Anal Fistulas & Surgery

8/9/2019 Anal Fistulas & Surgery

1/9

ANZ J. Surg. 2005; 75: 64–72

REVIEW ARTICLE

REVIEWARTICLE

ANAL ABSCESSES AND FISTULAS

MATTHEW J. F. X. R

ICKARD

Department of Colorectal Surgery, Concord Hospital, Sydney, New South Wales, Australia

Anal abscesses and fistulas are a common part of surgical practice. Most abscesses simply need to be drained and most fistulas can be

safely laid open. Excessive probing should not be attempted when draining abscesses as this may lead to iatrogenic fistulas. A small

percentage of fistulas are complex and very challenging to manage. Management involves an accurate diagnosis and a balance

between eradication of the fistula and maintenance of continence. A decision should be made, based on clinical evaluation and anal

ultrasound (if available), whether the fistula can be laid open. If it cannot be laid open, a loose seton is placed and the sepsis is

allowed to settle. Once the sepsis is quiescent, a definitive repair can be attempted. There are various techniques available including

rectal advancement flap, fibrin glue and cutaneous flaps all of which are discussed.

Key words: anal fistula, peri-anal abscess.

Abbreviations

: AIN, anal intraepithelial neoplasia; ATZ, anal transitional zone; ES, external sphincter; EUA, examination

under anaesthetic; EUS, endo-anal ultrasound; IS, internal sphincter; MRI, magnetic resonance imaging; PPV, positive predic-

tive value; RAF, rectal advancement flap; RCT, randomized controlled trial.

INTRODUCTION

Anal abscesses and fistulas are a common surgical condition.

Management of the majority of anal abscesses and fistulas is

straightforward and is based on a sound knowledge of the

anatomy of the anorectum and adherence to established surgical

principles. A small percentage of anal abscesses and fistulas are

complex and very challenging to manage surgically. The joint

aims of maintaining perfect continence and curing the fistula are

often difficult to achieve.

ANATOMY

An understanding of the anatomy of the anal canal is essential for

the appropriate management of anal fistulas. The external sphinc-

ter (ES) is a continuation of the pelvic floor musculature. The

internal sphincter (IS) is a continuation of the inner circular

muscle layer of the lower rectum (Fig. 1). These muscle layers are

easily appreciated on endo-anal ultrasound (EUS). The IS appears

as a hypo-echoic ring. The ES is identified by first identifying the

puborectalis sling at the anorectal junction. This is hyper-echoic

and U-shaped. Below this the ES commences when the open end

of the U begins to close to form a complete ring of muscle.

1–4

The mucocutaneous junction is the site of the dentate line (the

term pectinate line should be discarded).

5

The epithelium of the

anal canal is mucosa above the dentate line and stratified non-keratinized squamous epithelium below. The dentate line is the

site of the anal valves. Proximal to each anal valve is an anal

crypt or sinus, which macroscopically appears as a small pit. The

anal glands, which lie in the intersphincteric plane, empty into

these anal crypts. For a distance of 5–20 mm (varying with age)

above the dentate line, the mucosa is cuboidal and is known as

the anal transitional zone (ATZ).

6,7

This area is thought to be

important for discrimination between flatus and faeces. Above

the ATZ, the mucosa becomes columnar with typical goblet

mucus secreting cells.

AETIOLOGY

The cryptoglandular theory of Eisenhammer and Parks is now

widely accepted, although there have been very few subsequent

studies to confirm or refute it. Eisenhammer proposed that an

intramuscular anal gland became infected and because of subse-

quent infective obstruction of its connecting duct, it was unable todrain spontaneously into the anal canal.

8

Parks found cystic dila-

tation of anal glands in eight of 30 consecutive cases of anal fis-

tula. He attributes this to either acquired duct dilatation or a

congenital abnormality and suggested that it was a precursor to

infection within a mucin-filled cavity.

9

Infection begins in the

intersphincteric plane. If it tracks downwards in the intersphinc-

teric plane, a perianal abscess forms. If it tracks upwards in the

intersphincteric plane, a high intersphincteric abscess or supra-

levator abscess might form. Sepsis that tracks across the ES will

result in an ischio-rectal abscess. If this then extends upwards

across the levator plate, it may form a supralevator abscess

(Fig. 2). Horseshoe abscesses form from circumferential spread.

This is most common in the ischiorectal plane, but can occur in

the perianal and supralevator planes.

10

It has been suggested that most ‘high’ anal fistulas and most

degrees of incontinence represent iatrogenic and therefore avoid-

able complications of abscess or fistula surgery.

11

CLASSIFICATION

Initial classifications were made by Milligan and Morgan,

12

Steltzner,

13

Goligher,

14

and Eisenhammer

15

. These were refined

by Parks et al

. and his classification – intersphincteric, trans-

sphincteric, supra-sphincteric and extra-sphincteric (Fig. 3) – is

now the most widely used and taught.

16

Extensions (secondary

tracks) might occur in the intersphincteric plane, the ischiorectal

fossa and the pararectal (supralevator) space.

M. J. F. X. Rickard

MB BS, MMed (ClinEpi), DipPaed, FRACS.

Correspondence: Dr Matthew Rickard, Department of Colorectal Surgery,Concord Hospital, Sydney, NSW 2137, Australia.

Email: [email protected]

Accepted for publication 18 August 2004.


2/9

ANAL ABSCESSES AND FISTULAS 65

There are limitations to Parks’ classification. Superficial fistulas

are not classified because of the emphasis on the intersphincteric

plane. These are common and make up around 16% of one series.

17

The lowermost fibres of the ES curve in under the distal edge of the IS

and on EUS the lower third of the anal canal has no IS. In the author’s

experience, true intersphincteric fistulas, as classified by Parks, do not

occur. Even very low fistulas cross some of these lowest fibres of the

external sphincter and are, hence, trans-sphincteric.

The best treatment for an anal fistula is to lay it open. Obvi-

ously if this involves cutting a large amount of anal sphincter

this will cause incontinence. Bennett in 1962 said: ‘It is poor

consolation for the fastidious patient who, after 17 weeks off

work for treatment of his horseshoe fistula, finds that his under-

clothes are stained brown instead of yellow, even though his

fistula is healed’.

18

A more clinically useful classification is whether or not a

fistula can be laid open. This would be based on clinical and EUS

findings and depends not just on the type of fistula, but also the

sex of the patient and the extent of any pre-existing sphincter

abnormalities.

Fig. 2.

Supralevator extensions.

Fig. 1.

Anal canal anatomy.


3/9

66 RICKARD

ANORECTAL ABSCESS

Presentation of anorectal abscesses

Anorectal abscesses usually present with local pain and swelling.

Presenting features vary depending on the site of the abscess and

some patients may present simply with a pyrexia of unknown

origin. Abscesses can be classified as perianal, ischiorectal, inter-

sphincteric, or supralevator (Fig. 4).Perianal abscesses usually present with a tense, tender ery-

thematous swelling of short duration without any systemic signs

of toxicity. Ischiorectal abscesses often have a long history of

throbbing pain and can be associated with systemic toxicity.

Intersphincteric abscess presents with severe rectal pain and can

be mistaken for an anal fissure on history. There is often associ-

ated pyrexia (which should not occur in a simple fissure). Digital

rectal examination is extremely painful but might reveal a tender

nodule towards the upper end of the anal canal. A supralevator

abscess presents with pelvic pain and constitutional symptoms. It

might represent a true pelvic abscess from intra-abdominal

pathology or an intersphincteric or ischiorectal abscess that has

tracked upwards. Digital rectal examination may reveal a boggy

tender swelling in the rectum.

19

Management of anorectal abscesses

Abscesses require surgical drainage under general anaesthesia.

An intersphincteric abscess is readily identifiable on EUS and can

usually be drained endo-anally by excising a small disc of inter-

nal sphincter directly over the abscess. A supralevator abscess

that is an extension in the intersphincteric plane can often be

drained by opening into the intersphincteric plane in the upper

anal canal and then dissecting in the intersphincteric plane up to

the abscess. The abscess can be drained via this track and a small

mushroom tipped catheter (de pezzer) can be left in the cavity for

3–4 days to prevent recurrence.

Perianal and ischiorectal abscesses are much more accessible

and easy to drain. A supralevator extension of an ischiorectal

abscess can be drained through the ischiorectal fossa. Whether

an internal opening should be sought for at the time of draining

an anorectal abscess is a source of debate. The prevalence of anal

fistula formation ranges from 5 to 83%.

20–23

Subsequent fistula

formation is more likely with a perianal abscess than an ischio-

rectal abscess or if the abscess is complicated, e.g., horseshoe

abscess.

22,24

There have been six randomized clinical trials (RCT)

23,25–29

comparing immediate and delayed treatment of associated fistula

tracks. The first five of these were subjected to a crude meta-analy-

sis. The findings were that if the fistula track is not sought out and

treated when the abscess is drained, patients are 13 times (5.4–47

95%CI) more likely to have persistent problems with sepsis. How-

ever, if the track is sought out and treated there is an increased

incidence of incontinence (odds ratio 0.36 (0.13–0.93).

30

The sixth and most recent trial includes 100 patients in each

arm of the study.

23

In this series an anal fistula was identified in

83% of the group of patients where it was sought and treated

(group A). In the group of patients where only drainage of the

abscess was performed (group B), 29% went on to develop fistu-las. In groups A and B the overall incontinence rate at 1 year was

6 and 0%, respectively. The conclusion of the paper was that

‘drainage of anal abscess with fistulotomy can be safely per-

formed in cases of subcutaneous, intersphincteric, or low trans-

sphincteric fistulas with minimal recurrence’. However, if the

raw data is re-analysed, patients in group A were 2.8 times more

likely to have a fistula (relative risk 2.8 (2.1–3.9 95% CI),

P

< 0.0001) and infinitely more likely to have continence prob-

lems (relative risk = infinity, P

= 0.03). The obvious interpreta-

tion is that fistulas are present in the majority of cases of

anorectal abscess formation, but that many resolve spontaneously

such that aggressive early identification and treatment may be

unnecessary and lead to poorer outcomes.

Fig. 3.

Parks classification.


4/9


A reasonable approach is to inspect the anal canal with a suit-

able retractor before the abscess is drained. If moderate pressure

on the abscess leads to obvious extrusion of pus into the anal

canal at a readily identifiable internal opening, the abscess can be

drained and a Seton placed along the fistula track. It is always

safer to place a Seton in the track rather than lay open the track,

as the extent of sphincter involvement is difficult to define in the

presence of acute inflammation. If there is no obvious internal

opening with this simple manoeuver, the abscess should simply

be drained with an incision parallel with the anal canal (less likely

to cut through the lower most fibres of the ES). No attemptshould be made to look for a fistula with probes. In reality, inex-

perienced surgeons often perform this operation. Therefore, it

should be kept as simple as possible. It is the author’s practice to

use a 20–30 mm incision only, break down the loculations digit-

ally and place a de pezzer catheter into the cavity. The drain is

left until there is minimal purulent drainage, which may vary

from a few days to several weeks.

ASSESSMENT OF ANAL FISTULA

In 1900 Goodsall and Miles outlined five essential points in the

assessment of fistulas.

31

The principles remain unchanged.

●

The location of the internal opening (a fistula can never be

eradicated if this is not identified)

●

The location of the external opening

●

The course of the primary track

●

The presence of secondary extensions

●

The presence of other diseases complicating the fistula

Clinical assessment

In reality, the prevalence of fistula formation after drainage of an

anorectal abscess is around 30%. Persistent drainage from the

drainage site and/or recurrent abscess formation are usual indica-

tions that a fistula is present. The external opening is usually clear

and is at the site of previous drainage or at the site of spontaneous

drainage. The fistula track can often be felt between finger and

thumb when performing digital rectal examination. Any clinical

abnormalities of the anal sphincter, such as poor resting and/or

squeeze tone and deficiencies from obstetric or iatrogenic trauma

should be noted. No attempt should be made to pass probes in

an awake patient as it is very unpleasant and might create false

tracks. If there is no abscess formation requiring drainage, further

assessment either by EUS or examination under anaesthetic

(EUA) depending on available facilities is recommended.

Goodsall and Miles’ so-called ‘rule’ states that fistulas with an

external opening lying above a horizontal line drawn through the

centre of the anal canal, with the patient in lithotomy position,usually drain directly into the anal canal. Fistulas lying below this

horizontal line usually drain to the midline posteriorly.

31

Like all

rules, there are exceptions. Cirocco and Reilly put the rule to the

test in 1992. They reported a predictive accuracy of 90% for pos-

terior fistulas, but only 49% of anterior fistulas obeyed the rule.

32

Conversely, Gunawardhana and Deen, when comparing hydro-

gen peroxide installation with Goodsall’s rule, found a predictive

accuracy for posterior and anterior fistulas of 41 and 72%,

respectively.

33

Differential diagnoses should be considered in every case.

Hidradenitis suppurativa, tuberculosis, Bartholin’s abscess, actin-

omycosis, anal fissure, ulceration associated with Crohn’s dis-

ease, sexually transmitted diseases, and a low pilonidal sinus are

all possible differential diagnoses. A more important distinctionis the differentiation between a chronic abscess and a perirectal

dermoid cyst or a sacrococcygeal teratoma. If these lesions are

inadvertently opened total excision may prove difficult and a

sacrococcygeal fistula might develop, which is often impossible

to eradicate.

34

Imaging

In practice, most low fistulas do not require imaging. Assessment

and management is based on clinical evaluation and the findings

at EUA. In the current medico-legal environment an objective

assessment of the extent of muscle involved in a fistula is helpful

prior to laying it open. How this is done depends on the local

Fig. 4.

Abscess positions.


5/9

68 RICKARD

availability of imaging techniques. It is the author’s practice to

perform an EUS on all anal fistulas. This is performed either in

theatre under general anaesthesia or in the clinic without seda-

tion.

Endo-anal ultrasound

There are three criteria for identification of the internal opening

on EUS: (i) contact of the internal sphincter by the intersphinc-

teric track (positive predictive value (PPV) of 80%); (ii) an appar-

ent defect of the internal sphincter (PPV 79%); and (iii) a defined

subepithelial track associated with a localized sphincter defect

(PPV 94%).

35

The overall sensitivity using all three of these signs is 94%

with a specificity of 87% and a PPV of 81%. The gas reflections

from hydrogen peroxide highlight the track and may help to

locate the site of the internal opening.

36–40

Gold et al

. studied interobserver and intraobserver agreement

in assessing the anal sphincter and found interobserver agreement

to be very good (kappa = 0.84).

41

Conversely, Robinson des-

cribed error and variation in image interpretation as the Achilles’

heel of radiology. The human eye and brain have failed to keep

up with technological advances.

42

One of the limitations of EUS is the difficulty distinguishing

recrudescent sepsis from burnt-out fibrotic tracks in complicated

disease.

Magnetic resonance imaging

The advantages of magnetic resonance imaging (MRI) include

multiplanar imaging and high soft-tissue differentiation to

show the track system in relation to underlying anatomy in a

projection relevant to surgical exploration.

43

The accuracy of

MRI has been confirmed by comparison with surgery.

44–47

It is

especially helpful in assessing complex fistulas secondary to

Crohn’s disease.

48,49

Fistulography

Fistulography is certainly not the first line investigation. False

positive results for rectal openings and high extensions have been

reported

50

and might lead to inappropriate surgery. Fistulography

is an inaccurate, unreliable and possibly harmful procedure.

51

Chronic tracks may be patent, but many acute tracks are actually

just columns of inflamed granulation tissue without a patent

lumen. Contrast injection therefore may show only part of the

track system.

43

Nevertheless, for complex fistulas, particularly

those where repeated examination under anaesthesia has been

unsuccessful in defining the course of the track, fistulography

may be helpful in combination with either MRI and/or EUS.

Other investigations

Manometry

Manometry does not need to be done routinely, but might be

helpful on occasions to objectively assess sphincter function if a

decision is being made as to whether or not to lay open a fistula.

If the resting pressure appears abnormal clinically, manometry

may be used to confirm this clinical impression.

Examination under anaesthesia

If only an assessment is planned it is often more helpful to

perform the examination under light general anaesthetic without

muscle relaxation. If a definitive procedure is planned based on

EUS or previous EUA, muscle relaxation may be used. Usually

the lithotomy position is adequate. (For a rectal advancement flap

on an anterior fistula the prone position should be used.) Careful

palpation will usually identify induration along the fistula track.

Pressure over the indurated area might lead to extrusion of pus

through the internal opening. Traction on the external opening

sometimes defines the internal opening.

A probe can be inserted through the external opening (pro-

grade passage).

34

Most often, this probe will easily pass along the

track through the internal opening into the anal canal. A soft

blunt-ended probe should be used. (Lockhart-Mummery fistula

probes should probably be avoided as their tips are sharp.) A

probe passed progradely that follows a course parallel to the anal

canal is either in a secondary track from a transphincteric fistula

or a long intersphincteric fistula. Hydrogen peroxide can be

inserted via the external opening to identify the internal open-

ing.

37,52

Probes with curves of varying degrees can be used to

identify the internal opening. Anal crypts can be mistaken for

internal openings, but these are very shallow. If available, an

EUS performed during EUA can be very helpful in outlining the

track and identifying the internal opening.

TREATMENT OF ANAL FISTULA

The objectives of management have been outlined by Finlay:

53

●

to define the anatomy of the fistula

●

to drain any associated sepsis

●

to eradicate the fistula track

●

to prevent recurrence

●

to preserve continence and sphincter integrity

After full assessment it is the author’s practice to either lay

open the fistula or place a Seton and then re-assess after

2–3 months. At this time the fistula may now be able to be laid

open or a rectal advancement flap can be performed.

When can a fistula be laid open?

The best treatment, in terms of absolute cure, is to lay open the

fistula. It is usually clear that a very high fistula should not be laid

open and that a very low fistula can be laid open without any

functional sequelae. The low and mid trans-sphincteric fistulas

cause the most concern and difficulty in deciding whether to lay

open or to place a Seton. If in doubt, always place a loose Seton,

especially if a full discussion with the patient regarding possible

complications of muscle division has not taken place.

As a general rule of thumb, a fistula can be laid open if it

encompasses half of the external sphincter bulk posteriorly in a

man, one third of the sphincter bulk anteriorly in a man, one third

of the sphincter bulk posteriorly in a woman and never anteriorly

in a woman. However, this decision needs to be individualizedbased on the condition of the sphincter. This will be affected by

previous obstetric trauma, previous perianal operations (e.g.,

lateral sphincterotomy) and congenital abnormalities. It is essen-

tial to enquire regarding the above when taking the history.

‘Lay open’ technique (fistulotomy)

The majority of anal fistulas are successfully treated by this tech-

nique with little in the way of disturbed continence, however, this

risk should be discussed with every patient. The lithotomy posi-

tion is adequate to simply lay open a fistula. A Parks anal or Hill-

Ferguson retractor is used to visualize the anal canal. The track is

defined with a blunt ended probe. The tissue overlying the probe


6/9


is divided along its entire length. The granulation tissue along the

track is curetted. Any overhanging edges of skin can be excised

with diathermy. Haemostasis is achieved with diathermy. Any

extra tracks or undrained collections should also be opened. The

wound is packed loosely with saline soaked gauze and this is

removed when the patient washes the next day. Further packing

is usually painful and mostly unnecessary. Recurrence after this

technique ranges in the literature from 0 to 9% and disturbance in

continence varies from 0 to 33%.

17,21,24,54–59

Seton

The term seton is derived from the Latin word seta

meaning

bristle. Setons may be loose or cutting. Thin silastic tubing (‘ves-

sel loop’) is commonly used. Another option is monofilament

nonabsorbable nylon suture. When a loose seton is placed it is

better to tie the ends of the seton parallel to each other with heavy

silk (Fig. 5). When the seton is tied to itself, the knot is bulky and

unpleasant for the patient. Some surgeons will lay open the lateral

portion of the fistula track so that the seton encompasses only the

sphincter bulk.

60

This may not be necessary, as eventual eradica-

tion of the fistula depends only on elimination of the internalopening. It is the author’s practice to simply place the seton along

the track. At the time of placement, any extra tracks or undrained

cavities can be drained with a de pezzer catheter, which is left in

place for a few days or weeks.

Occasionally, when a seton has been present for 2–3 months,

it may be apparent that the fistula can be safely laid open. It has

been suggested that this is because of fibrosis associated with the

fistula.

61,62

It is more likely that it is simply that any associated

inflammation has settled and a more accurate assessment of the

fistula can be made.

Setons can be used long-term and may be the preferred option

in fistulas associated with Crohn’s

63

disease and those that have

recurred despite many attempts at repair.

Advancement flap

An advancement flap should be considered if a fistula cannot be

laid open. It should only be done when all signs of acute sepsis

have settled.

64

The term ‘mucosal advancement flap’ is a mis-

nomer, as the flap is raised in the intersphincteric plane and

includes both mucosa and IS. It is better described as ‘rectal

advancement flap’ (RAF).

Patients are given a full bowel preparation prior to the proce-

dure. For anterior fistulas the prone jack-knife position is used.

For posterior and lateral fistulas lithotomy is adequate. First

the flap is marked out with diathermy. It commences 5–10 mm

below the internal opening and around 10–15 mm either side of

the internal opening. The intersphincteric plane is infiltrated

with a 1:400 000 solution of adrenaline. Mobilization should

begin laterally away from the track on either side where the

planes are virginal. The intersphincteric plane is identified and

only then is the dissection continued towards the track. The flap

should be raised proximally to the low rectum and should be

under no tension when sutured. The portion of the flap contain-

ing the internal opening is excised and the flap is secured with

sutures (3/0 PDS) encompassing the internal sphincter in the flap

and the internal sphincter of the remaining anal canal. There is

no need to do anything to the external opening; however, a small

drain (infant feeding tube) should be placed through the external

opening to lie underneath the flap and should be removed after

2–3 days.

The success rate for RAF varies in the literature from 29 to

95%.

65–75

A realistic estimate of success rate from more recent

series is 60–70%.

74,75

In a recent series from the Cleveland Clinic,USA the overall primary success rate was 64%. When patients

with Crohn’s disease were excluded, the success rate rose to 77%.

As would be expected, the success rate decreases with each suc-

cessive attempt.

65,75

It is possible that the success rate might be

improved if RAF is combined with application of fibrin glue to

the track and the bed of the flap.

The functional results after RAF appear good. Finan

76

and

Kreis et al

.

77

report no change in anal manometry or rectal

compliance. Most papers report no deterioration in conti-

nence,

65,71–73,76

however, Mizrahi et al

.

78

report a 9% incidence

of deterioration in continence. Schouten et al

.

75

report that of

26 patients with normal continence preoperatively, 38% reported

flatus incontinence and 12% faecal incontinence.

Fibrin glue

Fibrin tissue adhesive has been used as a surgical sealant since

the early 1940s.

79

In the last 10–15 years there have been many

reports of the use of Fibrin glue in anal fistulas.

80–91

The success

rate varies from 14 to 85%. This wide variation is explained by

utilization of different types of glue, different techniques of inser-

tion, different types of fistulas and different lengths of follow-up.

The technique is easy and no division of muscle occurs. The only

potential disadvantage is that if the treatment fails, sepsis might

recur and the track might become more complex.

89

A randomized

controlled trial at Oxford found no advantage for fibrin glue over

fistulotomy for simple fistulas. Fibrin glue healed more complexfistulas than conventional treatment (which included RAF) with

higher patient satisfaction. In this trial 13 of the 16 patients

having conventional treatment of complex fistulas had loose

setons and healing would not be expected with the seton still in

place.

88

All acute sepsis should have settled and the track should be

debrided prior to placement of the glue. It appears to work better

in more complex fistulas and those with longer tracks. It is not

good for short recto-vaginal fistulas. The most recent publica-

tions have reported only moderate success (14%

87

and 33%

91

).

Enthusiasm for fibrin glue treatment of anal fistulas has increased

recently in Australia with the release of an easy to use product,

although anecdotal results have not been encouraging.

Fig. 5.

Loose seton.


7/9

70 RICKARD

Other options

Cutaneous flaps (island flap

92

or V-Y flap

93

), which involve

mobilization of skin from the anal verge and transposition of this

into the anal canal, have been described with good results. This is

surprising, given that biomechanically the forces of defecation

are directed to dislodging the flap rather than compressing it, as is

the case with RAF. This technique should be considered in fistu-

las with a very low internal opening, which cannot be laid openand a RAF would leave mucosa outside the anal canal.

Cutting setons are still used by some, but lead to a keyhole

deformity in the anal canal; they are poorly tolerated because of

pain and are associated with high rates of incontinence.

94,95

Chemical setons utilizing ayurvedic medicated thread (Kshara

sutra) are described in the Indian literature. This is a variation of

the cutting seton and involves a seton soaked in a caustic solution

of three herbs. The method of action is unknown, but alkaline pH

is critical. The herbs possess anti-inflammatory, antibacterial,

and wound-healing properties. In a randomized trial comparing

this method with conventional methods involving 500 patients,

chemical setons obtained better results than conventional methods

(recurrence 4% vs

11%, P

= 0.03).

96

A Martius flap involves mobilization of the fibro-fatty pad in

the labia. This can be used in recto-vaginal fistula to fill the

defect after mobilization of both rectum and vagina on either side

of the fistula.

97

Long-term setons are very well tolerated and often used in

Crohn’s disease and recurrent fistulas.

EXTRA-SPHINCTERIC FISTULAS

The primary pathology in a true extra-sphincteric fistula is intra-

abdominal. Common causes are diverticular disease, enteric

Crohn’s disease, malignancy and radiation damage to the bowel.

An abdominal approach is required.

FISTULAS ASSOCIATED WITHCROHN’S DISEASE

Full discussion of this topic is beyond the scope of the present article.

Usually a long-term seton is the safest option. This reduces perianal

drainage and pain without damaging the sphincter and minimizes the

risk of future abscesses arising from the fistula track.

63

In the pres-

ence of active proctitis, any definitive surgery is very unlikely to be

successful. Without proctitis, definitive surgery (laying open low fis-

tulas, RAF for high fistulas) can be attempted. Success rates will still

be lower than in patients without Crohn’s disease.

98

Infliximab is a murine chimeric monoclonal antibody against

tumour necrosis factor-

α

. It is the only medical treatment proven to

lead to a reduction in the number of perianal fistulas in patientswith Crohn’s disease.

99

However, in practice, recurrence after ces-

sation is common. There is emerging evidence that in patients with

setons in place who are undertaking a course of Infliximab for

enteric Crohn’s disease, it may be worthwhile to simply remove the

seton during the second course of Infliximab treatment.

100

ANORECTAL SEPSIS IN HIV

As a general principle, anorectal sepsis in an HIV positive patient

should be treated along the same principles as anorectal sepsis in

Crohn’s disease. Any tissue that is removed should be submitted

to histological analysis as the prevalence of anal intraepithelial

neoplasia (AIN) is high in this situation.101

IF NOTHING IS WORKING

It might be appropriate to place a loose seton and plan to leave

this in long-term. These are surprisingly well tolerated. Consider

a pelvic cause for the fistula. Is it, in fact, an extra-sphincteric fis-

tula? Perform an abdomino-pelvic computed tomography scan

and a fistulogram. Is it not a fistula, but something very uncom-

mon such as an infected epidermoid cyst?

SUMMARY

Most abscesses simply need to be drained and most fistulas can

be safely laid open. An understanding of the anatomy of the anal

canal and the aetiology of peri-anal abscesses and anal fistulas

is essential. Abscesses require surgical drainage under general

anaesthesia. Intersphincteric and supralevator abscesses may be

difficult to identify clinically. Anal ultrasound or MRI can be

used in these situations depending on local availability and exper-

tise. When an abscess is drained no attempt should be made to

probe the cavity as this can lead to iatrogenic fistulas. If an asso-

ciated fistula is obvious even without probing a seton should be

placed and left until the sepsis and associated inflammation and

swelling has settled.

It is usually clinically evident with very low and very high fistu-

las whether or not laying open is appropriate. Laying open a fistula

is the most successful method of treatment and in most situations

this can be done safely without any disturbance of continence. In

‘mid level’ fistulas radiological assessment (usually with anal

ultrasound) complements clinical examination and an objective

assessment of the extent of sphincter involvement can be made. If

it is not appropriate to lay open the fistula, another form of defini-

tive repair will be required. A rectal advancement flap is a reason-

able approach in this situation. The success rate for this technique is

around 60–70% and functional results are good. If it fails it can be

repeated, albeit with a lower success rate. When multiple attempts

at repair have failed a long-term seton may be the best option.

REFERENCES

1. Bollard RC, Gardiner A, Lindow S, Phillips K, Duthie GS.Normal female anal sphincter: difficulties in interpretationexplained. Dis. Colon Rectum 2002; 45: 171–5.

2. Godlewski G, Prudhomme M. Embryology and anatomy of theanorectum. Basis of Surgery Surg. Clin. North Am. 2000; 80:319–43.

3. Fritsch H, Brenner E, Lienemann A, Ludwikowski B. Analsphincter complex: reinterpreted morphology and its clinical rel-evance. Dis. Colon Rectum 2002; 45: 188–94.

4. Kaiser AM, Ortega AE. Anorectal anatomy. Surg. Clin. North Am. 2002; 82: 1125–38.

5. Wendell-Smith CP. Anorectal nomenclature: fundamental termi-nology. Dis. Colon Rectum 2000; 43: 1349–58.

6. Duthrie HL, Gairns FW. Sensory nerve endings and sensation inthe anal region of man. Br. J. Surg. 1960; 47: 585–95.

7. Thompson-Fawcett MW, Warren BF, Mortensen NJ. A newlook at the anal transitional zone with reference to restorativeproctocolectomy and the columnar cuff. Br. J. Surg. 1998; 85:1517–21.

8. Eisenhammer S. The internal anal sphincter and anorectalabscess. Surg. Gynaecol. Obstet. 1956; 103: 501–6.

9. Parks AG. The pathogenesis and treatment of fistula-in-ano. Br. Med. J. 1961; i: 463–9.

10. Held D, Khubchandani J, Sheets J, Stasik J, Rosen L, Wether R.Management of anorectal horseshoe abscess and fistula. Dis.Colon Rectum 1986; 29: 793–7.


8/9


11. Isbister WH. Fistula-in-ano: personal view. Aust. N. Z. J. Surg.1999; 69: 768–9.

12. Milligan ETC, Morgan CN. Surgical anatomy of the anal canalwith special reference to anorectal fistulae. Lancet 1934; 2.

13. Stelzner F. Die Anorectalen Fisteln. Berlin: Springer, 1959.14. Goligher JC. Surgery of the Anus, Rectum and Colon. London:

Bailliere, Tindall and Cassell, 1961.15. Eisenhammer S. The final evaluation and classification of the

surgical treatment of the primary anorectal cryptoglandular inter-muscular (intersphincteric) fistulous abscess and fistula. Dis.Colon Rectum 1978; 21: 237–54.

16. Parks AG, Gordon PH, Hardcastle JD. A classification of fistula-in-ano. Br. J. Surg. 1976; 63: 1–12.

17. Marks CG, Ritchie JK. Anal fistulas at St Mark’s Hospital. Br. J.Surg. 1977; 64: 84–91.

18. Bennett RC. A review of the results of orthodox treatment foranal fistulae. Proc. Royal Soc. Med. 1962; 55: 756–7.

19. Hughes F, Mehta S. Anorectal sepsis. Hosp. Med. 2002; 63:166–9.

20. Read DR, Abcarian H. A prospective survey of 474 patients withanorectal abscess. Dis. Colon Rectum 1979; 22: 566–8.

21. Vasilevsky CA, Gordon PH. The incidence of recurrentabscesses or fistula-in-ano following anorectal suppuration. Dis.

Colon Rectum 1984; 27: 126–30.22. Winslet MC, Allan A, Ambrose NS. Anorectal sepsis as a pres-

entation of occult rectal and systemic disease. Dis. Colon Rectum1988; 31: 597–600.

23. Oliver I, Lacueva FJ, Perez Vicente F et al. Randomized clinicaltrial comparing simple drainage of anorectal abscess with andwithout fistula track treatment. Int. J. Colorectal Dis. 2003; 18:107–10.

24. Ramanujam PS, Prasad ML, Abcarian H, Tan AB. Perianalabscesses and fistulas. A study of 1023 patients. Dis. Colon Rectum 1984; 27: 593–7.

25. Schouten WR, van Vroonhoven TJ. Treatment of anorectalabscess with and without primary fistulectomy. Results of a pro-spective randomized trial. Dis. Colon Rectum 1991; 34: 60–3.

26. Tang GL, Chew SP, Seow-Choen F. Prospective randomized

trial of drainage alone vs drainage and fistulotomy for acute peri-anal abscesses with proven internal opening. Dis. Colon Rectum1996; 39: 1415–17.

27. Ho YH, Tan M, Chui CH, Leong A, Eu KW, Seow-Choen F.Randomized controlled trial of primary fistulotomy with drain-age alone for perianal abscesses. Dis. Colon Rectum 1997; 40:1435–8.

28. Hebjorn M, Olsen O, Haakansson T, Andersen B. A randomizedtrial of fistulotomy in perianal abscess. Scand. J. Gastroenterol.1987; 22: 174–6.

29. Li D, Yu B. Primary curative incision in the treatment of peri-anorectal abscess. Zhonghua Wai Ke Za Zhi (Chinese J. Surg.)1997; 35: 539–40.

30. Nelson RL. Anorectal abscess fistula: what do we know? Surg.Clin. North Am. 2002; 82: 1139–51.

31. Goodsall DH, Miles WE. Diseases of the Anus and Rectum.London: Longman, 1900.

32. Cirocco WC, Reilly JC. Challenging the predictive accuracy of Goodsall’s rule for anal fistulas. Dis. Colon Rectum 1992; 35:537–42.

33. Gunawardhana PA, Deen KI. Comparison of hydrogen peroxideinstillation with Goodsall’s rule for fistula-in-ano. ANZ J. Surg.2001; 71: 472–4.

34. Keighley MR, Williams NS. Surgery of the Anus, Rectum and Colon, vol. 1. London: W. B. Saunders, 1999.

35. Cho DY. Endosonographic criteria for an internal opening of fistula-in-ano. Dis. Colon Rectum 1999; 42: 515–18.

36. Law PJ, Talbot RW, Bartram CI, Northover JM. Analendosonography in the evaluation of perianal sepsis and fistulain ano. Br. J. Surg. 1989; 76: 752–5.

37. Poen AC, Felt-Bersma RJ, Eijsbouts QA, Cuesta MA, Meuwis-sen SG. Hydrogen peroxide-enhanced transanal ultrasound inthe assessment of fistula-in-ano. Dis. Colon Rectum 1998; 41:1147–52.

38. Ratto C, Gentile E, Merico M et al. How can the assessmentof fistula-in-ano be improved? Dis. Colon Rectum 2000; 43:1375–82.

39. Sudol-Szopinska I, Jakubowski W, Szczepkowski M. Contrast-

enhanced endosonography for the diagnosis of anal and ano-vaginal fistulas. J. Clin. Ultrasound 2002; 30: 145–50.40. Sudol-Szopinska I, Jakubowski W, Szczepkowski M, Sarti D.

Usefulness of hydrogen peroxide enhancement in diagnosis of anal and ano-vaginal fistulas. Eur. Radiol. 2003; 13: 1080–4.

41. Gold DM, Halligan S, Kmiot WA, Bartram CI. Intraobserver andinterobserver agreement in anal endosonography. Br. J. Surg.1999; 86: 371–5.

42. Robinson PJA. Radiology’s Achilles’ heel: error and variation inthe interpretation of the Rontgen image. Br. J. Radiol. 1997; 70:1085–98.

43. Bartram C, Buchanan G. Imaging anal fistula. Radiologic Clin. North Am. 2003; 41: 443–57.

44. Lunniss PJ, Armstrong P, Barker PG, Reznek RH, Phillips RK.Magnetic resonance imaging of anal fistulae. Lancet 1992; 340:

394–6.45. Barker PG, Lunniss PJ, Armstrong P, Reznek RH, Cottam K,

Phillips RK. Magnetic resonance imaging of fistula-in-ano: tech-nique, interpretation and accuracy. Dis. Colon Rectum 1994; 37:288.

46. Madsen SM, Myschetzky PS, Heldmann U, Rasmussen OO,Thomsen HS. Fistula in ano: evaluation with low-field magneticresonance imaging (0.1 T). Scand. J. Gastroenterol. 1999; 34:1253–6.

47. Scholefield JH, Berry DP, Armitage NC, Wastie ML. Magneticresonance imaging in the management of fistula in ano. Int. J.Colorectal. Dis. 1997; 12: 276–9.

48. Zbar AP, de Souza NM, Piuni R, Kmiot WA. Comparison of endoanal magnetic resonance imaging with surgical findings inperirectal sepsis. Br. J. Surg. 1998; 85: 111–14.

49. Beets-Tan RG, Beets GL, van der Hoop AG et al. PreoperativeMR imaging of anal fistulas: does it really help the surgeon? Radiology 2001; 218: 75–84.

50. Kuijpers HC, Schulpen T. Fistulography for fistula-in-ano. Is ituseful? Dis. Colon Rectum 1985; 28: 103–4.

51. Kuijpers HC, Van Tets WF. Fistulography. In: Phillips, RK,Lunniss, PJ, eds. Anal Fistula. London: Chapman & Hall, 1996.

52. Rosen L. Anorectal abscess-fistula. Surg. Clin. North Am. 1998;68: 1293–308.

53. Finlay IG. Objectives in management. In: Phillips, RK,Lunniss, PJ, eds. Anal Fistula. London: Chapman & Hall, 1996.

54. Adams D, Kovalcik PJ. Fistula in ano. Surg. Gynecol. Obstet.1981; 153: 731–2.

55. Sainio P, Husa A. Fistula-in-ano. Clinical features and long-termresults of surgery in 199 adults. Acta Chirurgica Scand. 1985;

151: 169–76.56. Parks AG, Stitz RW. The treatment of high fistula-in-ano. Dis.

Colon Rectum 1976; 19: 487–99.57. Kuijpers HC. Diagnosis and treatment of fistula-in-ano. Neth. J.

Surg. 1982; 34: 147–52.58. Mazier WP. The treatment and care of anal fistulas: a study of

1,000 patients. Dis. Colon Rectum 1971; 14: 134–44.59. McElwain JW, McLean MD, Alexander RM. Anorectal prob-

lems. Experience with primary fistulectomy for anorectalabscess. Report 1000 cases. Dis. Colon Rectum 1975; 18:646–9.

60. Lunniss PJ, Thomson JP. The loose seton. In: Phillips, RK,Lunniss, PJ, eds. Anal Fistula. London: Chapman & Hall, 1996.

61. Gabriel WB. The Principles and Practice of Rectal Surgery.London: Lewis, 1963.


9/9

72 RICKARD

62. Ramanujam PS, Prasad ML, Abcarian H. The role of the setonin fistulotomy of the anus. Surg. Gynaecol. Obstet. 1983; 157:419–22.

63. Faucheron JL, Saint-Marc O, Guibert L, Parc R. Long-termseton drainage for high anal fistulas in Crohn’s disease – asphincter-saving operation? Dis. Colon Rectum 1996; 39:208–11.

64. Stone JM, Goldberg SM. The endorectal advancement flap

procedure. Int. J. Colorectal Dis. 1990; 5: 232–5.65. Ozuna G, Hull TL, Cartmill J, Fazio VF. Long-term analysis of the use of transanal rectal advancement flaps for complicatedanorectal/vaginal fistulas. Dis. Colon Rectum 1996; 39: 10–4.

66. Kodner IJ, Mazor A, Shemesh EI, Fry RD, Fleshman JW,Birnbaum EH. Endorectal advancement flap repair of recto-vaginal and other complicated anorectal fistulas. Surgery 1993;114: 682–9.

67. Jones IT, Fazio VW, Jagelman DG. The use of transanal rectaladvancement flaps in the management of fistulas involving theanorectum. Dis. Colon Rectum 1987; 30: 919–23.

68. Makowiec F, Jehle EC, Becker HD, Starlinger M. Clinicalcourse after transanal advancement flap in patients with Crohn’sdisease. Br. J. Surg. 1995; 82: 603–6.

69. Joo JS, Weiss EG, Nogueras JJ, Wexner SD. Endorectal

advancement flap in perianal Crohn’s disease. Am. Surgeon1998; 64: 147–50.

70. Rothenberger DA, Christenson CE, Balcos EG. Endorectaladvancement flap for treatment of simple rectovaginal fistula. Dis. Colon Rectum 1982; 25: 297–300.

71. Lewis WG, Finan PJ, Holdsworth PJ, Sagar PM, Stephenson BM.Clinical results and manometric studies after rectal flap advance-ment for infra-levator trans-sphincteric fistula-in-ano. Int. J.Colorectal Dis. 1995; 10: 189–92.

72. Miller GV, Finan PJ. Flap advancement and core fistulectomyfor complex rectal fistula. Br. J. Surg. 1998; 85: 108–10.

73. Hyman N. Endoanal advancement flap repair for complexanorectal fistulas: how I do it. Am. J. Surg. 1999; 178: 337–40.

74. Sonoda T, Hull TL, Piedmonte MR, Fazio VW. Outcomes of primary repair of anorectal and rectovaginal fistulas using the

endorectal advancement flap. Dis. Colon Rectum 2002; 45:1622–8.

75. Schouten WR, Zimmermann DDE, Briel JW. Transanaladvancement flap repair of transsphincteric fistulas. Dis. Colon Rectum 1999; 42: 1419–22.

76. Finan PJ. Management by advancement flap technique. In: Phil-lips, RK, Lunniss, PJ, eds. Anal Fistula. London: Chapman &Hall, 1996.

77. Kreis ME, Jehle EC, Ohlemann M, Becker HD, Starlinger MJ.Functional results after transanal rectal advancement flap repairof trans-sphincteric fistula. Br. J. Surg. 1998; 85: 240–2.

78. Mizrahi N, Wexner SD, Zmora O et al. Endorectal advancementflap: are there predictors of failure? Dis. Colon Rectum 2002; 45:1616–21.

79. Cronkite ET, Lozner EL, Deaver JM. Use of thrombin and

fibrinogen in skin grafting. J. Am. Med. Assoc. 1944; 124:976–8.

80. Hjortrup A, Moesgaard FA, Kjaergard J. Fibrin adhesive in thetreatment of perianal fistulas. Dis. Colon Rectum 1991; 34:752–4.

81. Abel ME, Chiu YS, Russell TR, Volpe PA. Autologous fibringlue in the treatment of rectovaginal and complex fistulas. Dis.Colon Rectum 1993; 36: 447–9.

82. Aitola P, Hiltunen KM, Matikainen M. Fibrin glue in perianalfistulas – a pilot study. Ann. Chir. Gynaecol. 1999; 88: 136–8.

83. Cintron J, Park JJ, Orsay CP. Repair of fistulas-in-ano usingfibrin adhesive: Long-term follow-up. Dis. Colon Rectum2000; 43: 944–50.

84. Venkatesh KS, Ramanujam PS. Fibrin glue application in thetreatment of recurrent anorectal fistulas. Dis. Colon Rectum1999; 42: 1136–9.

85. Park JJ, Cintron J, Orsay CP. Repair of chronic anorectal fistulaeusing commercial fibrin sealant. Arch. Surg. 2000; 135: 166–9.86. Patrlj L, Kocman B, Martinac M et al. Fibrin glue-antibiotic

mixture in the treatment of anal fistulae: experience with 69cases. Digestive Surg. 2000; 17: 77–80.

87. Buchanan GN, Bartram CI, Phillips RK et al. Efficacy of fibrinsealant in the management of complex anal fistula: a prospec-tive trial. Dis. Colon Rectum 2003; 46: 1167–74.

88. Lindsey I, Smilgin-Humphreys MM, Cunningham C, MortensenNJ, George BD. A randomized, controlled trial of fibrin glue vsconventional treatment for anal fistula. Dis. Colon Rectum2002; 45: 1608–15.

89. Sentovich SM. Fibrin glue for anal fistulas: long-term results. Dis. Colon Rectum 2003; 46: 498–502.

90. Sentovich SM. Fibrin glue for all anal fistulas. J. Gastrointesti-

nal Surg. 2001; 5: 158–61.91. Zmora O, Mizrahi N, Rotholtz N et al. Fibrin glue sealing in the

treatment of perineal fistulas. Dis. Colon Rectum 2003; 46:584–9.

92. Del Pino A, Nelson RL, Pearl RK, Abcarian H. Island flap ano-plasty for treatment of transsphincteric fistula-in-ano. DisColon Rectum 1996; 39: 224–6.

93. Amin SN, Tierney GM, Lund JN, Armitage NC. V-Y advance-ment flap for treatment of fistula-in-ano. Dis. Colon Rectum2003; 46: 540–3.

94. Christensen A, Nilas L, Christiansen J. Treatment of trans-sphincteric anal fistulas by the seton technique. Dis. Colon Rectum 1986; 29: 454–5.

95. McCourtney JS, Finlay IG. Setons in the management of fistula-in-ano. Br. J. Surg. 1995; 82: 448–52.

96. Shukla NK, Narang R, Nair NGK. Multicentric randomizedcontrolled clinical trial of Kshaarasootra (Ayurvedic medicatedthread) in the management of fistula-in-ano. Indian J. Med. Res. 1991; 94: 177–85.

97. Elkins TE, DeLancey JO, McGuire EJ. The use of modifiedMartius graft as an adjunctive technique in vesicovaginal andrectovaginal fistula repair. Obstetrics Gynecol. 1990; 75:727–33.

98. Schwartz DA, Pemberton JH, Sandborn WJ. Diagnosis andtreatment of perianal fistulas in Crohn disease. Ann. Intern. Med. 2001; 135: 906–18.

99. Present DH, Rutgeerts P, Targan S, Hanauer SB, Mayer L, vanHogezand RA. Infliximab for the treatment of patients withCrohn’s disease. N. Engl. J. Med. 1999; 340: 1398–405.

100. Topstad DR, Panaccione R, Heine JA, Johnson DRE, MacLean

AR, Buie WD. Combined seton placement, infliximab infusion,and maintenance immunosuppresives improve healing rate infistulizing anorectal Crohn’s disease: a single centre experi-ence. Dis. Colon Rectum 2003; 46: 577–83.

101. Zbar AP, Fenger C, Efron J, Beer-Gabel M, Wexner SD. Thepathology and molecular biology of anal intraepithelial neo-plasia: comparisons with cervical and vulvar intrepithelialcarcinoma. Int. J. Colorectal Dis. 2002; 17: 203–15.

Anal Fistulas & Surgery

Documents

Transcript of Anal Fistulas & Surgery