Serum 25-Hydroxyvitamin D Level and Influenza Vaccine Immunogenicity in Children and Adolescents

Background Vaccination is an important strategy in the prevention of influenza, but immunologic response to vaccination can vary widely. Recent studies have shown an association between serum 25-hydroxyvitamin D (25[OH]D) levels and immune function. The purpose of this study was to determine if serum 25(OH)D level correlates with influenza vaccine immunogenicity in children and adolescents. Methods We conducted a prospective cohort study of children age 3 to 15 years of age vaccinated with trivalent influenza vaccine (A/Brisbane/59/2007[H1N1]-like virus, A/Brisbane/10/2007 [H3N2]-like virus and B/Florida/4/2006-like virus) in Hutterite communities in Alberta, Saskatchewan and Manitoba. Serum 25(OH)D levels were measured at baseline and immunogenicity was assessed using hemagluttination inhibition (HAI) titers done at baseline and 3–5 weeks post vaccination. Logistic regression was used to assess the relationship between serum 25(OH)D level as both a continuous and dichotomous variable and seroprotection, seroconversion, fold increase in geometric mean titer (GMT) and post vaccination titer. Results A total of 391 children and adolescents were included in the study and 221 (57% had post-vaccination HAI titers. The median serum 25(OH)D level was 61.0 nmol/L (Interquartile range [IQR] 50.0, 71.0). No relationship was found between serum 25(OH)D level and seroprotection (post-vaccination titer ≥40 and ≥320) or seroconversion (post-vaccination titer ≥40 for participants with pre-vaccine titer <10 or four-fold rise in post-vaccination titer for those with a pre-vaccine titer ≥10). Conclusion Serum 25(OH)D level was not associated with influenza vaccine immunogenicity in otherwise healthy children and adolescents. Other strategies to enhance influenza vaccine response should continue to be evaluated in this population. The role of serum 25(OH)D level in vaccine responsiveness in other populations, especially those hyporesponsive to influenza vaccination, requires further study.


Introduction
Influenza is the cause of annual seasonal epidemics estimated to affect 5% to 10% of the population, posing an important public health burden [1]. Vaccination with influenza vaccine can reduce morbidity and mortality associated with influenza. Therefore, vaccination is a key strategy to prevent illness and transmission.
Protection obtained from influenza vaccination can vary widely depending on the vaccine match with circulating strains of influenza and the individual immune response [2]. The response is often reduced in immunocompromised patients, elderly patients, patients with chronic conditions and other individuals because of immunological changes of which many have yet to be fully elucidated [3]. Therefore, strategies to boost the effectiveness and duration of immunity after vaccination are of specific interest. Although strategies such as an increase in vaccine dose, intradermal injection, and use of adjuvants have been shown to increase immunogenicity [4][5][6], little is known about the effect of vitamins on influenza vaccine antibody response.
Recent evidence has shown that vitamin D is associated with both innate [7][8][9] and adaptive [10][11][12] immune responses and therefore may have a role in vaccine immunogenicity. However, the link between vitamin D levels and vaccine responsiveness remains theoretical and has received little focus in existing studies. Evidence to date includes mouse models which demonstrate that mucosal and systemic antibody responses to influenza are enhanced when the vaccine is co-administered with calcitriol (1,25-dihydroxyvitamin D) [13,14]. However, a study in healthy adults failed to demonstrate a benefit from calcitriol intramuscular co-administration with the vaccine [15]. Furthermore, post-hoc analysis of a prospective influenza vaccine trial in HIV positive individuals did not find a difference in vaccine responsiveness between those receiving routine vitamin D supplementation and those not on supplementation [16]. These studies were limited by their lack of baseline serum 25(OH)D levels, making it uncertain as to whether individuals had low serum 25(OH)D levels at baseline prior to supplementation. More recently, a small study of 35 patients with prostate cancer found an association between baseline vitamin D level and influenza vaccine response [17].
Children are an important potential target group to optimize vaccine response given their potential role in transmission to highrisk groups [18]. This is also a group where vitamin D deficiency can be common [19][20][21].
The purpose of this study was to determine if there is an association between baseline serum 25-hydroxyvitamin D [25(OH)D] level and inactivated trivalent influenza vaccine immunogenicity in children and adolescents.

Study Design
A prospective cohort of children and adolescents were given trivalent influenza vaccine as part of a cluster randomized controlled trial (RCT) evaluating the effect of influenza vaccination on infection rates in Hutterite communities (clinicaltrials.gov: NCT00877396; isrctn.org: ISRCTN15363571) [18]. The research protocol was approved by McMaster University Research Ethics Review Board. Study subjects included children and adolescents age 3 to 15 years. Participants were randomly assigned by colony (n = 46) to receive either inactivated seasonal influenza vaccine or hepatitis A vaccine. Research nurses monitored participants for influenza infection with twice weekly assessments during the influenza season and infection was confirmed by a positive nasopharyngeal polymerase chain reaction (PCR) result. Blood specimens for serum 25(OH)D levels were collected at baseline. Serum from venous blood was frozen until batched analysis was performed according to the manufacturer's instructions using the DiaSorin LIAISONH chemiluminescence assay. Blood specimens for influenza antibody titers were collected at baseline and at least 3-5 weeks after vaccination. The paired samples were then tested using hemaggluttination inhibition (HAI) using turkey erythrocytes and reference antigens a for A/ Only children and adolescents who received influenza vaccination and had vitamin D levels were included in the analysis. Covariables of interest included age, sex, presence of underlying comorbidities assessed by interview (heart/lung disease (including asthma), blood disorder, swallowing/choking disorder, ASA use, chronic metabolic condition, kidney/liver disease, immunodeficiency) and 25-hydroxyvitamin D level (nmol/L).

Outcomes
The primary outcome was immunogenicity, using criteria that included seroprotection (post-vaccination titer $40) and seroconversion (post-vaccination titer $40 for participants with prevaccine titer ,10 or four-fold rise in post-vaccination titer for those with a pre-vaccine titer $10), as defined by the Food and Drug Association (FDA) [23] and the Committee for Proprietary Medicinal Products (CPMP) [24]. We also explored the impact of defining seroprotection as a post-vaccination titer $320, a cutpoint that may be more appropriate for children [25]. Other outcomes measured included change in antibody titer (four-fold change and fold increase in geometric mean titers [GMT]) and antibody level post vaccination (log 2 -transformed). Fold increase in GMT was calculated by determining the ratio of raw postvaccination to pre-vaccination titer, calculating the arithmetic mean of the log of these ratios and calculating the exponent and associated 95% confidence interval.

Statistical Analysis
Baseline characteristics were described using mean and standard deviation for normally distributed data and median and interquartile range (IQR) for non-normal distributions. Characteristics were compared between those who had serology and those missing serology using the independent t-test or Wilcoxon rank sum test for continuous variables and chi-square or Fisher's exact test for dichotomous outcomes. Logistic regression was used to examine serum 25(OH)D level as a predictor of dichotomous outcomes (seroprotection, seroconversion and four-fold change in antibody titers). All relevant covariables (age, sex, presence of at least one comorbidity, serum 25(OH)D level) were first evaluated using univariable logistic regression. Vitamin D levels were analyzed as both a continuous variable (log transformed to correct positive skew) and dichotomized based on the American Academy of Pediatrics (AAP)(,50 nmol/L) and Canadian Pediatric Society (CPS)(,75 nmol/L) recommendations. Variables with a p value,0.1 were considered for inclusion in the multivariable model and the final model was determined using a step-wise backwards elimination method. It was decided a priori to adjust the final model for age and sex and to include serum 25(OH)D as a continuous variable.
Linear regression was used to examine the relationship between covariables and post vaccination HAI titers. Titers were log 2transformed using titer (transformed) = log 2 (titer/5) resulting in the following: 0 = no HAI activity, 1 = 1:10, 2 = 1:20, 3 = 1:40 etc. Covariables were analyzed and included in the final model as outlined for the logistic regression model. Generalized estimating equations (GEE) were used to account for clustering at the colony level for both regression analyses.
Geometric mean titers (GMT) were calculated at baseline and post vaccination and compared between subjects grouped by serum 25(OH)D levels based on the AAP (,50 nmol/L) and CPS (,75 nmol/L) cut-offs. Fold increase in GMT was calculated in each group (ratio of GMT pre and post vaccination) and the arithmetic mean of the log of the fold increase was compared using the independent t-test.
All estimates are presented with 95% confidence intervals. A p value,0.05 was considered significant. SPSS version 20 (SPSS Inc, Chicago, IL) was used to conduct the analyses.

Results
Baseline characteristics are summarized in Table 1. A total of 391 children from 21 communities received influenza vaccination; 221 (57%) had post-vaccination serology and could be included in the immunogenicity analyses. There were no significant differences between groups based on availability of serology aside from sex ( Table 1) In univariable analysis, there was no significant association between serum 25(OH)D level and seroprotection against any strain using a cut-off of 1:40 (Table 3) (Table 4). Similarly, there was no association between any of the covariables and presence of a four-fold change in titer or post-vaccination titer level (log 2 transformed).
The There was no significant difference in fold change in GMT between groups based on vitamin D level cut-offs of 50 nmol/l and 75 nmol/L (Table 5).

Sensitivity Analyses
Given the range in timing of post-vaccination titers (median 11 wks) and to account for possible waning immunity with time, a sensitivity analysis was conducted excluding subjects that had postvaccination titers more than 3 months after vaccination; this group also excluded the subjects that had late serum 25(OH)D levels. In this analysis, there was no change in the significance of covariables for any of the outcomes. Additionally, the analyses were conducted including the time from last vaccine dose as a variable in the analysis. This did not change the results appreciably. In a second sensitivity analysis, participants with proven influenza infection (n = 22) were excluded given the probability that the antibody change was related to natural infection not vaccine response. There was no appreciable change in the results. We also excluded participants who had titers done after the start of the influenza season (December 28, 2008). There was no association between vitamin D and seroprotection or seroconversion.
In order to account for pre-vaccination seroprotection, participants with baseline titers $40 were excluded from the analysis.
There was no association found between vitamin D and seroprotection.
Finally, participants who may not have had enough time to respond to the vaccine (less than 2 weeks between vaccine and serology) or were not appropriately vaccinated (i.e. only 1 vaccine dose in participants less than 9 years of age) were excluded (n = 11). There was no appreciable change in the results.

Discussion
The main finding of this study was that serum 25(OH)D level was not significantly associated with immunogenicity as measured by seroprotection (post-vaccination titer $40 or $320) and seroconversion. Serum 25(OH)D level was also not associated with other commonly reported measures of vaccine immunoge- nicity, including presence of a four-fold rise in antibody titer, the fold change in GMT or post vaccination titer. This finding is consistent with three randomized controlled trials of vitamin D supplementation in influenza vaccinated subjects, one in healthy adults [15], one in children [26] and the other in HIV-infected adults [16]. All three studies found no effect of supplementation on serologic responses. It is also consistent with a recent prospective cohort study in adults [27]. We also did not find an association between serum 25(OH)D level and any of the commonly used immunogenicity criteria. However, this failure to detect an association may be related to sample size and power.
Our findings are different from a recent study of influenza vaccination in prostate cancer patients (n = 35) which found that significantly more participants with a replete vitamin D status, defined as the upper quartile of serum 25(OH)D levels, had titers $1:40 at 3 months against any of the 3 strains [17]. They also reported an association between serum 25(OH)D levels and serologic response (p = 0.0446), but the magnitude of effect and confidence intervals were not provided. This trial was limited by the definition of serologic response (response to any of the three antigens) and the high serum 25(OH)D levels in the population (median 44.8 ng/mL<112 nmol/L). Another possible explanation for the divergent results is the different patient population. The effect of serum 25(OH)D level on vaccine immunogenicity may be different in the immunocompromised population, a group which tends to be more hyporesponsive to influenza vaccinations [3].
Overall, seroprotection and seroconversion proportions for the Influenza A strains were in keeping with the recommended standards [28]. A better seroconversion percentage and fold   [2]. However, we observed seroprotection and seroconversion percentages of 35% and 30%, respectively to B/Brisbane/60/2008-like virus. Only 2 subjects who had seroconversion had proven influenza B infection, therefore these changes in antibody titer represent either vaccine response or sub-clinical infection. Although these percentages are lower than recommended for vaccine licensure (seroprotection .60%, seroconversion .30%) [28], it does suggest some potential cross-protection if the antibody responses are related to vaccination.
Limitations of this study include the variability in the timing of both serum 25(OH)D levels and the post-vaccination titers. However, a recent study in a cohort of urban children in Toronto, found the variability in vitamin D levels from winter to summer months was less than 10 nmol/L [29]. In addition, two separate sensitivity analyses, first excluding participants with levels drawn more than 3 months after baseline levels/vaccination (n = 72, 32%) and second including the time from last vaccine dose as a variable in the analysis, revealed similar results to our primary analysis. Secondly, antibody titers were only available on study participants who agreed to the follow-up bloodwork (n = 221, 57%). This sample size limitation may have impacted our ability to detect a significant difference. With this sample size and 80% power, the minimum hazard ratio that the study was powered to detect was 2.78, a hazard ratio well above a clinically meaningful effect. In addition, we did not have the specimens to measure functional cell-mediated immune response. We also acknowledge that previous influenza vaccine history may have influenced the outcome of seroprotection. This was the reason for collecting baseline antibody levels and looking at multiple immunogenicity outcomes, including seroconversion. Finally, this study was conducted in Hutterite children and adolescents, which may limit generalizability to other populations.
In conclusion, we found that serum 25(OH)D level was not associated with immunogenicity in children and adolescents. However, sample size may have limited our ability to detect a significant difference and larger studies are warranted. Given the important role of children and adolescents in the spread of viral infections and the potential benefits from their immunization to protect other higher risk groups, other strategies should continue be evaluated to enhance the vaccine immune response. The role of serum 25(OH)D level in influenza vaccine immunogenicity in other populations, specifically immunocompromised patients, may warrant further study.