Bioterrorist attacks using spores derived from Bacillus anthracis have been identified as a high priority threat by the United States (US) Department of Homeland Security [1]. This issue was highlighted by bioterrorism-related cases of anthrax illnesses after envelopes containing spores of this organism were sent through the US mail [2]. Accordingly, the US Department of Health and Human Services (HHS) has been charged to address preparedness for such attacks. This preparedness includes providing guidance on the use of post-exposure prophylaxis (PEP) using anthrax vaccine and antibiotics [2]. Anthrax Vaccine Adsorbed (AVA, BioThrax^®) manufactured by Emergent BioSolutions Incorporated, was licensed in the US in 1970 for prevention of anthrax in adults aged 18 to 65 years. However, children and pregnant women are special populations for its use [3,4]. AVA is prepared from sterile culture filtrates of the toxigenic, nonencapsulated B. anthracis V770-NP1-R grown in a protein-free medium. The final product formulation contains aluminum hydroxide, sodium chloride, benzethonium chloride and formaldehyde [3]. The primary immunogen in AVA is anthrax toxin protective antigen (PA). Anti-PA IgG antibodies are considered to protect against anthrax by neutralizing the B. anthracis toxins, inhibiting spore germination, and enhancing phagocytosis and killing of spores by macrophages [5–13].

The current US Advisory Committee on Immunization Practices recommendation for PEP use of AVA is subcutaneous (SC) administration of three doses at 0, 2 and 4 weeks to be initiated within 10 days following an anthrax event [14]. The safety profile of AVA in adults 18–65 years of age is well established [15–22]. There is however, a paucity of data on AVA safety and immunogenicity in special populations, and none in children [23].

In 2011, a HHS interagency tabletop exercise, designated Dark Zephyr, was conducted to simulate an anthrax emergency [24]. During this exercise, it was estimated that up to 7.6 million people, of which approximately 25% would be children, could be exposed to B. anthracis spores [24]. If such a large-scale event actually happened, the absence of safety and immunogenicity data of AVA in pediatrics may result in concerns about the administration of this vaccine to individuals less than 18 years of age, a situation that could possibly deny children a potentially life-saving prophylactic countermeasure.

During the fall of 2011, the National Biodefense Science Board, now known as the National Preparedness and Response Science Board (NPRSB) was charged with assessing challenges in the use of AVA in the pediatric population in case of a large-scale anthrax emergency [24]. The NPRSB recognized that, in case of mass exposure of a population to B. anthracis spores, a FDA approved research investigational new drug protocol would allow the administration of AVA to children using a PEP regimen. However, this effort would require a research team to collect safety and immunogenicity data from these children after each AVA dose during this mass vaccination event. Consequently, the NPRSB noted that this type of post-event evaluation would pose major challenges to first responders, parents and research personnel in terms of mass vaccination of children during a large-scale anthrax spore exposure. Therefore, the NPRSB panel concluded that “HHS should develop a plan for and conduct a pre-event study of AVA in children, to include a research IND. HHS should submit a study protocol to one or more institutional review boards, and comply with the 21 CFR 50.54/45, CFR 46.407 federal review process." [24].

In response to the NPRSB report, the Secretary of HHS requested that the Chair of the Presidential Commission for the Study of Bioethical Issues convene a panel to review the ethical considerations of conducting clinical research studies of medical countermeasures in children. The Secretary went further to ask this panel to specifically include the ethics of conducting a pre-event AVA study in children. The Commission held four public forum meetings that addressed this issue directly and a summary report was issued on March of 2013 [25]. In that report, the Presidential Commission referred to the Code of Federal Regulations (CFR) regarding protections for children involved in research [26].

First, the Commission indicated that pre-event AVA studies could not be conducted in children in the US under 45 CFR 46.405, which specifies that studies above minimal risk require the possibility of direct benefit to the participating child. Second, pediatric studies may be possible using 45 CFR 46.407, which stipulates that a rarely utilized Presidential waiver could be sought if the information gained could possibly benefit children in general even if the study might not benefit the individual child. Finally, the Commission suggested a unique approach that would render a pre-event AVA study to “no more than a minor increase over minimum risk" by using a stepwise, age de-escalation approach. Specifically, the Commission suggested that a pre-event study of individuals 18 to 20 years of age might provide information to substantiate that such a study in 16 to 17 year-olds would involve no more than minimal risk. Consequently, the Commission indicated, with important caveats, that an age de-escalation pathway might be considered under the 45 CFR 46.404, as it poses no more than minimal risk.

In light of this Presidential Commission’s unique alternative pathway suggestion, we implemented a retrospective study with an exploratory objective to describe and compare safety and immunogenicity data from healthy individuals aged 18 to 20 years to the same type of data in individuals aged 21 to 29 years who participated in several HHS-sponsored AVA clinical trials. For this study the ’older’ age group of 21 to 29 years, albeit somewhat arbitrary, was chosen as the most appropriate age group to compare safety and immunogenicity data to the 18 to 20 year old group for several reasons. First, including data from subjects up to age 65 years would result in a markedly larger and very unbalanced sample size compared to the 18 to 20 year old group. Also, previous data has shown a decrease in antibody responses to AVA as age increases in ten year increments from 18 to 65 year olds [27]. Finally, injection site reactions to AVA significantly decrease with advancing age [20].

For all four studies combined there were 74 subjects 18 to 20 years of age and 243 subjects 21 through 29 years of age (Table 1). Table 1 also illustrates the demographics of subjects enrolled in each of the studies. When all study groups were combined, there were no significant differences in demographic characteristics of age, race or gender between the two age categories.

Table 2 shows the proportion of subjects in each age group who experienced any grade 1 to 3, or any maximum grade 3 elicited AEs. The denominators are the sum of the numbers of diaries collected after dose 1, dose 2 and dose 3. For example, for the younger age group, there are 74, 71 and 71 diaries collected for doses 1, 2 and 3, respectively. Therefore, the denominator used for this age group was 216 (= 74 + 71 + 71). There were no significant differences in overall combined local or systemic AEs between the two age groups for any grade 1 to 3, or of grade 3 local AEs within seven days after each vaccination.

With regard to individual local elicited AEs of any grade, Table 2 reveals that subjects aged 18 to 20 years had significantly lower rates of any local warmth (37.5% vs. 46.5%; P<0.05) and higher rates of maximum swelling (1.1% vs. 0%; P< 0.05) compared to older subjects. There were no significant differences in individual or of combined systemic elicited AEs of any or of maximum grade in this data set.

Table 3 displays the results of elicited AEs related to sequential vaccine dosing at the three time points 0 to 2 to 4 weeks. In general, there was a decline in the rates of elicited AEs with each subsequent dose. For both age groups, there were statistically significant declines in maximum local AEs and any grade systemic AEs.

No SAEs were recorded in subjects aged 18 to 29 years who received AVA by SC injection in the four combined studies within 30 days of any of the three AVA doses given at study weeks 0, 2, and 4.

Table 4 displays the seroresponse rates for the two age groups as measured by serum NF50 or anti-PA IgG concentrations. In general, the proportions of subjects with seroresponses were high after the second and third AVA doses. There were no significant differences in proportions of NF50 or anti-PA IgG antibody seroresponses after the second or third AVA doses between the 18 to 20 and 21 to 29 year olds. As noted in the methods, serologic data for both assays were only available on a subset of subjects included in the present study.

AVA appeared tolerable in both age groups in the present study, with few maximum grade 3 elicited AEs and no SAEs attributable to AVA observed in subjects that received AVA at 0, 2 or 4 weeks. Overall, the 18 to 20 year olds had similar rates of local and systemic AEs compared to the 21 to 29 year age group.

It is important to note that the overall proportions of elicited local and systemic AEs were similar to those reported for other vaccines, including diphtheria-tetanus-acellular pertussis vaccines (Tdap and DTaP), Hepatitis A and Hepatitis B, and others [17,35–39]. Relevant to the focus on the possibility of pediatric AVA studies in the future, the present study revealed proportions of elicited AEs that were similar to those seen in published studies of adolescents given booster doses of Tdap, or of 4 to 6 year-olds given DTap [40,41].

Finally, data from Table 3 regarding frequencies of AEs in each sequential AVA dosing week are somewhat reassuring in that the proportions do not increase and in fact generally appear to decrease with each subsequent vaccine dose. This information may be comforting for parents and children if pediatric studies involving AVA are ever initiated.

The immunogenicity results demonstrated that seroresponses measured in the 18 to 20 year-old subjects were not significantly different than those observed in the older subjects after the second and third AVA doses. We chose not to include subjects older than 29 years of age because the AVA000 study revealed that antibody responses decline with age [42].

A limitation of the present investigation was the small numbers of 18 to 20 year-old subjects studied. This situation resulted in unbalanced groups when comparing results between the small numbers of 18 to 20 year olds to the larger number of older subjects. To some degree, this unbalance was mitigated by limiting the older age group to 21 to 29 years of age resulting in a ratio of about 1 to 3.3 of the younger to the older age group. The small number of young subjects led to the decision not to compare study safety or immunogenicity data in additional smaller subgroups such as differences between subjects of different gender or race.

Another limitation was that numerical scoring criteria for mild, moderate and severe (grades 1, 2, or 3) local elicited AE safety data such as fever, lump, induration and swelling differed between the CDC and BARDA studies. These differences were somewhat mitigated analyzing just two categories as either the percentage that had any grade or only grade 3 elicited AE in the present study. Also, in the present study, the safety data likely varied between studies because of factors such as the use of different clinical research organizations, clinical research associates, and geographically clinical sites. However, this limitation is somewhat mitigated by the relative uniformity of type of local and systemic elicited AEs in all four of the studies utilized in these analyses.

Another study limitation was that the serologic data were available for only a subset of subjects in this study. Also, as previously mentioned, it was impractical to compare the actual NF50 measurements or anti-PA IgG concentrations between the four studies because of likely laboratory variability of assays that were conducted over more than a decade and tested in different laboratories. However, using the proportion of seroresponders, rather than comparing geometric mean values, helped mitigate immunogenicity variability between the studies included in this report.

In summary, the above limitations necessitate that well-designed prospective studies of balanced groups of 18 to 20 year-olds and older adults should be performed to validate our findings. Nevertheless, data from the present study suggests that AVA is as safe and immunogenic in the 18 to 20 year old when compared to the 21 to 29 year age group.

Current epidemiologic research does not support an association between AVA and severe, unusual, or chronic illnesses in adults [43,44]. These epidemiologic studies combined with the data from the present study may be useful in informing future discussions of conducting post-event or even pre-event AVA pediatric studies in the future. Certainly, immediate and long term safety data for AVA administered to children would be important information for medical response providers if this vaccine is to be given to this population during a large scale anthrax emergency. The ‘age de-escalation’ approach discussed in the Presidential Ethics Report raises significant ethical and regulatory issues that deserve further discussion.

As previously mentioned, this report involved only subjects that received AVA SC because this is what is currently recommended for PEP (14). However, it is likely that AVA administered SC or IM would result in similar antibody responses [45]. Additionally, it is clear that higher local reactogenicity rates have been observed when AVA is administered SC vs. IM [18,45]. Additionally, most health care personnel are more familiar with vaccine administered IM and thus this route may be preferable to use during an emergency PEP situation. Notably, pre-exposure prophylactic AVA is now recommended and approved by the FDA to be given IM [3] and this method of administration may be better tolerated in younger children with smaller limbs and less subcutaneous tissue than most adolescents and adults. It should be noted that the PEP AVA regimen for which FDA licensure is being sought involves SC administration not IM. However, it is anticipated that post-licensure studies directly comparing the safety and immunogenicity of SC vs. IM PEP regimens are planned.

In conclusion, the results of these exploratory analyses demonstrate that AVA appears to be safe and immunogenic in 18 to 20 year olds compared to individuals 21 to 29 years of age. Additionally, this investigation provides some information relevant to the ‘age de-escalation’ approach suggested by the Presidential Ethics Commission.