Clostridium botulinum (Botulism)

Organisms and Toxins

C. botulinum represents a heterogeneous group of gram-positive, anaerobic, spore-forming bacilli, individual strains of which can produce 1 of 8 serologically distinctive neurotoxins (BoNT) types A through H. ^, (A ninth botulinum toxin, named serotype “X”, identified in GenBank through its nucleic acid sequence, has not caused illness. ) More than 99% of US cases of infant botulism are caused by BoNT/A or BoNT/B. Neurotoxigenic Clostridium butyricum is responsible for almost all rare BoNT/E cases, ^, and neurotoxigenic Clostridium baratii is responsible for rare BoNT/F cases that occur predominantly in neonates. Clinical cases occasionally result from C. botulinum strains termed “bivalent” that can produce two toxins (e.g., BoNT/Ba or BoNT/Af). Amplified fragment length polymorphism (AFLP) analysis of a national collection from cases and related environmental samples identified 154 clades (with 52% of isolates clustered in 2 predominantly California clades) and co-location in clades of individual case isolates and their associated environmental isolates (e.g., home vacuum cleaner dust). A single clade could contain isolates that caused a wide spectrum of case severity.

Botulinum toxins are similarly structured, 150 kD, single di-chain polypeptides joined by a single disulfide bond and are the most potent toxins known; 10 ¹¹ molecules (25 ng) reaching neuromuscular junctions are enough to cause clinical botulism. BoNT/A is the most avidly binding and potent toxin and is associated with the most severe disease. BoNT is acid and digestive enzyme resistant but can be destroyed by heating of food to the boiling point (100°C) for 10 minutes. Certain environmental conditions, such as anaerobic conditions, pH >4.6, warm temperature (>39°C), and high moisture content, promote outgrowth of spores to vegetative forms and production of toxin in food. Spores of C. botulinum are heat resistant, withstanding 100°C for hours; food must be heated under pressure to temperatures of 116°C to destroy spores.

Pathophysiology

After systemic absorption, BoNT is concentrated at the plasma membrane of presynaptic peripheral cholinergic neurons. The toxin is then internalized via membrane recycling, at which point it becomes inaccessible to antitoxin treatment, and subsequently toxin enters into the cytosol. The active portion of botulinum toxin is a Zn ⁺⁺ -protease that causes flaccid paralysis because it cleaves one or more proteins of the so-called SNARE complex that enable acetylcholine-containing synaptic vesicles to fuse with the terminal neuronal membrane. With intoxication at 70% or more junctions, clinical voluntary muscular and autonomic function is impaired; with intoxication at 90%, diaphragmatic muscular function is impaired. Recovery results from regeneration of the demyelinated motoneuron twigs that then induce formation of new motor endplates.

In the absence of an identified source of swallowed C. botulinum spores (e.g., honey), it is assumed that almost all US infant botulism cases result from unrecognizable ingestion of spores transported by microscopic airborne dust. Although the recommendation not to feed honey to infants reduced US honey-associated cases, honey still is a significant source of spores for cases in other countries. The gastrointestinal tract of children and adults is resistant to outgrowth of swallowed C. botulinum spores, but in infants <12 months of age such outgrowth occasionally occurs and results in illness. Age-related vulnerability is related at least partially to lack of competitive flora and possibly to pH and motility. Experimentation and observation in an asymptomatic mouse model provided compelling evidence of the protective importance of a competitive gut flora that included (1) age-limited (7–13 days) natural susceptibility to intestinal colonization by C. botulinum spores; (2) ongoing susceptibility in germ-free adults; (3) acquisition of resistance in germ-free adults after placement in a normal cohort colony for 3 days; (4) resistance after establishment of selective, limited intestinal flora; and (5) loss of resistance after antibiotic suppression of intestinal flora. ^,

The necessary components of resistance are undoubtedly complex, but Bacteroides, Lactobacillus, and other Clostridium spp. are inhibitory to C. botulinum. Stool flora of 7 infants delineated at first manifestations of botulism showed a relative paucity of Bacteroides and Lactobacillus spp. The putrefactive intestinal flora of the infant who is fed formula and the fermentative flora of the infant who is fed human milk would be expected to be inhibitory. Epidemiologic data support the theory that the period of permissiveness of the gut of the infant fed formula is brief (i.e., competing flora is established rapidly ) and that of the infant fed human milk is more extended and is especially heightened during perturbation of flora during weaning. Infant botulism is more often associated with feeding of formula in infants <1 month of age, but by 5 months of age almost 100% of cases occur in infants being fed human milk. ^, The natural occurrence of infant botulism in horses (shaker foal syndrome) and ducks (limberneck) suggests a similar age-related vulnerability to colonization in animals. The rare cases of intestinal toxemia botulism in older persons almost always have been associated with an underlying gut anomaly or a presumed alteration of gut flora, pH, or motility by surgical procedures and drugs. ^,

Environmental Exposure

Soils and marine sediments worldwide are the natural habitat of C. botulinum. Spores can be transported by airborne microscopic dust and likely reach foodstuffs, blossoms (hence honey), and oral secretions in this way. C. botulinum strains producing BoNT/A, BoNT/B, and BoNT/F generally are found in areas of low rainfall and moderate temperature, whereas those producing BoNT/D and BoNT/E are associated with water. ^, The most recent soil survey in the US done in 1975 found that 23% of samples contained C. botulinum spores producing BoNT/A, BoNT/B, BoNT/C, BoNT/D and BoNT/E. In the US most C. botulinum type A strains are found west of the Mississippi River, where the Rocky Mountain cordillera rises out of the Great Plains ( Fig. 189.2 ). BoNT/B spores have a more general east-west distribution but are mostly confined to areas between 35–55 degrees of north latitude and occur in soil with high organic content. BoNT/E spores are found in marine life and in sediments around fresh water from the Pacific northwest and the Great Lakes region of the US, while C. botulinum BoNT/F spores have been isolated from marine sediment from the Pacific coast and from crabs in Chesapeake Bay. The reasons for the geographic distribution of C. botulinum that produce different neurotoxins are unknown. Soil sampling in Argentina showed that 23% of samples were positive for a variety of C. botulinum types; the distribution was uneven, with a higher prevalence in cultivated soil.

The geographic distribution of C. baratii– producing BoNT/F in the US has become evident through the rare cases of type F infant botulism and adult toxemia botulism that it causes. Infant and adult cases have been recognized in states that border the Atlantic (MA, VA) and Pacific (CA, OR, WA) oceans, the Canadian (ID, WA) and Mexican (NM, TX) borders, and also in inland (CO, IA, WI) areas. ^, Foodborne C. baratii type F botulism has occurred in Canada, France, and California. ^, The geographic distribution of C. butyricum —producing BoNT/E likewise has emerged as a cause of infant botulism, foodborne botulism, and adult toxemia botulism in Italy, Ireland, India, China, and the US. ^,

Botulism case density and toxin types in humans and animals mirror the geography and density of C. botulinum spores in the environment. Although infant botulism undoubtedly is under-recognized, cases have been reported from all US states and, with the November 2020 report of the first laboratory-confirmed case in Africa, from all inhabited continents. The highest reported incidence of disease in the US is the Delaware–Pennsylvania–New Jersey region arching Philadelphia, followed by California and Utah. Argentina, Australia, Canada, Italy, and Japan, respectively, have reported the next largest numbers of cases.

Among US cases of infant botulism overall, approximately one-half are caused by BoNT/A and one-half by BoNT/B, with rare cases usually caused by BoNT/F or BoNT/E. Cases in California are divided between BoNT/A and BoNT/B, ^, but BoNT/B accounts for >90% of cases in Pennsylvania. One or more samples from an affected infant’s environment, such as yard soil, vacuum cleaner dust, crib, father’s shoes, or consumed honey, may yield C. botulinum spores, always of matching toxin type. ^, Samples do not contain preformed toxin.

Unusual epidemiologic findings also suggest uneven distribution of spores within limited geographic areas. In one Pennsylvania study, residences of 83% of the 53 patients with botulism formed an arc around Philadelphia, with remarkable sparing of the city itself, counties surrounding the arc, and the western parts of the state. In Colorado, 3 of only 6 cases reported between 1977 and 1985 were from the same town of 800 people; 2 of 3 infants shared a crib (years apart), and all samples from their environments (including the crib) yielded C. botulinum –producing BoNT/A. Nearly all of 39 infants with botulism in Utah over a 20-year period resided within a 70-mile metropolitan corridor including Salt Lake City, Ogden, and Provo.

Many parents of infants with botulism report nearby soil disruption (e.g., home or commercial construction or road work) or a dusty “event” before the onset of the infant’s illness. ^{, , , ,} In Pennsylvania, 50% of the infants’ fathers had occupations that required daily contact with soil; on testing, spores were found on work shoes or at a work site in most cases. C. botulinum organisms with the same toxin type (A or B) as the case were found in yard soil and untreated cistern drinking water in Australia, in untreated well water in Japan, and in yard soil and household vacuum cleaner dust in Finland and California. ^,

In the 1970s and 1980s, honey and light and dark corn syrups were found to harbor C. botulinum spores and tested positive in 0.5%–20% of samples. ^, Following changes in corn syrup production, the FDA laboratory found in 1991 that samples tested negative for C. botulinum . Corn syrup has not been a proven vehicle for spores in any case of infant botulism and can be given to young infants safely. Consumption of honey is a clear risk factor for infant botulism (odds ratio [OR] = 9.8 in 1 study), but because of caregiver avoidance, honey accounts for <5% of cases currently in the US. A case of infant botulism BoNT/B in the UK was linked to isolation of C. botulinum spores from powdered infant formula. ^, Nonbotulinum clostridial spores have been found in commercial powdered formula in the US, a finding suggesting potential vulnerability of this foodsource. Some infant botulism cases have been linked to unusual environmental sources; for example, 2 BoNT/E cases in Ireland were related to pet terrapins or terrapin feed, a BoNT/B case was related to chamomile and fennel infusion in Spain, and a BoNT/A(B) case in Japan was linked to untreated well water. ^,

Host Susceptibility

Infant botulism has a striking age range for susceptibility and age-related associations with feeding type and bowel movement frequency. Approximately 90% of infant botulism cases occur between 2 weeks and 6 months of age, and 99.9% of patients are ≤12 months old. In the largest series yet reported, the youngest patient had illness onset at 38 hours of age, ^, while the two oldest patients (0.1% of cases) were 14 and 17 months old at onset. However, median age of onset was significantly younger in infants fed formula (1.7 months) than in infants fed human milk (4.4 months). A slower intestinal transit time (≤1 bowel movement/day) was significantly associated with acquisition of illness regardless of age. The median age at onset of BoNT/F caused by C. baratii and BoNT/E caused by C. butyricum is <14 days of age, which is significantly younger than the median age of 3.8 months for BoNT/A and 2.9 months for BoNT/B.

In bivariate analyses breastfeeding was associated with hospitalization from infant botulism, present in 66%–100% of cases in published series. ^, Multivariate analysis of a large (n=159) case-control study found no association with breastfeeding for patients ≤2 months old but did find a strong association for patients >2 months old. Long and colleagues found that 66% of breastfed infants with botulism who had received human milk exclusively had introduction of the first nonhuman food substances (cereal, formula, fruit) within 2–4 weeks before the onset of symptoms. Many sociodemographic factors bivariately associated with infant botulism probably are markers for breastfeeding. Additional multivariate-determined risk factors are age dependent. For infants ≤2 months old, cesarean delivery, birth order >1, and residing in a windy area were significant, while for infants >2 months of age, having dust exposure was additionally significant. Fig. 189.3 is a schematic summary of environmental factors, host factors, and pathophysiologic events in infant botulism.

Immunity

C. botulinum is not part of the normal intestinal microbiome. No apparent natural acquisition of immunity to the organism or its toxin occurs. Infants recovering from botulism continue to excrete C. botulinum organisms and toxin in stool for weeks to months, a consideration post-discharge. ^, Some infants develop demonstrable specific serum antibody to toxin, the extent and duration of protective effect of which is not known. Cases of relapsing hypotonia within weeks of hospital discharge in untreated infants have been reported, but on careful examination these appear to be instances of discharge before adequate, around-the-clock strength had returned.

Classic Manifestations

Based on last exposure to C. botulinum spore-containing honeys, the incubation period is estimated to be 3–30 days. However, knowledge of the incubation period does not assist in clinical management or prevention, as the environment via unavoidable airborne microscopic dust is the likely source of the causative spores.

Onset of illness almost always begins as constipation, although often this is clear only in retrospect. Poor feeding, weakness, drooling (inability to swallow secretions), shallow breathing, and loss of tone (floppiness) and head control are common presenting complaints. A descending symmetric paralysis typically progresses over 6 hours to 20 days (median, 4.2 days) before medical recognition or hospitalization. Cranial nerves are affected first, with a subtly flat facial expression, quiet voice, and less avid suck—usually noted only by parents. The relative frequency of symptoms and signs is shown in Table 189.1 . ^{, , ,} Young infants and those fed formula can have rapidly progressive paralysis. A small percentage of cases of sudden unexpected death in young infants have been shown to be rapidly progressive cases of infant botulism. ^,

The physical examination is most notable for the infant’s quiet, still demeanor. Poor head control and hypotonia (inability of sustained postural control and movement against gravity) are universal. Although frequently misinterpreted as “lethargic” and “septic,” the infant is alert but rendered (by motor weakness) unable to smile, regard, vocalize, or move; skin color is normal or robust ( Fig. 189.4 ). The diagnosis should not be made in the absence of head lag and cranial nerve palsies as first neurologic manifestations; swallowing function is most affected, and corneal reflexes are most preserved. Disconjugate gaze precedes ophthalmoplegia. Pupils become dilated or sluggishly reactive, especially when tested repeatedly without interruption over 1–2 minutes in a dark room. Ptosis may not be evident unless the eyelids have to work against gravity, but careful head and neck support may be needed. Hyporeflexia is unimpressive early in the course relative to profound hypotonia but progresses over time. Respiratory difficulty is present in approximately 50% of patients at or during hospitalization, but occurrence varies by toxin type, age, and promptness of diagnosis. In one study, approximately 40% of patients with type A intoxication and 60% with type B intoxication required mechanical ventilation.

Autonomic dysfunction is underrecognized. Decreased tearing and salivation can be misinterpreted as dehydration. Erythema of oral mucosa can be misinterpreted as pharyngitis. Blood pressure, heart rate, and skin tone fluctuate frequently. Harlequin flushing can occur. Decreased large intestinal motility, poor anal sphincter tone, and bladder atony are present commonly. No abnormality of sensation occurs.