Introduction

Infant botulism (IB), the intestinal toxemia form of the disease in a child aged less than one year, results when spores of the soil-dwelling anaerobic bacterium Clostridium botulinum or related neuro toxigenic clostridia temporarily colonize an infant’s large intestine and produce botulinum neurotoxin (BoNT) [1]. BoNT then enters the systemic circulation and binds to presynaptic cholinergic synapses within the peripheral and autonomic nervous system, blocking the release of acetylcholine which results in flaccid paralysis [2,3]. Most cases of infant botulism in the United States are caused by C. botulinum toxin type A and type B. Toxigenic strains of Clostridium butyricum and C. baratii produce botulinum neurotoxin type E and type F, respectively, and are rare causes of IB [2].

Although IB cases have been retrospectively identified in California as early as 1931[4], IB was first recognized as a novel form of human botulism in 1976 [2]. IB occurs globally and is the most common human form of botulism in the United States [5]. Approximately 40% of all IB cases in the United States have occurred in California. A forty-year longitudinal epidemiological study of California infant botulism cases found that 63.3 % of IB cases were attributed to toxin type A, 34.5% to type B, and only 0.2% to type F [3]. A 2018 review reported 18 IB cases caused by toxin type F in the United States from 1976-2016 including sporadic clusters [6]. According to the Centers for Disease Control and Prevention (CDC), in 2021 there were 214 laboratoryconfirmed cases of botulism including 183 IB cases [7]. Toxin types A and B were identified in all 183 IB cases including 2 cases with rare dual-toxin-producing or “bivalent” strains, type Ba and type Bf.

In comparison to toxin types A and B, infants infected with toxin type F typically have symptom onset at a younger age (median of 9 days compared to 1-6 months), with more rapid progression to severe paralysis, and a lower prevalence of constipation [6]. While common known risk factors for exposure to Clostridium spores include ingestion of honey [8] and environmental spores present in the soil, the source is not identified in the majority of infant botulism cases, including those caused by Clostridium baratii type toxin type F [5].

Following licensure in 2003 through 2015, human-derived botulism antitoxin, BIG-IV (BabyBIG) licensed for the treatment of infant botulism caused by toxin serotypes A or B, has been used to treat approximately 1,200 patients with infant botulism [9]. The equine-derived HBAT used to treat adult botulism patients is active against toxin types A,B,C,D,E,F,G; however, there is limited experience with its use for the treatment of infant botulism [10]. The article by Yu et al from 2017 [11] examined the use, clinical benefits, and safety of HBAT in patients younger than 17 years. The authors observed that those treated earlier in the disease course spent significantly fewer days in the hospital (P < 0.1) and intensive care unit (P = 0.04). They report one HBATrelated serious adverse event (< 1% of cases) where a 10-yearold child experienced hemodynamic instability characterized by tachycardia, bradycardia, and asystole during and following HBAT administration [10,12]. The article did, however, conclude that HBAT provided clinical benefits and was well-tolerated [11]. Given the limited safety data in the pediatric population, the short half-life of equine antitoxin, and concern for sensitization to equine proteins, there has been some hesitation for its use in infants. Notably, rapid recovery of infants with type F botulism without antitoxin treatment has also been reported [13].

Case presentation

In June 2020, a 10-day-old breastmilk and formula-fed male infant who was delivered post-dates was transferred to our level IV neonatal intensive care unit (NICU) for respiratory failure after a choking event. Oxygen saturations were noted to be in the 70 percent range upon arrival and the infant required positive pressure ventilation. He was described as lethargic and was intubated for airway protection. The examination was notable for generalized hypotonia with slightly dilated but reactive pupils and withdrawal of limbs to noxious stimuli. He had an intact gag reflex, palmar and plantar grasps, extensor plantar responses, and an incomplete Moro reflex. Within 12 hours, the infant became areflexic and paralyzed. He developed dilated and nonreactive pupils. Corneal, oculocephalic, and gag reflexes were absent. Muscle tone was diffusely flaccid with generalized areflexia and no spontaneous movements. There was no withdrawal to noxious stimuli despite cessation of sedative medications and administration of naloxone. 

An electroencephalogram (EEG) and magnetic resonance imaging (MRI) of the brain were performed to rule out anoxic brain injury or other processes, such as hemorrhage, infarction, encephalomyelitis, diffuse demyelination, or myelinolysis, that could affect the brainstem and both hemispheres and explain this clinical picture. EEG showed somewhat discontinuous activity with focal slowing, but normal amplitudes. MRI of the brain showed an incidental finding of a small frontal lobe calcification and scattered subdural blood products felt to be related to birth trauma. There was no evidence of structural lesions to explain the infant’s symptoms. Screening metabolic and infectious laboratory analyses were negative. Cerebrospinal fluid (CSF) analysis for cells, protein, and glucose was normal for age. Screening for infectious etiologies including Gram stain, cultures, and BioFire FilmArray® (BioFire Diagnostics, Durham, NC) multiplex PCR molecular testing meningitis/encephalitis panel were negative. Brainstem auditory-evoked responses and visual-evoked potentials were present, though delayed and with altered waveforms. 

Stool was collected for botulinum toxin testing, and the patient underwent nerve conduction studies (NCS) and needle electromyography (EMG). The NCS were abnormal and significant for absent motor responses from ulnar and peroneal nerves. Radial sensory nerve action potential was of normal latency and amplitude. Fast frequency (20 Hz) repetitive nerve stimulation elicited no motor responses. The needle EMG showed diffuse spontaneous activity (fibrillation potentials and positive waves without fasciculations) in all muscles tested. No motor unit

potentials were seen (Table 1). These findings were concerning for a process either affecting the anterior horn cells, peripheral nerves (severe motor axonal polyneuropathy), severe neuromuscular transmission defect, or possible severe myopathy.

Comment NCS: Absent CMAPs, normal left radial response. NR: non-reactive. CMAP: compound muscle action potential.

Comment EMG: Diffuse spontaneous activity (fibrillation potentials and positive waves). No motor unit potentials (MUPs) were observed in any of the muscles tested.

Table 1: Neurophysiology results.