Safety of Abatacept Compared With Other Biologic and Conventional Synthetic Disease-Modifying Antirheumatic Drugs in Patients With Rheumatoid Arthritis

Data From an Observational Study

Gulsen Ozen; Sofia Pedro; Rebecca Schumacher; Teresa A. Simon; Kaleb Michaud


Arthritis Res Ther. 2019;21(141) 

In This Article


The analysis included 1496 patients who initiated abatacept with 4896 patient-years of total follow-up and 2502 patient-years of abatacept exposure; 3490 patients who initiated another bDMARD with 11,777 patient-years of total follow-up (9658 patient-years of drug exposure); and 1520 patients who initiated a csDMARD with 4816 patient-years of total follow-up (4184 patient-years of drug exposure). Median follow-up was 2.5 years per patient (interquartile range [IQR] 1.0–4.5 years) for abatacept and bDMARDs and 2.0 years per patient (IQR 1.0–4.0 years) for csDMARDs.

Patient baseline characteristics by treatment group are presented in Table 1. At baseline, patients initiating abatacept versus other bDMARDs or csDMARDs tended to have higher HAQ-DI, pain and patient global scores, and greater proportions of patients had prior use of csDMARDs and concurrent use of a GC. Prior use of a bDMARD was more frequent in abatacept versus other bDMARD initiators (≥ 2 bDMARDs, 50.2% vs 27.8%, respectively), and most patients in both groups were on a concomitant csDMARD (73% and 76%, respectively).


In the fully adjusted MSMs, there were no differences in the risks of malignancies with abatacept relative to other bDMARDs or to csDMARDs (Figure 1). The relative risks of breast cancer, lung cancer and lymphoma could not be determined because none of these events was observed in any treatment group. Overall, abatacept treatment was associated with a lower risk of all hospitalized infections compared with other bDMARDs (adjusted HR [95% CI] 0.37 [0.18, 0.75]), and with a lower risk (wide CIs) compared with csDMARDs (0.31 [0.09, 1.05]). When analysis was limited to line of therapy (second-line, third-line or greater), the risk difference in hospitalized infections between abatacept and other bDMARDs and csDMARDs decreased (Table 2). The number of overall autoimmune events was low, and most models used to analyse the relative risks of these secondary outcomes did not converge. There was a numerical increase in the risk of psoriasis with abatacept compared with other bDMARDs and with csDMARDs. The relative risks of hospitalized infections, malignancies and NMSCs by abatacept treatment line are presented in Table 2.

Figure 1.

Association of treatment with hospitalized infections, malignancies and psoriasis in patients with RA. a Abatacept vs other bDMARDs. b Abatacept vs other csDMARDs. *Using inverse probability of treatment weights with further adjustment for time-varying age, disease duration, HAQ-DI, pain and patient global scores, RDCI, and GC treatment duration. bDMARD = biologic disease-modifying antirheumatic drug; CI = confidence interval; csDMARD = conventional synthetic disease-modifying antirheumatic drug; GC = glucocorticoid; HAQ-DI=Health Assessment Questionnaire-Disability Index; HR = hazard ratio; RA = rheumatoid arthritis; RCDI = Rheumatic Disease Comorbidity Index

For the secondary outcomes (except psoriasis), the results were presented as crude IRs due to the low number of events and inability to achieve convergence by the MSMs.

The IR of overall malignancy was higher in the abatacept group than in the other bDMARD group, and was lower than in the csDMARD group (Table 3). The most frequently observed malignancy was NMSC in all treatment groups.

The IRs of infections were similar across treatment groups, with lower rates of hospitalized infections or pneumonia with abatacept versus other bDMARDs or csDMARDs (Table 3). No cases of tuberculosis were observed in any treatment group. The IRs of autoimmune diseases were mostly similar across treatments; however, a higher incidence of psoriasis was reported in the abatacept versus other treatment groups (Table 3). The highest frequency of hospitalizations was observed in the csDMARD treatment group, and death rates were comparable across treatments (Table 3).

The IRs for any reaction to treatment infusion or injection were lower with abatacept than with other bDMARDs; severe administration reactions were less common with SC than with IV abatacept (Table 4).