PROJECT SUMMARY
Chronic pain is a highly prevalent problem in our society, is associated with significant personal suffering,
and disability, and incurs billions of dollars in cost annually. Despite its highly negative impact this medical
problem remains a clinical diagnosis relying on patients’ reports of pain intensity and the clinician’s physical
exam. The availability of reliable biomarkers for chronic pain, which do not rely on clinical diagnosis and
subjective pain reports, will increase our understanding of the pathophysiology of different chronic pain
conditions, improve care by providing clinicians with tools to classify and follow patients, and help greatly in
clinical trials developing novel analgesics where subjective pain reports remain the primary tool.
Neuroimaging results from our lab as well as others are starting to reveal reproducible brain signatures
of chronic pain involving changes in specific aspects of structural and functional cortico-striatal plasticity. As
such, we and others have shown that altered activity in, and functional connectivity between the nucleus
accumbens and medial prefrontal cortex, and loss of accumbens and dorso-lateral prefrontal cortex volume are
reproducible across laboratories and can potentially constitute an objective neural signature of chronic low back-
pain. However, it remains unknown whether this neural signature is generalizable to other musculoskeletal
chronic pain conditions like arthritis and/or chronic neuropathic pain like trigeminal neuralgia and/or chronic
affective conditions like major depression or is specific to CLBP. In addition, the effects of patients’ age and sex
on the neural signature of CLBP, which are major determinants of the clinical experience of chronic pain, remain
unknown. The overall aim of this application is to determine these two unknowns.
In Aim 1 we will test the specificity of the neural signature to musculoskeletal CLBP by testing its
predictive accuracy in chronic knee pain from osteoarthritis (KOA), which is a different musculoskeletal pain
condition, in chronic pain from trigeminal neuralgia (TN), which is a neuropathic pain condition, and in major
depressive disorder (MDD), which is a chronic negative affective condition. To that end, we will use machine
learning techniques to train a predictive model and classify CLBP patients from healthy controls (i.e., training
set) using a priori defined brain features drawn from reproducible findings based on our work and on the CLBP
neuroimaging literature. Next, this model will be validated on a new sample of CLBP patients and healthy
controls (i.e., test set) and tested in KOA, TN, and MDD patients to assess generalizability to other chronic pain
or affective conditions or specificity to CLBP. In Aim 2 we will study how age affects the predictive power of the
neural signature of CLBP and test how robust are the group differences between CLBP patients and healthy
controls in the brain features making up the neural signature across young (18-30 years old) and older (> 50
years) age groups. In Aim 3 we will explore the effects of sex on the neural signature of CLBP following a similar
approach to the one used in Aim 2 but using instead a grouping based on sex.