Revisiting an ancient treatment for transphincteric fistula-in-ano ‘There is nothing new under the sun’ Ecclesiastes 1v9

Abstract

Objective

The history of treatments for fistula-in-ano can be traced back to ancient times. Current treatment of transphincteric fistulae is controversial, with many options available. We reviewed the history of treatment using cutting setons and present our series of transphincteric fistulae in the light of the series in the literature.

Design

Literature review and case series.

Setting

Hospital based coloproctology service

Participants

140 consecutive patients presenting with fistula-in-ano were included.

Main Outcome Measures

The literature pertaining to treatment of transphincteric fistula was reviewed, along with the outcome of various treatment methods for this condition. Data were collected for 140 consecutive patients presenting with fistula- in-ano were assessed for fistula healing, recurrence and complications.

Results

A total of 140 consecutive patients with fistula-in-ano were identified, of which 111 were cryptoglandular (79.3%). Eighty-one of these 111 were transphincteric (73.0%). At a median follow-up of 35 months (range, 2–83 months), 70 transphincteric fistulae had healed (86.4%), 10 were still undergoing treatment (12.3%) and one patient was lost to follow-up prior to treatment (1.2%). Two patients in this group required a stoma (2.5%), six patients developed recurrence (7.4%); three ‘true’ recurrences (3.7%). One (1.2%) developed a chronic fissure. There were no reported cases of incontinence.

Conclusions

The management of transphincteric fistula-in-ano is complex and controversial, for which no clear surgical procedure has gained acceptance as the gold standard. This study demonstrates that transphincteric fistulae can be successfully treated using cutting setons. A high healing rate (86.4%), low recurrence rate (7.4%) and a low complication rate (3.7%) are shown, which compares favourably with published rates over a long follow-up.

Keywords

surgery history gastroenterology pathology

Introduction

Fistula-in-ano is a challenging pathology where treatment attempts have been traced back to the days of Hippocrates (Figure 1).¹ The pathophysiology is still not completely understood and can be complicated further by a variety of other underlying pathological conditions, or through the surgeon embarking on an inappropriate course of treatment.

Figure 1.

Hippocrates (460–370 BC). Image credit: Public domain.

Today, the most widely accepted classification system of fistulae is based on the work done by Sir Alan Parks leading to the accepted modified Parks classification (Figure 2)² in defining fistula anatomy. In the absence of an underlying condition such as Crohn’s disease or malignancy, most ‘idiopathic’ fistulae are thought to arise from sepsis in the intersphincteric glands as hypothesised by Parks,³ and are commonly described as ‘cryptoglandular’ in origin.

Figure 2.

Diagram of Parks classification of fistula-in-ano image taken from Parks 1976.² Type A Intersphincteric, Type B Transphincteric, Type C Suprasphincteric, Type D Extrasphincteric.

Transphincteric fistulae, which cross the external sphincter, are more challenging to treat. This is due to the need to balance eradication of the fistula with continence preservation.

History documents a wide range of treatments that have been undertaken over the years in an attempt to treat this problem, with no one particular treatment accepted as a gold standard.⁴ The documented history of fistula management with cutting setons initially started with Hippocrates,¹ who inserted horsehair with lint into the fistula and, which was periodically tightened.

The Middle Eastern physician Albucasais (936–1013 AD) (Figure 3)⁵ as well as the medieval physician John of Ardene (1307–1392 AD) (Figure 4) practised a variety of methods to treat fistulae. Patience was the cornerstone of their treatment; however, they found that their patients often wanted a quick fix to their ailments. The Treaties of Fistulae texts written by Ardene describe seton use for complex fistulae.⁵ Today the material may be different⁶ but the principle is the same.

Figure 3.

‘Abu al-Qasim Khalaf ibn al-Abbas Al-i‘: Known in the West as ‘Albucasis’ (936–1013 AD). Image credit: Peter Paul Rubens/Public Domain.

Figure 4.

John of Ardene (1307–1392 AD). Image credit: Treatises of fistula in ano, haemorrhoids and clysters 5/public domain.

In 1835, Fredrick Salmon (Figure 5) opened ‘The Infirmary for the Relief of the Poor Afflicted with Fistula and Other Diseases of the Rectum’ following the successful treatment of Charles Dickens. The fistula infirmary became popular and by 1854 the infirmary site was relocated and re-named ‘St Mark’s Hospital for Fistula and other diseases of the Rectum’.⁷

Figure 5.

Fredrick Salmon (1796–1868 AD). Image credit: George Brown/Public domain.

Sir Lockhart-Mummery summarised the difficulties that surgeons encountered in treating anal fistulae when he said ‘Probably, more surgical reputations have been damaged by the unsuccessful treatment of fistula than by excision of the rectum or gastroenterostomy’ and later went on to state ‘The bad results of laparotomy are generally buried with flowers, while fistulae go about the world exhibiting the unsuccessful results of the treatment’.^8,9

More recently, options for treating transphincteric fistulae have included anal advancement flaps,¹⁰ fistula plugs,^11,12 fibrin glue and collagen glue.¹³ More recently, the Ligation of Intersphincteric Fistula Tract procedure,^14,15 and stem cell injections¹⁶ have been introduced.

Proponents of the advancement flap have argued that it produces superior results, resulting in fewer procedures required to treat the condition.^17,18 However, it has also been counter argued that it demands more complicated skills and is associated with greater pain, bleeding and potentially more severe complications for the patient.¹⁹

Fistula plugs have been introduced and a UK-based randomised controlled trial FIAT-500 is currently underway comparing the SURGIS® AFP™ anal fistula plugs²⁰ to alternative fistula treatment of the surgeons’ choice. Current data do not seem to demonstrate any higher efficacy of fistula plug to that of conventional treatment options.^20–24

The Ligation of Intersphincteric Fistula Tractprocedure²⁵ and Modified Ligation of Intersphincteric Fistula Tract²⁶ Procedure have produced promising results from early studies and will need more long-term evaluation.

The desired outcomes following surgery to treat fistula-in-ano are threefold; the first aim is to get rid of the fistula, second to preserve sphincter function and the third is to prevent recurrence. Published results vary with respect to success of treatment, recurrence, technique used, patient cohort and length of follow-up.^{6,12–15,17–19} Prognosis is also affected by the complexity of the fistula and any secondary underlying disease that can affect fistula healing. However, treatment is considered successful and complete when the patient is free from fistula, has maintained continence and the fistula does not recur.

Methods

Patients presenting with fistula-in-ano were studied over an 11-year period within two health boards in South East Wales. Data were collected both retrospectively and prospectively during the study period.

Demographic data, fistula anatomy, investigations, treatment course, complications and subsequent follow-up were recorded. Cryptoglandular transphincteric fistulae were included and non-cryptoglandular and non-transphincteric fistulae were excluded from further analysis.

Outcome measures assessed healing, recurrence and complications. All patients were asked about continence to flatus and faeces and to report any soiling when attending clinic and at each subsequent day case visit. A subset of patients was formally continence scored using the Cleveland Clinic Faecal Incontinence Severity Index.²⁷

All procedures were carried out under general anaesthetic as day case Examination Under Anaesthetics (EUAs), where number 1 nylon cutting seton was hand tied in place. The anoderm from the external opening to the internal opening or dentate line, whichever was lower, was removed along with a Salmon back cut when necessary. This reduced the pain associated with tight seton placement across the sensate epithelium and allowed improved drainage at the external opening, respectively. The suture was subsequently changed and tightened, if necessary, on a six- to eight-week basis, following a re-examination under anaesthetic.

When final EUA demonstrated no residual fistula or sepsis following extrusion or removal of the seton, patients were followed up in clinic at approximately eight weeks to confirm healing.

Results

The series comprised 140 patients presenting with fistula-in-ano, of these 84 were male (60%) and 56 were female (40%). Men presented at a mean age of 44.3 years (range, 17–79 years) and women presented at a mean age of 41.9 years (range, 17–71years). The majority, 111 (79.3%), presented with cryptoglandular fistulae-in-ano. Twenty-nine patients were excluded from further analysis (20.7%); of these 21 (15.0%) were excluded due to a significant underlying pathology, e.g. Crohn’s disease and eight (5.7%) were excluded due to the patient records being unobtainable.

As shown in Table 1, 81 of the 111 patients in the cryptoglandular fistula group (73.0%) were identified as having transphincteric fistulae, 14 patients (12.6%) had intersphincteric fistulae, 15 patients (13.5%) had superficial fistulae, and one patient (0.9%) had an extrasphincteric fistula. The extrasphincteric, intersphincteric and superficial fistulae were excluded from further analysis.

Table 1.

Number of fistulae, aetiology, mean age of presentation and classification of fistulae in series.

Results	Male n(range)	Female n (range)	Total
Demographics
Number (ratio)	84	56	140 (3:2)
Age (years): mean (range)	44.8 (17–79) SD 13.8	41.9 (17–71) SD 12.4	43.4 (17–79)
Aetiology of fistula
Cryptoglandular	73	38	111
Crohn’s	4	10	14
Anal cancer	0	3	3
Fournier’s gangrene	1	0	1
Iatrogenic	0	3	3
Unknown (notes not obtainable)	6	2	8
Type of cryptoglandular fistula
Superficial	9	6	15
Intersphincteric	12	2	14
Transphincteric	52	29	81
Extrasphincteric	0	1	1

Of the 81 patients with transphincteric fistulae, 29 (35.8%) were women and 52 (64.2%) were men. Seven patients had two fistulae, of which three had two transphincteric fistulae. The remaining four had one transphincteric fistula plus one of either a superficial or intersphincteric fistula.

Magnetic resonance imaging was done in 30 patients (37.0%) during the management of their transphincteric fistula. No patient with cryptoglandular transphincteric fistula underwent endoanal ultrasound.

At the time of writing, 70 of the 81 patients (86.4%) had healed, 10 patients (12.3%) had a seton in situ and were still undergoing treatment and one patient (1.2%) was lost to follow-up prior to receiving treatment (Figure 6). The median number of procedures was four (range, 1–17). Median follow-up time was 35 months (range, 2–83 months) after the fistula had healed.

Figure 6.

Number of fistulae healed, number of patients with a seton in situ and number lost to follow-up.

Six patients (7.4%) developed recurrence; three ‘true’ recurrences (3.7%) and three ‘de novo’ fistulae (3.7%). Two (2.5%) patients required a stoma in an attempt to treat their sepsis; one had multiple co-morbidities including poorly controlled diabetes and a high body mass index and the other suffered with severe alcohol dependency. Both of these presented with large transphincteric fistulae and both were poorly compliant with scheduled procedure appointments. One patient (1.2%) developed a chronic fissure post treatment, as a result of a non-healing wound at the 6 o’clock site of the original fistula.

No patient reported a worsening of continence at any clinic or day surgery visit. Eleven patients (13.8%) completed the Cleveland faecal incontinence severity index.²⁷ Of these, one patient reported an improvement in their continence from preoperative Cleveland Incontinence Score of 3 to Cleveland Incontinence Score 0. One patient reported occasional incontinence of flatus (Cleveland Incontinence Score 1) both pre- and post-procedures. No patient has re-presented with incontinence at hospitals in either of the two South East Wales health boards.

Six (7.4%) patients required one or more of their EUAs at an interval of less than six weeks. This was due to either dislodgement of the previous seton or due to anal pain.

Discussion

In this study the age and sex distribution of patients is similar to that of other studies²⁸ (Table 2). However, there is a preponderance of transphincteric fistulae in this series compared to other series.² This could be attributed to a significant proportion of tertiary referrals comprising the case series. The total number of transphincteric fistulae in this series (81) is greater than many published series^{11,15,19,23,24} (Table 2).

Table 2.

Comparison of series against recent series treating fistulae through a variety of treatment modalities comparing healing rate, recurrence rate and follow-up time.

Author	Intervention	N	N = Healed (%)	N = Recurrence (%)	Median/mean follow-up (range)
Chung²³	(1) Fistula plug (2) Fibrin glue (3) Advancement flap (4) Seton	(1) 27 (2) 23 (3) 96 (4) 86	(1) 16 (59.3) (2) 9 (39.1) (3) 58 (60.4) (4) 28 (32.6)	Not stated	12 weeks (mean)
Ortiz⁴²	(1) Fistula plug (2) Advancement flap	(1) 15 (2) 16	Not stated	(1) 12 (80) (2) 2 (12.5)	12 months (mean)
Christoforidis et al.¹¹	(1) Advancement flap (2) Fistula plug	(1) 43 (2) 37	(1) 27 (63) (2) 12 (32)	(1) 3 (19) (2) 5 (20)	(1) 56 months (mean) (range, 6–136) (2) 14 months (mean) (range, 6–22)
Johnson²⁴	(1) Fibrin glue (2) Fistula plug	(1) 10 (2) 15	(1) 4 (40) (2) 13 (86)	(1) 6 (60) (2) 2 (13)	(1) 13.6 weeks (mean) (2) 13.8 weeks (mean)
Shanwani¹⁵	LIFT technique	45	37 (82.2)	8 (17.7)	9 months (median) (range, 2–16)
Theerapol¹⁹	Cutting seton	47	37 (78)	1 (2.1)	15 weeks (median)
This data series	Cutting seton	81	70 (86.4)	6 (7.4)	35 months (median) (range, 2–83)

In line with all these other series, the height of the fistula has not been presented as a separate entity in this report; however, the fact that a seton was inserted into the transphincteric fistula of 81 patients indicates that at the time of first EUA the operating surgeon did not have confidence that any of them were low enough to lay open without risk of incontinence.

Although treatment by cutting seton is a very old method of fistula management and in many centres has been superseded by more modern techniques, this case series suggests that transphincteric anal fistulae can be treated successfully by cutting setons alone. This causes minimal disruption to the pelvic anatomy and preserves continence. The technique fulfils the objectives in anal fistulae management; it defines the anatomy, drains sepsis, eradicates the fistulous track and preserves continence and sphincter integrity.²⁹ When compared to literature data^30,31 this study compares favourably in terms of healing (86.4%) to rates quoted for cutting setons as well as other techniques used in management (Table 2).

A significant duration of follow-up was achieved with a median 35 months (range, 2–83). This is an important aspect in assessing long-term outcome and success of treatment in terms of fistula recurrence and patients re-presenting to the hospital suffering from complications from fistula procedures. Many studies are noted for their relatively short follow-up time following treatment of fistulae and are unable to assess long-term outcome post-surgery. One study was identified as having comparable follow-up times when looking both at advancement flaps and fistula plugs¹¹ (Table 2).

Several papers suggest that seton use causes a functional defect leading patients to develop changes to continence, particularly to flatus,³² with some quoted figures greater than 60%.³³ It has also been theorised that defects in the internal sphincter contribute significantly to incontinence to either faeces or flatus.^34,35 There is an argument that division of the internal sphincter and eradication of sepsis in the intersphincteric space are important in the treatment of perianal sepsis.³⁶ However, internal sphincterotomy is not without its own risks, and avoidance of primary internal sphincterotomy may prevent deterioration of continence.^37,38 Therefore, for this reason the internal sphincter was not routinely divided, but if a few lower fibres of internal sphincter were still present inside the cutting seton at the time of final EUA, then they were divided with the final skin bridge to remove the seton. We were unable to formally continence score all patients with the Cleveland Incontinence Scoring System, but were able to follow up all patients in clinic and record continence issues in the clinical notes. No patient in this series reported a worsening of continence. This care around sphincter preservation may have contributed to a greater number of EUAs being required, in order to avoid the risk of incontinence.

The large number of EUAs for some patients also reflects the referral practice in this series. Several patients were referred after multiple EUAs, abscess drainages and other attempted treatments in other hospitals. In a few patients the seton fell out unintentionally and thus caused a slight delay in their treatment progression.

The choice of material for the cutting seton was influenced by several considerations. Training at St Marks Hospital where 1-nylon was regularly used provided the initial impetus. Many other materials have also been described. Steel wire³⁹ is not available in these health boards. Polyfilament, e.g. silk,⁴⁰ carries a potentially higher infection risk. Rubber, sialastic sloops or bands⁴¹ have been used by the authors, but mainly as loose setons in acute sepsis or in semi-permanent setons in complex Crohn’s/radiotherapy fistulae. The latter materials have more ‘elasticity’ and so a constant ‘cutting’ action is less predictable than with 1-nylon.

One downside of the use of cutting setons is the cost of repeated EUAs. Although change of seton can be done as an outpatient (office) procedure, many patients find this uncomfortable and prefer general anaesthetic (particularly in a free at the point of use service such as the National Health Service). Commonly used alternative strategies include the fistula plug, which currently have a list price of approximately £580. However, fistula plug use has declined over the last two to three years as surgeons see an increasing number of treatment failures and recurrences. The currently favoured option, the Ligation of Intersphincteric Fistula Tract procedure, also has the potential advantage of a single procedure, without the additional cost of a plug. However, follow-up is only short-term and its long-term success is yet to be determined.^14,15

The ideal treatment for transphincteric fistulae would be for a one-off intervention without the attendant costs of repeated procedures. Although some cases series in the literature do contain costings for individual procedures such as the fistula plug, it is difficult to do a formal health economic assessment due to the relatively short follow-up. Although we have not performed a formal health economic assessment ourselves, it is important to consider the relatively low recurrence rate at a median of 35 months.

Conclusions

The ideal surgery for transphincteric fistulae-in-ano management is still not known. Newer advances in technologies have added to the debate and allowed surgeons to explore if indeed, a holy grail to fistula-in-ano management can indeed ever be found. It is a challenging condition, which is likely to continue to plague us.

This relatively large series demonstrates that transphincteric fistulae can be successfully treated by the ancient method of cutting setons. It shows a high healing rate (86.4%), a low recurrence rate (7.4%) and a low complication rate (3.7%) and compares favourably with more modern techniques in the literature (Table 2).

Footnotes

Declarations

References

Francis Adams. The Genuine Works of Hippocrates. London: Kessinger Publishing, LLC. [Printed for the Sydenham Society], 1849: 811–822.

Parks

Gordon

Hardcastle

. A classification of fistula-in-ano. Br J Surg 1976; 63: 1–12.

Parks

. The pathogenesis and treatment of fistula-in-ano. Br Med J 1961; 1: 463–469.

Fucini

Giani

. Why do we have to review our experience in managing cases with idiopathic fistula in ano regularly? World J Gastroenterol 2011; 17: 3297–3299.

Arderne J. In: Power D (ed) Treatises of fistula in ano haemorrhoids and clysters. London: Trubner & co and Oxford University Press, 1910. https://archive.org/details/treatisesoffistu00ardeuoft.

McCourtney

Finlay

. Setons in the surgical management of fistula in ano. Br J Surg 1995; 82: 448–452.

Dukes

. Frederick Salmon: Founder of St Mark’s Hospital, London. J Med Hist 1959; 3: 312–316.

Lockhart-Mummery

. Discussion on fistula-in-ano. Proc R Soc Med 1929; 22: 1331–1358.

Lockhart-Mummery

. Fistula in ano. Lancet 1936; 227, 5873: 657–660.

10.

Rizzo

Naig

Johnson

. Anorectal abscess and fistula in ano: Evidence based management. Surg Clin North Am 2010; 90: 45–68.

11.

Christoforidis

Etzioni

Goldberg

Madoff

Mellgren

. Treatment of complex anal fistulas with the collagen fistula plug. Dis Colon Rec 2008; 51: 1482–1487.

12.

Ratto

Litta

Parello

Donisi

Zaccone

De Simone

. Gore bio-A fistula plug. A new sphincter sparing procedure for complex anal fistula. Colorectal Dis 2012; 14: 264–269.

13.

Swinscoe

Ventakasubramaniam

Jayne

. Fibrin glue for fistula-in-ano: The evidence reviewed. Tech Coloproctol 2005; 9: 89–94.

14.

Rojanasakul

Pattanaarun

Sahakitrungruang

Tantiphlachiva

. Total anal sphincter saving technique for fistula in ano; the ligation of intersphincteric fistula tract. J Med Assoc Thai 2007; 90: 581–586.

15.

Shanwani

Nor

Amri

. Ligation of the intersphincteric fistula tract (LIFT): A sphincter saving technique for fistula in ano. Dis Colon Rec 2010; 53: 39–42.

16.

Garcia-Olmo

Herreros

Pascual

Del-Valle

Zorrilla

. Expanded adipose-derived stem cells for the treatment of complex perianal fistula: A phase II clinical trial. Dis Colon Rectum 2009; 52: 79–86.

17.

Mitalas

Gosselink

Zimmerman

DDE

Schouten

. Repeat transanal advancement flap repair: Impact on the overall healing rate of high transphincteric fistulas and on faecal continence. Dis Colon Rectum 2007; 50: 1508–1511.

18.

Zimmerman

Briel

Schouten

. Endoanal advancement flap repair for complex anorectal fistulas. Am J Surg 2001; 181: 576–577.

19.

Theerapol

BYJ

Ngoi

. Routine use of setons for the treatment of anal fistulae. Singapore Med J 2002; 43: 305–307.

20.

The Surgisis® AFP™ anal fistula plug: A report of a consensus conference. The association of coloproctology of Great Britain and Ireland. Colorectal Dis 2008; 10: 17–20.

21.

McGee

Champagne

Stulberg

. Tract length predicts successful closure with anal fistula plug in cryptoglandular fistulas. Dis Colon Rectum 2010; 53: 1116–1120.

22.

Dudkgian

Abcarian

. Why do we have so much trouble treating anal fistula? World J Gastroenterol 2011; 17: 3292–3296.

23.

Chung

Kazemi

Sun

Brown

MCJ

Phang

. Anal fistula plug and fibrin glue versus conventional treatment in repair of complex anal fistulas. Am J Surg 2009; 197: 604–608.

24.

Johnson

Gaw

Armstrong

. Efficacy of anal fistula plug vs. fibrin glue in closure or anorectal fistulas. Dis Colon Rectum 2006; 49: 371–376.

25.

Sangar

Rojanasakul

. Fistula in ano: How I do the LIFT procedure. Sri Lanka J Surg 2011; 29: 97–99.

26.

Onkelen

Gosselink

Schouten

. Is it possible to improve the outcome of transanal advancement flap repair for high transphincteric fistulas by additional ligation of the intersphincteric fistula tract? Dis Colon Rectum 2012; 55: 163–166.

27.

Rockwood

Church

Fleshman

. Patient and surgeon ranking of the severity of symptoms associated with fecal incontinence; the fecal incontinence severity index. Dis Colon Rectum 1999; 42: 1525–1531.

28.

Seow-Choen

Nicholls

. Anal fistula. Br Surg 1992; 79: 197–205.

29.

Finlay

Objectives in management. In: Phillips

RKS

Lunniss

(eds). Anal Fistula, London: Chapman & Hall, 1996, pp. 69–72.

30.

Hyman

. Endoanal advancement flap repaid for complex anorectal fistulas. Am J Surg 1999; 178: 337–340.

31.

Kodner

Mazor

Shemesh

Fry

Fleshman

Birnbaum

. Endorectal advancement flap repair of rectovaginal and other complicated anorectal fistulas. Surgery 1993; 114: 682–690.

32.

Lunniss

Kamm

Phillips

RKS

. Factors affecting continence after surgery for anal fistula. Br J Surg 1994; 81: 1382–1385.

33.

Garcia-Aguilar

Belmonte

Wong

Goldberg

Madoff

. Cutting seton versus two stage fistulotomy in the surgical management of high anal fistula. Br J Surg 1998; 85: 243–245.

34.

Monson

JRT

. Should we care about the internal anal sphincter? Dis Colon Rectum 2012; 55: 105–108.

35.

Zbar

Jayne

Mathur

Ambrose

Guillou

. The importance of the internal anal sphincter (IAS) to maintaining continence: anatomical, physiological and pharmacological considerations. Colorectal Dis 2000; 2: 193–202.

36.

Williams

Macleod

Rothenberger

Goldberg

. Seton treatment of high anal fistulae. Br J Surg 1991; 78: 1159–1161.

37.

McCourtney

Finlay

. Cutting seton without preliminary internal sphincterotomy in management of complex high fistula in ano. Dis Colon Rectum 1996; 39: 55–58.

38.

Zbar

Ramesh

Beer-Gabel

. Conventional cutting vs. internal anal sphincter-preserving seton for high transphincteric fistula: A prospective randomized manometric and clinical trial. Tech Coloproctol 2003; 7: 89–94.

39.

Misra

Kapur

. A new non-operative approach to fistula in ano. Br J Surg 1988; 75: 1093–1094.

40.

Durgan

Perek

Kapan

. Partial fistulotomy and modified cutting seton procedure in the treatment of high extrasphincteric perianal fistulae. Dig Surg 2002; 19: 56–58.

41.

Joy

Williams

. The outcome of surgery for complex anal fistulas. Colorectal Dis 2002; 4: 254–261.

42.

Ortiz

Marzo

Ciga

. Randomized clinical trial of anal fistula plug versus endorectal advancement flap for the treatment of high cryptoglandular fistula in ano. Br J Surg 2009; 96: 608–612.