Anal incontinence (AI) is defined as the involuntary loss of flatus, liquid, or stool that is a social and hygienic problem.¹ It remains a complex and poorly understood condition with a multifactorial etiology. Several mechanisms, either alone or in combination, produce symptoms of AI: (a) consistency and amount of stool (eg, diarrhea), (b) damage to the mucosa of the colon and rectum (eg, colitis), (c) neurologic factors (eg, diabetes, Parkinson disease), (d) miscellaneous (eg, congenital disorders, rectocele, etc), and (e) injuries to the anal sphincter and pelvic floor muscles.

Childbirth and anorectal surgery are the main causes because the anal sphincters and the pudendal nerve may be damaged.^2,3 Minor degrees of fecal soiling due to internal anal sphincter (IAS) injuries have been reported after hemorrhoidectomy, mucoprolapsectomy, manual anal dilatation, or lateral internal sphincterotomy.^2,3 Obstetric trauma to the anal sphincter is invariably restricted to the area anterior to a horizontal line through the mid-canal. Injury of the anal sphincters posterior to this line is usually due to some other etiology such as trauma or fistula-in-ano. Obstetric anal sphincter trauma may involve part or the full length of the sphincter and can be partial or full thickness.^3-5 The majority of obstetric injuries are associated with a single, large defect in the external anal sphincter (EAS) between 9 and 3 o’clock but can also involve the internal sphincter. Fistula surgery can also be responsible for damage to the anal sphincters and up to 60% of patients can become incontinent following treatment of complex, high fistulas or after multiple operations for a recurrent or persistent fistula.⁶

A systematic evaluation is fundamental to reveal the underlying pathophysiology and lead to appropriate therapy. The first step in evaluating patients suffering from AI is always a careful history. Questions should focus on type and degree of incontinence as well as on changes in the patient’s lifestyle. A scoring system such as Williams, Pescatori, Wexner, and AMS is often used to rate incontinence more accurately. Assessment of patient’s quality of life (QoL), using specific questionnaires such as the Fecal Incontinence Quality of Life Scale should be considered useful parameters of this disorder.⁷

In daily clinical practice, endoanal ultrasonography (EAUS) is an important diagnostic tool to identify sphincter lesions and defects⁸ and it has been defined as the gold standard investigation in the assessment of anal sphincter integrity by the joint report of the International Urogynecological Association (IUGA)/International Continence Society (ICS) on the terminology for female pelvic floor dysfunction.⁹ Sphincter function is evaluated with anal manometry by measuring the resting and squeeze pressures.¹⁰ Evaluation of EAS innervation can be assessed with electromyography (EMG)¹¹ and a pudendal nerve terminal motor latency (PNTML) test.¹² Defecography can be useful when there is concomitant dysfunction such as rectal prolapse, rectocele, enterocele, and intussusception.¹³

Despite many therapeutic options, no single treatment is optimal. For patients with an anterior sphincter defect, an overlapping sphincter repair has traditionally been the treatment of choice. Unfortunately, the published long-term results have not been optimal.¹⁴ Other surgical options include dynamic graciloplasty (DG) and implantation of an artificial bowel sphincter (ABS). However, high complication rates such as infection and device malfunction have kept them from becoming mainstream treatments.¹⁵ Based on its success for urinary incontinence (UI), sacral nerve stimulation (SNS) has been used successfully. The indications for SNS are expanding while the technique continues to be refined.¹⁶ Other options have included bulking agent injection¹⁷ and radio-frequency (RF) energy¹⁸ applied to the anal sphincter. However, for some, a colostomy is the last resort to achieve an acceptable QoL without the uncertainty of stool loss. This manuscript will highlight the current published results, indications, and techniques in the surgical treatment of AI.

The aim of this procedure is to reapproximate the disrupted ends of the anal sphincter. In the acute situation, for example, immediately after childbirth, the freshly torn muscle ends can be identified and approximated by an experienced obstetrician.¹⁹ However, when women present later in life with AI, there is considerable scarring and it is more difficult to achieve successful repair especially of the internal sphincter.

Preoperative

Patient Evaluation

Patients best suited for an anterior sphincter repair (ASR) are those in whom AI is secondary to a sphincter defect usually following obstetric trauma. Anal endosonography is useful in delineating the disrupted ends of the IAS and EAS. Anal manometry allows for evaluation of the anal sphincter length, maximum resting pressure (reflecting mainly IAS function), and maximum squeeze pressure (reflecting mainly EAS function). Absent or weak contractility of the residual sphincter muscle is suggestive of a neuropathy that is associated with a poor outcome and therefore other options should be considered. Although some perform neurophysiological tests such as PNTML and EMG, these tests do not quantify the degree of neuropathy. They provide no additional clinically useful information over a clinical examination and are therefore not a prerequisite to sphincter repair.²⁰

Consent

It is important at the very outset to establish the patient’s expectations regarding surgery. The definition of success as viewed by the surgeon is not always the same as that of the patient. Although a success rate of 80% can be expected in the first year, this could fall to around 50% in five years.²¹ The patient needs to be aware that good control of flatus is not usually achieved and dietary modification may be necessary. Passive soiling may continue if there is a persistent IAS defect. The outcome of sphincter repair is affected by factors such as irritable bowel syndrome, rectal hypersensitivity, inflammatory bowel disease, diabetes mellitus, bowel transit, and bowel consistency. The patient should be made aware of possible complications such as infection, wound breakdown, fistula formation, and failure of surgery.

Intraoperative

The operation is usually performed under general anesthesia with the patient in lithotomy, although some surgeons prefer to use the prone jackknife position. A curvilinear incision is performed between the posterior fourchette and the anus (Figure 35-1). In the lithotomy position this incision would commence at the 9-o’clock position and extend to the 3-o’clock position. The incision is then deepened into the ischioanal fossa and using Metzenbaum scissors the sphincter muscles are identified and mobilized. Some surgeons use a nerve stimulator or a needle tip of an electrosurgical blade to identify the EAS. The next step is to transect the scar tissue in the midline and further mobilize the two ends of the EAS.

If the endoanal scan has revealed that the IAS was intact, then care should be taken to avoid surgical injury. Unlike primary repair of a freshly ruptured anal sphincter,¹⁹ the disrupted IAS is not always identifiable during secondary sphincter repair. The retracted scarred IAS is usually difficult to mobilize as a separate layer but when possible, repair should be attempted using a monofilament delayed absorbable suture such as PDS 3-0 with mattress sutures 0.5 cm apart.

Although the EAS can be repaired by a simple end-to-end apposition of the sphincter muscles, the most popular method of repair of the EAS is using an overlapping sphincter repair (Figures 35-1 and 35-2).^22,23 In a recent Cochrane review,²⁴ meta-analyses of primary repair following acute obstetric anal sphincter injury showed that there was no statistically significant difference in perineal pain, dyspareunia, flatus, and stool incontinence between the two repair techniques at 12 months but showed a statistically significant lower incidence in fecal urgency and lower AI scores in the overlap group. The overlap technique was also associated with a statistically significant lower risk of AI worsening over a 12-month period. Despite this, there was no significant difference in QoL. However, recent publications have reported far from perfect results with the overlapping technique as some patients developed new evacuation disorders.²¹ In a small randomized study, Tjandra et al.²⁵ found no significant difference in functional outcome of overlapping versus apposition of the sphincter ends. In general, a sphincter defect that exceeds three hours on the clock face or 90° (as identified by EAUS) could make overlapping technically difficult and place the repair under tension. However, a direct relationship between size of the tear and degree of dysfunction could not be confirmed.²⁶

After mobilization of the EAS, each end is grasped by an Allis clamp and 1 to 1.5 cm of the sphincter ends are overlapped with mattress sutures using a monofilament delayed absorbable suture such as polydioxanone PDS 2-0 (Ethicon, Somerville, NJ) (Figure 35-2). Although some surgeons prefer permanent sutures, there is a high incidence of suture erosion and wound dehiscence.²⁷ In a prospective study²⁸ correlating clinical signs with the postoperative endoanal scan image, there seems to be some benefit if the scar is not resected but used for overlapping.

For additional support some surgeons perform a levatorplasty by inserting two interrupted sutures using 2-0 delayed suture material. However, approximation of the levator muscle is not anatomically correct and has been associated with dyspareunia and complications when performed during primary repair.²⁹

After sphincter repair care must be taken to reconstruct the perineal body and bury the PDS suture knots so as to avoid any discomfort from migrating sutures. Finally a simple wound closure of the skin is performed with the midportion left open for drainage. Alternatively, a small suction drain could be inserted. The wound usually closes within four to six weeks (Figures 35-1 to 35-5). A Foley catheter can be inserted to avoid overdistension of the bladder. Intravenous antibiotics such as cefuroxime and metronidazole should be given.

Adequate pain relief is important and early ambulation should be encouraged. The Foley catheter can be removed the next day. Some advocate the use of oral antibiotics for a week but this is not essential provided personal hygiene is ensured. Sitz baths and the use of a bidet or handheld shower are recommended. Free fluids should be allowed after the surgery followed by a normal diet. Laxatives should be prescribed to keep the stools soft as passage of stool will be painful. Fecal impaction should be avoided at all costs. Follow-up is usually arranged for four to six weeks.

Role of Colostomy

The outcome of ASR is not improved significantly by a concurrent colostomy. In patients with a severe trauma to the perineum other than after delivery, a proximal colostomy is often constructed to avoid septic complications and to facilitate nursing management. However, stoma-related complications are reported in more than 50% of these patients.³⁰

Primary versus Secondary Repair

In the acute, emergency trauma situation, initial treatment consists of debridement of nonviable tissue, removal of foreign material, open drainage, and often proximal colostomy with distal washout. Depending on the extent of injury and the associated trauma, reconstructive surgery may be deferred. The approach in patients with obstetric trauma is somewhat different. A third- or fourth-degree perineal tear must be repaired immediately, although defects after repair are reported in up to 85% and about 40% of these women eventually develop incontinence.³¹ For secondary repair after obstetric injury, a delay of at least six months to one year has been recommended to allow the tissue to return to normal. However, Soerensen et al.³² prospectively followed up sphincter repairs done as a delayed primary (within 72 hours postpartum) or as an early secondary reconstruction (within 14 days after delivery) without a covering stoma in women who had sustained a third- or fourth-degree obstetric tear. They found equal results with acceptable long-term functional outcome in both groups.

Failed Primary Repair

There seems to be no difference in outcome in patients who had an unsuccessful primary repair and those who had no previous repair. In about 62%, a repeat repair can be expected to be successful, although patients who had undergone more than two previous repairs appear to have poorer clinical results.³³

Age

Simmang et al.³⁴ found no difference in outcome in patients with a mean age of 66 years compared with younger ones. This was confirmed in a recent study by Evans et al.³⁵ However, Nikiteas et al.³⁶ reported poorer results in patients older than 50 years, especially with concomitant obesity and perineal descent.

Pudendal Neuropathy

Some studies have shown that unilateral or bilateral pudendal neuropathy preoperatively (prolonged PNTML) is associated with a poor outcome after secondary repair.³⁷ However, other studies failed to show such association. In any event, provided there is evidence of contractility of the sphincter muscle, consideration would always be given to repair of a sphincter defect.

Biofeedback

In a one-year follow-up study of 48 patients after a third- or fourth-degree sphincter laceration, after one month ten patients (21%) complained of AI, eight of flatus only. After one year none complained of AI and three (7%) of flatus incontinence.³⁸ The authors concluded that pelvic floor exercises seem to suffice as first-line treatment. In light of poor long-term results with the overlapping ASR, pelvic floor exercises seem to be an appropriate first-line approach. It is of utmost importance that after successful repair, patients continue to perform pelvic floor exercises and biofeedback. Long-term results of electromyographic biofeedback training appear promising.³⁹

Combination with Other Perineal Operations

On occasion a sphincter defect can be diagnosed in combination with other perineal pathologies. Combining ASR with levatorplasty, procedures for UI and/or pelvic organ prolapse provide good outcome and are cost effective.⁴⁰ ASR can also be part of a more extensive perineal reconstruction of the pelvic floor for cloaca-like deformities.⁴¹

Outcome of Anterior Sphincter Repair

In the short term (<5 years), the results of ASR are usually quite good, with success rates of about 75%, although it is well known that a persistent defect after repair is associated with a poor immediate outcome.⁴² Technically, this can happen if the suture material cuts through the muscle, allowing the sphincter ends to retract.¹ Furthermore, isolated IAS defects often present as persistent AI.¹ Unfortunately, the long-term results of ASR are not so favorable as only one third of patients are totally continent, and about half are satisfied with their results provided they are not incontinent of feces.⁴³ Malouf et al.²¹ recently reported long-term results (>5 years) for patients after ASR. No patient was fully continent (stool and gas), with 52% of patients still wearing a pad and 66% of patients reporting lifestyle restriction. Although it is important to know the severity of AI, it is also important to understand and measure the impact of AI on patients, or rather the effect on QoL.

General questionnaires have a long history of use with established reliability, validity, and population norms. One of the newest measurement tools is the FIQL, which is very sensitive and appears to be useful.⁷ Halverson and Hull⁴⁴ nicely demonstrated the mismatch of full continence (14%) and excellent QoL (34%), which cannot be detected with other measurement tools. Recent publication⁴⁵ has addressed the issue of sexuality and sphincter repair. Interestingly, sexual activity and function were similar following ASR, compared with controls, despite more pronounced symptoms of AI. However, AI of solid stool and depression related to AI were correlated with poorer sexual function. Anal continence rates five years after ASR are disappointing and adversely impact QoL, yet do not appear to relate to sexual function.

Until the advent of EAUS, the etiology of AI was largely attributed to neuropathy and therefore referred to idiopathic or neurologic AI. One of the most often used surgical options was the postanal repair (PAR) as described by Sir Allan Parks to restore the anorectal angle, increase anal pressure, and lengthen the anal canal.⁴⁶ However, the short- and long-term results are not especially good when compared with sphincteroplasty. The problem of patient selection is of utmost importance. Patients with excessive posterior pelvic floor mobility are poor candidates for PAR and it has been suggested that they could be excluded with preoperative dynamic MRI.⁴⁷ The exact mechanism of action of PAR is unclear. Some speculate that success appears to be related more to improved sphincter pressure and anal sensation⁴⁸; others believe that the efficacy of PAR is more due to local scarring and anal stenosis than restoration of the anorectal angle.⁴⁹ Despite the low success rate, the absence of any mortality and the low morbidity suggest that PAR may be a valid therapeutic approach, especially as a second surgical approach after failed primary surgery. It should, however, be offered only to selected patients with persistent, severe AI despite an anatomically intact EAS who are not candidates for or refuse all other operative modalities. Interestingly, in a recent published series of 57 patients, although 48% complained of severe AI, in the long run (median, 9.1 years), 79% were satisfied with the outcome.⁵⁰

Intraoperative

The patient is positioned in the lithotomy position with the hips well flexed. A urinary catheter is passed into the bladder and the perineum shaved. A curved incision around the posterior and lateral aspects of the anus is made between 2 and 3 cm posterior to the anal verge. With scalpel or scissors an upper skin flap is dissected free toward the anal canal and lifted to achieve adequate exposure of, and access to, the intersphincteric plane. The dissection should start laterally as this is the easiest place to develop the intersphincteric plane at an early stage. It is important that the surgeon follows the natural slightly posterior direction of the intersphincteric groove as it nears the level of the pelvic floor. The dissection should proceed above the puborectalis muscle sling and above the levators to open up the postrectal space and also deepened laterally to a high level so that the ischial spines can be easily palpated. As the surgeon enters the supralevator plane, the fascia of Waldeyer is identified and then incised transversely, after which the dissection is nearly complete. The repair may now proceed. The rectum is displaced anteriorly with a retractor and the first polypropylene stitch of the repair is placed on both sides of the levator muscle as high and anteriorly as possible. Further four or six interrupted sutures are thus placed serially in the deep part of the levator muscle. The sutures are now tied, starting with the posterior stitch and working toward the anterior stitch. It is important not to put too much tension on the muscle that is being sewn together or the repair might cut through in the postoperative period. After the anorectal angle has been re-created, the repair is completed by additional layers of stitches further drawing the puborectalis and EAS across: this narrows and lengthens the anal canal. The “U”-shaped skin incision is changed as a “Y”-shaped closure. It may be desirable to leave a small triangle open at the point of greatest tension.

Postoperatively, the patient should avoid straining to open his or her bowels. The urinary catheter should be kept in place for at least four days. On the second postoperative day the patient should start to take a laxative to ensure that the first and subsequent bowel movements will be soft and pass easily without straining. There is no consequence if some separation of the wound edges occurs. There is often a lengthy period of postoperative adaptation, and the patient may not notice benefit from the operation especially if the stool is liquid. Frequent postoperative supervision is required to ensure maximum benefit.