About Us Physicians Education Members Patient and Public Corporate Partners DCR Research Foundation
Home > Physicians > Education > Core Subjects > Fistula-in-Ano, Abscess, ...

Fistula-in-Ano, Abscess, Pilonidal Cyst and Hidradenitis Suppurativa

Wayne De Vos, MD, PhD
West Reading, PA

The diagnosis and treatment of fistula-in-ano and anorectal infection is a cornerstone of any busy colon and rectal surgical practice. Also, while pilonidal disease is not strictly a disease of the anus, its proximity to the anus and the occasional difficulty in differentiation from fistula-in-ano often leads to referral to a colon and rectal surgeon for treatment. Finally, hidradenitis suppurativa, which occurs more frequently in other areas of the body, can be a diagnostic dilemma when suspected in the perianal area, and may present as a prolonged and long-suffering course of treatment for the patient and surgeon.

Perianal and Perirectal Abscess

Suppuration of the anal glands as the origin of most anorectal infections (cryptoglandular theory) is generally well accepted. The anal glands number 10-15, are more concentrated in the posterior midline, empty into the anal canal at the level of the dentate line, and penetrate into the surrounding sphincter to a variable depth, with the majority no deeper that the internal sphincter muscle or intersphincteric space. An abscess originating in one of these glands can extend in a variety of directions into one or more of the virtual perianorectal spaces, including the intersphincteric, perianal, ischiorectal, and supralevator spaces. Approximately 10% of anorectal abscesses will originate from a non-cryptoglandular source, including Crohn's disease, atypical infection (e.g., tuberculosis, lymphogranulum venereum), malignancy, or trauma.

Pain and swelling are the most common complaints, and may have been present for many days. Fevers and local bleeding or discharge may also be presenting symptoms. On evaluation, a tender swelling may be readily apparent, possibly accompanied by local erythema, warmth and fluctuance. An inflammatory process confined to the area of the anal verge most likely represents a perianal abscess, while involvement of the buttock beyond the sphincter complex suggests a perirectal abscess. Complaints of pain and findings of focal tenderness in the anal canal with or without local swelling or fluctuance may represent an intersphincteric abscess. Finally, a supralevator abscess may present with pelvic pain and dysuria, with or without obvious external physical findings; digital rectal exam in these patients may find induration or fluctuance in the deep perianal tissues. Certainly any patient with anorectal pain and inability to tolerate an exam in the office should undergo exam under anesthesia to rule out the presence of an infectious process.

Incision and drainage remains the cornerstone of treatment for anorectal abscess. With adequate drainage, antibiotics are generally not necessary, though in the immune-compromised patient, and in patients with prosthetic devices or symptomatic valvular heart disease, additional broad spectrum coverage should be considered. In rare circumstances, an abscess cavity cannot be identified by exam or needle aspiration of the inflamed and swollen area. In these unusual cases, broad spectrum antibiotics and re-examination in 24-48 hours may be appropriate.

A perianal abscess can generally be drained under local anesthesia. This is also true for most ischiorectal abscesses, though occasionally the extent of the abscess cavity and local swelling precludes effective exploration in the office, in which case the patient should be taken to the operating room for incision and drainage under anesthesia. In either case, after infiltration with lidocaine with epinephrine, a cruciate incision is made over the area of fluctuance as close to the anus as possible, so as to minimize the length of a potential fistulous tract. To ensure that the incision drains well, the "corners" of the cross are resected. If the cavity is large, digital exploration to break up possible loculations can be completed. The patient is instructed to take a sitz or tub bath 2-3 times a day and after bowel movements. Packing or antibiotics are not required.

An unusual form of ischiorectal abscess involves transphincteric progression of the infectious process posteriorly into the deep postanal space of Courtney. The patient may present with severe anal pain without obvious fluctuance, or may present with unilateral or bilateral ischiorectal abscesses or fistulae, which represent extrasphincteric extensions of the abscess through the ischiorectal fossa. In the patient with pain and tenderness but no obvious fluctuance, exam under anesthesia and needle aspiration of the deep postanal space will determine its involvement. Standard treatment for the abscess can then be completed. In the patient with uni- or bilateral ischiorectal abscesses, again standard treatment of the abscesses can be completed. In the event that drainage persists, and a fistulous tract is not easily identified, a posterior fistula must be ruled out.

Intersphincteric abscesses are often drained in the operating room. These abscesses can be approached through an incision at the level of the intersphincteric groove, followed by exploration of the intersphincteric space and division of the internal sphincter muscle to the level of the dentate line.

A supralevator abscess is best approached in the operating room to ensure correct identification of origin and adequate drainage. If cephalad extension of an intersphincteric process is determined, drainage should be into the rectum. If the process is thought to be an extension of an ischiorectal abscess through the levator floor, drainage should be through the skin overlying the ischiorectal abscess. Incorrect drainage in either case may lead to an extrasphincteric fistula.

Primary fistulotomy.
Controversy continues regarding the utility of primary fistulotomy in patients with acute abscess formation. Advantages include improved determination of the path of suppuration, improved healing time, and decreased incidence of recurrence . Disadvantages include possible creation of false passages through the inflamed tissue, unnecessary conversion of an office procedure into an operating room procedure, and over-treatment of abscesses which would heal without fistulous sequelae. In the general population, following incision and drainage of an abscess, only one third to one half of patients will show signs of a persistent internal opening as manifested by the presence of a fistula or the recurrence of an abscess in the same location. Therefore, in one half to two thirds of patients, a search for an internal opening is unnecessary and potentially costly. Given the generally benign nature of fistulae which present following uncomplicated incision and drainage of a perianal or perirectal abscess, watchful waiting seems appropriate in the majority of cases. Certainly, in the rare event of an obvious, readily identified internal opening at the time of initial examination, primary fistulotomy could be considered.

Recurrence. Following incision and drainage of a perirectal or perianal abscess, recurrence of the abscess or persistent drainage has been reported in 16-48%. Healing of the abscess without fistula formation may be due to blockage of the offending duct during the suppurative process, which removes the source of infection after adequate drainage through the perineum. Once a patient shows signs of persistent glandular infection (e.g., chronic drainage or recurrent abscess formation), the diagnosis of fistula-in-ano is reasonable, and exploration in the operating room is appropriate.

Incontinence. Of paramount importance in the treatment of any anorectal suppurative condition is the potential effect on continence. Incision and drainage should be completed near but not involving the sphincter musculature. Precise determination of the path of a supralevator infection will prevent creation of an extrasphincteric fistula. Adequate drainage, followed by daily local care by the patient with sitz- or tub baths following bowel movements will reduce the risk of further infection or inadequate healing.

Special Considerations.
Anorectal sepsis. Delay in presentation or inadequate treatment of an abscess may result in the extension of local infection into the superficial perianal tissues or deeper into the deep soft tissues, fascia or muscle layers. Additionally, particular virulence of the organism, immune compromise in the patient (e.g., poorly controlled diabetes), or local barriers to adequate healing (e.g., previous surgery) may play a role in this potentially life-threatening process. Local findings may be subtle, with only erythema, tenderness and swelling in the absence of fluctuance. Systemic findings of fever, chills, and shock may precede overt evidence of local gangrene, characterized by skin necrosis, blistering, and, in the presence of gas forming organisms, crepitus. Treatment of the patient in whom anorectal sepsis is suspected must be rapid and aggressive, with vigorous resuscitation and stabilization and initiation of broad spectrum antibiotics, followed by operative debridement of the affected tissue, including skin, subcutaneous fat, fascia and muscle back to healthy muscle. Gram stain of the suppurative fluid will assist in determining the presence of clostridia. Proximal fecal diversion should be strongly considered if tissue loss is extensive, or if the anal sphincter or rectal wall are involved. In most cases scheduled re-exploration under anesthesia is necessary to rule out extension of the necrotizing process. Despite aggressive treatment, mortality rates exceeding 50% have been reported, both secondary to the aggressive nature of the deep synergistic infection and to the presence of comorbid diseases.

Anorectal inflammation in the neutropenic patient. Profound neutropenia (absolute neutrophil count < 500) has been associated with an increased risk of potentially life-threatening anorectal suppurative disease. Initial signs and symptoms include fever, local pain and urinary retention. Local tenderness and swelling may follow, and fluctuance may or may not be present. Initial measures in patients with an ANC < 500 should include the discontinuation of any manipulation of the anorectum, including digital rectal exam, rectal temperatures or administration of suppositories/enemas. Broad spectrum intravenous antibiotics should be started, and are in the majority of cases successful in controlling the infectious process. Incision and drainage in the absence of frank pus is generally unrewarding, and should be reserved only for patients with persistent sepsis despite maximal antibiotic therapy.


A fistula can be defined as an abnormal connection between two epithelial surfaces. In the setting of the anorectum, a fistula-in-ano generally extends between the anal canal at the level of the dentate line and the skin around the anus. Most fistulae are thought to arise from suppurative disease of the anal glands (the cryptoglandular theory).

A well accepted classification scheme of anal fistulous disease was proposed by Parks, Gordon and Hardcastle in 1976. In this classification, fistulae are categorized into one of four general categories, with variations found in each category. They are labeled according to their relationship with the external sphincter, as essentially all anal fistulae involve the portion of the internal sphincter beyond the dentate line.

Intersphincteric fistula. Accounting for approximately 70% of cases, the tract in this fistula passes between the sphincters, with possible extension cephalad as a high blind intersphincteric tract. Additionally, it may originate as a pelvic abscess, or may extend into the supralevator space as a high blind tract.

Transsphincteric fistula. In Parks' study, approximately 23% presented with a fistula whose tract extended through the external sphincter at some level. High blind tracts can extend as far as the supralevator space.

Suprasphincteric fistula. Approximately 5% of patients will present with a fistula passing from the dentate line completely around the external sphincter via the supralevator space.

Extrasphincteric fistula. Only 2% of patients will present with this type of fistula. While rarely an extension of an ischiorectal abscess will extend and erode into the rectum above the level of the sphincter, more likely the majority of this uncommon type of fistula originate as a caudad extension of a pelvic infection or malignant process, or develop secondary to trauma, either spontaneous or iatrogenic, as may be seen with overly vigorous probing of high, blind tracts extending cephalad from an ischiorectal abscess cavity.

The majority of patients with fistula-in-ano will have a history of abscess development and drainage. They may report persistent drainage, pain and or bleeding. Physical exam in the office will generally reveal an external opening which on gentle probing will support the diagnosis of fistula-in-ano. Digital rectal exam, anoscopy and proctoscopy should be performed, and may reveal an area of induration in the anal canal corresponding to the internal opening, though an actual internal opening is rarely identifiable in the office. Goodsall's Rule suggests that external fistula openings posterior to a line drawn transversely across the perineum will curve posteriorly to open at the dentate line in the posterior midline, while anterior fistulous tracts tract radially directly to the anal canal.

Ultimately, once the diagnosis of fistula-in-ano has been made, the next step is an exam under anesthesia and appropriate treatment, though any preoperative suspicion of an underlying disease process such as Crohn's disease or immune suppression should prompt an appropriate evaluation before proceeding with treatment of the fistula.

Any operative intervention on the anorectum must proceed only after a frank discussion with the patient regarding probable early, and possibly permanent, alteration of fecal continence. While in many cases the potential for permanent loss of control of gas or stool will be small, ultimate outcome cannot always be predicted. Also, possible recurrence (0-18% in reported series) should be discussed. At surgery, several methods of determining the fistulous tract are possible. The common denominator is a gentle touch and respect for resistance of passage of the probe so as to decrease creation of false passages. A probe may be readily passed along the tract; if the internal opening eludes discovery, injection of dilute methylene blue dye, milk or hydrogen peroxide into the external opening using an angiocatheter while carefully observing the anal canal may assist in visualizing the target of the probe. If the internal opening is still not visible, careful unroofing of the identifiable tract may allow more successful probing of the deeper portion of the tract. Rarely, if the complete fistulous tract is still not identifiable despite these techniques, curettage of that part of the tract identified and discontinuation of the exam is reasonable, since further probing may lead to complications. Other tests may then be useful if evidence of a persistent fistula, e.g. trans-anal ultrasound or fistulogram, prior to re-exploration.

When a fistulous tract has been identified, the relationship of the tract with the external sphincter is determined. If little or no external sphincter muscle is involved, the external opening and skin overlying the tract may be excised, and the tract itself curetted. The tissue may be sent for pathological evaluation. The edges of the wound may then be marsupialized or excised so as to prevent the skin from healing prematurely. No packing is required. The patient is instructed to cleanse the area with sitz or tub baths following bowel movements and at least 2-3 times a day. Generally, complete healing can be expected in 3-4 weeks, occasionally longer.

When greater than half of the external sphincter muscle is involved, or in the patient where sphincter integrity is already at risk, a cutting seton can be employed. In this technique, the fistulous tract is curetted, the skin and involved internal sphincter are excised down to the external sphincter, and a synthetic strip of material is passed around the remaining sphincter and tied snugly though not tightly. Material commonly used includes silk suture, vessel loops or penrose drains. Over the ensuing 4-8 weeks, the seton will erode through the muscle without allowing significant retraction of the ends of the sphincter being divided. Ultimately, the seton will either completely erode through, or the remaining small amount of undivided sphincter still encompassed by the seton can be divided, either in the office or in the operating room.

In some cases, particularly when sphincter function is at risk, as in female patients with anterior fistula, closure of the internal opening and creation of a blind sinus by advancement of a rectal mucosal flap internally may be appropriate. This technique may be particularly useful in patients such as those with extrasphincteric fistulae, with previous anal surgery which has damaged the sphincter, and in patients with Crohn's disease.

Special Considerations.
Crohn's Disease. While an in-depth discussion of the treatment of anorectal Crohn's disease will not be pursued here, let it be said that anorectal fistulous disease must be approached extremely carefully in the patient with suspected or known Crohn's disease. Some important concepts include: asymptomatic fistulae do not necessarily require interventional treatment; draining (rather that cutting) setons can be used to control fistulous tracts and prevent further abscess formation; primary fistulotomy should not be attempted in patients with Crohn's disease. Alexander-Williams stated that "incontinence is likely to be the result of aggressive surgeons, not of aggressive disease."

A horseshoe fistula represents primary involvement of the deep postanal space of Courtney with secondary extension laterally into one or both ischiorectal spaces. For this reason, a horseshoe abscess often presents as a unilateral or bilateral ischiorectal abscesses, which, when drained, fail to heal, and often elude curative fistulotomy when a tract directly toward the anal canal is sought. Definitive treatment of a horseshoe abscess requires unroofing of the deep postanal space via an extrasphincteric incision posterior to the anus, division of the internal sphincter posteriorly to the level of the transphincteric fistula, and placement of counter-incisions over the involved ischiorectal spaces. A cutting seton should be employed to complete the treatment of the posterior trans- or suprasphincteric fistula, and until this is healed, draining setons can be used to keep the long lateral subcutaneous tracts from developing suppuration.

Pilonidal Cyst

The etiology of pilonidal disease remains controversial, and may in fact be a combination of factors. Although theories of congenital origin were popular in the past, generally an acquired etiology is more accepted at present. Two popular theories maintain that either 1) a small pit develops secondary to subcutaneous rupture of a follicle in the natal cleft, into which errant hairs may collect; or 2) errant hairs strands burrow into the skin at the level of the natal cleft secondary to their natural unidirectional scaled surface and the propensity of hair to collect in this region. Once a single hair has started the process, others follow. In either case, a subcutaneous cavity or sinus is created, not uncommonly with a surprisingly large amount of hair present in the space. This may smolder for months or years before becoming secondarily infected abscessed.

While a pilonidal cyst can be observed in both sexes and at any age, they predominate in men and in the second and third decade of life. Initially presenting as an abscess and cellulitis at the level of the sacrococcygeal area, spontaneous drainage often occurs, to be followed by generally painless chronic waxing and waning drainage from the secondary sinuses. The majority of secondary sinuses track cephalad, but some may track toward the anus, potentially being confused with fistula-in-ano or hidradenitis suppurativa.

The diagnosis is most often readily apparent, with an area of swelling and fluctuance over the sacrum characterizing the acute presentation of a pilonidal abscess, and one or multiple sinus opening in the midline over the sacrum in the chronic state. Midline pits are almost always identifiable inferior to the sinus. Differential diagnosis includes furuncle, anal disease and sacral osteomyelitis.

When presenting in the acute phase with redness, swelling, tenderness and fluctuance, drainage of the offending abscess can often be accomplished in the office. Antibiotics are not necessary. The incision should always be made off the midline, and drainage should be assured in the standard fashion as for any abscess. Importantly, any contained hair and debris should be removed from the abscess cavity, and any loculations should be disrupted. If possible, the chronic sinus should be debrided. The wound can be packed with gauze, and the patient is discharged and instructed to remove the packing the next day, shower or bathe at least twice daily thereafter. Close follow up in the office is important to keep hair from collecting in the healing wound. In many cases, such treated abscesses will heal without sequelae.

When a chronic sinus has developed, multiple options are available for definitive treatment. Generally, healing is expected in 30-60 days, and recurrence is in the range of approximately 15%, with the exception of marsupialization, which has the lowest recurrence rate of approximately 5%. In the majority of cases, treatment under some form of anesthesia will be required. The patient is placed in prone jackknife position, the area over the sacrum is shaved and prepped, and the limits of the pilonidal cavity are determined. This can be done by probing, or by using vital dye such as dilute methylene blue injected carefully into the sinus. This will stain the cyst but not the surrounding tissue, and be a helpful guide regarding completeness of excision of the cyst.

In perhaps the least invasive method, the sinuses are excised individually and closed primarily, and the deep cavity is curetted via a lateral incision, which is left open and packed. A second technique involves excising the anterior aspect of the cyst, with curettage of the base of the cyst. In a third technique, the entire cyst is excised down to sacral fascia and then packed open. In a fourth technique, after excision of the entire cyst, the edges of the wound are marsupialized to the deep tissue, which in most reports decreases healing time and recurrence. In rare circumstances, following complete excision of the cyst and creation of a relatively small wound with minimal tension, it is reasonable to consider primary closure, though this has been found to have a higher recurrence rate compared to secondary healing.

Postoperatively, the patients are instructed to change the dressing at least twice daily, use tub or sitz baths to bathe the area, and to return approximately weekly to the office for examination. At the time of weekly visits, careful removal of all the local hair is imperative to reduce the risk of recurrent cyst formation.

Complex or recurrent disease.
In a small subset of patients, despite adequate surgical technique, persistent or recurrent pilonidal disease defies cure. In these patients, more advanced plastic techniques aimed at replacement of the midline tissue with skin flaps and grafts (e.g., Z-plasty, myocutaneous flaps) has led to healing in the majority of cases.

Hidradenitis Suppurativa

Hidradenitis suppurativa (HS) is thought to be a chronic suppurative disease of the apocrine glands. While more common in women in other parts of the body, in the genital area it shows a male propensity. It is most often seen in the second to fourth generation, and will often "burn out" thereafter. Possible factors in the development of apocrine gland suppuration include poor hygiene, close shaving, hyperhidrosis, tight-fitting clothes, diabetes and smoking.

Patients will often present with multiple painful, swollen violaceous nodules, generally with little purulent discharge. The inflammatory process may resolve without treatment, but with or without treatment often pursues a waxing and waning course over many weeks, and will often smolder for years. Pitting and chronic ulceration may develop. The chronic discharge and discomfort can be socially limiting.

Differential diagnosis in the perianal or genital area is primarily between HS and other subcutaneous tunneling diseases, though, as with any non-healing cutaneous lesion, biopsy should be considered to rule out neoplastic disease. The lack of midline pits over the sacrum helps distinguish HS from pilonidal disease; the lack of involvement of the anal canal at the dentate line helps distinguish HS from Crohn's disease and benign anal fistula. Confounding the differentiation can be the coexistence of HS and Crohn's disease, which has been found to be true in up to 38% of patients with HS.

Opinions vary regarding the utility of oral antibiotics for acute HS; In the presence of significant cellulitis, antibiotics with activity toward cutaneous organisms seems warranted. Incision and drainage of any areas of fluctuance is appropriate, and can most often be accomplished in the office. Counseling regarding prevention of progression is appropriate at this time, with particular stress on local care and hygiene.

Surgical management of chronic HS involves one of four possible techniques: 1) unroofing and curettage of cavities and tracks with healing by secondary intention; 2) wide excision and healing by secondary intention, with or without marsupialization; 3) excision and primary closure; 4) excision and split thickness skin grafting or advancement flap closure. Of these techniques, wide local excision with healing by secondary intention is the most commonly used, though for very large wounds, some form of post-excision grafting is appropriate and may significantly accelerate healing.


  1. Vasilevsky, C-A. Fistula-in-Ano and Abscess. 1998. In Fundamentals of Anorectal Surgery, 2nd edition. Beck D and Wexner S, eds. Saunders, pub. Pp. 153-173.
  2. Karulf R and Perry WB. Pilonidal Disease. 1998. In Fundamentals of Anorectal Surgery, 2nd edition. Beck D and Wexner S, eds. Saunders, pub. Pp. 225-232.
  3. Parks AG, Gordon PH, Hardcastle JD. A classification of fistula-in-ano. 1976. Br J Surg; 63:1-12.
  4. Waters GS, Nelson H. Perianal Hidradenitis Suppurativa. 1998. In Fundamentals of Anorectal Surgery, 2nd edition. Beck D and Wexner S, eds. Saunders, pub. Pp. 233-236.
  5. Church JM, Fazio VW, Lavery IC, et al. The differential diagnosis and comorbidity of hidradenitis suppurativa and Crohn's disease. 1993. Intl J Colorectal Dis; 8: 117-119.
  6. Corman M. Anorectal abscess and Anal fistula. 1998. In Colon and Rectal Surgery, 4th edition. Lippincott-Raven, pub. Pp 224-271.
  7. Gordon PH. Anorectal abscesses and fistula-in-ano. 1999. In Principles and Practice of Surgery for the Colon, Rectum and Anus, 2nd edition. Gordon PH, Nivatvongs S, ed. Quality Medical Publishing, pub. Pp 241-286.
  8. Nivatvongs S. Pilonidal Disease. 1999. In Principles and Practice of Surgery for the Colon, Rectum and Anus, 2nd edition. Gordon PH, Nivatvongs S, ed. Quality Medical Publishing, pub. Pp 287-301.