Long-term anti-TNF therapy and risk of serious infections in a cohort of patients with rheumatoid arthritis: comparison among adalimumab, etanercept and infliximab

Background: TNF blockade in rheumatoid arthritis (RA) patients leads to an increased rate of infections.
Objectives: Aim of this study was to evaluate the risk of serious infections (SI) in patients with RA treated with anti-TNF therapy using data from the GISEA (Gruppo Italiano Studio Early Arthritis) Register.
Methods: Clinical records of 2769 adult patients with long-standing RA (mean age 53.2±13.4 yrs; mean disease duration: 9.0±8.3 yrs) included in the GISEA register were analyzed. The study included all patients who had been treated for at least six months with TNF-α inhibitors or had discontinued therapy earlier due to SI. Of the total number of patients included, 837 (30%) were treated with infliximab (IFN), 802 (29%) with adalimumab (ADA), and 1.130 (41%) with etanercept (ETN).
Results: 176 patients had at least one or more SI (total number 226). Overall incidence rates were 31.8/1000 patient-years of follow-up (95% CI 25.2-38.3), respectively 23.7/1000 patient-years of follow-up (95% CI 13.1-34.2) with ADA; 12.8/1000 patient-years of follow-up (95% CI 6.3-19.4) with ETN and 65.1/1000 patient-years of follow-up (95% CI 48.4-81.8) with IFN. The risk was higher in the first 12 months of treatment in comparison with the other 12-month interval, even though a statistically significant difference was not achieved (p=0.08) (38,9% of the total SI were registered in the first 12 months of treatment).
The most common sites of SI were upper and lower respiratory tract infections (nearly 50%), urinary tract infections (12%), skin infections (9,7%). Ten patients developed active tuberculosis. The risk of SI differs significantly among the three agents ADA, ETN and IFN (p<0.0001).
Univariate analysis identified taht the use of steroids (p<0.001), age at the start of anti-TNF treatment (p=0.004), presence of comorbidities (p<0.0001) and concomitant DMARD treatment during anti-TNF therapy (p=0.0204) are associated with a risk of SI. Multivariate analysis confirmed as significant predictors of infections concomitant use of steroids (p<0.001) and age at the start of anti-TNF treatment (p=0.003), with an accuracy of 71% of the model.
Conclusions: These data suggest that anti-TNF therapy is associated with a small but significant risk of SI. The risk is associated with concomitant use of steroid and age at the start of anti-TNF treatment. The risk of SI differs in relation to the type of anti-TNF agent.