Immunogenicity and safety of influenza vaccination in patients with juvenile idiopathic arthritis on biological therapy using the microneutralization assay

Children and adolescents with rheumatic diseases (RD) are at increased risk of infection compared to age- and gender-matched subjects without RD due to their aberrant immunity and frequent use of immunosuppressive drugs [1]. Annual vaccination against influenza is recommended for immunocompromised patients [2]. Efficacy and safety of vaccination remain to be defined in rheumatologic patients with immunomodulatory therapy, including high doses of steroids. Recommendations for vaccination in pediatric patients with RD were published in 2011 by the European League Against Rheumatism (EULAR). Seasonal influenza virus vaccination should be considered in all pediatric patients with RD [3, 4]. Few studies have addressed influenza vaccination safety and efficacy in this group, most of which have small sample sizes, include a wide variety of rheumatic diseases and have scarce data on new biological therapies [3‐6].

Juvenile idiopathic arthritis (JIA) is the most common inflammatory rheumatic chronic disease of childhood, with a prevalence of approximately 1 per 1000. JIA is defined as arthritis of unknown etiology beginning prior the age of 16 years persisting for at least 6 weeks with other known conditions excluded [7]. The incidence of influenza infection in this patient group is unknown and evidence is lacking whether children with JIA have an increased risk of severe influenza infection or its associated complications. However, influenza virus can trigger disease flare, treatment interruption and the use of antibiotic therapy for suspected bacterial co-infection, all of which are associated with unnecessary medical revisions, thus reducing the quality of life of these children and their families [8]. The development of therapeutic biological agents including anti IL-6 and anti IL-1 receptor (R) therapy, has markedly altered treatment strategies for RD and greatly improved the prognosis of affected individuals [9]. Whilst influenza vaccination in children with RD receiving non biological immunosuppressive agents results in similar serum antibody titers compared to those of healthy children, few data exist on the impact of biological therapy [10]. Here, we describe the immunogenicity and safety of influenza vaccination and long term seroprotection on a pediatric cohort diagnosed with JIA receiving immunomodulatory therapy including anti IL-6 and anti IL-1R therapy and review the published literature.

In this cohort of children with JIA receiving immunomodulatory therapy, influenza vaccination was safe and immunogenic; and the neutralizing antibody response was independent of the type of treatment. To date few influenza vaccine studies have included children with JIA, and biological therapy was specifically studied in only five of them (Table 3). In all but one (Toplak et al. [18]) the hemagglutination inhibition test (HAI) was used to determine antibody response to vaccination, which is known for its overall poor sensitivity, thereby potentially underestimating the seroprevalence in a given population [19]. The microneutralization assay, which was used in the present study, has been shown to enhance sensitivity and specificity [14]. No correlate of protection has been defined for microneutralization, but good correlation with HAI defined protection thresholds has been reported [16]. Furthermore, it has been proposed that HAI antibody seroprotection cut-offs in children aged 6-72 months may be 1:110 to achieve 50% protection [20], thus it is possible that children with RD may need antibody titers equal to or higher than those described. There is a need to review the current practice of evaluating vaccine responses and should include more advanced techniques to evaluate immunogenicity such as microneutralization assay, which identifies functional antibodies, rather than quantitative tests only [21].

Blockade of the IL-6R has been shown to decrease lymphocyte activation and to restore B and T cell homoeostasis in patients with SLE [22]. Anti-IL-6R treatment may diminish influenza vaccine antibody responses since IL-6 induces B cell differentiation into antibody-forming cells and T cells into effectors cells [22]. However, Shinoki et al. studied JIA patients treated with tocilizumab (anti IL-6R) and demonstrated that influenza vaccination was immunogenic and safe, as was the case in our study (Table 3) [9]. Whilst in healthy subjects therapy with anti-IL-1 did not affect the antibody responses after vaccination with influenza vaccine [23], the impact of its use in children with JIA is unknown. To the best of our knowledge, this is the first study evaluating influenza vaccine in children receiving IL-1R antagonist (anakinra) for JIA and was found to be safe and immunogenic. However, further studies are needed to confirm its safety in pediatric patients, as only four patients were included here.

Anti-TNF-α therapy in rheumatoid arthritis inhibits germinal center reactions, thereby leading to decrease peripheral blood memory B cell levels, potentially contributing to a poor response to vaccination [24]. Four pediatric studies (Table 3) have evaluated influenza vaccine in JIA patients receiving anti-TNF-α therapy. Antibody titers against A(H1N1), A(H3N2), and B vaccine viruses were found to be reduced in patient receiving anti-TNF-α although only four children were enrolled [18]. In another study, 30 JIA patients treated with anti-TNF-α showed significantly lower GMTs against the A/H1N1 strain compared to those treated with disease-modifying antirheumatic drugs (DMARD) and healthy controls. Furthermore, sero-conversion and -protection rates were significantly lower in patients receiving anti-TNF-α [25], as was also reported by Carvalho et al. in 5 patients and only for the A/H1N1 strain [8]. In the Aikawa et al. study, no differences were observed in immunogenicity parameters between patients with or without anti-TNF-α blockers however, the seroconversion rate for the A/H1N1 strain was significantly lower in JIA patients compared to controls (83.2% vs. 95.6%, p = 0.008) [26]_. In our study, no differences between the control group and patients with or without biological therapy were observed, thus biological therapy plus MTX did not alter the vaccine response. Although lack of differences between groups could be due to the small sample size, these findings are in contrast with previous results, where the use of anti-TNF-α blockers was associated with a reduced vaccine response [25, 26]. One year after vaccination seroprotection rates were similar with a slight and expectable reduction for all strains but A/H3N2, which was higher. This could be due to natural exposition to the virus, or test precision variability. Although it was evaluated in only few patients, biological treatment seems not to influence in long-term seroprotection rates after vaccination.

Like in other studies in children with RD, no SAEs were reported and vaccination using an inactivated influenza vaccine was shown to be safe with mild systemic and local reactions as most reported AEFI [26, 27]. Our patients remained clinically stable (JADAS-71:1 (0-5)) during the study period with only two patients receiving systemic corticoids at the time of vaccination. Lack of active inflammation at the time of vaccination could have influenced the safety and immunogenicity outcomes in our cohort. This is supported by the study of Toplak et al., where only 42% of the children had inactive disease at the time of vaccination and all children suffering from a flare post vaccination, had at least one criterion for active disease prior to vaccination [18]. This is important as high dose steroid therapy is known to be associated with a reduced immnunogenicity of the influenza vaccine resulting in lower antibody titers [28, 29]. We observed seroprotected patients to have lower inflammatory markers such as platelet count and ESR compared to non-seroprotected children. A low inflammatory state may result in an enhanced cellular immune response (Th1 vs Th2) thereby supporting B and T cell differentiation with subsequent improved vaccine response [30, 31]. In contrast to the studies of Dell’ Era and Toplak et al. where only 8.3% and 13% of the patients were previously vaccinated, 62.8% of children with JIA (and 72% of those receiving biological treatment) received the influenza vaccine the season prior to this study, and that has likely contributed to the overall high vaccine response shown here. This is also supported by previous results suggesting that baseline seroprotection promotes significantly higher GMT after vaccination [32]. Whilst over 60% (H1N1 68.6%, H3N2 62.9% and Flu B 68.6%) of our JIA children were seroprotected prior vaccination, only 20% of the patients described by Toplak and Aikawa et al. showed pre-vaccine seroprotection, further supporting the benefit of seasonal influenza vaccination. However, differences between participant’s age and previous exposure to the virus could alter pre-vaccination seroprotection rates. A limitation of this study is the overall low number of patients and healthy controls included, which may have limited the power to detect differences, especially when performing subgroup analysis. Although clinical, laboratory baseline parameters and GMT baseline titers were similar in the groups; there was heterogeneity in the treatment used and likely in the previous influenza exposure (artificial or natural) of the included patients. Further studies evaluating vaccine efficacy and effectiveness are necessary.

In this prospective study of children with JIA, the use of the trivalent inactivated influenza vaccine including A/H1N1, A/H3N2 and B strains was shown to be safe and immunogenic. The type of immunomodulatory therapy did not alter vaccine response. This patient group would likely benefit from seasonal influenza vaccination thereby supporting the annual immunization program.