VICTORIA, B.C. – Patients with rheumatoid arthritis who also have common comorbidities are at increased risk for infection when starting treatment with at least one class of biologic agents, based on the findings of a cohort study.
Investigators led by Dr. Joanne Homik, director of the division of rheumatology at the University of Alberta in Edmonton, studied 1,086 patients with rheumatoid arthritis (RA) who were initiating treatment, in most cases with a biologic agent targeting tumor necrosis factor (TNF).
Dr. Joanne Homik
In adjusted analyses, patients had more than triple the risk of a serious infection, defined as one requiring hospitalization, if they had anemia, according to data from the as-yet unpublished study. In other findings, patients were 97% more likely to develop an infection if they had lung disease and 39% more likely if they had heart disease.
"The anemia variable has an unclear relationship, but we are looking into that further. Our theory is that with lung disease, that has some biologic plausibility since bronchitis and pneumonia have prominence in our list of infections seen in this cohort," Dr. Homik commented at the annual meeting of the Canadian Rheumatology Association.
The incidence density of serious infection was 1.67 cases per 100 patient-years, about half that found in some other cohorts. "One of the things we thought of when trying to explain why our serious infection risk is so low is that health care restructuring in Alberta in the mid-90s resulted in bed closures," she explained. "So hospitalizations were reduced overall, and many patients who ... could be considered seriously ill are not admitted any longer, and it can be a bit more challenging to [identify] those serious infections."
The study occurred shortly after anti-TNF agents first became available, so the patients receiving them usually had the most severe, long-standing RA, Dr. Homik noted. Over time, as the demographic of the database changes to include more patients within the first 5 years of their RA, outcomes may differ and predictors may differ as well.
The risk of any infection among patients starting anti-TNF agents did not differ from that among patients starting the disease-modifying antirheumatic drug (DMARD) leflunomide instead. However, leflunomide confers a higher infection risk than some other DMARDs, she noted.
The investigators used registry data to study a cohort of all 943 patients with RA in Alberta who started a first anti-TNF agent between January 2004 and March 2009, and also studied a comparison cohort of 143 patients who started leflunomide (Arava). The patients completed the self-administered comorbidity questionnaire at baseline and were monitored prospectively for outcomes and adverse events. Their average age was 54 years, and 70% were female.
With a mean follow-up of 2.3 years, 70% of patients developed an infection as ascertained from physician claims data or self-report; the most common were bronchitis, cellulitis, and sinusitis. Nearly 4% of patients developed a serious infection; the most common were pneumonia, cellulitis, septicemia, and septic arthritis.
In the first multivariate analysis, patients had an elevated risk of infection if they had heart disease (hazard ratio, 1.39) or lung disease (1.97), used steroids (1.2), or currently smoked (1.24). Their risk was decreased risk if they were male (HR, 0.78). When analysis was restricted to the subset who started an anti-TNF agent, infection risk was increased for those with lung disease (2.02) and decreased for men (0.80).
In the second multivariate analysis, patients had an elevated risk of serious infection if they had anemia (HR, 3.26) or used steroids (3.21), and a decreased risk if they had more than a secondary education (0.33). In the subset who started an anti-TNF agent, risk was increased for patients who had anemia (HR, 3.7) or used steroids (3.5), and decreased for those having more than a secondary education (0.3).
The risks of infection and serious infection did not differ across subgroups when patients were stratified according to whether they stayed on their first treatment or switched from one kind of DMARD to their first anti-TNF agent (or from one anti-TNF agent to another), Dr. Homik reported.
"In the cohort of patients on their first anti-TNF agent, the specific anti-TNF agent used did not predict risk of infection," she noted.
Dr. Homik disclosed that she serves as a consultant to Abbott, Amgen, Pfizer, and Roche. The Alberta Biologics Registry has received unrestricted grant support from Abbott, Amgen, Bristol-Myers Squibb, Merck/Jansson, and Pfizer.