brook et al., 1994; Whalen et al., 1996; Kiyatkin et al., 1997; Bavari et al., 1998; Byrne et al., 1998; Smith, 1998). The monitoring period for adverse effects in these animal studies was no more than months. The authors of these studies did not mention adverse effects from vaccination with toxin fragments. Clayton and colleagues (1995) expressed a fragment of botulinum neurotoxin type A in Escherichia coli. Initial experiments immunized mice with a crude extract of E. coli expressing the gene of interest; several mice died following the second or third vaccination. The authors suspected endotoxin contamination. Subsequent experiments used purified neurotoxin fragments, and no toxicity occurred after repeated vaccination of these recombinant preparations.
Animal studies using botulinum toxoid vaccines have reported minimal transient local reactions and swelling at the injection site. Only short-term outcomes were reported, and in most studies, the reports did not mention adverse effects.
Only a few published peer-reviewed studies have examined the potential adverse health effects of the botulinum toxoid vaccine when administered to humans. The committee based its conclusions about possible associations between the toxoid and health effects solely on the peer-reviewed literature and included other studies when assessing research needs.
Early studies of the initial univalent botulinum toxoids in the 1940s reported a significant number of local and systemic reactions (Middlebrook and Brown, 1995). Toxoids were further purified in the 1950s, and a study published in 1962 (Fiock et al., 1962) reported on tests of bivalent toxoid preparations by Parke, Davis, and Company. This study of laboratory personnel at Fort Detrick focused on the efficacy of four different schedules of bivalent botulinum toxoid vaccination. Personnel, not previously immunized with botulinum toxoid, were assigned to four different vaccination schedules as they reported for initial vaccination (50 individuals followed a 0-, 2-, 4-, and 6-week schedule; 25 individuals followed a 0- and 8-week schedule; 50 individuals followed a 0-, 2-, and 10-week schedule; and 25 individuals followed a 0- and 10-week schedule). The study reports that after 800 injections, no systemic or severe local reactions had occurred. Some individuals (the number is not reported) had a small subcutaneous nodule that lasted 2–3 weeks. The report does not discuss the surveillance methods for monitoring adverse effects.
A subsequent study by Fiock and colleagues (1963) examined four different pentavalent (A–E) toxoid lots prepared by Parke, Davis, and Company. This