12 Testing for and the role of anal and rectal sensation JOHN ROGERS ANAL SENSATION Modalities of anal sensation The modalities of anal sensation can be precisely defined. Duthie and Gairns (1960) in a detailed examination of anorectal sensation tested for touch, pain and temperature sensation in normal subjects. Touch The anal canal is extremely sensitive to touch from the anal verge to a level between 0.25 and 0.75 cm above the anal valves. Although localization of the stimulus around the circumference of the canal cannot be made, the approximate level of the stimulus can be gauged by the subject. Pain Appreciation of pain by pin-prick is felt more distinctly in the anal canal compared with the perianal skin and is most acute in the region of the anal valves. The upper level of sensation to pain is between 0.5 and 1.5 cm above the anal valves, usually about 0.5 cm higher than the level for touch. Cold and heat These stimuli, like pain, are felt more distinctly in the anal canal compared with the perianal skin and the upper limit of sensation is also above the level of the anal valves (Table 1). Table 1. Upper limits of sensory discrimination. Modality Distance above anal valves Touch 0.25 to 0.75 cm Pain 0.5 to 1.5cm Temperature Level of anal valves BailliPre’s Clinical Gastroenterology- 179 Vol.6,No.l,March1992 Copyright 0 1992. by Baillike Tindall ISBN o-7020-1622-5 All rightsof reproduction in any form reserved
Transcript
12
Testing for and the role of anal and rectal sensation
JOHN ROGERS
ANAL SENSATION
Modalities of anal sensation
The modalities of anal sensation can be precisely defined. Duthie and Gairns (1960) in a detailed examination of anorectal sensation tested for touch, pain and temperature sensation in normal subjects.
Touch
The anal canal is extremely sensitive to touch from the anal verge to a level between 0.25 and 0.75 cm above the anal valves. Although localization of the stimulus around the circumference of the canal cannot be made, the approximate level of the stimulus can be gauged by the subject.
Pain
Appreciation of pain by pin-prick is felt more distinctly in the anal canal compared with the perianal skin and is most acute in the region of the anal valves. The upper level of sensation to pain is between 0.5 and 1.5 cm above the anal valves, usually about 0.5 cm higher than the level for touch.
Cold and heat
These stimuli, like pain, are felt more distinctly in the anal canal compared with the perianal skin and the upper limit of sensation is also above the level of the anal valves (Table 1).
Table 1. Upper limits of sensory discrimination.
Modality Distance above anal valves
Touch 0.25 to 0.75 cm Pain 0.5 to 1.5cm Temperature Level of anal valves
BailliPre’s Clinical Gastroenterology- 179 Vol.6,No.l,March1992 Copyright 0 1992. by Baillike Tindall ISBN o-7020-1622-5 All rightsof reproduction in any form reserved
180
Movement within the anal canal
J. ROGERS
Longitudinal and rotatory movement of an object in the anal canal can also be precisely defined.
Perianal skin has the same degree of sensitivity to these stimuli as the dorsum of the finger. In contrast, none of the above stimuli are appreciated by the rectal mucosa.
Receptors for anal sensation
In addition to defining the sensory modalities of the anal canal in normal subjects, Duthie and Gairns (1960) also described in detail the nature and distribution of specific sensory receptors in the anorectum following careful
Figure 1. Distribution of sensory nerve endings in the anal canal. From Duthie and Gairns (1960), with permission.
ANAL AND RECTAL SENSATION 181
histological examination of longitudinal post-mortem specimens of perianal skin, anal canal and rectum (Figure 1).
Perianal skin
The innervation of perianal skin resembles that of hairy skin found else- where on the body surface. Intraepithelial free nerve endings are present, but there are no organized nerve endings. There is profuse innervation of hair follicles by thick myelinated nerve fibres which, after division, form a cylindrical network around the sheath of the hair follicle.
Anal canal
The nature of the innervation and receptors of the anal canal varies with the exact region.
Anal margin. The skin of the anal margin contains only occasional hair follicles and the innervation is similar to perianal skin. Organized nerve endings similar to Krause end-bulbs which lie close to and touch hair follicles serve to distinguish this area.
Anal canal. The innervation of this region is richer than that of the anal margin, Some of the nerve bundles pass towards the epithelium before the nerve fibres separate and divide into fine branches to form free nerve endings below the epithelium.
Other nerves send their branches into the epithelium, where they ramify amongst the basal cells and terminate in discrete end-knobs. The numerous nerve bundles containing non-myelinated and myelinated fibres supply a profusion of encapsulated and uncapsulated organized nerve endings. Endings comprise Meissner’s corpuscles, Krause end-bulbs, Golgi-Mazzoni bodies, Genital corpuscles and Pacinian corpuscles (Table 2).
Touch Acute sensitivity to cold Tension and pressure Friction Tension and pressure
Meissner’s corpuscles are relatively infrequent, but typically lie in the dermal papillae close to the epithelial margin. They are less highly developed than those in the fingertip. They are believed to record touch sensation.
Krause end-bulbs are extremely numerous but vary somewhat in the size and calibre of their intrinsic fibres. They lie fairly deep in the dermis and are believed to respond to thermal stimuli; in particular they are acutely sensitive to cold.
182 J. ROGERS
Golgi-Mazzoni bodies are present in fairly large numbers and are characteristic of this region. They lie superficial to the Pacinian corpuscles and deep to the Meissner’s corpuscles or the Krause end-bulbs and have a prominent capsule. The long axis of the Golgi-Mazzoni endings is pre- dominantly parallel to the surface. They are thought to respond to changes in tension or pressure within the anal canal.
Genital corpuscles are also present in large numbers and are a main feature of this region. They are poorly encapsulated, roughly spherical in shape and frequently appear in clusters. The intrinsic branches to their supplying nerve fibres often pass outside the corpuscle into the surrounding tissue. They lie at varying depths in the dermis. The greatest number lie close to the epithelium in the dermal papillae and this suggests that they respond to friction along the anal canal.
Pacinian corpuscles are also found and are the deepest of all the receptor types. They appear to be closely related to the muscle bundles of the internal sphincter. Like Golgi-Mazzoni bodies, pacinian corpuscles respond to changes in tension and pressure.
In addition to the above named nerve endings, large diameter free nerve endings are also found in close relationship with the epithelium and are of a loosely coiled type. More complex coiled forms are seen in the transitional zone between the stratified squamous and the stratified columnar epi- thelium. Close to the region of the anal valves and crypts a large mass of compressed intertwining nerve fibres is often found filling the largest dermal papillae. These large endings are often found in adjacent papillae. The functions of these various nerve endings are likely to be as important as the eponymous forms as they are present in great numbers and are often served by large diameter nerve fibres and bundles, although no specific function can be ascribed to them at this time.
Anal crypts and valves. The endings in this region are similar to those of the anal canal. Free nerve endings now become the most numerous and the organized endings are in different proportions. There are fewer Golgi- Mazzoni bodies, but more genital corpuscles. The genital corpuscles are not only adjacent to the epithelial surface, but occur throughout the length of the crypts. The massive large diameter free nerve endings which occur in the dermal papillae of the anal canal are found in larger numbers than in any other region.
The nerve pathway for anal sensation
The nerve pathway for anal canal sensation is via the inferior haemorrhoidal branches of the pudendal nerve to the sacral roots of S2, S3 and S4. Gunterberg et al (1976) investigated the importance of the sacral root innervation of the anal canal in patients following radical pelvic surgery which involved unilateral or bilateral sacrifice of the sacral nerves. They demonstrated that the main sensory input was to the sacral roots below S2 and that at least one or both S3 nerves were required to subserve normal discriminatory function, continence and defecation.
ANAL AND RECTAL SENSATION 183
Testing for anal sensation
Tests of anal sensation may be qualitative or quantitative. The anal canal is able to perceive touch, pain, temperature and vibration, and qualitative testing of these sensory modalities is straightforward. Duthie and Gairns (1960) described the use of metal rods at different temperatures, von Frey hairs and pin prick applied to the anal mucosa through an anal speculum. Qualitative testing may be of use to confirm the presence or absence of normal sensory modalities in the anal canal with reference to anatomical structures such as the level of the anal valves, but is otherwise of little value.
Recently, two quantitative tests of anal sensation have been developed and assessed. Measurement of the sensitivity of the anal mucosa to electrical stimulation was originally described by Roe et al (1986a) and has been widely adopted. The technique involves the use of a specialized electrical generator the current from which can be steadily increased in 1 mA (Roe et al, 1986a) or 0.1 mA (Rogers et al, 1988a) increments independent of transient variations in tissue resistance. The output from this ‘constant current’ generator (Figure 2) is passed between two platinum wire electrodes 1 cm apart which are mounted on a 10 French gauge catheter graduated in centimetres (Figure 3). The usual protocol is to pass the catheter (lubricated with K-Y jelly) into the anal canal and position the electrodes in the upper, middle and lower thirds of the anal canal according to the functional length of the anal canal as measured by manometry. For anal sensation the generator is set to produce a 0.1 ms square wave stimulus
Figure 2. The constant current generator and bipolar electrode used for anal and rectal mucosal electrosensitivity.
184 J. ROGERS
Figure 3. Close up of the bipolar electrode used for mucosal electrosensitivity.
at a constant rate of 5 Hz. The current across the electrodes is increased in increments of 0.1 mA until the threshold of sensation is reported by the patient, usually as a burning or tingling sensation. Three measurements of the threshold are made at each site and the average reading or lowest reading recorded. Although it is not known which sensory fibres are stimulated, the technique has been shown to be a reliable and repeatable quantitative test of anal sensation (Rogers et al, 1989).
Quantitative measurement of anal temperature sensation was described by Miller et al (1987). Their technique involves the use of a specially constructed probe through which water at different temperatures is pumped while the temperature of the probe is constantly monitored. The probe is inserted into the upper, middle and lower thirds of the anal canal according to the functional length of the anal canal as measured by manometry. By using water pumped from three separate water baths thermostatically controlled at 32.5”C, 37°C and 41.5”C, they were able to determine the minimum detectable temperature change when the temperature of the probe was changed rapidly from normal to hot, hot to normal, normal to cold and cold to normal. They demonstrated that the anal canal was sensitive to changes in temperature.
The simplicity of the former technique and the general availability of the equipment make electrical stimulation the current gold standard in quanti- tative measurement of anal sensation.
RECTAL SENSATION
Modalities of rectal sensation
The modalities of rectal sensation are indistinct. In keeping with other parts of the gastrointestinal tract, with the exception of the anal canal, the rectum is insensitive to stimuli capable of producing pain and other sensations when applied to a somatic cutaneous surface. It is, however, sensitive to distension by a balloon introduced through the anus, though it is not known whether it
ANAL AND RECTAL SENSATION 185
is stretching or reflex contraction of the gut wall or the distortion of the mesentery and adjacent structures which is responsible for the sensation. Goligher and Hughes (1951) used balloon distension in a systematic investi- gation of rectal and colonic sensation. They demonstrated that distension of the bowel up to 1.5 cm from the anal verge caused a sensation of fullness in the rectum, giving the patient a desire to pass wind or motion (rectal-type sensation). Distension above this level produced a purely abdominal sensation, referred to the suprapubic or left iliac region and simulating a ‘wind-pain’ or intestinal colic (colonic-type sensation). An important distinction was that the pressure in the intraluminal balloon required to produce a rectal-type sensation was on average 20mmHg less than that required to produce a colonic-type sensation, suggesting that the rectum is more sensitive to this stimulus than the colon.
Receptors for rectal sensation
Duthie and Gairns (1960) were unable to define any specific receptor type on histological examination of the rectum in humans. The large numbers of free and organized nerve endings seen in the upper anal canal mucosa cease abruptly with the conversion to rectal mucosa. In multiple sections from ten post-mortem specimens they only found one organized nerve ending in the rectal mucosa adjacent to the transitional zone with the anal canal, and its supply came from the same leash of nerve fibres which supplied the organized endings of the adjacent anal canal. Myelinated and non- myelinated nerve fibres are seen adjacent to the rectal mucosa, but no intraepithelial fibres arise from these. Nerve fibres are also seen in associa- tion with blood vessels and submucous and myenteric plexuses, but no sensory function can be ascribed to these.
Possible mechanisms of rectal sensation
The absence of specific receptors in the rectal mucosa would account for the poor discriminatory quality of rectal sensation. An alternative explanation is that rectal-type sensation arises from stimulation of nerve endings and receptors in adjacent pelvic structures and the pelvic floor musculature. Evidence for this arises from the results of studies of anorectal function following surgery to the anorectum. Goligher and Hughes (1951) reported that rectal-type sensation was present to a reduced degree following distension of the colon onto adjoining pelvic structures in patients who had had sphincter-saving resections. Further evidence for this comes from Lane and Parks (1977) who investigated 12 patients following coloanal anasto- mosis. They demonstrated that all patients appreciated their normal feeling of perineal fullness and sensation of impending evacuation during balloon distension, and that the threshold of sensation did not alter over time. They also showed that sphincter inhibitory reflexes were preserved. Stephens and Durham-Smith (1971) observed in children who had supralevator and intermediate rectal deformities operated upon by drawing the upper rectum or colon through the puborectalis sling and anastomosing it to perianal skin
186 J. ROGERS
were capable of accurately discriminating the nature of neorectal contents. As the anal canal and anal sphincters were absent it suggests that receptors within the puborectalis may subserve a sensory function. Stretch receptors have been demonstrated in the levator ani (Winckler, 1958) and in the external anal sphincter in humans (Walls, 1959).
The nerve pathways for rectal sensation
The nerve supply to the rectum and sigmoid colon is derived from the sympathetic and parasympathetic nervous systems, and from inferior haemorrhoidal branches of the pudendal nerves. According to Gask and Ross (1937), sympathetic fibres arise from the lower six thoracic and upper two lumbar segments of the spinal cord. The ganglion cells are situated at the level of Ll and L2. The parasympathetic supply arises from the second, third and fourth sacral segments of the cord. The afferent impulses from the rectum and distal colon must travel by one or both of these pathways. Their relative importance was investigated using rectal and colonic distension following blockade of each of these routes. Goligher and Hughes (1951) showed that: following inferior haemorrhoidal nerve block by local anaesthesia the pressure threshold to induce colonic and rectal sensation was slightly reduced; following spinal anaesthesia, which successfully blocked the parasympathetic supply, colonic-type sensation was unaffected but rectal-type sensation was completely abolished; and following bilateral sympathectomy rectal-type sensation was preserved. This would suggest that the sensation of rectal distension travels with the parasympathetic system to S2, S3 and S4, whereas the sensation of colonic distension travels by the sympathetic system.
Testing for rectal sensation
The two commonly used methods for quantifying rectal sensation are rectal balloon distension and mucosal electrosensitivity. A variety of techniques have been described for rectal balloon distension. The techniques vary greatly according to the type of balloon used, and whether air or water is used to fill them. Balloons may be inflated in small or large increments or constantly filled by a pump. In addition, the definition of the thresholds of rectal sensation are imprecise. These factors make accurate and repeatable quantitative measurement of rectal sensation difficult and comparison between different centres impossible.
The usual method is to use either a condom (Farthing and Lennard-Jones, 1978) or party balloon (Rogers et al, 1988a) attached to a catheter which is inserted through the anal canal into the rectum. It is important that the balloon is placed above the anorectal junction to avoid sensory perception to distension being mediated by receptors in the anal canal. The balloon is usually filled with air or water by a hand-held syringe until certain thresholds of sensation are perceived by the patient. The volumes at which these thresholds are reached are then recorded. When a hand-held syringe is used it is important that the incremental volumes and intervals between incremental
ANAL AND RECTAL SENSATION 187
filling are kept constant. If water is used to inflate the balloon it must be drawn from a water bath heated to 37°C so that the subject does not confuse the stimulus of distension with that of temperature change. Three sensory thresholds are usually defined, although the exact definitions may vary between different centres:
Constant sensation of fullness is the threshold volume at which subjects feel a constant sensation of rectal distension over the whole interval between the filling increments. Sensation of rectal distension will occur before this but it will not remain as a constant sensation, due to accommodation of the rectum and adaptation of the sensory receptors to the stimulus.
Urge to defecate is the threshold volume at which subjects feel a sensation similar to that associated with their normal urge to defecate.
Maximum tolerated volume is the threshold volume at which subjects find rectal distension intolerable.
During rectal balloon distension the pressure in the balloon may also be monitored (Roe et al, 1986b; Varma and Smith, 1986). This permits an assessment of the compliance of the rectum, which can be a useful addition to the measurement of the sensory threshold volumes. Of the three levels of rectal sensation described, the maximum tolerated volume is the easiest to define and measure. It has also been shown to be highly repeatable (Varma and Smith, 1986).
Mucosal electrosensitivity , as described for the quantitative assessment of anal sensation above, may be used in the quantitative assessment of rectal sensation. Not surprisingly the current thresholds required to produce the sensation of electrostimulation are much higher because of the paucity of sensory endings in rectal mucosa. The optimum output characteristics required of the signal from the pulse generator are different for rectal sensation compared with anal sensation. J. Rogers et al (unpublished data) have found that a 0.3 ms square wave stimulus at a constant rate of 50 Hz gives consistent reproducible results.
THE ROLE OF ANAL AND RECTAL SENSATION
The sampling reflex
The upper limit of sensitive anal canal mucosa extends to a point 0.3 to 1.5 cm above the level of the anal valves, a mean distance of 3.1 cm from the anal verge (Duthie and Bennett, 1963). These authors postulated that if rectal contents could make contact with this sensitive part of the anal canal, the nature of the rectal contents may be discriminated, and that this sensory information would contribute to the maintenance of continence. In their experiment they measured the length of the functional anal sphincter and compared this with the length of the sensory zone of the anal canal in normal subjects. They were able to demonstrate that at rest the sphincteric zone was longer, preventing the rectal contents from encroaching on the sensitive anal mucosa. However, on balloon distension of the rectum a reflex decrease in the pressure of the anal sphincter permitted the intrarectal pressure to
188 J. ROGERS
extend downward to contact the sensory part of the anal canal. The reduction in the functional length of the canal extended to a point up to 2 cm below the upper limit of the sensory zone in some subjects, thus allowing rectal contents to come into contact with the sensitive anal mucosa. Anal sphincter relaxation in response to rectal distension was brisk and occurred mainly in the upper part of the anal canal; after the initial fall in anal canal pressure the sphincter pressure gradually returned to normal. This they termed the ‘sampling’ reflex. Anorectal sampling (equalization of rectal and upper anal canal pressures) occurring spontaneously as a normal physio- logical process has been recently confirmed using ambulatory manometry (Miller et al, 1988).
Sensation and continence
The importance of anal sensation in the continence mechanism remains controversial. For example, Read and Read (1982) showed no impairment of continence following application of local anaesthetic to the anal canal mucosa, and Keighley et al (1987) showed no functional impairment follow- ing the excision of the anal mucosa involved in the sampling reflex during proctocolectomy for ulcerative colitis. However, it should be pointed out that in the former study subjects were normal volunteers with normal anorectal motor function and the internal anal sphincter pressure was increased following the application of the local anaesthetic. In the latter study, despite there being no functional impairment, there was objective evidence of sensory impairment in the upper anal canal. These findings suggest that sensory impairment in the anal canal may lead to functional disturbance only when anorectal motor function is also compromised. This would certainly appear to be the case for patients with idiopathic faecal incontinence in whom abnormalities in motor function are well docu- mented. In addition, Rogers et al (1988a) have demonstrated a combined sensory and motor deficit in these patients. A pure sensory deficit has also been implicated by Lubowski and Nicholls (1988), who reported a small series of patients with faecal incontinence in whom no objective abnormality of motor function could be found but who all had significantly higher sensory thresholds to balloon distension of the rectum. In addition, Buser and Miner (1986) have demonstrated that delayed detection of balloon distension of the rectum in some patients with faecal incontinence can be improved by retraining techniques, which can result in the elimination of faecal inconti- nence and an improvement in sensory thresholds.
Discrimination between flatus, liquid and solid faeces
The precise mechanism by which the anorectum distinguishes the nature of rectal contents before evacuation is unknown. However, it is reasonable to assume that both the motor and sensory components of anorectal conti- nence play a part. The ‘sampling’ reflex is undoubtedly involved. The relative importance of each of the different and specific sensory modalities
ANAL AND RECTAL SENSATlON 189
present in the anal canal and their involvement in the mechanism of sensory perception of flatus, liquid and solid faeces remains speculative.
Recently it has been suggested that sensitivity to temperature change may be an important factor in the discrimination between flatus, liquid and solid faeces. Miller et al (1987) quantified temperature sensibility in the anal canal and rectum of normal subjects and patients with idiopathic faecal inconti- nence. They measured the smallest change in temperature that could be correctly detected in the rectum and in the upper, middle and lower thirds of the anal canal. In view of the absence of sensory nerve endings in the rectum it is not surprising that they found the rectum insensitive to temperature changes even in normal subjects. However, all regions of the anal canal were very sensitive to temperature changes and normal subjects could detect changes in temperature as little as O.@C in the lower third of the anal canal and l.l”C in the upper third of the anal canal. The acute awareness to temperature change in the anal canal led them to suggest that temperature sensation may aid discrimination between flatus, liquid and solid by detecting differences in the thermal capacity of rectal contents. They also demonstrated a significantly reduced sensitivity to temperature change in the anal canal of patients with faecal incontinence, and suggested that the reduced temperature sensitivity may have a role in the pathogenesis of faecal incontinence.
Their hypothesis of the role of temperature sensation in the discrimination of flatus, liquid and solid was challenged on theoretical grounds. Rogers (1987) pointed out that for temperature sensation to be of any physiological importance a temperature gradient must exist between the rectum and anal canal, and that the magnitude of the temperature gradient must be greater than the smallest temperature change detectable by the anal canal. Rogers et al (1988b) showed that the largest temperature gradient between the rectum and that part of the anal canalinvolved in the sampling reflex was only O.l3”C, with 95% confidence intervals of 0.06 to O.lYC. In other words, the temperature gradient between the rectum and anal canal is four times less than the temperature threshold necessary to produce a conscious appreci- ation of temperature change in the anal canal of normal subjects. Therefore, under normal physiological circumstances the conscious appreciation of the temperature of faecal material passing from the rectum to the anal canal is impossible during the anorectal sampling reflex.
Conscious discrimination of the nature of rectal contents as flatus, liquid or solid is probably dependent on the integration of information from all the sensory receptor types present in the anal canal mucosa, those within the muscles of the anal sphincters and pelvic floor, and those mediating rectal- type sensation.
SUMMARY
The rectum is insensitive to stimuli capable of causing pain and other sensations when applied to a somatic cutaneous surface. It is, however, sensitive to distension by an experimental balloon introduced through the
190 J. ROGERS
anus, though it is not known whether it is the stretching or reflex contraction of the gut wall, or the distortion of the mesentery and adjacent structures which induces the sensation. No specific sensory receptors are seen on careful histological examination of the rectum in humans. However, myelinated and non-myelinated nerve fibres are seen adjacent to the rectal mucosa, but no intraepithelial fibres arise from these. The sensation of rectal distension travels with the parasympathetic system to S2, S3 and S4. The two main methods for quantifying rectal sensation are rectal balloon distension and mucosal electrosensitivity. The balloon is progressively distended until particular sensations are perceived by the patient. The volumes at which these sensations are perceived are recorded. Three sensory thresholds are usually defined: constant sensation of fullness, urge to defecate, and maximum tolerated volume.
The modalities of anal sensation can be precisely defined. Touch, pain and temperature sensation exist in normal subjects. There is profuse innervation of the anal canal with a variety of specialized sensory nerve endings: Meissner’s corpuscles which record touch sensation, Krause end-bulbs which respond to thermal stimuli, Golgi-Mazzoni bodies and pacinian corpuscles which respond to changes in tension and pressure, and genital corpuscles which respond to friction. In addition, there are large diameter free nerve endings within the epithelium. The nerve pathway for anal canal sensation is via the inferior haemorrhoidal branches of the pudendal nerve to the sacral roots of S2, S3 and S4. Anal sensation may be quantitatively measured in response to electrical stimulation. The technique involves the use of a specialized constant current generator and bipolar electrode probe inserted in the anal canal. The equipment is generally available and the technique has been shown to be an accurate and repeatable quantitative test of anal sensation.
REFERENCES
Buser WD & Miner PB (1986) Delayed rectal sensation with fecal incontinence. Successful treatment using anorectal manometry. Gustroenterology 91: 1186-1191.
Duthie HL & Bennett RC (1963) The relation of sensation in the anal canal to the functional anal sphincter: a possible factor in anal continence. Gut 4: 179-182.
Duthie HL & Gairns FW (1960) Sensory nerve endings and sensation in the anal region of man. British Journal of Surgery 47: 585-595.
Farthing MJG & Lennard-Jones JE (1978) Sensibility of the rectum to distension and the anorectal distension reflex in ulcerative colitis. Gut 19: 6469.
Gask GE & Ross JP (1937) The Surgery of the Sympathetic Nervous System, 2nd edn. London: Bail&e.
Goligher JC & Hughes ESR (1951) Sensibility of the rectum and colon. Its role in the mechanism of anal continence. Lancet i: 543-547.
Gunterberg B, Kewenter J, Petersen I & Stener B (1976) Anorectal function after major resections of the sacrum with bilateral or unilateral sacrifice of sacral nerves. British Journal of Surgery 63: 546554.
Keighley MRB, Winslet MC, Yoshioka K & Lightwood R (1987) Discrimination is not impaired by excision of anal transition zone after restorative proctocolectomy. British Journal of Surgery 74: 1118-1121.
Lane RHS &Parks AG (1977) Function of the anal sphincters following cola-anal anastomosis. British Journal of Surgery 64: 59&599.
ANAL AND RECTAL SENSATION 191
Lubowski DZ & Nicholls RJ (1988) Faecal incontinence associated with reduced pelvic sensation. British Journal of Surgery 75: 10861088.
Miller R, Bartolo DCC, Cervero F & Mortensen NJMcC (1987) Anorectal temperature sensation: a comparison of normal and incontinent patients. British Journal ofSurgery 74: 511-515.
Miller R, Lewis GT, Bartolo DCC, Cervero F & Mortensen NJMcC (1988) Sensory discrimi- nation and dynamic activity in the anorectum: evidence using a new ambulatory technique. British Journal of Surgery 75: 1003-1007.
Read MG & Read NW (1982) Role of anorectal sensation in preserving continence. Gut 23: 345-341.
Roe AM, Bartolo DCC & Mortensen NJMcC (1986a) New method for assessment of anal sensation in various anorectal disorders. British Journal of Surgery 73: 310-312.
Roe AM, Bartolo DCC & Mortensen NJMcC (1986b) Diagnosis and management of intract- able constipation. British Journal of Surgery 73: 854-861.
Rogers J (1987) Anorectal temperature sensation: a comparison of normal and incontinent patients. British Journal of Surgery 74: 1189 (letter).
Rogers J, Henry MM & Misiewicz JJ (1988a) Combined sensory and motor deficit in primary neuropathic faecal incontinence. Gut 29: 5-9.
Rogers J, Hayward MP, Henry MM & Misiewicz JJ (1988b) Temperature gradient between the rectum and anal canal: evidence against the role of temperature sensation as a sensory modality in the anal canal of normal subjects. British Journal of Surgery 75: 1083-1085..
Stephens FD & Durham-Smith E (1971) Ano-rectal Malformations in Children. Chicago: Year Book Medical Publishers.
Varma JS & Smith AN (1986) Reproducibility of the proctometrogram. Gut 27: 288-292. Walls EW (1959) Recent observations on the anatomy of the anal canal. Proceedings of the
Royal Society of Medicine 52(supplement): 85-87. Winckler G (1958) Remarques sur la morphologie et l’innervation du muscle releveur de l’anus.
Archives d’dnatomie, d’Histologie et d’Ernbryologie 41: 77-95.
