Circulating levels of tumor necrosis factor receptors are highly predictive of mortality in patients with rheumatoid arthritis.

Objective: To investigate whether circulating levels of soluble tumor necrosis factor receptors (sTNFR) are predictive of mortality in rheumatoid arthritis (RA).

Methods: Levels of sTNFRI and sTNFRII at study entry were quantified using enzyme-linked immunosorbent assays in sera from 401 white patients with RA followed up for 13 years. Patients were tracked via the National Health Service Central Register, and the relationship between sTNFR levels and mortality was analyzed using a Cox proportional hazards regression model. Hazard ratios (HRs) and 95% confidence intervals (95% CIs) were calculated.

Results: At the end of the followup period, 132 (32.9%) of 401 patients had died. Of these, 64 (48.5%) died of cardiovascular disease (CVD). Significant associations between all-cause mortality and baseline levels of sTNFRI and sTNFRII were identified in men (HR 1.7 [95% CI 1.2-2.4] and HR 1.18 [95% CI 1.05-1.32], respectively) and women (HR 1.33 [95% CI 0.99-1.8] and HR 1.14 [95% CI 1.02-1.28], respectively). Analysis including levels of both sTNFRI and sTNFRII indicated that the sTNFRII level was the best overall predictor of mortality. Multivariate analysis also revealed that the sTNFRII level was a predictor of all-cause and CVD mortality independently of age, sex, disease duration, C-reactive protein level, erythrocyte sedimentation rate, rheumatoid factor, nodular disease, modified Health Assessment Questionnaire score, taking CVD drugs, and smoking.

Conclusion: Our data indicate that serum levels of sTNFR are powerful predictors of mortality in RA. Elevated levels are particularly associated with mortality due to CVD and may be useful for identifying patients at increased risk of premature death.