Completing the Pain Circuit: Recent Advances in Imaging Pain and Inflammation beyond the Central Nervous System.

Early work in the fields of neuroanatomy, neurophysiology, and clinical observations has provided a robust description of pain pathways. These pathways can now be evaluated with imaging to contribute to a more objective view of pain, where both the sensory and emotional experience may be assessed in health and disease. This review describes some recent advances in imaging of pain and inflammation-related processes below the level of the brain, that is, at the level of 1) the periphery; 2) the nerve; and 3) the nerve root. We discuss methods to measure neuroinflammation and future lines of inquiry linking peripheral markers to spine, brainstem, and brain functional imaging. An ultimate goal is a more mechanistic definition than the one currently offered by the International Association for the Study of Pain: “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage.” 1

An increasing body of literature has implicated inflammation as a contributor to the initiation but also to the maintenance of chronic pain, whether it be an obvious inflammatory process produced by specific disease (e.g. rheumatoid arthritis, chronic pancreatitis), following trauma (e.g. post-surgical neuroinflammation), or other neuropathic conditions such as complex regional pain syndrome (CRPS). It seems increasingly clear that peripheral inflammation may produce central inflammatory processes. 80 – 85 In addition, central inflammation—even in brain areas rarely considered to be involved in chronic pain, for example in the hippocampus—produce neuropathic-like behavior in rats. 86 In humans such changes in brain neuroinflammation contribute to altered pain 87 and mood changes. 88 , 89 Thus, the ability to measure neuroinflammation in humans with pain in both the peripheral and central nervous systems may provide objective indices for: 1) ongoing inflammation that may produce the maintenance of the disease either in the periphery 83 or centrally; 36 and 2) objective measures for treatment effects. While imaging markers may provide an initial definition of the status of inflammation, blood or serum markers may eventually be more sensitive and provide a more cost-effective use in the clinic.

The metabolic rate of glucose increases in the spinal cord during nociceptive in-flow, 57 , 58 affording a mechanism to image spinal pain signaling using 18 F-fluorodeoxyglucose. We found no studies demonstrating altered spinal PET ligand uptake in pain, but such an endeavor appears possible if there is massive peripheral signaling. FDG is routinely used in oncological staging, and a retrospective analysis of cancer pain patients might demonstrate elevated FDG uptake in corresponding segments of the spinal cord. Ideally, such a study would utilize high-resolution PET in combination with MR or CT to delineate the spinal cord cross-section in multiple voxels, allowing assessment of anterior and posterior segments, and possibly lateralization effects. To illustrate PET imaging of the spine, we present mean FDG standardized uptake values (SUV) obtained from two studies of 92 patients 59 and 30 patients 60 without spinal malignancy ( Figure 4 ). Midline 18 F-FDG PET/CT of a Healthy Spinal Cord. Bar plot illustrates weighted mean cord-to-background (CTB) ratio from 122 patients. 59 , 60 Structural MRI is used routinely to assess spinal cord injuries, but due to the spine’s small cross-section, and noise sources such as motion, cerebrospinal fluid (CSF) pulsation, and magnetic susceptibility, functional imaging of the spine is technically challenging. Recent developments in MR sequences and post-processing have opened up the field, and it is possible to define structure and function with greater specificity. 61 The first functional spinal cord imaging results were published in 1999, indicating that 3-tesla imaging of the cervical spinal cord showed that repeated hand exercise led to a blood-oxygenation level dependent (BOLD)-like increase in spinal cord signal, predominantly on the ipsilateral spinal cord between C6 and T1. 62 Since then, spinal fMRI has been reported using multiple paradigms (pain, motor, vibration, light touch) in healthy subjects and in patient populations including carpal tunnel syndrome, spinal cord injuries, and multiple sclerosis. These studies, along with methodological advances, are the subject of two excellent reviews on state-of-the-art spinal cord imaging methods 63 and applications 64 that we refer the reader to for full details.

Primary afferent nerves in the dorsal root ganglia convey pain information to the central nervous system. Both peripheral inflammation and nerve damage can lead to alterations in anatomy and function of neurons within the ganglion, alterations that contribute to persistent pain states. 46 , 47 While the dorsal roots are too small for standard neuroimaging approaches, the trigeminal ganglion serves an equivalent role for the trigeminal nerve. The trigeminal ganglion is located at the base of the brain in the posterior cranial fossa across the superior border of the petrous temporal bone. It comprises sensory neurons from the ophthalmic (V1), maxillary (V2), and mandibular (V3) divisions of the trigeminal nerve. We have demonstrated that fMRI can be used to assess both sensory (brush) and noxious thermal activation of the ganglion. Activation occurred ipsilaterally and somatotopically, as predicted by the known anatomical segregation of the neurons comprising the V1, V2, and V3 divisions of the nerve ( Figure 3 ). 46 We have further demonstrated that sensory processing in patients with trigeminal neuropathic pain is associated with distinct activation patterns consistent with sensitization within and outside of the primary sensory pathway, 48 and, in a case study, we demonstrated trigeminal ganglion activation in photophobia. 49 Using diffusion tensor imaging, we have further been able to segment the peripheral trigeminal circuitry, trigeminal nerve branches (ophthalmic, maxillary, and mandibular nerves), ganglion, and nerve root, and further segment the spinal trigeminal and trigeminal thalamic tracts, which, respectively, convey information to the spinal trigeminal nuclei and ventral thalamic regions. 50 Moreover, we have demonstrated a direct pathway from the optic nerve to the pulvinar nuclei in the posterior thalamus, providing a possible mechanism for exacerbation of pain by light in migraine. 51 Other groups have demonstrated alterations in trigeminal nerve diffusion in trigeminal neuralgia 52 – 55 and in temporomandibular disorder. 56 Somatotopically Organized Activation Patterns of the Human Trigeminal Ganglion Evoked by Noxious Heat to the Ophthalmic (V1), Maxillary (V2), and Mandibular (V3) Facial Regions. Adapted with permission from Borsook et al. 46 Taken together, these studies demonstrate that, at least for cranial nerves, functional and diffusion MRI can provide mechanistic insight into pain processes at the interphase of the peripheral and central nervous system.

There are only a handful of studies that have attempted to relate peripheral alterations to CNS dysfunction. An excellent example of such an approach is determining the relationship between carpal tunnel nerve conduction velocity and regional gray matter alterations in the brain. 65 This study found that patients with carpal tunnel syndrome had significant gray matter reductions in the hand area of the somatosensory cortex, a reduction that was correlated to lower median nerve conduction velocity. Of note, diffusion tensor imaging (DTI) of the medial nerve pre and post carpal tunnel surgery indicates that postoperative clinical improvement is related to nerve diffusivity but not anisotropy. 66 A next step may be to combine peripheral MR neurography with CNS imaging of brain morphology and function to evaluate how and when the periphery and CNS are affected by treatment. Another recent example used a combined analyte, behavioral, and imaging assessment of a rat sciatic nerve injury model to provide a “pathophysiological signature”; results indicate that the nerve injury was reflected in peripheral and central soft tissues, as well as in the expression of circulating cytokines, chemokines, and growth factors. 67 Functional MRI and machine-learning pattern recognition can be used to define neurologic signature of acute pain with high sensitivity and specificity. 68 The hardware (3T MRI, PET, and PET-MR), scanning sequences (structural, diffusion, BOLD, and spectroscopy), and analytical software now available have allowed the research community to quantify several aspects of the pain circuit, 69 – 75 as illustrated in Figure 5 . This circuitry is further linked to behavioral and psychological measures of pain experience, pain-related behaviors, and pain-induced co-morbidities and risk factors such as catastrophizing, fear of movement, and depression. The levels of inquiry range from genetic via neurophysiological to psychological and even sociological and anthropological domains (i.e. the perception, expression, and tolerance of pain are influenced by a variety of non-biological processes, such as disparities in work, economy, daily living, social life, gender norms, and cultural setting 76 – 78 ). An Illustration of the Levels of Inquiry in Pain Behavior That Imaging Has the Possibility to Inform Upon. The Fear-Avoidance Model is an adaptation from Vlaeyen et al. and Lethem et al. 69 , 70 Notably, inflammatory processes may interact at several levels of the pain behavior circuit. For example, the catechol-O-methyltransferase (COMT) val158met polymorphism may influence the neuronal 71 and opioidergic 72 response to pain (but see also Nicholl et al. 73 ). Furthermore, the 18-kDa translocator protein (TSPO) Ala147Thr genotype asserts a strong influence on the binding affinity of microglial PET tracers 74 and thus needs to be accounted for in between subject analyses. At the other end of the spectrum, imaging studies are beginning to explore how culture may interact with brain processing of perception and emotional valuation. 75 Clearly, the study of pain is and will remain a multidisciplinary field. Animal imaging of brain systems, reviewed by Borsook and Becerra, 79 offers the possibility of imaging awake animals and may serve as a “language of translation” between preclinical to clinical models. Human imaging, in turn, has strengthened and made objective the links between CNS neurophysiology and psychology of pain modulation. We foresee a similar development in the field of peripheral inflammation and spine imaging.