toxin type B (Myobloc, produced by Elan Pharmaceuticals, Inc) was approved by the FDA in 2000 for treatment of patients with cervical dystonia to reduce the severity of abnormal head position and neck pain (see References: FDA: Myobloc labeling information).
In April 2002, the FDA approved use of botulinum toxin type A to temporarily improve the appearance of frown lines between the eyebrows (see References: Allergan, Inc; FDA: Botox Cosmetic labeling information).
Therapeutic botulinum toxin contains about 0.3% of the estimated lethal human inhalational dose and only 0.005% of the estimated lethal human oral dose; therefore, this form of toxin is not likely to be used as a bioterrorist weapon (see References: Arnon 2001).
An unlicensed, highly concentrated preparation of botulinum toxin caused botulism in 4 adult patients undergoing cosmetic procedures. Affected patients may have received doses 2,857 times the estimated human lethal dose by injection. Pretreatment serum levels in 3 of the 4 patients were between 21 to 43 times the estimated human lethal dose (see References: Chertow 2006).
In rare instances, circumstances can mask the source of botulism. In botulism occurring in two sisters who had symptoms about 3 weeks apart, one presented with suspected botulism after receiving an injection of botulism toxin for cosmetic purposes. However, laboratory testing detected that type B rather than type A toxin was used in the cosmetic procedure. The patient's sister subsequently was evaluated for botulism and also tested positive for type B toxin. The true cause for both cases was type B toxin in home-preserved meat that both had eaten (see References: Fota-Markowska 2007).
Severe systemic effects have been noted on rare occasion after therapeutic use (see References: Borodic 1998). As noted above, iatrogenic botulism can occur from therapeutic or cosmetic use of botulinum toxin.

Clinical Features

Botulism is characterized by acute afebrile descending symmetric paralysis. Recovery occurs over weeks to months and often requires extensive supportive care.
Disease generally begins with evidence of cranial nerve dysfunction and then progresses to muscle weakness (proximal muscle groups are affected first and may be more severely involved).
Severity of disease ranges from mild cranial nerve dysfunction to complete flaccid paralysis. Paralysis of pharyngeal or respiratory muscles may result in the need for prolonged mechanical ventilation.

Severity of disease correlates with the amount of toxin absorbed into the circulation.
Several studies have shown that a shorter incubation period correlates with more severe disease (see References: MacDonald 1985: Type A botulism from sauteed onions; Tacket 1984). Similarly, a study of botulism cases in Japan revealed that patients who had shorter incubation periods had a significantly higher risk of death (see References: Nishiura 2007).
Disease caused by toxin type A tends to be more severe than disease caused by toxin type B or E (see References: Shapiro 1998).
Among more than 200 patients in an outbreak in Thailand, respiratory failure was less likely to develop in those who did not manifest nausea or vomiting and did not have urinary retention requiring catheterization. Nausea or vomiting and any cranial neuropathy with urinary retention or difficulty swallowing were symptoms most predictive of respiratory failure (see References: Wongtanate 2007).

Death can result from airway obstruction or paralysis of respiratory muscles. Death also can result from complications related to prolonged ventilatory support and intensive care, such as aspiration pneumonia and other infectious conditions.

Before mechanical ventilation was widely available, the case-fatality rate was about 60% (see References: Shapiro 1998).
The case-fatality rate currently is low owing to adequate supportive care; overall the rate is 5% to 10% for foodborne disease and somewhat higher for wound botulism (see References: Shapiro 1998, Werner 2000).
In the event of a mass exposure (such as a bioterrorism attack), clinical resources could be overwhelmed rapidly and the case-fatality rate could be much higher.
A recent retrospective study of hospitalized foodborne botulism cases in the Republic of Georgia, 1980-2002, found that patients with shortness of breath and impaired gag reflex and without diarrhea were 23 times more likely to die than were patients without this syndrome (see References: Varma 2004). In this case series, the incubation period was similar among those who died and those who survived, as was the likelihood of receiving antitoxin.

Clinical features are outlined in the table below.

Clinical Features of Foodborne and Wound Botulism
Note: Information presented is for foodborne and wound botulism; infant botulism is not included, since that condition is distinct from what would be expected in a bioterrorism attack. The presenting features of inhalational botulism likely would be comparable to those of foodborne and wound botulism.
Characteristic	Features
Incubation period*	—Dependent on level of toxin exposure —For foodborne botulism, 2 hr–8 days —For wound botulism, 4-14 days —Unknown for inhalational botulism; estimated to be 24-36 hr; the only three reported cases in humans had an incubation period of 72 hr
Symptoms (compiled from reports of foodborne botulism outbreaks caused by toxin types A, B, and E)†	—Nausea (88%)‡ —Dry mouth (82%) —Blurred vision (78%) —Dysphonia (76%) —Dysphagia (75%) —Weakness (72%) —Fatigue (69%) —Dyspnea (65%) —Dysarthria (63%) —Double vision (60%) —Dizziness (56%) —Vomiting (52%)‡ —Constipation (related to autonomic dysfunction) (45%) —Sore throat (40%) —Abdominal cramps or abdominal pain (40%)§ —Diarrhea (35%)‡ —Paresthesias (29%)
Signs (compiled from cases of types A and B botulism reported to CDC in 1973 and 1974)§	—Alert mental status (90%) —Weakness of upper extremities (75%) —Ptosis (73%) —Weakness of lower extremities (69%) —Extraocular muscle weakness (65%) —Diminished gag reflex (65%) —Facial nerve dysfunction (63%) —Dilated or fixed pupils (44%) —Diminished or absent deep tendon reflexes in affected groups (40%) —Nystagmus (22%) —Ataxia (17%) —Other considerations: ~Patients generally afebrile ~Mental status generally intact, although patients may appear lethargic or have difficulty communicating because of bulbar dysfunction ~Sensory exam generally normal
Laboratory features	—Normal CSF glucose, protein, cell count —Normal CBC —Normal imaging of brain and spine (ie, CT scan or MRI) —Characteristic EMG findings**: ~Incremental response (facilitation) to repetitive stimulation (not always present and often seen only at 50 Hz) ~Short duration of motor unit potentials (MUPs); polyphasic MUPs ~Decreased amplitude of compound muscle action potentials (CMAPs) after a single nerve stimulus (most prominent in proximal muscle groups) ~Normal sensory nerve function ~Normal nerve conduction velocity (motor and sensory)
Complications	—Respiratory failure (which may require prolonged ventilatory support); in some outbreak settings, up to 30%-40% of patients required mechanical ventilation —Aspiration pneumonia (among patients with respiratory failure)†† —Residual fatigue, dry mouth or eyes, dyspnea on exertion up to several years after initial presentation‡‡
Case-fatality rate§§	—5%-10% for foodborne botulism*** —15%-44% for wound botulism†††
Abbreviations: CSF, cerebrospinal fluid; CT, computed tomography; MRI, magnetic resonance imaging; CBC, complete blood count. See References: Arnon 2001, Franz 1997. †The percentages were derived from compiling information available from published reports of large foodborne outbreaks caused by toxin type A, B, or E. The number of cases in each denominator ranged from 30 to 180. See References: Angulo 1998; from s 1985: Type A botulism; St Louis 1988; Wainwright 1988; Weber 1993. ‡Gastrointestinal symptoms are uncommon in patients with wound botulism (see References: Merson 1973) and likely would be uncommon in the setting of inhalational exposure. §See References: Hughes 1981. See References: Cherington 1998, Maselli 2000. ††See References: Schmidt-Nowara 1983. ‡‡See References: Mann 1981, Mann 1983, Wilcox 1989. §§Before mechanical ventilation was widely available, the case-fatality rate was much higher (about 60%). In the setting of a mass exposure, where intensive-care resources could rapidly be overwhelmed, the case-fatality rate may be higher than that currently observed. **See References: Shapiro 1998. †††See References: Merson 1973, Shapiro 1998, Werner 2000.

Pediatric Considerations

Most pediatric cases of botulism occur in infants (ie, infant botulism), although foodborne and wound botulism also can affect the pediatric population.
In the event of an aerosol release of botulinum toxin, children may be at an even greater level of risk than adults, since children have a higher number of respirations per minute and consequently could have an increased level of exposure to toxin (see References: AAP 2000).
Signs and symptoms of botulism in children following a bioterrorist attack (ie, aerosol or foodborne exposure) would be similar to those seen in adults.
Assuring adequate intensive care resources for the pediatric population in the event of a bioterrorism attack involving an agent such as botulinum toxin should be an important priority in bioterrorism preparedness planning.

Differential Diagnosis

Differential Diagnosis of Botulism
Note: This differential diagnosis applies to botulism in adults and older children; infant botulism is not included, since that condition is distinct from what would be expected during a bioterrorism attack.
Condition	Features that distinguish each condition from botulism*
Guillain-Barre syndrome (GBS) (particularly Miller Fisher variant)	—Classic GBS results in ascending paralysis —Miller Fisher variant may be descending and may have pronounced cranial nerve involvement; it usually includes a triad of ophthalmoplegia, ataxia, and areflexia (5% of GBS cases are of the Miller Fisher variant)† —Abnormal CSF protein 1-6 wk after illness onset (although may be normal early in clinical course) —Paresthesias commonly occur (often stocking/glove pattern) —EMG shows abnormal nerve conduction velocity; facilitation with repetitive nerve stimulation does not occur (as with botulism) —History of antecedent diarrheal illness (suggestive of Campylobacter infection, which accounts for about one third of GBS cases) —Outbreaks of GBS do not occur (unlike botulism)
Myasthenia gravis	—Dramatic improvement with edrophonium chloride (ie, a positive Tensilon test), although some botulism patients may exhibit partial improvement following administration of edrophonium chloride (ie, a borderline Tensilon test) —EMG shows decrease in muscle action potentials with repetitive nerve stimulation
Tick paralysis‡	—Ascending paralysis —Paresthesias are common —Careful examination reveals presence of tick attached to skin —Recovery occurs within 24 hr after tick removal —EMG shows abnormal nerve conduction velocity and unresponsiveness to repetitive stimulation —Usually does not involve cranial nerves
Lambert-Eaton syndrome	—Commonly associated with carcinoma (often oat cell carcinoma of lung) —Although EMG findings are similar to those in botulism, repetitive nerve stimulation shows much greater augmentation of muscle action potentials, particularly at 20-50 Hz —Increased strength with sustained contraction —Deep tendon reflexes often absent; ataxia may be present —Usually does not involve cranial nerves
Stroke or CNS mass lesion	—Paralysis usually asymmetric —Brain imaging (CT or MRI) usually abnormal —Sensory deficits common —Altered mental status may be present
Poliomyelitis	—Febrile illness —CSF shows pleocytosis and increased protein —Altered mental status may be present —Paralysis often asymmetric
Paralytic shellfish poisoning or ingestion of puffer fish	—History of shellfish (ie, clams, mussels) or puffer fish ingestion within several hours before symptom onset —Paresthesias of mouth, face, lips, extremities commonly occur
Belladonna toxicity	—History of recent exposure to belladonna-like alkaloids —Fever —Tachycardia —Altered mental status
Aminoglycoside toxicity	—History of recent exposure to aminoglycoside antibiotics —More likely to occur in the setting of renal insufficiency —Most commonly seen with neomycin —Most commonly associated with other neuromuscular blocking agents such as succinylcholine and paralytics
Other toxicities (hypermagnesemia, organophosphates, nerve gas, carbon monoxide)	—History of exposure to toxic agents —Carbon monoxide toxicity: altered mental status may occur, cherry-colored skin —Hypermagnesemia: history of use of cathartics or antacids may be present, elevated serum magnesium level —Organophosphate toxicity: fever, excessive salivation, altered mental status, paresthesias, miosis
Other conditions	—CNS infections (particularly brainstem infections) —Inflammatory myopathy —Hypothyroidism —Diabetic neuropathy —Viral infections —Streptococcal pharyngitis (pharyngeal erythema and sore throat can occur in botulism owing to dryness caused by parasympathetic cholinergic blockade)
Abbreviations: CSF, cerebrospinal fluid; EMG, electromyogram; CT, computed tomography; MRI, magnetic resonance imaging. *See References: Arnon 2001, Campbell 1981, Cherington 1998, Werner 2000. †See References: Sobel 2005, Dorr 2006. ‡ See References: Felz 2000.

Laboratory Diagnosis

Specimen Collection and Transport

Specimen collection and transport procedures for testing related to diagnosing botulism are outlined in the following table.

Collection and Transport of Laboratory Specimens for the Diagnosis of Botulism
Note: A list of patient medications should accompany specimens, since some medications may be toxic to mice and can be removed by dialysis before testing is performed.*
Specimen	Clinical Indication	Collection and Transport
Serum	Intentional release, foodborne botulism, autopsy specimens	—Collect >20 mL whole blood before administration of antitoxin using red-top or separator tube (no anticoagulant) —Ship >10 mL serum at 4^oC —Do not ship whole blood, which tends to become hemolyzed during transit —Notify testing lab if patient has received "stigmine drugs" or a Tensilon test —Keep specimen refrigerated at all times
Serum	Wound botulism (critical specimen for confirmation)	—Collect 30 cc whole blood (before antitoxin administration) —Ship at 4^oC —Sera submitted for toxin detection should not be hemolyzed —Notify testing lab if patient has received "stigmine drugs" or a Tensilon test —Keep specimen refrigerated at all times
Wound/tissue	Wound botulism	—Collect exudate, tissue, or swabs —Ship at room temperature in anaerobic transport system
Stool, enema fluid, intestinal fluid	Intentional release, foodborne botulism, infant botulism, wound botulism†	—Obtain 10-50 g of stool (as little as "pea-size" for infant botulism); transport at 4^oC —Enema fluid (20 cc) can be collected as an alternative to stool, using minimal amount of sterile nonbacteriostatic water; ship at 4^oC —Intestinal fluid collected at autopsy (20 cc); ship at 4^oC
Gastric fluid, vomitus	Foodborne botulism, intentional release	—Collect within 72 hr of symptom onset —Obtain 20 cc of vomitus; ship at 4^oC —Obtain 20 cc of gastric fluid (living cases or at autopsy); ship at 4^oC
Specimens to collect at autopsy	Intentional release, foodborne botulism, infant botulism	—Serum, according to methods outlined above —Contents from different sections of small and large intestines (10 g per sample in separate containers) —Gastric contents as indicated, according to methods outlined above —Tissue samples as indicated, according to methods outlined above
Food samples (epidemiologically implicated)	Intentional release, foodborne botulism, infant botulism	—Obtain 10-50 g of implicated or suspect food; ship at 4^oC in original container —Place individually in leak-proof sealed transport devices
Nasal swab	Intentional release‡	—Obtain anaerobic swab; ship at room temperature
Environmental sample	Intentional release, infant botulism	—Collect as appropriate: ~Environmental swab; ship at room temperature ~Soil (50-100 g) ~Water (>100 mL)
*See References: Arnon 2001. †A wound may not be the actual source of infection/intoxication. ‡Toxin may be present on nasal mucosa for up to 24 hr after inhalational exposure (see References: Franz 1997). Adapted from the following sources (see References): ASM 2003: Sentinel laboratory guidelines for suspected agents of bioterrorism: botulinum toxin; CDC 1998: Botulism in the United States, 1899-1996; CDC: Specimen selection table.

Guidelines have been published for packing and shipping of infectious substances, diagnostic specimens, and biological agents from suspected bioterrorism (see References: ASM: Sentinel laboratory guidelines for suspected agents of bioterrorism: packing and shipping infectious substances, diagnostic specimens, and biological agents). C botulinum is classified under World Health Organization (WHO) risk group 4. Cultures that are reasonably suspected to contain C botulinum must be transported as "infectious substances." International Air Transport Association (IATA) rules require training of all individuals involved in the transport of dangerous goods, including infectious substances. Once botulinum toxin is identified, samples may be regulated as select agents and subject to additional transport requirements (see below). Chain of custody should be documented for material that may constitute evidence of criminal activity.

Laboratory Biosafety and Biosecurity

Botulinum toxin and Clostridium species that produce botulinum toxin are classified as select agents and therefore are regulated under 42 CFR part 73 (Possession, Use, and Transfer of Select Agents and Toxins), which was published as an Interim Final Rule in the Federal Register on December 13, 2002 (see References: HHS 2002). As specified in the Public Health Security and Bioterrorism Preparedness and Response Act of 2002, 42 CFR part 73 provides requirements for laboratories that handle select agents (including registration, security risk assessments, safety plans, security plans, emergency response plans, training, transfers, record keeping, inspections, and notifications). These new requirements went into effect on February 7, 2003, and override earlier government requirements regarding possession and transfer of select agents. For more information about CDC's Select Agent Program, see References: CDC: Select Agent Program. In addition, CDC recently published additional guidelines for enhancing laboratory security for laboratories working with select agents (see References: CDC 2002: Laboratory security and emergency response guidance for laboratories working with select agents).
C botulinum toxin detection should be performed only by trained individuals at laboratory response network (LRN) reference or higher laboratories.
Sodium hypochlorite (0.1%) or sodium hydroxide (0.1 N) inactivate the toxin and are recommended by CDC for decontaminating work surfaces and spills of cultures or toxin (see References: CDC 1999: Biosafety in microbiological and biomedical laboratories).