Global disruption of degree rank order: a hallmark of chronic pain

Abstract

Chronic pain remains poorly understood; yet it is associated with the reorganization of the nervous system. Here, we demonstrate that a unitary global measure of functional connectivity, defined as the extent of degree rank order disruption, kD, identifies the chronic pain state. In contrast, local degree disruption differentiates between chronic pain conditions. We used resting-state functional MRI data to analyze the brain connectome at varying scales and densities. In three chronic pain conditions, we observe disrupted kD, in proportion to individuals’ pain intensity, and associated with community membership disruption. Additionally, we observe regional degree changes, some of which were unique to each type of chronic pain. Subjects with recent onset of back pain exhibited emergence of kD only when the pain became chronic. Similarly, in neuropathic rats kD emerged weeks after injury, in proportion to pain-like behavior. Thus, we found comprehensive cross-species evidence for chronic pain being a state of global randomization of functional connectivity.

Pain Neuroscience Education: State of the Art and Application in Pediatrics

“…In order for this model to be successful, health-care providers need to convey the multi-dimensional causes of pain that are to be tackled in the intervention. The biological component of the biopsychosocial model largely revolves around the complex interplay of several cortical and subcortical brain regions involved in sensory, motor, cognitive, affective, and motivational functions [23]; with recent data suggestive of global brain connectivity reorganization among chronic pain patients [24]. This bombardment of neural input is a key mechanism that leads to chronic pain, as the brain keeps sending pain signals even in the absence of tissue damage.…”

Classification and characterisation of brain network changes in chronic back pain: A multicenter study

“…We also cannot determine the specificity of our results to chronic back pain, as opposed to other chronic pain conditions, and recent evidence suggests that many aspects of network changes may be common ( Baliki et al , 2014; Mansour et al , 2016). Hence future network studies would be greatly enhanced by longitudinal data (and pre-morbid data when available), better identifying correlations with pain severity, evaluation of response to drugs, and use of open data sources to provide larger data sets to test generalisation across diagnoses.…”

“…Of particular relevance is ’hub disruption’, which refers to a change in the nodal graph topology for any individual metric across the whole brain ( Achard et al , 2012). It has previously been shown that brain networks undergo hub disruption for degree (the number of connections for each node) in chronic pain, with evidence from both in human chronic back pain patients and rodent pain models ( Mansour et al , 2016). Here, we estimated Hub Disruption indices across all 3 data sets using a range of nodal graph metrics (see methods).…”

“…This value was chosen based on previous studies that have found such a link density to provide optimal discriminative ability ( Achard et al , 2012; Itahashi et al , 2014; Mansour et al , 2016; Termenon et al , 2016). Using the adjacency matrices, we calculated the Hub Disruption Index (HDI) - a well-recognized method that characterises functional reorganisation in resting-state brain networks in disease ( Achard et al , 2012; Itahashi et al , 2014; Mansour et al , 2016; Termenon et al , 2016). HDI is calculated based on the difference in a nodal graph-theoretic property of the network, and references the distribution of this metric across all nodes in a single subject, in comparison to the equivalent referential distribution in a cohort of healthy controls.…”

“…Using the Brain Connectivity Toolbox ( Rubinov & Sporns, 2010), we examined HDIs for degree, clustering coefficient, betweenness centrality, eigenvector centrality, K-coreness, flow Coefficient, local efficiency, and participation coefficient, and present results for those measures that were consistently significant across all data-sets. This choice is based on the measures that have been applied in previous studies ( Achard et al , 2012; Hashmi et al , 2014; Mansour et al , 2016). …”