PEDIATRICS

976

ORIGINAL CONTRIBUTIONS

nature publishing group

Long-Term Clinical Outcome After Botulinum Toxin

Injection in Children With Nonrelaxing Internal
Anal Sphincter
Bruno P. Chumpitazi, MD, MPH1,3, Steven J. Fishman, MD2 and Samuel Nurko, MD, MPH1

OBJECTIVES:

Children with surgically repaired Hirschsprungs disease (HD) and those with internal anal
sphincter (IAS) achalasia may develop obstructive gastrointestinal symptoms and/or enterocolitis
due to a functional obstruction caused by an inability of the IAS to relax. Anal sphincter
Clostridium botulinum toxin (BoTox) injections may provide a reversible therapy. However, there
is limited information regarding the long-term outcomes of children receiving this therapy. The
primary aim of this study was to determine the long-term clinical outcomes of BoTox therapy in
children with a nonrelaxing IAS. The secondary aim of this study was to determine prognostic
factors predicting a favorable outcome following BoTox IAS injection.

METHODS:

We conducted a retrospective review of children with nonrelaxing IAS who received anal sphincter
BoTox at a tertiary medical center. Children were classified into one of four long-term clinical
outcome groups (excellent, good, fair, poor).

RESULTS:

A total of 73 children (30 HD, 43 IAS achalasia) received anal sphincter BoTox injections
and had a mean follow-up of 32.12.9 (s.e.) months. A mean of 2.70.2 injections were
given to each child, with 56 (76.7%) children receiving multiple injections. An initial clinical
improvement was seen in 65 of 73 (89%) children after the first injection. A total of 39
(53.4%) children had an excellent or good long-term outcome that was maintained for a mean
of 17.13.1 months from the time of the last BoTox injection. Hospitalization rates significantly
decreased in those previously hospitalized before initial BoTox injection. Seven (9.5%) patients
developed transient fecal incontinence, and one (1.3%) developed significant pain after an
injection. Factors predicting a favorable long-term clinical outcome were initial short-term
improvement after the first BoTox injection and having IAS achalasia rather than HD.

CONCLUSIONS: Anal sphincter BoTox may be an effective and safe long-term therapy for children with

nonrelaxing IAS.
Am J Gastroenterol 2009; 104:976983; doi:10.1038/ajg.2008.110; published online 3 March 2009

INTRODUCTION
Children with surgically repaired Hirschsprungs disease (HD)
(13), and those with internal anal sphincter (IAS) achalasia
(47), may develop gastrointestinal obstructive symptoms
or enterocolitis. Obstructive symptoms include intractable constipation refractory to aggressive laxative regimens,

retentive fecal incontinence, and abdominal distention with

secondary vomiting (28). Enterocolitis episodes may be life
threatening, requiring aggressive treatment and hospitalization (3,8,9). The etiology of the obstructive gastrointestinal
symptoms and enterocolitis is frequently attributed to a functional obstruction caused by an inability of the IAS to relax

1
Center for Motility and Functional Gastrointestinal Disorders, Childrens Hospital Boston, Boston, Massachusetts, USA; 2Department of Surgery, Childrens
Hospital Boston, Boston, Massachusetts, USA; 3Present address: Neurogastroenterology and Motility Center, Texas Childrens Hospital, Houston, Texas, USA.
Correspondence: Samuel Nurko, MD, MPH, Department of Surgery, Center for Motility and Functional Gastrointestinal Disorders, Childrens Hospital Boston,
300 Longwood Avenue, Boston, Massachusetts 02115, USA. E-mail: Samuel.nurko@childrens.harvard.edu
Received 11 May 2008; accepted 4 October 2008

The American Journal of GASTROENTEROLOGY

VOLUME 104 | APRIL 2009 www.amjgastro.com

(absence of the rectoanal inhibitory reflex). A nonrelaxing IAS

is present in those with surgically repaired HD (10), and IAS
achalasia (4,5).
Traditionally, treatment for symptomatic children with a
nonrelaxing IAS, regardless of the etiology, involves surgical
procedures directed at the anal sphincter, such as myectomy
(11,12). However, these procedures are irreversible, and have
the potential to cause fecal incontinence, which in some longterm follow-up studies has been reported in 2040% cases
(13,14). Long-term studies demonstrate lower median anal
resting pressures and lower maximal anal squeeze pressures in
adult HD patients compared with controls (15). These findings
suggest there may be an increased baseline long-term risk for
fecal incontinence in patients after HD surgery. Many children
with surgically repaired HD who go on to develop obstructive
symptoms improve over time, particularly when late adolescence is reached (9). Therefore, given the risks associated with
surgical myectomy, and potential of improvement over time,
new nonsurgical options for the treatment of symptomatic children with a nonrelaxing IAS are being sought.
Recent years have seen the advent of chemical sphincterotomies which lower overall anal sphincter pressure through
pharmacologic means. Clostridium botulinum toxin (BoTox)
injection into the anal sphincter has been successfully used for
the treatment of chronic anal fissures in both adults and children (1621). BoTox has also been used successfully in limited
short-term studies of children with surgically repaired HD (22),
IAS achalasia (5,23), and IAS dysfunction (24). However, the
long-term clinical outcomes of children with nonrelaxing IAS
who receive anal sphincter BoTox therapy as a primary modality to address gastrointestinal obstructive symptoms and/or
enterocolitis are unknown.
The primary aim of this study was to determine the longterm clinical outcomes after the application of BoTox therapy
in children with nonrelaxing IAS. The secondary aim was to
elucidate factors which may be involved in influencing longterm clinical outcomes.

METHODS
A retrospective chart review of all children with nonrelaxing
IAS who received BoTox injections for obstructive symptoms
and/or enterocolitis at Childrens Hospital Boston from 1998
to 2007 was conducted. The Committee on Clinical Investigation of the Childrens Hospital Boston, approved this study. All
children with documented follow-up after a BoTox injection were included. Obstructive symptoms were defined by
the primary gastroenterologist, and confirmed by one of the
investigators (SN) and included constipation, abdominal distention, straining with defecation, and retentive fecal incontinence. Enterocolitis episodes were identified by hospital
discharge records, in-patient visits, and outpatient clinical visit
diagnoses.
Baseline symptoms were obtained by review of the history
and physical examination during the initial visit to the primary
2009 by the American College of Gastroenterology

gastroenterologist, and by documentation at the time of the

first BoTox injection by one of the investigators (SN). Symptoms were followed by documentation at subsequent clinical
visits or procedures. Rectal therapies were defined as any enemas, anal dilatations, irrigations, or suppositories. All findings
were recorded using a standardized form.
All children with HD included in the study had previously
undergone corrective surgery. All included HD patients had a
nonrelaxing IAS, and the presence of ganglion cells on repeat
rectal biopsies at the time of the BoTox injection.
IAS achalasia (IASA) was diagnosed when nonrelaxation
of the IAS occurred during manometry testing (as defined
below) and ganglion cells were found to be present on rectal
biopsy (5).
Anorectal manometry

On the night before the anorectal manometry, all children

received a phosphate enema. The manometric studies were
performed with a continuously perfused catheter using a
low-compliance pneumo-hydraulic system (Model ARM2;
Arndorfer Medical Specialties, Greendale, WI) as previously
described (5). The manometric catheter was made from polyvinyl tubing and contained four recording ports separated
by 1 cm at the distal end and staggered at a 90 angle. A latex
balloon was attached to the end of the anorectal catheter and
its port is connected to the recording system to register the
exact time of balloon distention. Pressure readings were transmitted to the Medtronics Polygraphic System (Medtronics,
Minneapolis, MN). All baseline measurements were set at
atmospheric pressure. Initially a slow pull-through was performed until the area of high-pressure zone was identified.
The balloon was then rapidly inflated with serial volumes that
started from 10 ml and increased by 10 ml up to at least 60 cm3.
In those children in whom there was no IAS relaxation even
with a balloon of 60 ml, the balloon volume was increased
until we found resistance to inflation, and evidence of discomfort was elicited in the patient. This was performed to ensure
that the lack of relaxation of the IAS sphincter was real, and
not a false positive result occurring because of a lack of rectal
wall stimulation. Children who were not initially cooperative
received oral midazolam (0.5 mg/kg) in order to aide in completion of the evaluation.
The intra-anal pressure was measured in all individuals
after giving appropriate time for accommodation using the
slow pull-through technique. In our laboratory normal values
range from 60 to 120 mm Hg. The presence of IAS relaxation
(rectoanal inhibitory reflex) during manometry testing was
then measured for each balloon distention. Relaxation of the
sphincter was defined as present when there was a consistent reproducible decrease in sphincter pressure of at least 5%
from the individuals baseline pressure within 15 s of the rectal
balloon distention with a subsequent return to the previous
baseline pressure. Children in whom the rectoanal inhibitory reflex could not be elicited even with maximum balloon
inflation were considered as having a nonrelaxing IAS.
The American Journal of GASTROENTEROLOGY

977

PEDIATRICS

Children With Nonrelaxing Internal Anal Sphincter

978

Chumpitazi et al.

PEDIATRICS

BoTox administration

BoTox (C. botulinum toxin A; Allergan, Irvine, CA) injections were performed under general anesthesia after cleaning
the rectum with an age-appropriate enema the night prior, as
described previously (5). Rectal cleanouts were completed both
manually and with irrigations in the operating room if necessary in order to visualize the field. BoTox was diluted in saline
to a concentration of 100 U/ml. Using a small nasal speculum,
the dentate line was identified by inspection and palpation.
The area was then cleaned with providoneiodine. With a 1 ml
syringe and a 2530 gauge needle, 6 U/kg (up to 100 U total)
was divided and injected equally into four quadrants.
In those children with HD that had undergone corrective surgery and had persistent symptoms, the BoTox often was given at
the time of repeated rectal biopsies to exclude persistent aganglionosis (25). If persistent aganglionosis was found, the patients
were excluded from the present study. All children with surgically
repaired HD had a barium enema performed before BoTox injection to evaluate for a transition zone or other anatomic abnormality, and if present underwent rectal biopsy directly without BoTox
injection. Further BoTox injections were given if obstructive
symptoms and/or enterocolitis continued or returned to baseline
after the initial injection. An anorectal manometry was always
performed before repeat BoTox injections to ensure a nonrelaxing IAS was still present, and to evaluate the individuals baseline
sphincter pressure. Colonic manometry was performed in selected
patients that continued to be intractable after BoTox injection.

Figure 1 shows the algorithm that we have followed in the

treatment of children with nonrelaxing IAS.
The decision to pursue medical therapy (BoTox) or perform
a surgical procedure was made by the family of the patient after
receiving recommendations from the childs gastroenterologist and
surgeon. Given a concern of the child having irreversible incontinence after a myectomy, some families chose to pursue multiple
BoTox injections rather than pursue a surgical option. In general,
if there was a positive response following BoTox, but the relapses
between BoTox injections occurred at intervals longer than 3
months, repeated BoTox injections were performed. If the response
was shorter than 3 months, a myectomy was suggested (Figure 1).
Outcome measures

Short-term improvement after the initial BoTox injection

was defined as an increase in bowel movement frequency, or
a decrease in the frequency of other obstructive symptoms
occurring within 2 weeks after the injection. Therefore this
measure reflects only symptomatic improvement, independently of the use of laxatives.
Long-term clinical outcome was determined by documented
symptom characterization, and laxative use at the time of the last
documented follow-up visit and was categorized into four groups.
Poor. No improvement in the number of bowel movements
or other obstructive symptoms as compared to baseline, or
underwent a subsequent therapeutic surgery, or hospitaliza-

Nonrelaxing IAS
In postoperative HD,
exclude anatomic problems and/or
persistent aganglionosis

BoTox

Normal or high
IAS pressure

No

Yes

Colonic manometry and

other motility studies.
Expand differential
diagnosis

Response
No

Repeat
anorectal
manometry

Yes

Dysmotility

Close follow-up
Normal or high
IAS pressure
Yes

Surgery
No

Relapse
Repeat BoTox

ACE
Segmental
resection

Response
Consider
myectomy

Yes

Redo pull
Colostomy through in
HD
Ileostomy

No

Figure 1. Algorithm for the evaluation and treatment of children with nonrelaxing internal anal sphincter (IAS). ACE, antegrade colonic enemas;
HD, Hirschsprungs disease.

The American Journal of GASTROENTEROLOGY

VOLUME 104 | APRIL 2009 www.amjgastro.com

Children With Nonrelaxing Internal Anal Sphincter

Fair. Improvement in the number of bowel movements or other

obstructive symptoms as compared to baseline but with retentive
fecal incontinence episodes greater than once a week, or consistent rectal therapy (e.g. enema, suppository, irrigation) usage.
Good. Improvement in the number of bowel movements
or other obstructive symptoms as compared to baseline with
continued usage of laxatives.
Excellent. Improvement in the number of bowel movements
or other obstructive symptoms as compared to baseline without the need for laxatives.
For analysis purposes excellent and good long-term
outcomes were classified as being favorable.
Hospitalization rates per year for enterocolitis or to manage
obstructive symptoms (e.g. constipation cleanout) were obtained
before and after the initial BoTox injection. Duration of BoTox
efficacy was determined based on the time interval between
initial injection of BoTox and one of the following: return to
baseline symptoms, hospitalization for obstructive symptoms
and/or enterocolitis, surgical procedure for obstructive symptoms,
documented clinical decision to proceed with another BoTox
injection, or time interval from injection to last documented followup visit. Fecal incontinence was determined to be a complication of
the BoTox injection if new fecal soiling developed during the first
week after injection, or if there was an increase by at least one in the
number of fecal accidents per day as compared to baseline.
Statistics

Data are reported as meanstandard error of the mean. Comparisons between children with surgically repaired HD and those
with IAS achalasia were performed with
2-analysis when comparing proportions or independent Students t-tests. Comparisons
between baseline and short and long-term results were performed
with
2-analysis when comparing proportions or paired t-test
analysis. Binary multivariate stepwise logistic regression was
completed with favorable and unfavorable long-term outcomes
being the dependent variable.
2-Analysis was used to compare
proportions in all other testing performed. Maximum resting
sphincter pressure, age at the time of the first BoTox, and age at
the time of the last follow-up were categorized into three evenly
distributed groups for regression analysis. A P value of < 0.05 was
considered as statistically significant. Survival analysis was used
to determine time to failure within each group. Log-rank test was
used to compare the time to failure between groups. All statistical
analyses were performed using SPSS (version 15.0).

RESULTS
Baseline characteristics

A total of 73 children (30 with HD, and 43 with IAS achalasia)

were included. Their main characteristics are shown in Table 1.
2009 by the American College of Gastroenterology

Table 1. Baseline characteristics of the patients

All
patients

Hirschsprungs
disease

IAS
achalasia

73

30

43

5.2 0.5
(Range:
0.118.8)

5.0 0.7
(Range:
0.116.0)

5.3 0.6
(Range:
0.518.8)

Male

42 (57.5%)

24 (80%)

18 (41.9%)a

Female

31 (42.5%)

6 (20%)

25 (58.1%)

Obstructive

60 (82.1%)

17 (56.6%)

43 (100%)a

Enterocolitis

7 (9.5%)

7 (23.3%)

0 (0%)

Both

6 (8.2%)

6 (20%)

0 (0%)

30 (41.1%)

18 (60%)

12 (27.9%)a

87.9 2.7

96.1 5.0

81.9 2.6a

Total (n)
Age (years)

Sex

Symptoms

Rectal therapies
Baseline anal sphincter
pressure (mm Hg)

Values are expressed as means.e. mean.

a
Denotes P < 0.05 comparing patients with Hirschsprungs and IASA.

Initial corrective operations in children with HD included

16 (53.3%) Soave, 9 (30%) Duhamel, 4 (13.3%) Swenson, and 1
(3.3%) unknown pull-through performed at an outside institution. The extent of the original aganglionosis included 19 (63.3%)
with short segment (rectum to sigmoid colon) involvement, 7
(23.3%) with long segment involvement (rectum to descending,
transverse or ascending colon), and 4 (13.3%) with total colonic
involvement with or without small bowel involvement.
Although similarities such as age at time of initial presentation were found in children with HD and IAS achalasia,
the two groups differed with respect to gender distribution,
anal sphincter resting pressure at the time of initial BoTox
injection, enterocolitis episodes, and baseline usage of rectal
therapies (Table 1). A total of 4 children in the entire group had
Downs syndrome, with 3 of these children having HD.
Short-term results

An initial clinical improvement in bowel movement frequency and

a decrease in obstructive symptoms within 2 weeks after the first
injection was seen in 65 of 73 (89%) of all patients. This included
27 of 30 (90%) of those with surgically repaired HD and 38 of 43
(88.3%) of those with IAS achalasia. Of 30 children who required
rectal therapies (e.g. enemas, irrigations, suppositories) before an
initial BoTox injection, only 13 (7 with HD, and 6 with IAS achalasia) children continued to use rectal therapies within 2 weeks after
the initial BoTox injection (P < 0.01) (Figure 2). In those who clinically responded to the initial BoTox injection, the mean duration
of improvement was 7.81.5 months (9.93.1 months with HD,
and 6.31.3 months with IAS achalasia, NS).
The original extent of the aganglionosis in those with surgically repaired HD did not determine initial improvement
The American Journal of GASTROENTEROLOGY

PEDIATRICS

tion for obstructive symptoms and/or an enterocolitis episode

within 3 months of the last documented follow-up.

979

980

Chumpitazi et al.

Pre-BoTox
30

HD

IASA

60

P<0.05

50
25
40

Percent

Number of patients

PEDIATRICS

All

After BoTox

20

15

30
20

10

10

Ex

Good

Fair

Poor

All

HD

IASA

Figure 2. Rectal therapies (enemas, irrigations, or suppositories) before

initial BoTox therapy and after initial BoTox therapy categorized in all
patients, those with Hirschsprungs disease (HD), and those with internal
anal sphincter achalasia (IASA).

Figure 3. Distribution of long-term clinical outcomes by percentage into

excellent, good, fair, and poor clinical long-term outcome categories in all
73 patients, and by Hirschsprungs disease (HD) or internal anal sphincter
achalasia (IASA). Ex, excellent.

Excellent /good
(n =25)

(P = 0.38) with 16 of 19 short segment, 7 of 7 long segment, and

4 of 4 total colonic having an initial improvement.

Multiple
injections
(n =50)

Fair/poor
(n =25)

Long-term results

The mean follow-up period after the initial BoTox injection was 32.12.9 months in all patients. Children with HD
(41.24.9 months) were followed for longer than those with
IAS achalasia (25.93.1 months), P < 0.05. A mean of 2.70.2
injections was given per child, with 56 (76.7%) children receiving multiple injections (range 18 injections).
A total of 39 (53.4%) children had a favorable (excellent or
good) clinical long-term outcome (Figures 3 and 4) at the time
of their last follow-up. These favorable outcomes were maintained for 17.13.1 months after the last injection. Of 43, 28
(65.1%) children with IAS achalasia had an excellent or good
long-term clinical outcome as compared with 11 of 30 (36.7) to
those with HD (P < 0.05). In the group of IASA the mean time
to failure was 45.76.1 weeks, and in the Hirschsprungs group
it was.33.85.3 weeks (P < 0.08) (Figure 5).
Predictors of long-term response

Only improvement to the initial BoTox injection predicted a successful long-term outcome. Overall, 38 of 65 (58.5%) children
that improved after the initial injection had a favorable longterm outcome even if they required more than one injection
(Figure 4). Only 13 (7 IASA and 6 HD) children had a favorable
long-term outcome following a single successful BoTox injection
(Figure 4) (mean follow-up 23.26.4 months). Having a shortterm improvement predicted a favorable long-term outcome
only in those with IAS achalasia in which 28 of 38 (73.6%) with
initial improvement, improved (P < 0.01) as compared with 10
of 27 (37.0%) of those with HD that had initial improvement.
Eight children did not have a short-term clinical improvement
after the initial BoTox injection (Figure 4). Only 1 (7.1%) of
these 14 children had a favorable long-term clinical outcome.
The American Journal of GASTROENTEROLOGY

Improved
(n =65)
Single
injection
only
(n =15)

Initial BoTox
injection
(n =73)

Did not
improve
(n =8)

Multiple
injections
(n =6)

Single
injection
only
(n =2)

Excellent/good
(n =13)
Fair
(n =2)

Fair/poor
(n =6)
Good
(n =1)

Fair/poor
(n =1)

Figure 4. Flow diagram delineating initial short-term outcome and

subsequent long-term clinical outcome based on having received only
an initial BoTox injection or multiple injections.

Age, sex, type of surgical correction, or length of the initial

aganglionosis did not predict long-term outcome.
Hospitalizations

A total of 40 (54.8%) children with nonrelaxing IAS had been hospitalized for obstructive symptoms and/or enterocolitis before BoTox
therapy. The hospitalization rate decreased from 1.90.4 hospitalizations per year to 0.50.8 hospitalizations per year (P < 0.01).
This decrease occurred in those previously hospitalized with primarily obstructive symptoms (e.g. constipation) with a decrease
from 1.60.6 hospitalizations per year to 0.20.1 hospitalizations
VOLUME 104 | APRIL 2009 www.amjgastro.com

Children With Nonrelaxing Internal Anal Sphincter

981

Table 2. Types of surgical procedure and the numbers of

children undergoing each procedure after the initial anal
sphincter BoTox injection therapy

IAS achalasia
Hirschsprungs

0.8

Type of surgery

All patients

Hirschsprungs
disease

10

Ileostomy/colostomy

Pull through

1a

Antegrade colonic enema

Partial colectomy

0.6

Anal dilation
0.4

Myectomy
P < 0.08

0.2

IAS
Achalasia

0.0
0

20

40

60

80

Time to failure (months)

Figure 5. Time to failure in children with internal anal sphincter achalasia
(IASA) and children with Hirschsprungs disease. In the group with IASA
the mean time to failure was 45.7 6.1 weeks, and in the Hirschsprungs
group it was 33.8 5.3 weeks (P < 0.08).

per year (P < 0.05). Those who had previously been hospitalized for
enterocolitis (n = 13) had a decrease from 2.30.6 hospitalizations
per year to 1.00.3 hospitalizations per year (P = 0.06).
Complications

Fecal incontinence occurred in 7 of 73 (9.5%) following

BoTox therapy. The average baseline anal sphincter baseline
pressure in these seven was 100.521.6 mm Hg, and was not
significantly different than the 86.42.0 mm Hg observed in
those without incontinence. The duration of incontinence
lasted for less than 3 days in 6 of the 7 children with this complication. One child (baseline maximal anal resting pressure
of 40 mm Hg) had fecal incontinence lasting 4 weeks. Incontinence was evenly distributed between both groups occurring
in 3 with HD, and 4 with IAS achalasia. Of the 3 with HD, none
had short-segment original aganglionosis (1 long segment and
2 total colonic). Transient rectal pain after waking from general anesthesia occurred in one patient with IASA.
Subsequent evaluations and surgical procedures

Nine children that did not respond to BoTox underwent

colonic manometry evaluations (two with HD and seven with
IASA). All 9 were abnormal, with 5 demonstrating a complete
lack of high amplitude propagating contractions and generalized neuropathy, and 4 having only high amplitude propagating contractions in the proximal colon, with absent activity in
the left side of the colon.
A total of 21 children (12 HD, 9 IASA) underwent 28 therapeutic surgical procedures during the follow-up period. Some
children underwent multiple surgical procedures (Table 2).
Children with HD (12 of 30) trended toward being more likely
than those with IAS achalasia (9 of 43) to undergo a subsequent
surgical procedure (P = 0.11).
A total of 10 children underwent a subsequent myectomy. In
seven, the myectomy was performed when symptoms recurred
2009 by the American College of Gastroenterology

Performed at an outside institution. Patient continued to have problems after

the surgery, and was found to have severe colonic dismotility.

following a successful initial response to BoTox. Three of these

seven remained asymptomatic without any further procedures.
Four became symptomatic again: one responded to a subsequent BoTox injection, and three went on to have a surgical
appendicostomy for the administration of antegrade colonic
enemas (ACE procedure).
None of the three children undergoing myectomy following
a lack of initial improvement with BoTox showed improvement
after the myectomy. They all had abnormal colonic motility. All
three underwent further surgical procedures. These included
one repeat myectomy, one ileostomy, and one ACE procedure.
Predicting long-term clinical outcome

With the use of multivariate regression, having IAS achalasia

(P = 0.01), and demonstrating short-term improvement after
an initial BoTox injection (P 0.05) predicted a favorable longterm outcome. Factors which did not predict long-term outcome
included gender, baseline maximal anal sphincter pressure, age
of first BoTox injection, previous hospitalization, rectal therapy
usage, developmental delay, having enterocolitis as a symptom,
and age at last follow-up. In children with HD, the extent of the
original aganglionosis, and the type of original surgical procedure did not significantly predict favorable vs. unfavorable longterm clinical outcome in the regression model. Colonic motility
was performed in selected intractable patients, and there was a
trend for colonic dysmotility (P = 0.064) to predict an unfavorable long-term result. However, we do not have information
regarding colonic motility in patients with a favorable outcome.

DISCUSSION
This study is the first to evaluate both the short and long-term
outcome of anal sphincter BoTox therapy in a large number of
children with nonrelaxing IAS. We found that BoTox was effective, as it was associated with an improvement in short and longterm gastrointestinal obstructive symptoms and/or enterocolitis
episodes, a short-term decrease in the need for rectal therapies,
and a significant decrease in hospitalization rates for those
The American Journal of GASTROENTEROLOGY

PEDIATRICS

Cumulative probablity of success

1.0

PEDIATRICS

982

Chumpitazi et al.

who had been previously hospitalized. We also found that it

was safe, with ~11% having transient, minor complications.
A total of 89% of the patients had an initial clinical improvement. This short-term success following BoTox injection was
comparable between those with IAS achalasia and those with
HD. The relatively high initial success corroborates the findings
seen in previous smaller studies with an initial improvement in
18 of 29 (90%) (5) of those with IAS achalasia, 14 of 18 (77.7%)
(22) with HD after corrective surgery, and 22 of 24 (91.7%) (24)
of those with IAS dysfunction.
In comparison, only 53% had a favorable long-term clinical
outcome. This decrease in long-term as compared to short-term
efficacy has been described in other gastrointestinal conditions
such as esophageal achalasia (26) in which BoTox therapy has
been used. Recognizing that BoTox provides a transient direct
effect estimated to last only for 34 months (19), the sustained
long-term response found in some children suggests that mechanisms beyond a transient decrease in anal sphincter pressure
are likely playing a role. It has been suggested (5) that secondary anatomic changes such as decreased rectosigmoid dilation
may improve function. Other mechanisms may involve permanent or more long-lasting unknown alterations in IAS function
with BoTox usage, interval development and maturity of defecation dynamics that may be facilitated by easier evacuation
after therapy, development of better behavioral coping methods
or strategies to address functional contributions, or intermittent
waxing and waning of symptoms. Given the chronic nature of
the symptoms in children with a nonrelaxing IAS, and that the
improvements occurred only after BoTox injection in otherwise
intractable patients suggests BoTox therapy played a strong role
in the improvements that occurred. Future studies following
functional and physiologic parameters in a prospective manner
may be able to provide further insight into the mechanisms of
success in those with favorable long-term clinical outcomes.
The long-term response to the BoTox varied according to the
underlying diagnosis as children with IAS responded better than
those with Hirschsprungs. The differences in baseline characteristics (e.g. gender distribution, sphincter pressure) and long-term
clinical outcomes between those with HD and IAS achalasia suggests different underlying pathophysiologic processes between the
two groups despite sharing a nonrelaxing IAS. Factors potentially
influencing unfavorable outcome in those with HD may include
a less functional post-surgical neorectum (or other surgical anatomic changes), a higher prevalence of colonic dysmotility (27), or
persistent enteric nervous system abnormalities. Higher symptom
severity before BoTox as reflected by factors such as increased rectal therapy usage may also have a function. The varied long-term
response may also be related to the natural history of IAS achalasia
as compared to HD, as the natural history of IAS achalasia is currently uncharacterized.
We also found that in the group with HD, the initial extent
of the aganglionosis was not a significant factor in the shortterm or long-term response to BoTox. This is in contrast to
the study by Minkes and Langer (22), in which a short-term
improvement was more likely in those with short-segment
The American Journal of GASTROENTEROLOGY

disease. Explanations for this difference may include having

more children with surgically repaired HD in our analysis, and
differences in regard to the types of corrective surgery and extent
of the original aganglionosis in our population. Nonetheless,
these results imply that BoTox may have equal efficacy in children who originally had short or long-segment aganglionosis.
Finding that the initial short-term response to the anal
sphincter BoTox injection, particularly in those with IAS achalasia, predicts long-term clinical outcome is not surprising as
one would not expect those who did not initially respond to do
well in long term. This finding is supported by the lack of a favorable long-term outcome seen in those children who received
multiple BoTox injections despite not improving after the first
injection. This may be due in part to the obstructive symptoms
and/or enterocolitis episodes not being solely related to the
nonrelaxing IAS. Other factors that cannot be addressed with
anal sphincter BoTox alone such as generalized colonic dysmotility (as was seen in intractable patients) may have played a
role, leading to a need for further interventions. The need to
consider more diffuse dysmotility is supported by the abnormal
colonic manometry evaluations found in all 9 children tested
with intractable symptoms following BoTox therapy.
The application of BoTox was safe, and there were no major
complications. Overall 10.8% had minor complications, with
the most common one being fecal incontinence, which was of
short duration and easy to manage in most patients. All fecal
incontinence events resolved within 4 weeks, comparing favorably to the long-term incontinence, which may result from anal
sphincter myectomy (13,14). We found that transient incontinence was more common in children with HD who originally
had long segment involvement. One child with long segment
HD experienced more prolonged fecal incontinence, and in
that case the anal sphincter pressure was low. No other patient
reviewed had an anal sphincter pressure less than 50 mm Hg.
It has been suggested that anal sphincter BoTox therapy may be
used as a diagnostic tool to predict which patients will respond to
a successful subsequent anal sphincter myectomy (3). This is only
partially supported by the results of this study. All children who
did not improve after an initial BoTox and subsequently underwent a myectomy continued to require further surgical interventions, suggesting that a lack of response to BoTox predicts failure
of myectomy. However, of those who improved with BoTox and
subsequently underwent myectomy, the majority 4 of 7 (57%)
also continued to require further interventions, indicating that
an initial response to BoTox does not ensure long-term response
after a myectomy. Prospective studies to address the relationship
between BoTox response and myectomy outcome are needed.
The primary weakness of this study is that it is retrospective.
As such, follow-up intervals and the number of BoTox injections given were not standardized. Factors such as concomitant laxative usage, dietary modifications, behavioral changes
or interventions, or other medical therapies such as outpatient
antibiotic regimens were not accounted for. The threshold for
seeking evaluation after BoTox therapy may have varied in the
studied population. Symptom improvement was subjective and
VOLUME 104 | APRIL 2009 www.amjgastro.com

Children With Nonrelaxing Internal Anal Sphincter

ACKNOWLEDGMENTS

We thank Dr William Klish, Dr Rachel Rosen, and Dr Robert

Shulman for their helpful criticism and thoughtful comments.
We also thank Jessica Lewis for her technical assistance.
CONFLICT OF INTEREST

Guarantor of the article: Samuel Nurko, MD, MPH.

Specific author contributions: Conception, design, and
performance of the study: Bruno P. Chumpitazi, Steven
J. Fishman, and Samuel Nurko; monitoring of data
acquisition, creation of database, and data cleaning: Bruno
P. Chumpitazi and Samuel Nurko; analysis and interpretation:
Bruno P. Chumpitazi, Steven J. Fishman, and Samuel Nurko;
critical revision of paper: Bruno P. Chumpitazi, Steven
J. Fishman, and Samuel Nurko; all listed authors have seen
and have approved the submitted paper. All authors take full
responsibility for the content of the paper.
Financial support: None.
Potential competing interests: None.

Study Highlights
WHAT IS CURRENT KNOWLEDGE
Children with a nonrelaxing internal anal sphincter (IAS)
may develop obstructive symptoms and/or enterocolitis.
Usual treatment includes surgical myectomy which may
have long-term side effects.
Anal sphincter Clostridium botulinum toxin (BoTox)
injection is a nonsurgical approach to decrease anal
sphincter pressure.
Preliminary studies show it provides short-term
improvement in these children.
However, the long-term results of BoTox therapy in
children with nonrelaxing IAS is unknown.
WHAT IS NEW HERE
BoTox decreased hospitalization rates in previously
hospitalized children with nonrelaxing IAS.
Complications of BoTox therapy in children with
nonrelaxing IAS are low.
Having an initial improvement after BoTox and having
IAS achalasia predicted a favorable long-term outcome.
BoTox may provide a safe, long-term therapy for children
with nonrelaxing IAS.

3
3
3
3
3
3
3
3
3

2009 by the American College of Gastroenterology

REFERENCES
1. Miele E, Tozzi A, Staiano A et al. Persistence of abnormal gastrointestinal
motility after operation for Hirschsprungs disease. Am J Gastroenterol
2000;95:122630.
2. Hyman PE. Defecation disorders after surgery for Hirschsprungs disease.
J Pediatr Gastroenterol Nutr 2005;41 (Suppl 1): S623.
3. Langer JC. Persistent obstructive symptoms after surgery for Hirschsprungs
disease: development of a diagnostic and therapeutic algorithm. J Pediatr
Surg 2004;39:145862.
4. Bentley JF. Faecal soiling and anal achalasia. Arch Dis Child 1978;53:
1856.
5. Ciamarra P, Nurko S, Barksdale E et al. Internal anal sphincter achalasia in
children: clinical characteristics and treatment with Clostridium botulinum
toxin. J Pediatr Gastroenterol Nutr 2003;37:3159.
6. Neilson IR, Yazbeck S. Ultrashort Hirschsprungs disease: myth or reality.
J Pediatr Surg 1990;25:11358.
7. Martucciello G, Pini Prato A, Puri P et al. Controversies concerning diagnostic guidelines for anomalies of the enteric nervous system: a report from
the fourth International Symposium on Hirschsprungs disease and related
neurocristopathies. J Pediatr Surg 2005;40:152731.
8. Engum SA, Grosfeld JL. Long-term results of treatment of Hirschsprungs
disease. Semin Pediatr Surg 2004;13:27385.
9. Yanchar NL, Soucy P. Long-term outcome after Hirschsprungs disease:
patients perspectives. J Pediatr Surg 1999;34:115260.
10. Zaslavsky C, Loening-Baucke V. Anorectal manometric evaluation of
children and adolescents postsurgery for Hirschsprungs disease. J Pediatr
Surg 2003;38:1915.
11. Blair GK, Murphy JJ, Fraser GC. Internal sphincterotomy in postpull-through Hirschsprungs disease. J Pediatr Surg 1996;31:8435.
12. De Caluwe D, Yoneda A, Akl U et al. Internal anal sphincter
achalasia: outcome after internal sphincter myectomy. J Pediatr Surg
2001;36:7368.
13. Heikkinen M, Lindahl H, Rintala RJ. Long-term outcome after internal
sphincter myectomy for internal sphincter achalasia. Pediatr Surg Int
2005;21:847.
14. Wildhaber BE, Pakarinen M, Rintala RJ et al. Posterior myotomy/myectomy for persistent stooling problems in Hirschsprungs disease. J Pediatr
Surg 2004;39:9206; discussion 9206.
15. Heikkinen M, Rintala R, Luukkonen P. Long-term anal sphincter
performance after surgery for Hirschsprungs disease. J Pediatr Surg
1997;32:14436.
16. Brisinda G, Cadeddu F, Brandara F et al. Randomized clinical trial comparing botulinum toxin injections with 0.2 per cent nitroglycerin ointment for
chronic anal fissure. Br J Surg 2007;94:1627.
17. Brisinda G, Maria G, Bentivoglio AR et al. A comparison of injections of
botulinum toxin and topical nitroglycerin ointment for the treatment of
chronic anal fissure. N Engl J Med 1999;341:659.
18. Cheng CM, Chen JS, Patel RP. Unlabeled uses of botulinum toxins: a
review, part 2. Am J Health Syst Pharm 2006;63:22532.
19. Cheng CM, Chen JS, Patel RP. Unlabeled uses of botulinum toxins: a
review, part 1. Am J Health Syst Pharm 2006;63:14552.
20. Gui D, Rossi S, Runfola M et al. Review article: botulinum toxin in the
therapy of gastrointestinal motility disorders. Aliment Pharmacol Ther
2003;18:116.
21. Jones OM, Brading AF, Mortensen NJ. Mechanism of action of botulinum
toxin on the internal anal sphincter. Br J Surg 2004;91:2248.
22. Minkes RK, Langer JC. A prospective study of botulinum toxin for internal
anal sphincter hypertonicity in children with Hirschsprungs disease.
J Pediatr Surg 2000;35:17336.
23. Messineo A, Codrich D, Monai M et al. The treatment of internal
anal sphincter achalasia with botulinum toxin. Pediatr Surg Int
2001;17:5213.
24. Irani K, Rodriguez L, Doody DP et al. Botulinum toxin for the treatment of
chronic constipation in children with internal anal sphincter dysfunction.
Pediatr Surg Int 2008;24:77983.
25. Nurko S. Complications after gastrointestinal surgery. A medical perspective. In: Walker WA (ed). Pediatric Gastrointestinal Disease, 4th edn. B.C.
Decker: Philadelphia, 2004, pp. 211138.
26. Pehlivanov N, Pasricha PJ. Achalasia: botox, dilatation or laparoscopic
surgery in 2006. Neurogastroenterol Motil 2006;18:799804.
27. Di Lorenzo C, Solzi GF, Flores AF et al. Colonic motility after
surgery for Hirschsprungs disease. Am J Gastroenterol 2000;95:
175964.

The American Journal of GASTROENTEROLOGY

PEDIATRICS

based on history. However, some of the nonstandardized factors

may have been ameliorated by the fact that one of the investigators (SN) was involved in the care of all the patients. Therefore,
medical management and the interventions used tended to be
uniform throughout the population studied.
The results suggest that BoTox therapy has long-term efficacy
in certain children with nonrelaxing IAS. Given a high initial success rate with the possibility of long-term therapeutic benefit, a low
number of transient adverse effects, and a significant decrease in
hospitalization rates, BoTox may provide a safe short-term and
long-term therapeutic option for children with a nonrelaxing IAS.

983