Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement
Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.
An Immune Signature Based on the Ex Vivo Responsiveness of Peripheral Blood Cells for Monitoring Disease Activity in Rheumatoid Arthritis.
M. Davis III, John, L. Knutson, Keith, A. Strausbauch, Michael, S. Crowson, Cynthia, M. Therneau, Terry, L. Matteson, Eric, E. Gabriel, Sherine
New techniques with enhanced capacity to detect systemic immune dysregulation are needed to enable tighter control of rheumatoid arthritis (RA). The objective was to identify an "immune signature" for predicting the level of disease activity in patients with RA by broadly assessing ex vivo cytokine release by PBMC in response to stimulation.
We conducted a correlative study of patients with early RA in an academic rheumatology practice. Data were collected for patient reported outcomes (PROs), including pain, morning stiffness, disability, and quality of life, using validated instruments. To establish immune signatures, we measured the release of 17 cytokines by PBMC in response to a panel of stimuli or in media alone using multiplexed immunoassays. The stimulant panel included anti-CD3/anti-CD28, CpG oligonucleotides (CpG), combined cytomegalovirus and Epstein Barr virus lysates (CMV/EBV), heat shock protein 60, phorbol myristate acetate with ionomycin (PMA/ionomycin), phytohemagglutinin, and combined Staphylococcal enterotoxins A and B. Mixed effects models were used to normalize the log-transformed cytokine data and adjust for assay effects. Factor analysis was used to combine multiple PROs into a single composite outcome. Gradient boosting models (GBM) were used to predict the PRO composite outcome using all of the 136 stimulated cytokine concentrations and thereby derive a multi-cytokine prediction score. Linear regression models were used to determine the associations of the score with validated disease activity measures, including the Disease Activity Score in 28 joints (DAS28), adjusting for clinical covariates.
The study included 98 patients (63% female; mean age 56 yrs; mean RA duration 2.9 mo) with a total of 136 visits. In the factor analysis, pain levels influenced the PRO composite outcome most heavily. Among patients with high PRO composite scores, we observed that the release of T cell-derived cytokines, including IFN-g, MIP-1b, IL-4, IL-5, IL-10, and IL-17, was significantly decreased. In contrast, we observed that the release of several innate cell-derived cytokines, including IL-1b, IL-6, and TNF-a, was significantly increased in these patients. In the GBM analysis, the most influential cytokines were MIP-1b, IL-7, IL-5, IFN-g, and IL-13, and the most influential stimuli were CMV/EBV, CpG, PMA/ionomycin, and anti-CD3/anti-CD28. The multi-cytokine prediction score was confirmed to be associated with the PROs used to train the score and also with established disease activity measures, demonstrating evidence of criterion validity. The score (0 100) was strongly associated with the DAS28 (b coefficient: 0.25 per 10 units; standard error: 0.063; p=0.0002), independent of clinical covariates, including the use of methotrexate, biologics, and prednisone.
A PBMC-based immune signature, reflecting aberrant responsiveness of the innate and adaptive immune systems, is able to quantify the clinical activity of RA. If validated and standardized in future studies, this approach could enable more sensitive monitoring of RA activity, facilitating strategies of tight control and improved patient outcomes.
To cite this abstract, please use the following information:
M. Davis III, John, L. Knutson, Keith, A. Strausbauch, Michael, S. Crowson, Cynthia, M. Therneau, Terry, L. Matteson, Eric, et al; An Immune Signature Based on the Ex Vivo Responsiveness of Peripheral Blood Cells for Monitoring Disease Activity in Rheumatoid Arthritis. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :1581