The effect of immunomodulators on the immunogenicity of TNF-blocking therapeutic monoclonal antibodies: a review

Therapeutic monoclonal antibodies have revolutionized the treatment of various inflammatory diseases. Immunogenicity against these antibodies has been shown to be clinically important: it is associated with shorter response duration because of diminishing concentrations in the blood and with infusion reactions. Concomitant immunomodulators in the form of methotrexate or azathioprine reduced the immunogenicity of therapeutic antibodies in rheumatoid arthritis, Crohn disease, and juvenile idiopathic arthritis. The occurrence of adverse events does not increase when immunomodulators are added to therapeutic antibodies. The mechanism whereby methotrexate and azathioprine influence immunogenicity remains unclear. Evidence-based consensus on prescribing concomitant immunomodulators is needed.


Immunogenicity of biologicals
Th erapeutic monoclonal antibodies (TmAbs) that block tumor necrosis factor are powerful modalities in the treatment of various infl ammatory diseases, but both chimeric and human TmAbs can induce anti-TmAb antibodies. Immunogenicity can change the pharmaco kinetics of biological therapeutics, resulting in suboptimal therapeutic levels of the drug in patient serum. Th e problem of immunogenicity against therapeutic antibodies has been described since TmAbs have been on the market for the treatment of various infl ammatory diseases, and knowledge regarding anti-TmAb antibodies is increasing. Nevertheless, technical factors, standard ization of the assays used to measure anti-TmAb anti bodies, and the timing of the measurements make immunogenicity a complex subject to investigate. Several studies in various infl ammatory diseases demonstrate the presence of anti-TmAb antibodies [1]. Table 1 gives an overview of the reported frequency of anti-TmAb antibodies in infl ixi mab (antibodies to infl iximab, or ATIs) and in adali mumab (anti-adalimumab antibodies, or AAAs) . Th e large variation in the percentages of anti-TmAb antibodies measured could be related to the diff erences in assays, duration of treatment, and the use of conco mitant immunosuppressive treatment.

Relevance of anti-TmAb antibodies
In studies in which trough serum adalimumab or infl iximab concentrations were measured, the presence of anti-TmAb antibodies was associated with decreased serum drug levels and a diminished response [2,[5][6][7]10,11,13,14]. Furthermore, anti-TmAb antibodies in the presence of TmAb concentrations in patients serum lead to the formation of immune complexes [23]. Th e continuous presence of immune complexes in the serum could lead to adverse events. Little is known about the safety of TmAb and anti-TmAb antibody immune complexes. Th e presence of ATIs and of immune complexes of various sizes might be associated with infusion-related hypersensitivity reactions [2,6,10,23,24]. In one study, higher concentrations of ATIs predicted a higher risk of infusion reactions [10]. Concomitant immunosuppressive therapy, in the form of methotrexate or azathioprine, was shown to be associated with a lower frequency of anti-TmAb antibodies compared with TmAb mono therapy in multiple studies [4,7,[10][11][12][13]15,16,18,25]. Th e administration of concomitant immunosuppressive therapy could be an opportunity to bypass the detri mental eff ect of immunogenicity on the effi cacy of biological therapeutics and possible immune complex-related adverse events. In rheumatoid arthritis (RA), biological therapeutics are preferably prescribed with concomitant disease-modifying antirheumatic drugs (DMARDs) since eff ectiveness is increased compared with monotherapy [26]. It is unclear whether this eff ect is related to a synergistic or an anti-immunogenic eff ect. However, in clinical practice, the decision to prescribe concomitant immunosuppressive treatment is deter mined by many factors: adverse events or intolerance, patient's preference, rheumatologist's preference, eff ect ive ness of immunouppressant monotherapy, and co morbidity. Also, daily practice diff ers among infl am matory diseases; for example, in RA, it is common to prescribe methotrexate together with biological treat ment, but in Crohn disease, the number of patients receiving concomitant immunomodulators is lower [13]. In psoriasis, methotrexate treatment is often discon tinued before the start with biological treatment, and in ankylosing spondylitis, eff ective therapeutic options (DMARDs) are lacking [22,27]. Furthermore, there are no clear guidelines on prescribing concomitant immuno suppressants.

Current knowledge
We performed a systematic PubMed search of articles on the subject of concomitant immunosuppressive therapy with TmAb treatment. Search terms were infl iximab, adalimumab, rheumatoid arthritis, ankylosing spondylitis, psoriatic arthritis, psoriasis, Crohn disease, juvenile idiopathic arthritis, juvenile rheumatoid arthritis, immunogenicity, antibodies, anti-adalimumab antibodies, antiinfl iximab antibodies, methotrexate, MTX, and immunomodulators. Articles were selected if a full text was available and if the formation of antibodies against adalimumab/infl iximab and the possible eff ect of immunomodulators on immunogenicity were described. CLMK and GMB performed the PubMed search and evaluated all of the articles.

Prospective studies
Almost 15 years ago, Maini and colleagues [4] investigated whether methotrexate could reduce the immunogenicity of infl iximab. Th e authors postulated that, if added to infl iximab in a dosage of 7.5 mg weekly, methotrexate itself would not be eff ective and toxicity would be minimized, but it would have an additive benefi t on decreasing immunogenicity, and toxicity would be minimized. Th ey performed a 26-week, double-blind, placebo-controlled, multicenter trial in which 101 patients with RA were randomly assigned to one of seven groups, receiving infl iximab at 1, 3, or 10 mg/kg or placebo with or without methotrexate 7.5 mg per week for 14 weeks. Th e overall incidence of ATIs after 26 weeks was 17.4%. Th e development of antibodies was inversely associated with the infl iximab dose: 53%, 21%, and 7% in patients receiving 1, 3, and 10 mg/kg monotherapy, respectively. Th e use of conco mitant methotrexate greatly diminished the appearance of ATIs, with incidence rates of 15%, 7%, and 0% at the three dose levels. Th e enzyme immunoassay used to measure the presence of ATIs was fully described in the article. Th e authors suggest that immunologic tolerance to infl iximab was induced by higher dosages of infl iximab, probably because of the maintenance of circulating levels of infl iximab, and that this tolerance was potentiated by the simultaneous administration of methotrexate at a dose of 7.5 mg per week. In a prospective proof-of-concept study in patients with Crohn disease, the concomitant use of immunosuppressive therapy was compared with infl iximab monotherapy in 174 patients [13]. In one study arm, 65 patients used concomitant azathioprine 2 to 2.5 mg/kg; in a second arm, 50 patients used concomitant intramuscular or subcutaneous methotrexate 15 mg weekly; and in a third arm, 59 patients received infl iximab monotherapy. Measurements of ATIs were performed by Prometheus Laboratories (San Diego, CA, USA) before and 4 weeks after each infusion. Again, the concomitant use of immunosuppressive therapy was associated with a lower incidence of ATIs compared with patients with infl iximab monotherapy (46% versus 73%, P <0.001). Th is diff erence was observed in both the methotrexate arm (44% ATIs, P = 0.002) and the azathioprine arm (48% ATIs, P = 0.004). When trough infl iximab levels were stratifi ed according to the presence or absence of ATIs, patients with ATIs had lower infl iximab levels than patients without ATIs and these levels were even lower when patients were not taking concomitant immunosuppressive treatment. Th ere was a trend toward Crohn disease 14-75 2.6-17 [10][11][12][13][14][15][16][17] Juvenile idiopathic arthritis NA 17 [18] Ankylosing spondylitis 29 31 [19,20] Psoriatic arthritis NA 18 [21] Psoriasis NA 45 [22] AAA, anti-adalimumab antibody; ATI, antibody to infl iximab; NA, not applicable.
signifi cance for the presence of ATIs being associated with a shorter duration of response in patients not taking concomitant immunosuppressive treatment compared with patients taking azathioprine or methotrexate (P = 0.06). Strikingly, when no ATIs were present, the duration of response was not infl uenced by immunosuppressive co-treatment. Th is suggests that the anti-immunogenic eff ect was more important than a possible synergistic eff ect.

Descriptive studies
Besides these prospective studies, a number of obser vational cohort studies on the subject of immunogenicity and studies with immunogenicity as secondary objective describe the postulated eff ect of immunosuppressive agents on immunogenicity of TmAbs [4,7,[10][11][12][13]15,16,18,25]. Th ese studies and the two studies described above are summarized in Table 2. During a 28-week cohort study, AAAs were detected in 21/121 of adalimumab-treated RA patients (17%) [7]. A radioimmunoassay designed by Sanquin (Amsterdam, Th e Netherlands) was used to measure the AAAs. Patients receiving concomitant methotrexate (mean dosage of 19.4 mg per week) had a lower rate of antibody development than patients receiving adalimumab monotherapy (12% versus 38%). EULAR (European League Against Rheumatism) non-responders had AAAs significantly more often than good responders did (34% versus 5%, P = 0.032). AAA formation corresponded with lower serum adalimumab concentrations at 28-week follow-up.
One hundred thirty-three patients with juvenile idiopathic arthritis (JIA) were randomly assigned to receive adalimumab or placebo [18]. In total, 16% of the patients had at least one positive test for AAAs during the study. Five of 85 patients (6%) receiving methotrexate and 22 of 86 patients (26%) not receiving methotrexate developed AAAs. Th e presence of AAAs did not lead to a greater rate of discontinuation of adalimumab and did not increase the incidence of serious adverse events. Th e assay used to measure AAAs was not described.
In Crohn disease, an anti-immunogenic eff ect of concomitant immunosuppressive therapy was shown in a cohort study of 125 patients [10]. Patients who were taking immunosuppressive agents had a lower incidence of ATIs (43%) and higher infl iximab concentrations than patients who were not taking immunosuppressive agents (75%, P <0.01). Tests were performed by Prometheus Laboratories. Th e incidence of infusion reactions was reduced and the duration of response increased in patients taking immunosuppressive agents. Th ere was a negative relation between the ATI concentration and the duration of response to infl iximab (P <0.001).
In the ACCENT I (Crohn disease without fi stulas) and II (fi stulizing Crohn disease) trials, patients received infl iximab induction therapy followed by placebo or maintenance therapy for up to 54 weeks [11,12]. In the ACCENT I trial, 442 patients were assessed for the presence of ATIs. Fourteen percent of patients developed ATIs, and 46% had an inconclusive result. Six percent of the patients receiving concomitant steroids and immunomodulators, 17% of the patients receiving concomitant steroids alone, 10% of the patients receiving concomitant immunomodulators alone, and 18% using infl iximab monotherapy developed ATIs. Median infl iximab concen tration in patients positive for antibodies was lower than in patients who had negative or inconclusive results. In the ACCENT II trial (n = 306 patients), response rates were similar among patients with (32%) or without (31%) ATIs. In this study, antibody status and effi cacy of infl iximab were not related. Four percent of patients receiving concomitant steroids and immunomodulators, 13% of patients receiving concomitant steroids alone, 11% of patients receiving concomitant immuno modulators alone, and 24% using infl iximab monotherapy developed ATIs. In both trials, infusion reactions occurred more often among patients with ATIs than among those without ATIs. In the ACCENT I trial as well as in the ACCENT II trial, assays used for the measurement of ATIs were not described in the text.
Recently, 508 biological and immunomodulator-naive Crohn disease patients were randomly assigned to receive azathioprine 2.5 mg/kg, infl iximab 5 mg/kg, or combination therapy with azathioprine and infl iximab for up to 26 weeks [25]. At 30 weeks, ATIs were detected in 1/116 patients (0.9%) receiving combination therapy and in 15/103 patients (14.6%) receiving infl iximab alone. Median trough infl iximab serum concentrations were higher for patients receiving combination therapy compared with patients receiving infl iximab mono therapy (1.6 versus 3.5 μg/mL, P <0.001). Th e assay used to measure ATIs was not described.
A small study of 30 adalimumab-treated patients with Crohn disease assessed whether AAAs aff ect adalimumab treatment outcome [16]. Seventeen percent of patients developed AAAs. Th e presence of AAAs was related to non-response to adalimumab (odds ratio 13.1, confi dence interval 1.7 to 99.2, P = 0.006). Of the 13 patients using concomitant medication (steroids or immunomodulators), only one patient (7.7%) developed AAAs, whereas 20% of patients without concomitant medication developed these antibodies. AAAs were detected with the radioimmunoassay developed by Landsteiner Laboratory Sanquin Research (Amsterdam, Th e Netherlands).
After induction therapy in the CLASSIC I trial [17], 276 patients with Crohn disease enrolled in the CLASSIC II trial and received open-label adalimumab 40 mg at weeks 0 and 2 [15]. Patients who were in remission at weeks 0 and 4 (55) in CLASSIC I were randomly assigned to receive adalimumab 40 mg every other week or weekly or placebo for 56 weeks. Patients not in remission enrolled in an open-label arm and received adalimumab 40 mg every other week. In these four groups, 17% to 33% of the patients were treated with concomitant immunosuppressive agents. Remission rates did not diff er between patients treated with or without concomitant immunosuppressants. Blood samples were collected for 269 out of 276 patients. Seven patients (2.6%) developed AAAs. Eighty-four out of 269 patients received concomitant immunosuppressants and none of them was positive for AAAs. Out of the 185 patients who did not receive concomitant immunosuppressive agents, 7 patients (3.8%) developed AAAs. Assays used for the measurement of AAAs were not described.

Unclear or no eff ect shown
Besides the studies described above (in which a benefi cial eff ect of concomitant immunosuppressive therapy on the immunogenicity of TmAbs was described), a few studies show less or no eff ect of immunomodulators on immunogenicity. In an observational cohort study on adalimumab therapy for Crohn disease (n = 168), concomitant immuno modulator therapy at baseline did not aff ect treatment outcome, trough serum concentration, or the development of antibodies against adalimumab and had no negative impact on serious adverse events. Only time to dose escalation was longer in patients who were treated with immunomodulators [14].
In a study of 106 patients with RA, 40% of antiinfl iximab antibody-positive patients were treated concomitantly with methotrexate and this frequency did not diff er signifi cantly from that of patients who were ATI-negative (50%). However, patients who were receiving methotrexate had antibody levels that were slightly lower than those of patients who were not receiving methotrexate [2]. In another study of 51 patients with RA, only three patients were not taking concomitant immunosuppressants. Antibodies were detected in two of these three patients [6].

Perspective
Since the eff ect of methotrexate on the immunogenicity of infl iximab in patients with RA was described by Maini and colleagues [4] almost 15 years ago, there has been only one other prospective study on the eff ect of concomitant medication on the immunogenicity of infl iximab. Both studies indicate a clear eff ect of methotrexate and azathioprine on the formation of ATIs in patients with RA or Crohn disease. Although no prospective studies of adalimumab on this subject have been performed, other cohort studies describing the eff ect of immunomodulator co-treatment on the immuno genicity of adalimumab show similar results. Th erefore, we conclude that there appears to be a favorable eff ect of immunosuppressive co-treatment on the immunogenicity of adalimumab and infl iximab.
Few data are available on the occurrence of adverse events associated with concomitant immunosuppressants, but even fewer data are available on the safety of anti-TmAb antibodies. Th e lack of known clinical manifestations associated with anti-drug antibodies does not imply that the continuous stimulation of the immune system and the development of immune complexes are harmless.
Th e occurrence of (serious) adverse events, or (S)AEs, did not increase when immunomodulators were added to TmAbs in Crohn disease and RA [28,29]. Only the proportion of patients with infusion reactions was lower in patients receiving immunomodulators (12.5%) compared with patients not receiving concomitant immunosuppressants (22.0%) [28]. Of 4,879 patients treated with adalimumab, 5.3% using at least one concomitant DMARD reported an SAE versus 7.3% of the patients using adalimumab monotherapy. Th is frequency did not diff er among various DMARDs [29]. Th e mechanism behind the eff ect of immunosuppressants on immunogenicity has not been elucidated. We hypothesize that by adding immuno modulators to the TmAbs, the immune response will be suppressed, leading to a decrease in antibody formation. In other words, methotrexate or azathioprine could block the expansion of the immune reactive cells, whereby the formation of anti-TmAb antibodies is reduced in quantity.
Optimization of treatment response should be the main goal when prescribing costly biological therapeutics. Especially in those infl ammatory diseases in which it is not common to prescribe concomitant immuno modulating therapy, great benefi ts in lowering the incidence of anti-TmAb antibodies could be achieved by the use of concomitant immunosuppressants, resulting in an increased portion of patients with therapeutic concentrations of TmAbs in their blood.
Th e concomitant use of immunosuppressants has not been associated with a higher incidence of (S)AEs; however, to minimize the risk of toxicity/intolerance, the minimal suffi cient dose of immunosuppressants to decrease the immunogenicity of TmAbs should be assessed. To facilitate an evidence-based consensus on prescribing concomitant immunosuppressive therapy in various infl ammatory diseases, a prospective, controlled, dose-fi nding trial is warranted.