Visceral Pain: Basic Mechanisms

Peripheral and Central Organization of Visceral Sensory Innervation

Among all tissues in the body, the viscera are unique in that each organ typically receives innervation from two sets of nerves, either vagal and spinal nerves or two anatomically distinct sets of spinal nerves. An older terminology described innervation of the viscera by the thoracolumbar spinal nerve as sympathetic because these afferent axons were anatomically associated with efferent axons of the sympathetic division of the autonomic nervous system; called them “afferent sympathetic fibers.” Vagus and pelvic nerve afferents were termed parasympathetic because of a similar anatomical association with the parasympathetic division of the autonomic nervous system. This terminology was also related to presumed function. It is still widely assumed that visceral pain is conveyed to the central nervous system (CNS) by afferent sympathetic fibers whereas parasympathetic afferent innervation regulates autonomic control. Contemporary evidence contradicts these assumed functions, and visceral afferent fibers are best described by the nerve’s name to avoid assumed functions as implied by confusing terminology.

Visceral afferent fibers are contained in nerves that terminate in the spinal cord except for those in the vagus and glossopharyngeal nerves, which terminate in the brain stem to provide a supraspinal, cranial component of visceral sensory innervation. The vagus nerve is undoubtedly the most far-reaching sensory nerve in the body. At least 80% of vagal axons are afferent, and most internal organs are innervated by the vagus nerve. The bilateral vagus nerves innervate the larynx, all the thoracic viscera (esophagus, heart, bronchopulmonary system), and most if not all of the abdominal viscera (stomach, small and large intestines, liver, proximal colon, etc.) ( ). The cell bodies of vagal afferent fibers are contained in the nodose ganglion (primarily) and a smaller, more proximally situated jugular ganglion with central terminals located principally in the nucleus of the solitary tract in the dorsal medulla. Not all vagal afferents terminate in the brain stem; about 5% project directly to and terminate in the upper cervical spinal cord (C1–2), where they are believed to contribute to referred sensations, as well as to propriospinal mechanisms of nociceptive modulation ( ). In support of the latter, electrical vagal afferent stimulation modulates spinal thoracic and lumbar nociceptive transmission and is analgesic in humans. Despite its widespread innervation of the internal organs, the vagus nerve has long been considered to play no role in the transmission of visceral nociceptive information, a role relegated to spinal afferent nerves, including the pelvic nerves. Growing evidence, however, suggests that vagal afferents are critical for chemonociception (see below) and, importantly, contribute to the affective dimensions and unpleasantness associated with visceral pain.

As illustrated in Figure 51-1 , spinal nerve innervation of the viscera is distributed from the cervical to the sacral spinal segments. All spinal nerves have their cell bodies in dorsal root ganglia (DRGs; not illustrated in Fig. 51-1 ), although unlike spinal somatic nerves, most visceral nerves pass through pre- and paravertebral ganglia en route to the spinal cord (see Fig. 51-2 for detail). In prevertebral (sympathetic) ganglia, axons of visceral nerves often give off collaterals that synapse on secretory or motor neurons contained in the ganglia and thus can influence organ function (e.g., motility). In addition, visceral afferent fibers that access the spinal cord through paravertebral ganglia can travel rostrally or caudally in the sympathetic trunk and enter distant spinal segments. Regardless of their route to the spinal cord, visceral afferents terminate in similar patterns within the spinal cord:

• Superficial dorsal horn (lamina I and II), which is also the site of termination of somatic nociceptors

• Intermediolateral cell column and sacral parasympathetic nucleus, where they influence sympathetic and parasympathetic efferent outflow to the viscera

• Around and dorsal to the central canal, an area termed lamina X

In addition to this extrinsic sensory innervation, most viscera (e.g., the gastrointestinal [GI] tract, heart) also possess an independent, intrinsic innervation. Best understood is the enteric nervous system of the GI tract, which encodes basic patterns regulating secretion, motility, and blood flow and interacts with the extrinsic innervation of the gut, but in ways that are poorly understood at present.

Visceral Afferent Fibers and Receptive Ending Structures

With the exception of a small number of Aβ fibers associated with pacinian corpuscles in the mesentery, the overwhelming majority of visceral afferent fibers are thinly myelinated Aδ or unmyelinated C fibers. Very little is known about the mechanisms of energy transduction or the structure of visceral afferent peripheral terminals, and it is assumed, as for somatic Aδ and C fibers, that most visceral afferent endings are unencapsulated, “free” nerve endings, but this has little direct experimental support. Most is known about mechanosensitive endings in the viscera, primarily the morphology of vagal afferents that innervate the GI tract. Mechanosensitive afferent endings in hollow organs are assumed to be frequently associated with the muscle layers and to be responsive to tension/stretch ( ). Best characterized are the intraganglionic laminar endings (IGLEs) and intramuscular arrays (IMAs) associated with vagal afferent fibers that innervate the stomach, both of which have been shown to be mechanosensitive. IMAs differ from IGLEs in morphology, in distribution, and probably in adequate stimulus for activation ( ). As the name implies, IMAs consist of long terminal arrays running within either the circular or longitudinal muscle layers of the organ. However, acute GI discomfort or pain produced by organ distention is not associated with vagal afferent input to the CNS, and the proposed sensitivity of IMAs to stretch (e.g., gastric distention, a noxious stimulus only at high intensity) is uncertain at present, thereby preventing definitive conclusions about their role in nociception.

Complementary morphological information about spinal visceral nerve mechanoreceptor endings in organs is limited ( ). Low-threshold, slowly adapting mechanoreceptors have been described in the rectal innervation of the guinea pig rectum ( ). Morphologically, these receptors resemble IGLEs; comparable structures have not been described in the spinal afferent innervation of the colon. The distribution of these receptive structures in the proximal and distal portions of the GI tract suggests a potential regulatory role in food intake and defecation. Most studies reveal single, small-sized receptive fields for mechanosensitive afferent fibers, although vagal and spinal mechanosensors in the stomach and colon, as well as afferent fibers innervating the bladder, have occasionally been reported in electrophysiological studies to have multiple receptive fields.

In addition to mechanosensitive endings, chemo- and thermoreceptive endings are present in the viscera. Virtually nothing is known about the morphology of thermoreceptive peripheral terminals, but the architecture of presumptive chemoreceptor endings in the vagal afferent innervation of the rodent pyloric antrum and proximal duodenum has been described. noted the presence of three vagal afferent specializations innervating the mucosa of the proximal part of the GI tract: villus afferents, crypt afferents, and antral gland afferents. Their location suggests roles in chemosensation (response to nutrients, content in chyme, etc.); whether this includes chemonociception remains to be determined. Functional characterization of mechano-, chemo- and thermosensitive endings (discussed below) reveals the presence of receptive endings in the mucosa and serosa, as well as in the muscle layers of hollow organs and mesenteric attachments.

Density and Complexity of Visceral Innervation

The number of axons that innervate the viscera is relatively small in comparison to somatic innervation. It has been estimated that 5–15% of the total afferent input to the spinal cord arises from the viscera, which is disproportionate to the greater than 50% of second-order spinal neurons estimated to respond to visceral afferent input. This apparent discrepancy is explained by the significant arborization and spread of visceral afferent terminals within the spinal cord. Whereas somatic input is commonly restricted to one or a few spinal segments, spinal visceral afferent input has been documented to spread several segments rostral and caudal from the spinal segment of entry and, moreover, to occasionally spread to the contralateral side of the spinal cord ( ). Similarly, vagal input to the medulla is amplified by the branching and widespread distribution of afferent terminals in the organs of innervation.

Visceral afferent input to the spinal cord is also characterized by convergence. That is, virtually all second-order spinal neurons that receive visceral input also receive convergent somatic input from skin and/or muscle (see Fig. 51-2 ), which provides an explanation for referral of visceral sensations to somatic sites (e.g., deep retrosternal pain that radiates to the neck, shoulder, or jaw with angina). Furthermore, viscerovisceral convergence of input onto second-order spinal neurons is also common (e.g., colon and urinary bladder, gallbladder, and heart). An older concept—dichotomizing visceral afferents—has recently been reinvigorated by reports that some visceral afferent endings innervating adjacent organs arise from a common cell body, thus revealing prespinal “convergence” in DRG somata. Morphological studies in the cat, rat, and mouse using retrogradely transported dyes clearly reveal that dichotomizing afferents innervate the pelvic organs (see for review). The reported proportion of such afferents among the total organ afferent innervation in rodents ranges widely (3–27%) but is generally a small fraction of the visceral innervation. Their role in cross-organ sensitization (see below) remains to be confirmed as being functionally significant.

Accordingly, the diffuse character and poor localization of visceral pain are contributed to by the widespread intraspinal arborization of visceral afferent terminals and by somatovisceral and viscerovisceral convergence onto second-order spinal neurons, thus challenging patients and physicians alike to easily identify the source or sources of visceral pain.

Neurochemistry of Visceral Primary Afferents

As indicated above, most visceral receptive endings in organs are typically non-encapsulated (“free”) and associated with slowly conducting unmyelinated (C) or thinly myelinated (Aδ) axons of generally small- to medium-diameter somata in the dorsal root and nodose ganglia. It would be convenient and useful if neuronal function could be assigned on the basis of cell size and myelination, but neither feature defines a nociceptor ( ). Because the somata of visceral nociceptors are generally larger than those of non-visceral nociceptors, they have been largely excluded from studies that focus on small-diameter somata in the DRG. Beyond cell size, cell content has been used to broadly segregate nociceptive afferents into peptidergic and non-peptidergic ( ). Peptidergic nociceptors contain the neuropeptide substance P and/or calcitonin gene–related peptide (CGRP) and express the nerve growth factor (NGF) receptor TrkA (tyrosine kinase receptor A). In contrast, non-peptidergic nociceptors express Ret, the receptor for the glial cell line–derived family of neurotrophic factors, and most also bind isolectin B4 (IB4) and express the purinergic P2X ₃ receptor. Even this classification, however, obscures differences between nociceptors innervating different tissues. For example, a significantly greater proportion of visceral than cutaneous nociceptors are immunoreactive for CGRP and express TrkA (see for an overview). The presence or absence of additional or alternative markers (e.g., the voltage-gated sodium channels Na _v 1.7 or Na _v 1.8 and transient receptor potential [TRP] channels A1 and V1) has been suggested as a means of identifying subsets of nociceptors. These additional markers also have limitations, principally that their distribution or presence varies among subpopulations of nociceptors defined by the target of innervation, which moreover changes as the state of the tissue changes (e.g., inflammation). Accordingly, the ability to classify a sensory neuron as nociceptive (which is a functional characterization) by cell content or expression of one or a combination of receptors is not possible at present.

Despite the lack of a definitive marker, it is known that more visceral afferent somata (than sensory neurons innervating skin or muscle) contain substance P and/or CGRP. This fraction increases further during inflammatory processes associated with pain. Similarly, more visceral than somatic sensory neurons express the high-affinity NGF receptor TrkA, with further increases occurring during inflammation. TRP channels have been a recent focus of study because many believe that expression of TRPV1 denotes a neuron as a nociceptor. Indeed, approximately 70–80% of visceral DRG somata are immunoreactive for TRPV1, whereas proportions of skin (35–60%) and muscle (≈40%) DRG somata are much less ( ). Not surprisingly, there also are significant differences in cell content markers between the two innervations of the same organ. For example, expression of TRPV1 mRNA in DRG neurons did not differ between the lumbar splanchnic (thoracolumbar [TL] DRG) and pelvic (lumbosacral [LS] DRG) nerve innervations of either colon or urinary bladder somata in TL and LS DRGs (≈45–60% of somata expressed TRPV1 mRNA). In contrast, whereas 60% of bladder TL neurons expressed the TRPA1 gene transcript, virtually no bladder LS neurons expressed this gene transcript; co-expression of TRPV1 and TRPA1 transcripts was high in colon DRG neurons ( ).

Sensory neurons release peptides (e.g., substance P and other neurokinins, CGRP) and other bioactive mediators from their peripheral and central terminals (e.g., in response to “mucosal noxae” or capsaicin, low pH, and distention of the colon; ). For example, substance P affects epithelial cells in the urinary bladder (urothelium) and GI tract, CGRP contributes to local vasodilatation, and mast cells, which in the viscera are found in close proximity to nerves and express neurokinin 1 (NK1) receptors, degranulate in response to substance P, which causes the release of additional mediators (histamine, tryptases). Mucosal noxae can thus lead to vasodilatation and hyperemia, inflammatory infiltration via chemotaxis, capillary leakage with exudate of plasma and swelling, and activation of smooth muscle cells, as well as increased secretion in epithelial cells, all of which are features generally associated with neurogenic inflammation. Emerging evidence suggests that this mechanism contributes to the pathogenesis of acute and chronic diseases involving the viscera (e.g., ).