Bacterial infections, including sepsis and pneumonia, invasive fungal infections, and other opportunistic infections (for example, pneumocystosis, candidiasis, listeriosis, aspergillosis), have been reported with the use of TNF inhibitors 91112. Reactivation of latent tuberculosis following treatment has led to the introduction of pre-initiation screening procedures, which have successfully reduced the number of reported cases 1617. The risk of reactivation of latent tuberculosis is, of course, dependent on the incidence of latent infection and is associated with all TNF inhibitors 1819. Some registry data, however, suggest that the risk may be lower with etanercept 202122. In RA patients, risk factors include active longstanding disease, age, country of origin, history of exposure to a person with tuberculosis, concomitant use of immunomodulators, and disease activity 23. Physicians should remain alert to the development of symptoms related to tuberculosis or other infections.

Owing to adverse effects observed during clinical trials, patients with congestive heart failure should be closely monitored if they are receiving TNF inhibitors 91112. Other rarely reported conditions possibly related to use of TNF inhibitors include demyelinating disease, seizures, aplastic anaemia, pancytopaenia, and drug-induced lupus 91112. Physicians should remain vigilant for the development of these conditions 16.

The formation of antibodies to biologic agents is a significant issue because antibodies have the potential to reduce the efficacy of the agent or to cause adverse events 10. All three TNF inhibitors have been associated with the development of antibodies, although etanercept does not appear to generate neutralising antibodies 9101012242526. The use of MTX in combination with TNF inhibitors appears to reduce the incidence of antibody formation 10111224.

In a cohort study of 53 patients receiving etanercept for AS without MTX, mean etanercept levels in responders and nonresponders at 12 and 24 weeks were similar, and no antibodies to etanercept were detected 27. No correlation was found among etanercept levels, formation of antibodies to etanercept, and clinical response. Conversely, in a 54-week cohort study of 38 patients receiving Infliximab for AS, detection of antibodies to Infliximab was associated with undetectable serum trough Infliximab levels and reduced response to treatment 28.

A look at the cellular and molecular levels of diseases in rheumatology demonstrates that such diseases share common mechanisms and may be more closely related than previously recognised. Rigorous studies have examined the mechanisms of action of the anti-TNF inhibitors, particularly Infliximab and etanercept; however, many questions remain unresolved 1. For example, although both Infliximab and etanercept are useful in the treatment of peripheral arthritis and AS, there appear to be differences in their effects at the cellular level 2930. Moreover, while their actions in AS have yet to be fully elucidated, the long-lasting suppression of T-cell function apparent during treatment with Infliximab suggests that neutralisation of soluble TNF cannot be the only mechanism 29. Possible mechanisms generally fall into two categories: those mediated by blockade of the TNF receptor, and those mediated by induction of trans-membrane TNF. Several mechanisms probably act simultaneously.

To what extent various mechanisms contribute to drug efficacy remains an open question. All of the anti-TNF agents bind to transmembrane TNF and could theoretically induce both complement-dependent cytotoxicity and antibody-dependent cellular cytotoxicity, although at lower levels for etanercept compared with the anti-TNF agents Infliximab and adalimumab 1. The roles of apoptosis and inflammation reversal for reducing cellularity in rheumatoid synovial tissue during anti-TNF therapy are unclear 1. A study by Wijbrandts and colleagues analysed apoptosis in peripheral blood and synovial tissue within 24 hours of treatment with Infliximab in patients with RA. There were no signs of apoptosis induction in peripheral blood monocytes or lymphocytes after Infliximab treatment. These results support the view that the rapid decrease in synovial cellularity observed after initiation of anti-TNF therapy cannot be explained by apoptosis induction at the site of inflammation 31.

The TNF inhibitors all require parenteral administration, either intravenously (Infliximab) or via subcutaneous injection (etanercept, adalimumab) 4. The availability of different formulations allows tailoring of treatment to the individual and ensures that the patient is receiving maximal benefit with minimal negative impact on their quality of life. Although some patients appreciate the control offered by self-administration of subcutaneous injections, others do not like to self-inject. Intravenous drugs can be inconvenient because of the need for regular hospital visits, but some patients desire regular contact with medical professionals. The decision on whether to use an intravenous or subcutaneous product should be based on the clinician’s and patient’s goals for treatment.

Intravenous administration allows high serum concentrations to be rapidly achieved, and therefore offers the potential for fast, complete suppression of inflammation 3233. Rapid improvement in signs and symptoms has been observed following the usual clinical dose of Infliximab (3 mg/kg) in RA patients 34. Within 48 hours of administration, patients experienced significant improvements in the mean duration of morning stiffness, patient assessment of pain, physician global assessment of arthritis, and patient global assessment of arthritis compared with baseline measurements. Studies using a high-dose infusion of Infliximab (10 mg/kg) in RA patients have shown significant reductions in C-reactive protein levels 3536, improvements in Disease Activity Score (DAS) and American College of Rheumatology (ACR) response 37, and significant reductions in bone resorption as measured by β-CrossLaps, a predictor of annual bone loss in RA, as soon as 24 hours post infusion 37. The benefits of higher doses, however, must be weighed against accompanying increases in side effects. Additionally, Infliximab therapy has demonstrated a reduction in the number of inflammatory cells, including intimal and sublining macrophages, T cells, and plasma cells, in rheumatoid synovial tissue as soon as 48 hours after initiation of treatment 33. Although unlicensed, intravenous administration of adalimumab also has demon strated a rapid onset of clinical effect 38. Whether intravenous administration of TNF antagonists has a faster effect than subcutaneous administration is not known presently, as no direct comparisons have been published.

Subcutaneous agents may be appropriate for and preferred by some patients. Although drug absorption into the bloodstream is slower and a delay of several days is possible before maximal concentrations are reached, desired outcomes can be achieved. While a rapid onset of effect for intravenous administration has been established, there is on average no clear-cut difference in long-term overall efficacy outcomes between subcutaneous and intravenous administration.

Although TNF inhibitors are currently the gold standard of biologics for patients with inflammatory arthritides, there are still a number of outstanding questions regarding how to gain the maximum benefit from these agents. The most recent ACR guidance stating that patients with early RA are not candidates for biologic therapy 18 is debatable. There are convincing data indicating that the use of biologics early in the course of the disease can be highly efficacious and may induce clinical remission in a certain percentage of patients 1315394041. Additional data may spur modification of guidelines and practice for those early RA patients who do not respond sufficiently to conventional treatment. Of importance, a well-defined referral pathway within health care systems is needed to identify patients early in the course of the disease. Also, family physicians and other healthcare professionals must be educated about the early symptoms of inflammatory arthritides, with an emphasis on the importance of early referral to rheumatologists for diagnosis and treatment 42.

Likewise, additional studies are needed to determine whether patients with co-morbidities or those taking concurrent medications require monitoring for specific toxicities 4. Several registries have reported a high prevalence of co-morbid conditions in RA patients who are commencing biologic therapy in routine practice 4344. Oldroyd and colleagues compared 354 patients with AS from the Australian Rheumatology Association Database who were commencing biologic therapy with more than 1,000 enrolees from four RCTs involving biologic therapy. At baseline, patients from the Australian Rheumatology Association Database – considered representative of the general population seeking clinical care – were found to have much higher levels of co-morbidity than the RCT subjects, as well as significantly greater disease activity. These findings have important implications for patient monitoring 45.

In a broader sense, RA trial inclusion criteria may need to be less restrictive 46. A comparison of 546 RA patients from the Dutch Rheumatoid Arthritis Monitoring registry with 1,223 RA patients from 11 RCTs showed much greater disease activity at baseline in RCT enrolees 47. The efficacy of TNF-blocking agents was lower in Dutch Rheumatoid Arthritis Monitoring registrants. For example, in 10 of the 11 comparisons, the ACR 20% improvement criteria (ACR20) response rate was lower in the registry cohort (again, representative of daily clinical practice) than in the RCT group, and the difference was significant in five of the 11 comparisons. These data indicate a smaller, real-world effect of anti-TNF treatment than the effect seen in trials. The discrepancy may be due to continued use of co-medication and selection toward greater disease activity in RCTs.

Zink and colleagues obtained similar results during their comparison of 1,458 patients from the Rheumatoid Arthritis Observation of Biologic Therapy registry with data from five major RCTs that led to approval of biologics for RA. Only 21 to 33% of Rheumatoid Arthritis Observation of Biologic Therapy registrants would have been eligible for the trials, and this ineligible group demonstrated lower TNF inhibitor response rates than RCT enrolees who received biologic therapy. The investigators concluded that observational cohort studies, which include a full spectrum of patients (for example, with various co-morbidities, taking assorted concomitant medications), are essential to complement RCT data 46. A study of 417 RA patients from the Danish Database for Biological Therapies in Rheumatology further supports these clinical practice data. In the majority of these routine care patients, TNF antagonists were not successful in controlling disease, although they did achieve moderate overall success in controlling clinical inflammation 48. Clearly, a bridge is needed between trial results and real-world results.

Some studies have hypothesised that TNF inhibitors may have the potential to repair RA joint damage 4950. The data to support this notion are currently negligible, however, and tools to measure and evaluate repair must be developed before in-depth investigations can be launched.

In one study, a small number of patients experiencing RA symptoms for <12 months but considered to have a poor prognosis were randomised to receive either Infliximab plus MTX (n = 10) or placebo plus MTX (n = 10) for 1 year 51. Patients receiving Infliximab experienced significant improvements in all measures at the end of year 1 compared with those receiving placebo. The Infliximab patients then received MTX alone for an additional year, and 70% of patients maintained the Infliximab responses, as measured by the C-reactive protein level, DAS in 28 joints (DAS28), and Health Assessment Questionnaire results 51.

van der Kooij and colleagues recently compared the clinical and radiological efficacy of initial (n = 117) versus delayed (n = 67) treatment with Infliximab plus MTX in patients with early RA in a post hoc analysis of the BeSt study 52. After 3 years of treatment, patients receiving initial Infliximab plus MTX demonstrated more improvement in functional ability over time, as measured by the Health Assessment Questionnaire, and were less likely to have radiological progression than patients treated with delayed Infliximab plus MTX. These results suggest that initial treatment with a biologic-plus-DMARD combination in patients with recent-onset RA is more beneficial than reserving such treatment for patients in whom traditional DMARDs have failed 52.

The PREMIER study compared the efficacy of early intervention with a combination of adalimumab and MTX versus either agent used alone as monotherapy in patients with early, aggressive RA 15. The primary end points in this 2-year, double-blind, controlled study (n = 799) were the percentage of patients in whom an ACR50 response was achieved and the mean change from baseline in the modified Total Sharp Score, which assesses bone erosion and joint space narrowing on radiographs. Combination therapy was superior to adalimumab and MTX mono therapy in all outcomes measured. At year 1, patients treated with combination therapy had a mean increase in Total Sharp Score of 1.3 units compared with 3.0 units in those receiving adalimumab monotherapy (P = 0.002) and of 5.7 units in those receiving MTX monotherapy (P <0.001). At year 2, patients receiving combination therapy continued to have significantly less radiographic progression (mean change 1.9 Sharp units) compared with those treated with either adalimumab (5.5 units) or MTX (10.4 units) monotherapy (P <0.001 for both comparisons). Although ACR responses were comparable in the two monotherapy arms, there was significantly less progression in the adalimumab arm compared with the MTX arm at 6 months (2.1 vs. 3.5), 1 year (3.0 vs. 5.7) and 2 years (5.5 vs. 10.4) (P <0.001 for all comparisons). This is another study suggesting the value of combination therapy in early RA 15.

Van der Heijde and colleagues have hypothesized that therapeutic intervention early in the disease course has a disproportionate benefit on outcome if treatment is started early in the disease course 51. Additionally, drug-free remission may be a realistic goal in some patients with early RA. In the BeSt study, 19% of patients who received Infliximab plus MTX in a DAS-steered, tightly controlled manner were in drug-free remission at 5 years, for a mean duration of 22 months. Infliximab had been successfully discontinued in 58% of patients, while 18% were still receiving combination therapy. Furthermore, compared with other treatment strategies, initial temporary treatment with Infliximab plus MTX resulted in significantly better functional ability over 5 years 53. These studies raise the possibility that if aggressive treatment to induce remission is instituted very early in the course of RA, more conservative management strategies may be sufficient to maintain that remission.

The use of TNF blockers for early-stage PsA is currently under discussion. For early-stage AS, one study showed Infliximab to be highly efficacious in patients who were positive for HLA-B27, had recent-onset inflammatory back pain, and had early sacroiliitis demonstrated by magnetic resonance imaging 54.

Additional unmet needs include: the ability to predict clinical response so that these drugs, which are expensive and have the potential for serious toxicity, can be targeted to patients who would most benefit 55; an understanding of acquired drug resistance to anti-TNF agents 56; a full explanation for why patients with spondyloarthritis (a group of disorders that includes AS and PsA) have a 20% lower probability of discontinuing TNF antagonists than patients with RA 57; and an understanding of reasons for and predictors of discontinuation.

Relative to the first point, the search for predictors of response is important in the context of personalised medicine, with the aim of increasing the percentage of patients exhibiting a robust response to a given treatment. Wijbrandts and colleagues recently studied arthroscopic synovial tissue in 143 patients with active RA prior to initiating treatment with Infliximab 58. Their analysis confirmed that the baseline level of TNF expression may be a significant predictor of response to anti-TNF therapy. At baseline, TNF expression in the intimal lining layer and synovial sublining was significantly higher in responders than in nonresponders (clinical response determined at week 16) (P = 0.047 and P = 0.008, respectively). The number of macrophages, macrophage sub sets, and T cells was also significantly higher in responders than in nonresponders 58. The relationship between synovial lymphocyte aggregates and the clinical response to Infliximab has also been studied in RA patients 59. Synovial tissue biopsy samples were obtained from 97 patients with active RA before initiation of Infliximab treatment. Lymphocyte aggregates were counted and graded for size, and logistic regression analysis identified whether the presence of lymphocyte aggregates could predict clinical response at week 16. The majority (57%) of RA synovial tissues contained lymphocyte aggregates. Additionally, aggregates were found in 67% of clinical responders compared with 38% of nonresponders. The presence of aggregates at baseline was a highly significant predictor of the clinical response to anti-TNF treatment (P = 0.008), demonstrating that RA patients with synovial lymphocyte aggregates may have a better response to Infliximab treatment than those with only diffuse leucocyte infiltration 59.

Relative to the fourth point, 21 to 35% of patients discontinue TNF-blocking agents within the first year 60. Reasons for discontinuation appear to include lack of response, loss of response, development of intolerance, partial efficacy, and adverse events 6162. Switching to a different TNF inhibitor may be an option for some patients 63. One limited study with 31 enrolees suggested that when etanercept is not efficacious, Infliximab may offer gains, and that when Infliximab fails due to adverse events, etanercept may allow continuation 61. Another larger study (complete data for 197 patients) in RA suggested that a second TNF inhibitor may be effective after failure of the first inhibitor, regardless of the reason for discontinuation of the first agent 60. Conceivably, efficacy of a second TNF blocker may be lower in primary nonresponders to a first TNF blocker (response being defined at 12 to 16 weeks after initiation of treatment). Switching to a different mechanism of action and agent, such as rituximab, abatacept, or tocilizumab, is also an option (see below).

Identifying predictors of discontinuation would be valuable in managing disease and targeting therapies to patients most likely to benefit. Currently, treatment choices are dominated by patient and physician preference, side-effect profiles, and cost 64. A cohort (n = 503) from the Brigham Rheumatoid Arthritis Sequential Study was examined to identify clinical predictors associated with discontinuation of TNF inhibitors 64. In this study, 210 out of 503 patients (42%) discontinued therapy. Unfortunately, only 63 patients gave a reason; the investigators therefore shifted to a model-based analysis. The results showed that higher risk of discontinuation was associated with prior use of another TNF agent. Lower risk of discontinuation was associated with longer disease duration, prior use of DMARDs, and longer MTX use.

More information is clearly needed with regard to individualising physician/patient decision-making about initiating anti-TNF agents, switching agents, and predicting efficacy and tolerability. Lowering the discontinuation rates is an important current goal.

Rituximab, a chimeric anti-CD20 monoclonal antibody, was the first B-cell agent approved for treatment of RA 82. This antibody was approved in combination with MTX in the United States and Europe in 2006 for adult patients with, respectively, moderate to severe active RA or severe active RA, after the failure of at least one TNF inhibitor. The agent targets B cells, rather than the entire immune system, and is administered by intravenous infusion to patients with an inadequate response to TNF inhibitors 83. Rituximab has been shown to inhibit progression of structural damage in RA over 2 years, and continues to inhibit joint damage with long-term treatment 3984.

In the event of inadequate efficacy with a TNF inhibitor, some have suggested that switching patients to rituximab is a more effective management strategy than switching to another TNF inhibitor 85. A prospective cohort study of 318 RA patients found that when the motive for switching to rituximab was TNF inhibitor ineffectiveness, disease improvement was significantly better than with an alternative TNF inhibitor 85. If the reason for switching is not lack of efficacy (for example, adverse events, patient preference), there is no advantage in switching to rituximab 85.

Immunoglobulin levels have been found to be lower in patients receiving rituximab in the long term for RA 86. An initial apparent trend toward higher rates of serious infection in this population may have been discounted by an open-label study of 1,039 RA patients 87. The serious infection rate was 5.0 per 100 patient-years, similar to that for etanercept, Infliximab, and adalimumab (5.3 per 100 patient-years) 88. There also have been reports of psoriasis and PsA developing in RA patients receiving rituximab 89; however, the same is true for TNF inhibitors 90. The development of progressive multifocal leukoencephalopathy or hepatitis B reactivation during rituximab treatment for RA is very rare.

Abatacept is a T-cell co-stimulation modulator administered by intravenous infusion. The modulator is thought to prevent the activation of T lymphocytes, including naïve T cells 9192. Abatacept was approved in the United States and Europe in 2005 for treatment of RA in adult patients with an inadequate response to DMARDs or TNF inhibitors. In January 2010 it was approved in Europe for moderate-to-severe active polyarticular juvenile idiopathic arthritis in patients 6 years of age and older. Because abatacept was the first therapy targeting the inhibition of co-stimulatory signals to prevent T-cell activation, its use in early disease 93 and in biologic-naïve patients with active RA 94 has generated particular interest and investigation 91959697. These data may support the use of abatacept in biologic-naïve patients with early disease who have had an inadequate response to MTX.

The magnitude of abatacept’s effect appears to increase over time. According to the initial report of the Abatacept in Inadequate Responders to Methotrexate, Abatacept or Infliximab versus Placebo, a Trial for Tolerability, Efficacy, and Safety in Treating Rheumatoid Arthritis study, clinical response and disease activity were not only maintained from 6 to 12 months, but also appeared to improve 98. The report containing 2-year results is currently only in abstract form but shows that reduced disease activity was maintained with ongoing abatacept treatment 9499. Abatacept has also demonstrated an increasing and significant degree of inhibition of structural damage progression in patients receiving treatment for 2 years 95. Abatacept may have an increasing disease-modifying effect on structural damage over time in the majority of patients who respond to treatment. To date, this is a unique observation among biologic treatments for RA.

The long-term efficacy and safety of abatacept have been demonstrated over 5 years with a dose of 10 mg/kg 97. In a long-term extension trial, abatacept was well tolerated and provided durable improvements in disease activity, with no unique safety events reported. These data, combined with relatively high retention rates, confirm that abatacept provides sustained clinical benefits in RA. Additionally, abatacept has been shown to provide clinical benefits in patients with RA who have previously failed TNF inhibitor treatment, regardless of the previous TNF inhibitor(s) used or the reason(s) for treatment failure 100. This finding suggests that switching to abatacept may be a useful option for patients who fail TNF inhibitor treatment.

Tocilizumab is a humanised anti-IL-6-receptor monoclonal antibody administered by intravenous infusion. This antibody inhibits signals through both membrane and soluble IL-6 receptors 101. Tocilizumab has received approval in Europe and the United States (January 2009 and 2010, respectively) for the treatment of moderate to severe RA in adult patients who have responded inadequately or have been intolerant to previous therapy with one or more DMARDs or TNF antagonists.

Tocilizumab used as monotherapy or in combination with MTX has demonstrated superiority over MTX mono therapy in reducing disease activity in RA over 24 weeks 102103. Furthermore, tocilizumab has resulted in significant improvements compared with placebo in physical function, fatigue, and physical and mental health scores over 24 weeks in patients who fail to respond to conventional DMARD therapy alone 104. Tocilizumab has also demonstrated efficacy in RA patients who fail to achieve an adequate response with or became refractory to TNF inhibitors 105.

There is a close relationship between normalisation of serum IL-6 levels following treatment with tocilizumab and clinical remission. In the phase III SATORI trial, patients whose serum IL-6 levels became normal tended to achieve DAS28 remission. Normal IL-6 levels may therefore provide a good marker to identify patients who can stop tocilizumab treatment without the risk of flaring 106107.

In the 3-year extension of the SAMURAI study, patients with early RA treated with tocilizumab exhibited strongly suppressed radiographic progression 108. Further more, radiographic progression was more effectively suppressed in patients who received tocilizumab at the start of the trial than in those who received conventional DMARDs at the start. Early introduction of tocilizumab treatment may therefore be more effective in preventing joint damage. The LITHE study in 1,196 patients who had inadequate responses to MTX further supports the potential for tocilizumab to suppress radiographic progression 109. Patients also demonstrated improvements in physical function.

Tocilizumab has a well-characterised safety profile, with infections being the most common adverse event in trials 101109. Safety data pooled from five pivotal tocilizumab studies demonstrate rates of serious infection of 3.5 per 100 patient-years for the 4 mg/kg dose and of 4.9 per 100 patient-years for the 8 mg/kg dose compared with 3.4 per 100 patient-years for the comparator groups over a median 3.1 years’ treatment duration 109. Physicians should also monitor for decreased neutrophil counts and increased lipid or liver enzyme levels, and manage appropriately 101109.

Certolizumab is a pegylated Fab fragment of a humanised anti-TNF monoclonal antibody that neutralises the activity of TNF 66. Certolizumab was approved for treatment of RA in combination with MTX in the United States and Europe in 2009. The use of pegylation increases the half-life of the molecule and eliminates the chimeric Fc portion. It is therefore hoped that adding poly ethylene glycol will produce a longer-lasting compound with fewer side effects, although it remains to be established whether pegylation does indeed confer these advantages in clinical practice 66.

Subcutaneous administration of 400 mg certolizumab every 4 weeks as monotherapy has demonstrated a rapid onset of response and reduction in RA disease activity as early as week 1 110. When used in combination with MTX, certolizumab (400 mg at baseline weeks 2 and 4, then 200 or 400 mg every 2 weeks) reduces radiographic progression compared with MTX alone over 1 year, and the difference is already significant at 6 months 111.

Golimumab is a fully human anti-TNF IgG₁ monoclonal antibody that targets and neutralises both the soluble and membrane-bound forms of TNF 66. Golimumab was recently approved for monthly subcutaneous treatment of adults with RA, PsA, and AS. A randomised, double-blind, placebo-controlled dose-ranging study compared subcutaneous injections of golimumab with placebo in patients with active RA despite treatment with MTX 112. In this study, greater efficacy was demonstrated for golimumab 50 mg every 4 weeks in addition to MTX compared with MTX plus placebo in terms of ACR responses. Furthermore, 20% of patients receiving golimumab achieved DAS28 remission at week 16, compared with only 5.7% (P = 0.074) of patients receiving MTX alone. Over a 52-week treatment period, all clinical responses achieved at week 16 were maintained and/or improved, and no unexpected safety issues were observed 112.

These results have been further confirmed in a phase III study in patients with established RA and disease activity despite treatment with MTX monotherapy 113. Additionally, golimumab demonstrated efficacy in patients with established RA who had previously received other TNF inhibitors and in MTX-naïve patients 114115.

Efficacy has also been demonstrated in patients with PsA and AS treated with golimumab 116, similar to that for currently available TNF inhibitors 117118. Furthermore, golimumab is capable of increasing function in patients with AS 118. In PsA, golimumab has also demonstrated improvements in psoriatic skin and nail disease 116.

Ustekinumab is a human monoclonal antibody directed against the p40 subunit of IL-12/IL-23 that has demonstrated efficacy in PsA 119. In a parallel-group crossover study involving 146 patients, a significantly higher proportion of ustekinumab-treated patients achieved a response using ACR criteria compared with placebo-treated patients at week 12. Ustekinumab was approved in 2009 in both the United States and Europe for treatment of patients with moderate-to-severe plaque psoriasis. Ustekinumab has not been approved for PsA.