Oslo University Hospital, Department of Pediatric surgery, P.O. Box 4950 Nydalen, 0424 Oslo, Norway
University of Oslo, Faculty of Medicine, P.O. Box 1078 Blidern, 0316 Oslo, Norway
a r t i c l e
i n f o
Article history:
Received 5 August 2014
Received in revised form 10 December 2014
Accepted 27 December 2014
Key words:
Hirschsprung disease
Anal endosonography
Anal manometry
Anal sphincter
Postoperative fecal incontinence
a b s t r a c t
Background: The reasons for fecal incontinence after surgery for Hirschsprung disease (HD) remain unclear.
The aim of this study was to examine the anal sphincters by anal endosonography and manometry after transanal
endorectal pull-through, with or without laparotomy or laparoscopy, in HD patients. Furthermore, we aimed to
correlate these ndings to bowel function.
Patients and methods: Fifty-two HD patients were followed after endorectal pull-through. Anal endosonography
and manometry were performed without sedation at the age of 3 to 16 years.
Results: Endosonographic internal anal sphincter (IAS) defects were found in 24/50 patients, more frequently
after transanal than transabdominal procedures (69 vs. 19%, p = 0.001). In a multiple variable logistic regression
model, operative approach was the only signicant predictor for IAS defects. Anal resting pressure (median
40 mm Hg, range 15120) was not correlated to presence of IAS defects. Daily fecal incontinence occurred
more often in patients with IAS defects (54 vs. 25%, p = 0.03).
Conclusions: Postoperative IAS defects were frequently detected and were associated with daily fecal incontinence. IAS defects occurred more often after solely transanal procedures. We propose that these surgical
approaches are compared in a randomized controlled trial before solely transanal endorectal pull-through is
performed as a routine procedure.
2015 Elsevier Inc. All rights reserved.
higher after TEPT than after other surgical procedures, where only a
minor part of the operation is performed through the anal opening
[46]. So far, few prospective studies have evaluated anorectal function
after introduction of TEPT, and the data are inconclusive [2,4,5,7].
Importantly, there are no agreed-on methods for assessing functional
outcome after HD operations, and this makes comparison of outcome
after different surgical techniques difcult. In addition, objective postoperative examinations of the anal sphincters in HD patients are almost
absent in the literature [6,811]. Postoperative evaluation of the anal
sphincters by anal endosonography (AES) and manometry may add important information about how surgery affects the anal sphincters and
make it possible to compare different surgical approaches with respect
to anal sphincter integrity and function.
The aim of this study was to examine the anal sphincters by AES
and anal manometry in HD patients operated with TEPT or endorectal
pull-through assisted by laparotomy or laparoscopy. Furthermore, we
aimed to correlate the endosonography and manometry ndings to
bowel function.
1. Patients and methods
This prospective cohort study was conducted in a tertiary pediatric
surgery center. The study was approved by the regional ethics
1342
(KJS & KB) to evaluate interrater reliability, and then by the three authors
in common if there was disagreement.
The thickness of the anal sphincters was measured in the mid anal
canal at the 3 and 9 o'clock positions, using image-processing software,
Image-Pro Express (Media Cybernetics, Inc, Bethesda, MD- USA). Mid
anal canal was dened as the most cranial level at which the external
anal sphincter (EAS) formed a complete ring anteriorly. The internal
anal sphincter (IAS) thickness was dened as the width of the
hyporeective band surrounding the subepithelium (Fig. 1a). IAS is
surrounded by a broader zone of moderate reectivity representing
two muscle layers, innermost the longitudinal smooth muscle layer
Fig. 1. a, Axial image from the mid anal canal in a 7 year old boy (anterior is uppermost).
The IAS can be seen as a hyporeective band (arrowheads) surrounding the subepithelium.
A small, relatively hyperreective defect (17o) can be seen in the posterior part. The surrounding broader zone of moderate reectivity (arrows), representing the longitudinal
smooth muscle layer and the EAS, is intact. b, In this 8 year old boy, nearly half of the IAS
is absent while the EAS is intact.
1343
and outermost the EAS [1517]. The EAS can be difcult to distinguish
from the longitudinal muscle layer, and in this study the longitudinal
muscle was included in the EAS measure. Thus, the EAS thickness was
dened as the width of the moderately reective zone or, when applicable, the distance between the interface reections at the inner margin of
the longitudinal muscle and the outer margin of the EAS (Fig. 1a).
A defect in the anal sphincters is visualized as a relatively
hyperreective area in the IAS and a relatively hypo- or hyperreective
area in the EAS. Lack of normal sphincter reectivity in a sector
representing 15 or more of the circumference was dened as a defect.
Both the localization and the angular distribution were recorded for
each defect.
Table 1
Background characteristics.
Down syndrome
Preoperative stoma
Age at operation (months)
Age at follow-up (years)
TEPT (n = 31)
TAA (n = 21)
p value
2 (6%)
4 (13%)
1.8 (0.4133)
6.4 (3.414.3)
1 (5%)
9 (43%)
13 (1.2100)
11.2 (4.416.6)
1.00
0.02
0.006
b0.001
Background characteristics for the different operative approaches, TEPT (transanal) or TAA
(laparotomy, n = 17, or laparoscopy, n = 4).
Age is expressed as median (range).
2. Results
Ninety-one children were operated with an endorectal pull-through
procedure in the study period. 52 children (42 boys) were included.
Reasons for exclusion were aganglionosis oral to the sigmoid colon
(4), lost for follow-up (4), child or parent refused examination with
AES and manometry (18), and age younger than 3 years at follow up
(13). The age at examination ranged from 3.4 to 16.6 years. TEPT had
been performed in 31 children. The TAA group consisted of 21 children
in whom the endorectal pull-through had been assisted by laparotomy
in 17 and laparoscopy in four. TEPT patients were operated at younger
age and were younger at the time of examination than TAA patients
(Table 1). AES was performed in 50 patients and anal manometry in
37. In two children, the manometry results were excluded because of
fecal masses detected by digital examination. Thus, valid manometry
results were achieved for 35 patients.
1344
Table 2
Predictors for IAS defectlogistic regression.
OR (95 % CI)
p value
p value
9.4 (2.536)
0.98 (0.951.00)
0.75 (0.620.92)
0.001
0.09
0.006
5.8 (1.7-29)
0.99 (0.96-1.02)
0.90 (0.69-1.17)
0.03
0.41
0.95
Odds ratio (OR) for IAS defects by binary logistic regression analyses. Adjusted OR was achieved by multiple logistic regression analysis.
complication [21], and IAS defects have also been described [22]. Ideally,
the transanal dissection during TEPT should be performed with minimal
or no anal dilatation. Our results indicate that too much dilatation of the
3. Discussion
In patients operated with endorectal pull-through procedures for
HD, postoperative AES demonstrated IAS defects in half of the patients
and a defect in the EAS in one patient. To the best of our knowledge,
postoperative AES examination has not been performed in such a patient group before. Three previous reports on postoperative AES results
after different operations for HD show contradictory results [8,9,11]. In a
study of 40 adults operated with the Duhamel technique, IAS defects
were found in 23 patients and EAS defects in 17 [8]. On the other
hand, in a study of 19 patients operated for HD (16 Duhamel, 2 Soave,
1 Rehbein procedure), Keshtgar et al. found intact IAS and EAS in all patients [11]. Kuhawara et al. performed AES after Duhamel and Rehbein
procedures, and described IAS and EAS defects corresponding to the
intended internal sphincterotomy or myectomy in all patients [9].
Since there are so few data from postoperative AES examinations in
HD patients, and the results are conicting, it is not possible to decide
whether endorectal pull-through is superior or inferior to other operations for HD with respect to avoiding anal sphincter damage. Nevertheless, it is a disturbing nding that half of the patients in this series were
found to have IAS defects. Since healthy children and adolescents do not
have IAS defects [20], these defects must be iatrogenic. We hypothesize
that the anal dilatation during the endorectal pull-through is an
important contributor to the defects observed in the IAS. In adults treated by anal dilatation for anal ssures, fecal incontinence is a known
Fig. 2. Mid anal canal in a healthy 3 year old boy with intact IAS and EAS. Anteriorly an
interdigitation of bers between EAS and the transverse perineal muscles (TP) causes
hyporeective areas of the EAS (arrows). Posteriorly the anococcygeal ligament (ACL)
can be seen as an area of EAS with lower reectivity.
Fig. 3. a, Axial image from mid anal canal in an 8 year old boy examined by one of the
authors (KJS) in 2010. In addition to a 90o IAS defect (arrows), a hyporeective defect of
60o can be seen in the EAS. b, This image was taken 4 years earlier from the same child
by a different author (KB) and a different AES system (B&K Medical). The same defects
both in the IAS and the EAS were seen.
Resting pressure
(mm Hg)
Squeeze pressure
(mm Hg)
Median
(range)
Median
(range)
p-Value
40 (15120)a
131 (60250)
40 (30120)
160 (60250)
0.56
42 (1560)
33 (1550)
0.54
125 (70250)
0.009
40 (1575)
115 (60250)
0.25
TAA (n = 10)
p-Value
48 (30120)
0.049
180 (100250)
a
One 14 year old boy had a reproducible resting pressure of 120 mm Hg. For the rest of
the patients the range was 1575 mm Hg.
b
p-Value when compared to patients with no defects or IAS defects b135.
anal canal may occur even if there is focus to avoid this. In addition,
incorrect endorectal dissection may also damage the IAS.
IAS defects were seen signicantly more frequently after the TEPT
procedure than after TAA. This effect of operative approach on IAS defects remained evident after controlling for background characteristics.
The principal difference between the TEPT and TAA procedures is that
the dissection of the colon and rectum is performed transanally or
transabdominally, respectively. In most patients, the transanal dissection during TEPT involves longer and more extensive anal dilatation
than TAA. We believe this contributes to more IAS defects and thereby
higher risk of fecal incontinence. Previous studies of long term results
after TEPT and TAA did not nd signicant difference in over all
functional outcome between the two operative approaches [2,4,5].
However, the TEPT group had inferior continence score or a possible
tendency for more daily incontinence in two of the three studies [2,4].
Fecal incontinence was reported by half of the patients. Daily fecal
incontinence occurred more frequently in children with IAS defects.
This result corresponds with ndings in HD patients operated with the
Duhamel procedure, where all patients with fecal soiling had scars in
both IAS and EAS [8]. We believe IAS defects contribute to fecal incontinence in many HD patients. In the present study, fecal incontinence was
also reported by one fourth of the patients with intact IAS. This is not
surprising since other factors than the anal sphincters also contribute
to fecal continence. These include colonic motility, anorectal sensation,
and the sub-epithelium of the anal canal [11,23,24]. All these factors
can be affected during surgery for HD. Especially, focus has been
drawn to the distance from the dentate line to the level of the coloanal
anastomosis [23].
This study was designed to seek correlations between the
endosonography and manometry ndings and the bowel function,
and not to compare results after endorectal pull-through with either
transanal or transabdominal approach. Age at operation and age at
Table 4
Fecal incontinence.
None or grade 1
(occasionally) a
Endosonographic appearance
Intact IAS (n = 26)
20 (77%)
Grades 23
(daily) a
p-Value
6 (23%)
0.045
12 (50%)
40 (2560)
130 (60230)
12 (50%)
45 (15120)
132 (90250)
0.5
0.8
a
Krickenbeck classication for fecal incontinence: grade 1occasionally (once or twice
a week). Grades 23every day or constant.
b
Anal pressures are expressed as median (range).
1345
follow-up were different in the two groups, and several selection biases
may be identied. Still, the effect of operative approach on IAS defects
remained evident after controlling for possible confounders. Therefore,
this association cannot be ignored. Loss to follow-up is unlikely to affect
this association, but may have caused an overestimation of occurrences
of fecal incontinence and constipation because patients with these
problems probably were more motivated for follow-up and invasive
examinations.
In children with idiopathic constipation, the width of IAS has
been reported to correlate with the severity of symptoms [25,26]. This
correlation could not be reproduced in HD patients, indicating that
other factors than IAS hypertrophy causes persistent constipation in
HD patients. Bowel dysmotility, intestinal neuronal dysplasia or reduced
numbers of interstitial cells of Cajal in the ganglionic bowel have been
proposed as explanations for persistent constipation [11,24,2729].
The median maximum anal resting and squeeze pressures were
within the range previously described in HD patients and healthy children [8,10,27,3032]. In one small series, manometry was performed
in ve children after TEPT and seven children after transabdominal
Soave procedure. Like us, they found similar resting pressures in the
two groups [6]. In the present study, the presence of IAS defects was
not associated with lower resting pressure, contrary to previous ndings after the Duhamel procedure for HD, where the numbers of IAS disruptions were inversely correlated to resting pressure [8]. Furthermore,
there were no associations between anal sphincter pressures and fecal
incontinence or constipation. Normal anal manometry ndings have
previously been reported both in HD patients with fecal incontinence
and with persisting obstructive symptoms [8,24,27]. Thus, the value
of anal manometry in the assessment of anorectal function after
endorectal pull-through procedures for HD may be questioned. However, associations between anal resting pressure and fecal incontinence
have been described after different operations for HD when the patients
were examined in ketamine anesthesia [10,11]. In the present study, the
examination was performed without preoperative bowel preparation
and without sedation. Even though most children cooperated well, the
anal resting pressure results may be less reliable because the conditions
for examining were not as standardized as under ketamine anesthesia
and after bowel preparation.
To conclude, this study shows frequent IAS defects after endorectal
pull-through for HD. Those operated with a solely transanal procedure
displayed more sphincter defects than those operated with a combination of transanal surgery and laparoscopy or laparotomy. These ndings
are worrying. IAS defects were associated with fecal incontinence, and
fecal incontinence is detrimental for psychosocial function. Therefore,
we propose that these operative techniques are compared in a randomized controlled trial with a sufcient number of patients before TEPT is
performed as a routine procedure.
Acknowledgments
The authors acknowledge research funding from South-Eastern
Norway Regional Health Authority and The Norwegian National Advisory
Unit on solid tumours in children (KSSB). The authors thank Professor
Lars Mrkrid for statistical and methodological assistance.
References
[1] Diseth TH, Egeland T, Emblem R. Effects of anal invasive treatment and incontinence
on mental health and psychosocial functioning of adolescents with Hirschsprung's
disease and low anorectal anomalies. J Pediatr Surg 1998;33(3):46875.
[2] Stensrud KJ, Emblem R, Bjornland K. Functional outcome after operation for
Hirschsprung diseasetransanal vs transabdominal approach. J Pediatr Surg 2010;
45(8):16404.
[3] Shankar KR, Losty PD, Lamont GL, et al. Transanal endorectal coloanal surgery for
Hirschsprung's disease: experience in two centers. J Pediatr Surg 2000;35(8):
120913.
[4] El-Sawaf MI, Drongowski RA, Chamberlain JN, et al. Are the long-term results of
the transanal pull-through equal to those of the transabdominal pullthrough? A
1346
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19] Holschneider A, Hutson J, Pena A, et al. Preliminary report on the International Conference for the Development of Standards for the Treatment of Anorectal Malformations. J
Pediatr Surg 2005;40(10):15216.
[20] Rehman Y, Stensrud KJ, Morkrid L, et al. Endosonographic evaluation of anal sphincters in healthy children. J Pediatr Surg 2011;46(8):158792.
[21] Nelson RL, Chattopadhyay A, Brooks W, et al. Operative procedures for ssure in ano.
Cochrane Database Syst Rev 2011;11:CD002199.
[22] Speakman CT, Burnett SJ, Kamm MA, et al. Sphincter injury after anal dilatation demonstrated by anal endosonography. Br J Surg 1991;78(12):142930.
[23] Levitt MA, Dickie B, Pena A. The Hirschsprungs patient who is soiling after
what was considered a successful pull-through. Semin Pediatr Surg 2012;21(4):
34453.
[24] Kaul A, Garza JM, Connor FL, et al. Colonic hyperactivity results in frequent fecal
soiling in a subset of children after surgery for Hirschsprung disease. J Pediatr
Gastroenterol Nutr 2011;52(4):4336.
[25] Keshtgar AS, Ward HC, Clayden GS, et al. Thickening of the internal anal sphincter in
idiopathic constipation in children. Pediatr Surg Int 2004;20(1112):81723.
[26] Hosie GP, Spitz L. Idiopathic constipation in childhood is associated with thickening
of the internal anal sphincter. J Pediatr Surg 1997;32(7):10413.
[27] Demirbag S, Tiryaki T, Purtuloglu T. Importance of anorectal manometry after denitive
surgery for Hirschsprung's disease in children. Afr J Paediatr Surg 2013;10(1):14.
[28] Medhus AW, Bjornland K, Emblem R, et al. Motility of the oesophagus and small
bowel in adults treated for Hirschsprung's disease during early childhood.
Neurogastroenterol Motil 2010;22(2):15460 [e49].
[29] Taguchi T, Suita S, Masumoto K, et al. An abnormal distribution of C-kit positive cells
in the normoganglionic segment can predict a poor clinical outcome in patients with
Hirschsprung's disease. Eur J Pediatr Surg 2005;15(3):1538.
[30] Benninga MA, Wijers OB, van der Hoeven CW, et al. Manometry, prolometry, and
endosonography: normal physiology and anatomy of the anal canal in healthy children. J Pediatr Gastroenterol Nutr 1994;18(1):6877.
[31] Sutphen J, Borowitz S, Ling W, et al. Anorectal manometric examination in
encopretic-constipated children. Dis Colon Rectum 1997;40(9):10515.
[32] Kumar S, Ramadan S, Gupta V, et al. Manometric tests of anorectal function in
90 healthy children: a clinical study from Kuwait. J Pediatr Surg 2009;44(9):
178690.