MACROSCOPIC STRUCTURE OF THE COLON

The human colon is just over one meter long and is divided anatomically into the cecum; the ascending, transverse, descending, and sigmoid colon; and the rectum, which lies between the rectosigmoid junction and the anal canal. The outer longitudinal smooth muscle layer forms three thick, cord-like structures called the teniae coli, which are spaced evenly around the circumference of the colon. Between the teniae, the longitudinal smooth muscle is much thinner, allowing the wall to bulge noticeably.

Irregularly spaced circumferential constrictions pinch the colon into a series of pockets, called haustra, which give the colon a sacculated appearance for much of its length. Haustra are not fixed structures and appear to be caused by sustained contractions of the circular muscle. Myogenic activity alone, however, does not seem sufficient to explain haustration, and neural input is likely to contribute to their formation; haustra move, disappear, and re-form during the propulsion of colonic contents.

The teniae fuse to form a continuous outer longitudinal smooth muscle layer at the rectosigmoid junction, which then continues down to the distal margin of the anal canal, insinuating itself between the internal and the external anal sphincters. Throughout the length of the colon, the circular smooth muscle layer consists of thick bundles of cells, which are separated by connective tissue septa. The internal anal sphincter consists of a thickening of the circular muscle layer over the last 2 to 4 cm of the anal canal.

Macroanatomy of the colon is also discussed in Chapter 96.

STRUCTURE AND ACTIVITY OF COLONIC SMOOTH MUSCLE

ION CHANNELS IN COLONIC SMOOTH MUSCLE

The membrane of colonic smooth muscle cells contains a variety of ion channels, including several types of potassium channels, calcium channels, chloride channels, and nonselective cation channels.² Although the exact physiologic roles of many of these ion channels are unknown, the high-threshold, voltage-operated calcium channels (l-type calcium channels) do play a crucial role in colonic muscle contractility. These channels open when the membrane potential of smooth muscle cells is depolarized beyond a voltage threshold, and they are responsible for the rapid upstroke of smooth muscle action potentials. The influx of calcium through l-type calcium channels during action potentials is a major trigger for activation of the contractile apparatus. It is not surprising that pharmacologic blockade of l-type calcium channels by dihydropyridine drugs such as nifedipine can reduce the contractility of colonic smooth muscle substantially. Release of calcium from intracellular stores, which is triggered by excitatory neurotransmitters, also may play a role in muscle contraction.

INTERSTITIAL CELLS OF CAJAL

Since 1991, the interstitial cells of Cajal (ICC) have been shown to play at least two important roles in the control of gastrointestinal motility: control of myogenic activity and mediating or amplifying the effects of motor neurons on the smooth muscle apparatus. ICC are non-neuronal in origin and are derived from common progenitors of smooth muscle cells. Mutant mice and rats that are deficient in ICC have profoundly disturbed intestinal motility, an observation that provides insight into the roles of ICC in the human gastrointestinal tract.

In the human colon, three types of ICC are recognized and are named according to their locations: ICC in the plane of the myenteric plexus (ICC_MY), ICC near the submucosal plexus (ICC_SM), and intramuscular ICC located between the circular and the longitudinal muscle layers (ICC_IM). ICC_MY and ICC_SM form extensive networks along the colon and are electrically coupled to one another and to the smooth muscle layers by gap junctions (Figs. 98-1 and 98-2). ICC_MY probably are the pacemakers for the small, rapid (12-20/min) oscillations in membrane potential (MPOs) of longitudinal and circular smooth muscle layers. ICC_SM are the pacemakers for the large-amplitude slow waves (2-4/min) originating in the plane of the submucosal plexus; these slow waves have a powerful influence on the patterning of circular muscle contraction.

Figure 98-1. Schematic cross section of the muscularis externa of the human colon. The outer longitudinal smooth muscle layer (LM) is thickened at the teniae. In the plane of the myenteric plexus (not shown) is a network of interstitial cells of Cajal (ICC), which generate a rapid myenteric potential oscillation (ICC_MY; see lower waveform on the right). The circular muscle layer (CM) is innervated by axons of enteric motor neurons with transmitter release sites (clusters of clear vesicles) that are associated with specialized intramuscular ICC (ICC_IM). At the outer border of the circular muscle is another network of submucosal ICC, which generate slow waves (ICC_SM; see upper waveform on the right). Also present are axons of motor neurons in the longitudinal muscles and ICC_IM (not shown in this cross section). The tiny white squares represent gap junctions, which electrically couple cells.

Figure 98-2. Micrographs of interstitial cells of Cajal (ICC) in the human colon, labeled by c-Kit immunohistochemistry. A, ICC in the plane of the myenteric plexus (ICC_MY) have an irregular shape, form a dense network of cells, and probably function as pacemakers. B, A different plane of focus of the same region shows spindle-shaped intramuscular ICC (ICC_IM) in the overlying circular muscle layer. These cells probably are involved in neuromuscular transmission to the smooth muscle.

(Courtesy of Liz Murphy and David Wattchow.)

The exact ionic basis of rhythmicity in ICC_MY and ICC_SM that gives rise to MPOs and slow waves is not entirely clear; however, oscillations in membrane potential are an intrinsic property of both ICC_MY and ICC_SM. Intramuscular ICC (ICC_IM) are a major target of neurotransmitters released from the axons of excitatory and inhibitory enteric motor neurons. Acetylcholine and nitric oxide (and probably several other motor neuron transmitters) evoke changes in the membrane potential of ICC_IM, which then spread through the smooth muscle by means of gap junctions. ICC_IM also may be involved in amplifying the slow waves as they spread through the muscle layers. Thus, these cells appear to be key players in integrating non-neuronal pacemaker activity and neuronal inputs to the smooth muscle.

The discovery that cellular mechanisms long considered to be the properties of smooth muscle cells actually are mediated by ICC may have important clinical implications. For example, in the distal bowel, reduced numbers of ICC, or a reduction in the total volume of ICC, has been associated with anorectal malformations, colonic manifestations of Chagas’ disease, and possibly some cases of slow-transit constipation.³ Some reports have suggested that the density of ICC may be reduced in aganglionic segments of colon in Hirschsprung’s disease and that this deficit might contribute to diminished propulsive activity; this finding, however, has not been consistent between studies.³

THE ENTERIC NERVOUS SYSTEM

Direct neuronal control of colonic motility is mediated mostly by the enteric nervous system (ENS). Although the ENS is capable of expressing a diverse repertoire of motor patterns, its functions are modulated by sympathetic, parasympathetic, and extrinsic afferent pathways (Fig. 98-3). In terms of numbers of nerve cells, the ENS is by far the largest component of the autonomic nervous system, with considerably more neurons than those of the parasympathetic and sympathetic divisions combined. The nerve cell bodies of the ENS are located in plexuses of myenteric ganglia (Auerbach’s plexus), which lie between the longitudinal and the circular muscle layers of the muscularis externa, or in the submucosal ganglia, which lie between the circular muscle and the mucosa (Fig. 98-4).

Figure 98-3. The extrinsic innervation of the human colon. Parasympathetic efferent pathways (filled cell bodies) arise from the dorsal motor nucleus of the vagus in the brainstem and pass through the vagus nerve and prevertebral sympathetic ganglia, through the lumbar colonic nerves to the proximal colon. Parasympathetic pathways also extend from nuclei in the sacral spinal cord that run through the pelvic nerves and either synapse in the pelvic plexus ganglia or run directly into the bowel wall. Sympathetic pathways (open cell bodies) consist of preganglionic neurons in the thoracic spinal cord that synapse with sympathetic postganglionic neurons either in the inferior mesenteric plexus or in the pelvic plexus. Enteric nerve cell bodies in the colon receive input from both parasympathetic and sympathetic pathways. Viscerofugal enteric neurons project out of the bowel to the prevertebral ganglia. Afferent pathways consist of vagal afferent neurons from the proximal colon with cell bodies in the nodose ganglia. In addition, spinal afferent neurons with cell bodies in lumbar dorsal root ganglia (DRG) run through the lesser splanchnic and colonic nerves to the colon and mediate nociception. Another population of spinal afferents, with cell bodies in the sacral DRG, runs through the pelvic nerves and pelvic ganglia to the rectum; these include sensory neurons that transmit non-nociceptive information about the distention of the rectum. The striated muscles of the pelvic floor (including the external anal sphincter) are supplied by motor neurons with cell bodies in the spinal cord and axons that run in the pudendal nerves. Triangles represent transmitter release sites; combs represent sensory transduction sites.

Figure 98-4. Diagram showing the layers and components of the intestinal wall. The lumen is at the top and the longitudinal muscle layer is at the bottom. Auerbach’s myenteric plexus and the submucosal plexuses (Meissner’s and Schabadasch’s plexuses) are shown, along with some of their major classes of enteric neurons. The networks of interstitial cells are shown in Figure 98-1.

The submucosal plexus is divisible into at least two networks: Meissner’s plexus, which lies closer to the mucosa, and Schabadasch’s plexus, which lies adjacent to the circular muscle. Some authors have identified an additional intermediate plexus. Internodal strands that contain hundreds of axons run within and between the different plexuses. Finer nerve trunks innervate the various target tissues of the intestinal wall, including the longitudinal muscle layer, circular muscle, muscularis mucosae, mucosal crypts, and mucosal epithelium. Within the ganglia of each plexus, different functional classes of enteric nerve cell bodies are intermingled, although differences in the proportions of cell types among the plexuses have been observed. It has become clear that an exquisite degree of organization is characteristic of the ENS, each class of nerve cell making highly specific and precise projections to its particular target.

The ENS uses many transmitters in addition to the major transmitters acetylcholine and nitric oxide, including tachykinins, purines, numerous other modulatory peptides, and some amines. Many other substances, released from neural and non-neural cells, also modulate neuronal and muscular excitability, including gaseous mediators (carbon monoxide and hydrogen sulfide) and, in inflammation, prostanoids, cytokines, purines, bradykinin, H⁺ ions, and neurotrophins.

SYMPATHETIC INNERVATION

The major sympathetic innervation of the proximal colon arises from the inferior mesenteric ganglion and projects through the lumbar colonic nerves to the ascending and transverse colon (see Fig. 98-3). A small number of sympathetic neurons in the celiac and superior mesenteric ganglia, in the paravertebral chain ganglia, and in the pelvic plexus ganglia also project to the colon (see Fig. 98-3). These neurons receive a powerful cholinergic drive from preganglionic nerve cell bodies in the intermediolateral column of the spinal cord (segments L2-L5). This is a major pathway by which the central nervous system modifies bowel activity, such as during exercise. Sympathetic efferent neurons also receive input from the enteric viscerofugal neurons and from extrinsic, spinal sensory neurons with cell bodies in the dorsal root ganglia, forming several reflex loops with the distal bowel.

Sympathetic nerve fibers from prevertebral ganglia cause vasoconstriction of the mucosal and submucosal blood vessels. Other cells project to the enteric ganglia, where they cause presynaptic inhibition of synaptic activity in the ENS and thus depress reflex motor activity. Another target for sympathetic axons is the circuitry of the submucosal plexus (largely Meissner’s plexus) involved in controlling epithelial secretion. Hence, these pathways inhibit colonic motor activity, reduce blood flow, and inhibit secretion to limit water loss from the body during times of sympathetic activation. In addition, some sympathetic axons innervate the smooth muscle directly, particularly the ileocecal junction and internal anal sphincter, where they cause contraction; these effects also are consistent with closing down enteric motor activity during sympathetic arousal.

PARASYMPATHETIC INNERVATION

The colon receives parasympathetic innervation from both the vagus nerve and pathways in the sacral spinal cord. Branches of the vagus nerve reach the prevertebral ganglia (superior hypogastric plexus) and then run with sympathetic axons to the cecum and the ascending and transverse colon. The distal colon is supplied largely by sacral parasympathetic axons via the pelvic nerves (pelvic splanchnic nerves). Some of these cholinergic spinal efferent neurons synapse first onto nerve cell bodies in the pelvic plexus (inferior hypogastric plexus), and others project directly to the colon. From their point of entry into the colon, many of the axons run in an oral direction and form thick trunks called shunt fascicles. Parasympathetic axons project to the enteric ganglia in the colon, where they make excitatory cholinergic synapses onto enteric nerve cell bodies. Sacral parasympathetic pathways play an important role in increasing the propulsive activity of the distal colon before defecation and also may be involved in triggering the propagating complexes that start more proximally before defecation.

EXTRINSIC AFFERENT PATHWAYS

Sensation from the colon is mediated by primary afferent neurons with cell bodies outside the bowel wall. Vagal afferent neurons, with nerve cell bodies located in the nodose and jugular ganglia, project to the proximal colon and run with the vagal efferent parasympathetic pathways. Currently, their exact role in reflex control and sensation is not clear, but they are unlikely to be involved in the transmission of pain sensation from the colon.

The entire colon also is innervated by spinal primary afferent neurons with nerve cell bodies in the lumbar dorsal root ganglia. Lumbar spinal afferents project along the lumbar splanchnic nerves, through the prevertebral inferior mesenteric ganglion, and through the lumbar colonic nerves to the colon, where they terminate in sensory endings in the mesentery, serosa, muscular layers, and mucosa throughout the entire colon and rectum. In addition, a population of spinal afferents, with cell bodies in the sacral dorsal root ganglia, projects along the pelvic nerves to the colon and traverses the pelvic plexus en route. Evidence indicates that some of these sacral spinal afferent neurons form a functionally different population from the lumbar spinal afferents, encoding different types of information, particularly from the rectum. Sacral afferents include many mechanoreceptors with a low threshold and wide dynamic range; these mechanoreceptors probably are responsible for graded sensations of rectal filling and for activating defecatory reflexes.⁵

By contrast, lumbar spinal afferents and some higher-threshold sacral afferents are responsible for generating pain sensations from all regions of the colon and rectum. They respond to gross distention of the bowel wall, traction on the mesenteric membranes, powerful colonic contractions, or chemical stimulation of the mucosa by bile acids, high osmolarity, and other stimuli. It is well established that the sensitivity of many spinal afferents is increased greatly by inflammation in the colon wall. In addition to their role in sensation, spinal afferents also have axon branches (collaterals) in enteric ganglia and prevertebral sympathetic ganglia and on mucosal blood vessels, where they might play a role in generating peripheral reflex responses to noxious stimuli.

In summary, sacral afferent and efferent (parasympathetic) pathways run in parallel and connect the distal bowel with neural circuitry in the sacral spinal cord via pelvic and rectal nerves. The important role of these pathways in both rectal sensation and in generating the enhanced motility required for defecation is clearly demonstrated by the effects of nerve lesions at several levels. Thus, severing of peripheral nerves and distal spinal cord injury can lead to loss of rectal sensation and to severely impaired defecatory ability.

NONPROPAGATING MOTOR PATTERNS

Nonpropagating, apparently random activity makes up the majority of the recorded colonic motor activity and is presumed to serve segmenting or mixing functions. The frequency of nonpropagating colonic contractions in vivo is generally 2 to 4 cycles per minute, similar to the frequency of the spontaneous myogenic slow waves generated by ICC_SM at the submucosal border of the circular muscle.¹ The timing of these nonpropagating contractions is affected relatively little by enteric motor neural activity but is very dependent on the degree of wall distention. This nonpropagating activity also displays a circadian rhythm, being significantly reduced during sleep.

PROPAGATING MOTOR PATTERNS

When excitatory motor neurons are active, contractions evoked by MPOs summate, giving rise to powerful lumen-occlusive contractions that can last longer than slow waves and that can propagate substantial distances along the colon. These identifiable colonic motor patterns are commonly referred to as propagating sequences or high-amplitude propagating sequences (also termed propagating contractions or high-amplitude propagating contractions [HAPCs]). Depending upon their direction of propagation, these patterns can be further qualified by the terms antegrade (aboral) or retrograde (orad). In the healthy colon, antegrade propagating sequences are recorded with a three-fold higher frequency than retrograde propagating sequences.⁶ As with nonpropagating activity, propagating sequences display nocturnal suppression and can be stimulated by a meal.

RECTAL MOTOR COMPLEXES

Periodic contractile activity predominates in the sigmoid colon and rectum. This activity is commonly termed the rectal motor complex (RMC) or periodic rectal motor activity (PRMA). The mean RMC amplitude ranges from 15 to 60 mm Hg with a duration of 3 to 30 minutes.⁶ In contrast to all other colonic contractile patterns, the circadian trend for RMCs is reversed, i.e., the RMC is more prevalent during sleep, suggesting the relevance of the extrinsic neural control of this pattern. The relationship between the RMC and flow is still incompletely understood. RMCs can be triggered by propagating pressure waves from the proximal colon and by the arrival of stool or gas from the sigmoid colon,⁷ suggesting the RMC provides a braking mechanism to keep the rectum empty.

REGIONAL VARIATION OF PROPAGATING SEQUENCES

Contractile activity in the human colon demonstrates marked regional variation. For example, propagating pressure waves originate nearly four times as frequently in the proximal colon than in the distal colon (Fig. 98-5). The mean distance covered by antegrade pressure waves arising from the cecum is 50 cm, compared with only 20 cm for sequences originating in the descending colon. Still, pressure waves arising proximally generally do not propagate beyond the mid-colon (see Fig. 98-5). It is now clear that slower propagation rates favor the effective propulsion of contents. The conduction velocity of pressure waves increases as the waves migrate caudally. Indeed, such events often accelerate to the point of synchronicity, which arrests the progress of content moving ahead of the contractile front. In addition, nonpropagating (segmenting) pressure waves make up a higher proportion of activity in the distal colon than in more proximal regions.⁶ Thus, most motor activity in the distal colon functions to retard forward flow, thereby minimizing challenges to continence.

Figure 98-5. Regional variation in the frequency of initiation and extent of propagation of antegrade propagating sequences. The blue histogram (bottom) shows the distribution of antegrade propagating sequences grouped according to the site of origin. The green horizontal bars at the top show the mean extent of propagation by sequences originating at the same site. Note that propagating sequences originate significantly more frequently in the proximal than in the distal colon, and the extent of propagation is much greater for sequences originating in the proximal colon. The solid purple line (bottom) is proportional to the propagating sequence frequency shown at the bottom and indicates that the density of component pressure waves is highest in the mid-colon and lowest at the ends of the colon.

REGIONAL LINKAGE AMONG PROPAGATING SEQUENCES

In addition to the regional variation in propagating sequence frequency, these sequential motor patterns are linked in an organized spatiotemporal pattern.⁸ Many of these regionally linked propagating sequences also form series in which three or more consecutive propagating sequences demonstrate a regional shift in the same direction. Each propagating sequence in a linked series originates in either a more proximal or more distal colonic location. Although most single propagating sequences do not span the length the colon, collectively a series of linked propagating sequences can do so (Fig. 98-6), and it is likely that such linkage is important for the transport of content over great lengths of the colon. The mechanisms underlying regional linkage are yet to be determined.

Figure 98-6. Six-hour spatiotemporal map of colonic propagating sequences in a female healthy control subject. In this map each individual ridge represents an antegrade propagating sequence. The start of each ridge indicates the site of origin and the time the propagating sequence occurred. The length of the ridge indicates the extent of propagation. The shading within the ridge indicates the amplitude of the component pressure waves. The hatched white arrows link the site of origin of sequential propagating sequences. Although no single propagating sequence spans the entire colon, a linked series of propagating sequences can do so.

ROLE OF THE ILEOCECAL JUNCTION

In humans, the ileocecal junction regulates colonic filling and prevents coloileal reflux, thereby preventing contamination of the small bowel with colonic bacteria.⁹ In the fasting state, cecal filling is slow and erratic, and chyme is retained in the distal ileum for prolonged periods.¹⁰ The close physical link between the terminal ileum and the cecum by the ileocecal ligaments behaves functionally as a valve and is responsible in part for continence of the ileocecal junction. A specialized band of muscle forms a low-pressure tonic sphincter¹¹ and prominent 6 cycles-per-minute (cpm) phasic contractions contribute to the regulatory function of the ileocecal junction. Phasic and tonic activity are inhibited concurrently with episodic terminal ileal flow or distention of the ileum, and the tone of the ileocecal junction increases in response to cecal distention.¹¹

Phase III of the interdigestive motor cycle (IDMC) (or migrating myoelectric-motor complex [MMC]), a motor pattern that occurs every 90 to 120 minutes in the upper intestine during fasting (see Chapter 97), does not contribute to ileocecal transit, because it rarely reaches the terminal ileum in the human. Most ileal chyme, driven by ileal propagating contractions in synchrony with inhibition of phasic contractions of the ileocecal junction, enters the cecum in a pulsatile fashion within 90 minutes of a meal. Prolonged studies, over several hours, correlating ileocecal movement of isotope with intraluminal pressures show that 72% of episodes of ileocecal transport result from monophasic, ileal propagating pressure waves.⁹ Furthermore, 93% of cecal propagating pressure waves were temporally associated with episodes of cecal filling, a finding that suggests episodic cecal filling is one of the triggers for proximal colonic propagating contractions (Fig. 98-7).⁹

Figure 98-7. Propagating pressure wave sequences identified in the terminal ileum and proximal colon during prolonged combined scintigraphic and manometric recordings. The bottom left corner of the far right box shows a scintigraphic image of technetium sulfur colloid in the terminal ileum and ascending colon of a healthy control subject. The pink circles indicate the location of the recording side holes, each spaced at 7.5 cm. The green hatched lines indicate the regions from which the luminal flow was recorded. Four scintigraphic images have been selected to indicate flow across the ileocolonic junction (solid bars 1 and 2) and mid-ascending colon (solid bar 3). The black arrows correspond to the time (horizontal axis) of acquisition of each 10-second scintigraphic frame. Small blue arrowheads on the scintiscans indicate the location of the manometric side hole from which the corresponding pressure tracing was recorded. Corresponding with the scintigraphic frame at T = 0, a cecal pressure wave is recorded. This cecal pressure wave initiates an ascending colonic propagating sequence that was temporally associated with coloileal reflux (solid bar 1) and flow across the mid-ascending colon (solid bar 2). During the coloileal reflux, an ileal propagating sequence is initiated (hatched black arrow), and this ileal propagating sequence is temporally associated with antegrade flow across the ileocolonic junction (solid bar 3). The red circle on the scintiscan images T = 0 to T = 40 follows the direction of the retrograde flow from cecum to ileum (T = 0 and T =10 sec) and then antegrade flow from the ileum to cecum (T = 20 sec and T = 40 sec).

THE COLON AS A STORAGE ORGAN

The region of preferential storage of colonic content is not entirely settled. In 1902, Cannon proposed on the basis of radiologic observations that the proximal colon is the site of storage and mixing, whereas the distal colon acts as a conduit for expulsion. Subsequent studies, however, found no difference in the dwell time for radiopaque markers in the proximal, middle, and distal colon: roughly 11 hours in each. Composition of the diet influences regional transit and probably accounts for some of the discrepancies among studies. With a liquid diet, the ascending colon empties rapidly, within one to two hours, whereas the transverse colon retains isotope for 20 to 40 hours.¹² Solid diet retards transit through the cecum and ascending colon. With a mixed diet, particulate matter and liquids are stored in both the ascending and transverse colon.¹³

RELATIONSHIPS BETWEEN COLONIC MOTOR PATTERNS AND FLOW

Emptying of the proximal colon occurs more rapidly when wall tone is increased (e.g., by intraluminal fatty acids) than when the tone is low; the volume and consistency of the contents also affect the rate of emptying. Isotonic fluid infused into the proximal colon stimulates proximal colonic emptying, suggesting that distention, per se, can activate propulsive motor patterns. Irritant laxatives (which act by stimulating mucosal receptors) in the proximal colon, however, trigger propagating contractions much more reliably than distention alone.¹⁴ Hence, proximal colonic emptying is influenced by a combination of increased wall tone and the initiation of propagating contractions, probably under the influence of both chemical and mechanical factors.

Mass movements, first detected radiologically, are infrequent movements of stool over long distances. More often, movement of colonic content occurs in a stepwise manner over short distances and in both antegrade and retrograde directions.¹⁵ Studies combining manometry with radiography in animals and with high frame rate scintigraphy in humans have shown that 93% of all propagating sequences in the proximal colon, regardless of amplitude or polarity, are temporally associated with discrete movements of isotope-labeled colonic contents within the unprepared colon (Fig. 98-8).¹⁵ The strength of this pressure-flow relationship is region dependent, being stronger in the transverse colon than in the cecum and ascending colon.

Figure 98-8. Intracolonic pressure measurements and corresponding scintiscans that show a clear correlation between a propagating pressure wave sequence and discrete movement of colonic contents from the cecum to the sigmoid colon. This particular movement of contents was not associated with defecation or sensation. The oblique lines ending in vertical arrows (scintiscan images) correspond to the time of acquisition of each 15-second scintigraphic frame. Small arrowheads on the scintiscans indicate the location of the manometric side hole from which the corresponding pressure tracing was recorded. In the proximal colon and mid-colon (channels 2, 3, and 4 from the top), a close temporal relationship exists between movement of the isotope and the onset of the propagating pressure wave upstroke. When the pressure wave reaches the splenic flexure, however, the proximal descending colon is seen to expand to accommodate the isotope, consistent with loss of lumen occlusion at this region. The pressure waves in channels 5 and 6 do not appear to correspond to lumen-occluding contractions. Note also that propagating pressure-wave amplitudes in channels 3 and 4 are only 30 and 39 mm Hg, respectively, yet the motor pattern is clearly propulsive.

(From Cook IJ, Furukawa Y, Panagopoulos V, et al. Relationships between spatial patterns of colonic pressure and individual movements of content. Am J Physiol 2000;278:G329.)

Most episodic antegrade movements of colonic content, which are not associated with propagating sequences, can be attributed to repetitive, nonpropagating pressure waves (over short distances in either direction), but there remain a number of movements of content that currently cannot be attributed to identifiable changes in intraluminal pressure. This observation might reflect the occurrence of contractions at points remote from the recording sites. Alternatively, propulsion is sometimes caused by motor events that do not significantly affect intraluminal pressure, such as longitudinal muscle shortening, non–lumen-occluding circular muscle contractions, or alterations in regional wall tone. Retrograde movements occur frequently. About half of retrograde contractions follow immediately after an antegrade movement, indicating frequent reflux of content back into the region from which it had just moved. This subtle to-and-fro motion is likely to help maintain maximal absorption, retard colonic transit, and, therefore, reduce stool frequency.

In summary, outside of the immediate predefecatory phase (see later), the distal colon displays a combination of fewer propagating sequences, shorter extent of propagation, higher conduction velocity, and lower probability of content propulsion than is observed in the proximal colon. In addition, proportionally more nonpropagating (segmenting) pressure waves occur in the distal colon than proximally. Considered together, these features would be expected to result in retardation of flow into the distal sigmoid and rectum, thus minimizing challenges to continence while maximizing the mixing of content more proximally.

RECTAL FILLING, CAPACITANCE, AND ACCOMMODATION AND MOTILITY OF THE ANAL SPHINCTERS

When stool or gas enters the rectum, the rectal wall is stretched, thereby simultaneously activating an enteric descending inhibitory reflex that causes transient relaxation of the internal anal sphincter and an extrinsic reflex pathway that leads to a brief contraction of the external anal sphincter. The anorectal inhibitory reflex can be demonstrated and tested by balloon distention of the rectum, and its presence reflects the integrity of enteric neural pathways. For example, the rectoanal inhibitory reflex is absent in Hirschsprung’s disease, which is characterized by loss of enteric ganglia in the rectal myenteric plexus. In health, this reflex permits entry of a small amount of content into the upper anal canal, and continence is maintained by the reflexive contraction of the external anal sphincter. This sampling of content by sensory receptors in the proximal anal canal permits the distinction between solid or liquid stool and gas. Sampling reflexes of this kind occur many times each day in response to low-volume rectal distentions, are not registered consciously, and do not cause an urge to defecate.

A large-volume rectal distention causes an internal sphincter relaxation of longer duration, which is registered consciously and which necessitates extra voluntary contraction of the external anal sphincter to maintain continence while the person decides how best to deal with the intraluminal content (stool or gas). Suppression of the defecation urge at this time, together with receptive accommodation of the rectum (see later), results in temporary storage of stool or gas in the rectum or retrograde transport of the stool or gas back to the sigmoid colon. Although the rectum is usually empty, it has the capacity to temporarily store feces until convenient evacuation can be arranged. More-prolonged rectal storage is made possible by the ability of the rectum to accommodate an increasing volume without a corresponding increase in intrarectal pressure, in a manner similar to gastric fundic relaxation.¹⁷ This adaptive increase in rectal compliance, mediated by inhibitory nerves, is important for maintaining continence by permitting prolonged fecal storage without a constant urge to defecate. Such rectal distention also has negative feedback effects on the proximal bowel and inhibits gastric emptying, slows small bowel transit, reduces the frequency of proximal colonic propagating pressure waves, and delays colonic transit.¹⁸ Typically, rectal tone is increased following a meal. A pathologic reduction of rectal compliance, such as after pelvic radiotherapy, causes rectal urgency. Conversely, excessive compliance, as in megarectum, attenuates the urge to defecate.

ANORECTAL MOTILITY DURING DEFECATION

If the processes just described give rise to the urge to defecate and the social circumstances are appropriate, the full defecation process is activated. This process involves a combination of pelvic reflexes coordinated in the medulla and pons. Rectal distention by stool stimulates reflex-induced complete relaxation of the internal anal sphincter, and the stool moves into the upper anal canal, heightening the sense of urge. Postural changes and straining facilitate this process in several ways: Sitting or squatting causes descent of the anorectal junction, and straining produces further rectal descent. Both activities serve to increase the anorectal angle, thereby reducing resistance to outflow. At this point, if the person wishes to proceed to expel stool, the external anal sphincter is relaxed voluntarily. At the same time, the puborectalis muscle is relaxed (further increasing the anorectal angle); the levator ani muscles contract; the perineum descends further; and stool is funneled into the anal canal and expelled by increasing strain-induced, intrarectal pressure (Fig. 98-10). Once the expulsion phase has commenced, evacuation of stool can proceed in some cases without further straining, as a consequence of colonic contractions propagating toward the anus (see Fig. 98-9).¹⁶ Expulsion of stool is possible in response to strain alone without rectosigmoid contractions, although a contribution from increased rectal wall tone cannot be excluded.

Figure 98-10. Some of the mechanical processes that facilitate stool expulsion, as illustrated by sequential films of a simulated defecation of thickened barium during defecation proctography. A, The rectum at rest with a normal resting angle of approximately 90 degrees; the anal canal is closed. B, On straining, as the anterior rectal wall begins to flatten, the proximal anal canal begins to funnel as barium contrast is forced into it. C, As more pressure is exerted, the anterior rectal wall flattens further, contrast fills the anal canal, and evacuation begins. At this time, the puborectalis muscle and external anal sphincter are relaxing, resulting in the onset of descent of the rectoanal junction. At the same time, the levator ani muscles are activated and help control the descent of the rectoanal junction (note the posterior indentation resulting from contraction of the pubococcygeus muscle). D, The puborectalis is fully relaxed; this, in combination with vigorous straining, has resulted in nearly complete descent of the rectoanal junction. Note the position of the rectoanal junction, which in this frame is well below the horizontal pale artifact (due to the water-filled toilet seat), compared with that in the previous frame, in which the junction is level with this artifact. This descent has now opened up the anorectal angle, thereby further reducing the resistance to outflow through the anal canal. E, Rectal emptying continues, and anterior rectal compression is more obvious. F, After evacuation, the anorectal junction has ascended to its original position, and the anorectal angle has returned to its more acute resting angle.

(Courtesy of Prof. D. Z. Lubowski.)

PHYSIOLOGIC

Twenty-four hour recordings of myoelectric activity or intraluminal pressure show that colonic phasic and tonic activity predictably are increased one to two hours after a meal (the gastrocolonic response) and are markedly suppressed at night.¹⁹ The entire colon responds to the meal, with an increase in colonic wall tone, migratory long spike-bursts, and propagating and segmenting contractile patterns. A minimum caloric load of approximately 300 kcal is required to generate the colonic response to a meal, and a meal of only 200 kcal increases rectal muscle tone.²⁰ The meal response also is highly dependent on the fat content of the caloric load. For example, 600 kcal of fat induces the response, whereas an equicaloric load of protein or carbohydrate does not.

The mechanism of the colonic meal-response remains unclear, although it is known that neither the stomach nor the spinal cord needs to be intact to display the response. Non-nutrient gastric distention, by balloon or water, also can stimulate rectosigmoid motility, yielding a similar response to that following intraduodenal lipid infusion. Both of these responses are markedly attenuated by prior intravenous administration of the 5-hydroxytryptamine-3 (5-HT₃) receptor antagonist granisetron, which suggests that 5-HT₃ receptors on vagal afferents may be involved in the gastrocolic response.²¹ Cholecystokinin (CCK), which is released by fats and fatty acids in the duodenum, can replicate the gastrocolic response, but only at doses exceeding those occurring postprandially. The CCK-A antagonist loxiglumide blocks the effects of CCK on the colon but does not abolish the gastrocolic response, thus making CCK an unlikely mediator of the response.

Colonic myoelectric and pressure activities are profoundly suppressed at night.¹⁹ During stable sleep, colonic motility virtually ceases (except for the antipropulsive rectal motor complexes, which increase), thereby reducing the challenges to continence at a time when anal sphincter tone and awareness of colorectal sensations are minimal. If the subject shifts to a lighter level of sleep, even without actually awakening, an immediate increase occurs in propagating and nonpropagating pressure waves (Fig. 98-11). Forced awakening at night and spontaneous early-morning awakening both stimulate an immediate increase in colonic propagating pressure waves. This phenomenon clearly is linked with the readily identifiable habit of defecation soon after awakening in the morning and demonstrates the potential for profound modulation of colonic motor activity by the central nervous system.

Figure 98-11. Relationship of propagating pressure wave sequences to nocturnal arousals. An arousal represents a lightening of the level of sleep and need not necessarily culminate in awakening. This example demonstrates an arousal-induced event propagating from the proximal to the distal sigmoid colon and followed by another arousal-induced event propagating from the ascending colon to the sigmoid. Only the second arousal culminated in a brief period of wakefulness. Repetitive propagating sequences of this type also are seen on early-morning awakening.

(From Furukawa Y, Cook IJ, Panagopoulos V, et al: Relationship between sleep patterns and human colonic motor patterns. Gastroenterology 1994;107:1372.)

Stress and emotional factors long have been believed to influence colonic motility, but experimental evidence for this is conflicting, possibly because of a reliance on measurements from the distal colon, which might not be representative. In light of the profound waking-response, it is likely, but unproved, that stress does induce propagating pressure waves. Due to technical difficulties associated with trying to record physical activity and colonic motility simultaneously, data on the colonic response to physical activity are sparse; however, physical exercise, perhaps through increased sympathetic tone, decreases colonic motility.²² The colonic response to stress and exercise highlight the importance of the autonomic nervous system in modulating colonic function. Similarly, autonomic dysfunction, resulting from pelvic surgery, childbirth, or neural degradation, has been implicated in several colonic disorders including slow-transit constipation and irritable bowel syndrome (IBS).²³

PHARMACOLOGIC

Laxatives exert their diarrheal actions by increasing mucosal secretion or by stimulating colonic propulsive activity. For example, the irritant laxative bisacodyl and the bile acid chenodeoxycholic acid both stimulate high-amplitude colonic propagating pressure wave sequences, thereby leading to mass movements. Bisacodyl exerts its motor effect through mucosal afferent nerve fibers, because the response can be blocked by topical mucosal application of lidocaine. In addition to the local response, these agents, when administered rectally, can stimulate motor activity in the proximal colon, thereby indicating the existence of long reflex pathways between the rectum and proximal colon.

Colchicine, a natural alkaloid, is well known to cause diarrhea. Colchicine increases the frequency of spontaneous bowel movements and accelerates colonic transit in patients with chronic constipation²⁴; the mode of action is not yet clear but it has been shown to increase prostaglandin synthesis and to promote intestinal secretion, the latter mediated through cyclic AMP. In the rat, small intestinal colchicine stimulates myoelectric activity.

Lubiprostone, a type 2 chloride channel (ClC2) activator, is a member of a new class of compounds known as prostones. Activation of ClC2 increases intestinal chloride secretion resulting in increased intraluminal fluid accumulation, which accelerates intestinal transit, softens stools, and increases spontaneous stool frequency in patients with constipation.²⁵

Serotonin (5-HT) is an important mediator of bowel physiology, and both 5-HT₃ and 5-HT₄ receptors play a role in colonic peristalsis and transit. For example, the 5-HT₃ receptor antagonists granisetron and ondansetron blunt the gastrocolic response and delay colonic transit, respectively.²¹ Alosetron, another antagonist of the 5-HT₃ receptor, exerts a significant constipating affect by slowing colonic transit.²⁶ In contrast, 5-HT₄ agonists (e.g., tegaserod, prucalopride, renzapride), act on presynaptic receptors and facilitate release of acetylcholine and CGRP (calcitonin gene-related peptide), thereby inducing colonic propagating contractions and accelerating colonic transit. Although this class of drug shows promise for the treatment of constipation,²⁷ tegaserod, a 5-HT₄ agonist, was withdrawn from the market because of concern about associated adverse cardiovascular events (see Chapter 118). Other highly selective 5-HT₄ agonists, such as prucalopride, might be attractive options because they do not interact with 5-HT₃ or 5-HT_1B receptors, and prucalopride does improve stool frequency and symptoms in severe constipation.²⁸ Further trials with these agents are awaited.

Opiates are well known to have an antidiarrheal effect, but their mechanism of action is less clear. In the human colon, morphine increases phasic segmenting activity, reduces colonic tone, and attenuates the bowel’s response to a meal.²⁹ Opiates are known to inhibit presynaptic and postsynaptic enteric neural circuitry. The reduction in neurally dependent propagating contractions and the enhancement of myogenic mixing movements and fluid absorption contribute to the constipating effect of the drug. Specific constipation syndromes, such as opiate-induced constipation or postsurgical ileus, might respond to opiate antagonists such as methylnaltrexone and alvimopan (see Chapter 120).³⁰

Nitric oxide is a potent endogenous inhibitor of colonic propulsive activity and the human colon appears to be under a state of tonic nitrergic inhibition. For example, infusion of the nitric oxide synthase inhibitor, l-NMMA (N^G-monomethyl l-arginine), is a potent stimulator of colonic propagating contractions.³¹ Alternatively, segmental lengthening of the colon induced by the entry of content triggers nitric oxide release from descending pathways, which in turn inhibits colonic propulsive activity.³²

NONPHARMACOLOGIC

Probiotics are living organisms that, when ingested in adequate amounts, are claimed to exert a health benefit to the host. Relatively few rigorously designed studies have been conducted with probiotics but some strains have been shown to have a beneficial effect in IBS (see Chapter 118), ulcerative colitis (see Chapter 112), and diarrhea.³³ In the colon, probiotics are likely to modulate the inflammatory response through activation of signals with the epithelium and immune system.³³ Probiotics may well inflence colonic motility, but this has not been systematically evaluated.

Sacral nerve stimulation modulates the extrinsic nerves innervating the pelvic floor and colon. Electrical stimulation of the S3 sacral root induces a modest increase in external anal sphincter tone and has been used successfully in the management of fecal incontinence. Stimulation of the S3 root also induces propulsive activity throughout the entire colon and has been shown to increase stool frequency in patients with slow-transit constipation.³⁴ Randomized trials of this promising technique for treating slow-transit constipation are in progress; the precise mode of this action remains unknown. The substantial latency between stimulus and pelvic floor or colonic contractile responses is longer than would be expected via a polysynaptic efferent pathway, which suggests possible involvement of extrinsic neural pathways.³⁵ Magnetic stimulation of the sacral nerve S3 also shows promise in modulating colonic and anorectal function.³⁶ Because this approach is less invasive than electrical stimulation of sacral nerves, it may be a reasonable treatment option in children with colonic or anorectal dysfunction.

Acupuncture has been shown to have significant effects upon upper gastrointestinal tract disorders such as nausea and vomiting. Only two studies have evaluated its potential in constipation, one in children and one in adults.³⁷ Acupuncture improved stool frequency in children, but these results weren’t replicated in adults; this warrants further study. Acupuncture is known to activate neural, opioid, humoral, and serotoninergic pathways and potentially has a clinical role in treating disorders such as constipation and IBS.³⁷

Biofeedback has been shown to improve stool frequency and rectal evacuation in patients with pelvic floor dyssynergia, and the technique has been shown to accelerate colonic transit in this subset of patients with constipation (see Chapter 18).³⁸ The mode of action of biofeedback is not fully understood, but evidence suggests that extrinsic autonomic efferent pathways mediate the response.³⁹

CONSTIPATION

Intuitively, one would expect that constipation and diarrhea should be manifestations of hypomotility and hypermotility, respectively. Sometimes this is true, but in the distal colon, at least, the converse may be true. A paradoxical increase in nonpropagating (segmenting) contractions and myoelectrical short spike-bursts has been reported in the rectosigmoid region in constipated patients. Conversely, patients with diarrhea have hypomotility in this region. It is likely that segmenting activity retards forward flow, whereas suppression of such activity permits unrestricted access of stool to the rectum, where a defecatory urge is initiated. Thus, constipation can result from either infrequent or ineffective propagating pressure waves, or from an increase in segmenting distal colonic pressure waves, or perhaps both.

In severe slow-transit constipation, prolonged manometric studies have confirmed a reduction in the overall number of high-amplitude propagating pressure waves¹⁹; however, the overall number of propagating pressure waves of any magnitude is often normal or increased. Studies examining the spatiotemporal patterning of propagating sequences have revealed colonic regions in which activity is diminished or absent, particularly within the vicinity of the splenic flexure (Fig. 98-12).⁴⁰ Furthermore, there appears to be a loss of the normal linkage in patients who have sequential progressive systemic sclerosis and constipation.⁸ The underlying pathogenesis of severe slow-transit constipation is unclear, but changes in enteric excitatory motor innervation of the smooth muscle in patients with severe slow-transit constipation are likely to contribute to this disorder.⁴¹

Figure 98-12. Twenty-four-hour spatiotemporal maps of colonic propagating sequences in a female healthy control subject (A) and a female patient with slow-transit constipation (B). Within these maps, each individual ridge represents a propagating sequence. Antegrade propagating sequences (green) originate at the orad end of the ridge, and retrograde propagated sequences (red) originate at the anal end of the retrograde ridge. The start of each antegrade and retrograde ridge indicates the site of origin and the time of day the propagating sequence occurred. The length of the ridge indicates the extent of propagation. The shading within the ridge indicates the amplitude of the component pressure waves. In health (top), the maps indicate several physiologic characteristics of colonic propagating sequences. These include increased frequency and amplitude before defecation, immediately after morning awakening, and in response to a high-calorie meal. The map also indicates nocturnal suppression of these motor patterns. In contrast, the patient with severe constipation (B) demonstrates a clear ability to generate propagating sequences, but there are notable differences in the characteristics of these motor events compared with those of healthy control subjects. For example, the patient demonstrates an increased frequency of low-amplitude antegrade and retrograde propagating sequences in the proximal and distal colon and few propagating sequences spanning the mid-colon. This patient also demonstrates a lack of the normal colonic nocturnal suppression of propagating sequences and a lack of the normal colonic response to a high-calorie meal.

(Courtesy of Dr. P.G. Dinning, Kogarah, New South Wales.)

Constipation is fully discussed in Chapter 18.

DIARRHEA

Detailed scintigraphic studies in patients with diarrhea have shown the dominant feature to be early and rapid transit through the ascending and transverse colon.⁴² Normally, propagating sequences are more frequent in these proximal regions than elsewhere. Manometric data from the entire colon in patients with diarrhea would help explain these observations, but are lacking. A relative lack of distal colonic segmenting activity, perhaps in combination with increased proximal colonic propagating pressure waves, might explain this preferential acceleration of proximal colonic transit, but proof of this hypothesis is awaited. Diarrhea is fully discussed in Chapter 15.

IRRITABLE BOWEL SYNDROME

Although colonic transit generally is slower in constipation-predominant IBS and faster in diarrhea-predominant IBS, no colonic motor pattern is specific for IBS.⁴³ Exaggerated responses to stimuli such as meals, CCK, and mechanical stimuli have been reported, but a consistent disturbance has not emerged, probably because of the heterogeneity of the disease and the methodologies used for characterization. In addition, remarkably little study of the proximal colon in IBS has been conducted to date. At present, compelling evidence regarding the pathophysiology of IBS suggests a major contribution by afferent hypersensitivity, in addition to a variable alteration in colonic motor function. IBS is fully discussed in Chapter 118.

COLONIC MOTILITY DISTURBANCES SECONDARY TO NONMOTOR INTESTINAL DISORDERS

Altered motility secondary to underlying inflammation or a hormonal disturbance can contribute to the colonic symptoms of a nonmotor disease. The diarrhea of idiopathic inflammatory bowel disease, for example, results from a combination of enhanced secretion, reduced absorption, and altered colonic motor function. In ulcerative colitis, rectosigmoid-segmenting, nonpropagating pressure waves are diminished, whereas postprandial propagating pressure waves are increased.¹³ Rectal compliance also is reduced, and together, these effects can exacerbate diarrhea, as suggested by studies demonstrating rapid rectosigmoid transit in ulcerative colitis.¹³ The motility of the healthy colon also can be perturbed by ileal diseases. For example, exposure of the healthy proximal colon to supranormal concentrations of bile salts, such as from terminal ileal disease or resection, not only stimulates net colonic secretion but also initiates high-amplitude propagating pressure waves, thereby accelerating colonic transit.¹³