Functional magnetic resonance imaging of peripheral neuropathic pain in the spinal cord and brainstem

To date, most studies investigating the neural signature of pain in humans have focused on the brain, and those studies concerned with more caudal areas (such as the spinal cord (SC) or brainstem) have used only experimental models of pain. The objectives of this study were (1) to determine the neural activity in the human brainstem and SC that is caused by a noxious mechanical stimulus and (2) to compare the neural response to noxious stimuli in healthy controls and a patient population diagnosed with peripheral neuropathic pain. The SC and brainstem contain important synaptic points in several major pain pathways, and comparing the neural response between a control and patient population in these areas provides a more complete picture of healthy and pathological pain processing.;We observed consistent signal intensity change in areas implicated in the transmission and modulation of pain in both control and CTS groups. Both groups showed a similar decrease in signal change with increasing pain, as results at pain level 2 are predominantly positive signal change and at pain level 6 are typically negative. This may indicate a reduction in the tonic inhibition of painful sensations. Differences between groups were readily visible in regions anatomically consistent with the dorsal horn (DH) of the cervical SC, rostral ventromedial medulla (RVM), dorsolateral pontine tegmentum (DLPT), and midbrain periaqudectal gray (PAG). The anatomical variation in signal change between groups may represent, for the first time, a visualization of the functional difference between healthy and pathological pain processing in the SC and brainstem using spinal fMRI.;Functional MRI studies of the SC and brainstem were carried out in healthy control subjects and patients diagnosed with carpal tunnel syndrome (CTS) in a 3T Siemens Magnetom Trio. Subjects reported the point at which the pressure (in mmHg, applied to the wrist at the location of the median nerve) corresponded to a pain level of 2, 4, and 6 on a numerical 11 point pain scale. Sagittal image slices were selected to span from the C7/T1 interface to the superior edge of the thalamus and were analyzed using a general linear model to improve the discrimination between physiological motion and signal intensity changes arising from neural activity. Spatially normalized group results superimposed on anatomical templates in the axial orientation were visually identified using several stereotaxic atlases.