Background/Purpose:

Inflammation is the hallmark of disease activity in rheumatoid arthritis (RA). RA is characterized by fatigue and cognitive/affective disturbances, and these symptoms often worsen during a disease flare. However, surprisingly little is known about the neural correlates of inflammation in autoimmune disease. Here we used MRI to investigate how neuroimaging outcomes are associated with levels of peripheral inflammation in this patient population.

Methods: RA patients, fulfilling ACR/EULAR 2012 criteria, were recruited and underwent a multi-modal MRI brain scan which was repeated 6 months later to validate our findings. Recruitment was stratified (1:1) by disease activity (swollen joint count ≥1). The clinical evaluation included measures of inflammation (ESR) and self-reported levels of fatigue and pain (VAS) as well as depression (Hospital Anxiety and Depression Scale). Images were acquired by a 3 Tesla, 8 channel phased array head coil using a T2*-weighted gradient-echo echo-planar imaging pulse sequence. Functional connectivity was assessed during a simple cognitive task and conducted through Independent Components Analysis (ICA) to investigate the relationship of established neural networks with other brain regions. Structural analyses were conducted using voxel-based morphometry (VBM). The primary outcomes of interest were brain regions where connectivity varied with levels of ESR. These regions were then examined post-hoc for associations with VBM findings, and clinical symptoms. Primary results were significant on the cluster level with a false discovery rate (FDR) p value <0.05 derived from an uncorrected voxel level p value <0.001.

Results: Fifty-four subjects participated (mean age 54.9 years; 75.9% female; mean disease duration 11.5 years; mean ESR 19.4 [2-62]; mean DAS28 3.6 [range 1.5-6.4]). Several regions were identified that showed positive connections to the Dorsal Attention Network (DAN), Salience Network (SLN), and Default Mode Network (DMN) as ESR increased. The right (R) frontal pole (r= .56, p=.002 FDR), and left (L) inferior/superior parietal junction (r=.58, p=.001) showed positive connections to the DAN while the L inferior parietal lobule, L mid-temporal gyrus and several frontal regions, demonstrated positive connections to the DMN and SLN (all p < .05 FDR). Positive connections between the R frontal pole and L parietal with the DAN were replicated at the second time point (p<.05). In the same parietal region, higher inflammation was associated with decreased grey matter volume. The degree of connectivity in most of the identified regions was associated with more severe fatigue and pain scores.

Conclusion: The neurobiology of fatigue, cognitive dysfunction, and pain in RA, and their relationship with inflammation are poorly understood, yet these are some of the most common and debilitating symptoms in rheumatic diseases. The present study identified brain regions that may serve as inflammatory “hubs” in RA. Several of these regions play important roles in higher-order cognitive processing and therefore may represent key areas underpinning the development of symptoms such as fatigue and cognitive impairment in RA.

« Back to 2017 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/the-neural-correlates-of-inflammation-in-ra-a-multi-modal-mri-study/