TNF-alpha antagonists may not increase risk for serious infections requiring hospitalization, study indicates

TNF-alpha antagonists may not increase risk for infection-related hospitalizations in patients with autoimmune diseases, according to a new study.

Researchers performed a retrospective cohort study using state and national databases to compare TNF-alpha antagonists and nonbiologic regimens in patients with rheumatoid arthritis, inflammatory bowel disease, and psoriasis, psoriatic arthritis or ankylosing spondylitis. Cohorts were matched using disease-specific propensity scores, and Cox regression models with nonbiologics were used as the reference. Overall, the study included 10,484 matched pairs with rheumatoid arthritis, 2,323 matched pairs with inflammatory bowel disease, and 3,215 matched pairs with psoriasis and spondyloarthropathies. The study involved patients treated from 1998 to 2007, 20% of whom were at least 65 years of age. The main outcome measure was serious infection requiring hospitalization in the first 12 months after nonbiologic regimens or TNF-alpha antagonists were started. Baseline use of glucocorticoids was evaluated separately. The study results were published online Nov. 6 by the Journal of the American Medical Association.

A total of 1,772 serious infections were identified, mostly pneumonia and skin and soft tissue infections (53%). Hospitalization rates for serious infection were 8.16 versus 7.78 per 100 person-years for patients with rheumatoid arthritis taking TNF-alpha antagonists versus those taking comparator regimens, respectively (adjusted hazard ratio, 1.05; 95% CI, 0.91 to 1.21). For patients with inflammatory bowel disease, these rates were 10.91 versus 9.60 per 100 person-years (adjusted hazard ratio, 1.10; 95% CI, 0.83 to 1.46), and for patients with psoriasis and spondyloarthropathies, these rates were 5.41 versus 5.37 per 100 person-years (adjusted hazard ratio, 1.05; 95% CI, 0.76 to 1.45). In patients with rheumatoid arthritis, infliximab was associated with an increased risk for serious infections compared with etanercept (adjusted hazard ratio, 1.26; 95% CI, 1.07 to 1.47) and adalimumab (adjusted hazard ratio, 1.23; 95% CI, 1.02 to 1.48). The case fatality ratio during hospitalization for serious infection was 3.6% in the rheumatoid arthritis group, 2.1% in the inflammatory bowel disease group and 7.1% for the psoriasis and spondyloarthropathies group. A dose-dependent increase in infections was associated with glucocorticoid use at baseline.

The authors acknowledged that their study used medication data from pharmacy files, which cannot confirm actual medication use, and that some study outcomes may have been misclassified because they were based on claims data, among other limitations. However, they concluded that TNF-alpha antagonists did not appear to increase risk of hospitalization for serious infections compared with nonbiologic regimens in patients with autoimmune diseases. Their results also indicate that infliximab may increase infection risk in patients with rheumatoid arthritis and that glucocorticoid use may have a strong dose-dependent relationship with infection risk, they wrote.

The authors of an accompanying editorial pointed out that residual confounding may have affected the study results, that propensity scores were generated by using data from the year before patients entered the study, and that more patients were censored in the comparison cohort than in the TNF-alpha cohort. They wrote that other trials will need to replicate the current findings but concluded that “the report. . .raises important questions about the comparative safety of immunosuppressant and biologic therapy and may prompt a reevaluation of anti-TNF safety.”