Abatacept is administered as a short intravenous infusion at a dose of approximately 10mg/kg. After the first infusion, abatacept is administered at 2 and 4 weeks and then every 4 weeks.5 Details on dosage, route of administration, and precautions to take before and during the administration are shown in Table 2. Intravenous administration of abatacept has the characteristic that is administered in 30min and rarely requires the use of premedication since infusion reactions are infrequent.6,22,23 In general, the dose and administration intervals of abatacept are often kept constant over time.24

Although abatacept is expected to be approved for subcutaneous administration within the next few months,25,26 at present in Europe it is only available for intravenous administration. Subcutaneous administration offers patients the convenience of being home administered. In any case, the patient's preferences with respect to the route of administration must be explored. It is reported that 50% of patients prefer the intravenous to subcutaneous administration due to the calming effect of the presence of medical personnel and security of hospital treatment, the distaste for self-administration and frequency of administration, generally more spaced intervals than with subcutaneous.27

There is a substudy of the AGREE trial in which patients who achieved remission at 2 years (ESR DAS28<2.6) were randomized in a double-blind fashion to receive abatacept at doses of 10mg/kg or 5mg/kg. One year after treatment, the proportion of patients with reactivation of the disease was similar in both groups.5

Meta-analysis of Maxwell et al. shows a clear efficacy of abatacept vs placebo.6 A phase IIb clinical trial in patients with established RA and inadequate response to methotrexate (MTX) showed that abatacept plus MTX was more effective than MTX monotherapy, with ACR20 response rates at 6 months of 60% with abatacept and 35% with placebo.28 This significant clinical improvement persisted at 12 months of treatment, with remission rates DAS28-CRP (<2.6) higher in the abatacept-treated patients than in those receiving placebo (35% vs 10%).29 The AIM30 study, a randomized phase III multicenter, double blind trial, compared abatacept (+MTX) vs placebo (+MTX) in 652 patients with an inadequate response to MTX. The ACR20 response rate at 6 months, the primary endpoint, was significantly higher in the abatacept-treated than in the placebo group (68% vs 40%). These differences were even slightly increased after one year of treatment (73% vs 40%). ACR50 and ACR70 responses were also significantly higher in the abatacept-treated group than in the placebo group at both 6 and 12 months. Significant differences between groups were observed after 30 days of treatment. There was also a significant improvement in HAQ-measured disability assessed in the active treatment group vs placebo at one year.

ATTEST31 study, a double-blind trial, separately analyzes the effectiveness of abatacept and infliximab vs placebo in patients with RA and inadequate response to MTX. At 6 months of treatment, both biological agents were superior to placebo. There is a greater speed of action of infliximab in the first 3 months and slower abatacept ACR20 responses at one year. The study was not designed to specifically evaluate the comparative efficacy of the 2 biological agents.

The ATTAIN32 study is a randomized, double-blind, randomized, placebo-controlled trial evaluating the efficacy of abatacept vs placebo in patients with failure to TNF inhibitors (infliximab or etanercept). 393 patients with active RA were included after a washout period of TNF inhibitors and were randomized 2:1 to treatment with abatacept (+MTX and/or FAME) or placebo (+MTX and/or FAME). At 6 months of treatment, ACR20 response rates, the primary endpoint, were significantly higher in the abatacept-treated group than in the placebo group (50% vs 20%), as were ACR50 (20% vs 4%) and ACR70 responses (10% vs 2%). There was also a significant improvement in disability (HAQ improvement>0.3) in the abatacept group compared to placebo (47% vs 23%).

The ARRIVE33 study not only is an open trial of abatacept in 1046 RA patients who had failed to TNF inhibitors, whose primary objective was to evaluate the safety of abatacept administered with or without washout of TNF inhibitors, but also analyzes their effectiveness, demonstrating a significant improvement of clinical activity of RA (reduction of DAS28-PCR greater than 1.2) in more than 50% of patients at 6 months of follow up.

In the AGREE study, abatacept with MTX also demonstrated more effectiveness than MTX monotherapy in patients with recent onset RA who had not previously received MTX and had poor prognostic factors, such as positive autoantibodies or x-ray34 erosions. The rate of remission (DAS28-CRP) at 12 months was significantly higher in patients treated with abatacept+MTX than in those assigned to placebo+MTX: 41% vs 23%. Abatacept, however, has no indication at present for the treatment of patients who have not received prior MTX or DMARD.

Before presenting the results of studies, it is important to note that in the assessment of structural damage with abatacept, the Genant-modified Sharp van der Heijde was used instead, which is usually the one used in the trials of inhibitors TNF, and the Genant index ranks damage lower (score of 0–290) than the van der Heijde (0–448), although both correlate.35

In the AIM30 study the difference in the rate of erosion and total Genant score after a year of treatment compared to baseline was significantly lower with abatacept+MTX than with placebo+MTX (erosion rate: 0.63 with abatacept compared to 1.14 in placebo, P=.029, total Genant index: 1.21 vs 2.32, P=.012),30 which means a reduction of radiological progression of 45% for erosions and 48% in the total index. This reduction is lower than that obtained with TNF inhibitors, which is approximated to 90%.36 The difference is probably not due to the use of the Genant index instead of the van der Heijde, but with abatacept improvement is slower.

These results are confirmed in the AGREE study performed in RA patients without prior MTX treatment, who showed a 39% reduction in the progression of total damage and 44% for erosions.34 After a year, 61% of patients treated with abatacept+MTX had not progressed radiographically compared to 53% among patients who received MTX alone, representing an estimated difference of 8% between groups.34 No details of radiological progression in patients with a previous TNF inhibitor were given, nor is it known whether there is significant radiological progression in patients with no clinical response to abatacept.

Regarding the long-term progression of the 547 patients who completed the AIM study, 539 (83%) were treated with abatacept during the open period. Mean radiographic progression was significantly reduced in the 2 groups of patients, but more so in those initially treated with abatacept: 1.07 and 0.46 vs 2.40 and 0.75 in the 1st and 2nd years, respectively.37 The rate of radiological progression was reduced even more during the 5 year open follow-up period however, until the 5th year, the rate of progression in the placebo group initially did not match those initially treated with abatacept (Fig. 2).38 The AGREE study data at 2 years confirmed these results. The progression of structural damage in the 2nd year was 57% lower than in the first year in patients initially treated with abatacept. At 2 years in patients who were treated with abatacept+MTX there was less radiographic progression than those who had been treated with MTX alone (0.84 vs 1.75 in the total score, P<.001).39 A recently study published analyzing all patients together, irrespective of initial treatment, showed that the total radiographic progression was significantly reduced in the 3rd year of treatment compared to the 2nd year (P=.022).40

Fig. 2.

Mean change in the Genant radiological score followed for 5 years in the AIM study. X-axis reflects absolute change of the Genant score with respect to the study time periods. The first year includes the blind abatacept or placebo period and afterwards patients on placebo received open label treatment with abatacept.91

(0.1MB).

Both phase IIb trials and AIM,40–43 as well as the ATTEST and AGREE38,39 trials have long-term extension studies(Table 4). It is important to understand that there is an intention to treat analysis unless the results of the patients completing the study are not discussed.

In the Phase IIb study of 220 patients treated with abatacept during the double-blind period, 152 entered the long-term extension phase, in which they received abatacept 10mg/kg every 28 days.41,42 Of these, 92 and 85 completed the 5th and 7th years of follow up with retention rates of 60% and 52%, respectively. Of the 73 patients who withdrew, 19 did so due to lack of effect, although this being a rare cause of withdrawal from the 2nd year of the extension study onward, with only 9 patients leaving the study during the last 6 years. The response obtained in the double-blind phase were low activity and improved physical function, as measured by mHAQ, maintained up to 7 years in patients who completed the study.39,41,42

Of the 385 patients initially treated with abatacept who completed the double-blind phase on AIM, 378 entered the extension phase, all receiving 10 abatacept mg/kg every 28 days and ending 266 after 5 years of follow-up, representing a retention rate of 70%.43 20% of the 112 patients who withdrew did so for lack of efficacy. The responses obtained in the remaining patients were maintained until the 5th year. A high proportion of patients achieved remission or low disease activity at the end of double-blind period, maintained throughout the extension period.44 At the end of double-blind study, 116 patients (31%) had a normal HAQ score. Of these, 74% remained normal throughout the length of the study.44,45

Both the ATTEST38 as well as the AGREE39 retention rate after one year was over 90%. The ACR20, 50 and 70 and the HAQ confirmed that efficacy was maintained over 2 years in both studies. After the double-blind, 6-month ATTAIN trial, 317 patients entered the extension phase,46 of which 79 (25%) and 73 (23%) reached the 5th year. During the extension period, 165 patients left the study, 69 of them (22%) due to lack of efficacy.45,46

Finally, in terms of quality of life, both extension studies of the AIM and ATTAIN trials have shown sustained improvement in various aspects of sleep quality using the MOS-Sleep and Sleep Problem Index, and in physical and mental subscales of the SF36, a fatigue scale, the number of days with limited activity and “participation in activities” (APAQ).47–49

The data on effectiveness of abatacept in clinical practice derived from registries have appeared mainly in the last 3 years, with most coming from conference proceedings (see supplementary material). In general we can say that the data correspond to what is already known through clinical trials, although such comparisons must always take into account the biases that involve treatment in clinical practice, where patients are very different from those usually in clinical trials.

A retrospective observational study analyzed the results of the administration of abatacept in 100 patients of whom 97% had previously received a TNF inhibitor, showing that at 6 months it was 80% in those treated with abatacept and 44% and 34% with low activity and those that achieved remission, respectively.50 The French ORA51,52 registry reported data on effectiveness in 133 patients treated with abatacept, of which only 9.5% had not received biological drugs previously, showing that at 6 months, 66% achieved a EULAR response (good or moderate). The DANBIO53 Danish registry presented data on 150 patients treated with abatacept. As in the previous registry, only a small proportion, 5%, received prior TNF inhibitors. The DAS28CRP at baseline and at 24 and 48 weeks was 5.3, 3.4 and 3.3, respectively. The American CORRONA54 registry analyzed patients treated with abatacept and TNF blockers, which had not previously received other biological drugs. After adjusting, the analysis at 2 years of treatment showed that the probability of discontinuation due to a loss of effectiveness was similar in patients receiving abatacept or anti-TNF (RR 0.88, 95% CI 0.51–1.52). There were no differences in clinical response assessed by the patient in what concerned the overall situation and pain.54 Monitoring of patients treated with abatacept in this registry has also shown a gradual improvement in fatigue and sleep disorders and a tendency to reduce the average working day loss.55 A Canadian registry also showed that a significant percentage of patients treated with abatacept achieved a clinically significant improvement in functional capacity assessed by the HAQ (≥0.3), which occurred after about 2 months of treatment and was not influenced by the number TNF inhibitors the patients had previously received.56 Finally, recent data from the French registry of patients treated with abatacept showed a superior response in patients who are RF/anti-CCP positive.57 However, in a previous study with a different analysis, Buch et al.58 found no differences between the two populations.

There are no randomized studies that allow a direct comparison of abatacept with other biologicals. Only the ATTEST31 study, in which abatacept and infliximab were compared individually with placebo, is able to extract some comparative data.

In the systematic review we performed (available in supplementary material) and analyzed, 7 studies included comparative efficacy or safety of abatacept with other biologics in RA. Among these, the ATTEST study showed similar efficacy between abatacept and infliximab (3mg/kg), with a favorable safety profile for abatacept.31 Systematic reviews and metaanalyses were also included2,3,59,60 as well as a literature review.4 From these reviews it can be concluded that abatacept is not very different from other biological regarding effectiveness. In terms of safety, the meta-analysis of abatacept 20114 by Singh, which also include infliximab, adalimumab, etanercept, anakinra, rituximab and tocilizumab, made indirect comparisons in cases where a statistical model that allowed for these. The result is that both abatacept and anakinra were significantly associated with fewer serious adverse effects compared with the others and abatacept was also associated with significantly fewer serious infections compared to infliximab and tocilizumab.

As for cost-effectiveness, Lopez-Olivo et al.61 show in a paper presented at the ACR conference of 2010 that 80% of studies funded by the pharmaceutical industry come to the conclusion that the intervention is cost effective, as opposed to 54% of studies funded by other agencies.

The recommendations of the SER for risk management in biologic therapy62 should be followed in all patients in whom treatment with abatacept is being considered and should include a clinical history and physical examination, in which data on relevant comorbidities, repeat infections and active infection (including TB) should be followed. A routine virus serology in addition to liver function tests, chest X-rays and Mantoux (+booster) testing must be routinely performed according to known recommendations.62 Before treatment onset, vaccination of patients against influenza and pneumococcus should be considered and other vaccines evaluated according to the situation of each patient. Avoid vaccination with attenuated or live germs. In patients of reproductive age pregnancy while abatacept is administered should be discouraged. During treatment with abatacept should be monitored for infections, cancer, or worsening of respiratory function in patients with chronic obstructive pulmonary disease (COPD) (for a detailed review on COPD, see below, following recommendations accordingly). In case of major elective surgery, abatacept treatment should be discontinued during the perioperatory period.62 Comorbidities are common in patients with RA and may be conditioned by treatment with biologics, such as abatacept.

In controlled clinical trials of abatacept with placebo (AIM,30 ATTAIN,32 ASSURE63 and ATTEST31), adverse reactions with abatacept were reported at rates similar to placebo-treated patients, between 55% and 80%. The adverse reactions most commonly reported (≥5%) were headache and nausea.5 The Cochrane meta-analysis of Maxwell and Singh shows that total adverse events were significantly higher in the abatacept group compared with the placebo group, but the relative risk was low (RR 1.05, 95% CI: 1.01–1.08).6

The rate in clinical trials of with drug administration associated reactions is about 7.7 (3.8–13.8) per 100 patient years, which is lower than that of other intravenous biological drugs.5 The acute events related to infusion occur with a frequency of 0.1%–1% of patients included cardiopulmonary symptoms such as hypotension, increased or decreased blood pressure, dyspnea, nausea, flushing, urticaria, cough, hypersensitivity, pruritus, rash, and wheezing. Most of these reactions were mild to moderate.5 In a clinical trial with abatacept 2 patients discontinued the drug due to serious adverse reactions that occurred in the first hour of infusion.30 Overall, the proportion of patients who discontinued treatment was slightly higher in patients treated with abatacept than with placebo (5% vs 2.2%).30 In another trial, in which a significant number of patients were treated concomitantly with MTX, the adverse reactions related to infusion in the first hour were more frequent in the abatacept group, although these were only mild to moderate.32 Only one patient developed anaphylaxis of 2688 patients treated with abatacept during 4764 patient-years.64

In clinical trials of abatacept, infection rates vary between 3.2% and 89.4%, and severe infections between 0.01% and 7.8% (or 1.8–5 episodes per 100 patient-years).4 These figures reflect the heterogeneity of the populations included in studies. In a meta-analysis of clinical trials of biologics, the risk of serious infections with abatacept was numerically lower than the rest, and significantly lower in comparison with infliximab, tocilizumab and certolizumab.4 Overall, it appears that the risk of serious infection is lower with abatacept than with other biologics. That assessment is further supported by work showing that the incidence rates of overall infections and pneumonia requiring hospitalization were similar in patients in clinical trials with abatacept than in cohorts treated with DMARD.65 In the meta-analysis by Singh et al., the odds ratio of serious infections with abatacept vs placebo is 0.57 (95% CI 0.30–1.08).4 As with other biologics, pneumonia is one of the most common serious infections in patients with RA treated with abatacept. They appear in varying rates between 0.36 and 1.43 per 100 patient-years in relation to the patient profile (see review in supplementary document). In any case, they are higher in patients who previously failed treatment with TNF inhibitors, which probably reflects rates in patients with more severity.66–68 The rates of opportunistic infections per 100 patient-years are between 0.01 and 0.36 and vary by prior therapy with abatacept.

In a study conducted in mice infected with M tuberculosis, evoking a model of chronic tuberculosis, abatacept or a murine anti-TNF was administered, with the primary endpoint being survival. While all mice treated with abatacept survived, those receiving anti-TNF died due to an exacerbation of tuberculosis.69

In all clinical trials with a total of 4149 patients and a rate of 12132 patients per year, we found 8 cases of tuberculosis, reflecting an incidence rate of 0.07 [0.03–0.13] per 100 patient-years.69 These cases were: infection with cervical lymphadenitis due to tuberculosis, 3 cases of pulmonary tuberculosis, a case of thoracic Pott's disease, a case of submandibular lymphadenitis, 1 case of latent tuberculosis, a case of suspected tuberculosis with placebo and 1 case of “suspected” tuberculosis. This represents a rate lower than that shown in patients with RA treated with TNF inhibitors. However, the number of patients treated with abatacept to date and the time they have undergone treatment after implementation of guidelines for the prevention of TB reactivation, prevents stronger conclusions.

The safety of abatacept in patients chronically infected with hepatitis viruses B and C has not been studied in depth, since these patients were excluded from clinical trials, and there are no consensus guidelines for use in these situations. There are only a few references to case reports in which abatacept has been used in patients with no active infection by hepatitis B virus, which is usually associated to prophylaxis. There are also isolated cases of RA associated with hepatitis C and treated with abatacept without any significant impairment of liver function.70–72 Given the lack of specific recommendations for treatment with abatacept in patients chronically infected with hepatitis viruses B and C, it is not recommended for use unless no other options are available and always with close monitoring of liver function and viral loads.

Clinical trials with abatacept excluded patients with a history of cancer in the previous 5 years and excluded those with mammograms suggestive of breast cancer. On the other hand it is known that patients with RA have a higher risk of developing lymphoma and lung cancer.73 Simon et al.74 performed a comparative analysis of all patients treated with abatacept during clinical development (4134 patients from 7 clinical trials) vs patients with RA from 5 cohorts (RA 41529, with a median follow up of 1, 8, and 9.3 years).74 In the population that was treated with abatacept, after excluding skin cancers other than melanoma, there was a total of 51 tumors: 7 breast, 2 colorectal, 13 lung and 5 lymphomas, whereas the control population showed 522 cancers, 275 of the breast, 470 colorectal and 184 lung lymphomas. The number of cancers in patients treated with abatacept was within the expected range in concordance with RA cohorts. The incidence rate of lung cancer with abatacept (0.15/100 person-years) is similar to that of the various RA cohorts (0.09–0.26) and for lymphoma occurred at a rate of 0.06/100 vs 0.06–0.11 person-years. Therefore, the data suggest that the risk of these tumors with abatacept is different from that expected in a population with RA.

Clinical trials excluded patients with NYHA class IV heart disease. The remaining patients without heart disease or below class IV had no evidence of worsening or the appearance of “de novo” heart failure.5

In the ASSURE63 study, patients with chronic obstructive pulmonary disease (COPD) treated with abatacept developed more respiratory adverse reactions than those treated with placebo: 43% vs 24%. Similarly, serious adverse events of respiratory origin were more frequent in patients treated with abatacept than in those receiving placebo: 10% vs 0%. Adverse reactions noted were relapses of COPD, bronchitis and pneumonia. Therefore in clinical practice, abatacept should be indicated only after a thorough review of risk/benefit.

Two systematic reviews of the literature aimed at assessing the occurrence of interstitial lung disease (ILD) following administration of different biological drugs found no published cases related to abatacept.75,76 On the other hand, it is unknown what effect abatacept would have in patients with prior interstitial lung disease. Despite all this, and given previous experience with biological therapies, administration of abatacept should be performed with very close monitoring of respiratory symptoms and lung function of patients.

So far a few cases of demyelinating disease, one in RA77 and another in juvenil78 chronic arthritis, have been reported in all clinical trials with abatacept and there is no clear increase over expected rates of demyelinating disease. Furthermore, treatment with abatacept is safe and could be effective in multiple79 sclerosis. However, we must await the results of longer-term pharmacovigilance to clarify this.

In 3985 patients with RA treated with abatacept for up to 8 years, 4.8% developed antibodies against abatacept during treatment.5 In patients who were evaluated after discontinuing abatacept for antibodies (>42 days after the last dose) 5.5% were positive.5 The presence of antibodies was not associated with adverse reactions, including those related to the administration, or with changes in efficacy or serum abatacept concentrations. The rates of ANA and anti-DNA in clinical trials were not different from those of the placebo5 treated group.

In clinical trials with abatacept have been cases of psoriasis 0.57 per 100 (95% CI 0.44–0.72), Sjögren's syndrome rate 0.19 (95% CI 0.12–0.29) and vasculitis reported rate 0.18 (95% CI 0.11–0.28).80 No cases of SLE have been reported.

The information on the efficacy and safety of different vaccines with abatacept therapy is limited. No data exist on their potential risk on vaccination with live attenuated organisms, such as oral polio, MMR, varicella and yellow fever. There are some facts about inactivated germs vaccines or cellular components. In a double-blind trial in healthy volunteers, there was a response to vaccination with tetanus and pneumococcal vaccines which was slightly lower after administration of a dose of abatacept in controls.81 A substudy of ARRIVE33 evaluated the efficacy of vaccination with 7 strains of Streptococcus pneumoniae in 21 patients with RA. The response was positive at least for one strain in 81% of cases, and 48% to 3 strains or more.82 The ARRIVE study also analyzed the response to influenza vaccination (strains H1N1, H3N2 and influenza B) in 20 patients with RA, 50% responded to at least 2 of these 3 strains.83 Moreover, in an observational study in Brazil, the rate of seroconversion in a group of 11 patients with abatacept, compared to that of 33 patients with RA treated with MTX and 33 healthy controls, was significantly inferior.84 Therefore, vaccination with live attenuated organisms is considered contraindicated until 3 months after cessation of treatment with abatacept, or at least 2 weeks before the start or its reintroductio.5 Inactivated vaccines or germ cell components can be administered at any time, although its efficacy may be somewhat diminished. For this reason, if possible, it also recommended that administration occurs 3 months after discontinuation.

Sequential administration of abatacept after failure of a TNF inhibitor is not associated with increased adverse event rates including serious infections, after 6 months of drug administration. This was demonstrated in the ARRIVE trial where prior to administration of abatacept a washout period to TNF antagonists was performed for at least 2 months in 449 patients and was not done in 597; there was no difference in rates of serious infections between groups.33 However, whenever you perform the transition from treatment with a TNF inhibitor to abatacept, patients should be monitored closely for signs of infection.

As is the case with other biologics, combination therapy of abatacept and a TNF inhibitor is associated with a significant increase in infections in general and specifically severe63 infections,85 which are a contraindication for this type of administration. There is no evidence on the safety and efficacy of abatacept in combination with anakinra or rituximab. As for the combination with other drugs, abatacept can be combined safely with MTX to the fact that the population pharmacokinetic analysis did not detect any effect of these on the clearance of abatacept, or found any significant safety issues with the use of abatacept in combination with DMARDs during development.5 No dose adjustment is necessary when used in combination with other DMARDs, glucocorticoids, salicylates, NSAIDs or analgesics.

Abatacept crosses the placental barrier, although preclinical studies showed no adverse effects at doses up to 29 embryos-times the human dose of 10mg/kg.5 Data on women who became pregnant accidentally treated with abatacept are very limited. Of 8 women in the 5 pivotal studies, 7 were receiving concomitant MTX and another leflunomide.5 In 3 cases there were spontaneous abortions in the first quarter (2 of them previously had abortions) and 2 pregnancy were terminated voluntarily. In the AGREE study, 2 accidental pregnancies ended in spontaneous and induced abortion.34 Therefore, pregnancy in women treated with abatacept should be avoided until 3 months after the suspension.

Other interesting aspects related to abatacept.

In a paper presented as a Congress summary there was an increase in mean HDL cholesterol at 24 weeks of starting treatment with abatacept in 17 patients: 1.7–2.1mmol/l (P=.008) and a decrease in the atherogenic index from 3.0 to 2.7 (P=.03), without changes in total cholesterol, triglycerides or LDL86,87 cholesterol. The increase in lipids with no significant increase in the atherogenic index seems a common occurrence to infliximab, etanercept, adalimumab, abatacept, rituximab and tocilizumab.88