Clinical Concerns

Typical patients are healthy full-term infants with painless constipation for days to months before onset of weakness. This weakness, poor feeding, a paucity of movement, and seeming lethargy are the most common initial concerns. At presentation, three cardinal features are recognized: symmetrical weakness of bulbar, face, and neck more than appendicular musculature; a nonirritable, awake sensorium; and the absence of fever. Bulbar weakness and poor head support are seen in all patients; ophthalmoparesis and pupillary sluggishness or dilatation are frequent and helpful findings. The face is typically expressionless, drooling is present, and a high-pitched, mewing cry develops that is recognized as characteristic by clinicians in areas of high prevalence. Some infants have a suggestion of fatigability, manifesting as bursts of movement amid a too-immobile restful state. Muscle stretch reflexes are often absent, but their presence does not rule out the diagnosis. Many infants progress rapidly to require assisted ventilation, usually initiated out of concern for a secure airway.

Once infant botulism is recognized and treated, patients with the classic syndrome may need ventilatory assistance or airway support for many months, with a mean of 3 weeks. Patients typically recover in the reverse order of their symptoms’ appearance, with limb movements reappearing before the infant has a competent airway. In exceptional cases, relapse may occur well into the course of recovery.

Pathogenesis

Infant botulism is caused by colonization of the large intestine by toxin-forming C. botulinum or related organisms. C. botulinum is not a normal constituent of gut microflora. Apparently, its absence is due to competition from other microorganisms: adult germ-free mice support large-intestine colonization, but immediately upon exposure to other normal flora, the colonization is cleared. An important exception that may have relevance for human infant botulism is seen during early postnatal development of mice: normal pups of 7 to 13 days can sustain enteric colonization, but older and younger mice cannot. Interestingly, like C. botulinum , other Clostridium species are excluded from the gut of breastfed infants, but the addition of any other food supplement results in the occasional appearance of colonization. This may explain the increased incidence of breastfeeding in older affected infants who, with even a small amount of supplementation from other food sources, may become more susceptible to colonization with C. botulinum .

Clostridia are obligatory anaerobic, Gram-positive, spore-forming rods. In general, they grow best in high pH media, although a combination of temperature, osmolarity, redox potential, the presence of food preservatives, and competing microorganisms affects growth in an interrelated fashion. Each strain produces a single toxin type, although the genes for multiple toxins sometimes exist within a strain, and occasionally a single strain produces two types of toxin. The C. botulinum species is an aggregate of multiple strains, likely the result of convergent evolutionary branches that have in common the production of botulinum toxins. Since this classification of species was developed, other Clostridium species have been found to produce botulinum toxin, thus demonstrating the difficulty of representing the complexity of bacterial phylogeny with a single nomenclature.

Botulinum toxin is the most toxic substance known. The lethal ingested dose is estimated to be less than 1 nanogram/kg body weight. Although abundant toxin is produced in the colon in infant botulism, apparently only a minute amount is absorbed. In the human infant, C. botulinum behaves more like the normal flora than an enteric pathogen; it does not invade the colonic mucosa, and there is no direct cytotoxic effect. The colon must therefore be a good barrier to toxin uptake, but more proximal regions of the gastrointestinal tract are likely not barriers to absorption, given the expression of symptoms with only minute toxin ingestion. The fact that many infants with symptomatic botulism have a history of chronic constipation suggests that diminished intestinal motility is a cause as well as a consequence of infant botulism. These infants may be at higher risk for colonization in the first place, they may be more likely to absorb the toxin once it is made by reflux through the length of the colon and the ileocecal junction, or both possibilities may be true. Breastfeeding is associated with increased colonic motility, which may explain the later onset and less severe expression, though not the increased incidence, of infant botulism in breastfed infants.

The various exquisitely tailored botulinum toxins are likely the consequence of complex pathways of convergent evolution. Seven different serotypes have evolved, designated by the letters A through G. The botulinum toxins are similar in action to tetanus toxin, although the targeted cell differs: tetanus toxin affects Renshaw cells of the spinal cord while botulinum toxin alters the secretive exocytosis of motor neurons of the brain stem and spinal cord. Botulinum toxin is synthesized initially as a large 150-kDa protein that is autocleaved into two fragments bonded by disulfide links. The larger 100-kDa heavy-chain fragment of each of the botulinum toxins is responsible for binding to a variety of specific gangliosides that characterize the axon terminus of cholinergic motor neurons. Once internalized by endosomic uptake, the light chain separates and escapes the endosome into the cytoplasm of the presynaptic terminal. Here, a zinc-dependent protease function of the light chain targets one of three proteins essential to the sequence of synaptic vesicle docking and exocytosis. These targeted proteins are enzymatically cleaved at a site distinct for each toxin type. Types A and E toxin cleave a protein known as SNAP 25. Types B, D, F, and G (as well as the short chain of the tetanus toxin) cleave a critical vesicle-associated membrane protein (VAMP, also known as synaptobrevin). Botulinum toxin type C cleaves the syntaxin molecule. The consequence of inactivation of any one of these crucial docking proteins is an inability to release quanta of acetylcholine from the presynaptic terminal in response to an action potential. The remarkable potency of botulinum toxin stems from this combination of high-affinity uptake at presynaptic terminals, mediated by the 100-kDa fragment, and highly efficient enzymatic protease activity mediated by the 50-kDa fragment.

Recovery from this intoxication requires the spontaneous degradation of the toxin and the synthesis and transport of new SNAP 25, VAMP, or syntaxin protein from the neuronal perikaryon. Morphologically, recovery is associated with new sprouting from the presynaptic terminal in the neuromuscular junction; whether this is a necessary feature of recovery is not known. While the muscle is denervated, there is spread of the acetylcholine receptor from the site of the original junction, but with recovery of synaptic quantal release, the regular arrangement of the synapse is reestablished. Subtle differences in the frequency of spontaneous quantal release of acetylcholine are apparent in the different forms of botulinum toxin. With type B cleavage of VAMP only, the stimulated release of acetylcholine quanta is inhibited. In contrast, with type A cleavage of SNAP 25, both stimulated and spontaneous quantal release of acetylcholine are inhibited. This difference may be associated with more significant denervation responses in muscle affected by type A toxin, with greater alteration of resting membrane potential and frequency of abnormal spontaneous activity. These differences may be reflected in single-fiber electrophysiologic studies.

Differential Diagnosis and Evaluation

The differential for subacute onset of hypotonia may be very broad, but due to the age at presentation the bias may lean towards infectious etiologies. Clinical suspicion for botulism must remain high in order to proceed toward appropriate evaluations in a timely manner. Actual clinical mimics were described by a review of 681 cases of infant botulism from 1992 to 2005 in which 32 patients (4.7%) met the clinical diagnosis of infant botulism but were not laboratory confirmed and subsequently were given a different diagnosis. Of those without laboratory confirmation, 28% had no other diagnosis and were felt to have clinical botulism. The remainder 23 patients fell into five diagnostic categories including metabolic disorders (25%, including glutaric aciduria type 1, maple syrup urine disease, Leigh’s syndrome), miscellaneous (22%, including GBS variants, central demyelinating disease, cerebral infarct), SMA type 1 (16%), and infectious diseases (9%, RSV, enterovirus encephalitis) ( Box 25.1 ).

Evaluations may include cerebrospinal fluid examination, but a spinal tap may be dangerous as there may be impending respiratory compromise. Other overlapping syndromes include electrolyte disturbances including hypo- or hypercalcemia, hypernatremia, hypo- or hyperkalemia, hypo- or hypermagnesemia, and hypoglycemia. The prospect of an inborn error of metabolism should be considered in very young infants who become lethargic after the initiation of enteral feedings, as well as in any infant newly challenged by fasting or intercurrent infection. In very young infants in whom the possibility of an inborn error is high, initial serum and urine samples should be saved for later, more definitive analysis, if indicated.

Other rare causes of weakness in previously healthy infants include organophosphate poisoning, which presents with additional signs of cholinergic excess. Poisoning with heavy metals may cause a chronic or acute flaccid state in infants. Now exceptionally rare, infantile poliomyelitis generally accompanies vaccination with live attenuated virus, although transmission from contact with another recently vaccinated infant can be a source of the virus. Vaccine-associated poliomyelitis usually involves back-mutation to a virulent state in the type 3 polio serotype. Alternatively, it can affect some infants with unsuspected immune deficiency in a more chronic fashion. The syndrome of infantile polio is quite different from that of infant botulism however, with recent diarrhea, a low fever on presentation, irritability, and aseptic meningitis evident. The weakness of infantile polio may or may not be asymmetrical during the acute phase. Another possible but unusual diagnosis is Guillain-Barré syndrome (GBS), which is exceptionally rare in infants, (see Chapter 23 ). GBS should be distinguished by electrophysiologic criteria and elevation of cerebrospinal fluid protein as the disease progresses.

One of the most difficult diagnoses to distinguish from infant botulism is an unrecognized heritable disorder of the neuromuscular junction, known as a congenital myasthenic syndrome. These infants present with increased weakness precipitated by some form of metabolic stress. Like infants with botulism, such infants characteristically have ocular involvement (though rarely pupillary sluggishness) and bulbar weakness greater than limb weakness. The use of edrophonium in the diagnosis can be confusing because infants with botulism may be transiently stronger. With the congenital myasthenic syndromes, the history is often one of a more indolent process, though this can overlap with milder, chronic forms of infant botulism. The two disorders are best distinguished by electrophysiologic studies and recovery of toxin or organism from the stool.

Nearly all infants suspected of having infant botulism must first be considered to have a serious systemic or central nervous system infection; appropriate cultures should be obtained, and prophylactic antibiotics administered. Early laboratory tests also need to include determination of glucose, electrolytes, calcium, magnesium, acid-base status, anion gap, and ammonia levels. Intravenous glucose can be started, but enteral feedings should be held pending results. Plasma and urine specimens for toxicology and metabolic studies must be obtained in case their evaluation is warranted by later developments.

The diagnosis can be made most rapidly by electrophysiologic studies. In the most carefully documented study, an incremental response to rapid repetitive stimulation (20 or 50 Hz), found in 92% of patients, was the most sensitive and specific finding ( Figure 25.1 ). Slower rates of repetitive stimulation produce variable results. Concentric needle electromyography is also sensitive but less specific, with many patients demonstrating short-duration, low-amplitude motor unit potentials and abnormal spontaneous activity. Several technical problems can undermine the sensitivity of the conventional rapid repetitive stimulation test for incremental compound muscle action potential (CMAP) amplitude, including inadequate immobilization of the limb, studying a nerve-muscle combination that has a very low CMAP amplitude below a physiologic threshold allowing increment, or studying a muscle that already has a normal CMAP amplitude and cannot increment further. One frequent error is to unwittingly study a nerve-muscle combination that has been transiently incremented by the baby’s exercising the muscle during the set-up for rapid repetitive stimulation. Concentric needle electromyographic studies demonstrate frequent brief, small-amplitude, polyphasic, voluntary motor unit potentials reflecting denervation of a portion of the array of muscle fibers innervated by each motor axon. This finding, along with the frequent presence of abnormal spontaneous fibrillation and positive sharp-wave potentials, is also seen in other disorders of infants and is not specific for infant botulism.

Tracing derived from first author’s patient with confirmed type A infantile botulism. Recording of the right abductor digiti minimi (ADM) muscle with 50 Hz repetitive nerve stimulation shows >50% facilitation of the baseline compound muscle action potential (CMAP).

Studies of jitter with a stimulated single-fiber electrode may have enhanced sensitivity over conventional rapid repetitive stimulation. This method is technically easy in weak infants and requires a much lower-level stimulus, making it less painful. A positive result is reported when abnormal jitter between the stimulus artifact and an isolated single muscle fiber potential decreases with an increasing stimulation rate, similar to that expected with Lambert-Eaton myasthenic syndrome. Autoimmune myasthenia and other postsynaptic disorders would be expected to have increasing jitter with an increased stimulation rate.

Most infants have both toxin and organism recoverable from the stool, although local logistical concerns may favor one test over the other. Because affected babies have a reduced stooling frequency, the fluid recovered from a gentle 20-mL saline enema contains enough material to test for the toxin. The blood is usually, but not always, negative for toxin. Stool can remain positive for culture and toxin for a long time, though others report that it is clear within 1 month of diagnosis. Polymerase chain reaction techniques have been developed that may be incrementally more sensitive but are not yet in routine use.

Therapy

Prior to the introduction of human botulism immune globulin (Baby-BIG), approved by the Food and Drug Administration in October 2003, treatment was largely supportive with respiratory and nutritional care. Baby-BIG is comprised of immunoglobulin isolated from donor plasma immunized with pentavalent botulinum toxoid (ABCDE). A 5-year randomized placebo-controlled trial treating with Baby-BIG vs. IVIg showed a significant reduction in morbidity with reduction in length of hospital stay and duration of intubation by a mean of 3 weeks. This study was followed by another 5-year open-label study and a 30-year retrospective review that recapitulated findings. Mean intensive care stay was reduced by 3.2 weeks, mechanical ventilation by 2.6 weeks, and tube feeding by 6.4 weeks. This resulted in significant cost reduction as well. Although the cost of BIG is high (approximately $45,000 in 2005), the overall savings in hospital stay and morbidity were far more significant with an estimated $34.2 million in hospital charges that were avoided. BIG has been demonstrated to be of benefit in infant botulism when given within the first 3 days after admission to the hospital. Equine-derived botulinum antitoxin is not used in infants because of the high incidence of rash and anaphylaxis and the potential for lifelong sensitization to horse serum products. A Cochrane review also supports BIG as the only medical treatment that showed significant effect compared to 3,4-diaminopyridine (3,4 DAP), plasma exchange, equine antitoxin, and guanidine.

Baby BIG is available worldwide by the California Department of Health Services Infant Botulism Prevention Program (telephone, 1-510-231-7600; up to date as of 8/2014). Because of the necessity for quick treatment, BIG can be administered on the strength of a classic history and physical examination (details at the California Department of Health Infant Botulism Prevention and Treatment Program website ), with laboratory confirmation of the diagnosis and botulinum toxin type to follow.

Other than administration of BIG, primary management consists of respiratory and nutritional management. At-risk infants should be watched carefully in the intensive care unit and intubated early when signs of airway compromise, bulbar dysfunction, or respiratory insufficiency arise. Infants who require an artificial airway do so for weeks and may acquire a respiratory virus that can prolong their ventilator requirement. Most now agree that despite this long-term requirement, tracheostomy is rarely indicated because normal lung compliance and a weak chest wall permit full ventilation even with a slightly small, uncuffed tube that has some degree of air leak. Under such conditions, few patients develop airway strictures, even with long periods of intubation.

Because infants often recover in the reverse order of the original progression, the clinician may be faced with the odd circumstance of an infant intubated for weakness who has good arm and leg muscle power. Some have advocated the return of head stability as a predictor of adequate airway competence. Others wait for the maximum inspiratory pressure to exceed −25 cm H ₂ O and the return of protective airway reflexes. In either case, it is just as important to follow the infant’s airway after extubation, during recovery from the disease, as it is when the child first presents. Throughout the illness, the capacity for enhanced fatigue of airway and respiratory muscles is a greater concern than is baseline weakness.

Given the severe constipation, concern about enteral feedings led many to give infants parenteral alimentation in the past. Many infants do not tolerate bolus feedings by nasogastric tube, but most can be adequately nourished by continuous nasogastric feedings, beginning slowly, as early as 3 days after presentation. Concern about gastroesophageal reflux with aspiration has prompted some to use the nasojejunal route for tube feeding. Caloric requirements in the absence of physical activity are significantly reduced.

Many infants experience altered autonomic function, including sudden unexplained alterations in heart rate, blood pressure, or skin color. These are usually mild, transient, or both and rarely require specific therapy. Some of these changes are undoubtedly due to supersensitivity of cholinergic autonomic terminals after functional denervation by the toxin. Urinary retention may require intermittent catheterization to prevent bladder infections. Some very young infants develop the syndrome of inappropriate antidiuretic hormone secretion, especially early in the disease, just after being placed on positive-pressure ventilators. This should respond to fluid restriction and usually resolves spontaneously over several days to a week. Left unrecognized or untreated, the low sodium values that result can cause serious, anticonvulsant-refractory seizures.

Although many infants are given antibiotics immediately upon presentation because of the possibility of unrecognized sepsis, further treatment with antibiotics has not been successful and might be harmful, either by perpetuating C. botulinum colonization of the gut with the alteration of normal enteric flora, or by increasing the release of botulinum toxin, which is in high concentration within the organism. In particular, treatment with aminoglycoside antibiotics may be contraindicated because of the potential for increased neuromuscular blockade. Giving specific foods such as yogurt in an attempt to alter the normal colonic flora has not been tried systematically, but the results with individual patients have not been striking. Although colonic enemas may theoretically reduce the enteric burden of toxin, there is concern that they may also increase absorption. Because recovery from infant botulism is very good once the diagnosis is made and appropriate supportive care is initiated, there is little enthusiasm for these treatments that may carry some risk. Treatment with acetylcholine esterase blockers produces mixed results.

Although some infants may worsen during convalescence, relapse after full recovery is not seen. Some infants may harbor toxin in the colon well after recovery, suggesting that some degree of circulating humeral antibody aids in absorption of residual toxin taken up systemically.

Very importantly, the parents need to be assured of the probable good outcome for their infant, despite the prospect of a prolonged intensive care hospitalization. Parents should be encouraged early on to establish a regular routine that includes time away from the hospital. This will enable them to better tolerate the personal stress associated with a prolonged intensive care stay.