ARTHRITIS & RHEUMATISM Vol. 65, No. 2, February 2013, pp 325–333 DOI 10.1002/art.37761 © 2013, American College of Rheumatology Association of Brain Functional Magnetic Resonance Activity With Response to Tumor Necrosis Factor Inhibition in Rheumatoid Arthritis Juergen Rech, Andreas Hess, Stephanie Finzel, Silke Kreitz, Marina Sergeeva, Matthias Englbrecht, Arnd Doerfler, Marc Saake, and Georg Schett Objective. To test whether brain activity predicts the response to tumor necrosis factor inhibitors (TNFi) in patients with rheumatoid arthritis (RA). Since clini- cal and laboratory parameters have proven unsuc- cessful in predicting response, we followed a radically different concept, hypothesizing that response to TNFi depends on central nervous system activity rather than the clinical signs of disease. Methods. Sequential testing by functional mag- netic resonance imaging (MRI) of the brain, anatomic MRI of the hand, and clinical assessment of arthritis were carried out in 10 patients with active RA before and 3, 7, and 28 days after the start of TNFi treatment. Results. Baseline demographic and disease- specific parameters were identical in TNFi responders and nonresponders. The mean  SEM decrease in the Disease Activity Score in 28 joints after 28 days was 1.8  0.3 in TNFi responders (n  5) and 0.2  0.1 in nonresponders (n  5). Responders showed signifi- cantly higher baseline activation in thalamic, limbic, and associative areas of the brain than nonresponders. Moreover, brain activity decreased within 3 days after TNFi exposure in the responders, preceding clinical responses (day 7) and responses observed on the ana- tomic hand MRI (day 28). Conclusion. These data suggest that response to TNFi depends on brain activity in RA patients, reflect- ing the subjective perception of disease. Rheumatoid arthritis (RA) is an autoimmune inflammatory disease affecting 1% of the population worldwide. The disease, which predominantly affects women, is characterized by loss of immune tolerance against self proteins followed by persistent inflammation of multiple joints. RA rapidly leads to the destruction of the cartilage and bone of the affected joints, which is associated with functional decline and premature death (1,2). Treatment of RA was extremely challenging be- fore the introduction of agents that neutralize the bio- logic activity of tumor necrosis factor (TNF), because standard antiinflammatory drugs have either shown lim- ited efficacy in reducing inflammation or proven to be rather toxic when used long term. The development of TNF inhibitors (TNFi) has substantially changed this unmet therapeutic need and dramatically improved the treatment of several different inflammatory diseases (3). In this context, it is particu- larly notable that neutralization of a single proinflam- matory cytokine, namely TNF, is highly effective in treating rather complex disease processes such as RA, psoriasis, and inflammatory bowel disease. Even more interesting, improvement of the signs and symptoms of disease occurs rapidly after the start of TNFi, even before remission of inflammation becomes clinically measurable. In this context, we have recently demon- strated that both mice with TNF-mediated arthritis and humans with RA show enhanced brain activity in centers involved in pain perception and the control of emotions (4). TNFi treatment reverses this enhanced central Supported by the DFG (FG 661/TP4) and the Immunopain project of the BMBF. Juergen Rech, MD, Andreas Hess, PhD, Stephanie Finzel, MD, Silke Kreitz, PhD, Marina Sergeeva, PhD, Matthias Englbrecht, Dipl. Psych., Arnd Doerfler, MD, PhD, Marc Saake, MD, Georg Schett, MD: University of Erlangen–Nuremberg, Erlangen, Germany. Drs. Rech and Hess contributed equally to this work. Dr. Englbrecht has received speaking fees from Pfizer, Ab- bott, and MSD (less than $10,000 each). Address correspondence to Georg Schett, MD, Department of Internal Medicine 3, Rheumatology and Immunology, University of Erlangen–Nuremberg, Krankenhausstrasse 12, Erlangen D-91054, Germany. E-mail: georg.schett@uk-erlangen.de. Submitted for publication April 21, 2012; accepted in revised form October 16, 2012. 325