Anal Fistula

Second half of ASCRS Ch13. The chronic sequela of Anorectal Abscess in a substantial minority of patients — by definition, 100% of cryptoglandular anorectal abscesses have a path from the dentate line to the drained cavity, but this only matures into a true fistula tract in ~30–35% of cases. Anorectal abscess persists as a fistula in ~30–35% of patients overall, more often in nondiabetics and those <40, with no significant difference by sex, smoking status, HIV status, or perioperative antibiotic use.

Presentation

Wide range of complaints: patients often self-describe symptoms as "hemorrhoids," history of spontaneous or surgical abscess drainage, an external bump that becomes irritated and bleeds, chronic drainage, and cyclical perianal pain/swelling relieved by expression of fluid. A careful physical exam is central to diagnosis, decision-making, and patient counseling given this symptom diversity.

Classification (Parks classification)

Fistula type is determined by the course of the offending gland/path of least resistance through the sphincter complex, and is categorized by degree of sphincter involvement:

• Subcutaneous/submucosal (2–3%) — begin at the dentate line, course just deep to the anoderm, no sphincter involvement.
• Intersphincteric (24–45%) — track through the internal sphincter alone before traveling through the intersphincteric space to reach the perianal skin.
• Transsphincteric (30–60%) — travel through both internal and external sphincter; subdivided by degree of external sphincter involvement into low (≤30% of external sphincter) and high (>30%) — a distinction that matters directly for the sphincter-dividing-vs.-sphincter-sparing treatment decision and incontinence risk (see Treatment below).
• Suprasphincteric (2–20%) — begin at the dentate line, course through the internal sphincter, travel cephalad in the intersphincteric space, and cross the skeletal muscle above the external sphincter to enter the ischiorectal space.
• Extrasphincteric — a fifth class, distinct from the other four in that it does not arise from cryptoglandular origin and does not involve the sphincter complex. Arises from an intraabdominal source (e.g. diverticulitis or malignancy), a separate fistulizing-disease etiology (Crohn's disease — see Crohn Disease), or iatrogenic injury/inappropriate drainage of a supralevator abscess. Mentioned here to complete the fistula-in-ano classification; management is addressed elsewhere (this chapter's cryptoglandular-fistula treatment ladder below does not apply).

Goodsall's rule: an external opening posterior to the transverse anal line will curve medially to an internal opening in the posterior midline; an external opening anterior to the transverse line corresponds to a radially located internal opening (exceptions: posteriorly-arising tracts that curve anteriorly before reaching the perianal skin). This principle has come into question more recently but remains a decent guideline — it correctly predicts actual disease in ~80% of cryptoglandular fistulas overall, most accurately for posterior (91%) and intersphincteric (93%) fistulas, less so for anterior (69%) and transsphincteric (68%) fistulas.

Preoperative imaging

Not needed for the majority of uncomplicated cryptoglandular fistulas diagnosed by symptoms/exam alone. Reasonable to consider for non-cryptoglandular disease (e.g. Crohn's), recurrent disease or prior fistula surgery, or when the tract isn't readily identifiable on exam under anesthesia. Modality choice depends heavily on surgeon preference/access/skill:

• Fistulography (water-soluble contrast injected into the external opening, plain X-ray/ fluoroscopy) — largely surpassed by other modalities due to radiologist-expertise dependence, discomfort, and inferior accuracy vs. MRI (73.1% vs. 92.7% fistula classification accuracy, P<0.001).
• CT — not typically helpful for fistula characterization specifically, though useful for complex abscess cavities or aberrant anatomy from prior surgery/infection/IBD; CT fistulography has reasonable tract-characterization accuracy with no instrumentation-related interobserver variability, but worse accuracy than MRI.
• MRI — the preferred/gold-standard modality: accurate, reproducible, no instrumentation of the tract needed (better patient tolerance), and localizes/characterizes both fistulas and abscesses with surrounding anatomy. Endoanal-coil MRI initially showed improved accuracy over an external coil, but with modern 1.5T/3.0T body-coil MRI, body coil findings now surpass endoanal coil in concordance with surgical findings. Notably, in one 136-patient study, 83.1% preop MRI/ operative concordance was found, but the finding only changed management in 33.8% of patients — more often in complex fistulas, external openings >2cm from the verge, or horseshoe fistulas — underscoring that when to image matters as much as which modality.
• Endoanal ultrasound (EAUS) — a reasonable alternative to MRI where cost/access are limiting; similar internal-opening accuracy to MRI (80–90%), less accurate for secondary extensions (67–80% vs. 90%). No significant EAUS-vs-MRI difference for simple fistula tracts specifically — most would argue neither imaging modality is indicated for simple tracts at all. Diluted hydrogen peroxide injected via the external opening is often used as ultrasonic contrast to improve visibility, internal-opening identification, classification accuracy, and supralevator extension/abscess detection. Highly operator-dependent — accuracy figures above reflect experienced-user performance.

Treatment strategies overview

Three main goals: (1) control of sepsis, (2) definitive repair without recurrent disease, (3) maintenance of continence — generally addressed in that priority order. First step: rectal exam under anesthesia to evaluate the fistula, surrounding anatomy, degree of ongoing infection, and classify sphincter involvement. There is little-to-no role for medical management alone in cryptoglandular fistula-in-ano (exceptions: minimally symptomatic patients with comorbidities precluding surgery, managed with sepsis control via draining seton as needed).

Intraoperative fistula identification

A blunt-tipped fistula probe (single/double-armed, straight/curved, malleable/nonmalleable) is passed along the tract to the internal opening, guided by opposite-hand palpation, moving smoothly without force to avoid creating a false passage. A common pitfall: the probe passes through the sphincter complex but not through the final mucosal layer — "popping through" this layer is tempting but inaccurate identification here increases recurrence risk and should be avoided. Intraoperative hydrogen peroxide or methylene blue injection via the external opening aids identification of a difficult-to-find internal opening (83% accurate in one study); intraoperative ultrasound ± peroxide enhancement is an alternative. If the internal opening cannot be identified, it is better to abandon the procedure (preserving a future opportunity for identification without injury to the anal canal/sphincter) than to risk iatrogenic injury.

Once identified, fistulas are classified as simple or complex based on incontinence risk with a sphincter-dividing operation. Per the ASCRS Standard Practice Task Force, complex features include:

30% external sphincter involvement (high transsphincteric, suprasphincteric, extrasphincteric), anterior location in a female, multiple tracts, recurrent fistula, preexisting incontinence, local irradiation, and Crohn's disease. Simple fistulas may be treated with definitive sphincter-dividing surgery at initial presentation; all others should undergo a sphincter-preserving technique, typically beginning with a draining seton.

Fistulotomy

Laying open the fistula tract (including secondary tracts) for complete/adequate drainage — the only recommended surgical option for simple fistulas. Recurrence 0–9%, incontinence 0–37% (wide range reflects variable simple-vs-complex inclusion criteria across studies) — overall safe and low recurrence/incontinence risk in appropriately selected (simple) fistulas. Marsupialization of wound edges after fistulotomy decreases resultant wound size, shortens healing time (6 vs. 10 weeks, P<0.001), and reduces postop bleeding (36% vs. 46%, P<0.05).

Concomitant fistulectomy (removing the fibrotic tract tissue, theorized to improve healing) is not recommended over fistulotomy alone for simple fistula — similar recurrence/incontinence rates but larger wound size, larger sphincter defect, and longer healing time. May have a role for chronic blind-ending sinus tracts not crossing the sphincter complex and extending cephalad into the ischiorectal space (where fistulotomy isn't feasible) — a drain can facilitate this scenario's narrow, deep tracts.

The modified Hanley procedure (posterior midline fistulotomy into the deep postanal space, via primary fistulotomy or cutting seton with counter-drainage of bilateral horseshoe tracts) has been associated with successful treatment of deep postanal space/horseshoe fistulas that would otherwise be considered complex — reserved by this chapter's authors for failures of sphincter-sparing approaches, given the significant external sphincter division involved (though relatively low observed incontinence rates).

Setons

From Latin seta (bristle) — historically horsehair ("seta equina"). Materials: nonabsorbable suture, vessel loops, Penrose drains, silastic catheters, rubber bands, wire, electrical cable tie.

• Draining seton: secured loosely, no tension/compression. Used for complex disease with significant sphincter involvement or inflammation precluding accurate assessment. Goal: sepsis control and tract maturation for later definitive repair. Well tolerated, low risk.
• Cutting seton: secured with tension/compression, requiring division of the anoderm overlying the tract; serially tightened in clinic (as often as every other day or as infrequently as once postop) until it extrudes, leaving intact scar. Healing time weeks to months depending on tissue volume divided. Low recurrence (0–10%) but significant, highly variable reported incontinence risk (0–67%) — variability attributed to differing technique, follow-up duration, and reporting quality. A large series (n=1460, Ritchie et al.) found 12% incontinence overall (rising to 32% when one-third of manuscripts not describing incontinence outcomes are excluded; 22% when limited to medicated cutting setons). Incontinence risk increases with more proximal internal-opening location. A review (Vial et al.) found similar recurrence with preserved vs. divided internal sphincter (5% vs. 3%) but significantly worse incontinence with division (5.6% vs. 25.2%) and no recurrence benefit — internal sphincter division is not necessary and should be avoided when using a cutting seton.
• Loose seton as definitive treatment (rather than bridge-to-repair): increasingly supported by recent studies despite an incompletely understood mechanism (proposed: eventual distal seton migration out of the high-pressure zone allows healing). Emile et al.: ~10% recurrence, 3% incontinence risk; recurrence risk factors were previously recurrent fistula, supralevator extension, and anterior fistula. Kelly et al. (200-patient multicenter review, setons changed electively every 3 months): 100% initial clearance, 6% overall recurrence, 96% patient tolerance (median 3 seton changes, range 1–8). Omar et al. RCT (60 complex-fistula patients, conventional drainage seton vs. external anal sphincter-sparing rerouting seton): 13% vs. 3% persistence/ recurrence (p=0.35, not significant), no difference in physical/social/sexual activity, but significantly faster healing with the external sphincter-sparing technique (46±18 vs. 103±47 days, p<0.0001) — promising for that specific variant.

Fibrin glue

First described in the 1990s for tract obliteration without incontinence risk. Initial series (Hjortrup et al.): 50% success — considered reasonable given the procedure's repeatability, ease, and minimal risk. Subsequent variations (preoperative setons, degree of tract debridement, intra-adhesive antibiotics, suture closure of openings) haven't shown recurrence benefit. Success rates remain widely variable (14–94%) and decrease with increasing fistula complexity. Not recommended as first-line for complex fistula-in-ano given variable success, but a reasonable low-risk second-line option/alternative when other surgical options aren't feasible.

Fistula plug

Developed with similar sphincter-preservation goals to fibrin glue: a biologic or synthetic plug is passed through the tract and secured to the internal opening. First bioprosthetic-mesh-plug series (Johnson et al. 2006, n=15): 87% closure. Subsequent studies show widely variable success (24–88%), attributed to inadequate tract debridement, inadequately secured plug, and preoperative-seton presence/absence — none proven as significant success factors. Mixed success plus higher cost have kept fistula plug from becoming a widely accepted first-line treatment for complex fistula-in-ano. Evaluated as an adjunct to ERAF and LIFT (below) with mixed results.

Endorectal advancement flap (ERAF)

Debridement/excision of the fistula tract plus mobilization of a wide-based mucosal/submucosal rectal flap, covering the internal opening after suture closure. The same core technique is reused for simple Rectourethral Fistula repair (ASCRS Ch14), which cites this chapter directly. Reproducible, reasonable success (60–100%) has made ERAF an accepted first-line treatment for complex anal fistulas. Success keys: adequate blood supply (wide submucosal-plexus base) and lack of tension (adequate mobilization length). Degree of circular muscle (internal sphincter) incorporated in the flap varies by series, with a direct correlation between muscle involvement and flap viability but an inverse relationship between muscle involvement and subsequent incontinence. Recurrence risk factors: smoking, recurrent disease, Crohn's disease, prior horseshoe abscess, elevated BMI. Contraindications: Crohn's disease, undrained sepsis, persistent secondary tracts, fistula diameter >3cm, malignancy/ radiation etiology, anorectal stricture (see Anal Stenosis). Flap configuration (rhomboid vs. elliptical) doesn't significantly change recurrence (64% vs. 62%); standard curette debridement vs. fistulectomy excision of the tract also shows no success difference. Repeat ERAF is feasible with good success rates, though higher recurrence than initial repair. Importantly, ERAF does not directly divide sphincter muscle, but incorporation of sphincter fibers in the flap still worsens continence in up to 35% of patients — a real caveat to the "sphincter-sparing" framing. Adding fibrin glue to ERAF has shown higher failure rates than flap alone (a negative combination result worth flagging); fistula plug likewise doesn't improve on flap alone. An alternative dermal (rather than mucosal) flap design carries a theoretical lower mucosal-ectropion risk, in "house," "diamond," or V-Y configurations — heterogeneous literature, no definitive recommendation possible; overall safe with low-to-moderate incontinence (10–20%) and moderate success (50–70%).

Ligation of intersphincteric fistula tract (LIFT)

Developed 2007 (Rojanasakul et al.) as a "total sphincter preserving" technique, now also widely accepted as first-line for complex fistula-in-ano. Dissection of the intersphincteric space to the mature fistula tract, which is divided and doubly ligated; internal/external wounds debrided and left open. Initial description: 94.4% healing, 0% incontinence. A subsequent 251-patient retrospective study by the same authors: 87.7% healing (a meaningful drop, flagged by the chapter as a limitation of relying on the initial small series) — the initial-study population (66.6% simple anal fistulas) underscored that broader/more-complex populations perform less well; overall reported success ranges 61–94% with rare continence changes. Recurrence associated with shorter fistula tract. When LIFT fails, it most often drains via the intersphincteric incision as a persistent intersphincteric fistula, subsequently manageable with simple fistulotomy. One RCT (Madbouly et al., 70 patients, LIFT vs. ERAF): initial success 94% (LIFT) and 91% (ERAF), falling to 74% and 66% respectively at 1-year follow-up — emphasizing both the importance of long follow-up and the risk of late failure. A meta-analysis found ERAF and LIFT results "quite similar" overall.

LIFT variants: BioLIFT (bioprosthetic graft placed in the intersphincteric plane to reduce communication between the fistula's two portions — not supported as advantageous given cost and equivocal results vs. standard LIFT, 80.2% vs. 81.9% success in one comparison); LIFT-PLUG (bioprosthetic plug through the debrided external tract — shorter healing time and higher primary healing than standard LIFT, 22 vs. 30 days and 94.0% vs. 83.9%, both P<0.001, in one comparison); LIFT plus (partial fistulotomy of the distal external tract to promote drainage — nonsignificant success advantage over standard LIFT, 85% vs. 81%, p=0.0529); LIFT + ERAF. With current data, none of these three variants can be confidently recommended over standard LIFT.

Novel surgical therapies

• Fistula tract laser closure (FiLaC™) — first described 2011/2014; radially emitting laser probe closes the tract, originally combined with ERAF, later used alone. Initial small-series success 77–82% with no incontinence; subsequent studies show lower success (33–71%) — reflecting typical early-adoption optimism vs. broader real-world results. Repeat FiLaC, or fistulotomy/fistulectomy with sphincter repair, achieved secondary success in some primary failures. Success favored by intersphincteric-type, short (<30mm) tracts and prior seton use; one study reported 1.7% minor mucous/gas incontinence at 25.4-month median follow-up.
• Video-assisted anal fistula treatment (VAAFT) — first described 2006 (Meinero and Mori); a fistuloscope with continuous glycine-mannitol irrigation directly visualizes the primary fistula, secondary tracts, and internal opening; a unipolar electrode fulgurates the tract walls, followed by debridement and closure of the internal opening (staple, suture, or advancement flap ± fibrin glue reinforcement). Similar in concept to FiLaC but with the added benefit of direct visualization. Initial series: 74% primary closure, 87% healing at 1 year. Meta-analyses: 76% success/16.2% complication rate (8 studies, 786 patients) and 86.8% success/4.8% complications (11 studies, 788 patients, 9-month median follow-up) in subsequent, larger pooled data. Recurrence rates vary by internal-opening closure method: staple lowest (15.3%), then suture (17.7%), advancement flap highest.
• Fistulotomy with primary anal sphincter reconstruction (FIPS) — historically fistulotomy was reserved for simple fistulas given rising incontinence risk with complexity; recent studies evaluate primary fistulotomy ± fistulectomy followed by end-to-end primary sphincteroplasty with dissolvable sutures for complex fistulas. High success (91–96%), low incontinence (2–13%, mostly minor post-defecation soiling). Recurrence/incontinence risk factors: prior recurrent fistula, complex fistula, secondary tracts, and prior seton drainage. Proponents argue a favorable success/complication profile vs. many other complex-fistula options.
• Stem cell therapy — active area of investigation, mixed results. Phase II RCT (Garcia-Olmo et al., n=35, fibrin glue alone vs. fibrin glue + 10 million adipose-derived stem cells): 4.43× relative healing rate increase with stem cells (71% vs. 16% at short-term assessment), but healing fell to 62.5% at 1-year follow-up. A subsequent phase III multicenter RCT (Herreros et al./FATT collaborative group, n=200, 19 centers; stem cells alone vs. stem cells + fibrin glue vs. fibrin glue alone): no significant between-group difference overall at 24–26 weeks (~40%) or 1 year (~50%) — but a much more favorable result at the pioneer/expert center specifically (54.56% / 83.33% / 18.18% respectively, P<0.001), suggesting a real center-experience effect not yet generalizable.
• Over the scope clip (OTSC® Proctology) — described 2012 (Prosst and Ehni): a super-elastic nitinol clip closes the internal fistula opening via a specialized endoscope. Initial small series: 60–93% healing. Discomfort reported as minimal, though the clip usually requires removal with a dedicated clip cutter. Promising but with inadequate evidence yet to support routine adoption; further study needed on success, failure risk, complications, and device cost.

Recommendation (per chapter's summary)

Surgical drainage is the standard for a demonstrable abscess and can often be done in-office under local anesthesia; antibiotics are reserved for special circumstances (cellulitis, sepsis). Cure and continence preservation are the overriding goals for fistula management, with continence often taking priority — patient QoL is generally better with an indwelling loose seton than with significant incontinence. The multitude of available methods reflects genuinely significant failure rates for sphincter-sparing approaches; when one method fails, it's often reasonable to try another. Informed consent — discussing failure rates, setting expectations, and ensuring patients understand their alternatives — is emphasized as paramount.

Related pages

• Anorectal Abscess — the acute precursor stage; shares this page's cryptoglandular pathophysiology and anatomic classification.
• Colorectal Anatomy & Embryology — anal gland/crypt and potential-space anatomy underlying this page's fistula classification.
• Crohn Disease — extrasphincteric/Crohn's fistulas are explicitly out of this chapter's scope; see that page's existing, more detailed perianal Crohn's fistula management content.
• Fecal Incontinence — fistulotomy, cutting seton, and (to a lesser degree) sphincter-fiber incorporation in advancement flaps as iatrogenic sphincter-defect causes.
• Anal Stenosis — anorectal stricture as an ERAF contraindication; cutting-seton/fistulotomy wound healing shares reconstructive principles with that page's flap techniques.
• Pruritus Ani — fistula is listed among the anorectal-disorder causes of anal pruritus in that page's differential.
• Anorectal Sexually Transmitted Infections — HIV-anorectal-disease framework (cryptoglandular fistulas in HIV-positive patients are treated in standard fashion per that chapter).

Open items / gaps

• No dedicated primary-literature ingest yet for any of the individual RCTs/meta-analyses cited (Madbouly LIFT-vs-ERAF trial, Garcia-Olmo/Herreros stem cell trials, Omar seton RCT) — all currently known only via this textbook chapter's synthesis.
• ~~Extrasphincteric/Crohn's fistula management is explicitly deferred by this chapter; not yet comprehensively reconciled with Crohn Disease's existing content beyond a pointer — worth revisiting when ASCRS Ch47 ("Anorectal Crohn's Disease") is ingested.~~ Resolved (2026-08-25) — ASCRS Ch47 - Anorectal Crohn's Disease ingested; full Crohn's-specific fistula classification, diagnosis, medical/surgical management now on Crohn Disease's "Perianal (anorectal) Crohn's disease" section.
• Novel therapies (FiLaC, VAAFT, stem cell, OTSC) are explicitly early-evidence/evolving per the chapter itself — worth updating if a definitive comparative trial or updated guideline is ingested later.