Serum soluble CD30 in early arthritis: a sign of inflammation but not a predictor of outcome.

ABSTRACT

OBJECTIVE: To evaluate serum soluble CD30 levels (sCD30) in an early arthritis series and assess their ability to predict the outcome in patients with rheumatoid arthritis (RA) and undifferentiated arthritis (UA) at one year follow-up. METHODS: Serum sCD30 levels were measured by ELISA from 92 adult patients with RA and UA at baseline and from 60 adult controls. The patients were followed up for one year in the Kuopio 2000 Arthritis Survey. Receiver operating characteristic (ROC) curves were constructed to determine cut off points of sCD30 in RA and UA that select the inflammatory disease from controls. Sensitivity, specificity and positive likelihood ratio, and their 95 % CIs were calculated for sCD30 levels in RA and UA. RESULTS: Median serum sCD30 levels were higher in RA 25.1 (IQ range 16.3-38.6) IU/ml (p<0.001) and in UA 23.4 (15.4-35.6) IU/ml (p<0.001) than in controls 15.1 (10.7-20.8) IU/ml. No differences were recorded between RA and UA (p=0.840). Serum sCD30 levels at baseline did not predict remission at one year follow-up. CONCLUSION: Serum sCD30 levels were higher in RA and UA than in controls at baseline but they did not predict remission at one year follow-up in this series.