Июнь

Int J Colorect Dis (1998) 13: 113115
Springer-Verlag 1998
O R I G I N A L A RT I C L E
G. Milito F. Cortese C. U. Casciani
Rhomboid flap procedure for pilonidal sinus:

results from 67 cases
Accepted: 6 February 1998
Abstract Sixty-seven patients with chronic pilonidal sinuses were treated by excision and rhomboid flap transposition (RFT). Primary healing was obtained in all patients
except two who developed a seroma and one who had a
partial dehiscence of the surgical wound due to a hematoma, which necessitated drainage through the margin of
the flap. The average stay was 5.3 days (range 116). All
patients returned to normal activities within 2 weeks of surgery. No late recurrence occurred after a mean follow-up
of 74.4 months (range 8137).
ter the age of 40 years [27]. Its cause is controversial, and

the disease is thought to originate in hair follicles [8]. Operations include excision and healing by open granulation
[9], excision with marsupialization [10], excision and primary closure [11, 12], excision and repair by plastic procedures [1315], cleaning out of hairs or the Lord-Millar
operation [16], and Bascoms operation of pit excision and
lateral drainage [8, 17]. We have been impressed with the
results of Dufourmentels operation [18, 19], and here report our own experience with this operation.
Key words Pilonidal sinus Dufourmentels technique

Plastic flaps Sacrococcygeal fistula
Materials and methods
Rsum 67 patients porteurs dun sinus pilonidal chronique ont t traits par excision et lambeau cutan rhombode. Une gurison premire a t obtenue chez tous les
patients lexception de deux qui ont dvelopp un srum
et un qui a dvelopp une dhiscence partielle de la plaie
en raison dun hmatome ayant ncessit un drainage par
la berge du lambeau. La dure moyenne de sjour est de
5.3 jours (116). Tous les patients ont repris une activit
normale dans les deux semaines qui ont suivi la chirurgie.
Aucune rcidive tardive na t observe au cours dun follow-up moyen de 74.4 mois (8137).
From April 1986 to April 1996, 67 patients with pilonidal sinus were
treated by excision and rhomboid flap transposition (RFT). Sixtyone (91%) were male of ages ranging from 1833 years (mean:
24 years). Fifty-one (76%) patients had one opening within the natal cleft, ten (15%) had two or three openings, always along the midline, while six (9%) also had one or more laterally situated granulation-lined openings. Six (9%) had previously been operated by excision with primary skin closure and had experienced recurrence
12 years after surgery. Swelling was the most frequent clinical manifestation (43 patients, 64%). Almost all were hirsute (65 patients,
97%). Clinical presentation included pilonidal abscess treated by
drainage (22 cases, 33%); chronic discharge (32, 48%); and single
uninfected sinus (13, 19%).
Surgical technique
Introduction
There are many ways to treat pilonidal sinus, perhaps reflecting the overall unsatisfactory results. Pilonidal sinus
or jeep disease [1] is common and afflicts young patients
after puberty with a male-female ratio of 3 : 1, it is rare afG. Milito () 1 F. Cortese C. U. Casciani
Cattedra di Clinica Chirurgica,
Universit di Roma Tor Vergata, Ospedale SantEugenio,
10 Piazzale dellUmanesimo, I-00144 Roma, Italy
Present address:
1
1530/b Via Cassia, I-00123 Roma, Italy
RFT was never performed during a period of acute inflammation,

but 54 patients (81%) had been treated before with antibiotics or had
undergone incision and drainage of an abscess. In the first 27 patients (40%), general anesthesia was used, but, in the remaining 40,
saddle spinal anesthesia with 1% hypertonic marcaine had been successfully employed.
With the patient in a prone jackknife position, a rhombus including the pilonidal sinus was marked (Fig. 1). It was excised down to
the presacral fascia in the midline and to the gluteal fascia laterally,
the main flap including skin and subcutaneous fat. Complete mobilization is essential to prevent tension. The wound was closed in two
layers, a subcutaneous adsorbable suture that incorporated a bite of
gluteal fascia was followed by skin closure. Suction drainage was
used and removed on the 2nd postoperative day. Skin sutures were
removed on the 7th or 8th postoperative day, and the patients were
114
AB = BC = CD = DA = AE = EF = AC = AF
ABC = BAC = BCA = AEF = 60
AI = LC = AH = GF
EM = BN
E
H
M
G
A2
C
A1
AC
BD
8
7.5
7
6.5
6
5.5
5
4.5
4
3.5
3
13.80
12.94
12.07
11.21
10.35
9.49
8.62
7.76
6.9
6.04
5.17
FGMHA
FE
A INLC
2
AE
may well be hope that it will shorten further. Compared

with most other techniques, wound healing is rapid and recurrence has not yet been seen.
The problem with excision and primary closure is that
suture-line tension sometimes occurs; and, combined with
quite frequent serous collections, this leads to wound infection rates as high as 47% [11, 20, 21]. Lay-open techniques are less likely to recur than with primary closure,
but healing times are prolonged. Other flap procedures
also have their advocates, although we have not found
them as easy to perform as this operation. Most depend
on avoidance of midline wounds, but, cosmetically, the resulting distorsion of the natal cleft is not always acceptable to the patient. Furthermore, the Z [22] and W
[15] plasty techniques leave acute angles, which can lead
to tip necrosis in up to 20% of cases, whereas
Dufourmentels technique has angles of 60 and 90 and,
in our experience, has not suffered from this problem.
Bascoms operation [8, 17, 23] of pit excision and lateral
drainage holds promise with most operations having been
performed on an ambulatory care basis. This probably reflects practice in the United States, as quite extensive flap
procedures were also treated by him as day cases [17]. It
may be that Dufourmentels and Bascoms operations
should be compared prospectively in a randomized, controlled clinical trial.
References
Fig. 1 Marking of a rhombus, including the pilonidal sinus
discharged with instructions about personal hygiene and the necessity for repeated shaving of the surrounding skin. All patients were
seen after 2 weeks and then after 3, 6, 8, and 12 months after discharge.
Results
Sixty-three of the 67 wounds healed by first intention. Of

the 67 patients, 64 (96%) had no complications. Two patients had a wound seroma, which healed with conventional
measures, and one formed an abscess, which was drained
through the margin of the flap, the opening healed within
four weeks. The average hospital stay was 5 days [In the
first 27 patients it was 8 days (range 716) while in the
next 40 patients it was 2 days (range 13)]. All but the patient with an infection returned to work 2 weeks after the
operation. There have been no recurrences over a mean follow-up period of 74.4 months (range 8137 months).
Discussion
This operation has led to remarkably good results in our

clinic. Hospital stay has shortened considerably, and there
1. Buie LA (1944) Jeep disease (pilonidal disease of mechanized

welfare). South Med J 37: 103109
2. Clothier PR, Haywood JR (1984) The natural history of the postanal (pilonidal) sinus. Ann R Coll Surg Engl 66: 201203
3. Monterola C, Barroso M, Araya JC, Fonseca L (1991) Pilonidal
disease: 25 cases treated by Dufourmentel technique. Dis Colon
Rectum 8: 649652
4. Sias F, Licheri S, Secci L, Loi R, Daniele GM (1994) Il trattamento ambulatoriale del sinus pilonidalis. Chirurgia 7: 639642
5. Hegge HG, Vos GA, Patka P, Oitsma HF (1987) Treatment of
complicated or infected pilonidal sinus disease by local application of phenol. Surgery 112: 5254
6. Price ML, Griffiths WAD (1985) Normal body hair a review.
Clin Exp Dermatol 10: 8797
7. Obeid SAF (1988) A new technique for treatment of pilonidal
sinus. Dis Colon Rectum 31: 879885
8. Bascom JU (1983) Pilonidal disease: long-term results of follicle removal. Dis Colon Rectum 23: 800807
9. Notaras MJ (1970) A review of three popular methods of treatment of postanal (pilonidal) sinus disease. Br J Surg 57: 886890
10. Abramson DJ (1960) A simple marsupialization technique for
treatment of pilonidal sinus: long-term follow-up. Ann Surg
151: 261263
11. Goligher JC (1984) Surgery of the anus, rectum and colon. 5th
edn, Baillire Tindall, London
12. Sarles JC, Sartre B, Delecourt P (1977) Kystes sacro-coccigiens
traites par excision chirurgicale et fermature immediate. Gastroenterol Clin Biol 1: 929931
13. Monro RS, McDermott T (1965) The elimination of causal factors in pilonidal sinus treated by Z-plasty. Br J Surg 52: 177181
14. Lalor JJ (1954) Plastic flap closure (25 cases). J Int Coll Surg
22: 181182
15. Roth RH, Moorman WL (1977) Treatment of pilonidal sinus and
cyst by conservative excision and W-plasty closure. Plast Reconstr Surg 60: 412415
115
16. Lord PH, Millar DM (1965) Pilonidal sinus: a simple treatment.
Br J Surg 52: 298300
17. Bascom JU (1987) Repeat pilonidal operations. Am J Surg
154: 118122
18. Dufourmentel C (1963) An L shaped flap for lozenge shaped
defects. Transactions of the Third International Congress of
Plastic Surgery. Excerpta Medica Foundation, Amsterdam,
pp 722726
19. Azab ASG, Kamal MS, Saad RA, Abou al Atta KA, Ali NA
(1984) Radical cure of pilonidal sinus by a transposition rhomboid flap. Br J Surg 71: 154155
20. Kronborg O, Christensen K, Zimmermann-Nielsen CE (1985)

Chronic pilonidal disease: a randomized trial with complete
3-year follow-up. Br J Surg 72: 303304
21. Bentivegna SS, Procario P (1967) Primary closure of pilonidal
cysts and sinuses. Am J Surg 43: 214216
22. McDermott FT (1967) Pilonidal sinus treated by Z-plasty. Aust
NZ J Surg 37: 6469
23. Mosquera DA, Quayle JB (1995) Bascoms operation for pilonidal sinus. J R Soc Med 88: 45P46P
J. L. Kelly D. S. ORiordain E. Jones E. Alawi

M. G. ORiordain W. O. Kirwan
The effect of hysterectomy on ano-rectal physiology
Abstract Hysterectomy is associated with severe constipation in a subgroup of patients, and an adverse effect on
colonic motility has been described in the literature. The
onset of irritable bowel syndrome and urinary bladder dysfunction has also been reported after hysterectomy. In this
prospective study, we investigated the effect of simple
hysterectomy on ano-rectal physiology and bowel function.
Thirty consecutive patients were assessed before and 16
weeks after operation. An abdominal hysterectomy was
performed in 16 patients, and a vaginal procedure was performed in 14. The parameters measured included the mean
resting, and maximal forced voluntary contraction anal
pressures, the recto-anal inhibitory reflex, and rectal sensation to distension. In 8 patients, the terminal motor
latency of the pudendal nerve was assessed bilaterally.
Pre-operatively, 8 patients were constipated. This improved
following hysterectomy in 4, worsened in 2, and was unchanged in 2. Symptomatology did not correlate with
changes in manometry. Although, the mean resting pressure was reduced after hysterectomy (57 mmHg
53 mmHg, P = 0.0541), the maximal forced voluntary contraction pressure was significantly decreased (115 mmHg
105 mmHg, P = 0.029). This effect was more pronounced
in those with five or more previous vaginal deliveries
(P = 0.0244, n = 9). There was no significant change in the
number of patients with an intact ano-rectal inhibitory reflex after hysterectomy. There was no change in rectal sensation to distension, and the right and left pudendal nerve
terminal motor latencies were unaltered at follow-up. Our
results demonstrate that hysterectomy causes a decrease in
the maximal forced voluntary contraction and pressure, and
this appears to be due to a large decrease in a small group
of patients with previous multiple vaginal deliveries.
Rsum Lhystrectomie est associe une constipation

svre chez un sous-groupe de patientes et lon a dcrit un
effet ngatif de lhystrectomie sur la motilit colique. On
a galement rapport aprs hystrectomie le dbut dun
syndrome du clon irritable et des troubles de la fonction
urinaire. Dans une tude prospective, nous avons tudi
leffet dune hystrectomie simple sur la physiologie anorectale et la fonction colique. 30 patientes conscutives ont
t investigues avant et 16 semaines aprs lopration.
Une hystrectomie abdominale a t ralise chez 16 patientes et une hystrectomie par voie vaginale chez 14. Les
paramtres mesurs comportent la pression de repos rsiduelle, la pression de contraction volontaire maximale, le
rflexe recto-anal inhibiteur, la perception de la distension
rectale. Chez 8 patientes, le taux de latence du nerf honteux interne a t mesur des deux cts. En propratoire,
8 patientes taient constipes. A la suite de lhystrectomie, la constipation sest amende chez 4, pjore chez 2
et demeure inchange chez 2. La symptomatologie nest
pas corrle avec des changements manomtriques. Bien
que la pression de repos soit rduite aprs hystrectomie
(57 53 mmHg, P = 0.0541), la pression de contraction
volontaire maximale est significativement abaisse (115
105 mmHg, P = 0.029). Cet effet est encore plus prononc
chez les patientes qui ont eu 5 ou plus daccouchements
par voie vaginale (P = 0.0244, n = 9). Il ny a pas de changement significatif pour le nombre de patientes qui prsentent un rflexe recto-anal inhibiteur intact aprs hystrectomie. Il ny a pas de changement dans la perception
de la distension rectale.
Introduction
Key words Hysterectomy Anal sphincter

Incontinence Resting anal pressure Maximum
voluntary contraction
J. L. Kelly D. S. ORiordain E. Jones E. Alawi
M. G. ORiordain W. O. Kirwan ()
Department of General Surgery, Cork University Hospital,
Wilton, Cork, Ireland
Hysterectomy is the second most frequently performed major surgical procedure in the developed world [1], and although most operations are uneventful, the complication
rate during hospital admission is approximately 7% [2, 3].
Aside from the initial complications, post-hysterectomy irritable blowel syndrome develops de novo in 13% of pa-
117
tients at 6 weeks [4]. A small group of patients also proceed to develop idiopathic slow-transit constipation [5].
This is believed to be the result of damage to the autonomic
nerve supply of the hindgut, resulting in functional obstriction [6], and may necessitate left hemicolectomy [7]. Furthermore, altered relaxation of the internal anal sphincter
and decreased rectal sensation have been described following radical hysterectomy for carcinoma of the cervix [8].
Therefore, in order to evaluate fully the effect on ano-rectal physiology, we examined a group of patients, before
and after simple abdominal and vaginal hysterectomy.
Patients and methods

Thirty patients were prospectively examined. Five gynaecologists
participated in the study, and consecutive patients were referred. All
patients gave informed consent, and permission was obtained from
the ethics committee at Cork University Hospital. Each patient was
assessed before, and 16 weeks after, hysterectomy. Exclusion criteria included: (1) previous ano-rectal surgery, (2) neurological disorders, (3) faecal incontinence, and (4) active ano-rectal disease, such
as haemorrhoids or anal fissures. Patients with neoplasia were also
excluded, except those with cervical intra-epithelial neoplasia.
A detailed questionnaire was completed before each visit, and a
full examination was routinely performed. This involved details of
bowel function, medications, and previous surgery, including all obstetric treatment and deliveries.
Five patients were classified as having constipation (2 bowel
movements per week, or straining at stool >25% of the time) and one
as having diarrhoea (>21 movements per week, or loose watery stools
>25% of the time) before surgery. No patient was taking anti-spasmodic, anti-diarrhoeal, or stool softening/bulking agents. The decision to perform vaginal or abdominal hysterectomy was made by the
individual gynaecologist in consultation with each patient.
After screening, the patient was placed in the left lateral position
and a standard rectal probe (Gaeltec, Isle of Skye, Scotland, UK) was
inserted into the anal canal, and manipulated until the high-pressure
zone was located. No bowel preparation was given beforehand. Mean
resting anal pressure was taken as an average reading during a 30-s
interval. The maximal pressure generated by forced voluntary contraction of the external anal sphincter was measured, and the highest of three separate recordings noted. The presence or absence of
the recto-anal inhibitory reflex was determined in standard fashion
[7]. Rectal sensation was measured using an air-filled, intra-rectal
balloon. The lowest of three volumes required to produce a sensation of gas and a desire to defecate was recorded.
The terminal motor latency of the pudendal nerve was measured
bilaterally in eight patients using a St. Marks pudendal nerve electrode (Dantec, Bristol, UK).
Results
The median age was 46 years (range 30 64). Sixteen patients had an abdominal, and fourteen had a vaginal hysterectomy, with preservation of ovarian function. The most
frequent indications for operation were abnormal bleeding
(40%), chronic pelvic pain (30%) and leiomyomas (14%).
Age showed a negative correlation with mean resting anal
pressure (correlation coefficient, r = 0.4241; P = 0.019),
but not with maximal forced voluntary contraction anal
pressure (correlation coefficient, r = 0.24; P = 0.19). The
patiens who had a vaginal hysterectomy were significantly
older (mean age 44 years) than those who had an abdominal hysterectomy (mean age 54 years). However, there was
no difference between the groups regarding the number of
vaginal deliveries, or the mean resting, or forced voluntary
contraction anal pressures.
Seventeen patients reported an unchanged bowel pattern at follow-up. In nine, there was a slight increase in the
frequency of bowel motions, and a decrease in four. Only
one patient required stool softening agents post-operatively.
The mean resting anal pressure was unchanged by hysterectomy, but the forced voluntary contraction anal pressure was significantly reduced at 16 week (115 mmHg
105 mmHg, P = 0.029). This effect was due to a large decrease in nine patients with a history of five or more vaginal deliveries. In this group, the mean drop in forced
voluntary contraction pressure after hysterectomy was
21 mmHg (112 mmHg 91 mmHg, P = 0.048). In those
who had had fewer than five vaginal deliveries, the mean
drop in forced voluntary contraction pressure was 3 mmHg
(115 mmHg 112 mmHg, not significant). There was no
difference between these two groups regarding the age of
the patients or the number who had a vaginal hysterectomy
(4/9 and 10/21), or those who had an abdominal hysterectomy (5/9 and 11/21). The number of vaginal deliveries
did not correlate with the mean resting (correlation coefficient, r = 0.21) or maximal forced voluntary contraction
pressures (correlation coefficient, r = 0.027).
The threshold volume and the urge volume were unchanged at follow-up. The terminal motor latency of the
pudendal nerve was not effected by hysterectomy, and no
difference was found in the number of patients with an intact recto-anal inhibitory reflex.
Statistical analysis
Correlation coefficients were calculated by linear regression analysis and ANOVA. Paired, and unpaired, Students
t-tests were used to compare variables between the various groups, and significance was set at P 0.05.
Discussion
Pelvic surgery can damage the autonomic innervation of organs leading to dysfunction of the urinary bladder [9] and,
in a small group of patients, to intractable constipation [5,
6]. Hysterectomy has been reported to result in increased
rectal sensitivity [4], and impaired motility of the distal sigmoid colon has also been described [7]. In association with
abnormalities of bladder function, this is most likely due to
damaged fibres of the inferior hypogastric plexus.
It is believed that traction injury to the parasympathetic
nerves (S2, S3 and S4) lying lateral to the vaginal fornix,
and in the broad ligament of the uterus, is responsible for
post-hysterectomy constipation [10]. Although cases of faecal incontinence and rectal prolapse have been reported [11],
118
no adverse effects on the pudendal nerve-innervated, external and sphincter mechanism have yet been described [12].
In this study, we found a significant decrease in the maximal forced voluntary contraction pressure, generated by
the external and sphincter, at 16 weeks. This was due to
the large decrease recorded for nine patients who had had
five or more previous vaginal deliveries. The mechanism
responsible may be direct trauma to the external and
sphincter at the time of operation, or damage to the S3 and
S4 nerves supplying the puborectalis part of the external
sphincter. It is worth noting that the external urethral
sphincter, which is innervated by the pudendal nerve, is
particularly susceptible to injury following hysterectomy
[9, 13]. In this study, no patient reported urinary incontinence, and we found no difference in the terminal motor
latency of the pudendal nerve at follow-up. The resting anal
pressure in our study showed a negative correlation with
age, and in contrast to previous reports, was unchanged
following hysterectomy [8].
In the past, hysterectomy has been associated with the
onset of irritable bowel syndrome in up to 13% of cases
[4], and we found a significant change in bowel pattern at
16 weeks in 13 of our patients. In agreement with previous reports, rectal sensitivity was unchanged following
hysterectomy [10]. It has been suggested that hysterectomy
may have a greater effect on the act of defecation than on
colonic motility, as up to 25% of patients report increased
straining at stool at follow-up [14]. No patient in this study
developed faecal incontinence and the number of patients
with an intact recto-anal inhibitory reflex remained the
same at follow-up.
Although no patient in this study developed faecal incontinence, we believe that hysterectomy may have an adverse effect on the external sphincter mechanism in a subgroup of patients. It appears that patients with a history of
multiple vaginal deliveries are more prone to external
sphincter injury after hysterectomy. The mechanism remains unclear, but the terminal motor latency of the pudendal nerve seems to be unaffected. This appears to indicate either direct trauma to the sphincter itself, or to the
nerves supplying the puborectalis.
In summary, we have demonstrated that following hysterectomy there is a decrease in the maximal pressure gen-
erated by the external anal sphincter. This effect is pronounced in a small group of patients with previous multiple vaginal deliveries. Although this decrease did not have
any adverse effect at 16 weeks, it may have a detrimental
effect on those patients with an already compromised
sphincter.
References
1. Graves EJ (1992) National hospital discharge survey: annual
summary, 1990. Vital and health statistics, series 13, no. 112.
National Centre for Health Statistics, Hyattsville, Maryland
2. Carlson KJ, Miller BA, Fowler FJ (1994) The Maine womens
health study outcomes of hysterectomy. Obstet Gynecol
83: 556 564
3. Clarke A, Black N, Rowe P, Mott S, Howle K (1995) Indications
for and outcome of total abdominal hysterectomy for benign
disease: a prospective cohort study. Br J Obstet Gynaecol
102: 611 620
4. Prior A, Stanley KM, Smith ARB, Read NW (1992) Relation
between hysterectomy and the irritable bowel: a prospective
study. Gut 33: 814 817
5. Roe AM, Bartolo DCC, McMortensen NJ (1988) Slow transit
constipation comparison between patients with or without previous hysterectomy. Dig Dis Sci 33: 1159 1163
6. Varma JS (1992) Autonomic influences on colorectal motility
and pelvic surgery. World J Surg 16: 811 819
7. Vierhout ME, Schreuder HWB, Veen HF (1993) Severe slow
transit constipation following radical hysterectomy case report. Gynecol Oncol 51: 401 403
8. Barnes W, Waggoner S, Delgado G, et al (1991) Manometric
characterisation of rectal dysfunction following radical hysterectomy. Gynecol Oncol 42: 116 119
9. Taylor T, Smith AN, Fulton PM (1990) Effects of hysterectomy
on bowel and bladder function. Int J Colorectal Dis 5: 228
10. Devrode G, Lamarche J (1984) Functional importance of extrinsic parasympathetic innervation of the distal rectum and colon
in man. Gastroenterology 86: 287
11. Schwartz S (1994) Textbook of surgery. 6th edn. McGraw-Hill,
New York, p 1220
12. Prior A, Stanley K, Smith ARB, et al (1992) Effect of hysterectomy on anorectal and urethrovesical physiology. Gut 33: 264
267
13. Bartolo DCC, Jarratt JA, Read MG, Donnelly TC, Read NW
(1983) The role of partial denervation of the puborectalis in idiopathic faecal incontinence. Br J Surg 70: 664 667
14. Heaton KW, Parker D, Cripps H (1993) Bowel function and irritable bowel symptoms after hysterectomy and cholecystectomy a population based study. Gut 34: 1108 1111
P. J. Hainsworth D. C. C. Bartolo
Selective omission of loop ileostomy in restorative proctocolectomy
Abstract Omission of a temporary ileostomy in patients

undergoing restorative proctocolectomy is controversial.
Although fewer operations may be required and some complications avoided, the risks of anastomotic dehiscence and
pelvic sepsis may be greater. Patients undergoing restorative proctocolectomy with no ileostomy (Group NI, n = 72)
were compared retrospectively with patients given a conventional loop ileostomy (Group I, n = 30). Criteria for
avoiding faecal diversion included: absence of severe acute
colitis, good nutritional status and favourable surgery with
creation of a sound, tension-free anastomosis. Steroid intake was not a contraindication to single-stage surgery. Delayed stomas were necessary in 8% of Group NI. For
Groups NI and I, the rates of anastomotic leak (3% vs 3%),
pelvic sepsis without demonstrable leak (3% vs 0%), pouch
fistula (3% vs 10%) and intestinal obstruction (8% vs 3%)
were similar. Closure of the temporary ileostomy in Group
I was associated with a 10% complication rate. Cumulative post-operative hospital stay was significantly less in
Group NI (median 11 vs 16 days). Functional results at 1 year
were similar. A temporary loop ileostomy can be safely
avoided in carefully selected patients undergoing restorative proctocolectomy.
Key words Ileoanal anastomosis Ileostomy
Restorative proctocolectomy Ulcerative colitis
Rsum Faut-il renoncer une ilostomie temporaire
chez des patients subissant une proctocolectomie avec
anastomose ilo-anale? Bien que le nombre des temps
opratoires puissent tre rduits et quun certain nombre
This paper was presented in part to the Association of
Coloproctology of Great Britain & Ireland, Tripartite Meeting,
London, 810 July 1996
P. J. Hainsworth () 1 D. C. C. Bartolo
Department of Colorectal Surgery,
The Royal Infirmary of Edinburgh, EH3 9YW, UK
Present address:
1
Ward 5, Freeman Hospital, High Heaton, Newcastle-upon-Tyne,
NE7 7DN, UK
de complications puissent tre vites, le risque de

dsunion anastomotique et de sepsis du pelvis est accru.
Des patients subissant une proctocolectomie restorative
sans ilostomie (groupe NI, n = 72) ont t compars
rtrospectivement avec des patients chez lesquels une
ilostomie latrale conventionnelle avait t ralise
(groupe I, n = 30). Les critres pour renoncer une diversion fcale comportaient: labsence de colite aigu svre,
un bon tat nutritionnel et des conditions chirurgicales favorables avec la confection dune anastomose sans tension.
La prise de strodes ne constitue pas une contre-indication une intervention en un temps. La confection dune
stomie de manire diffre a t ncessaire chez 8% du
groupe NI. Dans les groupes NI et I, le taux de fuite anastomotique (3% vs 3%), de sepsis du pelvis sans fuite
vidente (3% vs 0%), le taux de fistule sur la poche (3%
vs 10%) et le taux docclusion intestinale (8% vs 3%) sont
similaires. La fermeture de lilostomie provisoire dans le
groupe I tait associe avec un taux de complications de
10%. La dure cumule des sjours hospitaliers
postopratoires tait significativement plus rduite dans le
groupe NI (valeur mdiane 11 vs 16 jours). Des rsultats
fonctionnels une anne sont identiques. On peut renoncer de manire sre une ilostomie latrale temporaire
dans un groupe de patients correctement slectionns et
devant subir une proctocolectomie avec rtablissement de
la continuit.
Introduction
Omission of a temporary loop ileostomy in selected patients undergoing restorative proctocolectomy is not new.
First advocated more than 10 years ago [1], avoidance of
an ileostomy remains controversial. Perceptions of an
easier, quicker and more reliable anastomosis with stapled
versus hand-sewn construction may account for the increased popularity of stoma avoidance [2]. Opponents cite
the relative safety of a temporary ileostomy [3] and express concerns about a possible increase in frequency of
120
Table 1 Published series of
restorative proctocolectomy
with ileostomy (I) and with no
ileostomy (NI). (Numbers in
parentheses represent I group).
All complications shown as
percentages. (Sel selected,
Mat matched controls,
Rand randomised, Con consecutive, Hist historic controls)
Ref/Year
1. 8083
4. 8190
5. 8488
6. 8388
7. 8891
8. 8489
9. 8788
10. 8789
11. 8991
12. 8991
13. 8993
14. 8993
15. 8292
Present
a
b
c
d
e
Numbera
NI (I)
21
37 (37)
38
32 (53)
22 (23)
29 (35)
16 (15)
25 (21)
30 (28)
50 (50)
68 (63)
68
71 (87)
72 (30)
Study
designb
Sel
Sel/Mat
Sel
Sel
Rand
Sel
Con (Hist)
Con (Hist)
Sel
Sel (Mat)
Sel (Hist)
Sel
Sel
Sel
IPAAc
Hand
Hand
Stapled
Both
Stapled
Hand
Hand
Hand
Stapled
Stapled
Stapled
Stapled
Stapled
Both
Complication rates (%)

Leaksd
abscess
Bowel
obstruction
Pouch
faile
10
8 (14)
11
18 (53)
5 (9)
17 (11)
6 (0)
4 (0)
10 (11)
18 (6)
15 (24)
5
30 (13)
6 (3)
19
?
?
6 (23)
9 (22)
7 (14)
13 (13)
24 (16)
7 (29)
0 (2)
7 (10)
?
?
8 (3)
10
3 (8)
0
6 (17)
5 (0)
?
0 (0)
4 (0)
0 (0)
0 (0)
0 (19)
3
1 (3)
0 (0)
Numbers of patients in NI&(I) groups

Selection of patients in NI&(I) groups
Method of IPAA (ileal pouch anal anastomosis)
Anastomotic leaks, excluding closure of ileostomy
Pouch failure: pouch excised or diverted at last follow-up
pelvic sepsis or more serious consequences if sepsis occurs. Table 1 shows the salient findings from reports comparing patients with and without temporary diversion.
In 1990 we reported a series of ileostomy closures after restorative proctocolectomy and found the morbidity to
be low with most patients able to leave hospital soon after
the procedure [16]. Therefore, we initially adopted a very
liberal approach to the use of temporary ileostomies. Between 1990 and 1995 a stoma was increasingly avoided
providing the operation proceeded satisfactorily and that
experience forms the basis of this report.
without a proximal stoma. A temporary ileostomy was avoided providing the patient was not acutely unwell (fever, tachycardia, abdominal tenderness, toxic dilatation) and not malnourished. Further criteria for stoma avoidance included performance of a straightforward
operation with creation of a sound, tension-free anastomosis. Confidence in omitting an ileostomy increased during the study period.
Recent steroid intake or immunosuppression did not dictate the need
for a stoma.
Median follow-up was 17 months (051 months). Outpatient follow-up at regular intervals assessed 24-hour stool frequency, nocturnal emptying, continence and use of constipating medication.
Results
Between June 1990 and March 1995, 102 patients (57 male, 45 female) underwent restorative proctocolectomy. Age ranged between
12 and 76 (median 36) years. Preoperative diagnoses were ulcerative colitis (n = 83); indeterminate colitis (n = 4); familial adenomatous polyposis (FAP) (n = 5); idiopathic megacolon and/or slow transit constipation (n = 6); Hirschsprungs disease (n = 1); diffuse angiodysplasia (n = 1) and colon cancer (n = 2).
Patients who were severely ill underwent colectomy and ileostomy with a view to delayed reconstructive surgery. Other patients
underwent immediate restorative proctocolectomy or, if the colectomy had been carried out elsewhere, rectal excision and reconstruction.
All operations were performed or directly supervised by the senior author (D. C. C. B.). Broad-spectrum antibiotics and subcutaneous heparin were routinely given. The small bowel was mobilised to
the origin of the superior mesenteric artery, preserving the ileal
branch of the ileocolic artery. The preferred method of pouch construction was the stapled J with 20 cm limbs. Early on in the series, W pouches were created in 7 patients. The pouch-anal anastomosis was single-stapled in 91 patients preserving 12 cm of mucosa above the dentate line. The remaining 11 patients underwent
mucosectomy with hand-sewn anastomosis to the dentate line.
Those in Group I (temporary ileostomy) were defunctioned at a
convenient point proximal to the pouch. Group NI (no ileostomy)
patients underwent restorative proctocolectomy or proctectomy
A temporary loop ileostomy was added at the time of fashioning the pouch in 30 out of 102 patients (29%) (Table 2).
All those with toxic megacolon or acute colonic bleeding
underwent initial subtotal colectomy with delayed pouch
construction. Among 25 other patients with acute colitis,
7 underwent immediate restorative proctocolectomy, three
without a covering ileostomy. All those undergoing restorative proctocolectomy for chronic relapsing colitis, dysplasia or adenomatous polyps had immediate pouch construction, mostly without a covering stoma. Patients undergoing pouch surgery without an ileostomy were contrasted
with those given an elective ileostomy (Table 3). There
were no significant differences between the groups although trends towards low body weight and low serum albumin were present in Group I.
A late stoma was required by 6 out of 72 patients (8%)
in Group NI for the following reasons: pouch anastomotic
leak (2 cases), pelvic sepsis, small bowel perforation,
pouch-vaginal fistula and one instance of staple gun
sphincter injury. None of these six cases had acute colitis
at the time of creating the pouch. One elderly male patient
who suffered a leak from the top of the pouch was taking
121
Table 2 Use of diverting ileostomy at time of pouch procedure: no
ileostomy (NI); proximal ileostomy (I)
Indication for pouch procedure
NI (n=72)
I (n=30)
Chronic relapsing colitis

Dysplasia/Polyps
Acute colitis
Non-colitis
Previous colectomy and end ileostomya
Revisional pouchb
34
6
3
10
19
0
10
1
4
5
9
1
Indications: acute colitis (n=18), toxic megacolon (n=8), rectal

bleeding (n=2)
b
First pouch created elsewhere
Table 3 Comparison of patients initially without (NI) or with (I)
proximal ileostomy
Age (years)
Sex
Weight (kg)
History (years)
Hb
WBC
Serum albumin
On steroids
median (range)
M:F
median (range)
median (range)
<110 g/l
12 109/l
<36 g/l
NI (n=72)
I (n=30)
39 (1265)
41 : 31
68 (44105)
5 (025)
9 (13%)
7 (10%)
6 (8%)
22 (31%)
32.5 (1576)
16 : 14
60.5 (4895)a
4 (024)
4 (13%)
4 (13%)
7 (23%)b
13 (43%)
Table 4 Early and late complications for pouch procedures in those

initially without (NI) and with (I) proximal ileostomy
NI (n=72)
Pelvic sepsis:
Anastomotic leak
No leak demonstrated
Small bowel perforation
Fistulae:
Pouch-vaginal
Pouch-cutaneousb
Anastomotic stricture
Dilatation
Pouch advancement
Sphincter injury
Small bowel obstructionc
Laparotomy
Conservative
Wound infection
Incisional hernia
Pouchitis
Closure of ileostomy:d
Small-bowel perforation
Wound infection
Incisional hernia
I (n=30)
2a
2 (1a)
1a
1
0
0
1a
1
1
2
1
1
1a
0
1
0
3
3
1
1
13
1
0
1
0
8
0
0
0
1
1
1
a
a
Mann-Whitney U test, 2 P=0.067, NS

b 2
X (1df)=2.32 (with Yates correction) P=NS
high dose steroids and azathioprine at the time of surgery.

Of the other five, none were taking steroids or immunosuppressants. Steroid therapy before pouch surgery was not
associated with complications leading to a late stoma.
Overall, 65% of patients underwent pouch surgery without the need for an immediate or late stoma.
Early and late surgical complications for the two groups
are shown in Table 4. Subgroup numbers are small but there
are no significant differences in the incidence of pelvic sepsis, post-operative fistulation or obstructive episodes in
those undergoing pouch surgery with and without an ileostomy.
Pelvic sepsis occurred in four patients in Group NI, two
of whom had a demonstrable leak. One other patient with
a radiological leak but no sepsis is not included in this
group. One patient settled with antibiotics and three required late ileostomies which were subsequently closed.
In Group I, one patient developed a leak complicated by a
pelvic abscess and subsequently a pouch-cutaneous fistula
and anastomotic stricture which required revisional surgery.
There were two early post-operative deaths from massive pulmonary embolism despite standard subcutaneous
heparin prophylaxis. Post-mortem failed to identify any
underlying sepsis or other risk factors. A further at-risk patient, who was prepared with a caval filter and full-dose
heparin, suffered a caval thrombosis and survived. Two
other patients experienced pulmonary emboli after discharge from hospital with no sequelae. All patients now
receive high-dose subcutaneous heparin and calf stimulation.
Delayed ileostomy
All late events
After pouch procedure or closure of any ileostomy
d
Including late ileostomies
b
c
Table 5 Pouch function at intervals after establishing intestinal continuity: no ileostomy (NI); proximal ileostomy (I)
Time after
surgery
NI (n=72)
I (n=30)
Freq/24 h
Med (range)
6 weeks
6 months
12 months
7 (220)
5 (120)
4 (211)
Noctural
emptying
6 weeks
6 months
12 months
42/67 (63%)
16/55 (29%)
9/42 (21%)
16/27 (59%)
6/23 (26%)
7/21 (33%)
Soiling
6 weeks
6 months
12 months
30/68 (44%)
19/54 (35%)
11/43 (26%)
12/28 (43%)
5/23 (22%)
6/21 (29%)
7 (312)
5 (310)
5 (210)
The incidence of pouchitis appeared more common in

Group I (27% vs 18%) but this is not significant and is
probably a reflection of longer median follow-up in Group
I (28 vs 16 months). Extracolonic manifestations of colitis
were equally common in Groups I (3/30) and NI (6/72).
Pouchitis only developed in one patient without colitis,
who had FAP. Pouchitis occurred more frequently in those
with colitis extending to at least the transverse colon
(19/69) compared with left-sided colitis (0/13) (2 P=0.047,
Fisher Exact Probability test).
Overall, 59% patients underwent pouch surgery without developing any early or late complications.
Post-operative stay was calculated on an intention-totreat basis and included subsequent admissions for related
complications. Median cumulative post-operative stay for
122
Group NI was 11 (664) days and for Group I 16 (10160)

days (Mann-Whitney, U=530.5, 2 P<0.001).
The functional results were similar in both groups and
improved with time (Table 5). Anal seepage was commonly
experienced in the early post-operative phase but reduced
to an overall rate of 27% at 12 months and was mainly minor mucous leakage. Stoma closure was delayed in one patient with a staple gun sphincter injury but eventual function was satisfactory. No patient experienced major faecal
incontinence. Antidiarrhoeals were frequently prescribed
but at 12 months only 27% required treatment.
Discussion
Numerous reports indicate that restorative proctocolectomy may be safely undertaken without a covering ileostomy in selected patients [48, 10, 11, 13, 14]. Potential
benefits of stoma avoidance include the need for just one
operation, reduced operating time, shorter hospital stay and
avoidance of ileostomy related complications [4, 8, 10, 11,
13, 17]. The crux of the argument for and against temporary ileostomy rests on the relative incidence of complications and their severity in each group.
All available studies are limited by a variable degree of
patient selection bias. The only randomised study excluded
patients taking steroids [7] and in the 45 low-risk patients
who were randomised, only two cases of pelvic sepsis occurred. Even in one apparently consecutive series, the authors excluded acute and emergency cases [10]. Others
have included the vast majority of patients although those
with acute fulminant colitis are usually excluded [13]. One
study apparently included patients with acute fulminant
colitis [14].
In selecting patients for stoma avoidance, Galundiuk et
al. [4] stress the importance of an absolute lack of tension
on the anastomosis, good blood supply to the terminal ileum, good general health and absence of recent steroid intake. In the present series, fewer patients were on steroids
in Group NI (31% vs 43%) but most serious complications
occurred in those not taking steroids. The literature is divided on the risks of steroids in those without a covering
stoma and although several series show no increase in complications [1012, 17], one study reports an increase in pelvic sepsis with 20 mg/day prednisolone [12].
Because of the potentially disastrous consequences of
pelvic sepsis and the relative safety of creating and closing a temporary ileostomy, many consider temporary diversion to be preferable [3, 12, 1820]. The overall incidence of pelvic sepsis among 1218 patients treated at the
Majo clinic was 5% [21]. The collected data in Table 1
show no obvious increase in pelvic sepsis or pouch failure
in those without ileostomies. However, if an anastomotic
leak occurs, are the consequences worse in the absence of
a proximal stoma? In this series, pelvic sepsis occurred in
4/72 patients in Group NI, three of whom required late ileostomies. In Group I, one of 30 patients developed a leak
complicated by a pelvic abscess, fistula and anastomotic
stricture which required revisional surgery. The presence

of an elective stoma does not necessarily prevent serious
sequelae but, in patients without a stoma, a proximal ileostomy should be raised at the first sign of sepsis.
Pouch-vaginal fistula is particularly troublesome to treat.
Among 17 cases from St. Marks Hospital [22], 13 occurred
1144 (median 7) months after ileostomy closure, 1 in a
patient who had no ileostomy, and 3 before ileostomy closure. The prognosis was worse in the group of 13 which
appeared after ileostomy closure. In the present series, the
incidence of late fistulae, including pouch-vaginal fistula,
was no more common in those without an elective ileostomy (3% vs 10%).
Complications associated with the ileostomy occur in
about 10% of patients [3, 16, 19] but are usually easily
remedied. Several groups report more frequent smallbowel obstruction after diversion (Table 1) although we
did not see this. Rarely, the blood supply to the pouch
may be compromised by creating a proximal stoma under
tension.
An argument often advanced in favour of a period of
temporary diversion is that patients can later compare
pouch function with ileostomy function. Tjandra et al. [12]
suggested that quality of life (assessed by patients on an
analogue scale) was worse at 6 weeks if a stoma was omitted but this difference later disappeared. The evidence from
our series and published reports [7, 12] indicates that longterm function is identical in the two groups.
Reports of restorative proctocolectomy without temporary diversion have included patients having hand-sewn as
well as stapled procedures (Table 1). Our preference is for
a stapled anastomosis since it is simple, quick and effective. It avoids prolonged anal dilatation, allows preservation of the anal transitional zone and has been associated
with good short-term results in our hands.
In summary, we have shown that in a consecutive series of over 100 patients undergoing pouch surgery, a stoma
was safely avoided in about two-thirds. Patients thought to
carry a particularly high risk of complications were excluded. Most patients referred for total colectomy are not
acutely unwell and a temporary stoma can be safely
avoided. Nevertheless, any anastomotic leak is likely to be
more serious in the absence of a proximal stoma and, in
our opinion, a stoma should only be avoided where the
pouch operation has proceeded uneventfully with creation
of sound, tension-free anastomoses in well perfused bowel.
If there is any technical uncertainty about the pouch or
pouch-anal anastomosis then a stoma is indicated. If a patient who is not defunctioned develops pelvic sepsis, then
early intervention is mandatory: the pouch can usually be
salvaged with immediate intervention and creation of an
ileostomy.
The incidence of complications in this series is very low
compared with other reports and stoma avoidance may be
less safe where the background morbidity is higher. Although a selective policy of stoma avoidance has proved
safe on this unit, caution dictates that individual units
should ensure a low incidence of complications before
adopting a policy of stoma avoidance.
123
References
1. Thow GB (1985) Single-stage colectomy and mucosal proctectomy with stapled antiperistaltic ileoanal reservoir. In: Dozois RR
(ed) Alternatives to conventional ileostomy. Year Book Medical
Publishers, Chicago, 1985: 420432
2. Winslet MC, Barsoum G, Pringle W, Fox K, Keighley MRB
(1991) Loop ileostomy after ileal pouch-anal anastomosis is it
necessary? Dis Colon Rectum 34: 267270
3. Khoo RE, Cohen MM, Chapman GM, Jenken DA, Langevin JM
(1994) Loop ileostomy for temporary fecal diversion. Am J Surg
167: 519522
4. Galandiuk S, Wolff BG, Dozois RR, Beart RW (1991) Ileal
pouch-anal anastomosis without ileostomy. Dis Colon Rectum
34: 870873
5. Peck DA (1988) Stapled ileal reservoir to anal anastomosis. Surg
Gynecol Obstet 166: 562564
6. Hosie KB, Grobler SP, Keighley RMB (1992) Temporary loop
ileostomy following restorative proctocolectomy. Br J Surg 79:
3334
7. Grobler SP, Hosie KB, Keighley MRB (1992) Randomized trial
of loop ileostomy in restorative proctocolectomy. Br J Surg
79: 903906
8. Everett WG, Pollard SG (1990) Restorative proctocolectomy
without temporary ileostomy. Br J Surg 77: 621622
9. Jrvinen HJ, Luukkonen P (1991) Comparison of restorative
proctocolectomy with and without covering ileostomy in ulcerative colitis. Br J Surg 78: 199201
10. Matikainen M, Santavirta J, Hiltunen K (1990) Ileoanal anastomosis without covering ileostomy. Dis Colon Rectum 33:
384388
11. Sagar PM, Lewis W, Holdsworth PJ, Johnston D (1992) Onestage restorative proctocolectomy without temporary defunctioning ileostomy. Dis Colon Rectum 35: 582588
12. Tjandra JJ, Fazio VW, Milsom JW, Lavery IC, Oakley JR, Fabre
JM (1993) Omission of temporary diversion in restorative proctocolectomy is it safe? Dis Colon Rectum 36: 10071014
13. Sugerman HJ, Newsome HH (1994) Stapled ileoanal anastomosis without a temporary ileostomy. Am J Surg 167: 5866
14. Mowschenson PM, Critchlow JF (1995) Outcome of early surgical complications following ileoanal pouch operation without
diverting ileostomy. Am J Surg 169: 143145
15. Cohen Z, McLeod RS, Stephen W, Stern HS, OConnor B, Reznick R (1992) Continuing evolution of the pelvic pouch procedure. Ann Surg 216: 506511
16. Lewis P, Bartolo DCC (1990) Closure of loop ileostomy after
restorative proctocolectomy. Ann R Coll Surg Engl 72: 263265
17. Jrvinen HJ, Luukkonen P (1993) Coloproctectomie et anastomose ilo-anale en un temps: bnfices et facteurs de risques.
Ann Chir 47: 971975
18. Wong WD, Rothenberger DA, Goldberg SM (1985) Ileoanal
pouch procedures. Curr Probl Surg 22 (3): 978
19. Wexner SD, Taranow DA, Johansen OB, Itzkowitz F, Daniel N,
Nogueras JJ, Jagelman DG (1993) Loop ileostomy is a safe option for fecal diversion. Dis Colon Rectum 36: 349354
20. Metcalf AM, Dozois RR, Kelly KA,Wolff BG (1986) Ileal
pouch-anal anastomosis without temporary ileostomy. Dis Colon Rectum 29: 3335
21. Pemperton JH (1985) Complications, management, failure and
revisions. In: Nicholls RJ, Bartolo D, Mortensen N (eds) Restorative proctocolectomy. Blackwell Scientific Oxford pp 3452
22. Groom JS, Nicholls RJ, Hawley PR, Phillips RKS (1993) Pouchvaginal fistula. Br J Surg 80: 936940
F. Pucciani M. L. Rottoli A. Bologna F. Cianchi

S. Forconi M. Cutell C. Cortesini
Pelvic floor dyssynergia and bimodal rehabilitation:

results of combined pelviperineal kinesitherapy and biofeedback training
Abstract Dyschezia may be caused by pelvic floor dyssynergia, which takes place when a paradoxical contraction or a failure to relax the pelvic floor muscles occurs
during attempts to defecate. The aim of our study was to
set up a new bimodal rehabilitation programme for pelvic
floor dyssynergia, which combined pelviperineal kinesitherapy and biofeedback, and to evaluate the results of this
treatment. Thirty-five patients (age range: 2864 years;
mean age: 42.5 years) from the outpatient unit of the Clinica Chirurgica of the University of Florence, Italy, and an
age-matched group of 10 healthy control subjects (age
range: 3159 years; mean age 45.7 years) with normal
bowel habits and without any defecatory disorders, were
studied. The 35 patients were symptomatic for dyschezia
without slow colonic transit and had been diagnosed as being affected by pelvic floor dyssynergia. No evidence of
any organic aetiology was present but all demonstrated
both manometric and radiological evidence of inappropriate function of the pelvic floor. All of the patients underwent bimodal rehabilitation, using the combined training
programme Clinical evaluation, computerized anorectal
manometry and defecography were carried out 1 week before and 1 week after a completed course in bimodal rehabilitation. The control group underwent manometric and
defecographic examination. Their results were compared
with those of the 35 patients before and after training. After the programme, all 35 patients had a very significant
increase in stool frequency (P<0.001), while laxative and
enema-induced bowel movements had become significantly less frequent (P<0.001). After bimodal rehabilitation, computerized anorectal manometry showed some peF. Pucciani () A. Bologna F. Cianchi C. Cortesini
Istituto di Clinica Chirurgica Generale e Discipline Chirurgiche,
Universit degli Studi di Firenze, Unit di Coloproctologia (UCP),
Viale Morgagni 85, I-50134 Firenze, Italy
M. L. Rottoli
Radiodiagnostica 1, Dipartimento Fisiopatologia Clinica,
Universit degli Studi di Firenze, Italy
S. Forconi M. Cutell
U.O. Riabilitazione, Ospedale Mauriziano, Torino, Italy
culiar results. Resting anal canal pressure had increased

but not significantly. Pre-programme values that indicated
a shorter duration (exhaustio) of maximal voluntary contraction than found in the controls had returned to normal
values. The rectoanal inhibitory reflex (RAIR), with incomplete relaxation, which had been shorter than that of
controls, became normal by the end of the rehabilitation.
All RAIR parameters were significantly different especially when pre- and post-treatment values were compared
(P<0.001). No differences were found as regards rectal
sensation parameters and rectal compliance between those
before or after bimodal rehabilitation. Defecographic pretreatment X-ray films showed indentation of the puborectalis and poor anorectal angle (ARA) opening, at evacuation, with trapping barium of at 50%. After pelviperineal
kinesitherapy and biofeedback training, the indentation
had disappeared and the ARA had become significantly
larger (P<0.001) during evacuation. No differences were
found after rehabilitation, when both were compared with
those of controls. The pelvic floor descent was also significantly deeper (P<0.001) than before the start of the programme. The bimodal rehabilitation technique can be considered a useful therapeutic option for functional dyschezia as shown by our clinical evaluations, manometric data
and defecographic reports.
Key words Dyschezia Arismus Rehabilitation
Kinesitherapy
Rsum Une dyschzie peut tre cause par une dyssynergie du plancher pelvien qui survient lorsque se produit
une contraction paradoxale ou, linverse, un dfaut de relaxation de la musculature du plancher pelvien au cours
dun effort de dfcation. Certaines formes de training, tel
le biofeedback, sont connues pour pouvoir aider certains
patients relcher la musculature strie du plancher pelvien au cours de lexonration. Le but de notre tude a t
de mettre au point un nouveau programme de rhabilitation bimodale en cas de dyssynergie pelvienne qui combine de la kinsithrapie pelvi-prinale et do biofeedback
ainsi que dvaluer le rsultat de ce traitement. Trente-cinq
125
patients (ge: 28 64 ans; ge moyen: 42.5 ans) de lunit

ambulatoire de la Clinique de Chirurgie de lUniversit de
Florence en Italie et un groupe contrle de 10 sujets de
mme ge (ge: 31 59 ans; moyen dge: 45.7) avec des
habitudes dexonration normales et sans trouble de la
dfcation ont t tudis. Les 35 patients taient symptomatiques pur une dyschzie sans inertie colique et ont t
diagnostiqus comme tant porteurs dune dyssynergie du
plancher pelvien (PFD): aucune vidence de lsion organique nest prsente mais tous dmontrent, tant la manomtrie que lors dexamens radiologiques, des vidences
de fonctions inappropries du plancher pelvien. Tous les
patients ont subi une rhabilitation bimodale laide dun
programme dentranement combin. Une valuation clinique, une manomtrie anorectale digitalise et une dfcographie ont t ralises une semaine avant et un mois
aprs la fin du traitement de rhabilitation bimodale. Le
groupe de contrle a t soumis une manomtrie et une
dfcographie. Les rsultats ont t compars ceux des
35 patients avant et aprs traitement. Aprs le programme,
tous les 35 patients ont montr une augmentation significative de la frquence des selles (P<0.001), une frquence
significativement abaisse de la consommation de laxatifs
et de lavements pour induire lexonration (P<0.001).
Aprs la rhabilitation bimodale, la manomtrie anorectale digitalise montre des rsultats particuliers. La pression de repos du canal anal est augmente mais pas de manire significative. Les donnes avant traitement qui montraient une dure rduite de la contraction volontaire maximale comparativement aux tmoins taient retournes
des valeurs normales. Le rflexe recto-anal inhibiteur avec
une relaxation incomplte qui tait plus bref que chez les
contrles sest normalis la fin du traitement. Toutes les
valeurs prthrapeutiques de rflexe anorectal inhibiteur
montrent des diffrences significatives si on les compare
ceux des contrles. Ces valeurs sont toutefois significativement diffrentes, particulirement lorsque les valeurs
pr- et postthrapeutiques sont compares (P<0.001). Aucune diffrence nest retrouve en ce qui concerne les paramtres de perception rectale et la compliance rectale de
mme queen ce qui concerne les donnes avant et aprs
la rhabilitation bimodale. Les dfcographies prthrapeutiques montrent une indentation de la sangle puborectale et un dfaut de louverture de langle anorectal au cours
de lvacuation lorsque 50% du baryum est retenu. Aprs
la kinsithrapie pelvi-prinale et le biofeedback, lindentation disparat et le rflexe anorectal inhibiteur est allong au cours de lexonration (P<0.001). Aucune diffrence nest retrouve aprs rhabilitation lorsque ces
donnes sont compares celles de sujets contrles. Le
plancher prinal est significativement abaiss (P<0.001)
quavant le dbut di programme. La technique de rhabilitation bimodale peut tre considre comme une option
thrapeutique utile en cas de dyschzie fonctionelle. Les
donnes cliniques et des valuations manomtriques et
dfcographiques confirment ces observations.
Introduction
Pelvic floor dyssnergia may be defined as a faecal evacuation disorder which is a consequence of some functional
outlet obstruction [1]. It is characterized by a paradoxical
contraction or failure to relax the pelvic floor muscles during attempts at defecating. Clinical manifestations can include straining, feeling of incomplete evacuation and/or
the need to digitally evacuate the rectum [2]. D. M. Preston and J. E. Lennard-Jones [3] suggested that some form
or retraining might help patients to relax the striated muscles of the pelvic floor during defecation. Biofeedback appears to be effective for this form of dyschezia [4], even
though the questions of how, when and why it works remain unanswered [5]. Pelviperineal kinesitherapy, which
has been tried occasionally in cases of dyschezia [6], is a
specific muscular re-education technique for the pelvic
floor muscles [7].
The aims of this study were to set up a new bimodal rehabilitation technique using both pelviperineal kinesitherapy and biofeedback, for pelvic floor dyssynergia,
and to evaluate the results of this treatment.

Thirty-five women (age range: 2864 years; mean age: 42.5 years)
from the outpatient unit of the Clinica Chirurgica of the University
of Florence (Italy) and one group of 10 healthy age-matched female
subjects (age range: 3159 years; mean age 45.7 years) were studied and compared. All 45 women were multipara.
A the start of the study, the 35 patients were symptomatic for dyschezia, without slow colonic transit, and had been diagnosed as being affected by pelvic floor dyssynergia (PFD). Dyschezia was defined as difficult defecation during at least 25% of bowel movements
over a period of at least 3 months. Diagnostic criteria for PFD were
those of the Working Team Report on the Functional disorders of
the anorectum. There was no evidence of any organic aetiology. All
patients showed both manometric and radiological evidence of inappropriate function of the pelvic floor [2]. All patients underwent
bimodal rehabilitation, using both pelviperineal kinesitherapy and
biofeedback training as defined below.
Computerized anorectal manometry and defecography were carried out 1 week before and 1 week after a completed course of bimodal rehabilitation.
The 10 control subjects had normal bowel habits and no defecatory disorders. All 10 also had computerized anorectal manometry
and defecography. Their results were compared to those of the 35
patients both before and after rehabilitation.
Written consent had been obtained from all the participants before the start of the study.
Clinical evaluation
Information regarding bowel movements and concomitant diseases
had been gathered from previously completed patient charts. Patients
with any psychiatric illness, metabolic and/or endocrine disease, neurological diseases, obstetric sphincteric lesions, and those who had
undergone anorectal and/or pelvic surgery were excluded from the
study.
126
Computerized anorectal manometry
Anorectal manometry was performed using standard techniques [8].
Recordings and analyses of the tracings were made using a computerized system (Dyno System, Menfis s.r.l., Bologna, Italy) as previously described [9].
Anal Resting Pressure (ARP) was recorded in mmHg with the
stationary pull-through technique and the computer identified the
maximal pressure (Pmax), the mean pressure (Pm) and the high pressure zone area (HPZ area), where HPZ area was = (Pi Li): Pi was
the pressure value in mmHg at each sampling, and Li was the length
in millimeters between two successive samplings. The maximal voluntary contraction (MVC) was examined by evaluating the voluntary contractions of the anal sphincter; amplitude was expressed in
mmHg, duration in seconds. The rectoanal inhibitory reflex (RAIR)
was elicited by inflating a soft balloon in the rectum at 10 cm from
the anal verge wiith 40 ml of air, our normal value to induce complete relaxation [9]. The computer quantified the total duration of reflex (TDR) in seconds: TDR was equal to the complete amount of
relaxation time (RT) in seconds plus contraction time (CT) in seconds, as suggested by Martelli [10]. The computer also quantified
the maximal amplitude of relaxation (MAR) expressed in percent,
the residual pressure at the lowest point of the RAIR (Pres in mmHg),
the mean RAIR pressure (Pm RAIR in mmHg) and the area of the reflex where the RAIR area was = (Pi Li): Pi was the pressure value in mmHg at each sampling and Ti was the time in seconds between
two successive samplings.
The first distension volume at which internal anal sphincter relaxation had occurred, i.e. the RAIR threshold, (RAIRT), and the distension volume at which an initial transient sensation had taken place,
i.e. the conscious rectal sensitivity threshold, (CRST), were determined in all patients and controls. The maximal tolerated volume
(MTV) was also measured in all subjects and it was considered an
expression of rectal reservoir capacity. Compliance of the rectum
was expressed by the ratio mmHg/ml of inflated air, measured by
means of the pressure/volume curve. Manometric signs of PFD were
high anal canal pressure (Pm; HPZ area) and impaired RAIR, i.e. incomplete relaxation with a short duration of the reflex [9].
Defecography
All patients and controls underwent defecography according to the
methods suggested by the Italian Working Team [11]. The radiological assessment was carried out at rest, during contraction, and during expulsion of the barium. The anorectal angle (ARA) was measured between the longitudinal axis of the anal canal and the tangential line to the posterior rectal wall, and was expressed in degrees.
The pelvic floor descent (PFDe), which was defined as the vertical
distance between the pubococcygeal line and the anorectal junction,
was expressed in millimetres. A qualitative evaluation, diagnostic
for pelvic floor dyssynergia, was made by noting the persistence of
the puborectalis indentation during evacuation.
Bimodal rehabilitation
All patients underwent bimodal rehabilitation. A single cycle consisted of ten outpatient sessions. Each session lasted 1 hour and took
place twice a week. The first step was pelviperineal kinesitherapy.
This was then combined with biofeedback training from the fifth session until the end of the cycle.
From the start to the end of treatment, patients were asked to
record stool frequency. They were also asked to indicate whether laxatives and/or enema assistance had been required.
a) Pelviperineal kinesitherapy
The cycle of pelviperineal kinesitherapy followed the standard sequence listed below, but was adapted to the individual woman. During each session two essential steps were taken: the exercises of the
last lesson were reviewed and new exercises were introduced so as

to ensure continuing and accurate patient response.
1st session:
preliminary lesson on relaxed breathing and corporeal consciousness (used at the start of all sessions),
diaphragmatic breathing,
marking of perineal area, made easier by peri- and
intra-anal digital manipulation,
location and focusing of agonist, antagonist and synergic muscles on the perineal plane.
2nd session:
antiversion and retroversion pelvic movements,

short anal contractions,
some exercises of short anal relaxation,
perianal and perivaginal stretching,
stretch reflexes of the puborectalis, elicited by the
therapist but contra a simultaneous voluntary anal
contraction.
3rd session: perianal and perivaginal stretching,

stretch reflexes of the puborectalis,
the learning of abdominal press principles (diaphragm, pelvic floor, abdominal wall, paravertebral
muscles, iliopsoas).
4th session: perianal and perivaginal stretching,
stretch reflexes of the puborectalis,
abdominopelvic synergy (the abdominal press force
vectors are directed to the posterior perineum while
simultaneous voluntary sphincterial anal relaxation
occurs),
simulation of defecation by expelling the therapists
forefinger, but without any abnormal muscular recruitment.
5th session: abdominopelvic synergy and simulation of defecation with slight pelvic floor descent (used from this
session until the end of the cycle),
consciousness reinforcement with the correct execution of anal relaxation,
start of biofeedback (learning of techniques and
some exercises regarding anal contractions/relaxation).
6th session: visual control of pelvic floor descent using a mirror,
anal corticalization stage: some anal contraction exercises are introduced (bending down, coughing, or
the use of Valsalvas manoeuvre in supine, upright,
sitting positions),
biofeedback (some exercises of anal relaxation).
7th session: response modulation: gradualness in sphincteric recruitment/inhibition,
biofeedback (response modulation).
8th session: response modulation in sphincter inhibition with
slight pelvic floor descent,
biofeedback (some exercise of anal relaxation with
modulation technique).
9th session: revision exercises on abdominopelvic synergy and
gradual anal relaxation,
biofeedback (revision exercises).
10th session: revision exercises,
biofeedback,
final interview (stool frequency, laxative-enema assistance).
b) Biofeedback
Biofeedback (BF) was performed using Contimed manometric BF
equipment (Hollister, Libertyville, Ill., USA). Once its function had
been explained to the patients, the equipment was used from the fifth
session to the end of rehabilitation. While the patient was lying in
the left lateral position, a BF probe was introduced into the anorectum. Patients were required to use their anal muscles as they had
127
Table 1 Clinical evaluation
Stool frequency (n/week)

Laxative assistance (n/week)
Enema assistance (n/week)
Digital evacuation
(n patients/total patients)
Obstructed micturition
a
Controls
(10)
Patients
Patients
(35) (pre) (35) (post)
6.1 1.2
2.7 1.8 a
1.2 0.3 a
1.6 0.6 a
26/35
8/35
6.3 2.0
0.4 0.2
0.6 0.1
3/35
Patients (pre) vs Patients (post) or vs Controls: P<0.001
learned to do during kinesitherapy. Feedback was noted by changes

in the coloured lights on the Contimed meter.
Statistical analysis
Results were expressed as mean standard deviations (SD).
Students t-test for paired and unpaired samples was used for statistical analysis.
Results
duration, which had been significantly (P < 0.001) shorter

(exhaustio) in patients than in controls before kinesitherapy, returned to significantly normal values after the
rehabilitation (Table 2). The RAIR parameters are reported
in Table 3.
TDR was statistically shorter in women with pelvic
floor dyssynergia (P < 0.01) given that they had had a
shorter CT (P < 0.001) before rehabilitation when compared to controls. After rehabilitation treatment, no significant differences were found as regards either TDR or CT
in patients and controls. In the patient group, the residual
pressure at the lowest point of reflex was statistically
higher (P < 0.001) than in controls while MAR, RAIR area
and mean RAIR pressure were statistically lower
(P < 0.001). At the end of rehabilitation, the 35 patients had
Pres, MAR, RAIR area and Pm RAIR values which were
similar to those of the controls. These values were also significantly different (P < 0.001) from those noted before the
treatment. RAIRT was not significantly different when patients and controls were compared. No differences were
found as regards CRST, MTV or rectal compliance, before
and after bimodal rehabilitation. All 35 patients had tolerated a volume of 180200 ml of air and all had compliance
of the rectum.
Clinical symptoms
Defecography
Table 1 shows a comparison of clinical characteristics of
controls and patients.
All 35 patients had a very significant increase in bowel
movement frequency (P < 0.001) after bimodal rehabilitation. The frequency of laxative-induced (P < 0.001) and enema-induced (P < 0.001) bowel movements was significantly reduced, and the need to digitally evacuate the rectum (74.2% of our patients had used this technique) was
no longer necessary after the therapeutic treatment. Only
3 women still had obstructed micturition at the end of the
ten sessions.
Computerized anorectal manometry
After bimodal rehabilitation, anal canal pressure had increased, but not significantly. Maximal pressure (P < 0.01),
mean pressure (P < 0.01), and HPZ area (P < 0.05) were significantly higher than those of the controls (Table 2). After treatment, MVC had no significant differences in amplitude when compared with pre-treatment values. MVC
Table 2 Anal canal pressures
Pmax
mmHg
Controls
Patients (pre)
Patients (post)
a
b
c
78.0 10.2
97.1 11.1 b
110.2 17.4 b
Defecographic results are reported in Table 4.

Before the rehabilitation treatment, all patients had
shown indentation of the puborectalis and a poor ARA
opening at evacuation (P < 0.001). Rectal emptying had
also been impaired to some degrees, as verified by barium
trapping results (50%). At the end of treatment, ARA was
significantly higher (P < 0.001) during evacuation, and was
then similar to that of controls. Indentation of the puborectalis had disappeared, and barium trapping was lower
(25%). After bimodal rehabilitation, the patients showed a
significantly deeper PFDe at evacuation (P < 0.01) than
controls. This was also a statistically significant difference
between pre- and post-course values.
Discussion
Dyschezia, defined as difficult bowel movements, may be

brought on by pelvic floor dyssynergia as a consequence
Pm
mmHg
41.8 6.6
51.6 8.6 b
52.8 6.8 b
HPZ area
(Pi Li)
1678.4 279.3
2685.1 374.3 a
2871.6 178.9 a
Patients (pre) or Patients (post) vs Controls: P<0.05

MVC
P (mmHg)
T (s)
124.1 3.7
151.1 15.6 a
163.7 20.4 a
24.3 7.3
17.3 5.1 c
24.5 6.8
128
Table 3 RAIR
Controls
Patients (pre)
Patients (post)
a
b
c
RT (s)
CT (s)
TDR (s)
Pm
mmHg
Pres
mmHg
MAR
(%)
RAIR area
(P Ti)
9.1 0.2
7.8 1.6
7.7 3.1
20.4 2.5
12.5 0.3 c
19.9 4,8
29.5 2.7
20.2 1.9 b
26.6 8.7
22.6 2.4
15.9 9.8 a
19.4 7.3
1.3 0.7
12.0 3.7 c
2.1 0.4
97.4 2.2
79.0 3.0 c
96.8 3.1
654.7 39
365.1 32 c
612.6 56

Table 4 Defecography (R resting, E evacuation)
Controls
Patients (pre)
Patients (post)
a
b
ARA (degrees)
Pelvic flor descent (mm)
94 3
85 6 a
87 2 a
110 3
87 4 b
106 9
39.4 11
32.6 18
37.6 15
69.4 11
49.5 12 a
82.3 13 a
Barium
trapping (%)
Puborectalis indentation
10
50
25
35/35

of obstructed defecation, which occurs when pelvic floor

muscles fail to relax during attempts to defecate. Biofeedback has been considered the best training method for this
common functional disorder [4]. Biofeedback is an operant conditioning; it is voluntary, employs a trial and error
process by which learning takes place, and the subject
must be aware of the desired response (graphic drawings
or coloured lights for the normal response). However, success is not always guaranteed and there is an absence of
any predictive factor. Moreover, it is not certain when and
how it works [5]. Even in view of these factors, it has been
considered a cortical reconditioning method for the defecation reflex [12]. Pelviperineal kinesitherapy is a specific
muscular re-education technique for the uncoordinated
pelvic floor muscles. This muscular training programme
works particularly well on elevator ani, by improving performance, extension and elasticity. However, since the
pelviperineal area has poor sensorial consciousness it
has been hypothesized that pelviperineal kinesitherapy
might require biofeedback to strengthen the re-education
process. Moreover, once this method has been learnt, it
might make it easier, in a few sessions, to achieve the biofeedback response desired. Therefore, a new bimodal rehabilitation procedure using both techniques was set up
for the treatment of pelvic floor dyssynergia and, for the
reasons stated above, we did not compare bimodal rehabilitation with biofeedback alone. Before therapy, all of
our patients had been affected by dyschezia and had had
diagnostic signs of PFD, such as higher than normal anal
canal pressure with impaired RAIR. Defecographic images at evacuation with indentation of the puborectalis and
poor ARA opening had also been characteristic of the 35
patients.
After bimodal rehabilitation, all the women showed improvement: they had a significant increase in bowel movement frequency (P < 0.001) and use of laxatives and ene-
mas had decreased (P < 0.001). The need to digitally evacuate the rectum disappeared and only three patients showed
obstructed micturition, with hesitancy. These symptomatic
improvements were confirmed by manometric-proctographic data.
After treatment, manometric results showed that anal
canal pressure was higher (Table 2) and that there was a
normal rectoanal inhibitory reflex (Table 3). Anal resting
pressure was statistically higher than that of the controls,
but it was not significantly different from pre-treatment
values. This was also shown by the defecographic resting
ARA which remained unchanged and was statistically
more narrow than that of controls (P < 0.05). It is difficult
to say what might have been the functional mechanism
which determined the anal canal hypertonia. The failure of
anorectal myectomy to resolve anismus [3], the absence of
any anal fissure, and recent advances in morphological
evaluation using sonography [13] suggest that the internal
anal sphincter cannot be responsible. On the other hand,
since the influence of striated sphincter muscles on resting
anal tone values is well known (1530%) [14, 15], it is possible that stronger contractions of the puborectalis muscle
and the external anal sphincter might have induced the increased anal canal pressure. But we do not know what the
aetiological factor is. Furthermore, the fact that there were
minimally higher changes in ARP (Table 2) with increased
anal squeeze pressure has also been reported by others after the use of biofeedback [5]. Nevertheless, anal hypertonia does not seem to be an important factor for defecation
retraining if correct pelvic floor function can be obtained.
Anal canal hypertonia is a physical and/or manometric sign
which indicates increased anal sphincteric strength: it does
not offer any information on anal sphincteric muscle coordination. Such strength is only one aspect of anal function.
On the contrary, the coordination of agonist, antagonist and
synergic muscles of the pelviperineal plane is the crucial
129
Fig. 1 P1010 = manometric
channel. t1; t2 = manometric
sample window: times at the
start (t1), at the end (t2). RAIR
elicited: r/40 = inflation of balloon 40 ml air. Upper tracing:
normal subject at r/40 (empty
arrow). Lower tracing: patient
with pelvic floor dyssynergia.
Note the high excitatory response at r/40 (full arrow) and
the incomplete relaxation during reflex
voluntary step for arriving at normal defecation. Bimodal

rehabilitation exercises did improve voluntary coordination. This can be demonstrated in our patients by the longer
duration of MVC after the sessions (P < 0.001), and by the
fact that the sphincteric recruitment had improved and
there had been longer endurence without any further assisted-influence, given that these changes left the MVC amplitude unchanged.
Therefore, the higher resting anal canal pressures at the
start of the study might not have been specific signs of pelvic floor dyssynergia; however, the impairment of RAIR
most probably was.
All the patients who had been affected by failure to relax the pelvic floor muscles during attempts to defecate
had impaired RAIR, with incomplete relaxation (MAR:
P < 0.001; RAIR area: P < 0.001) and with a duration which
had been shorter than that of controls (TDR: P < 0.01). After rehabilitation, patient RAIR became normal, and the
RAIR parameters showed no significant differences in relation to those of controls. On the other hand, there was a
significant difference between pre- and post-treatment values (P < 0.001) (Table 3). This seems to be the key to understanding some aspects of the pathophysiology of pelvic
floor dyssynergia. RAIR impairment cannot be caused by
any organic internal anal sphincter damage, for the reasons

stated above. RAIR is an intramural rectal reflex with an
off-on response: it is independent of any cortical influence.
In fact, to understand this anal dysfunction better, it is important to remember that, in normal subjects, the response
to rectal distension includes two recto-anal reflexes: relaxation of the internal anal sphincter (RAIR), and contractions of the puborectalis and the external anal sphincter
(Recto-Anal Excitatory Reflex: RAER). Both striated muscles act, functionally, as one unit [1618]. Relaxation of
the internal sphincter allows for the rectal content to come
into contact with the sensory epithelium of the anal canal
(anal sampling) [19]. The reflex excitatory response,
which proximally involves the puborectalis in the upper
anal canal and distally the external anal sphincter in the
lower anal canal [20], prevents the loss of faecal or flatus
sampling. Concerning defecation, after anal sampling, voluntary inhibition of both the striated muscles occurs and
rectal evacuation takes place. In our patients with pelvic
floor dyssynergia, proximal RAER might be higher than
in normal subjects as regards the paradoxical puborectalis
activity and thus might overcome the inhibitory response
(Fig. 1), causing RAIR impairment and outlet obstruction.
In fact, impaired RAIR weakens the anal sampling and in-
130
creases the pressure gradient between the anal canal and

rectum at evacuation, so that defecation does not begin.
After rehabilitation treatment, the coordinated activity of
the puborectalis seemed to be restored, as suggested by the
normal defecographic anorectal angle at evacuation. The
proximal excitatory response had improved and had less
influence on the pressure profile of RAIR. Finally, distal
RAER, which is not considered be a true reflex but an autonomic learned response [21] given that it is under voluntary control [20], could be modified during the muscular retraining sessions of the bimodal rehabilitation programme.
Defecographic X-ray films were a determining factor
for the evaluation of the therapeutic results. After rehabilitation, the anorectal angle showed a normal excursion at
evacuation. There was a significant difference when compared with pre-treatment values (P < 0.001) (Table 4). The
indentation of the puborectalis was shown to have disappeared in all patients. The mechanical obstruction, induced
by the paradoxical contraction or the failure to relax this
muscle, was no longer blocked, as evidenced by the lower
percentage of barium trapping. Pelvic floor descent at
evacuation was significantly deeper (P < 0.01) after rehabilitation and the dynamic excursion of the puborectalis
sling was restored during straining [13].
In conclusion, our bimodal rehabilitation techniques
seems to be a successful therapeutic option for functional
dyschezia. However, long-term follow up is necessary to
evaluate long-term results and validation in comparison to
other retraining programme could be the next step. Moreover, studies on the influence of the puborectalis on anal
canal pressure, the role of RAER, and the function of the
elevator any muscle may be necessary to arrive at a fuller
explanation of the multiple, functional and anatomical factors most probably involved in PFD.
Acknowledgements This work was supported in part by a grant
from Ministero dellUniversit e della Ricerca Scientifica e Tecnologica. We would like to thank Attilio Masi. He is considered by our
group to be the most patient, the most humane, and the best physical therapist we have ever met.
References
1. Badiali D, Habib FI, Corazziari E, Viscardi A, Primerano L, Anzini F, Torsoli A (1991) Manometric and defaecographic patterns of straining. J Gastroint Mot 3: 171
2. Whitehead WE, Devroede G, Habib FI, Meunier P, Wald A

(1992) Functional disorders of the anorectum. Gastroenterol Int
5: 92108
3. Preston DM, Lennard-Jones JE (1985) Anismus in chronic constipation. Dig Dis Sci 30: 413418
4. Park UC, Choi SK, Piccirillo MF, Verzaro R, Wexner S (1996)
Patterns of anismus and the relation to biofeedback therapy. Dis
Colon Rectum 39: 768773
5. Hong Ho Y, Tan M, Su Goh H (1996) Clinical and physiologic
effects of biofeedback in outlet obstruction constipation. Dis Colon Rectum 39: 520524
6. Louis D (1988) Incontinence anale et dyschezie: bases et
principes de la rducation. In: Guarzi C, Spreafico L (eds)
Rieducazione in uroginecologia. Tecnograph, Reggio Emilia,
pp 147156
7. Di Benedetto P (1995) Chinesiterapia pelvi-perineale: generalit. In: Di Benedetto P (ed) Riabilitazione uro-ginecologica.
Edizioni Minerva Medica, Torino, pp 147149
8. Cortesini C, Pucciani F, Carassale GL, Paparozzi C (1983) Anorectal physiology after anterior resection and pull-through operation. Eur Surg Res 15: 176183
9. Pucciani F, Rottoli ML, Bologna A, Buri M, Cianchi F, Pagliai
P, Cortesini C (1996) Anterior rectocele and anorectal dysfunction. Int J Colorect Dis 11: 19
10. Martelli H, Devroede G, Arhan P, Duguay C, Dornic C, Faverdin C (1978) Some parameters of large bowel motility in normal man. Gastroenterology 75: 612618
11. Piloni V, Genovesi N, Grassi R, Lazzini S, Pieri L, Pomerri F
(1993) Italian Working Team report on defecography. Radiol
Med 85: 784793
12. Papachrysostomou M, Smith AN (1994) Effects of biofeedback
on obstructive defecation. Reconditioning of the defecation reflex? Gut 35: 252256
13. Fransioli A, Weber B, Cunningham M, Roche B, Marti MC, Hadengue A (1996) Dynamic evaluation of puborectalis muscle
function by external perineal sonography. Tech Coloproctol
3: 125129
14. Duthie HL, Watts JM (1965) Contribution of the external anal
sphincter to the pressure zone in the anal canal. Gut 6: 6468
15. Frencker B, Euler B (1975) Influence of pudendal block on the
function of the anal sphincters. Gut 16: 482489
16. Denny Brown D, Robertson EG (1935) An investigation of the
nervous control of defaecation. Brain 58: 256310
17. Alva J, Mendeloff AI, Schuster MM (1967) Reflex and electromyographic abnormalities associated with fecal incontinence.
Gastroenterology 53: 101106
18. Scharli AF, Kiesewetter WB (1970) Defecation and continence:
some new concepts. Dis Colon Rectum 13: 81107
19. 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: 179182
20. Sangwan YP, Coller JA, Schoetz DJ, Murray JJ, Roberts PL
(1995) Latency measurements of rectoanal reflexes. Dis Colon
Rectum 38: 12811285
21. Whitehead WE, Orr WE, Engel BT, Schuster MM (1982) External anal sphincter response to rectal distension: learned response
or reflex. Psychophysiology 19: 5762
J. M. Gattuso H. S. Debinski H. O. Kangro

D. Jeffries M. A. Kamm
Evaluation of specific herpes DNA viruses in idiopathic megarectum

and idiopathic megacolon
Abstract Purpose: The aetiology of idiopathic megarectum and idiopathic megacolon is unknown. A previous study in patients with chronic idiopathic intestinal
pseudo-obstruction, a condition also associated with a dilated gut, identified the possible involvement of herpes
viruses. This study therefore aimed to determine whether
these viruses may also be implicated in the pathogenesis
of these conditions. Methods: Resected large bowel from
three patients with idiopathic megarectum and three patients with idiopathic megacolon were studied. Histology
for viral inclusions and nested polymerase chain reaction
(PCR) using specific primers for cytomegalovirus, Epstein-Barr virus, herpes simplex virus type 1 and varicella
zoster virus was performed. DNA was extracted from paraffin-embedded blocks by proteinase K and phenol chloroform extraction. Results: Viral inclusions were not
seen. PCR failed to identify DNA of the four herpes viruses tested. Conclusion: Patients with idiopathic megarectum or idiopathic megacolon may have subtle abnormalities of the enteric innervation, but these do not appear to be attributable to the neurotropic effects of the
herpes viruses studied.
Key words Idiopathic megarectum Megacolon
Herpes DNA viruses
Rsum But: Ltiologie du mgarectum idiopathique et
du mgaclon idiopathique est inconnue. Une tude pralable chez des patients avec une pseudo-obstruction intestinale idiopathique chronique, une condition associe avec
une dilatation du clon, a montr le rle potentiel dune atteinte virale herptique. Cette tude a t entreprise pour
dterminer si ces virus peuvent galement tre impliqus
J. M. Gattuso H. S. Debinski M. A. Kamm ()
St. Marks Hospital, Northwick Park,
Watford Road, Harrow,
Middlesex HA1 3UJ UK
H. O. Kangro D. Jeffries
Department of Virology,
St. Bartholomews Hospital,
London, UK
dans la pathognse de cette condition clinique. Mthode:

Les pices opratoires de trois patients avec un mgarectum idiopathique et de trois patients avec un mgaclon
idiopathique ont t tudies. Lexamen histologique la
recherche dinclusion virale et une raction en chane la
polymrase ont t raliss laide de sondes spcifiques
pour les cytomgalovirus, le virus de Epstein Barr, le virus de lherps simplex type I et le virus de la varicelle. Le
DNA a t extrait des prlvements inclus en paraffine au
moyen de la protinase K et par extraction au phnol chloroforme. Rsultats: Des inclusions virales nont pas t
vues. La PCR na pas permis didentifier le DNA du virus
de lherps ni des quatre autres virus tests. Conclusion:
Les patients porteurs dun mgaclon idiopathique ou dun
mgarectum idiopathique peuvent avoir des anomalies
subtiles de linnervation entrique mais celles-ci toutefois
ne semblent pas pouvoir tre attribues leffet neurotropique des virus herptiques tudis.
Idiopathic megarectum and idiopathic megacolon appear

to be distinct clinical conditions [1]. Patients with idiopathic megarectum present with constipation, abdominal pain, bloating and faecal incontinence and have a rectal diameter of more than 6.5 cm [2] which may be associated with a variable extent of proximal colonic dilatation. Patients with idiopathic magacolon have a normal rectum diameter, are older and have a functional large bowel
disorder, but no faecal incontinence [1].
The cause of both conditions is unknown. Routine histology of resected specimens of the gut from such patients
reveals no abnormalities. The architecture of the enteric
innervation is intact, but some abnormalities of the enteric
innervation have been reported [37].
In a previous study of patients with chronic idiopathic
intestinal pseudo-obstruction, Epstein-Barr virus (EBV)
was identified in the gut of one patient with an inflammatory plexitis using polymerase chain reaction (PCR). The
localisation of this virus to the myenteric plexus was confirmed by in situ hybridisation [8]. Two other patients had
132
cytomegalovirus, (CMV) identified by PCR, but no

changes were found by in situ studies. The resected gut of
the last two of these patients was normal histologically.
On account of these previous findings and the unknown
aetiology of idiopathic megarectum and megacolon, we
wished to examine the possibility of a viral cause for these
conditions. We chose to study the herpes DNA viruses because of our previous findings in patients with pseudo-obstruction and because of their known neurophilic properties.

Patient groups
Patients were selected on the basis of an acute or adult onset of their
condition as it was considered more likely that their disorders would
have a viral aetiology. All had previous surgery for intractable symptoms and failed medical therapy. We retrieved tissue from the colon
or rectum of three patients with idiopathic megarectum and three patients with idiopathic megacolon (Table 1). The following full-thickness specimens were obtained: (a) sigmoid colon from all six patients, all of whom had a megasigmoid; (b) rectum from one patient
with idiopathic megarectum and one with idiopathic megacolon; (c)
proximal colonic tissue from five patients (two with idiopathic megarectum).
Controls
Positive control material consisted of DNA prepared from human
foreskin fibroblast infected with CMV (strain C-81), varicella zoster virus (VZV; H-551) or herpes simplex virus type 1 (HSV 1;
TC-50). EBV was prepared from chronically infected B95-8 cells
(obtained from ECACC, Porton Down). Apart from EBV all were
clinical isolates routinely used in the Department of Virology at
St. Bartholomews Hospital. The CMV isolate was originally from
an aborted fetus with congenital CMV infection. VZV was from a
patient with herpes zoster, and the HSV 1 isolate was from a patient
with gingivostomatitis.
crofuge at 13 000 rpm for 10 min and the xylene discarded. The resulting pellet was rinsed in 95% ethanol and spun at 13 000 rpm for
a further 10 min. This was repeated with clean ethanol to ensure complete removal of the xylene. The pellet was dried in a heating block
at 45 C, resuspended in sterile distilled water and incubated with 0.5
mg/ml proteinase K at 37 C for 12 h. Phenol chloroform was added to the samples, which were mixed by inversion and centrifuged
at 13 000 rpm for 15 min. The phenol layer was discarded and 0.3 M
sodium acetate, 3 l of tRNA and ethanol was added to the aqueous
phase. This was mixed and left to precipitate at 70 C for 20 min.
The samples were spun for 15 min at 4 C at 13 000 rpm to pellet the
DNA. The pellet was washed in 70% ethanol, centrifuged for 5 min
at 13 000 rpm, dried, resuspended in sterile distilled water and boiled
for 15 min.
The PCR was performed as described by Saiki [9] using recombinant DNA polymerase from Thermus aquaticus (Amplitaq; Perkin-Elmer Cetus) according to the manufacturers protocol. All preparative work for the PCR was carried out in an ultraviolet-irradiated class II microbiological safety cabinet, using autoclaved plastic
tubes and pipette tips throughout.
Amplification reactions were carried out in a total volume of
100 l comprising 50 l test sample and 50 l reaction buffer in
0.6-ml microcentrifuge tubes. The reaction buffer was prepared in
bulk mastermix at double strength to give final concentrations of
50 mM KCl, 10 mM Tris-HCl (pH 8.3), 2.0 mM MgCl2, 0.01% gelatin, 0.2 mM of each deoxynucleoside triphosphate and 0.2 M of
each oligonucleotide primer and stored at 20 C.
The mastermix was thawed and the Amplitaq was added to give
a final concentration of 2.5 U/reaction tube. The reaction mixture
was dispensed into 0.6-ml microcentrifuge tubes and the test sample
was added, either neat or diluted with distilled water, using sterile
50-l glass capillaries. The reaction mixture was overlaid with mineral oil to prevent evaporation. Reaction tubes containing positive
and negative viral DNA control samples were included in each batch
of tests. Nested PCR for CMV, EBV, HSV 1 and VZV was performed
in a DNA thermal cycler (Perkin-Elmer Cetus).
The samples were amplified through 20 cycles consisting of denaturation at 94 C for 1 min, annealing at 60 C for 1 min and primer extension at 72 C for 2 min. Aliquots of 5 l of amplified products were added to 45 l of sterile distilled water and transferred in
a separate room to fresh reaction tubes containing standard mastermix with 2.5 units of Taq polymerase and 1.0 M of each of the nested primers. The reactions were then done according to the standard
protocol for 30 cycles.
Gel electrophoresis of the PCR products
Routine histology
Longitudinally orientated blocks of the full thickness of the bowel wall
were taken for processing into paraffin wax after fixation in 10% formal saline. Sections were cut at a thickness of 6 m and stained with
haematoxylin and eosin (H & E) and periodic acid Schiff (PAS).
Polymerase chain reaction
Paraffin-embedded blocks were cut at a thickness of 10 m and four
sections from each block were dewaxed with xylene, spun in a miTable 1 Details of three patients with idiopathic megarectum and
three with idiopathic megacolon
Idiopathic megarectum Idiopathic megacolon
Pat. 1 Pat. 2 Pat. 3
Pat. 1
Pat. 2 Pat. 3
12
20
42
40
34
37
Duration of
symptoms (years) 10
17
Age at operation
(years)
After PCR amplification the products were transferred to a separate

laboratory where they were collected with sterile 100-l glass capillaries and 8 l aliquots were mixed with 2 l of loading buffer
(0.25% bromophenol blue, 30% (w/v) glycerol in water). Ten microlitres of each mixture was electrophoresed through a 2% agarose gel
(Sigma, UK) at 100 volts. A sample of X174 RF DNA digested
by HaeIII (Gibco, BRL) was included on each gel as a DNA size
marker. After electrophoresis, the gels were stained with ethidium
bromide for 15 min at room temperature, destained with distilled water for 1530 min and visualised under a UV transilluminator (UVP
Inc, USA). The gels were photographed using a Polaroid DS-$ camera and Polaroid T667 film (Sigma).
Oligonucleotide primers
The oligonucleotide primers were custom synthesised using the
380 B or 391 DNA synthesizers (Applied Biosystems). The oligonucleotides were deprotected after synthesis, precipitated with ethanol
and used without further purification.
133
Routine histology
Validation of the PCR technique with control tissue would

suggest that although there may be morphological evidence
of neural damage this is not attributable to these herpes viruses. Evaluation for other viruses may be warranted.
Routine H & E staining did not show changes suggestive

of viral infection. PAS staining did not demonstrate inclusion bodies.
References
Results
PCR
No herpes virus DNA for the four viruses tested was identified in tissue from patients with either idiopathic megarectum or idiopathic megacolon. Positive results were obtained with the control material.
Discussion
Subtle abnormalities of the enteric nervous system are observed in patients with idiopathically dilated large bowel
[37]. We postulated that there may be an underlying viral aetiology in some of these patients, specifically a neurotropic effect of herpes viruses.
PAS staining did not demonstrate any inclusion bodies,
and no herpes virus DNA for EBV, HSV type 1, VZV or
CMV was detected using the PCR technique described.
1. Gattuso JM, Kamm MA (1997) Clinical features of idiopathic

megarectum and idiopathic megacolon. Gut 41: 9399
2. Preston DM, Lennard-Jones JE, Thomas BM (1985) Towards a
radiologic definition of idiopathic megacolon. Gastrointest Radiol 10: 167169
3. Palmer JA, McBirnie JE (1967) Atonic megacolon. Can J Surg
10: 1520
4. Smith B (1972) The neuropathology of the alimentary tract.
Arnold, London, pp 6567
5. Smith B, Grace RH, Todd IP (1977) Organic constipation in
adults. Br J Surg 64: 313314
6. Krishnamurthy S, Schuffler MD, Rohrmann CA, Pope CE (1985)
Severe idiopathic constipation is associated with a distinctive abnormality of the colonic myenteric plexus. Gastroenterology
88: 2634
7. Barnes PRH, Lennard-Jones JE, Hawley PR, Todd IP (1986)
Hirschsprungs disease and idiopathic megacolon in adults and
adolescents. Gut 27: 534541
8. Debinski HS, Khan G, Kangro HO, Jeffries D, Kamm MA (1997)
DNA viruses in the pathogenesis of sporadic chronic idiopathic
intestinal pseudo-obstruction. Gut 41: 100106
9. Saiki RK, Gelfand DH, Stoffell S, Scharf SJ, Higachi K, Horn
GT, Mullis KB, Erlich HA (1988) Primer directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239: 487491
D. Hrsch J. J. Kirsch E. Weihe
Elevated density and plasticity of nerve fibres in anal fissures
Abstract Neural proliferative processes are regarded as

a contributing factor in chronic inflammatory diseases and
chronic pain. To elucidate whether neural proliferations occur in tissues surrounding chronic anal fissures and in the
normal anal canal, the nerve fibre density was examined
with the pan-neural marker protein gene product 9.5 (PGP)
and the neural proliferative marker growth-associated
protein 43 (GAP) by immunohistochemistry. GAP-immunoreactive nerve fibres in the uninflamed anal canal were
distributed region specifically. The proportion of GAP-immunoreactive nerves in relation to the PGP-immunoreactive innervation exhibited regional differences. In tissue
sections of chronic anal fissures, a marked increase in the
density of PGP- and GAP-immunoreactive nerve fibres
was noted, and PGP- and GAP-immunopositive nerve fibres displayed a neuroma-like appearance. Image analysis
revealed that PGP- and GAP-immunoreactive innervation
represented an area fraction of 0.5% (0.49 0.052; mean
and SEM) and 0.1% (0.11 0.013) in the normal anal canal, respectively. In tissue sections of chronic anal fissures,
PGP- and GAP-immunostained nerve fibres represented
area fractions of 1.3% (1.32 0.12) and 0.6% (0.56 0.15),
respectively. The increases in PGP- and GAP-immunopositive area fractions were highly significant (P > 0.01). The
mean ratio of GAP to PGP immunoreactivities was not significantly increased in chronic anal fissures. The increase
in pan-neural innervation and neuronal GAP immunoreactivity in tissues of anal fissures may imply that neuronal
proliferation is involved in the pathogenesis of anal fissures. Neuronal proliferations may also be responsible for
pruritus and severe pain in chronic anal fissures.
D. Hrsch () E. Weihe
Institute of Anatomy and Cell Biology, Philipps University,
D-35033 Marburg, Germany
D. Hrsch
Department of Internal Medicine,
Division of Gastroenterology and Metabolism, Philipps University,
D-35033 Marburg, Germany
J. J. Kirsch
Center of Proctology, Clinic,
D-68159 Mannheim, Germany
Key words Anal fissures Neural plasticity

Chronic inflammation Pruritus Pain
Protein gene product 9.5 Growth associated protein-43
Rsum Des processus de prolifration neurale peuvent
tre considrs comme des facteurs dinflammation chronique et de douleurs chroniques. Afin de dterminer si des
prolifrations neurales surviennent dans le tissu entourant
les fissures anales chroniques en comparaison des
prlvements du canal anal normal, la densit des fibres
nerveuses a t examine avec la protine gnique marquant les nerfs 9.5 (PGP) et la protine 43 (GAP), associes
un marqueur de la prolifration neurale par immunohistochimie. Les fibres nerveuses immunoractives GAP
sont distribues en tissu anal non enflamm de manire
rgionale spcifique. La proportion de nerfs ragissant
GAP 43 en relation avec linnervation immunoractive
PGP montre des diffrences rgionales. Dans le tissu provenant de fissures anales chroniques, on note une lvation
prononce de la densit de fibres nerveuses immunoractives PGP et GAP et les fibres nerveuses immunopositives pour PGP et GAP prsentent un aspect de type
neurinome. Lanalyse des images montre que linnervation
immunoractive PGP et GAP reprsente une surface de
0.5% (0.49 0.052) et 0.1% (0.11 0.013) dans le canal
anal normal. Dans du tissu de fissures anales chroniques,
les fibres nerveuses colores immunohistochimiquement
pour PGP et GAP reprsentent des surfaces de 1.3%
(1.32 0.12) et 0.6% (0.56 0.15). Laugmentation des
fractions immunopositives pour PGP et GAP est hautement
significative (P > 0.01). La moyenne du rapport de GAP/
PGP nest pas significativement augmente en cas de fissures anales chroniques. Laugmentation de linnervation
et laugmentation de tissu immunoractif pour GAP dans
le tissu provenant des fissures anales peut signifier quune
prolifration neuronale est implique dans la pathognse
des fissures anales chroniques. La prolifration neuronale
peut aussi tre responsable du prurit et des douleurs aigus
en cas de fissures anales chroniques.
135
Introduction
Methods
The human anal canal is a complex organ and comprises

several histotopological compartments [1, 3]. The innervation of the anal canal is ubiquitously dense, especially
in the smooth muscle layers, around the anal glands and in
the anal transitional zone [2, 3]. It has been shown that neuropeptidergic nerve fibres are distributed region specifically in distinct histotopological zones [2]. Furthermore,
abundant neuroendocrine cells have been detected in the
epithelial layer of the anal canal and characterized by their
epithelial-type specific neurochemical phenotypes [4].
These data imply that anal functions such as continence
and defecation are regulated by interactive neural and neuroendocrine mechanisms [2, 5].
Anal disorders haemorrhoids, fistulas and fissures have
a high prevalence and incidence. They are characterized
by severe pain, pruritus and bleeding. Anal fissures appear
histopathologically as ulceration in the squamous zone of
the anal canal below the dentate line. Acute anal fissures
may develop into chronic anal fissures, a chronic inflammation, which is difficult to treat pharmacologically. Often, surgery is the ultimate treatment [69].
Various factors have been proposed to play a role in the
pathogenesis of chronic anal fissures, such as hypertrophy
and increased sphincter tone of the internal anal sphincter,
persistence of anal crypts and low perfusion of the posterior comissure of the anal canal, a region where fissures
normally develop [611]. In earlier studies, unusual concentrations of nerve fibres in the dermis of the squamous
zone were described [2, 12]. These nerve fibre concentrations were referred to as neuronal proliferations [12] or
areas of high nerve fibre density [2] and exhibit dense
immunoreactivity for proinflammatory neural messengers
such as substance P and calcitonin gene related peptide [2].
Proliferations of nerve fibres in distinct subsets of the peripheral nervous system have been shown to occur in
chronic inflammatory disorders, e.g. chronic pancreatitis,
chronic appendicitis and prurigo nodularis of the skin.
Hence, neuronal proliferations have been discussed as
pathogenic promotors of chronic inflammation and chronic
pain [1319]. In this context it is especially interesting that
increased immunoreactivity for the neuronal plasticity
marker growth-associated peptide 43 (GAP) or neuromodulin has been detected in proliferated nerve fibres of chronically inflammed pancreas and intestine. This implicates
elevated plasticity of the peripheral nervous system in
chronic inflammation [16, 19, 20].
In this context it was interesting for us to determine
whether areas of high nerve fibre density are upregulated
in chronic anal fissures by a neural proliferative process.
To mark all nerve fibres, we used an antiserum for the panneural marker protein gene product 9.5 (PGP) [2]. Here
we report a marked increase in PGP- and GAP-immunoreactive nerve fibres in chronic anal fissures. Parts of this
study have been previously published in abstract form
[21].
Patients and tissue processing

Control tissue specimens from the anal canal of adult patients (n = 6;
four men and two women; mean age 55 years) were obtained within 30 min in the course of abdomino-perianal resections for the therapy of colorectal carcinoma. Tissue specimens of chronic anal fissures (n = 6; four men and two women; mean age 49 years) were obtained in the course of surgery for long-standing anal fissures. Tissue samples of chronic anal fissure and controls were taken from the
dorsal comissure of the anal canal, fixed by immersion in Bouins
solution, acid-free Bouins solution or Bouin-Hollande for 24 h and
embedded in paraffin. All tissue specimens were independently assessed by a histopathologist. Squamous metaplasia was not detected in the specimens. Processing of the tissue specimens in the control group and in the anal fissure group was the same. Section thickness was controlled by staining every tenth tissue section with haematoxylin and eosin. Immunohistochemistry was performed on adjacent sections (57 m thick).
Nickel-enhanced streptavidin -biotin -peroxidase

immunohistochemistry
To reveal the distribution of PGP- and GAP-immunoreactive nerve
fibres nickel-enhanced streptavidin -biotin -peroxidase immunohistochemistry was used. Deparaffinized sections were incubated overnight at room temperature and for 2 h at 37 C with the appropriately diluted primary antisera. Rabbit antiserum against PGP (Ultraclone, Isle of Wight, UK) was applied in a dilution of 1 : 1000. The
murine monoclonal antibody against GAP (Boehringer Mannheim,
Germany) was applied in a dilution of 1 : 500. Immunoreactions were
visualized by biotinylated secondary antisera from donkey, preabsorbed against serum proteins of multiple species (Dianova/Jackson,
Hamburg, Germany) and horseradish-peroxidase complex (Amersham/Buchler, Braunschweig, Germany). Both systems were applied
in a dilution of 1 : 200. The diaminobenzidine (Sigma, Deisenhofen,
Germany) reaction was performed in 10 mM phosphate-buffered saline and enhanced by addition of 0.08% ammonium nickel sulphate
(Fluka, Buchs, Switzerland) [2, 4].
Double immunohistochemistry
Double labelling with fluorescence and enzymatic detection was
used to assess the coexistence of GAP- and PGP-immunoreactive
nerve fibres. The primary antisera were used at concentrations three
to five times greater for the fluorescence immunolabelling than for
streptavidin-biotin-peroxidase immunohistochemistry due to the
lower sensitivity of the immunofluorescence technique. Two different protocols were used, which proved to be equally sensitive and
specific. Dichlorotriazinyl amino fluorescein (DTAF) and Cy3 were
used as fluorochromes. In protocol 1, the primary mouse antibodywas detected by Cy3-labelled donkey anti-mouse IgG (diluted 1 : 50,
Jackson/Dianova, Hamburg, Germany), and the rabbit primary antisera were detected by DTAF-labelled donkey anti-rabbit IgG (diluted 1 : 25, Jackson/Dianova). In protocol 2, the primary murine antibody was detected by DTAF-labelled donkey anti-mouse IgG (diluted 1 : 25, Jackson/Dianova), and the rabbit primary antiserum by
Cy3-labelled donkey anti-rabbit IgG (diluted 1 : 50, Jackson/Dianova). The tissue sections were incubated with the detection systems
for 2 h at 37 C. Secondary antisera were preabsorbed against serum
proteins of several species including immunoglobulins of the donor
species of the primary antisera to avoid cross reactions. All detection systems were filtered with 0.45-nm disposable filter units (Millipore, Eschborn, Germany). Tests for unspecific binding of the secondary antisera included omission of the primary antibody/antiserum and the detection of the primary antibody/antiserum with noncorresponding detection systems [21]. There was no evidence for
cross reactions. The sections were examined and photographed with
136
a Zeiss axiophot microscope equipped with epi-illumination. To obtain DTAF fluorescence, a 450- to 490-nm filter was used for excitation and a 520-nm filter for the specific emission. A 546-nm filter
was used for Cy3 excitation and a 590-nm filter for the specific emission [2, 21].
Image analysis
The CUE-3 System (Galai Productions, Migdal Haemek, Israel) connected to an Olympus BH-2 microscope via a solid-state video camera (Sony, 3CCD) was used for comparative image analysis, which
was performed on adjacent sections stained for PGP and GAP. In
control sections of uninflamed anal canal, the dermis of the squamous zone and perianal epidermis was analysed because these regions were found to correspond to the main location of the inflammatory process in chronic anal fissures. The system was calibrated
for a 20 objective. All artefacts were excluded from the scanning
area. A threshold was defined for each individual section in order to
distinguish between specific immunostaining and nonspecific background staining. Sections were first scanned for PGP-immunoreactive nerve fibres, followed by an analysis of GAP-immunoreactive
nerve fibres in the same regions on consecutive sections. Hence, the
mean ratio of PGP to GAP staining could be assessed. For the dermis of the normal anal canal, 60 high-power fields were examined
to determine the density of PGP- and GAP-immunoreactive nerve
fibres. To assess the density of PGP- and GAP-immunoreactive innervation of anal fissures to fields were examined under high power. The density of immunoreactive nerve fibres was expressed as
area fraction. To compare the density of PGP- and GAP-immunoreactive innervation, the Wilcoxon rank order test was used for statistical analysis of all measured high-power fields. To compare the
GAP to PGP ratio, the unpaired t-test was used.
Results
Distribution of PGP- and GAP-immunoreactive

nerve fibres in the normal anal canal
In the colorectal zone of the anal canal, uniform and dense
PGP-immunoreactive nerve fibre plexus were detected in
the mucosal and submucosal plexus, in muscle layers (circular muscle, longitudinal muscle and muscularis mucosae), in the submucosa and in the mucosa. Neurons of the
enteric nerve plexus were strongly PGP-immunopositive.
In the anal transitional zone, PGP-immunostained nerve fibres were most prominent around the anal glands and ducts,
in the submucosa and in the cranial part of the internal anal
sphincter. In the squamous zone and perianal epidermis,
PGP-immunoreactive innervation was dense in the dermis
(Fig. 1 a), especially in areas of high nerve fibre density,
and around hair plexus and sweat glands. The epidermis
and the caudal internal anal sphincter were less densely
supplied by PGP-immunopositive nerve fibres. Nerve bundles and perivascular nerve fibre plexus were PGP immunoreactive in all zones and regions of the anal canal.
Schwann cells were not labelled by the PGP antiserum as
assessed by comparison with staining of Schwann cells for
S-100.
The nerve fibre population staining for GAP constituted
region-specific subpopulations of the pan-neural PGP-immunoreactive innervation. All regions supplied by PGPimmunoreactive nerve fibres were also supplied by GAP-
immunoreactive nerve fibres, although quantitative differences were obvious in the various regions. In the colorectal zone, GAP-immunoreactive nerve fibres were ubiquitous but less numerous than the PGP-immunopositive innervation. The density of GAP-immunoreactive nerve fibres decreased caudally towards the anal transitional zone,
the squamous zone and perianal epidermis. However, the
density of GAP-immunoreactive innervation of the internal anal sphincter, the musculus canalis ani and the longitudinal muscle was comparable to that of the colorectal
zone. In the dermis of the squamous zone and perianal epidermis, GAP-immunoreactive nerve fibres were contained
in a very minor portion of the pan-neural innervation revealed by PGP (Figs. 1 b, 2 a). In areas of high nerve fibre
density, the GAP-immunopositive innervation constituted
a minor fraction of the PGP-positive nerve fibres.
Changes in PGP- and GAP-immunoreactive innervation
in chronic anal fissures
Histological examination of tissue specimens of chronic
anal fissures revealed that the major inflammatory reaction was located in the dermis of the squamous zone and
perianal epidermis. Portions of the anal transitional zone
were rarely contained in tissue specimens of chronic anal
fissures and were excluded from the analysis. Chronic anal
fissures appeared as ulceration with missing or severely
damaged epidermal layer. All tissue specimens of chronic
anal fissures contained characteristic features of longstanding chronic inflammation, such as fibrosis and infiltrates of lymphocytes.
Changes in PGP- and GAP-immunoreactive innervation were seen in all patients suffering from chronic anal
fissures. In all tissue sections of chronic anal fissures, a
marked increase in the innervation density of PGP- and
GAP-immunoreactive nerve fibres was noted. The hyperinnervation in chronic anal fissures was visible as concentrations of nerve fibre bundles in the dermis (Fig. 3 a, b).
Nerve fibre bundles in chronic anal fissures were enlarged
and in the majority of analysed tissue sections had a neuroma-like appearance (Fig. 2 b). In contrast, the PGP- and
GAP-immunoreactive innervation of the epidermis, if
present in chronic anal fissures, was decreased compared
with control sections. Likewise, the portion of the ulcerations immediately adjacent to the anal canal was not innervated by PGP- or GAP-immunopositive nerve fibres. In
adjacent and identical sections of chronic anal fissures double immunohistochemistry revealed a proportional increase in the GAP-immunoreactive nerve fibre subpopulation compared with the pan-neural PGP-positive nerve
fibres in tissue sections of chronic anal fissures (Fig. 3 a, b).
Quantitative image analysis of PGPand GAP-immunoreactive nerve fibres
Image analysis of adjacent sections alternately stained for
PGP or GAP revealed that the PGP-immunoreactive inner-
137
Fig. 1 Distribution of PGP(A) and GAP- (B) immunoreactive nerve fibres in the dermis
of the uninflamed anal canal revealed by double immunofluorescence. The same section is
shown in A and B. GAP-immunoreactive nerve fibres (arrows
in B) constitute a minor subpopulation of the general innervation (arrows in A), as visualized by staining for PGP. Bar,
50 m
Fig. 2 A GAP-immunoreactive nerve fibres in the upper

dermis of the squamous zone of
the uninflamed anal canal
(arrows). e Epidermal layer.
B GAP immunoreactivity in a
neuroma-like formation in a tissue section of chronic anal fissure. Bar, 50 m
vation represented an area fraction of 0.5% (0.49 0.052;

mean SEM) in the dermis of the uninflamed anal canal.
GAP-immunoreactive nerve fibres were present in an
area fraction of 0.1% (0.11 0.013). In tissue sections of
anal fissures, PGP- and GAP-immunostained nerve fibres
represented an area fraction of 1.3% (1.32 0.12) and

0.6% (0.56 0.15), respectively. The increase in PGP- and
GAP-immunoreactive area fraction was highly significant
(P < 0.01) as calculated using the Wilcoxon rank order
test. For each patient, the mean GAP to PGP ratio was de-
138
Fig. 3 Almost complete colocalization of PGP (A) and GAP (B)

immunoreactivity in proliferated nerve fibres (arrows) in a tissue
section of chronic anal fissure. The same section is shown in A and
B. Double immunofluorescence was performed using DTAF (A) and
Cy3 (B) as fluorochromes. Bar, 50 m
termined, which reflects the proportion of GAP-immunoreactive nerve fibres in relation to the pan-neural PGPimmunoreactive area fraction. The mean ratio of GAP to
PGP immunoreactivity was not significantly increased in
chronic anal fissures.
Discussion
In this study, we established the region-specific distribution pattern of GAP-immunoreactive nerve fibres in the
normal anal canal and chronic anal fissures in relation to
the general innervation, as revealed by staining for the panneural marker PGP. In addition, we demonstrated a highly
significant increase in the PGP- and GAP-immunoreactive
innervation in chronic anal fissures as compared with control sections.
The distribution of PGP- and GAP-immunoreactive
nerve fibres in the uninflamed anal canal is in agreement
with previous studies [2, 22, 23]. Interestingly, the proportional distribution of GAP-immunoreactive nerve fibres
was region specific. In compartments innervated by the enteric nervous system, GAP-immunoreactive nerve fibres
constituted a higher proportion of the general innervation
than in compartments innervated by the somatic nervous
system. These differential distribution patterns may indicate elevated plasticity and synaptic remodelling in the enteric nervous system compared with the somatic nervous
system [19, 2224].
In previous studies we reported a very high density of
neuropeptidergic nerve fibres in the dermis of the squamous zone and in the perianal epidermis. This has been
discussed as a possible pathogenic factor for anorectal diseases [2]. In the present study we provide evidence for a
pronounced increase in nerve fibre density in chronic anal
fissures, reflected by an increase in the density of GAPand PGP-immunoreactive nerves. GAP is a member of a
group of brain acid-soluble proteins [2529] and is widely
distributed in the central and peripheral nervous system
[19, 22, 23]. A number of studies has substantiated the
role of GAP in neuronal proliferation and regeneration
[24, 3032]. Recent studies indicate a close association
between elevated levels of GAP in neurons of the peripheral nervous system and chronic inflammation [19, 20]. In
addition, GAP may function as a neuromodulator in the
neuronal release of classical and peptidergic transmitters
[33, 34]. Thus, the pronounced elevation of neuronal GAP
expression in tissue specimens of chronic anal fissures may
not only implicate neuronal proliferation and abortive
sprouting mechanisms in this chronic inflammation, but
also facilitate release of presumed proinflammatory neuronal messengers. Furthermore, the increased nerve fibre
densities observed in anal tissues surrounding chronic fissures might be the result of pathological processes, such
as increased sphincter tone or reduced blood flow [611]
and may thus be secondary in nature.
Nerve fibres in tissue sections of chronic anal fissures
often were neuroma-like in appearance. Neuroma formation occurs during denervation and subsequent loss of neuronal target tissue. Histologically, neuromas are characterized by disrupted perineurium and numerous growth cones.
Hyperalgesia has been found to be associated with neuroma formation [13, 15, 35]. Since neuromas have been
found to have an elevated neuronal discharge [36], neuroma formation in chronic anal fissures may explain the
severe pain that accompanies this chronic inflammation.
Furthermore, alterations in nerve fibre qualities have been
liked to pruritus [37]. In this context it appears likely that
neuroma formation in chronic anal fissures is involved in
the pruritus that accompanies this disorder [6, 9]. Recent
investigations of human chronic pancreatitis lend support
to the view that a possible causal relationship exists
between pain, neuronal plasticity of GAP expression and
immune cell infiltration [38].
Most intriguingly, the putative involvement of the peripheral nervous system in the pathogenesis of anal fissures
suggests new strategies for the therapy of chronic anal fis-
139
sures. Anal fissures are often unresponsive to treatment

with corticosteroids or nonsteroidal anti-inflammatory
drugs [611]. New concepts to treat chronic anal fissures
include the appliance of selective agents to relax the tone
of the internal anal sphincter, e.g. nitric oxide donors or
botulinum toxin [39, 40]. These agents act on the motoric
innervation. In addition, it was shown that application of
selective sensory neurotoxins, such as capsaicin, were able
to reduce chronic inflammatory reactions and pain [41, 42].
In order to find novel conservative treatments for chronic
anal fissures, it may be worthwhile to examine the ability
of capsaicin or capsaicin agonists as anti-inflammatory
agents to diminish inflammation, pain and pruritus.
Acknowledgements This study was supported by the BMFT and
the Deutsche Forschungsgemeinschaft (WE 910/2-3). The help of
B. Michel and R. Eissele (Department of Internal Medicine, Division of Gastroenterology and Metabolism, Philipps-Universitt,
Marburg, Germany) with image analysis is gratefully acknowledged.
References
1. Fenger C (1992) Anal canal. In: Sternberger S (ed) Histology
for pathologists. Raven, New York, pp 607623
2. Hrsch D, Fink T, Bchler M, Weihe E (1993) Regional specificities in the distribution, chemical phenotypes, and coexistence
patterns of neuropeptide containing nerve fibres in the human
anal canal. J Comp Neurol 335: 381401
3. Duthie HL, Gairns FW (1960) Sensory nerve-endings and sensation in the anal region of man. Br J Surg 206: 585595
4. Hrsch D, Fink T, Gke B, Arnold R, Bchler M, Weihe E
(1994) Distribution and chemical phenotypes of neuroendocrine
cells in the human anal canal. Regul Pept 54: 527542
5. Christiansen J, Lorentzen M, Holst J (1990) Influence of peptides on anorectal function. Ann Med 22: 413418
6. Brhl W (1982) tiologie und Therapie der Analfissur. Coloproctology 2: 114115
7. Kirsch JJ (1989) Out-patient treatment of haemorrhoids
risks and benefits. Results from 16,349 patients. Akt Chir
24: 253259
8. Brown AC, Sumfest JM, Rozwadowski JV (1989) Histopathology of the internal anal sphincter in chronic anal fissure. Dis
9. Reinhard F, Kirsch JJ (1993) Improved results in the treatment
of anal fissures by combined conservative and surgical therapy.
Coloproctology 6: 370373
10. Vogel P, Klosterhalfen B (1988) Die chirurgische Anatomie der
Rectum- und Analgefe. Langenbecks Arch Chir 373: 264269
11. Schouten WR, Briel JW, Auwerda JJA (1994) Relationship
between anal pressure and anodermal blood flow: the vascular
pathogenesis of anal fissures. Dis Colon Rectum 37: 664669
12. Fenger C, Schrder HD (1990) Neuronal hyperplasia in the anal
canal. Histopathology 16: 481485
13. Weihe E, Bchler M, Mller S, Friess H, Zentel HJ, Yanaihara
N (1990) Peptidergic innervation in chronic pancreatitis. In:
Berger HG, Ditschuneit H, Malfertheimer P (eds) Chronic pancreatitis. Springer, Berlin Heidelberg New York, pp 83105
14. Weihe E, Hartschuh W (1988) Multiple peptides in cutaneous
nerves: regulators under physiological conditions and a pathogenic role in skin disease? Semin Dermatol 7: 284300
15. Bchler M, Weihe E, Friess H, Malfertheimer P, Bockmann DE,
Mller S, Nohr D, Beger HG (1992) Changes in peptidergic innervation in chronic pancreatitis. Pancreas 7: 183192
16. Weihe E, Nohr D, Michel S; Mller S, Zentel H-J, Fink T,
Krekel J (1991) Molecular anatomy of the neuro-immune connection. Int J Neurosci 59: 123
17. Abadia Molina F, Burrows NP, Russell Jones R, Terenghi T,

Polak JM (1992) Increased sensory neuropeptides in nodular
prurigo: a quantitative immunohistochemical analysis. Br J Dermatol 127: 344351
18. Di Sebastiano P, Fink T, Friess H, Weihe E, Beger HG, Bchler M (1995) Changes of protein gene product 9.5 (PGP 9.5)
immunoreactive nerves in the inflamed appendix. Dig Dis Sci
40: 366372
19. Fink T, Di Sebastiano P, Bchler M, Beger HG, Weihe E (1994)
Growth-associated protein-43 and protein gene-product 9.5 innervation in human pancreas: changes in chronic pancreatitis.
Neuroscience 63: 249266
20. Stead RH, Kosecka-Janiszewska U, Oestreicher AB, Dixon MF,
Bienenstock J (1991) Remodeling of B-50 (GAP-43)- and NSEimmunoreactive mucosal nerves in the intestines of rats infected with nippostrongylus brasiliensis. J Neurosci 11: 3809
3821
21. Hrsch D, Weihe E (1994) Profuse peptidergic innervation of
the anal canal: a pathogenic factor for anorectal diseases? Gut
35 [Suppl 4]: A 12 (abstract)
22. Sharkey KA, Coggins PJ, Tetztlaff W, Zwiers H, Bisby MA,
Davison JS (1990) Distribution of growth-associated protein
B-50, (GAP-43), in the mammalian enteric nervous system.
Neuroscience 38: 1320
23. Fantini F, Johansson O (1992) Expression of growth associated
protein 43 and nerve growth factor receptor in human skin: a
comparative immunohistochemical investigation. J Invest Dermatol 99: 734742
24. Gispen WH, Nielander HB, De Graan PNE, Oestreicher AB,
Schrama LH, Schotman P (1991) Role of the growth associated
protein B-50 (GAP-43) in neuronal plasticity. Mol Neurobiol
5: 6185
25. Jacobsen DJ, Virag I, Skene JHP (1986) A protein associated
with axon growth, GAP-43, is widely distributed and developmentally regulated in the rat CNS. J Neurosci 6: 18431855
26. Oestreicher AB, Dekker LV, Gispen WH (1986) A radioimmunoassay for the phosphoprotein B-50: distribution in rat brain.
J Neurochem 46: 13661369
27. Basi GS, Jacobson RD, Virag I, Schilling J, Skene JHP (1987)
Primary structure and transcriptional regulation of GAP-43, a
protein associated with cell growth. Cell 49: 785791
28. Benowitz LI, Routtenberg A (1987) A membrane phosphoprotein associated with neural development, axonal regeneration,
phospholipid metabolism and synaptic plasticity. Trends Neurosci 10: 527532
29. Skene JHP (1989) Axonal growth associated proteins. Annu Rev
Neurosci 12: 127156
30. Chong MS, Fitzgerald M, Winter J, Hu-Tsai M, Emson PC,
Wiese U, Woolf CJ (1992) GAP-43 mRNA in rat spinal cord and
dorsal root ganglia neurons: developmental changes and reexpression following peripheral nerve injury. Eur J Neurosci
4: 883895
31. Mehta A, Reynolds ML, Woolf CJ (1993) Partial denervation of
the medial gastrocnemius muscle results in growth-associated
protein-43 immunoreactivity in sprouting axons and Schwann
cells. Neuroscience 57: 433442
32. Mearow KM, Kril Y, Gloster A, Diamond J (1994) Expression
of NGF receptor and GAP-43 mRNA in DRG neurons during
collateral sprouting and regeneration of dorsal cutaneous nerves.
J Neurobiol 25: 127142
33. Dekker LV, De Graan PNE, Oestreicher AB, Versteeg DHG, Gispen WH (1989) Inhibition of noradrenaline release by antibodies to B-50 (GAP-43). Nature 342: 7476
34. Hens JJH, Ghijsen WEJM, Dimjati W, Wiegant VM, Oestreicher AB, Gispen WH, De Graan PN (1993) Evidence for a role of
protein kinase C substrate B-50 (GAP-43) in Ca2+-induced neuropeptide cholecystokinin-8 release from permeated synaptosomes. J Neurochem 61: 602609
35. Carlton SM, Dougherty PM, Pover CM, Coggeshall RE
(1991) Neuroma formation and numbers of axons in a rat model of experimental peripheral neuropathy. Neurosci Lett 131:
8892
140
36. Devor M (1984) The pathophysiology and anatomy of damaged
nerve. In: Wall PD, Melzack R (eds) Textbook of pain. Churchill Livingstone, New York, pp 4964
37. Winkelmann RK (1988) Pruritus. Semin Dermatol 7: 233235
38. Di Sebastiano P, Fink T, Weihe E, Friess H, Innocenti P, Beger
HG, Bchler W (1997) Immune cell infiltration and growthassociated protein 43 expression correlate with pain in chronic
pancreatitis. Gastroenterology 112: 16481655
39. Jost WH, Schimrigk K (1994) Therapy of anal fissure using botulinium toxin. Dis Colon Rectum 37: 13211324
40. Lund JN, Scholefield JH (1997) Glyceryl trinitrate is an effective treatment for anal fissure. Dis Colon Rectum 40: 468470
41. Jansco G, Lynn B (1987) Possible role of capsaicin in pain therapy. Clin J Pain 3: 123126
42. Maggi CA (1991) Capsaicin and primary afferent neurons: from
basic science to human therapy. J Auton Nerv Syst 33: 114
U. Karlbom K. Edebol Eeg-Olofsson W. Graf

S. Nilsson L. Phlman
Paradoxical puborectalis contraction is associated

with impaired rectal evacuation
Abstract The role of paradoxical puborectalis contraction in the aetiology of constipation and how to best diagnose this condition is controversial. The aims of this study
were to investigate whether absolute or relative paradoxical electrical activity during electromyography (EMG) are
related to rectal emptying and to compare EMG, defecography and digital examination in the diagnosis of paradoxical puborectalis contraction. Included in the study were
171 consecutive patients with idiopathic constipation; 136
of these cases were also classified as paradoxical or unclear or not paradoxical at digital examination. Absolute
amplitudes and a strain/squeeze index were used to grade
the EMG activity in the puborectalis and external sphincter muscle. Rectal evacuation was analysed by defecography with image analysis of rectal area. The results showed
that 142 patients had paradoxical EMG activity during
straining. There was a correlation between rectal evacuation and amplitudes (r = 0.20 to 0.03, P < 0.01) and
between evacuation and index (r = 0.34 to 0.39,
P < 0.0001). Forty-two patients with an index of >50 had
impaired rectal evacuation compared with those with an
index 50 (P < 0.0001). Thirty-three of 34 cases (n = 136)
with an index of > 50 also were paradoxical at defecography whereas 19 were diagnosed digitally. In conclusion,
paradoxical puborectalis contraction is associated with impaired rectal evacuation. The activity seems to be best reflected by a strain/squeeze index. The best correlation in
diagnostic methods was between EMG and defecography.
Key words Anismus Constipation Electromyography
Defecography Paradoxical puborectalis contraction
U. Karlbom () W. Graf L. Phlman
Department of Surgery, University Hospital,
S-75185 Uppsala, Sweden
Rsum Le rle dune contraction paradoxale de la

sangle puborectale dans ltiologie de la constipation et le
meilleur moyen pour diagnostiquer cette condition sont sujets controverse. Le but de cette tude est de dterminer
se une activit lectrique paradoxale absolute ou relative
sur llectromyographie peut tre mise en relation avec
lvacuation rectale et de comparer llectromyographie
la dfcographie et au toucher rectal dans le diagnostic de
la contraction paradoxale de la sangle puborectale. 171 patients conscutifs avec une constipation idiopathique ont
t tudis. 136 de ces patients ont t classs comme prsentant une contraction soit paradoxale soit spcifique soit
non paradoxale lors du toucher rectal. Lamplitude absolue et un index contraction/pousse ont t utiliss pour
graduer lactivit lectromyographique dans la sangle puborectale et dans le sphincter externe. Lvacuation rectale
a t analyse sur la dfcographie par mesure de la surface rectale. Les rsultats ont montr que 142 patients avaient une activit paradoxale lectromyographique durant la
pousse. Il y a une corrlation entre lvacuation rectale et
lamplitude (r = 0.2023, P < 0.01) en entre lvacuation
rectale et lindex dactivit (r = 0.3439, P < 0.0001). 42
patients avec un index suprieur 50 avaient une vacuation rectale entrave en comparaison ceux avec un index
plus petit ou gal 50 (P < 0.0001). 33 des 34 patients
(n = 136) avec un index suprieur 50 avaient galement
une contraction paradoxale sur la dfcographie alors que
19 seulement taient diagnostiqus au toucher rectal. En
conclusion, la contraction paradoxale de la sangle puborectale est associe avec une vacuation incomplte. Lactivit semble tre le mieux apprcie par un index contraction/pousse. La meilleure corrlation dans lapproche
diagnostique est obtenue entre llectromyographie et la
dfcographie.
K. Edebol Eeg-Olofsson
Department of Neurophysiology, University Hospital,
Uppsala, Sweden
Introduction
S. Nilsson
Department of Radiology, University Hospital,
Uppsala, Sweden
Paradoxical puborectalis contraction (anismus, spastic pelvic floor syndrome) has been discussed as a pathogenic
142
factor in constipation, whereby a contracting or nonrelaxing puborectalis muscle and external sphincter muscle
would cause an outlet obstruction during attempted rectal
emptying. The paradoxical contraction has been associated
with increased electromyographie (EMG) activity [1] and
anal pressure during straining [2], inability to expel a rectal balloon [3], and prolonged segmental colonic transit
[4]. Failure to increase the anorectal angle [5] or a distinct
impression of the puborectalis muscle during evacuation
have been used as diagnostic criteria on defecography [6].
These criteria have also been associated with impaired rectal evacuation at defecography [5, 6].
Sphincter EMG has been considered a sensitive method
detect paradoxical puborectalis contraction [7] and is used
in many centres in the work-up of severe constipation. In
recent years, biofeedback training has been used most in
the treatment of patients with a paradoxical sphincter response. However, the functional importance of these findings has been questioned, since it has been observed in
other anorectal conditions [8] and in healthy subjects
[9, 10]. It is also questionable whether the laboratory setting is suitable for the evaluation of paradoxical sphincter
contraction. Measurements with ambulatory equipment in
the home environment have led to a reduction in the proportion of patients with paradoxical EMG activity [11].
However, biofeedback training based upon laboratory test
results and directed towards relaxing the sphincter muscles has been successful in several studies [1214] and also
associated with a reduction in sphincter activity during
straining [15]. If paradoxical puborectalis contraction is
clinically relevant, it would affect rectal emptying; however, the relationship between paradoxical EMG activity
and rectal emptying is unclear.
The principal aim of this study was to investigate
whether paradoxical EMG activity is related to rectal evacuation and to evaluate two different grading systems of paradoxical activity. The second aim was to compare EMG,
defecography and digital examination in the diagnosis of
paradoxical puborectalis contraction.
lapse in one, rectopexy in five, rectopexy and sigmoid resection in

five and rectocele repair in five patients). Twenty-one patients had
a previous hysterectomy.
The work-up included a rigid proctoscopy and a barium enema
or colonoscopy. The specific investigations included defecography,
colonic transit time [16], EMG and anorectal manovolumetry [17].
Fifty-nine patients had prolonged colonic transit (exceeding the 95th
percentile of controls), and 20 of them had markedly slow transit (retaining 90% of the markers after 7 days). The rectoanal inhibition
reflex could be elicited in all patients.
Electromyography
Integrated EMG was recorded with hook electrodes (isolated lacquered steel wire, WE-100, Life-Tech, Houston, Texas, USA) in the
external sphincter and puborectalis muscle. The electrodes were inserted posterolaterally on each side of the midline, 10 mm from the
anal verge to a depth of about 15 mm in the external sphincter
muscle. The corresponding distances in the puborectalis muscle were
12 and 35 mm, respectively. Both electrodes in each muscle were active. The muscles were examined sequentially; first the external
sphincter was recorded in the left lateral position followed by the puborectalis muscle in the same position. The patient was then seated
on a commode and a new recording was done in a similar manner.
The puborectalis muscle was also examined while the subject was in
the sitting position trying to expel a balloon (Foley catheter Ch 18
filled with 40 ml body-tempered water).
The recordings were performed with a standard EMG equipment
(Neuromatic 2000, Dantec, Skovlunde, Denmark) and graphically
displayed on a printset with a time constant of 50 ms. The registration represented the sum of recruitment of motor potential and the
size of recruited motor units. The EMG was measured in millivolts
(mV) relative to a baseline representing activity at rest. Negative values were not registered. A strain/squeeze index (100 strain amplitude/squeeze amplitude) was calculated for each muscle and position, thus yielding five indices for each patient. The mean of the five
indices and the peak index were used to grade the paradoxical sphincter activity. An index level above 50 has previously been used as a
criterion of significant paradoxical activity [12] and this level was
also used in this study. Mean and peak amplitudes were also calculated in a corresponding manner. The cut-off level for pathologic amplitudes was adjusted to create groups of similar numbers of patients
when comparing indices and amplitudes. All EMGs were done by
one investigator and analysed without knowledge of other test results.
Defecography
Materials and methods

Patients
A consecutive series of 171 patients with constipation referred to the
Department of Surgery, University Hospital, Uppsala, during the
time period 19921997 constituted the study population. The median age was 51 (range 2093) years and 152 patients (89%) were
women. The symptom duration varied between 0.5 and 60 years (median 10 years). Patients with neurological, connective tissue or active proctological disease were not included. Each patient had tried
fibre supplements or other bulking agents without satisfactory results. At the first visit 24 patients did not use any laxatives regularly while 113 used bulking agents, 66 motorstimulants and 57 enemas. Manually assisted defecation was used by 109 patients. Thirty-four patients presented with infrequent defecation only, whereas
the rest stated various degrees of emptying difficulties. There was a
mixed symptomatology of both infrequent defecation and emptying
difficulties in 51 patients. Seventeen patients had undergone surgery
related to constipation (colectomy in one, anterior resection for pro-
Our routine for defecography have been previously described in detail [6]. Briefly, the anorectal angle was measured as the angle
between the axis of the anal canal and the posterior border of rectum. The size of a rectocele was calculated during straining as the
distance between the maximal anterior outbulge and the axis of the
anal canal. The length of a circular intussusception and the position
of the anorectal junction relative to the ischial tuberosities was measured. Perineal descent was defined as the change in position of the
anorectal junction during straining compared with rest. The defecographic evaluations were performed blindly and X-ray magnification was corrected for in all measurements.
A paradoxical puborectalis contraction was diagnosed when there
was a marked impression of the puborectalis muscle and/or a failure
to increase the anorectal angle during straining compared with rest.
Rectal emptying was evaluated with a computer-based image analysis method [6, 18]. The area with homogeneous contrast in the lower 8 cm of rectum was calculated at rest, after initial evacuation (initial or first straining episode, 0 30 s) and after the total evacuation
period. The time was noted and rectal emptying was expressed as:
(1) percentage evacuated area per second during the initial evacuation; (2) percentage evacuated area per second during the total evacuation period; (3) percentage evacuated area.
143
Fig. 1 Distribution of EMG activity during straining measured with

a strain/squeeze index in 171 constipated patients
Digital examination
One hundred and thirty-six of the patients were prospectively assessed with a digital examination according to a protocol. The examination was performed with the patient in the left lateral position.
After identifying the puborectalis and external anal sphincter muscles during squeeze and relaxation, the patient was instructed to strain
as to evacuate. The procedure was repeated at least three times. The
muscle activity during straining was classified as: (1) paradoxical
sphincter contraction; (2) unclear; or (3) not paradoxical. All digital
examinations were performed by three surgeons with a special interest in the field.
Statistical methods
Nonparametrical methods were used. The Kruskall-Wallis test was
used for comparison of several independent groups of patients and
Mann-Whitney U test was used when comparing two groups of patients. Proportions were analysed with Fishers exact test.
Spearmans rank correlation test was used for analysis of correlations.
In a comparison of diagnostic methods, sensitivity was calculated as (true positive/true positive + false negative) 100; specificity
as (true negative/true negative + false positive) 100; positive predictive value as (true positive/true positive + false positive) 100;
and negative predictive value as (true negative/true negative + false
negative) 100. The assumed best method was used to compare
the three diagnostic methods.
Results
Electromyography
Paradoxical activity was found in 142 patients (83%) i.e.
an increment over baseline, during stain. There were 68
patients (39%) with a peak index > 50 and 42 patients (25%)
with a mean index > 50 (Fig. 1). The corresponding amplitude levels, discriminating equal number of patients, were
> 0.9 mV (n = 68) and > 0.76 mV (n = 42). There was an
overall correlation between amplitudes and rectal emptying, but the correlation was stronger between indices and
emptying (Table 1). There was also an inverse relation
Table 1 Overall correlations between EMG activity during straining (amplitude and index) and rectal evacuation measured at defecography in 171 constipated patients
Amplitude (mV)
Rectal evacuation
% Area evacuated
%/s (initial)
%/s (total)
Index
Mean
Peak
Mean
Peak
r = 0.23, P < 0.01

r = 0.20, P < 0.01
r = 0.22, P < 0.01
r = 0.18, P < 0.05

r = 0.04, P = ns
r = 0.16, P < 0.05
r = 0.39, P < 0.00001

r = 0.34, P < 0.00001
r = 0.39, P < 0.00001
r = 0.39, P < 0.00001

r = 0.36, P < 0.00001
r = 0.39, P < 0.00001
Correlations were calculated using Spearmans rank correlation test
144
Table 2 Defecographic findings and rectal evacuation according to sphincter muscle activity during straining measured
as mean amplitude or mean index levels at EMG (n = 171)
Mean amplitude (mV)
Anorectal angles
Rest
Strain
Perineal descent (cm)
Mean index
> 0.76
(n = 42)
0.76
(n = 129)
> 50
(n = 42)
50
(n = 129
97 (68 135)
119 (53 150)
100 (44 143)

125 (25 160
95 (44 124) a
92 (28 138) b
101 (57 143)

128 (25 160)
2.3 (0 4.8)
2.5 (0 5.7)
2.0 (0 4.9) a
2.7 (0 5.7)
Intussusception (cm)
0 (0 3.6)
0.7 (0 3.9)
0 (0 2.1)
Rectocele (cm)
2.5 (0 3.9)
2.1 (0 5.1)
2.4 (0 4.2)
2.1 (0 5.1)
74 (0 100)
2.5 (0 12.5)
1.3 (0 12.5)
83 (0 100)
4.2 (0 16.6)
2.1 (0 16.6)
47 (0 100) b
1.2 (0 10.3) b
0.6 (0 5.6) b
90 (0 100)
5.2 (0 16.6)
2.7 (0 16.6)
Rectal evacuation
% Area evacuated
%/s (initial)
%/s (total)
0.9 (0 3.6)
Values are median and range; a P < 0.05, b P < 0.001, Mann-Whitney U test
between indices (mean and peak) and length of intussusception (r = 0.23 to 0.22, P< 0.01), perineal descent
(r = 0.25 to 0.22, P< 0.01) and anorectal angle during
straining (r = 0.39 to 0.37, P< 0.001). The anorectal angle during straining was also inversely related to mean amplitude (r = 0.20 P< 0.05), otherwise there were no statistically significant relations between amplitudes and defecographic findings. Those with a mean index > 50 had significantly impaired rectal emptying, less intussusception,
perineal descent and more acute anorectal angles at rest
and at straining, whereas no particular defecographic features were found in those with the highest mean amplitudes
(Table 2). In a corresponding analysis, peak index levels
> 50 related significantly to the same defecographic parameters, whereas peak amplitudes (> 0.9) did not (data not
shown). Expressing the paradoxical puborectalis contraction as index levels rather than amplitudes was more in
agreement with the defecographic diagnosis of paradoxical contraction (proportion positive at defecography: mean
index > 50, 41/42 vs. mean amplitude > 0.76, 19/42;
P < 0.0001, Fishers exact test).
Table 3 Rectal evacuation and results of EMG in patients with and

without a paradoxical puborectalis contraction at defecography
Paradoxical
(n = 50)
Not paradoxical
(n = 121)
Rectal evacuation
% Evacuated area
%/s (initial)
%/s (total)
48 (0 100)
1.1 (0 7.2)
0.7 (0 2.0)
92 (10 100) a
5.6 (0 16.6) a
2.9 (0.1 16.6) a
EMG
Mean index
Peak index
Mean amplitude
Peak amplitude
56 (4 306)
85 (10 420)
0.7 (0.05 3.7)
1.0 (0.1 6.7)
16 (0 67) a
30 (0 129) a
0.2 (0 5.9) a
0.5 (0 17.8) a
Values are median and range; a P < 0.001, Mann-Whitney U test
ing could be explained by a nonemptying rectocele in

seven, whereas there was no clear reason for the poor emptying in one patient.
Digital examination
Defecography
A paradoxical puborectalis contraction was diagnosed in
50 patients (29%). The diagnosis was related to a short circular intussusception (P < 0.00001), little perineal descent
(P < 0.008) and an acute anorectal angle at rest (P < 0.002,
data not shown). Paradoxical puborectalis contraction was
also highly related to all rectal evacuation parameters,
EMG amplitudes and EMG indices (Table 3).
Fifty-four patients evacuated all contrast whereas five
patients could not evacuate at all. All patients in the latter
group had a paradoxical puborectalis contraction. Twenty
patients did not evacuate anything during the first 30 s
(initial %/s = 0) and 18 of them were diagnosed as having
a paradoxical contraction at defecography. Of the 25 patients that evacuated less than 0.5%/s during the whole investigation, there were 17 with a paradoxical puborectalis
contraction. In the remaining eight cases the poor empty-
Of 136 patients, 31 (23%) were judged to have a paradoxical puborectalis contraction at rectal examination, and
15 cases (11%) were classified as unclear. The digital examination separated patients with higher indices and
higher amplitudes (Table 4). A digitally diagnosed paradoxical contraction was associated with the same defecographic findings as a paradoxical contraction diagnosed
with EMG indices or with defecography (Fig. 2, Table 4).
Comparison of diagnostic methods
A comparison between diagnostic methods was made
undertaken in the 136 patients who were prospectively assessed by EMG, defecography and digital examination.
The proportions of patients with a paradoxical puborectalis contraction were: defecography 30%, EMG (mean index > 50) 25%, and at digital examination (clearly para-
145
Table 4 Digital classification
in relation to results of EMG
and defecography (n = 136)
Anorectal angles
Rest
Strain
Paradoxical
(n = 31)
Unclear
(n = 15)
Not paradoxical
(n = 90)
Kruskall-Wallis
P value
91 (56 120)
88 (28 150)
100 (74 133)

134 (25 134)
102 (44 137)

126 (37 160)
0.008
< 0.0001
Perineal descent (cm)
1.9 (0 4.2)
3.2 (0.6 5.7)
2.7 (0 5.1)
Intussusception (cm)
0 (0 1.8)
1.2 (0 2.7)
0.6 (0 3.6)
0.0009
Rectocele (cm)
1.8 (0 4.2)
2.4 (0 4.8)
2.3 (0 5.1)
0.58
53 (0 306)
0.7 (0 3.7)
37 (0 94)
0.4 (0 1.2)
17 (0 81)
0.3 (0 5.0)
< 0.0001
0.003
EMG
Mean index
Mean amplitude
0.017
Values are median and range

Table 5 The sensitivity, specificity and positive and negative
predictive values of EMG, defecography and digital examination versus and assumed best
method in the diagnosis of paradoxical puborectalis contraction
Reference method
Sensitivity
Specificity
Positive
predictive
value
Negative
predictive
value
Defecography a
EMG (mean index > 50)
Digital examination
80
54
99
91
97
71
92
82
EMG (mean index > 50) a

Defecography
Digital examination
97
56
92
88
80
61
99
86
Digital examination a
Defecography
EMG (mean index > 50)
71
61
82
86
54
56
91
88
Values are percentages; a assumed best method
Fig. 3 Distribution of patients diagnosed by defecography, EMG

(mean index > 50) and digital examination as having a paradoxical
puborectalis contraction (n = 136)
Fig. 2 Relationships between rectal evacuation and digital classification of paradoxical puborectalis contraction in 136 patients
(Kruskall-Wallis test P < 0.001)
doxical) 23%. Nineteen patients were diagnosed as having

a paradoxical contraction with all three methods and 17
with two methods, while 14 were diagnosed with one of
the methods (Fig. 3).
Sensitivity, specificity and positive and negative predictive values for each method are depicted in Table 5. In
these comparisons each method is also taken as a reference
method (assumed best method). Sensitivity and negative
predictive value for digital examination increased to 66%
and 84% (vs defecography) and 68% and 88% (vs EMG),
respectively, if patients with an unclear muscle activity
were included in the paradoxical group. Correspondingly,
specificity and positive predictive value decreased to 80%
and 59% (vs defecography) and to 77% and 50% (vs EMG),
respectively.
Discussion
The high incidence of patients with a paradoxical sphincter activity at EMG during straining and the wide range in
146
amplitudes suggest that grading needs to be more detailed

instead of just considering an increase as a pathological response. The observation of a paradoxical puborectalis contraction in healthy subjects and in patients with various anorectal conditions [810] supports this view. Use of absolute amplitudes for grading makes the placement of electrodes crucial; a suboptimal positioning might give false
low amplitudes. The electrode placement is corrected for
when using the strain/squeeze index. The muscle strength,
another potential source or error, is also adjusted for with
the use of an index. In the present study, indices had a closer
overall relation than amplitudes to all rectal evacuation parameters. A comparison between a number of patients with
an index cut-off level of 50, which has been used previously [12], and the same number of patients with the highest amplitudes showed that amplitudes were not as clearly
related to evacuation as mean and peak index. Furthermore,
the index levels were related to other defecographic findings such as more acute anorectal angles and less perineal
descent. There was also a close correlation between a defecographic diagnosis of paradoxical puborectalis contraction and high index levels. A corresponding correlation
was not found regarding amplitudes.
The incidence of defecographic paradoxical contraction
was 30%, which is higher than in a previous study [6]. This
might be explained by increased referrals of patients with
outlet obstruction and by the fact that failure to increase
the anorectal angle during straining was accepted as a diagnostic criterion in this study. A prominent impression of
puborectalis is not alway present or detectable in patients
with a more vertically positioned rectum even if the anorectal angle during strain actually decreases. The incidence
of paradoxical puborectalis contraction in other consecutive series of constipated patients has ranged between 29%
and 37% [1921].
Intussusception has been associated with the solitary
ulcer syndrome in which paradoxical puborectalis contraction has also been implicated as a contributing factor
[22]. In the present study, a pronounced intussusception
was uncommon with paradoxical puborectalis contraction, irrespective of whether diagnosis was made with defecography, EMG or digital examination. It seems logical that contracting sphincter muscles cause a reduced
mobility of the perineum, and if the contraction grossly
impairs rectal emptying, this will conteract intussusception, since intussusception at defecography usually implies some emptying. A rectocele has also been indicated
as a cause of impaired rectal emptying. The size of the
rectocele did not differ between patients with or without
paradoxical puborectalis contraction, suggesting that rectocele formation is independent of paradoxical contraction.
From these results it is concluded that paradoxical
sphincter activity at EMG is related to rectal evacuation
and that this is more accurately reflected by the use of a
strain/ squeeze index than by absolute amplitudes. Both
mean and peak index also correlated better than amplitudes
with the diagnosis of paradoxical puborectalis contraction
at defecography.
A paradoxical puborectalis contraction assessed by digital examination was related to higher indices and also to
higher amplitudes, which makes the digital assessment
clinically relevant. It was also clearly related to impaired
rectal evacuation. It has been suggested that clinical examination is sufficient in most patients with defecatory difficulties [23]. In this study, 15 patients (11%) were difficult
to classify because of varying contractionrelaxation patterns or not fully relaxing muscles during straining. There
were also two patients with weak muscles and were therefore difficult to categorise. However, an unclear diagnosis
might also have some relevance since patients with this
finding had intermediate levels of indices and rectal evacuation.
Although both defecography and EMG are established
methods, none of them can act as a reference method. All
comparisons must be done with the assumption that each
method could be the best. Poor agreement between defecography and EMG has been reported [24, 25]; however,
only a few studies have focused on the methods of diagnosing paradoxical puborectalis contraction. Jorge et al.
[26] found suboptimal correlations between defecography
and EMG when defining the paradoxical contraction as a
failure to achieve a descrease in electrical activity during
attempted defecation. They found a sensitivity and specificity of about 70% and 80%, respectively, for each method.
With the use of a strain/squeeze index (mean index > 50)
as in the present study, the sensitivity and specificity was
improved. Correlations between manometry, EMG and defecography in the diagnosis of paradoxical puborectalis
contraction, which have been studied by Ger et al. [21],
were found to be suboptimal but improved with the use of
a manometric strain/squeeze index.
The agreement between digital examination and
EMG/defecography was not as good as between EMG and
defecography. Embarressment at first visit may explain
some of the false positive results at digital examination.
Siproudis et al. [23] found a negative predictive value of
96% for clinical diagnosis of anismus vs manometric anismus. The corresponding figures in this study were 82%
(vs defecography) and 86% (vs EMG). If unclear diagnoses
were included in the paradoxical group the figures were
84% and 88%, respectively. The results suggest that a finding of normal relaxing sphincter muscles at digital examination can be used to exclude the diagnosis of paradoxical puborectalis contraction in most patients. A positive
diagnosis or an unclear finding requires further evaluation
with other methods.
References
1. Turnbull GK, Lennard-Jones JE, Bartram CI (1986) Failure of
rectal expulsion as a cause of constipation: why fibre and laxatives sometime fail. Lancet: 767769
2. Preston DM, Lennard-Jones JE (1985) Anismus in chronic constipation. Dig Dis Sci 30: 413418
3. Barnes PRH, Lennard-Jones JE (1985) Balloon expulsion from
rectum in constipation of different types. Gut 26: 10491052
147
4. Kuijpers HC, Bleijenberg G, Moree (1986) The spastic pelvic
floor syndrome. Large bowel outlet obstruction caused by pelvic floor dysfunction: a radiological study. Int J Colorectal Dis
1: 4448
5. Papachrysostomou M, Smith AN, Merrick MV (1994) Obstructive defecation and slow transit constipation: the proctographic
parameters. Int J Colorectal Dis 9: 115120
6. Karlbom U, Nilsson S, Phlman L, Graf W (1995) The relationships between defecographic findings, rectal evacuation and colonic tranisit in constipated patients. Gut 36: 907912
7. Wexner SD, Marchetti F, Salanga VD, Corredor C, Jagelman DG
(1991) Neurophysiology assessment of the anal sphincters. Dis
8. Jones PN, Lubowski DZ, Swash M, Henry MM (1987) Is paradoxical contraction of puborectalis muscle of functional importance? Dis Colon Rectum 30: 667670
9. Barnes PRH, Lennard-Jones JE (1988) Function of striated anal
sphincter during straining in control subjects and constipated patients with a radiologically normal rectum or idiopathic megacolon. Int J Colorectal Dis 3: 207209
10. Lubowski DZ, King DW, Finlay IG (1992) Electromyography
of the pubococygeus muscles in patients with obstructed defecation. Int J Colorectal Dis 7: 184187
11. Duthie GS, Bartolo DCC (1992) Anismus: the cause of constipation? Results of investigation and treatment. World J Surg
16: 831835
12. Dahl J, Lindquist B, Tysk C, Leissner P, Philipson L, Jrnerot G
(1991) Behavioral medicine treatment in chronic constipation
with paradoxical anal sphincter contraction. Dis Colon Rectum
34: 769776
13. Wexner SD, Cheape JD, Jorge JMN, Heymen S, Jagelman DG
(1992) Prospective assessment of biofeedback for treatment of
paradoxical puborectalis contraction. Dis Colon Rectum
35: 145150
14. Fleshman JW, Dreznik Z, Meyer K, Fry RD, Carney R, Kodner
IJ (1992) Outpatient protocol for biofeedback therapy of pelvic
floor outlet obstruction. Dis Colon Rectum 34: 17
15. Kawimbe BM, Papachrysostomou M, Binnie NR, Clare N,
Smith AN (1991) Outlet obstruction constipation (anismus)
managed by biofeedback. Gut 32: 11751179
16. Abrahamsson H, Antov S, Bosaeus I (1988) Gastrointestinal and

colonic segmental transit time evaluated by a single abdominal
x-ray in healthy subjects and constipated patients. Scand J Gastroenterol 23 [Suppl 152]: 7280
17. Holmberg A, Graf W, sterberg A, Phlman L (1995) Anorectal manovolumetry in the diagnosis of fecal incontinence. Dis
18. Karlbom U, Graf W, Nilsson S, Phlman L (1996) Does surgical repair of a rectocele improve rectal emtying? Dis Colon Rectum 39: 12961302
19. Turnbull GK, Bartram CI, Lennard-Jones JE (1988) Radiologic
studies of rectal evacuation in adults with idiopathic constipation. Dis Colon Rectum 31: 190197
20. Siproudhis L, Ropert A, Lucas J, Raoul JL, Heresbach D, Bretagne JF, Gosselin M (1992) Defecatory disorders, anorectal and
pelvic floor dysfunction: a polygamy? Int J Colorect Dis
7: 102107
21. Ger GC, Wexner SD, Jorge JMN, Salanga VD (1993) Anorectal manometry in the diagnosis of paradoxical syndrome. Dis
22. Womack NR, Williams NS, Holmfield JH, Morrison JF (1987)
Anorectal function in the solitary ulcer syndrome. Dis Colon
Rectum 30: 319323
23. Siproudhis L, Ropert A, Vilotte J, Bretagne JF, Heresbach D,
Raoul JL, Gosselin M (1993) How accurate is clinical examination in diagnosing and quantifying pelvirectal disorders? A prospective study in a group of 50 patients complaining of defecatory diffuiculties. Dis Colon Rectum 36: 430438
24. Miller R, Duthie GS, Bartolo DCC, Roe AM, Locke-Edmunds,
Mortensen NJM (1991) Anismus in patients with normal and
slow transit constipation. Br J Surg 78: 690692
25. Johansson C, Ihre T, Holmstrm B, Nordstrm E, Dolk A, Broden G (1990) A combined electromyographic and cineradiologic investigation in patients with defecation disorders. Dis Colon
Rectum 33: 10091013
26. Jorge JMN, Wexner SD, Ger GC, Salanga VD, Nogueras JJ, Jagelman DG (1993) Cineradiography and electromyography in the
diagnosis of nonrelaxing puborectalis syndrome. Dis Colon Rectum 36: 668676
L E T T E R TO T H E E D I TO R
A place for the pascal in anorectal physiology?
Dear Sir,
Amongst coloproctologists there is
no consensus on the optimal unit to
express anorectal luminal pressures.
American and continental European
workers tend to use millimetres of
mercury whereas their British counterparts prefer centimetres of water.
To illustrate this, at the ECCP meeting in June 1997 nine research groups
(six European, two North American
and one United Kingdom) presented
13 abstracts with anorectal pressures
measured in millimetres of mercury.
In contrast, five other groups presented five abstracts (one European,
four United Kingdom) in centimetres
of water.
Clearly there is a need to adopt
one unit of pressure measurement,
but which is optimal? It is difficult to
be certain, but the likely answer is
that the pascal (Pa) (1 newton/square
metre) is superior to either of the tra-
ditional units currently used. The pascal is the unit of pressure for the
system internationale (SI system).
The SI system is coherent. This means
that any parameter within the system
can be entered into a calculation with
any other parameter within the system
without the need to use mathematical
constants.
Further advantages include the
wide geographical understanding of
the SI system throughout the world
and its increasing utilization within
everyday medicine, for example, in
the expression of blood gas analysis.
Commercially available pressure
manometers are currently calibrated
in traditional units, but it would be
a relatively simple and inexpensive
step to recalibrate in kPa. In order
to field test the use of Pa in the
anorectal physiology laboratory, we
have carried out 30 consecutive
anal pressure measurements expressed in millimetres of mercury and
converted these to kPa using the formula:

mmHg
kPa
1 mmHg = 0.13 kPa .
(cm H2O kPa

1 cm H2O = 0.10 kPa)
We have had no problems with this
system of expressing anal or rectal
pressures. The time taken to do this is
only a few seconds.
In conclusion, one unit of pressure
measurement for all anorectal physiology laboratories is desirable in order to compare results between various research groups. We suggest from
our recent experience as well as on
theoretical considerations that the
time has come for the pascal to fulfil
this role.
A. Mellon () A. Allan P. E. Bearn
Anorectal Physiology Unit
Good Hope Hospital Trust
Rectory Road
Sutton Coldfield B75 7RR, UK
L E T T E R TO T H E E D I TO R
Anal pressures after hemorrhoidectomy

Dear Sir,
We read with interest the article by
Ho and Tan on ambulant anal manometry before and after hemorrhoidectomy [1]. As mentioned by the
authors, anal pressures are often
raised in patients with piles. It may
be due either to an associated spasm
of the internal anal sphincter (IAS)
or to an increased blood flow and hypertension in the anal cushions forming the piles [2]. The authors state
that, after hemorrhoidectomy without sphincterotomy, there was a significant reduction in the maximum
anal pressures of their patients, consistent with the findings reported by
others [3]. Nevertheless, many surgeons still advocate an internal
sphincterotomy associated with hemorrhoidectomy in patients with
anal spasm evaluated either clinically or manometrically, and some
do it routinely in young adult males
with the aim of reducing postoperative pain.
A randomized prospective study

of closed versus open hemorrhoidectomy is in progress in our unit, and
we thought that our preliminary data
on the correlation between morphology and function of the IAS before
and after surgery would be of some
interest for this controversial point.
Endosonographic IAS thickness and
manometric resting tone, mainly related to IAS activity, were evaluated
as previously reported [4]. No sphincterotomy and a minimal stretch of the
anus were performed by anal retractor in a group of ten patients. A significant reduction of resting tone at
manometry from 694.6 to 54.93.8
mmHg was found (P<0.01; values
expressed as meanSEM), whereas
IAS thickness at anal ultrasound was
2.30.62 and 2.10.67 cm before
and after surgery, respectively (n.s.).
These findings seem to support
the theory that high preoperative
anal pressures are mainly due to the
bulk of enlarged anal cushions and
may represent a warning against the
routine use of internal sphincterotomy, which potentially can lead to

faecal soiling, in these patients.
Yours sincerely,
M. Pescatori () U. Favetta L. Navarra
Coloproctology Unit
Villa Flaminia
Via Luigi Bodio, 58
I-00191 Rome, Italy
References
1. Ho YH, Tan M (1997) Ambulatory anorectal manometric findings in patients
before and after haemorrhoidectomy. Int
J Colorect Dis 12:296297
2. Sun WM, Read NW, Shorthouse AJ (1990)
Hypertensive anal cushions as a cause of
the high anal canal pressures in patients
with haemorrhoids Br J Surg 77:458462
3. Ho YH, Seow Choen F, Goh HS (1995)
Haemorrhoidectomy and disordered rectal and anal physiology in patients with
prolapsed haemorrhoids. Br J Surg 82:
596598
4. Favetta U, Amato A, Interisano A, Pescatori M (1996) Clinical, manometric
and sonographic assessment of the anal
sphincters. A comparative prospective
study. Int J Colorect Dis 4:163166

Июнь

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Июнь

Загружено:

Авторское право:

Доступные форматы

Int J Colorect Dis (1998) 13: 113115

G. Milito F. Cortese C. U. Casciani

Rhomboid flap procedure for pilonidal sinus:

Accepted: 6 February 1998

ter the age of 40 years [27]. Its cause is controversial, and

Key words Pilonidal sinus Dufourmentels technique

RFT was never performed during a period of acute inflammation,

may well be hope that it will shorten further. Compared

Sixty-three of the 67 wounds healed by first intention. Of

This operation has led to remarkably good results in our

1. Buie LA (1944) Jeep disease (pilonidal disease of mechanized

20. Kronborg O, Christensen K, Zimmermann-Nielsen CE (1985)

Int J Colorect Dis (1998) 13: 116118

J. L. Kelly D. S. ORiordain E. Jones E. Alawi

The effect of hysterectomy on ano-rectal physiology

Accepted: 10 February 1998

Rsum Lhystrectomie est associe une constipation

Key words Hysterectomy Anal sphincter

Patients and methods

Int J Colorect Dis (1998) 13: 119123

Selective omission of loop ileostomy in restorative proctocolectomy

Accepted: 10 February 1998

Abstract Omission of a temporary ileostomy in patients

de complications puissent tre vites, le risque de

Complication rates (%)

Numbers of patients in NI&(I) groups

Chronic relapsing colitis

Indications: acute colitis (n=18), toxic megacolon (n=8), rectal

Table 4 Early and late complications for pouch procedures in those

Mann-Whitney U test, 2 P=0.067, NS

high dose steroids and azathioprine at the time of surgery.

The incidence of pouchitis appeared more common in

Group NI was 11 (664) days and for Group I 16 (10160)

stricture which required revisional surgery. The presence

Int J Colorect Dis (1998) 13: 124130

F. Pucciani M. L. Rottoli A. Bologna F. Cianchi

Pelvic floor dyssynergia and bimodal rehabilitation:

Accepted: 10 February 1998

culiar results. Resting anal canal pressure had increased

patients (ge: 28 64 ans; ge moyen: 42.5 ans) de lunit

Patients and methods

last lesson were reviewed and new exercises were introduced so as

antiversion and retroversion pelvic movements,

3rd session: perianal and perivaginal stretching,

Stool frequency (n/week)

Patients (pre) vs Patients (post) or vs Controls: P<0.001

learned to do during kinesitherapy. Feedback was noted by changes

duration, which had been significantly (P < 0.001) shorter

Defecographic results are reported in Table 4.

Dyschezia, defined as difficult bowel movements, may be

Patients (pre) or Patients (post) vs Controls: P<0.05

Patients (pre) vs Patients (post) or vs Controls: P<0.05

Table 4 Defecography (R resting, E evacuation)

Pelvic flor descent (mm)

Patients (pre) or Patients (post) vs Controls: P<0.05

of obstructed defecation, which occurs when pelvic floor

voluntary step for arriving at normal defecation. Bimodal

any organic internal anal sphincter damage, for the reasons

creases the pressure gradient between the anal canal and

2. Whitehead WE, Devroede G, Habib FI, Meunier P, Wald A

Int J Colorect Dis (1998) 13: 131133

J. M. Gattuso H. S. Debinski H. O. Kangro

Evaluation of specific herpes DNA viruses in idiopathic megarectum

Accepted: 10 February 1998