Background Despite the WHO recommendation that the 2010–2011 trivalent seasonal flu vaccine must contain A/California/7/2009/H1N1-like virus there is no consistent data regarding its immunogenicity and safety in a large autoimmune rheumatic disease (ARD) population.
Methods 1668 ARD patients (systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), ankylosing spondylitis (AS), systemic sclerosis, psoriatic arthritis (PsA), Behçet's disease (BD), mixed connective tissue disease, primary antiphospholipid syndrome (PAPS), dermatomyositis (DM), primary Sjögren's syndrome, Takayasu's arteritis, polymyositis and Granulomatosis with polyangiitis (Wegener's) (GPA)) and 234 healthy controls were vaccinated with a non-adjuvanted influenza A/California/7/2009(H1N1) virus-like strain flu. Subjects were evaluated before vaccination and 21 days post-vaccination. The percentage of seroprotection, seroconversion and the factor increase in geometric mean titre (GMT) were calculated.
Results After immunisation, seroprotection rates (68.5% vs 82.9% p<0.0001), seroconversion rates (63.4% vs 76.9%, p<0.001) and the factor increase in GMT (8.9 vs 13.2 p<0.0001) were significantly lower in ARD than controls. Analysis of specific diseases revealed that seroprotection significantly reduced in SLE (p<0.0001), RA (p<0.0001), PsA (p=0.0006), AS (p=0.04), BD (p=0.04) and DM (p=0.04) patients than controls. The seroconversion rates in SLE (p<0.0001), RA (p<0.0001) and PsA (p=0.0006) patients and the increase in GMTs in SLE (p<0.0001), RA (p<0.0001) and PsA (p<0.0001) patients were also reduced compared with controls. Moderate and severe side effects were not reported.
Statistics from Altmetric.com
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.
Since the identification of a novel influenza A H1N1 virus in Mexico and the USA in April 2009, the virus has spread throughout the world, prompting the WHO to raise the pandemic alert level to phase VI on 11 June 2009.1 This flu pandemic caused by the new A/H1N1 virus led to a higher incidence of hospitalisation and death than the annual rates associated with seasonal flu viruses.1 Recently, the WHO recommended that the Northern Hemisphere's 2010–2011 and Southern Hemisphere's 2011 trivalent seasonal flu vaccine must contain A/California/7/2009(H1N1)-like virus.2
According to the 2010 Recommendations of the Advisory Committee on Immunization Practices and the recent European League Against Rheumatism recommendation,3 immunocompromised patients are recommended to receive the flu vaccine (A/California/7/2009), as these patients are likely to have a higher risk and a more severe course of H1N1 infection.4 Despite this recommendation, the immunogenicity of this vaccine in this patient population remains unclear.
Previous studies that evaluated non-rheumatic immunosuppressed patients lacked a normal control population for comparison. Nevertheless, the immune response to the pandemic H1N1 flu vaccine seems to be low for HIV-infected patients5 6 and recipients of stem cell transplants.7 However, pre-vaccination data were not available for the latter group, and safety assessments were not uniformly reported for either of these populations, thus preventing a definitive conclusion to be drawn from these findings.6 7
With regard to autoimmune rheumatic diseases (ARDs), there is only one published study that evaluated the efficacy and safety of the vaccine against pandemic influenza A H1N1/2009 vaccine in a very limited number of systemic lupus patients.8 The dearth of data regarding other rheumatic diseases and, more importantly, the need for an appropriate patient sample size led to the development of this research.
The objectives of this study are to evaluate the humoral response and safety of the pandemic influenza A H1N1/2009 vaccine in immunosuppressed patients with ARDs compared with healthy controls.
This is a prospective study conducted at a single site in São Paulo, Brazil (Outpatient Clinics of Rheumatology Division, Hospital das Clínicas da Universidade de São Paulo) between March 2010 and April 2010. The protocol was approved by the Local Institutional Review Board and registered with Clinicaltrials.gov under #NCT01151644.
All ARD patients who were regularly followed at the rheumatology outpatient clinics were invited by letter to participate in the Public Health influenza A H1N1/2009 vaccine campaign at the Immunization Center of the Hospital das Clínicas da Universidade de São Paulo. Healthy individuals who came to this centre seeking vaccination in response to the Public Health National Campaign were invited to participate in the control group.
The study included two phases: entry (vaccination from 22 March to 2 April) and a follow-up period of 21 days with a personal diary card of side effects. Blood samples were obtained from each participant immediately before and 21 days after vaccination.
A total of 1668 patients with ARD were included in this study. All patients fulfilled the international classification criteria: systemic lupus erythematosus (SLE),9 rheumatoid arthritis (RA),10 ankylosing spondylitis (AS),11 systemic sclerosis (SSc),12 psoriatic arthritis (PsA),13 Behçet's disease (BD),14 mixed connective tissue disease (MCTD),15 primary antiphospholipid syndrome (PAPS),16 dermatomyositis (DM),17 primary Sjögren's syndrome (pSS),18 Takayasu's arteritis (TA),19 polymyositis (PM)17 and Granulomatosis with polyangiitis (Wegener's) (GPA).20 A total of 234 healthy subjects were concomitantly included in the control group.
Inclusion and exclusion criteria
All participants were ≥18 years old and provided written informed consent. Exclusion criteria were as follows: previous known infection with influenza A (H1N1) 2009; anaphylactic response to vaccine components or to egg; acute infection resulting in fever over 38°C at the time of vaccination; history of Guillain–Barré syndrome or demyelinating syndromes; previous vaccination with any live vaccine 4 weeks before or any inactivated vaccine 2 weeks before the study; 2010 seasonal flu vaccination; or blood transfusion within 6 months and hospitalised patients.
The H1N1 vaccine, a novel monovalent, unadjuvanted, inactivated, split-virus vaccine, was produced by Butantan Institute/Sanofi Pasteur (São Paulo, Brazil). The active substance is an inactivated split flu virus containing antigen equivalent to the A/California/7/2009(H1N1) virus-like strain (NYMCx-179A), one of the candidate reassortant vaccine viruses recommended by the WHO. The vaccine was prepared in embryonated chicken eggs using the same standard techniques for the production of seasonal, trivalent, inactivated vaccine and was presented in 5 ml multi-dose vials with thimerosal added as a preservative (45 μg per 0.5 ml dose).
All participants received a single intramuscular dose (0.5 ml) of 15 μg of haemagglutinin antigen specific for pandemic H1N1 A/California/7/2009-like virus (A/California/7/2009/Butantan Institute/Sanofi Pasteur).
A 21-day diary card was given to each participant upon entry into the study which included 13 established side effects requiring yes/no responses. This written card included the following: local reactions (pain, redness, swelling and itching) and systemic adverse events, such as arthralgia, fever, headache, myalgia, sore throat, cough, diarrhoea, rhinorrhoea and nasal congestion. Participants were asked to return their diary cards at the end of the follow-up period (21 days after vaccination). All local reactions were considered to be related to the H1N1 vaccine. Recorded systemic symptoms were checked by the investigators to determine the causality of solicited systemic adverse events. Unsolicited adverse events were also assessed. Severe side effects were defined as those requiring hospitalisation or death.
Pre-vaccination disease activity was assessed in SLE patients by the SLE Disease Activity Index (SLEDAI)21 and laboratory assessment of RA activity was done using erythrocyte sedimentation rate (ESR) and C reactive protein (CRP) levels.
Blood samples were collected at baseline and 3 weeks after vaccination. The immunogenicity of the H1N1 A/California/7/2009-like virus vaccine was evaluated using the haemagglutination inhibition assay (HIA) at the Adolfo Lutz Institute.
Haemagglutination inhibition assay
The flu virus antigen used in this study was the H1N1 A/California/7/2009 supplied by Butantan Institute. Virus concentrations were previously determined by haemagglutinin antigen titration, and the HIA test was performed after removing the naturally occurring non-specific inhibitors from the sera as previously described.22 The immune response to H1N1 vaccination was evaluated by determining the levels of antibodies by HIA. Anti-H1N1 titre was determined by influenza HIA. Percentage of seroprotection (titre ≥1:40), seroconversion (pre-vaccination titre <1:10 and post-vaccination HIA titre ≥1:40 or pre-vaccination titres ≥1:10 and post-vaccination titres ≥4-fold increase), geometric mean titres (GMTs) and factor increase in GMTs were calculated.
The sample size was chosen practically rather than statistically because of the need to obtain robust estimates of vaccine immunogenicity and safety in ARD. The large sample size of ARD population and controls gave the study a power to find differences between frequencies inferior to 1% (with a power of performed test >80%).
A subanalysis of patients and age-matched controls was performed by a random selection of ARD patients using the SPSS (version 15).
The analyses were descriptive, with calculation of two-sided 95% CI, assuming binomial distributions for dichotomous variables and log-normal distribution for haemagglutination inhibition titres. For categorical variables, statistical summaries included the rate of seroconversion and seroprotection; these were compared using Fisher's exact test. Every subgroup of volunteers had the haemagglutination inhibition GMT calculated before and 21 days after vaccination. Compared between each subgroup of patients with ARD and the group of healthy volunteers using a two-sided Student's t test with an α level of 0.05.
Multivariate logistic regression analyses were performed to analyse possible factors that influenced the vaccine performance.
A total of 1862 patients were recruited of whom 1668 (89.6%) completed the study. The different categories of ARD patients included those with SLE (n=572), RA (n=343), AS (n=152), SSc (n=127), PsA (n=101), BD (n=85), MCTD (n=69), PAPS (n=54), DM (n=45), pSS (n=36), TA (n=30), PM (n=28) and WG (n=26) who were compared with 234 healthy controls (table 1).
A significantly higher mean age (47.1±14.1 vs 38.7±12.5 years, p<0.0001) and a predominance of women (80.4% vs 65.8%, p<0.0001) were observed in ARD patients compared with controls.
Approximately half of the patients were under glucocorticoid therapy, receiving a mean daily dose of 5.4±9.3 mg prednisone/day, and less than 10% of these patients were under high doses (≥20 mg/day). More than 50% of all ARD patients were using immunosuppressive drugs (table 1).
The overall vaccine immune response is illustrated in table 2. Prior to immunisation, the seroprotection rate was comparable between ARD patients and controls (p=0.57). After immunisation, the seroprotection (68.5%, 95% CI 66.3% to 70.7% vs 82.9%, 95% CI 78.1% to 87.7%, p<0.0001) and seroconversion rates (63.4%, 95% CI 61.1% to 65.7% vs 76.9%, 95% CI 71.5% to 82.3%, p<0.001) were significantly lower in ARD patients compared with healthy subjects. Subanalysis of 234 randomly selected patients and 234 age-matched controls (mean age of 38.7±12.6 years) revealed that results still hold with comparable seroprotection prior to immunisation (p=0.28) and lower seroprotection after immunisation (p=0.001) as well as lower seroconversion rates (p=0.01) in ARD patients compared with controls.
Analysis of each disease subgroup revealed that pre-vaccination seroprotection was similar in patients and controls (p>0.05). After vaccination, seroprotection rates were significantly reduced in SLE (p<0.0001), RA (p<0.0001), PsA (p=0.0006), AS (p=0.04), BD (p=0.04) and DM (p=0.04) patients. Seroconversion rates were also reduced in SLE (p<0.0001), RA (p<0.0001) and PsA (p=0.0006) patients compared with healthy subjects.
GMTs before immunisation (8.0, 95% CI 7.7 to 8.4 vs 9.3, 95% CI 8.2 to 10.5, p=0.016), GMTs after immunisation (71.5, 95% CI 66.2 to 77.3 vs 122.9, 95% CI 103.4 to 146.1, p<0.0001) and the factor increase in GMTs (8.9, 95% CI 8.3 to 9.6 vs 13.2, 95% CI 11.1 to 15.8, p<0.0001) were significantly lower in the ARD group compared with the control group. Disease evaluations for specific patient subgroups revealed lower GMTs after immunisation for SLE (p<0.0001), RA (p<0.0001), PsA (p<0.0001) and AS (p=0.004) patients compared with healthy subjects. The increase in post-vaccination titres in GMTs was also significantly lower in SLE (p<0.0001), RA (p<0.0001) and PsA (p<0.0001) patients compared with normal controls (table 3).
Multivariate logistic regression was performed to analyse the possible influence of relevant parameters for seroconversion rates (disease (RA, SLE and PsA), seroprotection rate before immunisation, age >60 years, glucocorticoid and immunosuppressive use). Only SLE (p<0.001), RA (p<0.001) and age >60 years (p=0.004) remained significant.
Regarding disease activity, the mean SLEDAI pre-vaccination scores were similar in seroconverter and non-seroconverter SLE patients (p=0.08). Mean levels of pre-vaccination inflammatory markers were also comparable in seroconverter and non-seroconverter RA patients (ESR, p=0.11 and CRP, p=0.82).
All participants who completed the study have returned their adverse events questionnaire. No severe side effects were reported in patients and controls up to 3 weeks of follow-up. Minor local reactions were more frequently observed in controls than in patients (8.3% vs 14.1%, p=0.007), whereas the frequency of mild systemic reactions were comparable between both groups (24.6% vs 25.6%, p=0.75), with a significant higher frequency of arthralgia and fever and a lower frequency of sore throat in ARD patients compared with control group (table 4).
To our knowledge, this is the largest evaluation of the pandemic unadjuvanted influenza A (H1N1) 2009 virus immunisation safety in ARDs reported to date. The vaccine was generally well tolerated and had a favourable safety profile. A diverse pattern of immune responses was observed among patients with different diseases, with substandard antibody production in SLE, RA, PsA, AS, BD and DM patients.
Pre-existing immunity to the pandemic virus evaluated herein was comparable in patients and controls with an overall seroprotective rate higher than that reported in 12 691 people recruited in China.23 The recommendation for voluntary annual seasonal flu vaccination in ARD patients with pulmonary involvement and those receiving immunosuppressive therapy may not account for this difference, as contemporary seasonal flu vaccines offer little or no advantage regarding antibody responses to the pandemic influenza A H1N1/2009 vaccine.24
In the present study, the non-adjuvant preparation was chosen to prevent triggering an ‘adjuvant disease’ in genetically susceptible individuals.25 26 Because adjuvants may act as ligands for Toll-like receptors or stimulate innate immune responses, molecular mimicry is a potential risk in autoimmune-prone individuals. In this regard, aluminium adjuvants commonly used in human vaccines were found to be associated with macrophagic myofasciitis, an autoimmune-related disease.25 26 In addition, for the pandemic influenza A H1N1 vaccine tested in a large population in China, the non-adjuvanted formulation was more effective compared with adjuvant formulations with equivalent antigen content.23
Immunogenicity against the 2009 pandemic H1N1 influenza vaccination was significantly lower in the present group of patients with ARDs. This reduction in antibody production to viral proteins is likely associated with the impaired immune state of patients with these illnesses, as a previous study reported that T lymphocyte-dependent responses to protein antigens are compromised in these patients.27 This feature may be an inherent characteristic of rheumatic diseases, whereas in other immunosuppressive conditions such as long-term haemodialysis28 and renal transplant,29 a high rate of protective immunity was observed for the seasonal flu vaccine. Alternatively, this discrepancy in renal patients may reflect frequent seasonal immunisations in previous years with vaccines of similar composition, which is not the case for the pandemic H1N1 flu vaccine.29 Reinforcing this latter possibility, recent studies on stem cell transplantation and HIV patients reported a reduced response in these patients to the pandemic vaccine.5,–,7
Of note, the higher mean age of our patient group does not seem to explain the reduced immunogenicity observed since the subgroup analysis performed herein including patients and age-matched controls confirmed the reduced response in ARD patients. Likewise, the female predominance of our rheumatic disease patients cannot contribute to the observed decrease in humoral responses in light of previous findings that women have higher antibody titres to a large number of viral and bacterial vaccines.30
Importantly, attenuated protective immunogenicity against the pandemic influenza A H1N1/2009 vaccine was not a universal finding in rheumatic diseases and most likely reflects the specific immune alterations associated with therapy of each illness that ultimately affect antibody production. In this regard, this is the first and largest prospective study performed demonstrating a high 2009 pandemic H1N1 vaccine seroprotection rate, seroconversion rate and a significant increase in GMT for adults with MCTD, SSc, PAPS, pSS, PM, WG and TA. Reinforcing these findings, two recent studies with a small number of WG patients reported an adequate humoral and cellular response to the seasonal flu non-pandemic virus.31 An elegant study on SSc patients reported effective humoral and cellular responses to an adjuvanted virosomal non-pandemic flu vaccine.32 As for other connective tissue diseases such as MCTD, PAPS, pSS, PM and TA, no data are available regarding immune responses to flu vaccination, but the adequate seroprotection and seroconversion rates observed herein for the 2009 H1N1 flu virus may support extending the recommendation for seasonal flu vaccination in these patients.
In the present study, vaccine immune responses were significantly decreased in SLE, RA and PsA patients and reflected the low seroprotection and seroconversion rates as well as the inadequate rise in factor increase in GMT. A reduction in seroprotection with adequate seroconversion was also observed in DM, AS and BD patients. Our finding settles the controversy raised by previous reports on SLE patients, one of which reported normal immune responses in a small number of patients immunised with the seasonal flu vaccine33,–,37 and another that reported reduced immune responses in patients immunised with the 2009 H1N1 pandemic vaccine.8 This reduced immunogenicity is likely due to impairments in humoral and cellular immunity in SLE patients that may ultimately affect the response to antigenic challenge.37 The unexpectedly low immune responses in RA patients contrast with previous studies, including two with randomised designs reporting normal immunogenicity to the seasonal vaccine in RA patients.38 39 The use of non-adjuvanted vaccine in the present study may partly explain this discrepancy, as the use of adjuvants may be required in RA to achieve maximum immunogenicity.39
Alternatively, glucocorticoid and immunosuppressive drugs may account for the deficient vaccine responses in rheumatic disease patients. However, several reports have demonstrated that, with the exception of rituximab,40 41 glucocorticoids, methotrexate and tumour necrosis factor blockers do not have a deleterious effect on the immunogenicity of the seasonal flu vaccine in RA37 38 42 43 and AS.43 The small number of patients using rituximab in our group of patients (0.8%) cannot explain the deficient immune responses observed in some rheumatologic disease patients.
It is important to emphasise that immunisation had an excellent safety profile and that only mild reactions were observed. The possibility of autoantibodies produced post-vaccination, the exacerbation of established rheumatic diseases and the influence of therapy38 44 are currently being evaluated in different subgroups and will be the subject of another publication. Importantly, neurological autoimmune diseases following flu immunisation, such as Guillain–Barré syndrome, acute encephalomyelitis or transverse myelitis,44 were not observed in this population, which is theoretically more prone to develop autoimmunity. We cannot exclude, however, the possibility that a severe side effect might have been missed since not all participants returned to the second phase. This possibility is unlikely since all patients are still followed in our Outpatient Rheumatology Clinic.
In summary, this is the largest prospective study of pandemic unadjuvanted influenza A (H1N1) 2009 virus immunisation in ARDs patients that provides clear evidence of its safety. A distinct disease profile of immunogenicity was identified, and further studies are necessary to determine if a booster dose will be effective for those with suboptimal immune responses.
This study was supported by grants from Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP #2010/10749-0 to EB), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPQ #303165/2008-1 to EFB, #300248/2008-3 to CAS, #300559/2009-7 to RMRP, #300665/2009-1 to JFC and #301411/2009-3 to EB), Federico Foundation (to EFB, CAS, RMRP, JFC, PDS-B and EB) and Butantan Foundation. The authors thank the subjects for their critical roles in this study, the staff of Hospital das Clinicas FMUSP, Adolfo Lutz Institute, Faculdade de Medicina da USP and other participants including the following: Dr Ricardo Pietrobon (Duke University, Durham, North Carolina, USA) for revising the manuscript and Dr João Miraglia (Butantan Institute) for statistical analysis.
Ethics approval This study was conducted with the approval of the Cappesq.
Provenance and peer review Not commissioned; externally peer reviewed.