OBJECTIVE

Methotrexate (MTX) is an important drug for treatment of rheumatoid arthritis; however, there is variation in the clinical response. MTX inhibits T cell cytokine production, with significant interindividual variability in the dose required. We investigated if the variability in clinical response was related to variability in the in vitro assay.

METHODS

Patients with disease modifying antirheumatic drug-naive, active RA [1982 American College of Rheumatology (ACR) criteria] seen from September 2005 through January 2006 were enrolled. MTX was started at 10 mg/week and increased monthly by 2.5 mg/week. Baseline whole-blood cultures were set up with anti-CD3, anti-CD28, and increasing doses of MTX. Supernatants were harvested at 96 hours and tumor necrosis factor-alpha (TNF-alpha), interferon-gamma (IFN-gamma), and interleukin 10 (IL-10) concentrations were estimated by ELISA. The dose of MTX (ID50) required for 50% suppression of production of cytokines and the change in Disease Activity Score-28 (DeltaDAS) at 4 months were noted.

RESULTS

T cell stimulation resulted in significant increase in cytokine release, and addition of MTX led to a dose-dependent suppression of all 3 cytokines. There was significant negative correlation of DeltaDAS with ID50 values for TNF-alpha (R = -0.62, p < 0.01) and IFN-gamma (R = -0.43, p = 0.04). At 4 months, EULAR moderate and ACR 20% responses were achieved by 13 and 16 patients, respectively. EULAR moderate response could be predicted using ROC curves for TNF-alpha (sensitivity 93%, specificity 86%) and IFN-gamma (60% specificity, 71% sensitivity). ACR response was correctly predicted in 14 of 16 ACR 20% responders and in all ACR 50% and ACR 70% responders.

CONCLUSION

An in vitro TNF-alpha suppression assay may help predict clinical response to MTX in RA.