Inflammatory bowel disease (IBD) is a chronic immune-mediated condition of the gastrointestinal tract resulting from a complex interplay between the environment, immune system, and susceptibility genes.1 Immunosuppressive therapy is a mainstay of treatment for IBD. However, immunosuppressive therapy adds an increased risk of more frequent and severe infections2, 3 to the immune dysregulation already inherent in IBD. As infections may adversely impact the course of IBD, protection against vaccine-preventable infections is an important component of health maintenance. However, immunosuppression from medical therapy may alter vaccine response and clinical protection from infection. In addition, there is limited evidence about the effect of immune dysregulation inherent to IBD on vaccine response.
Influenza is usually an acute, self-limited, respiratory illness that occurs in annual outbreaks. As a leading infectious cause of death, seasonal influenza along with pneumonia kills about 4000 Canadians and 36,000 Americans annually.4–6 Influenza also leads to many thousands of hospitalizations and visits to emergency departments and physician offices as well as substantial absenteeism from work and school.7–9 Rates of morbidity and complications are higher among immunosuppressed populations such as transplant or chemotherapy recipients.10–12
The Public Health Agency of Canada recommends annual inactivated influenza vaccination for individuals at risk of complications of influenza infection,13 including children with IBD on immunosuppressive therapy; the Advisory Committee on Immunization Practices (USA) recommends annual influenza vaccination for all persons age 6 months and older, especially including immunosuppressed individuals.14 Vaccination usually results in an adaptive immune response of antibody production (including hemagglutination-inhibition [HAI] antibodies) 7 days after vaccination, peaking at ≈2-4 weeks.15 The main concern regarding vaccination of IBD patients on immunosuppressive therapy is the ability of a medically suppressed immune system to mount an appropriate serological response; however, as some IBD patients have concerns about vaccination side effects,16 the monitoring of adverse events from immunization is also important.
We evaluated immunogenic response and serologic protection to each component of the inactivated influenza vaccine in children with IBD and conducted comparisons of immunogenic response and serologic protection between children with IBD and healthy controls and between immunosuppressed children with IBD and nonimmunosuppressed children with IBD. We also explored the safety of influenza vaccination in children with IBD by recording adverse reactions and disease activity.
MATERIALS AND METHODS
This prospective, open-label, cohort study recruited participants (children with IBD and healthy sibling controls) from September 1, 2008 to January 1, 2009 through the Pediatric Gastroenterology Clinic at the Alberta Children's Hospital. The inclusion criteria for the IBD cohort were a diagnosis of IBD by accepted criteria17 and age between 2–17 years. The inclusion criteria for the control cohort were age between 2–17 years and no current use of immunosuppressive therapy or history of immunosuppressive medical condition. The exclusion criteria for both cohorts were: anaphylactic reaction to previous influenza vaccine; known hypersensitivity to eggs, chicken, or other components of the influenza vaccine; serious acute febrile illness; previous severe lower respiratory symptoms within 24 hours after influenza vaccination, an apparent allergic reaction to influenza vaccine or any other symptoms raising concern for reimmunization; already received 2008 seasonal influenza vaccine; or parent or legal guardian unwilling or unable to provide signed informed consent.
IBD participants were classified into two groups based on whether their IBD therapy included any immunosuppressive medications at the time of vaccination. Nonimmunosuppressed participants were on no medication, or receiving only antibiotics, 5-aminosalicylates, or probiotics. Immunosuppressed participants were receiving any of the following classes of medications: systemic corticosteroids (prednisone, prednisolone, methylprednisolone, budesonide), immunomodulators (azathioprine, 6-mercaptopurine, methotrexate, cyclosporine, tacrolimus), or biologics (infliximab, adalimumab). All immunosuppressed participants were on immunosuppressive therapy for at least 2 weeks prior to vaccination.
At the baseline visit for IBD participants, we collected demographic information, medical history (including IBD-related medications and influenza vaccination history), and physical exam data. Disease location and behavior were characterized according to the Montreal Classification.18 The Pediatric Crohn's Disease Activity Index19, 20 and Pediatric Ulcerative Colitis Activity Index21 measured disease activity for participants with Crohn's disease or ulcerative colitis / indeterminate colitis, respectively.
Baseline sera were collected for preimmunization HAI titers from all participants. In addition, blood was collected for complete blood count, erythrocyte sedimentation rate, C-reactive protein, and albumin from the IBD cohort.
All participants received the 2008 inactivated influenza vaccine (Fluviral®, GlaxoSmithKline, Mississauga, Ontario, or Vaxigrip®, Sanofi Pasteur, Toronto, Ontario) intramuscularly. This vaccine included A/Brisbane/59/2007 (H1N1)-like, A/Brisbane/10/2007 (H3N2)-like, and B/Florida/4/2006 antigens. A standard 0.5-mL dose of vaccine contained 15 μg hemagglutinin of each strain.22 In accordance with Public Health Agency of Canada recommendations, children <9 years of age received Vaxigrip for thimerosal reduction and children ≥9 years of age received either Fluviral® or Vaxigrip® according to availability.22 Children <9 years of age required two doses given ≥4 weeks apart if they were receiving the influenza vaccine for the first time, or if they had received only one dose during the previous influenza season as their first dose.22
We contacted all participants by telephone 72 hours postimmunization to document any adverse reactions. An adverse reaction was defined as a response which was more severe than usual, or was unusual or unexpected in character according to the Adverse Event Reporting Criteria of Alberta Health and Wellness operationalized in the Calgary Health Region manual.23
Participants returned for a follow-up visit 3–5 weeks after vaccination for sera collection for postimmunization HAI titers, along with measurement of disease activity and collection of IBD-related laboratory studies for IBD participants. For participants requiring two doses, the follow-up visit occurred 3–5 weeks after the second dose.
The Provincial Laboratory for Public Health (Microbiology), Calgary, Alberta, stored serum samples for HAI titers at −20°C before parallel testing against all three components in the 2008 vaccine. Unaware of the participant's diagnosis, immune status, or treatment, laboratory technicians and virologists performed the HAI assay according to standard methods described in the World Health Organization protocol.24 In brief, sera were treated with receptor-destroying enzyme and titrated using 4 HAI units of the appropriate influenza antigen with fowl erythrocytes as the indicator.
A 4-fold or greater increase between pre- and postimmunization titers indicated an immunogenic response; a postimmunization titer of 1:40 or greater indicated serologic protection.25
For the primary objective of estimating the proportion of children with IBD mounting an immunogenic response, a sample size of 80 participants with IBD would yield an estimate with a 95% confidence interval (CI) width of 21%; this was based on an estimate of 72% mounting an immunogenic response derived from studies on influenza vaccination in children with IBD or organ transplant recipients.26–30
For a secondary objective of comparing the proportions mounting an immunogenic response between children with IBD and controls, an estimated sample size was calculated based on the assumption of a 23% difference between the proportions of children with IBD (72%) and controls (95%) mounting an immunogenic response.26, 27 Based on a power of 80% and an alpha of 0.05, 48 participants would be required in each group.
This study was not powered to assess adverse events or compare immunogenic response or serologic protection between different classes of immunosuppressive therapy.
All statistical tests of significance were two-sided with an alpha of 0.05. Analysis was performed using Intercooled Stata software (College Station, TX). We described demographic characteristics, diagnosis, previous influenza vaccination history, medications, disease activity, laboratory data, and Montreal Classification for all participants with IBD. The t-test or Mann–Whitney U-test (continuous variables) and the χ2 test or Fisher's exact test (dichotomous variables) were used to compare IBD and control cohorts and also to compare immunosuppressed and nonimmunosuppressed IBD groups.
The proportion of participants with IBD mounting an immunogenic response to each vaccine component was estimated along with the corresponding exact binomial 95% CI. We compared the proportions mounting an immunogenic response between IBD and control cohorts and between immunosuppressed and nonimmunosuppressed IBD groups using Fisher's exact test or χ2 test. We repeated these analyses with serologic protection as the outcome measure. Multiple comparisons were conducted without adjustment of the P-value. Where adequate numbers permitted, multivariate logistic regression was performed to evaluate the effect of the variables of age, IBD type, and vaccine type on immunogenic response and serologic protection.
To explore the impact of influenza vaccination on disease activity and IBD-related parameters, we compared pre- and postimmunization levels with the paired t-test or Wilcoxon signed-rank test. For the exploratory analyses, all P-values were presented for description only.
The Conjoint Health Research Ethics Board of the Faculty of Medicine, University of Calgary approved our study.
We enrolled 61 IBD participants and 55 sibling controls; 60 IBD participants (98%) and 53 sibling controls (96%) completed pre- and postimmunization HAI titers (Fig. 1).
In the IBD cohort, only 30 participants (50%) reported any previous influenza vaccinations (Table 1). The median age at enrolment in the IBD cohort was 15.4 years and only one participant with IBD required two vaccine doses. At enrolment and during the study, 42 IBD participants (70%) were on immunosuppressive therapy: two on systemic corticosteroids alone (for at least 1.5 months prior to enrolment); 32 on azathioprine, 6-mercaptopurine, or methotrexate; none on cyclosporine or tacrolimus; and eight on biologic therapy (including two on combination infliximab and methotrexate therapy). All participants remained in their respective immunosuppression categories throughout the duration of the study. Immunosuppressed participants were more likely to have Crohn's disease and nonimmunosuppressed participants were more likely to have ulcerative colitis (P = 0.003).
Table 1. Baseline Characteristics of Participants with Inflammatory Bowel Disease
|Gender male, n (%)||34 (57%)||24 (57%)||10 (56%)||0.9|
|Median age in years (Q1, Q3)|| || || || |
| At enrolment||15.4 (10.5, 16.4)||13.2 (10.4, 15.7)||15.2 (13.8, 17.2)||0.9|
| At diagnosis||11.7 (9.7, 14.7)||12.2 (10.3, 15.3)||11.1 (8.5, 14.3)||0.2|
|Previous influenza vaccination, n (%)||30 (50%)||19 (45%)||11 (61%)||0.4|
|Prevaccine serologic protection, n (%)|| || || || |
| H3N2||24 (40%)||14 (33%)||10 (56%)||0.1|
| H1N1||30 (50%)||20 (48%)||10 (56%)||0.6|
| B/Florida||22 (36%)||14 (33%)||8 (44%)||0.4|
|Diagnosis|| || || || |
| Crohn's disease||26 (43%)||23 (55%)||3 (17%)||0.003|
| Ulcerative colitis||24 (40%)||11 (26%)||13 (72%)|| |
| Indeterminate colitis||10 (17%)||8 (14%)||2 (11%)|| |
|Montreal classification|| || || || |
|Crohn's disease (n = 26)|| || || || |
| Disease location|| || || || |
| L1 (terminal ileum)||2 (8%)||1 (4%)||1 (33%)|| |
| L2 (colon)||9 (35%)||7 (30%)||2 (67%)|| |
| L3 (ileocolon)||15 (58%)||15 (65%)||0|| |
| L4 (isolated upper disease)||0||0||0|| |
| Disease behavior|| || || || |
| B1 (nonstricturing, non penetrating)||22 (85%)||19(83%)||3 (100%)|| |
| B2 (stricturing)||4 (15%)||4 (17%)||0|| |
| B3 (penetrating)||0||0||0|| |
| Perianal disease||13 (50%)||12 (52%)||1 (33%)|| |
|UC or IC (n = 34)|| || || || |
| Disease extent|| || || || |
| E1 (proctitis)||0||0||0|| |
| E2 (left-sided colitis)||4 (12%)||3 (16%)||1 (7%)|| |
| E3 (pancolitis)||30 (88%)||16 (84%)||14 (93%)|| |
|Disease Activity Index, n (%)|| || || || |
| Inactive||32 (53%)||22 (52%)||10 (56%)||0.6|
| Mild||17 (28%)||13 (31%)||4 (22%)|| |
| Moderate||9 (15%)||5 (12%)||4 (22%)|| |
| Severe||2 (3%)||2 (5%)||0|| |
In the control cohort, only 24 participants (45%) reported any previous influenza vaccinations (Table 2). The control cohort was younger (median age at enrolment 12.3 years) than the IBD cohort (median age at enrolment 15.4 years, P < 0.0001). Five controls required two vaccine doses.
Table 2. Comparison of Baseline Characteristics Between Participants with Inflammatory Bowel Disease and Controls
|Gender male, n (%)||34 (57%)||34 (64%)||0.4|
|Median age in years (Q1, Q3)|| || || |
| At enrolment||15.4 (10.5, 16.4)||12.3 (9.9, 13.9)||<0.0001|
|Previous influenza vaccination, n (%)||30 (50%)||24 (45%)||0.8|
|Prevaccine serologic protection, n (%)|| || || |
| H3N2||24 (50%)||21 (40%)||0.3|
| H1N1||30 (50%)||28 (53%)||0.8|
| B / Florida||22 (36%)||10 (19%)||0.04|
Forty-nine and nine participants with IBD received the Fluviral® and Vaxigrip® vaccines, respectively; 19 and 33 controls received the Fluviral® and Vaxigrip® vaccines, respectively.
The proportion of IBD participants who achieved an immunogenic response to the A/Brisbane/10/2007 (H3N2), A/Brisbane/59/2007 (H1N1), and B/Florida/4/2006 components were 70% (95% CI 57%–81%), 72% (95% CI 59%–83%), and 53% (95% CI 40%–66%), respectively (Table 3). IBD participants were less likely to achieve an immunogenic response to B/Florida/4/2006 than to A/Brisbane/59/2007 (H1N1) (P = 0.03).
Table 3. Immunogenic Response and Serologic Protection
|All IBD (n = 60)||A/Brisbane/10/2007 (H3N2)||42 (70%)||57 (95%)|
|A/Brisbane/59/2007 (H1N1)||43 (72%)||59 (98%)|
|B/Florida/4/2006||32 (53%)||51 (85%)|
| IBD, No IT (n = 18)||A/Brisbane/10/2007 (H3N2)||12 (67%)||17 (94%)|
|A/Brisbane/59/2007 (H1N1)||11 (61%)||17 (94%)|
|B/Florida/4/2006||9 (50%)||18 (100%)|
| IBD, Any IT (n = 42)||A/Brisbane/10/2007 (H3N2)||30 (71%)||40 (95%)|
|A/Brisbane/59/2007 (H1N1)||32 (76%)||42 (100%)|
|B/Florida/4/2006||23 (55%)||33 (79%)|
|Controls (n = 53)||A/Brisbane/10/2007 (H3N2)||44 (83 %)||51 (96%)|
|A/Brisbane/59/2007 (H1N1)||40 (76%)||52 (98%)|
|B/Florida/4/2006||43 (81%)||50 (94%)|
Comparing IBD participants to controls, there was no difference in the proportion achieving an immunogenic response for A/Brisbane/10/2007 (H3N2) or A/Brisbane/59/2007 (H1N1) components (Table 3). However, a lower proportion of IBD participants (53%, 95% CI 40%–66%) developed an immunogenic response to B/Florida/4/2006 component compared to controls (81%, 95% CI 68%–91%, P = 0.002) (Table 3). Comparing immunosuppressed to nonimmunosuppressed IBD participants, there was no significant difference in the proportion achieving immunogenic response for any of the vaccine components (Table 3). In the multivariate analysis, age, IBD type, and vaccine type were not predictors of immunogenic response.
The proportion of IBD participants who achieved postimmunization serologic protection against A/Brisbane/10/2007 (H3N2), A/Brisbane/59/2007 (H1N1), and B/Florida/4/2006 components were 95% (95% CI 86%–99%), 98% (95% CI 91%–100%), and 85% (95% CI 73%–93%), respectively (Table 3). Participants with IBD were less likely to develop serologic protection to B/Florida/4/2006 than to A/Brisbane/59/2007 (H1N1) (P = 0.01).
The proportion of IBD participants who achieved serologic protection against any of the vaccine components was not significantly different than controls. Comparing immunosuppressed to nonimmunosuppressed IBD participants, there was no significant difference in the proportion achieving serologic protection for A/Brisbane/10/2007 (H3N2) and A/Brisbane/59/2007 (H1N1) components. However, a lower proportion of immunosuppressed IBD participants (79%, 95% CI 63%–90%) achieved serologic protection to B/Florida/4/2006 compared to nonimmunosuppressed IBD participants (100%, 95% CI 90%–100%, P = 0.03). Prevaccine disease activity did not impact the relationship between impaired serologic protection and immunosuppression for children with IBD. For IBD participants, prevaccine serologic protection was present against A/Brisbane/10/2007 (H3N2), A/Brisbane/59/2007 (H1N1), and B/Florida/4/2006 components in 40%, 50%, and 36%, respectively (Table 1). IBD participants were more likely to have prevaccine serologic protection against the B/Florida/4/2006 component compared to controls (36% versus 19%, P = 0.04) (Table 2).
Impact of Influenza Vaccine on Disease Activity
Only one participant with IBD experienced a clinically significant increase in disease activity score (from 15 prevaccination to 35 postvaccination). She was a 14-year-old with pancolonic ulcerative colitis whose disease worsened within 6 days postvaccination, requiring an increased prednisone dose and start of oral 5-aminosalicylate. She had a recent IBD exacerbation 1.5 months prior to vaccination which was treated with prednisone. At the time of vaccination, she was on a prednisone tapering schedule (20 mg orally once daily) without any maintenance medication. Because of her recent exacerbation, tapering prednisone dose, and lack of maintenance medication, it is difficult to determine the degree of contribution of the influenza vaccination to the worsening disease activity.
No IBD participants required hospitalization or surgery related to IBD in the 4 weeks postvaccination.
Prevaccination Pediatric Crohn's Disease Activity Index and Pediatric Ulcerative Colitis Activity Index scores were not significantly different from postvaccination disease activity scores (Table 4). There was no significant change from pre- to postvaccination IBD-related laboratory parameters for participants with IBD (Table 4).
Table 4. Comparison of Pre- and Postvaccine Disease Activity and Laboratory Parameters
| PCDAI, median (Q1, Q3) [n = 21]||10 (5, 15)||10 (5, 15)||0.7|
| PUCAI, median (Q1, Q3) [n = 33]||10 (0, 25)||5 (0, 20)||0.08|
| Hemoglobin (mg/dL), mean (95% CI)||132 (129–136)||133 (130–137)||0.3|
| Platelet count* (x 109 /L), mean (95% CI)||333 (309–357)||344 (319–369)||0.2|
| Erythrocyte sedimentation rate* (mm/h), median (Q1, Q3)||9 (2, 20)||8 (3, 17)||0.8|
| C-reactive protein* (mg/L), median (Q1, Q3)||1.6 (0.4, 6.7)||1.5 (0.5, 4.8)||0.9|
Adverse Reactions of Influenza Vaccine
Only one IBD participant developed a reportable adverse reaction. A 15-year-old male with ileocolonic Crohn's disease on maintenance adalimumab developed clinically significant pancreatitis 84 hours postvaccination requiring 4 days of hospitalization with supportive care. As he had two prior episodes of idiopathic pancreatitis, the postvaccination episode was thought to be temporally associated but likely not causally related to the influenza vaccine. No controls developed any reportable adverse reactions.
A high proportion of children with IBD, similar to that observed in controls, mounted an appropriate immunogenic response to the two influenza A subtype components of the 2008 influenza vaccine (70% for H3N2, 72% for H1N1). However, only 53% of children with IBD mounted an immunogenic response to the influenza B component compared to 81% of controls. For serologic protection, a high proportion of children with IBD achieved serologic protection to all three components of the influenza vaccine (95% for H3N2, 98% for H1N1, 85% for influenza B). For the influenza B component only, immunosuppressive medication use in children with IBD affected the achievement of serologic protection.
For the influenza B component, the control cohort was less likely to have prevaccine serologic protection compared to the IBD cohort, as might have been expected given the younger age of the control cohort (i.e., less opportunity for prior exposure to disease or vaccine). Despite a higher proportion of children with IBD possessing prevaccine serologic protection, a similar proportion of controls achieved postvaccine serologic protection compared to children with IBD.
We selected immunogenic response (4-fold or greater increase in HAI titers from pre- to postimmunization) and serologic protection (postimmunization HAI titer ≥1:40) as outcome measures of immunogenicity because of their indication of the immune system's ability to respond to vaccination and their use as markers for protection against disease.25 Indeed, the ideal outcome for a vaccine study is clinical protection from disease (i.e., an efficacy study); however, the requirements for an efficacy study include a much larger sample size and longer duration of follow-up. Currently, no study has evaluated the frequency of influenza infections postvaccination in children with IBD, although decreased immunogenic response to the influenza B vaccine component may be suspected to increase susceptibility to influenza B infection.
Although two studies26, 27 recently evaluated immune responses to influenza vaccination in children and young adults with IBD, the unique qualities of our study include a larger control group of children without IBD or immunosuppression and the addition of the outcome immunogenic response, as defined by a 4-fold or greater increase in HAI titer. Similar to our study, prior studies found that children with IBD mounted appropriate serologic protection to influenza A components. However, Mamula et al26 demonstrated that children with IBD were less likely to mount serologic protection against the B/Hong Kong/330/2001 component of the 2002–2004 influenza vaccine compared to healthy controls (64% versus 90%). Likewise, Lu et al27 showed that only 39% of children with IBD developed serologic protection against the B/Malaysia/2506/2004 component of the 2007 influenza vaccine compared to 96% and 88% for A/Solomon Islands/3/2006 (H1N1) and A/Wisconsin/67/2005 (H3N2), respectively.
Although the differing vaccines of each influenza season in our study and the two prior studies26, 27 limit the ability to make direct comparisons, the results taken together suggest that children with IBD have a decreased ability to mount serologic protection against the B component of the influenza vaccine. This appears to be further diminished with the use of immunosuppressive medications.
The decreased immunogenicity of the influenza B vaccine component compared to the A subtype components is not unique to children with IBD, as similar results have been shown in healthy children31, 32 and in immune-mediated conditions such as rheumatoid arthritis.33 HAI titers may be affected by previous exposure to the same or similar strains of influenza virus. The influenza vaccines in the previous 5 years did not contain any of the strains of the 2008 vaccine; however, the most common circulating strains in the 2007 influenza season in Canada identified by the National Microbiology Laboratory included all three 2008 vaccine strains.22 As influenza A subtypes exhibit greater crossreactivity, individuals may be more primed to develop immunogenicity due to previous exposure to not only the same, but also to similar influenza A subtypes by vaccination or infection; this may explain the higher rates of immunogenicity against influenza A compared to influenza B.
Other studies have also demonstrated decreased immunogenicity to influenza vaccines in individuals with immune-mediated conditions or on immunosuppressive medications, including adults with rheumatoid arthritis, adults with systemic lupus erythematosus, adult liver transplant recipients, and children with liver, kidney, or heart transplants.28–30, 34–37
Studies of other vaccines in IBD have shown impaired adaptive immunity to vaccination, including impaired antibody responses to tetanus toxoid booster in adults with IBD, and to oral cholera and S. typhi vaccinations in adults with ulcerative colitis postcolectomy.38–41 Impaired immune response to the pneumococcal polysaccharide vaccine in adults with IBD on immunosuppressive therapies has also been reported.42 Therefore, immune dysregulation inherent to IBD or from immunosuppressive medications or surgical resections altering anatomical exposures to oral vaccines may weaken immune responses to a variety of vaccines in IBD.
A potential future strategy for individuals with IBD, especially those on immunosuppressive therapy, may be administration of two doses of influenza vaccine during an influenza season to boost immune response and achieve serologic protection. In immunosuppressed adult liver transplant recipients, two doses of influenza vaccine administered 28 days apart improved HAI titers; serologic protection against all three influenza strains among recipients increased from 68% after the first dose to 80% after the second dose.37 A recent review on vaccination strategies for patients with IBD recommends ascertainment of an adequate immune response for immunosuppressed patients whenever possible, and consideration of repeat dosing when immune response to immunization is insufficient.43 However, although the Public Health Agency of Canada advises for close monitoring of immunocompromised individuals after vaccination and aggressive boosting to optimize magnitude and duration of vaccine-induced immunity, it currently reports insufficient evidence to recommend two doses of influenza vaccine during an influenza season to boost immunity in immunocompromised individuals.13
In our exploration of the safety of influenza vaccination in children with IBD, the vaccine appeared to be well tolerated. Only one child developed a reportable adverse reaction which required hospitalization. However, because of his previous history of pancreatitis, the adverse event was not conclusively attributed to influenza vaccination. Prior studies reported no serious adverse events in individuals with IBD.26, 27 The remainder of participants did not require any IBD-related hospitalizations or surgeries in the 4 weeks postvaccination. We observed only one child with a clinically significant worsening of disease activity, similar to observations made by others.26, 27 However, because of our participant's recent disease flare, tapering prednisone schedule, and lack of maintenance medication, it is difficult to determine whether the worsening disease activity was attributed to the influenza vaccination or to the aforementioned factors.
Although we aimed to enroll 80 participants with IBD, 60 participants with IBD ultimately completed the study. The reasons for refusal of study participation included: unable to allocate time for clinic visits, fear of bloodwork, lack of belief in effectiveness of influenza vaccination, and stable or unstable health status. In particular, 51 of the children with IBD refusing study participation (50%) resided outside of the City of Calgary. However, the 95% CI width for estimates of proportion achieving immunogenic response or serologic protection varied minimally for a sample size of 60 subjects compared to 80 subjects. Our findings may be generalized to the wider population of interest, children with IBD, on account of similarities in characteristics of diagnosis, disease classification and activity, and medication use. However, although we attempted to enroll children between 2 to 18 years of age, the majority of participants with IBD were over 9 years of age. Children less than 9 years of age have specific requirements of two doses of influenza vaccine based on influenza immunization history according to current Canadian guidelines22; only 1 of 60 children with IBD required two doses in our study. Therefore, because of small numbers and special dosage schedules for children less than 9 years of age, our findings should be generalized only to children with IBD over 9 years of age.
The proportion of participants reporting any previous influenza vaccination was low: 50% in children with IBD (including 45% and 61% in those on and not on immunosuppressive therapy, respectively) and 45% in controls. This occurred despite recommendations for annual influenza vaccination by the Public Health Agency of Canada for immunosuppressed individuals13 and by the Advisory Committee on Immunization Practices (USA) for all persons 6 months of age and older.14 In addition, guidelines for individuals with IBD proposed that immunizations do not deviate from recommended schedules for the general population and that most immunizations, except for live agent vaccines, may be safely administered, even when immunosuppressed.43–45 We were unable to determine the proportion of patients with IBD with any previous influenza vaccinations among those followed at the Pediatric Gastroenterology Clinic; however, vaccination records from 34 nonparticipants with IBD followed at the Pediatric Gastroenterology Clinic revealed that only nine (26%) had ever received any prior influenza vaccination. Among adults with IBD, only 28% reported receiving regular influenza vaccinations.16 In contrast, among children with IBD, other investigators observed that 73% reported receiving any previous influenza vaccine27; however this higher proportion may be due to self-selection into the study sample. Nevertheless, the low proportion of IBD patients with previous influenza vaccinations demonstrates the need for physicians to address the risk of infections and benefit of immunizations in patients with IBD, especially those on immunosuppressive medications.
In conclusion, children with IBD develop appropriate immunogenic responses and serologic protection to the influenza A vaccine components; however, immunogenic responses and serologic protection to the influenza B component appear to be impaired, especially with use of immunosuppressive therapy. The influenza vaccination was safe and well tolerated.
The authors thank the families who participated in this study. The influenza vaccine was publicly funded and provided by Alberta Health and Wellness through the Calgary Health Region. Author contributions: J.C.D. was involved in study concept and design, data acquisition, analysis and interpretation of data, statistical analysis, drafting of the article critical revision of the article for important intellectual content, and obtaining funding. R.H., M.L.R., and G.K. were involved in study concept and design, drafting of article, analysis and interpretation of data, statistical analysis, and critical revision of the article for important intellectual content. K.F. was involved in study concept and design, acquisition of data, analysis and interpretation of data, and critical revision of the article for important intellectual content. OV was involved in study concept and design, drafting of article, critical revision of the article for important intellectual content, and obtaining funding. IW was involved in study concept and design, data acquisition, analysis and interpretation of data, drafting of article, critical revision of the article for important intellectual content, obtaining funding, and study supervision.