Gilt, 1987, 28, 122 l1233

Pressure and prolapse the cause of solitary rectalulcerationN R WOMACK, N S WILLIAMS, J H M HOLMFIELD, ANDJ F B MORRISON

From the Sturgicail Unit, the London Hospital, Whitechapel, London, the University Depairtmeit of Surgery,the General Infirmary at Leeds, Leeds, and the Dep(artment of Physiology, the University of Leeds, Leeds

SUMMARY The cause of solitary rectal ulceration has been investigated using a method thatradiologically visualises rectal voiding whilst simultaneously measuring intrarectal pressure andexternal anal sphincter electromyographic activity. Control subjects and patients with the solitaryrectal ulcer syndrome, both with and without mucosal ulceration, have been studied. A highincidence of rectal prolapse (94%) was present in the patients who voided. Overactivity of the analsphincter during evacuation contributed to the fact that patients with mucosal ulceration requiredhigher intrarectal pressures to void than the controls and the patients without mucosal ulceration.The results indicate that a combination of rectal prolapse and a high voiding pressure may act tocause the mucosal ulceration in this syndrome by exposing the rectal wall to a high transmuralpressure gradient.

It is suggested that the solitary rectal ulcer syndrome(SRUS) forms part of a spectrum of disorders (the'mucosal prolapse syndrome') that is characterisedby histological changes in the rectal mucosa,' and iscaused by prolapse of the rectal mucosa.' A relation-ship between solitary rectal ulceration and rectalprolapse, which is often clinically occult, has beenwell demonstrated" but if the mucosal prolapsetheory of the aetiology of SRUS is to be accepted it isnecessary to be able to explain why only a smallpercentage of patients with rectal prolapse ulcerate,and the majority do not. It may be that other factorsact in combination with the prolapse to cause themucosal damage seen in patients with ulceration.The objectives of this study were, therefore, to test

the mucosal prolapse theory of the aetiology of theSRUS, and should it prove correct, to identify theadditional factors responsible for the mucosal ulcer-ation. To do this we have investigated the anorectalfunction of control subjects and patients with adiagnosis of SRUS confirmed by rectal biopsy, usinga technique that radiologically visualises the

Address for corrcsp) oldiclcc: i'rofcssor N S Willlid s I- R(F C. Surgic.ill lilitLondlonl l-iospit.II fldonEfl 1 1 3i13

Rcccivcdi or publici.tionl 19 Mircvh 1987

anorectum during voiding, whilst simultaneouslymeasuring intrarectal pressure and anal sphincterEMG activity.

Methods

SRUS PATIENTS AND CONTROI. SUBJECTSEighteen patients with SRUS (12 women, six men,age 30 years (19-64) (median and range)) confirmedby the presence of the characteristic changes of thesyndrome on rectal biopsy and nine control subjects(six women, three men, age 44 years (35-52)) wereinvestigated. Eleven of the patients (SRUS-U, sevenfemale, four male, 27 years (19-59)) had ulceration,seen on sigmoidoscopy done in the week before theinvestigation, in addition to the biopsy changes.Seven patients (SRUS-NU, five women, two men, 46years (22-64)) had an abnormal biopsy but noulceration on sigmoidoscopy. The patients withulcers were significantly younger than those withoutulceration (p<0-01) and the control subjects(p<0-01). The latter, because of a restriction placedupon the selection of control subjects of childbearing age by the Ethical Committee of TheGeneral Infirmary at Leeds on approval of this studyin March 1985.

1228

on July 12, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.28.10.1228 on 1 October 1987. D

ownloaded from

Presslsre and prolapse - the caulse of'solitarv rec(tal ulceration1

Table 1 Cliniicalfeatlures oftie SR US patients

SRUS-U SRUS-NU 1) ('isher)

Rectal bleeding 10(. 71"% NSRectal paiin 910" 28"0/. <()-(0Rcpeated straining at stool 82"%, 43" NSTenesmus 73% 57"/% NSDifficulty initiating dcfccation 82X'°/ 28%3 <005tSelf digitation 73" 43"0/. NSAwarc of prolapsc at prcscntattion 45"/. 14'S NS

The patients' symptoms are detailed in Table 1.Rectal bleeding was a common symptom and was thereason for presentation in five of the SRUS-NUpatients and six of the SRUS-U patients. Pain in theperineum or lower abdomen was more common in

the SRUS-U patients and was the reason for pre-

sentation in two of them. Repeated straining at stoolwas also common in both groups. It was related totenesmus in the SRUS-NU patients as they were ableto initiate defecation without difficulty at the first callto stool. This was different from the SRUS-U patientswho commonly could not initiate defecation, withoutdifficulty, on any occasion. Nearly half of theSRUS-U patients admitted being aware of some

degree of rectal prolapse on direct questioning,although it was the reason for presentation in onlyone patient from each group.The control subjects were volunteers free from

anorectal symptoms who had been admitted to theinfirmary for minor surgical procedures.Our technique for the dynamic assessment of

anorectal function has been described in detailelsewhere.7 Briefly it entails delineation of therectum by 300 ml of a semisolid contrast mediummade from barium sulphate, porridge oats and water.The contrast medium also simulated a soft stool to bevoided by the subject. The intrarectal pressure was

measured via a pressure sensitive radiotelemetrycapsule placed within the rectum. The electromyo-graphic activity of the puborectalis and superficialparts of the external anal sphincter was measured bythe insertion of fine stainless steel wire electrodes. Agraphic display of the pressure and EMG data was

produced by a BBC B microcomputer. This image

was mixed with the video output of the imageintensifier and the composite picture, containing theradiological, pressure and EMG data was stored on

videotape. Lateral flouroscopy of the anorectum was

done with the patient seated on a commode. Thepatient first maximally contracted the sphinctermuscles, and then attempted to void the contrastmedium.Measurements of radiation exposure were made

using thermoluminescent dosemeters. With expo-

sure factors for flouroscopy of 90- 110 kVp, 2-3 mAand screening times of between one and two minutesskin exposure was measured at 1-4 rads. From thesedata the total body dose was calculated to be 30- i 20mrem, and the dose to the gonads 5(M-200 mrem.

STATISTICAI. ANAL YSISFisher's exact test has been used for the analysis ofdifferences in the patients' symptomatology. Theother data were not normally distributed and valueshave been quoted as median and range. The MannWhitney U test has been used for the analysis ofdifferences between groups.

Results

Voiding was achieved in all control subjects and in 16(8911) of the SRUS patients. The radiological abnor-malities revealed on proctography are listed inTable 2. The incidence of lesions involving prolapseof the rectal mucosa into the anal canal was 94'%0 inthe SRUS patients who voided. The incidence of thetypes of lesion shown was similar in the two subgroupsof SRUS patients. Figure 1 shows an example of anintra-anal intussusception of the rectum, this was themost common lesion seen in both of the groups ofSRUS patients.

During voiding in both SRUS and control subjectsthe puborectalis and superficial parts of the externalanal sphincter always behaved in a similar manner,either increasing or decreasing in activity together, ifnot by exactly the same degree. Expressing sphincteractivity as the mean of the puborectalis and super-ficialis activities voiding in the control subjects wasassociated with inhibition of the sphincters to medianvalue of 25%'o of the resting level of activity (Fig. 2).Inhibition occurred with the onset of straining in fourof the subjects and after a period of increasedsphincter activity in the others (Fig. 3a). In the SRUSpatients voiding was accompanied by increased

Table 2 Radiological abnormlalities on)proctograph.v,

(Controls SRUS S?US- U SRUS-NU

n 9 18 11 7Voldcd ') 16 9 7Rlidiological

abnormaliticsAntcrior rcctal wall

prolapsc O 2Initrat-anal

intussusccption (O 8 3Full thickncss rcctal

prolapsc O) 4 1Rectococic I O) O OWith prolapsc OS () 94 89 1(00

I229'

on July 12, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.28.10.1228 on 1 October 1987. D

ownloaded from

Womack, Williams, Hol/rnfield, and Morrison

| _ _ l | | |- -- b _ -

Fig. 1 Initra-anal intussusception, the head ofthe intussusception (*) has eitered the anal canal (-) during voiding, therehtas beenI soOm'e recruitment oJ'anal sphincter EM(G activity. (Upper trace, initrarectalpressuire, bar= 75 mmHg; middle trace,p,uborectalis EMGIactivity; lower trace, superficial extern(tlanal sphincter EMG activit.y.)

activity in the sphincters that persisted throughoutvoiding (Fig. 3b). This increased activity, with a

mediain value of 24% of the activity of a maximalvoluntary contrac-;on, was significantly different tothe activity in the control subjects (p<0(02).

In the SRUS-U patients voiding was accompaniedby overactivity of the sphincters to a median level of30%, of the activity of a maximal voluntary contrac-tion. This wiIs significantly different from thebehaiviour of the sphincters in the controls (p<0-0l)and the SRUS-NU patients (p<()05), in whomthe mediain level of activity during voiding, was

unchanged from the resting level of activity (Fig. 2).The rise in intrarectal pressure measured during

voiding, or maximal attempts to void, was signific-antly higher in the SRUS patients (70 mmHg(45- 155)) than in the controls (44 (37-103), p<0-02).The highest voiding pressures were recorded in the

SRUS-U patients (74 (53-155)), their voiding pres-sures being significantly greater than the controls(p<(-01) and the SRUS-NU patients (48 (45-70),p<0-01). In the 13 SRUS patients (nine SRUS-U,four SRUS-NU) in whom sphincter activity increasedor remained unchanged during voiding (Fig. 2) therewas a significant inverse correlation (Rs=0-63,p<0.05) between the age of the patients and thepressure required to void contrast and a significantdirect correlation (Rs=0 68, p<0).01) between thepressure required for voiding and the activity of theanal sphincter.

Discussion

Before discussing the results of these investigationsthe definition of rectal prolapse needs to be con-sidered. The generally accepted definition is a con-

1 230)

on July 12, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.28.10.1228 on 1 October 1987. D

ownloaded from

Pressure and prolapse - the caluse ofsolitary rectal ullceration 3

Pu borectalis200'/o r

100%/..

Recruitment

Inhibition

100*/.

.*

0

a

aO-

a

............. ..... .....|

_

:us

* *

* a

-U *-_ .

* -

.

Controls SRUS-U SRUS-NU

Maximumvoluntaryactivity

RPctinn

Superficial externalanal sphincter

. *

0

--

IA..1LwIl1 y ...............

activity

No activity L .

* a*

o..

M... _100°/.Controls SRUS-U SRUS-NU

*v Controls p<001, v SRUS-NU p<005, Mann-Whitney u test

Fig. 2 Anal sphincter activity during voiding. (U patient voided; O patientfailed to void.)

dition where part or all of the rectal wall protrudesthrough the anal orifice.' The introduction of tech-niques involving radiological visualisation of rectalvoiding, as used in this study, however, has necessi-

tated a reappraisal of the definition of rectal pro-lapse, since degrees of prolapse which cannot bediagnosed clinically become apparent with thesetechniques. We have, therefore, considered anycondition in which rectal mucosa enters the analcanal to be one of rectal prolapse because, forreasons explained below, these apparently minor

abnormalities may have detrimental effects on theprolapsing mucosa and be the cause of much dis-comfort to the patient. Using this definition of rectalprolapse the results of this study support the 'mucosalprolapse syndrome' hypothesis of the aetiology of thesolitary rectal ulcer syndrome as defecographyrevealed rectal prolapse in 94% of the SRUS patientswho voided.The presence of rectal prolapse in all the SRUS-

NU patients studied implies that rectal prolapse, atleast when acting alone, is not the cause of mucosalulceration in this syndrome. The results suggest thatit is the combination of rectal prolapse and a highvoiding pressure, caused in most cases by overactivityof the external anal sphincter during voiding, thatleads to mucosal ulceration. It would seem, there-fore, that a mechanical explanation for the ulcerationis most likely. Previous authors (reporting over-

activity of the puborectalis muscle during straining,

in the SRUS) have explained the ulceration in termsof abrasion of the prolapsed mucosa by the unyield-ing puborectalis muscle bar," or as a result ofischaemia in the prolapse caused by its compressionwithin the anal canal."' Both of these hypothesespropound that the damage to the prolapsed mucosa

occurs from without, as a result of muscle over-

activity, and, therefore, they cannot satisfactorilyexplain the presence of ulceration in those patientswhere the sphincters relax during voiding. Wesuggest an alternative mechanism whereby the dis-ruptive force acts from within the prolapsed mucosa

and is not entirely dependent upon the activity of thesphincters during voiding.To consider the normal situation first, during

voiding the rectal mucosa is supported by the rectalcontents which are at a similar pressure to the intra-abdominal pressure (Fig. 4a). The pressure gradientacross the rectal wall will be small, being the intra-abdominal pressure - intrarectal pressure. The direc-tion of the gradient may vary according to whether or

not there is contraction of the rectal wall muscle. Inthe patient with a rectal prolapse the situation is

different, here the prolapsed mucosa loses the sup-port of the rectal contents, and as the prolapsereaches the outer end of the anal canal it comes in

contact with air at atmospheric pressure. Under thesecircumstances the pressure gradient across the pro-lapsed rectal wall will be great (Fig. 4). This highpressure may lead to distension of the submucosal

200%/

100°/.

1231

on July 12, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.28.10.1228 on 1 October 1987. D

ownloaded from

Wo,nuack, William s, Holnllfield, auid Morrison

VoidingIRP lOO cm water

__ . o_ ^~~W

PUBO

SEAS 4

MVC 5 seconds

IRP lOOcm water, - /

Voiding

lit ^ PP kC \,k'k1\1/VX

PUBO

I'o

SEAS

. 5 seconds

Fig. 3 Intraire(ctal pres.sulre( datnal sp)hinc ter aclivit yduring voiding. (a) normnal subject, the s/)hincter is inhiited

(Iiuri,ig v'oiding. (h) SRUS.subject, activity in the sphinclerr('riUit% with sstraining and persi.ss throu(ghout voiding.(IRP= iral r(ect(alj(l.resllire, PUBO= puhorectalis EMGacti i'it, SLA .S superficial external atial sphincter EMGactlivl, M V('= ti(virial5 olliaitarx'r contlractionl oft/ilc.S)pliiletrs, /)00(m waetr= 75 mtitnmHg.)

blood vessels resulting in rupture with rectal bleedingand devitalisation of the rectal mucosa. As thishypothesis depends upon the voiding pressure, in

combination with rectal prolapse, to explain themucosal ulceration it does not require the sphinctersto be overactive during voiding, although this may bethe most frequent cause of a high voiding pressure in

these patients. Other causes of a high voiding pres-sure such as hard constipated stool, or obstructeddefecaition due to the presence of the head of an

intussusception in the anal canal m-ay be sufficient tocreate the correct conditions for ulceration to occur.

This proposed aetiological mechanism for solitaryulceration miay explain the non-random situation ofthese lesions within the rectum.' Solitary ulcerscharacteristically occur at a distance of 7- 10 cm fromthe ainail margin. This wias the usual level of origin ofthe investigations of the rectal wall that prolapsedinto the anal canal as an anterior wall prolapse,

intussusception or full thickness rectal prolapse. Theusual position of solitary ulceration, therefore, coin-cides with that part of the rectal mucosa subjected toa high transmural pressure gradient during voiding.The characteristic anterior position of solitary ulcersmay result from selective exposure of the anteriorrectal wall to a high transmural pressure gradientoccurring as a result of the way the rectum prolapses.Prolapse usually starts with invaginiation of theanterior rectal wall at the initiation of voiding. Insome patients the lesion progresses no further alndonly the anterior rectal wall is exposed to the highpressure gradient. In other patients the prolapsingmucosa becomes circumferential as voiding pro-ceeds. The anterior component, however, is oftenthe largest (Fig. 1) and it forms the leading part of theintussusception as it descends the anal canal. In suchan intra-anal intussusception the anterior rectalmucosa would be exposed to the greatest pressuregradient and, therefore, would be the preferentialsite for ulceration. In a full thickness rectal prolapsethe prolapsed mucosa would be exposed to a highpressure gradient circumferentially and ulcerationcould occur at any point. In this study the occultforms of prolapse, that would favour anterior ulcer-ation, were more common (Table 2) and it would beexpected, therefore, that the most frequent site forulceration would be anterior.To analyse the data from this study we have

divided the SRUS patients into two groups accordingto whether ulceration was present or absent onsigmoidoscopy at the time of the study. Such adivision might be considered to be artificial as thereare a few patients with this condition in whom theulcer may heal and recur. Whilst we accept that theremay be some overlap between the two groups, thefew longitudinal studies of this condition that havebeen published' "' suggest that the ulceration is in-dolent- that is, once the ulcer has formed it remains.None of our patients showed any evidence of relaps-ing ulceration during the period of managementbefore the study. It may be that the factors which webelieve are the cause of ulceration act intermittentlyand this may explain the relapsing nature of thedisease in some patients.

Overactivity of the sphincters during voiding was acommon factor contributing to a high voiding pres-sure, occurring in approximately 70% of the patientsin this study. The cause of the overactivity is un-known but the correlation between the intrarectalpressure and the degree of sphincter overactivityimplies that it is because of a failure to inhibit themyogenic stretch reflex that normally modulates theactivity of the pelvic floor muscles."' It results inexcessive straining at stool, which, over a period oftime, is thought to cause neurogenic damage to the

1 232

i ,1.

on July 12, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.28.10.1228 on 1 October 1987. D

ownloaded from

Pressure andprolapse - the cause ofsolitary rectal ulceration 1233

aD Intra-abdominalIntra-abdominal pressure 1 ta /1 1pressure I ntrcirectIa I

pressure p pessre2 p presr

a atmospheric pressure a atmospheric pressure

Transmural pressure = p-p= p Transmural pressure = P-a=P

Fig. 4 Diagrammatic representation ofthe pressures involved in rectal voiding. Ap=pressure gradient across the rectal wall.(a) normal. (b) with intra-anal intussusception.

sphincters resulting in weakness.' This in turn wouldbe expected to result in lower voiding pressures andindeed an inverse correlation between age andvoiding pressure was shown in this study. This mayexplain why the peak incidence of solitary rectalulceration occurs in younger subjects, where voidingpressures are highest, and decreases with increasingage. "'The findings of our study have implications for the

management of this condition. If the dual factorhypothesis of the aetiology of solitary rectal ulcer-ation is correct removal of one of the factors shouldbe sufficient to cure the ulceration. There is atpresent no successful treatment for sphincter over-activity during voiding and therapy, therefore,should be directed at the rectal prolapse. Suchtreatment has been reported'3 to improve symptomsthat can be related to the presence of ulceration orthe prolapse itself. Such treatment, however, wouldnot be expected to remove all obstructive symptomsas some of these will be due to the pelvic floor muscleabnormality which will be unaffected by rectopexy.These residual symptoms are probably best dealtwith by explanation of the problem and advice toresist the urge to excessive straining as much aspossible.

This work was supported by a grant from the MedicalResearch Council. The authors wish to thank Pro-fessor D Johnston, Department of Surgery, TheGeneral Infirmary at Leeds, for atllowing us to studyhis patients and Janice Womack for secretarialassistance.

References1 Madigan MR, Morson BC. Solitary ulcer of the rectum.Gut 1969; 10: 871-81.

2 Du Boulay CE, Fairbrother J, Isaacson PG. Mucosalprolapse syndrome - a unifying concept for solitary ulcerand related disorders. J Clin Pathol 1983; 36: 1264-8.

3 Schweiger M, Alexander-Williams J. Solitary rectalulcer syndrome of the rectum - its association withoccult rectal prolapse. Lancet 1977; i: 170-1.

4 White CM, Findlay JM, Price JJ. The occult rectalprolapse syndrome. BrJ Surg 1980; 67: 528-30.

5 Womack NR, Holmfield JHM, Morrison JFB, WilliamsNS. Anorectal function in the solitary rectal ulcersyndrome (SRUS). Br J Surg 1986; 73: 49 1.

6 Mahieu P, Pringot J, Vanheuverzwijn R, Gancette L.Les prolapsus du rectum. Acta Gastro-Enterolog Belg1981; 44: 502-12.

7 Womack NR, Williams NS, Holmfield JHM, MorrisonJFB and Simpkins KC. New method for the dynamicassessment of anorectal function in constipation. Br JSurg 1985; 72: 994-8.

8 Goligher J. Rectal prolapse. In: Surgery of the anus,rectum and colon. London: Bailliere-Tindall, 1984: 246.

9 Rutter KRP. Electromyographic changes in certainpelvic floor abnormalities. Proc R Soc Med 1974; 67:53-6.

10 Rutter KRP, Riddel RH. The solitary rectal ulcersyndrome. Clin Gastroenterol 1975; 4: 505-30.

11 Parks AG, Porter NH, Melzak J. Experimental study ofthe reflex mechanism controlling the muscles of thepelvic floor. Dis Colon Rectum 1962; 5: 407-14.

12 Kiff ES, Barnes P, Swash M. Evidence of pudendalneuropathy in patients with perineal descent and strain-ing at stool. Gut 1984; 25: 1279-82.

13 Nicholls RJ, Simson JNL. Anteroposterior rectopexy inthe treatment of solitary rectal ulcer syndrome withoutovert rectal prolapse. BrJ Surg 1986; 73: 222-4.

on July 12, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.28.10.1228 on 1 October 1987. D

ownloaded from