Background Influenza vaccination is recommended for patients with inflammatory bowel disease (IBD). The 2009 H1N1 influenza vaccine produced seroprotection rates of >85% in the general population but there are no data on the immunogenicity of the vaccine in patients with IBD.
Methods An observational prospective open-label study was conducted to examine the immunogenicity of the 2009 H1N1 influenza vaccine in 108 patients with IBD. Patient details, medications and disease activity were recorded. Pre- and post-vaccination haemagglutinin inhibition titres and geometric mean titres were measured. A functional assay of T lymphocyte activity was measured at vaccination in a subset of patients as an alternative measure of immunosuppression. Subjects were followed for 6 months post-vaccination.
Results Of 108 patients enrolled, 105 completed the study. The post-vaccination seroprotection rate was 50%. Immunosuppressed subjects had a lower rate of seroprotection than non-immunosuppressed subjects (44% vs 64%, p=0.06). The proportion with seroprotection was significantly lower in subjects on combination immunosuppression than in those receiving no immunosuppression (36% vs 64%, p=0.02). Patients receiving combined immunosuppression had a significantly lower fold increase in geometric mean titres than those on monotherapy immunosuppression (3.5 vs 11.5, p=0.03). An assay of T lymphocyte activity was performed in a subgroup of 48 subjects. Those with intermediate activity had lower seroprotection than those with high activity (28% vs 61%, p=0.02). The vaccine was well tolerated.
Conclusions Patients with IBD vaccinated with the 2009 H1N1 influenza vaccine had a low rate of seroprotection, particularly among those who were immunosuppressed. Although there is a need for studies of the clinical benefit of vaccines in this population, patients with IBD need to be aware of this reduced immunogenicity.
- Crohn's disease
- ulcerative colitis
- inflammatory bowel disease
- IBD clinical
Statistics from Altmetric.com
- Crohn's disease
- ulcerative colitis
- inflammatory bowel disease
- IBD clinical
Significance of this study
What is already known about this subject?
When the WHO declared a global pandemic of novel influenza A (H1N1) in 2009, there was particular concern for infection in immunocompromised patients, including those with inflammatory bowel disease (IBD).
Routine influenza vaccination is recommended for all patients with IBD despite variable data on serological response and no data on clinical response in this group.
What are the new findings?
Patients with IBD vaccinated with the 2009 H1N1 influenza vaccine had a low rate of seroprotection, particularly those who were immunosuppressed.
Patients with IBD receiving combination immunosuppression had the lowest rates of serological response to the vaccine.
An assay of T lymphocyte function may indicate those at risk of low rates of vaccine response.
How might it impact on clinical practice in the foreseeable future?
Patients with IBD and the physicians who care for them need to be aware of the potential for a suboptimal response to influenza vaccine. Although the level of serological response required to prevent clinical infection in this population has yet to be defined, strategies for assessing vaccine response in high-risk individuals may be necessary, particularly in the event of a pandemic infection. Assays of T lymphocyte activity may have a role to play in identifying patients at increased risk.
The World Health Organization (WHO) declared a global pandemic of novel influenza A (H1N1) in 2009 with over 70 countries reporting H1N1 outbreaks. The spread of H1N1 influenza was expected to increase dramatically during the influenza season of 2009/10. The WHO specifically identified age <65 years, immunosuppression and chronic medical conditions as being specific risks for H1N1 influenza infection (http://www.who.int). All three risk factors are relevant to inflammatory bowel disease (IBD).
Patients with IBD are at increased risk of infection for a number of reasons. The initiating events in IBD may be due to immune deficiency.1 Distinct immunological abnormalities play a role in both the initiation and perpetuation of Crohn's disease (CD) and ulcerative colitis (UC). Some patients with IBD have impaired cell-mediated immunity2 as well as impaired responses to tetanus toxoid.3 In addition to innate immune defects in IBD, patients with these conditions may be treated with medications that further alter the immune response. Drugs such as corticosteroids, thiopurines, methotrexate and tumour necrosis factor (TNF) antagonists (infliximab, adalimumab and certolizumab) are associated with increased risk of infection.4 This infection risk is not confined to bacterial infection and a number of severe and even fatal viral infections have been reported in patients taking these agents.5–7
These concerns lead to the recommendation that patients with IBD receive routine vaccinations,8 although the immunogenicity and clinical efficacy of specific vaccines in the adult IBD population are unknown. Immunosuppressed patients may have a suboptimal response to the influenza vaccine. There are data illustrating this in both transplant and chemotherapy patients.9–11 Mamula et al reported that children with IBD treated with immunosuppressants mounted an inadequate immune response to one of the three strains in the 2002–4 inactivated influenza vaccines compared with healthy controls.12 This has not been a universal finding; a study by Lu et al, also in children, identified a high level of seroprotection in immunised patients with IBD, an effect that was not altered by immunosuppressant medications.13 A Dutch study that included patients with rheumatoid arthritis and CD identified lower post-vaccination geometric mean titres (GMTs) against influenza (A/H3N2 and B) in those taking anti-TNF agents than in non-immunosuppressed individuals, but the overall seroprotection rates were similar between groups, indicating a quantitative but not a qualitative effect.14
The H1N1 pandemic offered an opportunity to examine the functional level of immunosuppression as well as the immunogenicity of a monovalent vaccine in an IBD population. We conducted an open-label prospective study to examine the serological response to the H1N1 influenza vaccine in a cohort of adult patients with IBD.
This was an observational prospective study of adult (≥18 years) patients with IBD vaccinated for the 2009 H1N1 influenza virus at the Crohn's and Colitis Center at the Massachusetts General Hospital between December 2009 and March 2010. As standard of care, patients with IBD attending the Center for a physician visit, routine laboratory tests or infliximab infusion were offered vaccination against H1N1 influenza. In addition, patients enrolled in the Center's prospective IBD database of almost 1000 subjects (PRISM) were emailed and offered vaccination.
All patients receiving the vaccine at the Center were eligible for inclusion in the study. Patients who had previously been vaccinated against 2009 H1N1 influenza or those with a documented H1N1 influenza infection were excluded. Patients with a known allergy to eggs or other components of the vaccine were excluded.
Written informed consent was obtained from all subjects at enrolment and the study was approved by the Massachusetts General Hospital Institutional Review Board. Patients were asked to give details regarding previous influenza vaccinations or influenza infection. Clinical activity scores (Harvey Bradshaw Index (HBI) for CD or Simple Clinical Colitis Activity Index (SCCAI) for UC) were recorded for each subject. An HBI score of ≥8 for CD and a SCCAI of ≥5 for UC were taken as indicators of active disease. Medical records were reviewed to retrieve information on IBD diagnosis, duration of disease, current and past treatments and significant medical history.
Prior to vaccination, blood was drawn from the antecubital fossa for measurement of haemagglutinin inhibition (HI) titre and functional T lymphocyte activity. Serum was stored at −80°C for HI measurement. Whole blood for the T lymphocyte activity assay was stored at room temperature and sent for analysis within 24 h of the blood being taken. Subjects were then administered a single 0.5 ml dose of non-adjuvant influenza A (H1N1) 2009 monovalent vaccine (Sanofi Pasteur, Lyon, France) by intramuscular (deltoid) injection. This vaccine contains 15 μg haemagglutinin of influenza A/California/07/2009 (H1N1) v-like virus per 0.5 ml dose. In addition, each 0.5 ml dose contains residual amounts of formaldehyde (not more than 100 μg), polyethylene glycol p-isooctylphenyl ether (not more than 0.02%) and sucrose (not more than 2%). Thimerosal is not used in the manufacture of this vaccine.
Subjects returned between 4 and 10 weeks after vaccination for repeat HI measurement and to provide details on adverse events related to the vaccine, influenza-like symptoms and IBD activity (repeat HBI or SCCAI). Blood was drawn for repeat HI measurement and stored at −80°C.
Six-month post-vaccination contact
After 6 months, subjects were contacted by telephone or email to enquire about adverse events related to the vaccine, episodes of influenza and IBD activity. In addition, subjects were asked to notify the investigators at any time during the 6-month follow-up period if they experienced any of these events.
Definition of immunosuppression
Immunosuppression was defined by the level of clinical immunosuppression in all subjects and a functional assay of T lymphocyte activity in a subset of 48 subjects. Clinical immunosuppression was defined as treatment with glucocorticoids (prednisone 20 mg/day equivalent for 2 weeks or more or within 3 months of stopping), weight-appropriate doses of 6-mercaptopurine/azathioprine (or within 3 months of stopping), 15 mg (or more) methotrexate per week (or within 3 months of stopping) and, finally, treatment with a biological agent (infliximab, adalimumab, certolizumab pegol or natalizumab) or within 3 months of stopping one of these agents.
In addition to using patients' medications as an indicator of immunosuppression, we measured a functional assay of T lymphocyte activity (Immuknow, Cylex Inc, Columbia, Maryland, USA) in a subset of subjects. This assay is used in the transplant population where it is proposed as a surrogate measure of immune cell function to monitor immunosuppressive therapy, independent of the particular medication used (data at http://www.cylex.net). The assay is a measure of CD4+ T lymphocyte ATP levels. In transplant populations, low or intermediate levels (defined as ≤225 ng/ml and 226–524 ng/ml, respectively) indicate immunosuppression, whereas high levels (≥525 ng/ml) are defined as normal.
Haemagglutinin inhibition and functional T lymphocyte activity
All samples were labelled with unique patient identifiers so that both laboratories were blinded to the identity and clinical details of the subjects. Samples were sent on dry ice in a batch to the Laboratory for Specialized Clinical Studies at Cincinnati Children's Hospital Medical Center for the HI. Pre- and post-vaccination serum samples for each individual study subject were tested simultaneously. The serum samples were assessed for antibody to the component of the vaccine by HI assay using standard methods.13 15 Egg-derived inactivated viral antigen (pdH1N1) representative of the 2009/10 vaccine were obtained from the Centers for Disease Control and Prevention. Serum samples without reactivity were assigned a value of <10. Those with initial titres of ≥2560 were retested at a higher starting dilution to obtain a reportable titre.
The Immuknow assay of functional T lymphocyte activity was measured in 50 subjects at the time of the initial visit. The initial plan had been to perform this in all enrolled subjects but, due to a delay in acquiring the assay and the knowledge that we could not delay patients' vaccination, we proceeded with the study and the assay became available after 58 subjects had been enrolled. Two of the three subjects excluded for lack of follow-up had Immuknow performed. The assay was performed by the Molecular Diagnostics Center, Tufts Medical Center, Boston, Massachusetts, USA. The Immuknow kit was provided by Cylex Inc.
The primary outcome was seroprotection (defined as HI titre ≥40) to A/California/07/2009 (H1N1) virus. The secondary outcome was the post-vaccination GMTs, which provide a quantitative measure of vaccine response. Additional outcomes included the effect of vaccination on disease activity (measured by increases in HBI or SCCAI of >2 points or a requirement for additional treatment) and adverse events related to vaccine administration.
Demographic characteristics—including IBD history, treatment, current disease activity and Montreal classification—were recorded for the non-immunosuppressed (NIS) and immunosuppressed (IS) groups. For continuous variables, differences between groups were measured using t tests. Tests for normality (Kolmogorov–Smirnov) were applied and median values and non-parametric tests (Mann–Whitney U) were used for data that were not normally distributed. For dichotomous variables, differences between groups were compared using χ2 or Fisher exact tests. Multivariate analysis was used to identify factors that predict response or lack of response to the vaccine.
Evidence of serological response to the vaccine (HI ≥40) was recorded for each subject and comparison made between the NIS and IS groups. We examined the difference in proportions between the two groups using the Fisher exact test in all subjects and also (on post-hoc analysis) in subjects who did not have evidence of pre-vaccination seroprotection. The geometric mean of the HI titres was calculated pre- and post-vaccination. The fold increase in the HI titre was calculated by dividing the post-vaccination titre by the pre-vaccination titre and calculating the geometric mean of these values. As with the HI analysis, GMTs were examined in the total group and in the subgroup that did not have pre-vaccination seroprotection.
Based on a χ2 test, an α level of 0.05 and a SD for seroprotection in the normal population of 40%, 150 subjects were planned to yield a power of 80% to detect a difference of 16%. A difference of 16% was calculated based on responses in NIS and IS patients with IBD in previous studies.12 13 The aim was to recruit 75 patients with IBD taking immunosuppressants (as defined above) and 75 patients with IBD taking either no medications or medications that do not have a direct effect on the immune system (eg, aminosalicylates, aminosalicylate or steroid enemas, antibiotics).
One hundred and eight subjects were enrolled in the study between December 2009 and March 2010. Three did not return for the follow-up HI and were excluded, leaving 105 subjects (55 men) for the analysis. We failed to reach the target of 150 subjects, primarily because the primary care centres in the surrounding area had access to the H1N1 influenza vaccine for almost a month before the study centre. Fifty-seven patients had CD and 48 had UC. Seven subjects with CD and four with UC had active disease at the time of enrolment (by disease activity indices). Seventy-seven subjects were taking immunosuppressant medication at the time of vaccination (IS group) while 28 were taking no medication or aminosalicylates or antibiotics (NIS group). Although our definition of immunosuppression allowed for patients who had been on immunosuppressants in the previous 3 months, all those in the IS group were actively taking the immunosuppressant medication at the time of enrolment. The median time between vaccination and the follow-up visit at which the HI was measured was 5.3 weeks (range 3–11). One patient had follow-up at 3 weeks and another at 11 weeks, while the remaining 103 were within the range specified in the protocol. The baseline characteristics of the total group and of the IS and NIS groups are shown in table 1.
Proportion of subjects with seroprotection after vaccination
The overall proportion with seroprotection (HI ≥40) was 50% (52/105); 64% (18/28) in the NIS group and 44% (34/77) in the IS group (p=0.06, table 2). The percentage of post-vaccination seroprotection was significantly lower in subjects on combination immunosuppression than in those on no immunosuppression (p=0.02). The proportion with seroprotection was lower in those taking combined immunosuppression (two or more of prednisone, thiopurine, methotrexate or a biological drug) than in those on monotherapy immunosuppression (36% (5/14) vs 42–47% (11/26–16/34)), but this did not reach statistical significance (p=0.06).
Three subjects had an HI titre ≥40 pre-vaccination. A subgroup analysis was performed to examine post-vaccination seroprotection rates in the 102 subjects who did not have pre-vaccination seroprotection. In this group the post-vaccination seroprotection rate was 48% (49/102), 64% (18/28) in the NIS group and 42% (31/74) in the IS group (p=0.04). As before, those on combined immunosuppression had a lower response rate than those on a single immunosuppressive agent (42–46% (11/26–15/33) for monotherapy vs 25% (3/12) for combined immunosuppression), but this did not reach statistical significance (p=0.2).
Post-vaccination geometric mean titres
The overall post-vaccination GMT was 23.8 (95% CI 15.9 to 35.7), representing a fold increase from the pre-vaccination level of 11.4 (table 3). The post-vaccination GMT was higher in the NIS than in the IS group (39.1 (95% CI 15.6 to 97.9) vs 19.9 (95% CI 12.8 to 31.1)), but this was not statistically significant (p=0.2). The fold increase in GMT was 20.4 (95% CI 9.2 to 45.5) in the NIS group and 9.3 (95% CI 6.1 to 14) in the IS group (p=0.06). The post-vaccination GMTs were not significantly different between immunomodulator/methotrexate monotherapy, biological monotherapy and combined immunosuppression (18.2, 27.2 and 11.3, respectively), but the fold increase was significantly lower in those on combined immunosuppression compared with those on monotherapy immunosuppression (3.5 (95% CI 1.3 to 9.2) vs 11.5 (95% CI 7.3 to 18.1), p=0.03). In addition, the fold increase in GMT was significantly lower in the combination immunosuppression group than in the biological monotherapy group (3.5 (95% CI 1.3 to 9.2) vs 16.7 (95% CI 8.8 to 31.7), p=0.009, table 4).
As before, a subgroup analysis was performed to analyse the post-vaccination GMTs after excluding the three subjects who had pre-vaccination seroprotection. The post-vaccination GMT was not significantly higher in the NIS group than in the IS group (39.1 (95% CI 15.6 to 97.9) vs 18.1 (95% CI 11.6 to 28.4), p=0.1). The mean fold increase in GMT was 20.4 (95% CI 9.2 to 45.5) in the NIS group compared with 9.6 (6.3 to 14.5) in the IS group (p=0.08). The post-vaccination GMTs for immunomodulator/methotrexate monotherapy, biological monotherapy and combined immunosuppression were 18.2 (95% CI 8.7 to 37.9), 25.8 (95% CI 12.8 to 51.8) and 6.8 (95% CI 2.1 to 21.9), respectively, with fold increases of 7.3 (95% CI 3.8 to 13.9), 17.4 (95% CI 9.1 to 33.5) and 3.4 (95% CI 1.1 to 10.2), respectively. The post-vaccination GMT and fold increase in GMT were significantly lower in the combination group than in the biological monotherapy group (p=0.05 and p=0.01, respectively).
Post-vaccination seroprotection and geometric mean titres by functional T lymphocyte activity
None of the subjects had a low functional T lymphocyte activity but 25 had intermediate levels and 23 had high levels. An intermediate level indicates immunosuppression relative to a high level. Apart from their vaccine response, there were no significant differences between these two groups (table 5). The patients with an intermediate level had a significantly lower post-vaccination seroprotection rate than those with a high level of functional T lymphocyte activity (28% vs 61%, p=0.02). The post-vaccination GMT was lower in those with intermediate level activity, but this was not statistically significant (10 (95% CI 4.2 to 23.8) vs 27.4 (95% CI 10.5 to 71), table 6).
Adverse events to vaccine and effects on IBD activity
The vaccine was well tolerated with 12% (13/105) reporting minor post-vaccination symptoms (table 7). These consisted of local discomfort at the injection site for <24 h or mild upper respiratory tract symptoms such as rhinorrhoea or sore throat. There was no difference in post-vaccination events between the IS and NIS groups. Six per cent (6/105) had influenza-like illness during the 6-month period after the vaccination (three of these had post-vaccination seroprotection).
The median time between vaccination and the follow-up visit at which the HI was measured was 5.3 weeks (range 3–11). This did not differ between the IS and NIS groups (table 7). Twelve subjects (11%) had an increase of >2 points in either the HBI (CD) or SCCAI (UC) in the period of time between the initial and follow-up visits, but only three required an alteration in their IBD medications to treat these additional symptoms. Eleven of the 12 subjects (and all three who required medication adjustment) were in the immunosuppressed group.
In this study of the 2009 H1N1 influenza vaccine in IBD, patients had an overall post-vaccination seroprotection rate of 50%. Immunosuppressed patients with IBD had a lower rate of seroprotection following vaccination against the H1N1 influenza virus than non-immunosuppressed patients. This was true for both definitions of immunosuppression used in this study (defined by immunosuppressant medication or by measurement of functional T lymphocyte activity). Patients on more than one immunosuppressant agent had a lower response than those on single-agent immunosuppression.
Influenza infection is a significant global health problem. As at August 2010, more than 214 countries had reported confirmed cases of H1N1 influenza 2009, including over 18 449 deaths (WHO global alert and response report: http://www.who.int). Published data on the immunogenicity of the 2009 H1N1 vaccine in the general population have shown seroprotection rates in excess of 85%.16 17 Patients with IBD theoretically represent a significant at-risk group due to the combination of innate immune defects and immunosuppressive therapy. There are data to suggest that the severity of influenza infection is greater in immunocompromised individuals.18 This risk—and the knowledge that patients with IBD underutilise available vaccines19—has been recognised, with leading societies and experts in the field advocating routine vaccination for all patients with IBD.8 20 It is not known if these patients respond to influenza vaccines in a similar fashion to their healthy peers. Data from the paediatric literature suggest reduced response rates to certain influenza strains in those taking anti-TNF therapy, but the adult population has not been well studied.12 13 Some studies indicate reduced rates of influenza and pneumococcal vaccine response in immunosuppressed transplant and oncology patients.9 11 There are data showing a reduced response to pneumococcal vaccine in immunosuppressed patients with IBD.21 Recent publications have examined the serological response to the 2009 H1N1 vaccine in immunosuppressed subjects. Reduced rates of seroprotection following H1N1 immunisation have been reported in patients receiving chemotherapy (post-vaccination seroprotection 27–50%).22 Data on the serological response to the H1N1 influenza vaccine in a rheumatological population may be more relevant to IBD.23 A recent study of patients with rheumatoid arthritis (n=41), systemic lupus erythematosus (n=21), psoriatic arthritis (n=17) and ankylosing spondylitis (n=15) showed reduced seroprotection and GMT levels (56%, 67%, 59% and 53% respectively) compared with healthy controls (82%). The authors also identified a reduced rate of seroprotection among 19 patients receiving infliximab (53%) compared with those not taking infliximab (77%).
Our data show a markedly reduced response rate to this particular influenza vaccine in patients with IBD compared with the general population. The fact that this response was also low in non-immunosuppressed subjects was striking. Although we did not have a non-IBD control group in this study, published data on single-dose non-adjuvant 2009 H1N1 vaccines show response rates in excess of 95% (in over 350 subjects in different countries). Within the immunosuppressed group in our cohort, the rates of post-vaccination seroprotection were particularly low for those on two or more immunosuppressants (36%). In addition, the GMT titres were lower in the combination therapy group than in the anti-TNF monotherapy group. We did not identify any other factors associated with diminished seroprotection. Age, gender, type of IBD, disease activity and time from vaccination to assessment of serological response were not associated with lower immunogenicity, although the study was not powered to identify these factors. There was a trend towards lower GMTs in the immunosuppressed patients but this did not reach statistical significance, presumably because of the spread of the data. Although immunosuppressed patients were less likely to develop protective titres, a number of those who did respond to the vaccine were able to produce high titre levels.
Overall, the vaccine was well tolerated with minor adverse reactions consistent with recently published observational data in a European IBD population.24 Just over 10% of the subjects in our study had an increase in their IBD activity scores but, if change in treatment is taken as a marker of clinically relevant disease activity, this was only 3%. More patients in the immunosuppressed group had increased disease activity in the post-vaccination period but this would be expected as, by definition, this group had more severe disease.
The functional T lymphocyte activity assay we used provided an alternative indicator of immunosuppression in the subgroup examined. None of the subjects had a low level, but there has been little study of this assay in the IBD population and the levels and cut-offs proposed for a transplant population may not be applicable. Of those with an intermediate level of T lymphocyte ATP production, 28% had seroprotection post-vaccination compared with 61% of patients with a high level. There was no difference in the use of immunosuppressant medications between these two groups (68% (17/25) vs 61% (14/23), respectively). This assay may potentially identify impaired lymphocyte function associated with the functional level of immunosuppression regardless of medications, but this requires further study.
The major limitation of this study is that it was not possible to use clinical occurrence of H1N1 influenza infection as an endpoint. Six subjects reported a mild influenza-like illness during the 6 months post-vaccination. This low incidence of mild illness meant that the serological response was our only measure of vaccine effectiveness. At the time of study design we (and the WHO) predicted a high attack rate which did not materialise. There have been studies in immunocompromised subjects showing low rates of influenza infection post-vaccination despite inadequate serological titres, suggesting that HI titres may not be reliable markers of seroprotection in these patients.25 26 However, the results are still important as they examine a number of measures of vaccine response (HI and GMT) and show significant variations from published data in the general population. The study did not recruit the 150 subjects calculated for 80% power, but the unexpectedly low vaccine response rate coupled with a very low pre-vaccination seroprotection rate meant that statistically significant differences still emerged. A larger cohort may have facilitated more detailed subgroup analysis. Finally, we did not use a non-IBD control group and instead compared our overall response rates to previously published data on the 2009 H1N1 vaccine in the general population. While this is not ideal, the short time frame of a global pandemic meant that a number of high-quality randomised controlled trials were performed at multiple sites around the world and there is no reason to suspect that the response rates in our local population would be any different from these subjects. We only included studies that used a similar non-adjuvant vaccine, and draw particular attention to the study of more than 200 subjects aged 18–64 years that identified a 98% serological response to the identical vaccine to the one we used.17
To our knowledge, this is the first study to examine the immunogenicity of influenza vaccination in the adults with IBD. We have identified a low serological response to the vaccine among patients with IBD, particularly those on immunosuppressant medications. It is not known if a second dose of the vaccine could improve the seroprotection rate in these patients and this is worthy of future study. It may be that patients with IBD should be routinely checked for evidence of post-vaccination seroprotection, but the standards by which we currently measure seroprotection are derived from studies of healthy individuals and need to be defined for immunosuppressed patients. Functional measurement of T lymphocyte activity may have a role in identifying those at increased risk, but this also requires further study. Large prospective studies using clinical occurrence of influenza infection as an endpoint are desirable, but are difficult to design given the proven efficacy of influenza vaccination in the general population. Patients with IBD should be informed that vaccination may not provide as high a degree of protection as it does in healthy individuals, particularly in those receiving immunosuppressant medications.
The authors thank Michael D Decker, Vice President, Scientific and Medical Affairs, and Chief Medical Officer, Sanofi Pasteur, USA who arranged part-funding of the cost of the haemagglutinin inhibition assays; Cylex Inc who provided the Immuknow kits to the Molecular Diagnostics Center, Tufts Medical Center, Boston; Dr Monica McNeal, Associate Director, Laboratory for Specialized Clinical Studies at Cincinnati Children's Hospital Medical Center who coordinated and performed the haemagglutinin inhibition assays; Dr Massimo Mangiola, HLA Technical Supervisor and Director-in-training at Molecular Diagnostics Center, Tufts Medical Center who coordinated and performed the Immunkow assay; and Elisabeth Cole and Courtney LaPierre at MGH Crohn's and Colitis Center for their work in coordinating the delivery of samples for measurement of haemagglutinin inhibition.
Funding Sanofi Pasteur, USA, part-funded the cost of the hemagglutinin inhibition assay.
Competing interests Sanofi Pasteur, USA provided an educational grant to the Crohn's and Colitis Center, Massachusetts General Hospital, Boston to part-fund the cost of the haemagglutinin inhibition assays that were performed at the Laboratory for Specialized Clinical Studies at Cincinnati Children's Hospital Medical Center. BES received consulting fees from Centocor and Abbott Immunology.
Ethics approval This study was approved by Partners Healthcare (Massachusetts General Hospital) Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.