Journal of Pediatric Surgery (2012) 47, 2096–2100

www.elsevier.com/locate/jpedsurg

Surgical treatment of perianal abscess and fistula-in-ano in

childhood, with emphasis in children older than 2 years☆
Anestis Charalampopoulos a,⁎, Nikolaos Zavras b , Emmanouil I. Kapetanakis a ,
Kostantinos Kopanakis a , Evangelos Misiakos a , Pavlos Patapis a ,
Georgios Martikos a , Anastasios Machairas a
Third Department of General Surgery, Athens University Medical School, “Attikon” University Hospital, Athens, Greece
a

Department of Pediatric Surgery, Athens University Medical School, “Attikon” University Hospital, Athens, Greece
b

Received 14 November 2011; revised 18 May 2012; accepted 25 June 2012

Key words: Abstract

Perianal sepsis;
Background: Anal sepsis in children ranges from perianal abscess to fistula-in-ano. It is mostly observed
Children;
in boys younger than 2 years. Most are treated conservatively. In contrast, anal sepsis in older children
Fistula-in-ano;
presents significant similarities to that of adults and is predominantly treated surgically. We report our
Perianal abscess;
outcomes after surgical treatment of anal abscess and fistula-in-ano in children older than 2 years.
Surgical drainage;
Patients and Methods: Ninety-eight (98) children were operated on for anal abscess (46 patients; 47%)
Fistulotomy
and/or fistula-in-ano (52 patients; 53%). Incision and drainage of the abscess was performed as
outpatients. In patients with fistulas, fistulotomy was the main treatment approach. All patients were
healthy without risk factors for anal sepsis.
Results: In patients with anal abscess treated with incision and drainage, low recurrence (13%) or fistula
formation rates were observed. Most anal fistulas were simple entities. Significant involvement of the
anal sphincter was found in 3 (6%) of 52 patients. An abscess cavity between the anal canal and the
perianal skin was found in 4 (8%) of 52 patients, and an enlarged cryptic gland was found in 5 (10%) of
52 cases. Fistulotomy was performed in all patients with additional seton placement in 3 (6%) of 52 and
a cryptotomy in 5 (10%) of 52 patients.
Conclusions: Anal abscesses in children are easily treated by incision and drainage with low recurrence
of perianal sepsis. Fistulas can be treated successfully in most patients with a fistulotomy, whereas
complex fistulas are uncommon.
© 2012 Elsevier Inc. All rights reserved.

Perianal abscess (PA) and fistula-in-ano (FIA) in children

☆
Institution: Third Department of General and Paediatric Surgery,
Athens University Medical School, “Attikon” University Hospital, Athens,
Greece.
⁎ Corresponding author. Tel.: + 30 210 2844894; fax: + 30 210
2844894.
E-mail address: chalaral@med.uoa.gr (A. Charalampopoulos).
are rare compared with adults. Most individuals affected
are young boys, usually younger than 1 year. Conservative
treatment plays a significant role in disease management
because spontaneous resolution is usually the norm [1]. Older
children are not, however, excluded from this pathology, and

0022-3468/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jpedsurg.2012.06.032
Anal abscess and fistula-in-ano in childhood
if enteric flora is present in the pus of the abscess, a fistula
may form [2]. Overall, PA and FIA seem to be a different
entity in infants when compared with older children or adults
[3]. The characterization of PA and FIA in children is
descriptive, based mainly on clinical features [1]. They are
usually superficial entities close to the anus; however the
variability in size, precise location, and relation to the anal
sphincter system, combined with the lack of information on
possible fistulous tract routes and the lack of contemporane-
ous imaging studies such as endoscopic ultrasound or nuclear
magnetic resonance imaging, have contributed to the
confusion and variability regarding conservative vs surgical
treatment [3]. Therefore, at the present time, PA and FIA in
children remain unclassified, without consensus data regard-
ing treatment and surgical techniques that should be utilized.
Surgical treatment remains empiric, based on the skills and
experience of the surgeon in anorectal surgery. The
recurrence of an abscess or the development of a fistula,
after the initial treatment of PA (conservative with antibiotics
or surgical drainage), is a possibility and is reported in the
literature to range from 5% to 80% [3,4]. The goal of this
study was to retrospectively analyze our experience and
present the outcomes of surgical treatment of PA and FIA in
older children.
1. Patients and methods
This is a retrospective consecutive case-series analysis.
The medical records of children ranging from 0 to 15 years of
age presenting with a PA or an FIA over the last decade
(January 1, 2001–November 31, 2011), were reviewed. The
study received approval from the local institutional ethical
committee, but individual patient consent was waived
because the data collected initially and during follow-up
were part of routine clinical practice.
All children presenting with a PA and treated by surgical
drainage were included in the study. A perianal abscess (PA)
was defined as a superficial, inflamed mass near the anus.
Children were referred as surgical outpatients either “in-
house” by the pediatric service or by their own pediatrician.
All had received antibiotic therapy for a median period of 6
(range, 4-15) days. Surgical drainage was performed, under
local anesthesia, in an outpatient setting. All parents were
informed postprocedure about the possibility of a recurrence
in the form of a new abscess or a FIA. Patients were
monitored and reexamined at least twice during a period of
2 to 10 months for the presence of a fistula. A personal
history and clinical presentation were recorded for all
children. All children in the study were healthy, and indi-
viduals with predisposing diseases to abscess or fistula
formation such as Crohn disease, diabetes, immunosuppres-
sant conditions, previous anal disease (eg, anal fissure) and
Hirschsprung disease were excluded from the analysis. Such
conditions necessitate a specific management approach to the
2097
main disease, and the results of surgery vary in each case.
Patients with recurrent anal sepsis were also excluded from
the analysis.
The incision of the abscess was executed over the
dome of the pus collection (abscess), and the cavity was
gently explored with sinus forceps. If possible, microbial
cultures were obtained. The cavity was unified and washed
with an antiseptic solution. Discharge of the antiseptic
solution through the anal canal revealed the existence of a
fistulous tract joining the abscess cavity with the lumen of
the anal canal.
A fistula-in-ano was defined as a fistulous track with 2
openings: 1 internal in the anorectal lumen and 1 external in
the perianal skin. The diagnosis of a fistula was clinical, by
careful macroscopic examination of the perianal skin and by
identifying the external opening of a fistulous track. Such
openings could have been active or inactive, depending on
the presence of secretions or pus. Inactive external openings
were probed carefully and gently with small sized probes, to
discover the initial external segment of the fistulous track and
differentiate an inactive opening from a skin deformity crypt.
The complete probing of a fistula and the location of the
internal anorectal opening was possible only under general
anesthesia at the time of surgery. Patients were candidates for
surgery only when there was a long-standing fistula of more
than 3 to 4 months duration.
A fistulotomy with/or without the placement of a seton
and/or a cryptotomy was the routine surgical management of
long-standing fistulas. During surgery, a number of data
were recorded to assess the severity, the length and the
course of the fistula, and the efficacy of the procedure
performed. A literature-based comparison with adult fistula
disease was done, and fistulas were classified as follows:
1. According to their external-internal opening and the
route of the fistulous tract, which was categorized
either as anterior or as posterior based on Salmon and
Goodssal's rule, with anterior fistulas representing
simple, easily probed, and surgically treated entities
2. According to the coexistence of an abscess cavity in
the fistulous tract's course
3. By the fistulous track's site and direction
4. By the inclusion of the anal sphincter system in the
fistulous track (All children were observed for 1 year in
case of fistula recurrence.)
2. Results
A total of 98 patients (89 boys [91%] and 9 girls [9%])
were operated on for a PA or FIA. Of those, 46 (47%)
presented with only a PA, whereas 52 (53%) presented with
an FIA. Six patients with FIA were referred in-house and 46
were referred and diagnosed in the outpatient clinic. The
median age of all patients was 7 (range, 0-15) years. In the
2098 A. Charalampopoulos et al.

group of patients with PA, there were only 2 infants (mean
age, 4.8 months) needing treatment by incision and drainage,
whereas the rest of the children were older than 2 years.
Complete patient characteristics are presented in Table.
After drainage of the PA, 6 patients (13%) developed a
subsequent FIA. None of the infants previously mentioned
develop an FIA in the follow-up period. The median follow-
up was 6 (range, 2-10) months, with 9 children lost to follow-
up (91% complete).
In patients with PA, no severe clinicolaboratory signs of
sepsis were found, and only 7 patients (7/46; 15%) had
moderate fever of 38.0°C to 38.8°C. Microbial cultures grew
a pathogen in 28 cases (Table). The most common microbe
was Escherichia coli (20/28; 71%). In the rest (6/28; 22%),
mixed enteric flora (E coli, Enterococci, Bacteroides fragilis)
was found, whereas skin flora (Staphylococcus aureus,
Streptococcus group C) were grown in 2 cases (7%). Oral
antibiotics were administered postdrainage in all children.
Patients with FIA had, in their history, a long-standing
fistula, spanning from 4 months to 4 years. The main symp-
tom was the intermittent excretion of small amounts of pus or
liquid secretions from the external opening of the fistula. All
cases underwent fistulotomy under general anesthesia.
During surgery, careful probing of the fistulas revealed the
complete length and route of their fistulous tract. No false
route was formed while probing the fistulas. According to
Salmon and Goodssal's classification, 14 fistulas were
anterior (14/52; 27%) and 38 were posterior (38/52; 73%),
(Table). Anterior fistulas were probed more easily than
posterior ones because of their direct fistulous tract, and a
fistulotomy was thus easily performed. Posterior fistulas
were probed with more difficulty, often presenting with a
curved tract. In 4 cases (4/52; 8%), an abscess cavity with
small amounts of pus was interposed in the route of
the fistula. Microbial cultures obtained during surgery grew
E coli in 3 cases and mixed enteric flora (E coli and Enter-
ococci) in 1.
Two FIAs were intersphincteric, whereas 2 were in the
perianal area in an extrasphincteric location, on the internal
and lower ischioanal fossa, close to the anus. A partial
involvement of the sphincter system in the fistulous tract was
found mainly in posterior fistulas. In those cases, a muscle
fiber incision of the sphincter system was carefully
performed so as to avoid incontinence. Three patients
(3/52; 6%) presented with a transsphincteric fistula with
significant involvement of the muscular sphincter (Table). In
these cases, a seton was placed to avoid incontinence
postoperatively. The seton was progressively and gradually
tightened over a course of 20 to 30 days. In 5 patients (5/52;
10%) an inflamed cryptic gland could not be excluded, and a
small cryptotomy was therefore performed (Table).
All fistulas were short (b 3 cm in length). No fistula
recurrence was observed during a median follow-up period
of 10 months (range, 9-12 months). No cases of incontinence
were diagnosed postoperatively.

3. Discussion
Table Patient and disease characteristics
n (%) Range Anal sepsis in children ranges from PA to FIA. Most
Male sex 89 (91) cases are infants younger than 1 year with a clear male sex
Female sex 9 (9) predominance, demonstrating a congenital predisposition to
Median age (y) 7 0-15 the disease [3,5]. However, older children are not excluded.
Perianal Abscess 46 (47) In infants, the disease is most often treated conservatively
Symptoms of sepsis (fever) 7 (15) with a low incidence of recurrence or FIA development. In
Antibiotic treatment duration (d) 7 4-15 older children, however, the disease presentation is more
Pus cultures
akin to that of adults. Our retrospective review focused
E coli 20/28 (71)
Mixed enteric flora 6/28 (22)
therefore on evaluating the outcome of surgical treatment in
Skin flora 2/28 (7) children older than 2 years.
Postdrainage fistula-in-ano 6 (13) Our study also included 2 infants with PA, which
Length of follow-up (mo) 6 2-10 required incision and drainage because of abscess size.
Fistula-in-ano 52 (53) Currently, PA treatment in infants is nonoperative because
Fistula characteristics surgical drainage produces a high rate of recurrence, either
Anterior fistulas 14 (27) as an abscess or as a fistula, ranging from 28% to 85% [6-8].
Posterior fistulas 38 (73) Surgical drainage is usually considered a second-line
Concurrent abscess cavity 4 (8) therapy and performed only rarely, when there is extreme
Transsphincter tract 3 (6) local discomfort, a large abscess, or conservative treatment
Operative characteristics
has failed. Infant abscesses have no significant clinicola-
Simple fistulotomy 44 (84)
Fistulotomy with seton placement 3 (6)
boratory signs of sepsis because the principal pathogenic
Fistulotomy with cryptotomy 5 (10) microorganism, lactobacilli, is not considered especially
Post–fistulotomy recurrence 0 pathogenic. The use of antibiotics, despite the lack of
Length of follow-up (mo) 10 9-12 microbial-enteric flora participation, is usually justified for
surgical prophylaxis after incision and drainage and because
Anal abscess and fistula-in-ano in childhood
there are strong indications that synchronous antibiotic
administration is associated with decreased rates of fistula
formation [6,9]. None of the infants that underwent drainage
in our study developed a subsequent fistula during the
follow-up period.
In our patient population with a PA, a perianal sepsis
recurrence rate of 13% was demonstrated after surgical
drainage. This rate is significantly lower when compared
with adults, where a rate of more than 35% is expected [10].
There were 52 cases with FIA, all of which were older
than 2 years. All had a history of anal abscesses, which had
either spontaneously drained or were treated conservatively
with antibiotics or/and needle aspiration. Surgery in these
children was necessary because it offers optimal therapeutic
results comparable with adult fistulas.
There are a number of proposed theories to explain the
etiology and FIA formation in children; however, the exact
pathophysiologic mechanism remains obscure. The prevail-
ing theories are the congenital theory [11], which implicates
the excess production of androgens, and the one proposed by
Shafer et al [12], which is based on the existence of abnormal
crypts of Morgagni. The cryptoglandular theory, which is
valid for the adult population, may also apply to older
children. In our 52 patients with FIA, cryptic infection was
supported in 9 cases: 4 with an abscess cavity interposed in
the fistulous tract's route and a positive enteric flora culture
and 5 with an enlarged cryptic gland demonstrated on
anoscopy before fistulectomy. Therefore, an acute infection
of the anal glands with transmission of the infection in the
intersphincteric space, as an initial event before fistula
formation, seems possible. This is similar to the disease
process found in adults. In contrast, the progression of
infection, first with formation of an intersphincteric abscess
and the subsequent evolution to an FIA, seems to be quite
different in children compared with adults.
All our fistulas were treated by fistulotomy and were
characterized as simple ones, of short length (b 3 cm), a low
transsphincter route, and minimal involvement of the
sphincter system. Therefore, surgical incision of some mus-
cular fibers of the lower internal sphincter and the superficial
segment of the external sphincter during fistulotomy could
be performed safely, without the fear of postprocedural
incontinence. In transsphincter fistulas, the surgeon must
carefully estimate the amount of muscle fibers contained in
the fistulous tract, in relation to the sphincter system, and
decide to either cut the fibers or place a cutting seton [13].
Noncutting setons should be used only in complex fistulas
such as high transsphincteric or suprasphincteric perianal
fistulas with sepsis whose management remains a major
surgical challenge [14].
Fistulotomy with placement of a cutting seton was
performed only in 3 cases (3/52; 6%) because the thickness
of the involved sphincter muscle meant that surgical incision
could be unsafe and result in incontinence. Despite the
significant involvement of the sphincter system in these 3
cases, they would not be normally characterized as high
2099
transsphincteric. Seton tightening was intermittent and
gradual with complete fall-off achieved in 20 to 30 days.
The disadvantage of using a seton in children is the
associated anal pain, but because tightening is slight and
gradual, it usually disappears after some hours. Currently, the
use of setons remains empiric, and there are no data available
for the efficacy of this surgical technique in children.
However, the slow, gradual tightening approach used and the
delay in seton fall-off seem to be safe for the child's anal
continence. All of these 3 cases were posterior fistulas
according to Salmon and Goodsall's rule.
In adults, the most common FIA classification system
used by surgeons is that of Park et al [15], which classifies
perianal fistulas as intersphincter, transsphincter, supras-
phincter, and extrasphinchter. Most cases (N 70%) are simple
intersphincter fistulas, with the remainder being more
complex. The recent widespread use of nuclear magnetic
resonance imaging has proved a useful tool for the study and
classification of complex fistulas, aiding the decision for
surgical management [16]. Complex fistulas account for
more than 20% to 30% in the adult population and may be
high transsphincter, supralevator, extrasphincter, transleva-
tor, or horseshoe fistulas. These necessitate more advanced
surgical techniques and skills. Fortunately, no complex
fistulas were found in our study population because they are
more prevalent in patients with inflammatory bowel
diseases, patients with immunosuppressant conditions,
patients with previous anorectal diseases, or patients with
recurrent anal sepsis, which were excluded from this
analysis. The lack of FIA recurrence after surgery in the
patients studied demonstrates that simple fistulotomy and the
placement of a seton or a cryptotomy when applicable are
sufficient surgical treatments in children. This is another
significant difference compared with adults, where a
recurrence rate for perianal sepsis of more than 10% to
20% is expected after surgical treatment [17]. Fistulotomy
seems to be an excellent and safe surgical technique for most
children with anal fistulas.
4. Conclusion
Perianal sepsis in children older than 2 years necessitates
surgical treatment. In cases of PA, incision and drainage is an
easy, fast, and safe therapy with a low recurrence rate. In
children with FIA, most are simple presentations without any
significant involvement of the sphincter system. Simple
fistulotomy is the most common surgical technique used and,
when required fistulotomy with seton placement or cryptot-
omy, more than meet the needs for surgical treatment.
References
[1] Nelson R. Anorectal abscess fistula: what do we know? Surg Clin
North Am 2002;8(2):1139-51.
2100
[2] Nix P, Stringer M. Perianal sepsis in children. Br J Surg 1997;84:
819-21.
[3] Serour F, Somekh E, Gorenstein A. Perianal abscess and fistula-
in-ano in infants: a different entity? Dis Colon Rectum 2005:
359-64.
[4] Ezer SS, Oguzkurt P, Ince E, et al. Perianal abscess and fistula in ano in
children: aetiology, management and outcome. J Paediatr Child Health
2010;46:273-4.
[5] Murthi GVS, Okoye BO, Spicer RD, et al. Perianal abscess in
childhood. Pediatr Surg Int 2002;18:689-901.
[6] Christison-Lagay ER, Hall JF, Wales PW, et al. Non-operative
management of perianal abscess in infants is associated with
decreased risk for fistula formation. Pediatrics 2007;120:e548-52.
[7] Abott MB. Surgeons say no to surgery for perianal abscesses in infants.
Pediatrics 2007;120:646.
[8] Fasten C, van Harten H. Perianal abscess and fistula in ano in infants.
J Pediatr Surg 1998;33:711-3.
[9] Bruzzese E, Canani RB, De Marco G, et al. Microflora in inflam-
matory bowel diseases: a pediatric perspective. J Clin Gastroenterol
2004;38:S9.
A. Charalampopoulos et al.
[10] Hamadani A, Haigh PI, Liu IL, et al. Who is at risk for developing
chronic anal fistula or recurrent anal sepsis after initial perianal
abscess? Dis Colon Rectum 2009;52:217-21.
[11] Fitzerald RJ, Harding B, Ryan W. Fistula in ano in childhood: a
congenital aetiology. J Pediatr Surg 1985;20:80-1.
[12] Shafer AD, McGlone TP, Flanagan RA. Abnormal crypts of
Morgagni: the cause of perianal abscess and fistula-in-ano. J Pediatr
Surg 1997;22:203-4.
[13] Ikeda T, Inoue M, Sugitou K, et al. Treatment of fistula in ani in infants
with a seton. J Pediatr Surg 2007;42:1095-7.
[14] Subhas G, Gupta A, Balaraman S, et al. Non-cutting setons for
progressive migration of complex fistulas tracts: a new spin on an old
technique. Int J Colorectal Dis 2011;26:793-8.
[15] Parks AG, Gordon PH, Hardcastle JD. A classification of fistula in
ano. Br J Surg 1976;63:1-12.
[16] Morris J, Spencer JA, Ambrose NS. MR imaging classification of
perianal fistulas and its implications for patient management.
Radiographics 2000;203:623-35.
[17] Fisher JE. Mastery of surgery, 5th ed, Vol. II. Philadelphia, Lippincott
Williams & Wilkins, Treatment of anorectal fistulae. 2007:1611–1612.