Patients with rheumatoid arthritis (RA) may not respond to treatment with disease-modifying anti-rheumatic drugs (DMARDs) alone 1 2 3 4 . In patients who have failed DMARD therapy for RA, clinical studies have demonstrated the effectiveness of drugs directed against tumour necrosis factor (TNF) as monotherapy or when used in combination with DMARDs 2 5 6 7 8 9 10 . Adalimumab (ADA) is a fully human anti-TNF monoclonal antibody for the treatment of moderate to severe RA. Initial clinical trials of ADA in patients with RA demonstrated a good safety profile, with improvements in disease signs and symptoms and functional ability, achievement of clinical remission and inhibition of radiographic disease progression 2 3 7 .

The Research in Active Rheumatoid Arthritis (ReAct) phase 3b study was initiated in 2002 to assess the safety and effectiveness of ADA in RA patients who had failed treatment with at least one traditional DMARD 11 12 . ADA was well tolerated and effective, alone or with DMARDs, in 6,610 patients with active RA over a mean treatment duration of 233 days 11 12 . To evaluate the long-term safety and effectiveness of ADA in clinical practice settings over 5 years in patients who completed ReAct, the REgistry of HUMIRA™ in RA: a Long-Term Investigation of Safety and Efficacy (ReAlise) observational follow-up study was conducted (NCT00234884). The primary objectives of this analysis include examination of adverse events (AEs) and the temporal pattern of their occurrence and maintenance of response through 5 years of ADA treatment (i.e., from the first injection received in ReAct through the last observation in ReAlise).

The combined data from ReAct and ReAlise include more than 6,600 RA patients and provide one of the largest well-monitored evaluations of the long-term safety and effectiveness of a single TNF antagonist in routine clinical practice settings. Other reports involving thousands of RA patients come from country registries that record and compare use of multiple TNF antagonists 18 19 20 21 22 . The observational study design allows for inclusion of patients receiving routine care for RA and is not restricted to patients who meet the stricter entry criteria of randomised controlled trials, including fewer comorbidities. Patients enrolled in ReAct and ReAlise reflect typical RA patients treated with TNF antagonists, who have long-standing RA with moderate to severe baseline disease activity despite treatment with DMARDs. Some patients also had previously received infliximab and/or etanercept.

ADA was well tolerated, and no new or unexpected safety concerns were identified during more than 5 years of therapy, representing nearly 20,000 PYs. The pattern and rates of AEs were comparable with those observed in double-blind studies of ADA with or without methotrexate 2 3 7 . SAEs, including serious infections and TB, were reported more frequently during the first year, after which the incidences decreased and remained stable. This finding is expected because of positive selection bias of the patients who tolerate the drug and less frequent visits in ReAlise, and generally aligns with data from ADA clinical trials and registries 4 7 18 23 24 25 . A recent review of safety data from ADA clinical studies across indications found a stable rate of serious infections throughout ADA exposure, which may be attributed to differences in populations and methodology 26 . Beyond the positive selection bias of the patients who tolerate the drug, the other possible explanations for the decrease in infection risk with ongoing TNF antagonist therapy may include decreased use of corticosteroids and better control of the RA disease process that reduces disease-associated alterations in natural immunity 18 19 . The overall incidence rates of SAEs and serious infections with ADA (13.8 and 2.8 E/100 PYs) were found to be within the range of those previously reported in long-term clinical trials of other TNF antagonists and biological DMARDs. Rates of serious infections for infliximab and etanercept after a median follow-up of 3.9 years in the British registry were 4.6 and 3.8 E/100 PYs, respectively 18 . Five-year data on abatacept and tocilizumab report serious infection rates of 3.0 and 5.7 E/100 PYs, respectively 27 28 . SAEs of interest other than serious infections occurred at a low frequency and decreased over time with ADA. Similarly, the risk of TB decreased over time, with an incidence rate <0.1% in the patients treated more than 5 years. Still, five cases occurred in patients treated for more than 3 years, underscoring the need for continued attention to the risk of infection. There was no overall increased risk of malignancies; however, use of a US database (NCI) as a comparator for SIRs in this largely European population may limit interpretation of results. The observed risk of lymphomas was consistent with that for RA patients. The risk of lymphoma, both Hodgkin’s and non-Hodgkin’s, is increased by approximately two-fold in patients with RA, and this increase has been associated with the chronic inflammation of RA 23 29 30 . Use of disease-modifying treatment that controls chronic inflammation may reduce the elevated risk associated with uncontrolled RA. Recent long-term studies have failed to identify an effect of TNF antagonists on the incidence of cancer, including lymphoma (with exception of NMSC), and also determined that the relative risk of cancer did not increase with time since first starting a TNF antagonist or with cumulative duration of TNF antagonist therapy 31 32 33 . Lastly, mortality rates were lower in patients receiving ADA therapy in ReAct and ReAlise than observed in a matched population without RA, and there was no increase in the overall mortality rate over the course of the studies. The decreased rates of cancer and mortality must be interpreted with caution because patients with previous cancers may have been excluded from receiving a TNF antagonist.

Patients were allowed to continue concomitant DMARD therapy and/or corticosteroids with ADA in ReAct and ReAlise. Generally, the incidences of SAEs were similar among subgroups based on concomitant therapy, with a slightly higher rate among patients who received ADA monotherapy (no DMARDs) and those who received corticosteroids (17.4 and 14.9 E/100 PYs, respectively). Patients who had received etanercept and/or infliximab before enrolling in ReAct showed a higher rate of SAEs and infections than patients without prior history of TNF antagonist therapy, but no differences in the rates of malignancies or lymphoma were observed. Statistical analyses of these subgroups were not performed in this observational study; however, no statistical differences have been identified in controlled clinical studies that evaluated AEs in subgroups based on prior or concomitant treatment 2 3 12 .

The effectiveness of ADA in reducing disease activity and inflammation was maintained over more than 5 years of treatment, as evidenced by the percentages of patients achieving ACR20, 50 and 70 responses and at least moderate EULAR responses. The initial 12-week improvements observed in ReAct included ACR20 in 69% and moderate and good EULAR responses in 83% and 33%, respectively 11 . For patients who continued in ReAlise, these response rates were 66%, 80% and 42%, respectively, at last observation. LDA and REM rates also were maintained at last observation, with half of the patients achieving LDA, and one in five achieving REM. Clinically relevant functional improvement was reported for approximately two-thirds of patients, and one-third of patients reported normal function at last observation.

Approximately 70% of patients who received prior TNF antagonists discontinued that therapy because of loss or lack of response before enrolling in ReAct/ReAlise. At week 12 of ReAct, substantial clinical benefit was achieved with ADA in patients previously treated with infliximab and/or etanercept 12 , and response rates were maintained through 3 years of ADA 34 . In agreement with our data, clinical studies that have evaluated the effectiveness of switching TNF antagonists have generally found that treatment with a second TNF antagonist can offer comparable or slightly lower response rates than those observed in TNF antagonist-naive patients 20 35 36 . However, as the number of TNF antagonists increase to three or more, response rates decrease. In a study of more than 2,000 RA patients, discontinuation rates were similar for infliximab, etanercept and ADA, and factors predictive of discontinuation were increased disease severity and comorbidities 36 . Delaying administration of TNF antagonists in patients who fail to achieve an adequate response with methotrexate has been associated with poorer clinical, functional and radiographic outcomes 4 37 . The findings from ReAct/ReAlise indicate that switching from failed therapy with a traditional DMARD or TNF antagonist to ADA (with or without concomitant DMARD) is a reasonable strategy.

Limitations of long-term observational studies include inherent bias. Patients who experience effective or tolerable therapy are likely to continue treatment. In ReAct/ReAlise, the most common reasons for patients discontinuing the study were loss of efficacy and AEs. To address potential bias, assessment of effectiveness over 5 years included results for all efficacy parameters at last observation. Results at last observation were consistent with results at the end of 12-week treatment in ReAct 11 .

In a clinical practice setting, ADA was well tolerated, and no new safety concerns were identified during nearly 20,000 PYs of exposure. The incidence of SAEs and serious infections decreased with ongoing ADA treatment over 5 years. The incidences of deaths and malignancies were lower than expected for the general population, and the incidence of lymphoma was within expected rates for patients with RA. Clinical response to ADA, as measured by reduced disease activity and functional improvements, was maintained through more than 5 years of observation in patients with long-standing, severe RA. Safety and effectiveness observed in more than 6,600 RA patients over 5 years were consistent with results of randomised clinical trials of ADA 2 3 7 .

^aFrom a total of 129 cases (114 cases of malignancy (excluding lymphomas, and NMSC) plus 15 cases of lymphoma), 11 were excluded (9 cases of carcinoma in situ and 2 of metastases for which the primary cancer was already counted), in keeping with the SEER database criteria for comparison.

ACR: American College of Rheumatology; ADA: adalimumab; AE: adverse event; CI: confidence interval; DAS28: 28-joint Disease Activity Score based on ESR; DMARD: disease-modifying anti-rheumatic drug; E: event; EMEA: European Medicines Agency; ESR: erythrocyte sedimentation rate; EULAR: European League Against Rheumatism; HAQ DI: Health Assessment Questionnaire Disability Index; ITT: intention-to-treat; LDA: low disease activity; NCI: National Cancer Institute; NMSC: non-melanoma skin cancer; NSAID: non-steroidal anti-inflammatory drug; PtGA: patient global assessment; PYs: patient-years; RA: rheumatoid arthritis; REM: remission; SAE: serious adverse event; SD: standard deviation; SDAI: Simplified Disease Activity Index; SEER: Surveillance, Epidemiology and End Results; SIR: standardised incidence ratio; SJC: swollen joint count; SMR: standardised mortality ratio; TB: tuberculosis; TJC: tender joint count; TNF: tumour necrosis factor.

The ReAct and ReAlise studies were funded by AbbVie Inc, which participated in designing and conducting the study; collection, analyses, and interpretation of data; and writing, reviewing, and approval of the publication. GRB received consulting fees and payment for speakers bureaus and board membership from AbbVie, Bristol-Myers Squibb (BMS), MSD, Pfizer, Roche, and UCB. GRB’s institution has received research grants from AbbVie, BMS, MSD, Pfizer, Roche, and UCB. MM-C declares that he has no competing interests. XM received research grants from Pfizer and Roche and consulting fees from BMS, GSK, LFB, Pfizer, Roche and UCB. FN-B received research grants from Roche. SK is a former contract employee for AbbVie. KU is an employee of AbbVie and may hold AbbVie stock. HK is an employee of AbbVie, may hold AbbVie stock, and is named an inventor on the following patent applications assigned to AbbVie Biotechnology Ltd: WO2011/097301 published 8/11/2011, WO2007/120656 published 10/25/2007, WO2007/120626 published 10/25/2007, US20120171123A published 1/5/2012.

GRB made substantial contributions to conception and design, acquisition of data, analysis and interpretation of data; was involved in drafting the manuscript and revising the manuscript critically for important intellectual content; and has given final approval of the version to be published. MM-C made substantial contributions to acquisition of data and analysis and interpretation of data; was involved in revising the manuscript critically for important intellectual content; and has given final approval of the version to be published. XM made substantial contributions to acquisition of data and analysis and interpretation of data; was involved in revising the manuscript critically for important intellectual content; and has given final approval of the version to be published. FN-B made substantial contributions to acquisition of data; was involved in drafting the manuscript and revising the manuscript critically for important intellectual content; and has given final approval of the version to be published. SK made substantial contributions to conception and design, analysis and interpretation of data; was involved in drafting the manuscript and revising the manuscript critically for important intellectual content; and has given final approval of the version to be published. KU made substantial contributions to conception and design, acquisition of data, analysis and interpretation of data; was involved in drafting the manuscript and revising the manuscript critically for important intellectual content; and has given final approval of the version to be published. HK made substantial contributions to conception and design, and analysis and interpretation of data; was involved in drafting the manuscript and revising the manuscript critically for important intellectual content; and has given final approval of the version to be published. All authors read and approved the final manuscript.