1. Tumour Necrosis Factor-α in Rheumatoid Arthritis

The joints of patients with rheumatoid arthritis are characterised by synovitis, inflammatory exudate in the joint cavity and erosions of the cartilage and bone. The lining layer is thicker than in a normal joint and has an increased number of small vessels with infiltrating cells. Cells contributing to the thickened lining include T lymphocytes, macrophages, plasma cells, dendritic cells, fibroblasts and synoviocytes. Therefore, cell migration from the plasma to the synovium and neovasularisation seem to be important pathogenic factors in this disease.[1]

Erosion occurs where the synovial tissue lies adjacent to the cartilage and bone (sometimes called the pannus). Inflammation, proliferation and immune responses within the pannus lead to the release of metalloproteinases and other mediators responsible for the destruction of cartilage and connective tissue.[1] Erosive damage occurs early in the course of rheumatoid arthritis: 60 to 98% of patients have joint erosions within 2 years of disease onset.[2]

There is extensive expression of cytokines by the rheumatoid synovium, the most common of which are interleukin (IL)-1, IL-6 and tumour necrosis factor-α (TNFα). IL-1 and TNFα are also found at the junction of the pannus and cartilage and are responsible for:

  • controlling the secretion of multiple inflammatory cytokines

  • up-regulation of adhesion molecule expression on endothelial cells, so enhancing the influx of immune and inflammatory cells into the joint

  • induction of the synthesis and release of inflammatory mediators, e.g. metalloproteinases, prostaglandins and nitric oxide, leading to destruction of the cartilage matrix. These mediators also inhibit synthesis of collagen and proteoglycans and so hinder repair of the matrix.[1,3]

Although other cytokines are important in rheumatoid arthritis, evidence accumulated over the last decade indicates that TNFα, a proinflammatory cytokine, plays a key role in the pathological inflammatory process of the disease.[3,4] Data from in vitro experiments, animal models and, most recently, from clinical trials show that therapy directed against TNFα is associated with improvement in symptoms of the disease.[35]

To date, 2 anti-TNFα agents are available for clinical use: infliximab, an anti-TNFα antibody, and etanercept, a soluble TNF receptor (for review see Jarvis & Faulds[6]). This review focuses on infliximab (cA2), a chimaeric monoclonal antibody directed against human TNFα (fig. 1)[7] and its use in the treatment of patients with rheumatoid arthritis. It updates the previous review of this agent published in BioDrugs.[8] Infliximab is also indicated for the treatment of patients with Crohn’s disease; its use in this indication will not be covered in this review.

Fig. 1
figure 1

Schematic diagram of infliximab, a chimaeric monoclonal antibody to human tumour necrosis factor-α (TNFα). Infliximab is constructed by linking the variable (V) κ light and γ1 heavy chain regions of a murine anti-human TNFα monoclonal antibody to the constant (C) human κ light and γ1 heavy chains which are joined together by disulfide bonds.[7] Infliximab is 75% human and 25% murine protein.[9]

2. Pharmacodynamic Properties

Where possible, this section focuses on data that define the pharmacodynamic effects of infliximab in patients with rheumatoid arthritis. Importantly, however, most studies used only single doses with a follow-up period of 1 to 4 weeks. Data concerning the 3 mg/kg dose recommended for patients with rheumatoid arthritis (section 6) are limited.

2.1 Mechanism of Action

The primary mechanism of action of infliximab is through inhibition of TNFα. Infliximab binds to both soluble and transmembrane forms of TNFα at picomolar concentrations in vitro[1,7] but not to lymphotoxin (TNFβ).[9] Binding to soluble TNFα results in loss of bioactivity[1] whereas binding to membrane-bound TNFα leads to cytotoxicity by complement- and/or antibody-dependent cell-mediated mechanisms (see also section 2.2).[1,10] However, in vivo evidence defining which mechanism(s) is relevant for infliximab is lacking (for discussion see Feldmann et al.[1]).

Once the activity of TNFα has been blocked, it has been proposed that anti-TNFα agents, such as infliximab, bring about clinical benefit in patients with rheumatoid arthritis by:

  • down-regulation of local and systemic proinflammatory cytokines

  • reduction of lymphocyte migration (or trafficking) into the joints and macrophage-T-cell interactions

  • reduction of angiogenesis in the joints.[11,12]

A summary of the pharmacodynamic effects of infliximab in patients with rheumatoid arthritis is provided in table I.

Table I
figure Tab1

Effects of infliximab on leucocytes and TNFα-induced factors in patients with rheumatoid arthritis. Infliximab 1, 10 or 20 mg/kg was administered as a single intravenous infusion. Four studies were placebo-controlled,[1316] although in 1 of these[16] the statistical analysis versus baseline only was presented. Patient numbers per study ranged from 5 to 68

2.1.1 Effects on Proinflammatory Cytokines

Following an infusion of infliximab, serum levels of inflammatory mediators and the expression of some chemokines in the synovial tissue are reduced in patients with rheumatoid arthritis (table I).

Serum levels of IL-6 were significantly reduced 1 day after administration of single doses of infliximab 1 or 10 mg/kg (p < 0.05 vs placebo)[13] and the effects were sustained for 3 or 4 weeks depending on the dose (table I).[1] Of those with elevated pretreatment serum levels of IL-6, levels normalised in 70% of patients receiving a 10 mg/kg dose.[1] After a repeat dose of infliximab 3 or 10 mg/kg, decreases in IL-6 levels were again observed, although the effects were evident later (day 7), were less marked and lasted for a shorter duration.[20] Serum levels of IL-1β were also reduced 1 day after administration of infliximab 10 mg/kg (p < 0.05 vs placebo).[13]

Expression of chemokines in synovial tissue, specifically IL-8 and monocyte chemoattractant protein-1 (MCP-1), is also significantly decreased 2 weeks after administration of infliximab 10 mg/kg. Other chemokines, however, do not appear to be affected (table I).

2.1.2 Effects on Lymphocyte Migration

Infliximab is associated with a reduction of lymphocyte migration into the joints of patients with rheumatoid arthritis. Using radiolabelled granulocytes, Taylor et al.[17] demonstrated that infliximab 10 mg/kg significantly reduced cell movement into the affected joints of 10 patients with rheumatoid arthritis (p < 0.05 vs baseline; fig. 2). Over a similar time-frame, significant increases in peripheral lymphocyte counts were observed following administration of infliximab 1 and 10 mg/kg (p ≤ 0.05 vs placebo); the effects were evident after 1 day and were sustained for 1 month (table I).[14] Analysis of peripheral T cell subsets revealed that the changes were attributable mainly to increases in CD4+ and CD8+ T lymphocytes.[13,21]

Fig. 2
figure 2

Effects of infliximab on granulocyte traffic into the joints. Ten patients with rheumatoid arthritis received a single intravenous infusion of infliximab 10 mg/kg. Patients also received a bolus injection of 111In-labelled autologous granulocytes the day before and 2 weeks after the dose of infliximab. Imaging of the knees (n = 20) and hands/wrists (n = 18) was performed 22 hours after each of the radiolabelled marker injections;[17] * p < 0.05 vs baseline.

These effects are thought to be secondary to a reduction in the expression of endothelial adhesion molecules in the synovial tissue and in serum levels of soluble forms of the same molecules associated with infliximab 1 to 20 mg/kg (table I).

The ability of infliximab to downregulate lymphocyte migration into the joints may account for its protracted duration of clinical effect;[1,14] the median duration of a Paulus 20% response after a single infusion of infliximab 3 mg/kg was 6 weeks.[22]

2.1.3 Effects on Neovascularisation

Vascular endothelial growth factor (VEGF) is a potent endothelial cell-specific angiogenic factor. It is produced in the synovium and is an important regulator of neovascularisation in the pannus. In patients with rheumatoid arthritis, serum levels of VEGF are elevated by approximately 3-fold versus healthy individuals.[23]

Three weeks after a single dose of infliximab 10 mg/kg, median serum levels of VEGF were reduced by 42% in patients with rheumatoid arthritis; the decrease correlated significantly with the clinical benefit observed in these patients (p < 0.01).[23] When administered in combination with methotrexate, the decrease in VEGF (41%) was sustained for 12 weeks after the last infusion of infliximab.[23]

E-selectin and VCAM-1 also possess angiogenic activity which may be relevant in patients with rheumatoid arthritis. It is possible that the reduction achieved by infliximab in the expression of these endothelial adhesion molecules (table I) may also reduce synovial vascularity,[1] but there are no data to support this, as yet.

2.1.4 Other Effects

In patients with rheumatoid arthritis, infliximab also influences other factors which may be involved in the pathogenesis of the disease. Single infusions of infliximab 1 to 20 mg/kg significantly reduced serum levels of matrix metalloproteinases-1 (MMP-1 or interstitial collagenase), MMP-3 (or stromelysin-1), nitric oxide synthase (NOS) activity and expression of its type 2 protein product for up to 4 weeks (table I). Reductions in MMP-3 and NOS activity correlated significantly with clinical benefit in these patients (p < 0.05).[15,16]

2.2 Other Systemic Effects

In addition to the observed increases in lymphocyte counts (section 2.1.2), infliximab is also associated with transient reductions in peripheral blood monocyte and neutrophil counts. These effects may be mediated by complement- or antibody-dependent cellular toxicity of membrane TNFα-expressing monocytes.[13] Since peripheral blood monocyte and neutrophil counts can be slightly elevated in this patient group, these changes may result in normalisation of abnormalities in some patients.

Monocyte counts decreased by 20 and 29% 3 days after single doses of infliximab 1 (n = 23) and 10 mg/kg (n = 21), respectively (p < 0.01 vs placebo), and returned to baseline levels within 28 days.[14] Neither monocyte or granulocyte function appeared to be altered by infliximab.[13] It has been suggested that monocytopenia may be a mechanism by which infliximab produces clinical benefit;[24] however, in patients with Crohn’s disease, there was no difference between responders and nonresponders with regard to changes in monocyte counts from baseline.[25]

Neutrophil counts also decreased with infliximab 1 and 10 mg/kg. Reductions were evident on day 1 after treatment and were sustained for 4 weeks (p ≤ 0.05 vs placebo with both doses). Median neutrophil counts did not, however, reach the neutropenic range (<1.5 × 109/L) with either dose of infliximab.[14] Neutrophil counts in the synovial fluid were also deceased in 3 of 4 patients after an infusion of infliximab.[14]

Infliximab 10 mg/kg was also associated with significant increases in haemoglobin levels 2 and 4 weeks postdose (p < 0.001 vs placebo); haemoglobin levels increased from baseline values of 11 g/dl to 11.2 and 11.4 g/dl, respectively.[26] It has been suggested that infliximab may act directly to reverse TNFα-induced suppression of erythroid development in the marrow.[1]

3. Pharmacokinetic Properties

To date, only limited data concerning the pharmacokinetics of infliximab are available. This section focuses primarily on data pertaining to patients with rheumatoid arthritis who were also receiving concurrent methotrexate.[2729] In these studies, infliximab was administered as a 2-hour intravenous infusion.

Infliximab demonstrates dose-dependent pharmacokinetics; maximum serum concentrations (Cmax) and area under the concentration-time curve values were linearly related to dose over the range 1 to 20 mg/kg.[8,29,30] Median serum concentrations of infliximab remained steady (7.9 to 9.9 mg/L) after each of three 10 mg/kg doses given 8 weeks apart, suggesting that the pharmacokinetics of infliximab does not appear to be altered by repeat dosing.[29] No systemic accumulation of infliximab was observed when administered at a dosage of 3 or 10 mg/kg at 4- or 8-week intervals.[30]

The mean serum concentration in patients receiving the recommended dosage regimen of infliximab 3 mg/kg at weeks 0, 2 and 6 and every 8 weeks thereafter was 1.5 mg/L at week 30 (i.e. 8 weeks after the last dose).[28] Clinical response reportedly declines rapidly after serum infliximab concentrations drop below 1 mg/L.[22]

The effect of the disease-modifying antirheumatic drug (DMARD) methotrexate on the pharmacokinetics of infliximab has not been formally investigated. However, serum concentration versus time profiles suggest that while infliximab Cmax values are similar when infliximab is administered with or without methotrexate, serum concentrations of infliximab decline more slowly when methotrexate is present. Eight weeks after the last of 5 × 3 mg/kg doses, serum concentrations of infliximab were approximately 2 and <0.1 mg/L in those receiving concomitant methotrexate (n = 15 enrolled) and those not (n = 14; data estimated from graph).[27] It is possible that methotrexate, by decreasing the immunogenic potential of infliximab (section 5.1.2), may slow its rate of clearance from the blood,[31,32] although this has yet to be formally investigated.

The volume of distribution (Vd) of infliximab at steady state is independent of dose, suggesting intravascular distribution.[28,29] Median Vd ranges from 3 to 5L.[8,29] However, the degree to which infliximab distributes into tissues, specifically the joints, is unknown. Infliximab is not metabolised by the cytochrome P450 system so lessening the potential for drug interactions and variability in metabolism secondary to genetic polymorphisms of P450 isoenzymes.[9]

The clearance of infliximab is approximately 0.01 L/h[29] and it has an estimated elimination half-life of 8 to 9.5 days at the 3 mg/kg dose,[30] although longer values have been reported at higher doses (5 to 20 mg/kg; 8.9 to 12.3 days).[29] The effects of marked hepatic or renal dysfunction on clearance and Vd of infliximab are currently unknown.[30]

4. Therapeutic Use

The efficacy of infliximab in patients with rheumatoid arthritis has been assessed in several clinical trials. In order to be eligible for entry, patients had to have active rheumatoid arthritis, diagnosed according to American College of Rheumatology (ACR) criteria, that had not responded adequately to treatment with ≥1 DMARD.

Infliximab was administered either alone (section 4.1) or with background methotrexate therapy (section 4.2). Infliximab doses usually ranged from 1 to 10 mg/kg,[2629] although higher doses were used in some of the earlier trials.[29,33] In longer term trials,[2729] doses were spaced at either 4- or 8-week intervals. In all trials, infliximab was administered as an intravenous infusion.

In studies which investigated the use of infliximab with background methotrexate (section 4.2), oral methotrexate was administered at dosages of 7.5 to 15 mg/week.[2729] All patients had received methotrexate for ≥3 to 6 months at doses greater than or equal to the trial dose. Stable doses of non-steroidal anti-inflammatory agents (NSAIDs) and oral corticosteroids (≤10 mg/day prednisone equivalent) were permitted.

Many studies used ACR[34] or Paulus[35] criteria as the primary measure of efficacy. These are broadly similar composite measures of response to treatment. Assessments contributing to these criteria include tender and swollen joint counts, duration of morning stiffness, clinician and patient assessments of disease severity, functional indices and acute phase reactants [erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP)].[34,35]

4.1 Infliximab Without Concomitant DMARDs

The role of infliximab in the treatment of rheumatoid arthritis was evaluated in an 8-week phase I/II trial in 20 patients with active disease despite treatment with DMARDs. All DMARDs were discontinued ≥1 month prior to entry into the study; infliximab was given either as two 10 mg/kg infusions (1 at baseline and the second 2 weeks later) or four 5 mg/kg infusions (the first at baseline and the remainder at 4-day intervals thereafter).[33] Median swollen joint count (maximum 28) was reduced from 18 at baseline to 6 after week 6 (p < 0.001). A similarly dramatic reduction was seen in serum CRP levels (from 39.5 to 8 mg/L). A rapid response to treatment was noted, with significant reductions in swollen joint count and CRP levels evident after 1 week.[33]

The dosage regimens chosen for this initial trial were based on the premise that a short course of treatment would induce remission. However, it became evident during follow-up that patients eventually relapsed. Thus, 7 patients were retreated with up to 4 cycles of infliximab, a cycle being defined as a single 10 mg/kg infusion of the drug. Cycles were usually administered within 5 weeks of relapse occurring.[36] All 7 patients responded to retreatment, with a >80% median maximum improvement in disease activity assessments including swollen joint count and CRP. There was an overall tendency towards a shorter duration of response with successive infliximab doses, but this finding should be interpreted with caution given the small patient sample and the reduced infliximab dose used in retreatment cycles.[36]

A subsequent 4-centre randomised placebo-controlled double-blind trial evaluated the short term (4-week) effects of infliximab 1 or 10 mg/kg administered as a single infusion.[26] 73 patients with active rheumatoid arthritis were enrolled. According to the primary efficacy criterion (Paulus 20% response rate), infliximab 1 and 10 mg/kg was significantly more effective than placebo from week 1 until the final assessment at week 4 (p ≤ 0.008). The response rate with the 10 mg/kg dose was consistently greater than that achieved with the 1 mg/kg dose.[26]

4.2 Infliximab With Methotrexate

4.2.1 Single Dose Study

In patients receiving background DMARD therapy (methotrexate 10 mg/week), the efficacy of single dose infliximab 5, 10 or 20 mg/kg was compared with placebo according to a randomised double-blind protocol.[29] Only patients with active disease despite ≥3 months’ treatment with methotrexate were enrolled. 71 to 86% of patients receiving infliximab 5 to 20 mg/kg plus methotrexate achieved a clinical response (ACR 20%) at any visit during the 12-week follow-up compared with 14% of placebo plus methotrexate recipients (n = 7/group). Respective ACR 50% response rates were 43 to 71% compared with 14%. A clinical response was generally observed within 1 to 2 weeks of initiating treatment.[29]

4.2.2 Multiple Dose Studies

Subsequent multicentre randomised double-blind trials have evaluated the effects of multiple doses of infliximab over longer time periods (26[27] and 30[28] weeks) in patients receiving background methotrexate 7.5[27] or 15[28] mg/week (table II). The more recent of the 2 studies, the ATTRACT (Anti-TNF Trial in Rheumatoid Arthritis with Concomitant Therapy) study, is the most comprehensive analysis to date of the efficacy of infliximab in patients with refractory rheumatoid arthritis.[28]

Table II
figure Tab2

Multicentre randomised double-blind placebo (PLA)-controlled studies evaluating the efficacy of various intravenous infliximab (INF) schedules with or without concomitant methotrexate (MTX) in patients (pts) with refractory rheumatoid arthritis. All patients were refractory to MTX at dosages greater than or equal to those used in these studies; MTX was administered orally in the studies at dosages of 7.5[27] or 15[28] mg/week

Different primary efficacy measures were presented for each study. The ATTRACT trial[28] used ACR criteria, whereas the earlier trial[27] evaluated response according to Paulus criteria. ACR response was also measured in the latter study but, other than an assertion that results were similar to those obtained using the Paulus criteria, data were not presented.[27]

In addition to composite measures, both studies evaluated specific response parameters including swollen and tender joint counts, health assessment questionnaire (HAQ) scores, physician- and patient-assessed global scores, pain score, morning stiffness, CRP and rheumatoid factor (RF) levels and ESR. However, data for most measured parameters were not presented in the earlier study;[27] data that were presented, along with corresponding results from the ATTRACT study, are shown in table II.

In each trial, infliximab plus methotrexate was significantly more effective than placebo plus methotrexate according to primary response criteria (figs. 3 and 4). An ACR 20% response rate of 51% was observed after 30 weeks (estimated from graph) with the dosage regimen recommended for patients with rheumatoid arthritis (infliximab 3 mg/kg administered every 8 weeks plus methotrexate; section 6) in the ATTRACT trial (p < 0.001 vs placebo plus methotrexate).[28]

A substantial response to infliximab-containing regimens was evident early (within 2 weeks) in both studies.[27,28] The duration of response improved as the dosage of infliximab increased; this effect appeared more marked in the ATTRACT study.[28]

Fig. 3
figure 3

Response according to ACR criteria in patients with rheumatoid arthritis after 30 and 54 weeks’ treatment with infliximab (INF). Patients were received one of 4 different schedules of intravenous INF or placebo (PLA) in a randomised double-blind trial. All participants received concomitant oral methotrexate therapy (MTX; 15 mg/week).[28,37] ACR = American College of Rheumatology; qxwk = every × weeks; * p ≤ 0.007; ** p < 0.05 vs PLA.

Fig. 4
figure 4

Response according to Paulus criteria with infliximab (INF) in patients with rheumatoid arthritis. Patients received intravenous INF at various dosages (n = 16 to 20/group) or placebo (PLA; n = 17) with or without oral methotrexate (MTX; 7.5 mg/week) in a randomised, double-blind trial.[27] * p < 0.01 vs MTX + PLA.

The clinical response rate achieved with infliximab 1 to 10 mg/kg in combination with methotrexate was consistently greater than that achieved with infliximab alone (fig. 4).[27] Although the mechanism for this effect has yet to be formally investigated, it is possible that methotrexate partially or completely suppresses the immunogenicity of infliximab.[31,32] Combination treatment is also associated with a slower decline of serum concentrations of infliximab (section 3) and a decreased occurrence of human anti-infliximab antibodies (section 5.1.2) versus infliximab alone.[27,32]

Subgroup analysis of the ATTRACT study indicated a consistent treatment benefit with infliximab when patients were grouped according to gender, age, weight, geographic location, disease duration, anatomical class, functional class, previous joint surgery, RF, swollen or tender joint count, HAQ score, CRP, duration of methotrexate therapy, methotrexate dose or concomitant steroid treatment. Odds ratios for these subgroups ranged between 1.6 and 13.3 (with a value >1 indicative of treatment benefit relative to placebo).[38]

4.2.3 Long Term Extension Studies

In a 28-week nonblind extension of the single dose trial described in section 4.2.1, Kavanaugh et al.[29] treated 19 patients with 3 further doses of infliximab (10 mg/kg) at weeks 12, 20 and 28. Methotrexate was continued at a dosage of 10 mg/week. ACR 20% and 50% responses were sustained for 40 weeks in 11 (58%) and 7 (37%) patients, respectively. Improvements in the median number of tender and swollen joint counts, HAQ scores and CRP levels were also maintained over the study follow-up period.[29]

Results of the extension phase of the ATTRACT trial (section 4.2.2) provide more substantial evidence of sustained clinical response to continued infliximab therapy although, again, only preliminary data are available. After 54 weeks of follow-up, 42% of recipients of infliximab 3 mg/kg every 8 weeks plus methotrexate achieved an ACR 20% response versus 17% of patients treated with placebo plus methotrexate (p < 0.001 vs placebo in each treatment group; fig. 3).[37] Of considerable importance are the preliminary 1-year radiographic findings that show zero median progression of joint damage in infliximab recipients (n = 285) compared with a 7 to 8% deterioration in placebo plus methotrexate recipients (n = 63).[39]

5. Tolerability

Tolerability data from several studies, including a number of placebo-controlled trials, evaluating the therapeutic efficacy of infliximab have been pooled in a single report.[40] In order to increase the precision of this analysis,[40] patients with rheumatoid arthritis (n = 203), Crohn’s disease (n = 199) and ulcerative colitis (n = 51) were included. This strategy was justified by the supposition that patients in these trials shared certain clinical features; namely an immune system basis for disease, refractoriness to standard therapies and the receipt of similar types of concomitant medication. Thus, tolerability data are available for 453 patients treated with single or multiple doses of infliximab and 109 patients who received placebo. The dose range of infliximab was 1.5 to 81 mg/kg in patients with rheumatoid arthritis or ulcerative colitis and 5.3 to 63 mg/kg in patients with Crohn’s disease. Because placebo recipients were withdrawn or lost to follow-up more frequently than infliximab recipients, the mean duration of follow-up was somewhat longer in the latter group (22 vs 12 weeks).[40]

Tolerability data for infliximab are also available from the manufacturer’s prescribing information.[30] This includes adverse event data from 555 infliximab and 133 placebo recipients with rheumatoid arthritis over respective follow-up periods of 26.9 and 22.3 weeks (no dosages given).[30]

Overall, 76% of infliximab versus 57% of placebo recipients experienced adverse events,[40] most frequently headache, nausea, upper respiratory tract infection (fig. 5)[30,40] and infusion-related reactions (see section 5.1.1). Less frequently reported adverse events (occurring in 5 to ≤10% of infliximab recipients) included abdominal pain, pharyngitis, fever, vomiting, coughing, rash, pain, rhinitis, sinusitis, urinary tract infection, fatigue and pruritus (fig. 5).[30]

Fig. 5
figure 5

Incidence of specific adverse events in infliximab (n = 555) and placebo (n = 133) recipients during respective mean follow-up periods of 26.9 and 22.3 weeks. Data are pooled from clinical trials in patients with rheumatoid arthritis. URTI = upper respiratory tract infection; UTI = urinary tract infection.[30]

Serious adverse events likely to be associated with treatment occurred in 4.4% of infliximab versus 1.8% of placebo recipients in the pooled report.[40] These included pneumonia and fever (incidence in infliximab recipients of 0.9% each), dyspnoea (0.7%) and rash (0.4%). Prescribing information for infliximab lists a wider range of serious adverse events in patients treated with the drug; however actual incidences were not included.[30]

5.1 Sensitisation Reactions

5.1.1 Infusion-Related Reactions

The pooled analysis reported 58 (5%) nonspecific (characterised by fever or chills) reactions accompanying a total of 1207 infliximab infusions.[40] A similar incidence of nonspecific infusion reactions (4%) was reported in the prescribing information (n = 3284 infusions).[30] On a per-patient basis, the overall incidence of infusion reactions was 17% in infliximab recipients versus 7% of placebo recipients.[30] Symptoms documented were fever or chills (1%), cardiopulmonary events (including chest pain, hypotension, hypertension and dyspnoea; 1%) or both (0.1%).[30] Infusion reactions necessitating withdrawal of infliximab occurred in fewer than 2% of patients.[30]

In a subset of patients with Crohn’s disease, infusion reactions occurred more frequently in those who had developed antibodies to infliximab (ATI; see section 5.1.2) than in those who had not (36 vs 11%).[30]

5.1.2 Development of Antibodies to Infliximab

Administration of murine monoclonal antibodies can lead to a human anti-mouse antibody response. Although the potential for an immune response with chimaeric or humanised antibodies is reduced, formation of anti-idiotype of antivariable antibodies still does occur (reviewed by Feldmann et al.[1]). However, the exact clinical relevance of these antibodies is currently unknown. Also, the assay used to determine ATI levels is affected by the presence of infliximab itself.[1] For this reason, ATI levels have to be measured after drug discontinuation.

The frequency of ATI formation may be inversely related to infliximab dose. ATI formation was observed in 53, 21 and 7% of patients with rheumatoid arthritis receiving infliximab 1, 3 or 10 mg/kg, respectively, 12 weeks after the last of 5 infusions of the drug.[27] It has been suggested that higher doses may be associated with immunologic tolerance.[1,27] However, final per-dose data are not available for 27 of 87 patients in this study because infliximab was detected in the serum at sampling.[32]

Concomitant administration of methotrexate appears to reduce ATI formation with infliximab. In the same study, the rates of ATI formation were 15, 7 and 0% with the 1, 3 and 10 mg/kg doses in those receiving concurrent methotrexate.[27] In the ATTRACT study, 27 patients who discontinued infliximab treatment were tested for the presence of ATI; 3 were positive, 2 with a titre of 1 : 10 and 1 with a titre of 1 : 40.[28]

5.2 Events Related to Immune Suppression

5.2.1 Infection

The pooled analysis reported a 21% incidence of infection among 453 patients treated with infliximab compared with an 11% incidence in 109 placebo recipients.[40] Infections considered serious occurred in 3.4 and 1.8% of patients, respectively.[40] Prescribing information reports a 26% incidence of infection in infliximab recipients versus 16% in placebo recipients after respective follow-up periods of 27 and 20 weeks. Upper respiratory and urinary tract infections were most frequently reported (fig. 5).[30]

The mortality rate in the 30-week ATTRACT trial was 3% among placebo recipients and 0.6% among infliximab recipients. Two patients included in the ATTRACT study died from infectious disease during the second 6 months of the trial (disseminated tuberculosis and coccidomycosis). Although the relationship between infliximab therapy and these events is not known,[28] TNFα has an important role in the defence against intracellular pathogens.[41] Prescribing information for infliximab carries a warning regarding the risk of infection (see section 6).[30]

5.2.2 Development of Autoantibodies

Antinuclear antibodies (ANA) and antibodies to double-stranded DNA (anti-dsDNA) have been reported in association with anti-TNFα agents, including infliximab.[6,8] However, these antibodies are present in patients with certain rheumatic diseases, including rheumatoid arthritis. ANA occur in 50 to 75% of patients and anti-dsDNA in 3% of patients with rheumatoid arthritis.[42] Anti-dsDNA are relatively disease specific for systemic lupus erythematosus and occur in 70 to 88% of patients with this disorder.[42,43]

ANA were detected in 23% of infliximab versus 6% of placebo recipients during the ATTRACT study, whereas anti-dsDNA antibodies were detected in 16% (54 of 342) of infliximab recipients but not in patients who received placebo.[28] This incidence of anti-dsDNA antibodies is higher than observed in previous studies or analyses of infliximab data (8%).[27,44] Antibodies tended to appear 4 to 10 (mean 6.5) weeks after initiating treatment[44] and disappeared spontaneously during or after treatment.[27] There was no evident relationship between infliximab dosage and antibody development in the ATTRACT study.[28]

Of approximately 880 patients treated with infliximab in clinical trials, 3 (<0.5%) have developed drug-induced lupus.[45] Two of these patients had rheumatoid arthritis and received infliximab 10 mg/kg every 8 weeks or 3 mg/kg every ≤4 weeks in addition to background methotrexate; the remaining patient had Crohn’s disease. Of the 2 patients with rheumatoid arthritis, 1 experienced serositis and fever (with high levels of both ANA and anti-dsDNA antibodies) and the other experienced rash with evidence of mild perivascular infiltrate; symptoms resolved on drug withdrawal and administration of steroids.[45]

5.3 Malignancies/Lymphoproliferative Disorders

Seven malignancies (5 new and 2 recurrent) occurred among 6 of 771 (0.8%) patients treated with infliximab for up to 36 weeks in clinical trials. In patients with rheumatoid arthritis, these included B cell lymphoma (n = 1), recurrent breast cancer (n = 1) and melanoma and squamous cell cancer (n = 1).[28] It is not clear whether infliximab contributed to the development of these malignancies; however, rates are within expectations for the populations involved.[30]

6. Dosage and Administration

The recommended dose of infliximab for the treatment of rheumatoid arthritis is 3 mg/kg given as an intravenous infusion. Additional doses should be administered 2 and 6 weeks after the initial dose, and every 8 weeks thereafter. The drug should be administered in combination with methotrexate. Recommendations concerning the use of infliximab in children are, as yet, unavailable.

Infliximab carries a warning regarding the risk of infections in patients treated with anti-TNFα therapy. Serious infections, fatal in a small number of cases, have been reported (see section 5.2.1). It is noted that candidates for infliximab therapy are likely to be receiving concomitant immunosuppressive therapy that, in addition to rheumatoid arthritis, could predispose them to infections. Caution is advised regarding the use of infliximab in patients with chronic infection or a history of recurrent infection, and the drug is contraindicated in patients with clinically important active infections. Prescribers should be vigilant about the development of new infections in infliximab recipients and the drug should be discontinued if serious infection or sepsis develop.[30]

Infliximab therapy should be discontinued in patients who develop symptoms suggestive of a lupus-like syndrome.[30]

7. Place of Infliximab in the Management of Rheumatoid Arthritis

Rheumatoid arthritis is widely acknowledged as a severe and debilitating condition that considerably affects the quality of life of affected patients and leads to premature mortality.[46] Perhaps the most compelling evidence of this remains an analysis of 1972 data showing the 5-year survival rate among patients with particularly severe rheumatoid arthritis to be similar to that in patients with stage IV Hodgkin’s disease or 3-vessel coronary artery disease.[47]

The cause of rheumatoid arthritis is unclear but it is widely hypothesised that an autoreactive T cell recognises an autoantigen, initiating an inflammatory process which results in the development of inflamed synovial tissue (section 1). Rheumatoid synovial tissue then begins to proliferate and invades adjacent structures, resulting in eventual joint destruction (reviewed by Huizinga & Breedveld[48]). Despite this widespread awareness of the severity and consequences of rheumatoid arthritis, defining treatment regimens able to effectively arrest progression of the disease has proven elusive.

Historically, patients with rheumatoid arthritis were initially treated with NSAIDs, even though this drug class offers symptomatic relief only; second-line drugs (DMARDs), were withheld until joint damage became evident. However, longitudinal experience with this approach shows that the long term outcome is poor (reviewed by Pincus et al.[46]). Although NSAIDs continue to be commonly used as initial therapy in patients with rheumatoid arthritis, DMARDs are now introduced earlier in the disease course, often within 3 months of diagnosis.[49]

The primary reason for withholding DMARDs until disease progression was concern regarding toxicity. During the late 1980s, however, the DMARD methotrexate was found to have a more favourable tolerability profile than previously thought. This drug has also been shown to have greater effectiveness in the long term than other DMARDs, such as injectable gold salts and penicillamine (reviewed by Markham & Faulds[50]), and appears to have become the most commonly used DMARD in patients with rheumatoid arthritis.[6,51] Beyond this, though, there has been no real consensus as to the optimal pharmacological therapy for rheumatoid arthritis.

It appears this may change, however, with the advent of anti-TNFα therapy. Treatment with infliximab, and also etanercept,[6] has demonstrated unequivocal efficacy in controlled clinical trials in patients with progressive disease. Indeed, in all clinical trials conducted to date, patients with previous refractory rheumatoid arthritis have experienced marked improvements in functional assessments of disease severity during treatment with infliximab (section 4). Importantly, this beneficial effect occurred relatively rapidly, with significant improvement evident as early as 1 week after initial administration of the drug. This attribute represents an important advance over all available DMARDs which are slow to exert their efficacy.

Also of considerable importance is preliminary data suggesting infliximab therapy may arrest joint destruction. One-year x-ray data from the pivotal ATTRACT trial showed no median progression of joint damage in patients treated with infliximab plus methotrexate. In contrast, a 7 to 8% deterioration in radiographic scores was reported in methotrexate plus placebo recipients.[39] Indeed, the longest term follow-up data available show that a response to infliximab plus methotrexate is sustained for up to 54 weeks with repeated administration according to standard ACR criteria.

Opinion leaders in the rheumatology field have responded with cautious optimism to clinical trial results with TNFα-blocking drugs and a number of position statements on how these drugs might be rationally introduced into clinical practice have recently appeared.[5,52,53] The most recent of these, a consensus statement representing the views of ≈80 rheumatologists,[53] suggests TNFα-blocking agents should be used only by experienced rheumatologists who have the skills necessary to conduct long term efficacy and tolerability assessments. The statement acknowledges cost considerations and reinforces the fact that subpopulation data from clinical studies are still being acquired.[53]

Infliximab has been approved by the US Food and Drug Administration for the treatment of rheumatoid arthritis based on the results of the ATTRACT study. The drug is, however, only approved for use in rheumatoid arthritis when combined with methotrexate. The latter agent, as described above, is widely accepted as the DMARD of choice. Thus, for now at least, patients with disease of sufficient severity to warrant treatment with anti-TNFα therapy will most likely already be receiving methotrexate.

Unlike infliximab, etanercept may be prescribed in the absence of background methotrexate therapy (table III). The dosage regimens of the 2 agents also differ. A shorter elimination half-life (median 4.8 days) means that etanercept is administered twice weekly[54] compared with the less frequent 8-weekly dosing schedule required for infliximab. Etanercept is administered by the subcutaneous route whereas infliximab is given as an intravenous infusion (table III).

Table III
figure Tab3

Comparison of the anti-tumour necrosis factor-α agents infliximab and etanercept[6,30,54]

Overall infliximab has been well tolerated in the clinical trials programme. Initial toxicity concerns raised during small scale preliminary studies have been mitigated somewhat by longer term data. Importantly, the overall incidence of acute infusion-related events has proven considerably lower than that seen in early studies. Nevertheless, serious adverse events, including fatal infections, have occurred. Drug-induced lupus has been observed in <0.5% of patients and at dosages higher than those recommended for the treatment of rheumatoid arthritis. The precise clinical relevance of the formation of ATI is, as yet, unknown.

Like infliximab, etanercept is associated with infections, most commonly upper respiratory tract reactions, and rare cases of serious infections (table III).[6] Both agents are subject to several precautions concerning these events.[30,54] Autoantibodies, ANA and, more rarely, anti-etanercept antibodies have also been reported with etanercept (table III)[6] suggesting that anti-TNFα agents may have similar tolerability profiles. Until a direct comparison of the 2 agents is available, however, the relative frequency and intensity of associated adverse events cannot be assessed.

Thus, infliximab represents an important advance in the treatment of rheumatoid arthritis, with tolerability concerns raised by early studies having been eased somewhat by more recent data in larger patient numbers. If preliminary results indicating that infliximab is able to arrest joint destruction in patients with rheumatoid arthritis are corroborated, the drug will likely become an integral component of future management strategies for this difficult-to-treat condition.