We have demonstrated that RA patients treated with anti-TNF have impaired B cell and antibody responses to seasonal influenza vaccine. This suboptimal response includes lower serum HAI antibodies, which is consistent with previous studies [4
] and also diminished induction of influenza-specific effector and memory B cell responses, which have not been previously recognized.
The impact of anti-TNF therapy was most evident during the effector phase (day 5 to day 10) than the memory phase (one to six months) of the B cell response. This difference most likely reflects the sensitive resolution of vaccine response provided at these early time-points. Whereas resolution of the influenza-specific B cell and antibody responses at one month or later after vaccination is confounded by variation in amounts of pre-existing memory B cells and antibody present among individuals, potentially underestimating the responsiveness to a single vaccination event. The strong positive correlation observed between the effector B cell response from day five to day seven after vaccination and the memory B cell and antibody response at one month suggests a predictive value for examining the effector B cell response as an early biomarker for the development of memory B cells and antibody. The reduced effector and memory B cell response we observed in anti-TNF-treated patients, may contribute to the increased risk and severity of infection observed in anti-TNF-treated patients.
Treatment with anti-TNF is likely to target multiple aspects of the vaccine response that are contributing to the decreased B cell response observed. Previously, we demonstrated that lymphoid architecture is altered in RA patients treated with etanercept, including the destruction of the FDC network and disruption of germinal centers and corresponding decrease in peripheral blood memory B cells [13
]. This dysregulation may be preventing influenza-specific B cells from participating in optimal germinal center reactions, thereby compromising the magnitude and quality of the effector and memory influenza-specific B cell response. Furthermore, as autocrine TNF enhances B cell proliferation [23
], and TNF is produced in abundance by plasma cells [24
]; TNF blockade may inhibit this process, contributing to the reduced development of influenza-specific memory B cells and plasmablasts that we observed. A possible additional and not mutually exclusive point of impact for TNF blockade is limiting the maintenance of influenza-specific plasma cells, because TNF enhances plasma cell survival in vitro
], and has been suggested to be important for plasma cell survival in vivo
]. It would be of particular interest to determine if TNF blockade is merely blocking the activation and expansion of influenza-specific plasmablasts and memory B cells or promoting the development of a distinct short-lived effector population that does not produce antibody, nor can develop adequately into antibody-producing cells.
Diminished B cell and antibody responses may also result from impaired monocyte, dendritic, and T cell function by TNF-blockade that has been reported, limiting the supportive, yet potentially critical roles these cells play in the induction of a robust B cell response. The known suppressive effect of anti-TNF on memory and effector T cell functions such as IFNγ and IL-2 production [28
], and increased Treg function [29
] may negatively impact the B cell response. A limitation of the current study is the inability to adequately assess the various anti-TNF agents individually for their impact on influenza vaccine response. We expect that slight differences in the B cell response to vaccination may become apparent in a large multi-anti-TNF agent comparison study; however, it is uncertain that it would be feasible to dissect differences that occur as a result of mode of action, from those that could result from the different routes of administration or dosing schedules for the various agents.
Interestingly, we observed subtle indications of enhanced B cell responsiveness to the influenza vaccine in RA patients not treated with MTX or anti-TNF as compared with the healthy control subjects. This outcome may reflect an inflammatory environment that could be present in these patients, resulting from the disease process, contributing in an adjuvant-like manner, thereby enhancing the germinal center reaction, and consequently the plasmablast and memory B cell responses. Additionally, rheumatoid factor-expressing B cells exhibit an enhanced ability to internalize antigen-immunoglobulin complexes and subsequently present the antigen to specific T cells [31
], raising this possibility as a potential mechanism of enhanced vaccine responses in some RA patients. Future studies with greater numbers of untreated RA patients are warranted to adequately substantiate and dissect these questions further. Such studies may elucidate additional mechanistic aspects of the inflammatory processes in RA disease and provide strategies to enhance vaccine responsiveness in the general population.
Although a decreased response to the influenza vaccine was observed in RA patients treated with anti-TNF, it is important to emphasize that vaccination did increase serum antibody and presumed protection from infection. Furthermore, recent European League Against Rheumatism (EULAR) recommendations indicate that influenza vaccination should be strongly considered, and that vaccination can be administered during the use of anti-TNF therapy [32
]. The anti-TNF treated patient population may benefit from the high-dose influenza vaccine recently licensed for individuals aged 65 years and older, and warrants further investigation. Moreover, our results indicate that every effort should be made to provide annual influenza immunization and indicated vaccination updates to RA patients before anti-TNF therapy is initiated. It is also possible that the efficacy of annual influenza vaccination in anti-TNF treated patients could be improved by timing immunization with the decline in biological effect of the corresponding agent. This question could be addressed by studying the relative efficacy of immunization provided just before the next dose of anti-TNF. The clinical value of examining anti-TNF and vaccination timing is substantiated by a recent study by Elkayam at al. which demonstrated that RA patients treated with infliximab three weeks prior to influenza vaccination had a reduced serum antibody response as compared with those patients treated with infliximab on the day of vaccination [33
]. In patients with particularly high risk (i.e. those patients with previous history of severe infection, low baseline HAI titers, and/or other co-morbidities) it may be reasonable to delay the next dose of anti-TNF drug to provide a longer window of developing a protective response.