Introduction
Current therapies for RA are aimed at inhibiting inflammatory cytokines, especially tumor necrosis factor (TNF)-α biotherapies, such as antibodies (infliximab, adalimumab) and soluble receptor (etanercept) specific for TNF. These three commercially available TNF antagonists have been tested in established and in early diseases. They effectively improved disease activity and significantly slowed radiologic deterioration [
1,
2]. However, serious infections are a major concern in patients with rheumatic diseases, and inhibition of TNF-α increases the risk of serious and benign infections [
3]. The role played by TNF-α in the body's defense against bacterial and viral invasion is multiple: recruitment of neutrophils, eosinophils, and macrophages; release of cytokines and local chemokines; attraction and activation of phagocytes; increased T-cell adhesion; enhanced antigen presentation; and recruitment and proliferation of T and B cells [
4]. Moreover, TNF-α is also involved in the formation and sustainment of Mycobacterium granulomatous lesions [
5]. Neutralization of TNF-α for treating rheumatic diseases increases the risk of reactivation and outbreak of tuberculosis and other opportunistic infections [
6,
7]. A decrease of the tuberculosis-specific CD4
pos T-cell response in patients treated with anti-TNF was found [
8]. In addition, anti-TNF treatment induced a reduction in effector memory CD8
pos T cells specific for mycobacteria [
9].
In contrast, less is known about viral infections. Herpesviruses can persist in patients in a latent state and be reactivated under situations of immunosuppression. Although cases of lymphoproliferative disorders have been reported in RA, the role of TNF-α antagonists in Epstein-Barr virus (EBV)-related lymphomas is not clear [
3,
10,
11]. Conversely, inflammation, a hallmark of RA, might be associated with the risk of lymphoma [
12]. Regarding EBV infections, data are rather reassuring. In a recent study, no impairment of the anti-EBV CD8
pos T-cell response was found in patients treated with anti-TNF, and the EBV viral load was not increased [
13]. However, regarding varicella zoster virus, another herpesvirus, a recent publication suggested that anti-TNF-α antibodies could be associated with increased risk of reactivation, responsible for an increased rate of herpes zoster events in patients treated with these biologic agents [
14].
Cytomegalovirus (CMV) is a member of the β-herpesvirus subfamily, which infects 50% to 60% of the European population. Primary infections are mostly unnoticed. However, the virus reactivates from latent infections on immunosuppression, leading to graft rejection and severe pathology, such as pneumonitis in bone marrow transplantation, colitis, and retinitis in AIDS [
15]. CD4
pos and CD8
pos T-cell responses against CMV have been studied by using peptides, recombinant proteins, or lysates of infected cells [
16‐
19]. Frequencies of CMV-specific CD4
pos and CD8
pos T-cells have been shown to be extremely high in immunocompetent persons [
17], and to be maintained throughout life [
20]. Contributions of CD4
pos and CD8
pos T cells have been demonstrated both
in vitro [
21,
22] and
in vivo, [
23,
24].
Although CD4
pos T cells possess their own capacity to inhibit CMV replication [
21,
22,
25], they also contribute to the differentiation and maintenance of CMV-specific CD8
pos T cells [
23]. Moreover, anti-CMV specific effectors are increased in CD28
neg CD4
pos T cells [
17,
26], a population that is expanded in RA, because of TNF-α [
27,
28]. TNF-α has been demonstrated not only to play a prominent role in RA but also to diminish the intensity of the T-cell response [
29]. Moreover, anergy of T cells was observed in RA patients [
30]. Therefore, the outcome of the anti-CMV CD4
pos T-cell response in RA patients treated with anti-TNF-α is of interest.
Case reports have mentioned the reactivation of CMV in anti-TNF-treated patients [
3]. It is thus important to know more precisely the persistence of anti-CMV memory CD4
pos T cells in RA. The high proportion of CMV-seropositive individuals and the high frequencies of CMV-specific T cells allow the follow-up of the Ag response
ex vivo [
17,
18]. We thus have chosen to test the anti-CMV CD4
pos T-cell response as a model for the study of the antiviral response in RA patients whose TNF-α is neutralized with anti-TNF. We previously showed that TNF-α participates in the control of infection [
25]. Because neutralization of TNF-α may alter the control of CMV, we thus measured the CD4
pos T-cell response in RA patients treated with anti-TNF.
Because IFN-γ produced by CD4
pos T cells is important in the control of CMV
in vitro [
25] and
in vivo [
21], we tested the intracellular production of IFN-γ in CD4
pos T cells in response to total CMV Ags in patients with RA, before and after initiation of treatment with an anti-TNF-α agent. Specific proliferation in response to CMV Ags also was investigated. Our data show that anti-TNF treatments do not impair the CD4
pos anti-CMV response and suggest that CMV infection remains controlled in treated RA patients latently infected with CMV.
Discussion
In this work, we analyzed the CD4pos T-cell response to CMV total Ags to evaluate the consequences of treatments with anti-TNF agents on the viral immune memory response. We observed that the CD4pos T-cell response toward CMV Ags was not altered by anti-TNF antagonists, whether soluble receptor or antibodies.
Production of IFN-γ is a marker and potent effecter of the antiviral response, especially against CMV [
21,
25]. In this work, total CMV Ags were used to monitor the global CD4
pos immune response through the IFN-γ production. It appeared that the immunity to viral CMV Ags was conserved during the course of the treatment. Primary infections by CMV are accompanied by the appearance of a high response and frequency of specific CD4
pos T cells, which are maintained durably [
18,
21,
36]. The absence of modification of the anti-CMV CD4
pos T-cell response in most patients and slight variations in some others observed in our present study are similar to those in previous studies in normal blood donors regarding proliferation [
18] and cytokine production by flow cytometry in transplant recipients [
37] and HIV patients [
36]. Contrary to the anti-mycobacteria CD8
pos T-cell response [
9], the frequency of anti-CMV CD4pos T cells was not impaired by anti-TNF treatments. Our data suggest that the anti-CMV response is sufficient to control the latent CMV infection during the course of the anti-TNF treatments.
As observed in previous reports [
17,
26], the percentages of CMV-specific CD4
pos T cells in the CD28
neg population were high, as compared with those observed in the general CD4
pos T cell population. We observed that those percentages, obtained by using infected cell lysates, were lower than those observed when using synthetic peptides [
17]. However, they were within the range of and in accordance with those reported with whole-cell lysate [
17]. Thus, all peptides may not be available as epitopes through processing of total CMV Ags by APC.
The high proportion of CMV-specific CD4
pos T cells in the CD28
neg population was reported earlier [
17,
26], but had not been studied in RA. The percentage of CD28
neg CD4
pos T cells observed in the present study was similar to that observed by Schmidt
et al. [
27] in RA patients.
However, the scope of this present study was not to compare the percentage of CD28
neg CD4
pos T cells in the normal population and RA patients. We followed the proportion of CD28
neg CD4
pos T cells during the course of anti-TNF treatment. Despite the neutralization of TNF-α, the percentage of CD28
neg CD4
pos cells did not vary during the course of anti-TNF treatment in our study. The role of CD28
neg CD4
pos cells in RA is not elucidated, but it has been suggested that they do not play an aggressive role in autoimmunity and may not play a specific role in RA [
34].
Hyporesponsiveness was reported in RA patients in T cells from synovial fluid [
30]. The significant response of CD4
pos T cells from peripheral blood at day 0 of treatment and the relatively stable anti-CMV response over the course of treatment in our current study suggest (a) that anergy to anti-CMV Ags, if any, was not a prominent feature of RA patients; and (b) that anti-TNF treatments did not restore function from CD4
pos T cells putatively engaged in anergy or in TNF-α-induced hyporesponsiveness [
29].
Several studies have analyzed the
in vitro responses to pathogens that are at risk in RA patients treated with anti-TNF. The
ex vivo anti-mycobacteria IFN-γ response was found to be impaired by infliximab and adalimumab [
8] and was in accordance with the risk of reactivation of tuberculosis, especially with antibodies [
38]. The situation regarding the antiviral immunity may be more complex and has been less explored. Although HBV has been described to reactivate on anti-TNF treatment [
3], long-term safety of TNF blockers requires longer follow-up regarding HCV [
3,
39].
Regarding EBV, a member of the herpes family, the risk of lymphoma has been debated [
3]. The anti-EBV response was found to be maintained [
13], suggesting that no short-term (3 month) defect in EBV-immune surveillance occurs in patients receiving MTX or anti-TNF drugs.
A risk of varicella-zoster virus infection, another member of the herpes virus family, may be present, as reported by Strangfeld
et al. [
14]. However,
in vitro studies of the CD4
pos T-cell response against varicella-zoster Ags have not been performed. Several cases of CMV infection have been reported during the course of anti-TNF treatments [
3,
40,
41]. However, because patients received concomitant immunosuppressive treatments, it is difficult to establish a link between anti-TNF treatment and reactivation of CMV.
Our present data argue in favor of the maintenance of anti-CMV immunity during anti-TNF treatments. This is of importance in light of previous observations that TNF-α is an important component of the anti-CMV control
in vitro [
25,
42]. Our current study and former published reports [
8,
13] suggest that
in vitro responses to viral proteins or peptides are of help to identify risks of viral infection in patients treated with anti-TNF. In addition, the conservation of anti-CMV CD4
pos T cell immunity during anti-TNF treatment suggests that vaccinations can be envisaged during treatment by anti-TNF. However, until safety data are available, live attenuated virus vaccines should be contraindicated in RA patients.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
J-LD, ACo, and ACa designed the study. DN participated in the design of the study and performed the statistical analysis. J-LD and J-FB performed the experiments. J-FB, BJ, ACo, and ACa enrolled patients. J-LD, ACo, and ACa wrote the manuscript. All authors helped to draft the manuscript. All authors read and approved the final manuscript.