Introduction

Fecal incontinence is defined as the recurrent, uncontrolled passage of solid or liquid stool or gas for at least 1 month in an individual, with a developmental age of at least 4 years. Its consequences on physical health are considered not dangerous, but its impact on the emotional and social lives of patients and caretakers is devastating. Anxiety of having unexpected episodes leads to avoiding normal everyday activities, low self-esteem, and social isolation, thus diminishing the quality of life. Furthermore, fecal incontinence is a substantial and continuous financial burden to patients, their families, and the health care system [1].

Fecal incontinence is a common problem. This disorder is present in all age groups and both genders, with prevalence varying from 1.5% in children to more than 50% in nursing home residents [2]. Prevalence rates increase with age and, in contrast to past beliefs, appear to be similar in men and women [2, 3]. However, true prevalence remains unknown and hard to establish due to the lack of standard scoring scales, differences in data collection, under-reporting of symptoms by patients, and variations in the populations sampled [4].

The causes of fecal incontinence, either passive or urgent, are diverse (Table 1) [4, 5]. It occurs when one or more of the continence-maintaining mechanisms are compromised to an extent that other mechanisms are unable to compensate. Anal sphincter disruption, pudendal neuropathy, impaired anorectal sensation, and incomplete evacuation may all contribute to the pathogenesis of fecal incontinence. Thus it appears to be a cumulative, multifactorial process. Obstetric sphincter disruption is the most common cause. It is found in nearly one third of the vaginal deliveries in primiparous women with a resulting group risk of incontinence equal to 76.8–82.8% [6].

Table 1 Summary of the varied causes of fecal incontinence
Full size table

Fecal incontinence is a symptom, and as such, its measurement should be based on the subjective perception of the patient [7]. A substantial number of severity-grading scales exist now. Although not perfect, there are some measures such as the Jorge/Wexner Continence Grading Scale, the American Medical Systems Scale, and the Vaizey Scale that have proven to be useful and acceptably reliable (intra-class correlation coefficient=0.75–0.87) (Table 2) [8–10].

Table 2 Cleveland clinic incontinence scoring system developed by Jorge and Wexner
Full size table

Exact identification of the underlying cause and severity of fecal incontinence is the most important prerequisite for selecting an adequate treatment. The diagnostic process begins with an accurate interview aiming a full obstetric, previous surgery, neurological, dietary and medication history; objective severity scoring; a careful inspection of the anus and perineum; and a digital rectal examination. Rectosigmoidoscopy is always performed to rule out organic conditions such as neoplasms or proctitis. Although there is a large armamentarium of specialized investigations nowadays, it is clear that no single objective study can be considered as the definitive test of anorectal function. Anorectal physiology testing includes anorectal manometry, endoanal ultrasonography, electromyography, defecography, endoluminal and dynamic pelvic magnetic resonance imaging, saline continence test, and balloon-retaining test. Anorectal manometry assesses anal sphincter function, rectal sensation, reflexes, and compliance. Endoanal ultrasonography is the simplest, most widely available, and least expensive test that provides information about the integrity of internal and external anal sphincters, the puborectalis muscle, and the rectovaginal septum, with an accuracy of 90–100% [11, 12].

Treatment options

The goal of treatment for patients with fecal incontinence is to restore continence and to improve the quality of life. The therapeutic approach is strongly individualized and depends mainly on the severity of incontinence, the available treatment options, and the common health status of the patient. Despite the potential appeal of primary surgical intervention, the majority of patients would greatly benefit from conservative measures. Initial specific treatment should be medical, including biofeedback in combination with a bowel management program (avoiding offending foods, ritualizing bowel habit, improving skin hygiene). Biofeedback is believed to improve continence in up to 80% of patients; it is simple, relatively cheap (though labor-intensive), and without adverse physical effects. It can be used alone or as an adjunct to surgical treatment. However, the mechanism of improvement from biofeedback is not clearly known, and a recent study suggested that the beneficial results may be due to the motivation of the patient and the enthusiasm of the therapist, rather than to the technical aspects of therapy [13].

Surgery should be considered in carefully selected patients who have failed conservative therapy. The available options fall into several major groups:

Techniques of direct repair of the anatomical defect

Sphincter repair

Patients best suited are those in whom the incontinence is secondary to an anterior sphincter defect. Obstetric and iatrogenic etiologies are the most common examples of surgically correctable fecal incontinence. An overlapping repair is propagandised as preferable to a simple end-to-end apposition of the sphincter muscles and achieves short-term improvement in up to 85% of cases [14]. However, recent studies have shown that continence deteriorates thereafter, falling to a drastic level of 6% in 10 years [15–17]. Until now, there is no good evidence to show the advantage of the overlapping method over the end-to-end technique. The causes of long-term deterioration in function remain uncertain. Breakdown of the repair is a common cause of early failure but seems an unlikely mechanism many months and years after initial clinical success. Other hypotheses include aging, scarring, and progressive pudendal neuropathy related either to the initial injury or to the subsequent repair. The optimal salvage therapy for patients whose sphincteroplasty has failed or for those who are not candidates for the procedure remains uncertain. Current options include muscle transfer procedures, artificial sphincters, and sacral nerve stimulation (SNS) [18].

Postanal repair, total pelvic floor repair

The first method was popularized by Sir Alan Parks from St. Mark’s Hospital, London, in the 1980s [19]. The idea was to use an intersphincteric approach to sharpen the anorectal angle and thus improve continence. Total pelvic floor repair, a combination of postanal repair and preanal levatoroplasty (reefing) recommended by Deen et al. and Korsgen et al. is of no further benefit in the long term [20, 21].

Neosphincter-formation techniques

Dynamic graciloplasty (DGP)

Dynamic graciloplasty (DGP) involves the transposition of the gracilis muscle to construct a neosphincter (Fig. 1a). Gracilis transposition without electrical stimulation was described first by Pickrell et al. in 1952 [22]. Stimulated graciloplasty was first performed by Baeten and colleagues in 1986 [23, 24]. The distal portion of the muscle encircles the anus, whereas its tendon is tightly sutured to the ascending part of os pubis. The neuromuscular bundle remains preserved. Wraps can be made in three different fashions (Fig. 1b). The first turn can be made anterior or posterior of the anal canal with contralateral tendon attachment (gamma or epsilon configuration). If the length of the muscle appeared to be insufficient, an ipsilateral attachment could be made (alpha configuration) [25]. A low-frequency neurostimulator (pulse generator) is implanted into the abdominal wall beneath the rectus fascia, and electrodes are then connected to an implanted battery. Chronic stimulation of the muscle via this neurostimulator converts it from a fast-twitch, fatigue-prone muscle into a slow-twitch, fatigue-resistant muscle, giving it the properties to function as a sphincter (allowing prolonged states of contraction). An external magnet allows patients to control stimulation. The muscle contracts when the neurotransmitter is activated and relaxes, thereby allowing defecation, when the transmitter is deactivated. DGP is associated with substantial morbidity, infection being the most frequent, followed by disturbed evacuation, pain caused by stimulation, and pulse generator displacement. A specific complication with this method is the late anorectal perforation, caused by erosion of the tendon during stimulation [25, 26]. According to most recent publications, complication rate is more than 50%, and overall success rate is below 35% [27, 28]. DGP is largely limited to a small number of centers in which adequate patient volume and surgical experience help assure acceptably low morbidity and satisfactory functional outcomes [18]. Patients with lack, atrophy, or denervation of the gracilis muscle are contraindicated for DGP and should be offered artificial sphincter implantation.

Fig. 1
figure 1

a Mobilization of the right gracilis muscle by ligation of the minor vascular pedicles; proximally the major vascular pedicle and the obturator nerve with its branches are located. On the left side, the most common gamma configuration of dynamic graciloplasty is shown after transposition of the muscle and implantation of the stimulation device. N Neurostimulator, G gracilis muscle, T ischial spine. (From [55], copyright British Journal of Surgery Society Ltd., reproduced with permission. Permission is granted by John Wiley & Sons Ltd on behalf of the BJSS Ltd.). b Epsilon, gamma, and alpha configurations of the graciloplasty are possible, the choice being dependent on anatomical considerations. (From [56] with permission)

Full size image

Artificial bowel sphincter (ABS or Acticon Neosphincter, American Medical Systems, Minneapolis, MN)

This method consists of an inflatable cuff (mimicking the function of the anal sphincter) placed around the upper anal canal usually through a transverse perineal incision (preferred in women) or bilateral vertical perianal incisions, a pressure regulating balloon (filled with 40 ml of normal saline or radiopaque liquid) implanted extraperitoneally in the abdominal wall (in the iliac fossa or the retropubic space of Retzius), and a subcutaneously positioned control pump in either the scrotum or the labia (Fig. 2). The implanted device remains inactivated to allow wound healing for up to 2 months [29]. The procedure was first reported in 1987 by Christiansen and Lorentzen [30]. It is technically easier, less surgeon-dependent, and more cosmetic than DGP. Contraindications for this method are a destroyed or severely scarred perineum, perineal radiation-induced lesions, pelvic sepsis, pregnancy, receptive anal intercourse, Crohn’s disease, mild forms of fecal incontinence, diarrheic transit, and the absence of rectal reservoir (less than 150 ml maximal tolerable volume) [31]. Infection, skin erosion, device malfunction, and pain are the most frequent and fearsome complications because revisional procedures and explantation will be necessary nearly always [32]. Once successfully placed, however, the ABS results in excellent functional outcome for most patients, and its function remains stable for many years. Implantation success rate is still under 50%, but those who retain the device recover continence in up to 100% of the cases at the second year after implantation [33]. Although ABS met the enthusiasm of surgeons initially, the high explantation rate and the frequency of adverse events have shown that its implantation may be harmful. The prognosis for those who have undergone explantation is unclear, as their long-term outcomes have not ever been reported. Results in all published studies are not analyzed on an intention-to-treat basis and so misrepresent the effectiveness of the procedure. They describe the effect solely on responders and cannot be generalized to all patients implanted with ABS. There is insufficient evidence on the safety and effectiveness of ABS implantation for fecal incontinence to date. From the low-level evidence available, it is obvious that for most patients the procedure is of uncertain benefit. Until long-term data on an intention-to-treat basis are made available, the implantation of ABS should be considered experimental [34].

Fig. 2
figure 2

Artificial bowel sphincter (ABS) implanted in a male. The inflatable cuff around the anal canal is connected to the subcutaneously positioned control pump (here, in the scrotum; in female, in the labia), which is itself connected to the regulating balloon, placed either in the prevesical space as shown here or extraperitoneally in the iliac fossa. (Modified from [56])

Full size image

Prosthetic anal sphincter (PAS)

An alternative to ABS, PAS consists of a sphincter element placed around the anorectal junction via a transabdominal approach, a constant-pressure (60–70 mmHg) balloon reservoir, and a control pump (Fig. 3). The sphincter component comprises an inflatable linear expander that, when inflated, flattens and angulates the bowel against a soft-gel-filled pillow, thus reproducing the action of the puborectalis muscle. The balloon reservoir provides the hydraulic pressure to drive the system. The control pump is placed in a subcutaneous pouch in the right iliac fossa, where it is operated by the patient. The device is activated 6 weeks after surgery [35].

Fig. 3
figure 3

The prosthetic anal sphincter (PAS) is placed at the anorectal junction, above the pelvic floor. The figure shows the resultant bowel angulation when the device is activated. The PAS in the open (left) and closed position (right). (From [35], copyright British Journal of Surgery Society Ltd., reproduced with permission. Permission is granted by John Wiley & Sons Ltd on behalf of the BJSS Ltd.)

Full size image

Minimally invasive, innovative, and experimental techniques

Sacral nerve stimulation (SNS)

This is an innovative three-staged procedure. In the first stage, the sacral nerves that best correlate with anal sphincter contraction are located (usually in the S3 foramen) with the patient under anesthesia. Then a trial of temporary SNS is conducted over a period of 1 to 2 weeks to assess the therapeutic potential with regard to the patient’s continence. Finally, a permanent electrode is implanted, and the leads are tunnelled to a permanent impulse generator in the ipsilateral buttock or the anterior abdominal wall for ease of control [36]. The mechanism of action remains uncertain. Possible mechanisms include stimulation of sensory afferents or motor efferents, modulation of neural reflexes, and a combination of these effects [18]. Indications are still not clearly defined, but it seems most applicable in patients with internal sphincter defects or degeneration, or in patients with a weak but structurally intact external anal sphincter who are not amenable to simple surgical treatments [37]. A new percutaneous technique, the insertion of a tined lead, is becoming more widely available, and this may remove the need for the sacral wound and dissection and allow the procedure to be performed under local anesthesia and sedation [38]. Current data show SNS to be a successful technique for the treatment of fecal incontinence, with an 80% approximate overall success rate. In clinical surgery, it is extremely rare to have a minimally invasive, low morbidity predictor of success prior to surgical intervention, and this is a major advantage of this technique. Implantation of a permanent sacral nerve stimulator is also a relatively minor procedure, especially compared to the alternative surgical options for these patients. The operative site is distant to the bowel, so previous anal procedures do not complicate the surgery. Overall morbidity rate is 5–10%, with the major complication being infection. Patients with bilaterally absent pudendal nerves showed no benefit [39]. Patients who have a neurologically intact sacral plexus, an elicitable pudendal nerve latency, no evidence of spinal cord injury, and an anatomically intact anal sphincter and rectum appear to have a favorable prognosis. Those with the converse appear to have a poorer success rate. Currently, SNS is still used on a research basis. Contraindications for the procedure are pathologic conditions of the sacrum that prevent adequate electrode placement (such as spina bifida), skin disease at the area of implantation, anal sphincter damage amenable to direct repair or requiring a sphincter substitute (DGP or ABS), trauma sequelae with micturition disorders or low bladder capacity, pregnancy, bleeding complications, psychologic instability, low mental capacity, and the presence of a cardiac pacemaker or implantable defibrillator. The method of choice for permanent implantation is the unilateral one, although when it is insufficient, a bilateral implantation should be considered [40].

Secca procedure

This technique uses a temperature-controlled radio-frequency (RF) energy delivery to the anal canal muscles, which is believed to cause heat-induced tissue contraction and remodelling of the anal canal and distal rectum. It was first used by Takahashi et al. in 1999 [41]. It is a safe, well-tolerated outpatient procedure that can be performed without need for general anesthesia. No long-term adverse reactions were observed. It may be appropriate for those patients who have failed conservative measures and are neither candidates for overlapping sphincter repair nor willing to undergo artificial bowel sphincter implantation or colostomy. It may be of benefit also to those with recurrent incontinence after previous overlapping sphincteroplasty. The procedure is performed under sedation and local perianal anesthesia. A specifically designed anoscopic instrument (a barrel with four nickel–titanium curved-needle electrodes deployed from within; 22-gauge, 7-mm length) and RF generator (Secca System, Curon Medical, Inc., Sunnyvale, CA) are used to deliver temperature-controlled RF to the muscle of the anal canal (selected target temperature of 85°C) while preserving anoderm and mucosal integrity with irrigation coolant. Lesions are placed in all quadrants in the anal canal, beginning just distal to the dentate and progressing in a stepwise manner about 15 mm proximally [41]. Tissue heating with the delivery of radio-frequency energy results in immediate collagen contraction, followed by focal wound healing, remodelling, and tightening. Mean procedure duration is less than 40 min, and most patients have 48 to 64 lesions created. Patients are discharged approximately 2 h after treatment. There are presently few additional options, besides formation of a stoma, for those individuals in whom conservative therapy fails and who either are not eligible for sphincter repair or ABS or are not interested in such surgery. It is these patients, specifically, who may benefit from the Secca procedure as an alternative to a more invasive surgical intervention. The procedure does not preclude the possibility of further surgery [42].

Anal sphincter bulking

Variety of agents have been used, including autologous fat, glutaraldehyde-treated collagen (GAX), synthetic macromolecules, or silicon. These materials are usually injected submucosally either at the site where the sphincter is deficient or circumferentially if the whole muscle is degenerated or fragmented. There is limited short-term experience and no long-term data with these techniques [43]. Recently, a biocompatible agent (Durasphere) composed of pyrolytic carbon-coated beads suspended in a water-based carrier gel containing beta glucan was used with good results. The main advantages of this method are its simplicity, speed, and the fact that it does not need anesthesia or antibiotic coverage and highly qualified personnel [44].

Perineal puborectalis sling operation

This technique was first applied by Yamana et al. in 2001 [45]. A specially designed polyester mesh sling is introduced along the puborectalis muscle, from a posterior perianal incision running to a small suprapubic incision. The ends are tied together with moderate tension [45]. This operation is very similar to the original puborectoplasty reported by Shafik in 1991 [46]. His procedure consists of routing a Teflon sling from behind the anorectal angle to the pubic bone. Both techniques rely on Parks’ theory, emphasizing the importance of anorectal angle to maintain continence. Few patients have been evaluated up to now, although these procedures gave promising initial results in the treatment of neurologic passive fecal incontinence.

Pudendal nerve anastomosis

Pudendal nerve anastomosis is an encouraging, although still experimental, technique. The pudendal nerve is anastomosed to a peripheral nerve (innervating the biceps femoris or the semitendinosus muscle). Then the respective peripheral muscle is wrapped around the anal canal and electrically stimulated for several months till anal function is recovered [47, 48].

German Artificial Sphincter System (GASS)

This is an experimental high-tech sphincter made of polyurethane. It consists of a support ring including two cuff elements: a fluid reservoir fixed on its outer diameter and a multichamber occluding cuff on the inside diameter. Both cuffs are interconnected by an integrated piezoelectric micropump/valve unit. This entirely new device was evaluated around the external sphincter of isolated porcine anal canals and achieved adequate continence at very low working pressures, thus promising a correspondent low risk of intestinal ischemic injury and erosion bleeding. Future in vivo studies should prove the efficacy and safety of GASS as therapy for major fecal incontinence [49].

Stoma-formation techniques

Antegrade continent enema (Malone procedure)

The appendix is detached from the cecum and reversed so that the cecal end of the appendix is anastomosed to the right lower quadrant abdominal wall. The tip of the appendix is excised, and a tunnel is made in the tenia of the cecum to create a one-way valve. The patient would then intubate the appendicostomy regularly and flush the colon, thus clearing it from stool. This method is suitable for children and patients with neurological lesions. In the long term, fibrosis at the stoma site with stoma stenosis may lead to loss of response. This most frequent complication is treated with dilatation [50].

Continent stoma

This is probably the safest and easiest surgical way to handle patients with severe fecal incontinence. Irrigation and gas-filtering plugs can improve life quality. Fecal diversion is a good alternative in cases where all other therapies have failed or if comorbidities preclude more aggressive therapy. It is particularly suitable for patients with spinal cord injury and those immobilized with skin problems or other complications. The use of laparoscopic-assisted approach helps for minimal morbidity for the patient. Bardoel et al. fashioned a dynamic rectus muscle stoma sphincter in dogs, which seems a promising technique yet to be reported in humans [51]. As patients naturally express reluctance to live with a stoma, when diversion is inevitable, it is of utmost importance to create a well-constructed and appropriately sited stoma to limit difficulty with its management [4].

Discussion

Surgical treatment of fecal incontinence comprises a broad spectrum of methods whose diversity expands from mini-invasive ambulatory procedures to technically sophisticated device-implantation techniques. The indications for all these methods are often overlapping and remain undefined. There is no perfect generally applicable surgical technique that is both safe and effective enough. This means that the end-point of surgical therapy should be recovery of adequate quality of life at the price of possibly minimal morbidity and mortality. However, today we seem still far from achieving this goal. Well-established procedures such as the sphincter and pelvic floor repairs in their different variants sound logical and are still the most widely spread, although recent studies revealed their poor long-term outcomes. Other methods including the Thiersch wire encirclement and the passive muscle neosphincters (bilateral nonstimulated gracilis or gluteus transpositions) were abandoned and belong to history. Enthusiasm for any new technique often leads to overemphasis of the results, and early reports are usually good. The history of ABS is no exception, and now we have seen that results deteriorate with time and do not correspond to initial encouraging data [52]. The same is true for the DGP. Both methods are technically demanding, with considerable morbidity, substantial learning curve, and low success rate. Despite these obvious disadvantages, both DGP and ABS remain attractive for surgeons because once successful, they provide outstanding and long-lasting improvement. One should mention here the innovative prosthetic anal sphincter. It produces continence without risking ischemic injury to the bowel. Thus it may overcome the major problem of the ABS, which caused ischemia at operating pressures required to maintain continence, a problem not shared by its urinary analogue. This is explained by the fact that bowel is less tolerant to ischemia than the urethra and is not “circular.” Compared to PAS, ABS has a higher failure rate related to both the site of implantation and the mode of action. Perineal placement increases the risk of infection and erosion. The circular design has the potential to damage bowel in the high-pressure zones between the cusps. The PAS in turn has two disadvantages not shared by the ABS. First, the abdominal approach necessitates surgery of a greater magnitude than with the ABS, although in the future it could be performed laparoscopically. Second, the PAS is placed around the lower rectum (anorectal junction) resulting in a short “dead space” below the implant. This produces a certain degree of postevacuational fecal leakage. It is unlikely that control of liquid stool will be achieved using an artificial device without compromising blood supply because the control of liquid stool requires a constant sphincter occlusion pressure equal to the intralumenal pressure of the bowel [35]. The availability of an effective and relatively complication-free artificial anal sphincter would be a major advance in the management of patients with fecal incontinence who, at present, have few successful surgical options other than colostomy. Unsatisfactory results from neosphincters lead to the development and introduction of new, less traumatic and minimally invasive surgical methods, SNS being the most impressive of them. The Secca procedure and the injection of anal bulking agents are performed on an outpatient basis and should find their appropriate niche in the treatment scheme of fecal incontinence in the future.

Additional problems are the lack of validated and standardized scoring system and the absence of reliable diagnostic methods possessing high preoperative value for the prediction of outcome. Although there is no hesitation about the need of endoanal ultrasonography and anorectal manometry, the place of other less frequently used methods of physiology testing in the diagnostic plan is not clear. For example pudendal nerve terminal motor latency is used to preoperatively distinguish whether a weak sphincter is due to muscle or nerve injury [53]. However, this study was proven to have poor sensitivity and specificity and does not predict surgical outcome [15, 54].

In conclusion of this short critical review, a relatively simple and up-to-date treatment algorithm for fecal incontinence is desirable (Fig. 4). However, the construction of such a guide is very difficult. It demands the availability of randomized controlled trials studying large number of patients on an intention-to-treat basis, with appropriate control groups and not biased by investigators’ own preferences. Until then, the therapy for fecal incontinence remains stepwise and based to a great extent on the “trial and error” principle. Its last step should be surgical treatment. It improves symptoms but does not cure, and its results deteriorate with time, so surgeons have to be realistic about its final outcome. An individual approach should be favored, based on exact diagnosis and extremely careful patient selection.

Fig. 4
figure 4

A simplified scheme representing the major steps of an algorithm for the surgical treatment of severe fecal incontinence

Full size image